UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Information control as a bargaining tactic in social exchange networks Foddy, Margaret Lynn 1975

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1975_A1 F63_3.pdf [ 11.16MB ]
Metadata
JSON: 831-1.0100036.json
JSON-LD: 831-1.0100036-ld.json
RDF/XML (Pretty): 831-1.0100036-rdf.xml
RDF/JSON: 831-1.0100036-rdf.json
Turtle: 831-1.0100036-turtle.txt
N-Triples: 831-1.0100036-rdf-ntriples.txt
Original Record: 831-1.0100036-source.json
Full Text
831-1.0100036-fulltext.txt
Citation
831-1.0100036.ris

Full Text

INFORMATION CONTROL AS A BARGAINING TACTIC IN SOCIAL EXCHANGE NETWORKS by MARGARET LYNN FODDY B.A., Univ e r s i t y of Saskatchewan,1967 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n the Department of Anthropology and Sociology We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA February, 1975 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Soc.-o/.^y and rHtC0CCu( °^ ? The University of British Columbia Vancouver 8, Canada Date / / ^ ^ ^ c r y ^ ^ / r i ABSTRACT This d i s s e r t a t i o n i s concerned with the process by which s o c i a l actors conceal information about the true l e v e l of t h e i r p r o f i t s i n ex-change i n t e r a c t i o n s , so that they may deviate from a norm of fairness c a l l i n g for equality of subjective p r o f i t s to the parties i n an exchange. Two factors are posited to act as constraints on the p o t e n t i a l advantage of information control — a preference by s o c i a l actors f o r r e l i a b l e i n f o r -mation that allows comparison with exchange partners; and the a v a i l a b i l i t y of a l t e r n a t i v e exchange partners who do not conceal t h e i r resources. In this context, we outline s i x exchange s i t u a t i o n s characterized by d i f f e r e n t d i s t r i b u t i o n s of information (symmetric and asymmetric), and by d i f f e r e n t numbers of a l t e r n a t i v e s . One case, i n v o l v i n g asymmetric i n -formation and several a l t e r n a t i v e exchange partners, i s selected as the focus of t h i s study. A theory i s constructed to make predictions concern-ing the nature and d i r e c t i o n of i n i t i a t i o n s of exchange, the perception of advantage i n information c o n t r o l , and the l i k e l y success of t a c t i c s of con-cealing information about resources from p o t e n t i a l exchange partners. The predictions are subjected to test i n an experimental study, involving 336 subjects i n 42 experiments. The r e s u l t s are l a r g e l y supportive of the predictions that: 1) people who can conceal t h e i r resources make more attempts to gain advantageous exchanges; 2) people p r e f e r to enter exchange transactions i n which they have r e l i a b l e information about t h e i r partners; and 3) people d i r e c t more i n i t i a t i o n s of exchange to others who possess r e l a t i v e l y large amounts of desired resources. The few cases i n which negative evidence arose are evaluated, and attention i s drawn to aspects of the theory and experimental design i n need of further i n v e s t i g a t i o n . i i Bibliography V TABLE OF CONTENTS Page Abstract i Table of Contents i i L i s t of Tables i i i L i s t of Figures i v Acknowledgements v i i Chapter 1 Introduction 1 2 Theory 39 3 Research Design 77 4 Results and Evaluation of Results 106 5 Evaluation and Suggestions for Future Research 148 168 Appendices I Laboratory Set-Up 175 II Experimental Instructions and Questionnaires, Set A III Experimental Instructions and Questionnaires, Set B IV P i l o t Work 180 186 192 Data Referred to But Not Included i n Main Text 208 VI Raw Data 222 i i i LIST OF FIGURES Figure Page 3.1 D i s t r i b u t i o n of Resources and Information i n Set A 89 3.2 D i s t r i b u t i o n of Resources and Information i n Set B 92 3.3 Summary of Features of Experimental Design, Sets A and B 100 A . l Main Features of Booths Used i n Experiments 175 A. 2 I n i t i a t i o n Forms Used i n Set A 176 A. 3 I n i t i a t i o n Forms Used i n Set B and P i l o t Study 177 A.4 Sample of Card Showing Subject h i s Resource Base 177 A. 5 Value Chart Used f o r Set B and P i l o t 178 A. 6 D i s t r i b u t i o n of Resources and Information, P i l o t Set 193 LIST OF TABLES Table 4.1 Frequencies of in i t i a t i o n - - to high and low resource, v i s i b l e s and to non-vi s i b l e s , T r i a l 1, Set A 4.2 Frequencies of i n i t i a t i o n to balanced and unbalanced v i s i b l e s , and to no n - v i s i b l e s , T r i a l 1, P i l o t Set 4.3 Proportions of Advantageous o f f e r s to v i s i b l e s by v i s i b l e and non-visible i n i t i a t o r s , T r i a l 1, Set B 4.4 Proportion of high non-visibles and v i s i b l e s attempting advantageous exchanges, T r i a l 1, Set B l 4.5 Proportions of low non-visibles and v i s i b l e s attempting advantageous exchanges, T r i a l 1, Set BII 4.6 F a i r exchange r a t i o s and obtained average exchange r a t i o s for i n i t i a t i o n s to v i s i b l e s , T r i a l 1, Set B 4.7 Average exchange r a t i o s for i n i t i a t i o n s to n o n - v i s i b l e s , T r i a l 1, Set B 4.8 Proportion of v i s i b l e s and non-visibles who want to re t a i n covers on resources of non-v i s i b l e s , T r i a l 1, Set B 4.9 I n i t i a t o r s ' evaluation of l i k e l i h o o d that t h e i r o f f e r s would be accepted, T r i a l 1, Set B 4.10 Proportions of i n i t i a t i o n s to v i s i b l e s , T r i a l 1, Set B 4.11 Proportions of o f f e r s to v i s i b l e s , T r i a l 1 and 2, Set A 4.12 Acceptance of r e a l o f f e r s , T r i a l 1, Set A 4.13 Frequency of i n i t i a t i o n s to v i s i b l e s and no n - v i s i b l e s , T r i a l 2, Set A 4.14 Acceptance of f a l s e o f f e r s by high n o n - v i s i b l e s , Set B l 4.15 Acceptance of f a l s e o f f e r s by v i s i b l e s and low non-v i s i b l e s , Set BII V L i s t of Tables (Continued) Table Page 4.16 Number of r e a l o f f e r s accepted per number received, v i s i b l e r e c i p i e n t s , T r i a l 1, Set A 134 4.17 Acceptance of r e a l o f f e r s to v i s i b l e s , T r i a l 1, Set BI 134 4.18 Acceptance of f a l s e o f f e r s by v i s i b l e s , T r i a l 2, Set A 135 4.19 Non-visibles' preference f o r removing covers, T r i a l 1 and T r i a l 2, Set A 137 4.20 Proportions of subjects making same type of choice ( v i s i b l e s versus non-visibles) on T r i a l s 1 and 2, Set A 138 A.1 F a i r exchange r a t i o s and average exchange r a t i o s f o r i n i t i a t i o n s to v i s i b l e s , P i l o t Set, T r i a l 1 196 A.2 Directions of i n i t i a t i o n s on T r i a l 1, by v i s i b i l i t y of i n i t i a t o r s and target, P i l o t Set 198 A.3 D i r e c t i o n of i n i t i a t i o n s , T r i a l 2, by v i s i b i l i t y of i n i t i a t o r and target, P i l o t Set 199 A.4 Acceptance of f a l s e o f f e r s by V i s i b l e s , T r i a l 1, P i l o t Set 200 A.5 Relative frequencies of acceptance of r e a l o f f e r s on T r i a l 2, by value p o s i t i o n and v i s i b i l i t y of i n i t i a t o r and target, P i l o t Set 201 A.6 Types of o f f e r s by resource l e v e l and v i s i b i l i t y of i n i t i a t o r , Set A 208 A.7 High n o n - v i s i b l e s ' perception of covers a f t e r T r i a l 1, Set BI 210 A. 8 Low non- v i s i b l e s ' perception of covers a f t e r T r i a l 1, Set BII 212 A.9 V i s i b l e s ' perception of having ri£ covers, T r i a l 1, Set B 214. A. 10 Perceived p r o b a b i l i t y that T r i a l 1 o f f e r w i l l be accepted, Set A 217 v i L i s t of Tables (Continued) Table Page A.11 High n o n - v i s i b l e s ' reasons f o r i n i t i a t i o n s to v i s i b l e s and non-visibles, Set BI 218 A. 12 Low n o n - v i s i b l e s ' reasons for i n i t i a t i o n s to v i s i b l e s and non- v i s i b l e s , Set BII 219 A.13 V i s i b l e s ' reasons for i n i t i a t i o n s to v i s i b l e s and no n - v i s i b l e s , Set BI 220 A.14 V i s i b l e s ' reasons f o r i n i t i a t i o n s to v i s i b l e s and n o n - v i s i b l e s , Set BII 221 v i i ACKNOWLEDGEMENTS I would l i k e to express my indebtedness to my supervisor, Dr. R-A£.H$. Robson, who has given me professional and personal guidance from the beginning of my graduate t r a i n i n g . I would also l i k e to acknowledge the advice of the members of my d i s s e r t a t i o n committee: Martha Foschi George Gray Jean LaPonce Ken.MacCrimmon I am gra t e f u l to Don Earner, and e s p e c i a l l y to my husband, B i l l Foddy, for t h e i r suggestions, c r i t i c i s m s , and moral support during the f o r -mulation and w r i t i n g of th i s d i s s e r t a t i o n . I would l i k e to thank the Canada Council for providing f i n a n c i a l assistance i n the form of Doctoral Fellowships i n the period 1969-73, and the University of B r i t i s h Columbia f o r support for a Graduate Fellowship i n 1968-69. CHAPTER 1 INTRODUCTION Most people have expectations that s o c i a l exchange t y p i c a l l y involves a trade of equally valued u n i t s . However, as the value assigned to units of goods and services i s s u b j e c t i v e l y determined, t h i s leads to the p o t e n t i a l for s o c i a l actors to conceal and d i s t o r t the true value to them of what i s traded. Can people c o n t r o l information about the r e -sources they possess, and those they wish to possess 5to t h e i r own advan-tage? What happens to them i f they try? Is such an advantage l i m i t e d to the case i n which the person being taken advantage of has no a l t e r n a t i v e s ? These are the questions with which t h i s study i s concerned. F i r s t , we wish to examine the process by which actors conceal information about the true l e v e l of t h e i r p r o f i t s i n exchange transactions, so that they may deviate from a norm of fairne s s c a l l i n g f o r equality of subjective p r o f i t s to the p a r t i e s i n the exchange. Second, we are i n t e r -ested i n whether s o c i a l actors display a preference for exchange r e l a t i o n -ships i n which the partner does not control information about his p r o f i t s , and the consequences of such a preference. To t h i s end, we w i l l construct a set of hypotheses which have t h e i r basis i n the more general theory of s o c i a l exchange. We w i l l then see how our ideas stand up when we subject them to test i n an experimental representation of a s o c i a l exchange s i t -uation. 2 . In the present chapter, a review w i l l be provided of the s o c i a l exchange framework, elaborating those concepts from i t that are relevant to the present study. These w i l l include the concepts of r e c i p r o c i t y , f a i r -ness, s o c i a l p r i c e s , subjective p r o f i t , and resources. In t h i s context, the advantages w i l l be considered that might accrue to the actor who can control information about his needs and resources. Attention w i l l be given to the questions of when an actor i s l i k e l y to desire advantage i n exchange, and under what conditions he i s l i k e l y to s u c c e e d w i t h the use of i n f o r -mation co n t r o l as a t a c t i c f o r gaining advantage. In o u t l i n i n g t h i s l a t t e r issue, we w i l l extend a psychological p r i n c i p l e of a 'preference for avoid-ing uncertainty', to argue for the existence of a preference for exchange re l a t i o n s h i p s i n which partners to an exchange know the true l e v e l of each other's p r o f i t s . We s h a l l argue that i n the presence of a l t e r n a t i v e s about whom information i s a v a i l a b l e , t h i s preference places severe constraints on the success of c o n t r o l l i n g information to gain advantage. S o c i a l Exchange Theory 1 In the framework of s o c i a l exchange, i n t e r a c t i o n between two or more actors i s conceptualized as a mutually r e i n f o r c i n g transfer of com-modities and behaviours that are valued i n s p e c i f i a b l e ways by the p a r t i -cipants. An actor, Person, engages i n goal-directed behaviour, requiring i n t e r a c t i o n with others who are s i m i l a r l y engaged i n meeting t h e i r goals. To s a t i s f y h i s needs, Person must f i n d an Other who has what Person wants, and who i s w i l l i n g to give i t up i n . r e t u r n for something Person has. 3 An Example Consider two students, Peter and Olga. Peter i s good at English, but can barely scrape through Math. Olga, on the other hand, i s a whiz at Math, but has never done well at English. These two students are i n a p o s i t i o n to set up an exchange r e l a t i o n s h i p . Peter can write Olga's essays, and Olga can l e t Peter copy her Math problems, or, i f they are more honest, they can help one another study. But how many essays should Peter write i n return for a set of Math problems? A l o t w i l l depend on how bad Peter perceives Olga to be at English, and on whether she can get help elsewhere. And Peter's perception of these things can vary. I f he has never seen Olga i n an English c l a s s , and does not know her grades, Olga may be able to convince Peter that she needs less help at English than he does at Math, or that i t takes her a great deal of e f f o r t to help him with Math problems. I f , i n addition, Olga i s sure that Peter, i s an utter dunce at Math, she may have him at an advantage. Accordingly, she may be able to demand a l o t of help i n English to make i t worth her while, or to demand some addi-t i o n a l reward , such as rides to school. Even more than i n economic ex-change, where prices tend to be standardized, the perceived worth of r e -sources (and these may range from s k i l l s and deference to concrete goods) i s something that can be influenced by how the actors present themselves, and what information they make a v a i l a b l e . Let us now review how exchange theory would describe the i n t e r a c t i o n between actors such as our imaginary students. 4 The Actors i n S o c i a l Exchange Statements making up an.exchange theory of i n t e r a c t i o n are meant to apply equally to both p a r t i e s to a r e l a t i o n s h i p , but i t i s easier to frame the statements i n terms of a f o c a l person, whom we s h a l l c a l l P. 2 Although someeexchange th e o r i s t s would question the importance of giving the actor the capacity to c a l c u l a t e rewards, costs, and the p r o b a b i l i t y of rec e i v i n g both, most researchers would usually employ some model of P as a decision maker (acknowledging that there are many s i m i l a r i t i e s between an exchange and a decision theoretic model of man). Person i s assumed to be a maximizer. He has several goals or needs, and pursues things which give maximum s a t i s f a c t i o n of these needs. In th i s context, P i s faced with two general problems: which needs s h a l l he t r y to s a t i s f y , i n the near future; and where s h a l l he go to get the things he needs. The f i r s t involves the way i n which P assigns value to resources, and we s h a l l address t h i s ques-t i o n below. The second involves a l t e r n a t i v e s , or opportunities for meeting P's needs, and we s h a l l delay t h i s problem to a l a t e r point i n the chapter. Valuation of Resources For purposes of discussion, we w i l l f i r s t consider the actor, P, and two commodities or behaviours X and Y (such as help i n Math and help i n En g l i s h ) . So long as X and Y can s a t i s f y some of P's needs, they constitute resources. Reward refe r s to resources gained; costs are the resources f o r e -gone i n obtaining the reward. We designate p r o f i t as the net r e s u l t when costs 3 are subtracted from rewards. Because P i s a maximizer, so long as the reward 5 value of Y exceeds the costs i n X l o s t i n obtaining Y ( i . e . , so long as h i s p r o f i t i s p o s i t i v e ) , P w i l l take actions to obtain Y. Such a c t i o n i n d i c a t e s 4 a preference f o r Y over X. What i s the basis f o r saying that the rewards to P of Y exceed the costs to him of X? The o r i g i n a l source of such valuations i s that P has learned to need and l i k e c e r t a i n resources more than others, and can order them according to how much s a t i s f a c t i o n they give. We w i l l not be concerned here with how and why P has been s o c i a l i z e d to need Y more than X, and thus gets more reward from Y than X (as, f o r example, when we say P values a career more than marriage, or coffee more than t e a ) . We w i l l assume that although i t may be d i f f i c u l t to a s c e r t a i n what kinds of things P wants, the researcher, and other actors, can f i n d t h i s out by looking at P's past behaviour, to see what he and other people l i k e P have pursued, and by asking P what he values. Peter's bad grades i n Math-would give us an i n -d i c a t i o n that he would value help i n Math, and value i t more than help i n English. General preference orderings based on a h i e r a r c h i c a l ordering of needs are l i k e l y to be r e l a t i v e l y constant f o r the short run, and may be 5 regarded as given. We w i l l simply assume that we have discovered a small set of resources that P does value, and that he ranks each about equal i n importance. I t i s then possible to focus on a second sort of value, which hinges on the quantity of a given resource P has received i n the recent past, r e l a t i v e to other resources. 6 From the psychology-of reinforcement, we know that the more P has obtained of resource Y i n the immediate past, the le s s value he w i l l receive from successive equal increments of Y. Our student Peter could not spend 6 endless hours getting help on Math problems from Olga, as he would eventu-a l l y become sati a t e d . Thus, the rewards to P of successive increments of Y are assessed r e l a t i v e to the amounts of Y already possessed. Further, i f P i s giving up X (costs) to get Y, as h i s store of X diminishes, successive increment's of X w i l l constitute higher and higher costs to P. P w i l l reach a point when the costs of what he i s giving up w i l l equal the gain of the l a t e s t increment of Y. When he reaches t h i s point, he gains nothing i n trading X f o r Y, or v i c e versa. P has reached an equilibrium, and w i l l 7 stop, or pursue a t h i r d resource. I t i s a hard fa c t of l i f e that P-cannot have a l l the X, Y, Z, etc., that he wants, because resources are l i m i t e d . Since P desires several resources, when he has reached a c e r t a i n l e v e l of X, he w i l l gain more by d i v e r t i n g h i s a c t i v i t i e s to obtaining another valued resource. Thus, maximum p r o f i t i s obtained by getting an optimum balance of resources. In the cases of i n t e r e s t to us, he does t h i s by entering 8 intiboexchanges. Interdependence of P and 0 Exchange theory makes e x p l i c i t recognition of the fact that many of the events that a f f e c t P's a b i l i t y to get what he wants depend on the' 9 preference and behaviour of another actor. In the case of Peter and Olga, whether Peter gets h i s desired help i n Math, and how much;,depends on whe-10 ther he can mesh his needs and resources with hers. I t i s generally assumed that an exchange transaction w i l l only take place i f both parti e s perceive, that they w i l l be better o f f a f t e r the exchange has taken place than they were before i t . In other words, they must both receive a p o s i t i v e p r o f i t . 7 An exchange transaction requires two decisions: P's decision to give up X for Y, and O's decision to give up Y for X. Once P and 0 decide they can profit from an exchange, the basic problem is to agree on how much profit each should receive. It is obvious that the basis for exchange lies 11 in a difference in the relative valuation of each of two things by each of two parties. We take as given that P and 0 both want X and Y. The i n -verse relationship between the amount of a resource possessed (the resource  base) , and the subjective value of further increments of that resource, des-cribes a value function, mapping one set of values (here, units of resources), into another set (here, subjective rewards and costs to the actors). The particular value function we have described allows us to assume that the same objective amount of X or Y may have a different subjective worth to different actors. If this were not true, there would be no room for an increase in total subjective value through exchange — P would gain as much from having a lot of X, as from having a balance of X and Y. If we further assume that the actors are aware of the value func-tion by which rewards and costs are evaluated relative to resource base, we have a means of handling the question of the comparison of profits in exchange. So long as P also knows O's resource base (i.e., in addition to knowing the value function), he is able to take the other person's point  of view to assess how much 0 w i l l value given increments of X and Y in an exchange. This process is sometimes referred to as empathy, or role taking. The process of role taking is made easier for P i f he has had experience in a position similar to O's, or i f he has had an opportunity to observe similar 8 actors i n s i m i l a r s i t u a t i o n s to O's. The simplest case holds i f P can assume that 0 values things i n the same way P does ( i . e . , that 0 i s l i k e 12 P). The more d i s s i m i l a r 0 i s , the less accurate P i s l i k e l y to be when 13 he assumes the Other's point of view. I t i s undoubtedly true that a c e r t a i n amount of error i s usually involved i n P's determination of O's value function, but we are more interested i n the consequences of P's know-ing how 0 values resources, when he may or may not know O's resource base. P cannot e a s i l y use role-taking to determine the l a t t e r , as i t i s more sub-j e c t to v a r i a t i o n i n short time periods. Lack of information about some-one's resource base i n t e r f e r e s with the comparison of p r o f i t s and the s e t t i n g of terms i n an exchange. Thus, i n t h i s study, we w i l l begin with the case i n which P's and O's value functions are known to each other, and these value functions aret'fche same, except that there i s complementarity of resources possessed. This case lends i t s e l f to experimental c o n t r o l . What w i l l vary i s the information actors have about resource bases, and therefore, what would be the r e l a t i v e l e v e l s of s a t i a t i o n f or d i f f e r e n t resources. I t i s intended that the theory w i l l eventually be shown to apply to s o c i a l i n t e r a c t i o n s where P and 0 do not necessarily value commodities i n the same way, and to s i t u a t i o n s i n which P and 0 must f i r s t discover one another's 14 value functions. Most exchange relationships permit a v a r i e t y of terms, i n which the p r o f i t to P and 0 v a r i e s . The terms of the exchange can give equal subjec-t i v e p r o f i t to both, most of the p r o f i t to P and l i t t l e to 0, or t h i s may 15 be reversed. I t i s important for P and 0 to have information about how 9 much each has of X and Y, to determine how f a r the other can be pushed before the transaction collapses ( i . e . , i f P or 0 receives zero p r o f i t ) . According to Kuhn, P and 0 w i l l s e t t l e at a point mid-way i n the overlap 16 of t h e i r preferences. This process i s captured by the concept of f a i r , or equitable exchange. F a i r Exchange i s No Robbery While every actor prefers to gain as much as he can i n an ex-change, the very fact that both interactants are s t r i v i n g to get as much as possible , means that there are opposing forces on any given P: P's own preferences about h i s p r o f i t s lead him to attempt to pay as l i t t l e as. pos-s i b l e to 0 for as much as he can get, without l o s i n g the transaction. Other, however, has the same preference, but with regard to O's own p r o f i t s , which can be r e a l i z e d only i f P gets l e s s than, not more than 0. Under f u l l i n -formation, we expect the r e s u l t to be some sort of compromise. This has the net e f f e c t of pushing negotiated s o c i a l prices to a point where each i s 17 s a t i s f i e d that he i s doing as well as the other. This point we c a l l ' f airness'. Thus, r e c i p r o c i t y and fairness are not seen as moral o b l i g a -tions on:,the part of actors, but rather as a prudential r e a l i z a t i o n of what 18 i s possible and necessary to accomplish one's goal. We would a l l like to get something f o r nothing (as maximizers), but we do not usually act on the expectation that t h i s w i l l occur. In a s i m i l a r vein, Emerson stat e s : "What Gouldner c a l l s the norm of r e c i p r o c i t y may be l i t t l e more than the widespread human recognition of the contingencies i n t r i n s i c to a l l s o c i a l 19 exchange." 10 The expectation of f a i r exchange i s well documented i n the l i t e r -20 ature on equity. In addition, studies of b i l a t e r a l monopoly show that witLft f u l l information, subjects tend to a stalemate around a f a i r d i v i s i o n 21 of p r o f i t . In these studies, subjects had no resource base from which to 22 begin making o f f e r s , and therefore no costs. A study by W. Foddy showed that subjects with d i f f e r i n g resource bases tended to trade i n such a way that both partie s to the transaction received equal subjective p r o f i t s , de-fined r e l a t i v e to the"resources they possessed. Both of these studies lend support to the view taken here that P does not simply assess outcomes on the basis of his own p r o f i t s , but that a comparison with O's p r o f i t also 23 enters into h i s evaluation. As we have noted, such comparisons w i l l take into account the non-linear r e l a t i o n s h i p between the value of an exchange 24 un i t , and the objective amounts of such u n i t s . E m p i r i c a l l y t h i s means that i f P can exchange with 0^, who has a great deal of Y, and O2, who has less Y, P may pay more X to 0^ than to 0^, yet both transactions could meet the c r i t e r i o n of f a i r n e s s . Advantageous Exchange I t has been suggested, on the one hand, that P sets out to maximize his p r o f i t , and, on the other, that as a prudential person, he i s l i m i t e d by r e s t r a i n t s of .fairness. We have assumed that exchange i s t y p i f i e d by 25 'antagonistic cooperation', where the actors must exchange to p r o f i t , but where the terms of the transaction are i n c o n f l i c t . This structure of r e -wards i n exchange re l a t i o n s h i p s encourages i n P a desire to p r o f i t more than 1 1 0, since t h i s i s equivalent to a desire to maximize. (There i s evidence that people are more uncomfortable when an u n f a i r exchange favours Other, 26 than when i t favours s e l f . ) Given that P w i l l mainly have a v a i l a b l e exchange r e l a t i o n s h i p s which tend toward a balance of p r o f i t s , what options are open to him to r e a l i z e his desire to obtain more" than would be possible on a s t r i c t l y f a i r exchange? One obvious way i s to a f f e c t r e a l changes i n the valuation of behavioural products (e.g., P can decide he r e a l l y does not need Y) , or changes i n the s c a r c i t y of goods (for example, by l i m i t i n g Other's sources 27 of X). This topic has been dealt with i n d e t a i l by other researchers. If such r e a l changes are not p o s s i b l e , are there any other ways that P can' obtain more than a f a i r exchange would allow? Since i n d i v i d u a l s act on the perceived rates of supply and demand, P may be able to a f f e c t the per-ceived values and the perceived a v a i l a b i l i t y of resources. I f so, he could a f f e c t the perception of what i s , i n f a c t , f a i r . This i s possible i f he can control information that i s used to assess the balance- of subjective p r o f i t s i n an exchange. Control of Information The control of information as a t a c t i c f o r increasing one's out-28 29 comes i s of i n t e r e s t to symbolic i n t e r a c t i o n i s t s and exchange th e o r i s t s a l i k e . Both frameworks are interested i n the manner i n which Person acts to l i m i t the range of behaviours emitted by 0, so that 0 w i l l do those things P most prefers. P's shaping of O's perception or d e f i n i t i o n of the s i t u a t i o n 12 figures large among t a c t i c s open to the actor. This may involve conceal-ing the ' t r u t h 1 about oneself from other, while t r y i n g to discover as much 30 about other as possible. Advantageous exchange means maximization of outcome i n terms of the balance of rewards and costs for the i n d i v i d u a l . Maxi-mization, i n turn, i s l i k e l i e s t when the other's outcome values f o r various acts of yours are known to you, but your rewards and costs are not known to him. (Thibaut and K e l l e y , 1959). Peer r e l a t i o n s provide a good t r a i n i n g ground for learning to conceal one's rewards. Others should now know how important i t i s to you to have a p a r t i c u l a r baseball card, or that you have a duplicate of the one you are trading. Being too eager can r e -s u l t i n the other's demanding higher payment for the resources under h i s c o n t r o l , so that the a b i l i t y to keep one's 'cool' comes to have t a c t i c a l value for the child.31 Thus Weinstein argues that i n an exchange r e l a t i o n s h i p , i f P can conceal h i s reward and cost l e v e l s from Other, then 0 cannot assess what would constitute a f a i r exchange, even when 0 knows how P assigns value to X and Y, when he has c e r t a i n amounts of X and Y ( i . e . , even i f he knows P's value function). This p r i n c i p l e has been accepted by researchers such as Emerson, Thibaut and K e l l e y , S c h e l l i n g , and Kuhn, although systematic 32 -empirical support for i t i s lacking. But i s concealment always an advan-tage? How does i t a c t u a l l y work? Our purpose i n the rest of t h i s chapter w i l l be to explore the process i n more d e t a i l , i n order to delineate the conditions under which t a c t i c s involving control of information are l i k e l y to succeed. In order to do t h i s , i t i s f i r s t necessary to further develop the following three areas: l),_the mechanics of concealing information about p r o f i t i n exchange s i t u a t i o n s ; 2) the importance of a l t e r n a t i v e exchange rel a t i o n s h i p s to the process of a r r i v i n g at the d i v i s i o n of p r o f i t i n a transaction; and 3) the desire by Other for information about P's p r o f i t s . 13 As we s h a l l see, 2) and 3) can severely l i m i t the advantage claimed for concealing information. Concealing P r o f i t i n Exchange Transactions Let us assume that P and 0 have discovered that t h e i r r e l a t i v e valuation of X and Y are such that an eexchange w i l l provide p o s i t i v e pro-f i t to both, and f u r t h e r , that the overlap of t h e i r preferences i s large enough to allow more than one set of terms. The more P knows about how 0 values X*and Y, and what amounts of each he has, the better i s he able to compare his own gain with O's, £0 know the value 0 w i l l attach to increments of X and Y, and to a n t i c i p a t e the sort of terms i n a transaction 0 i s l i k e l y to consider f a i r and acceptable. I f P can at the same time conceal his true i n t e r e s t s , he can argue for a larger share of the t o t a l p r o f i t , as i f i t were the f a i r s o l u t i o n . In the language of negotiation, we can say that Other would be forced to what Bartos c a l l s a 'soft' bargaining p o s i t i o n , because Person knows the point of minimum p o s i t i v e p r o f i t to Other. P can l i m i t 0 to t h i s point, by pretending that he (P) w i l l not p r o f i t by any other agreement than the one y i e l d i n g minimum p r o f i t to 0. I f P i s not hampered by Other having r e c i p r o c a l information, he can adopt a "hard1- strategy, involving 33 high i n i t i a l demands, and small concessions i n negotiation. Where there i s f u l l symmetric information, of course, a hard bargaining strategy on P's part i s l i k e l y to be met by an equally intransigent p o s i t i o n from 0. Emerson states that i f P conceals h i s p r o f i t s from 0, P w i l l attempt to obtain an advantageous exchange using whatever means are a v a i l -14 able, but he claims that "...when the par t i e s make assumptions or have knowledge about X (the p r o f i t to be shared), t h e i r judgements concerning equity and d i s t r i b u t i v e j u s t i c e enter, probably operating as r e s t r a i n t s 34 upon the use of bargaining power." This argues that deviation from norms of fairness i s more l i k e l y i f there i s secrecy, because the p o s s i b i l i t y of sanction i s lower i f the deviation i s not recognized. Another aspect of the advantage for P i n having asymmetric i n f o r -mation i s that Other i s not aware that P i s being a hard bargainer, t r y i n g 35 to gain an u n f a i r p r o f i t . Cummings, et a l . , and others have shown that a hard i n i t i a l bargaining stance by P has the e f f e c t of lowering the l e v e l of a s p i r a t i o n of h i s opponent, and of leading to higher payoffs for P than 36 does a f a i r strategy. L i e b e r t , et a l . also found that a bargainer unin-formed of P's payoffs for various agreements made use of P's i n i t i a l bids -to decide what was a reasonable agreement. In the absence of f a c t u a l i n -formation about P's resource base, 0 has to r e l y on cues emitted by P as to P's p r o f i t s , and.the minimum terms of a transaction that P w i l l f i n d acceptable. Of course, i t i s even better f o r P i f 0 simply assumes that P w i l l suggest f a i r terms, but t h i s i s not necessary f o r P to have an advan-tage, so long as he i s the only exchange partner a v a i l a b l e to 0. 37 A f t e r Kuhn, we w i l l r e f e r to the concealment of information about resources and p r o f i t s to gain advantage as 'informational t a c t i c s . ' Because p r o f i t varies with both rewards and costs, t h i s suggests two pos-s i b i l i t i e s f o r P •— he can conceal h i s true rewards, his true cost, or both. Let us examine these t a c t i d s i n more d e t a i l . 15 Concealing true rewards: Weinstein's idea that P should not l e t 0 know 'how important i t i s to him to have a p a r t i c u l a r baseball card' i s rela t e d to the idea that i f P has very l i t t l e Y, any small amount more w i l l be highly valued, and w i l l produce a r e l a t i v e l y large subjective reward.-. I f he canconceal h i s lack of Y (and therefore the true value of Y to him), he may be able to get more absolute units of Y by giving the impression that he does not value Y as much, as i n fact he does. I f 0 does not know how much reward P gets from a given increment of Y, 0 may f e e l that he must, give up more Y for a given increment of X, i n order to get into i n t e r a c t i o n with P. P then ends up with more p r o f i t than does 0 — an advantageous exchange. The t y p i c a l ploy here i s to create the impression of le a s t i n t e r e s t , exemplified by Tom Sawyer's s k i l l f u l handling of the whitewash-ing. In this case, P (Tom) a c t i v e l y engaged i n deception concerning the 38 reward value of whitewashing. Further, Kelley claims that "concealing information about one's s i t u a t i o n has possible advantages of successful deception", because P can delay the decision to deceive, without giving away h i s true p o s i t i o n . In this d i s s e r t a t i o n , we w i l l focus on the process of withb'l.dingginformation. Both witholding and concealing information are used with the same intent by P — to manipulate O's perceptions i n a manner that serves P's i n t e r e s t s . Concealing true costs: Concealing the subjective value of rewards i s not the only way advantageous exchange can occur. P may conceal h i s true costs. The idea here i s that i f 0 thinks P has a l o t of X, i . e . , the commodity 0 i s asking f o r , that P therefore assigns r e l a t i v e l y low value to 39 an increment of X. 0 w i l l not f e e l he has to give up a large amount of Y 16 (what P wants)_ to balance the exchange, because P can a f f o r d to be generous. The less Y an Other has, the more he w i l l f e e l free to ask f o r a greater absolute amount of X r e l a t i v e to the Y he gives up. However, i f 0 does not know that P has a l o t of X, or can be led to believe that P has very l i t t l e X, then even a small increment of X has high subjective cost to P. As a r e s u l t , Other w i l l f e e l he cannot ask f o r as much X for a given amount of Y. P again makes an advantageous exchange. Small boys trading hockey 40 cards w i l l frequently use a t a c t i c c a l l e d 'begging' — getting cards from someone who has a l o t , and giving none i n return. Thee'begger' often has as many cards as the person he begs from, but a successful begger knows enough to leave these cards at home. While the discussion above has implied that p a r t i e s trade i n ob-j e c t i v e l y comparable units (cards f o r cards, time for time, etc.) i n many cases where the units are d i s s i m i l a r , i t i s the function of prices to de-f i n e what can be given i n exchange f o r what, and from t h i s base can be de-41 fined what constitutes excessive wealth,poverty, etc. Having established our d e s c r i p t i o n of exchange r e l a t i o n s h i p s i n which no one has a power advantage, where an expectation of f a i r exchange e x i s t s , and where there i s some opportunity for P to manipulate the kinds of exchanges that are perceived to meet the requirement of f a i r n e s s , we can turn our attention to the t a c t i c s that Other can employ. An exchange i n t e r -a c t i o n , as pointed out e a r l i e r , cannot be thought of simply as something P does with, and to, Other.pOther i s at the same time t r y i n g to do something with, and to, P. We would l i k e to argue that while people w i l l indeed attempt to work advantageous exchanges when t h e i r resources are concealed, 17 that the failure to provide 0 with information places limits on the success of P's tactics. This leads us f i r s t to the question of how alternatives affect the terms of a transaction. Alternative Exchange Relationships So far we have pictured P and 0 in an isolated dyad;with the pos-sible exception of romantic love dyads, this is not a very accurate picture of relationships in the real world. Actors usually have more than one po-tential relationship into which they could enter. Whether a particular Other is chosen by P depends on whether P prefers that relationship to alternate transactions. Since we claim that P assesses exchange relationships in terms of his own profit, and how his profit compares with O's, the principle of maximization of profit says P w i l l prefer the alternative that is most appealing on these two dimensions. In any P-0 dyad, P w i l l not have to give any better terms than he would give in his next best alternative. The same limit holds with respect to O's alternatives. If one party in a dyad has more or better alternatives than the other, we say he has a bargaining 42 advantage, or bargaining power. In this dissertation, however, we are not interested in real power that results from an unequal distribution of alternatives. Rather, we wish to ask what w i l l happen when P and 0 have equal numbers of alternatives (some or none), which differ in the amount of information available concerning the resource bases. Commenting on the failure to provide alternatives in experimental studies of bargaining, Kelley and Schenitzki note: 1 8 It i s probably most common i n bargaining s i t u a t i o n s , including the s t r i c t l y economic ones, that each party has one or more a l t e r n a t i v e sources with which to deal, should he be unable to reach agreement with the party at hand.43 As soon as there i s more than one p o t e n t i a l partner, we may ask what sorts of people w i l l be preferred as partners, and what e f f e c t s hiding resources i s l i k e l y to have. Other's Preference for Information about P Consider Other, now, as the f o c a l person. What we have said about the importance of a l t e r n a t i v e s applies with equal force to 0. We s t i p u l a t e that 0 i s unable to conceal information from P. As an i n i t i a t o r (designa-ted 0^ ) Other w i l l want to assess p o t e n t i a l transactions i n terms of his own p r o f i t r e l a t i v e to P^ (where P r i s the r e c i p i e n t of an i n i t i a t i o n from 0^), and i n terms of how much i t i s necessary to o f f e r P r to ensure a transaction. As a r e c i p i e n t ( c a l l him 0 Other w i l l want to see whether he i s being asked for a f a i r trade, or whether he i s paying excessively for a resource. In either r o l e , 0 w i l l want information about P. As a r e s u l t , i n the context of asymmetric information, where O's p o s i t i o n i s known to various P's, 0 may avoid contact with those P's about whom he has l i t t l e information. If he does, t h i s w i l l surely constrain the advantage P finds i n concealment. If people t y p i c a l l y have an expectation of f a i r exchange.yis there any basis f o r p r e d i c t i n g that P w i l l avoid exchange r e l a t i o n s i n which he does not know P's resource base? If 0 i s unable to assess whe-ther a transaction has been/is f a i r , he may assume that P i s constrained 19 to act f a i r l y , j u s t as 0 i s ; or he may assume that P, whose p r o f i t s he 44 cannot see, i s motivated to deceive him; or he may assume that either guess i s equally l i k e l y . Or he may assume nothing. In the l a t t e r three cases, 0 i s l i k e l y to prefer a l t e r n a t e , d i f f e r e n t exchange r e l a t i o n s h i p s , where he can assess the balance of p r o f i t s . An exchange partner about whom 0/ has no information might turn out to be a better a l t e r n a t i v e i n f a c t , i . e . , he might be w i l l i n g to pay a better p r i c e for Y. However, i f the l a t t e r were true, 0,v/may think that P has no reason to conceal h i s r e -s' / sources i n the f i r s t place. While i t seems reasonable to posit a p r e f e r -4 5 ence for information, we would l i k e to have some evidence that such a preference does e x i s t , because t h i s i s the basis on which we argue that information control has l i m i t s as a t a c t i c a l advantage. Studies of bar-gaining and exchange do not give any hard evidence f o r a preference f o r information, though i t s importance i s i m p l i c i t i n the desire to avoid un-f a i r exchanges. However, there are some studies concerning the v a l i d i t y of r a t i o n a l i t y postulates i n decision making that i n d i c a t e a preference of the sort we have described. 4 6 Research i n i t i a t e d by Daniel E l l s b e r g has shown that subjects often prefer a bet for which the p r o b a b i l i t i e s of events are known unambi-guously (r i s k y b e t s ) , to bets f o r which the p r o b a b i l i t i e s of events and outcomes are completely unknown (uncertain). T y p i c a l l y , subjects i n these experiments have to predict which colour would be sampled from e i t h e r : a) an urn with 50 red and 50 black b a l l s (risky) b) an urn with a t o t a l of 100 red and black b a l l s i n unknown propor-tions (an uncertain bet). 20 The subject could decide to make either bet a or bet b. He won $1 i f he c o r r e c t l y predicted the colour of a b a l l that would be drawn. Making the assumption that the best estimate of the proportions for the second urn i s .5 black and .5 red, the expected value for a bet on "red" i n either a or b i s .5 x $1 = $0.50. (Where the expected value i s calculated as the value 4 7 of an outcome times the probability of the event leading to the outcome.) Subjects should therefore be i n d i f f e r e n t between the two urns. However, El l s b e r g and others found that subjects tended to prefer the bet with the urn i n a, where the proportions of red and black were known ( r i s k y ) , and 4 8 would pay up to $0.36 to avoid the uncertain bet. The preference for the r i s k y a l t e r n a t i v e held even when i t s expected value f e l l a b i t below $0.50, and the expected value of the uncertain bet remained the same. E l l s b e r g claimed that the ambiguity of information a v a i l a b l e about outcomes (which varies with the type, q u a l i t y , amount, and source of information), affected the confidence a person had i n his estimates of expected value. The more ambiguous the information used to c a l c u l a t e expected value, the lower the person's confidence i n the estimate. This led subjects to devalue i n some manner the ambiguous a l t e r n a t i v e . While such a conservative evaluation may v i o l a t e r a t i o n a l i t y p r i n c i p l e s , i t appears to come c l o s e r to representing ac t u a l choice behaviour i n such s i t u a t i o n s . E l l s b e r g claimed that ambi-guity i s a matter of degree, and w i l l vary with how much information P has, whether i t i s v e r i f i a b l e , whether the source i s trusted, and so on. He did not attempt to test the r e l a t i o n s h i p between perceived ambiguity and these other v a r i a b l e s . 21 Ellsberg's study points to a perceived r e l a t i o n s h i p between the quantity and q u a l i t y of information about a l t e r n a t i v e courses of a c t i o n , and O's assessment of how an a l t e r n a t i v e w i l l further h i s attempts to maxi-mize his gain. It i s not a major extension of the argument to claim that i f none of P's a l t e r n a t i v e s have well-defined p r o b a b i l i t i e s f o r events ( i . e . , they are a l l uncertain), they may vary i n the degree of ambiguity, and that P w i l l prefer the l e a s t ambiguous a l t e r n a t i v e , i . e . , that for which he has r e l a t i v e l y more and better information. The extension to s o c i a l i n t e r a c t i o n i s obvious. O.'s information 1 about P's resources and value function may c e r t a i n l y vary i n both q u a l i t y and quantity, and r e l i a b i l i t y of source. Direct information, or information not under P's c o n t r o l , i s l i k e l y to be considered l e s s ambiguous than cues 49 which P i s able to manipulate. The le s s ambiguous 0_^'s information about P's resources, the more confidence he w i l l have i n both h i s assess-ment of the fairn e s s ( r e l a t i v e p r o f i t ) of a transaction, and i n h i s e s t i -mate of the p r o b a b i l i t y that a given i n i t i a t i o n of exchange he makes to P w i l l be accepted. If the q u a l i t y and quantity of information do a f f e c t 0, 's estimates of the expected value of exchange transactions, i t should show i n h i s behaviour, i . e . , i n h i s i n i t i a t i o n s and acceptances of exchange. Under the assumption that 0^ i n i t i a t e s where he ant i c i p a t e s the best p r o f i t , then, other things being equal, we expect him to choose an a l t e r n a t i v e with the least ambiguous information, or to seek more information before he pro-ceeds. 0 may estimate that a P with concealed resources i s as l i k e l y to be a better as a worse partner than someone whose resource base i s known, 22 i n the sense that an urn i s as l i k e l y to have more than .5 red b a l l s , or le s s than .5 red b a l l s . This may make the i n i t i a l estimate of the d e s i r e -a b i l i t y of an exchange with an 'unknown' partner equal to that of an ex-change with a partner whose resource base i s known. Even so, the greater ambiguity of the information about the P with a concealed resource base w i l l depress O's estimate of the expected value of that a l t e r n a t i v e . I f two a l t e r n a t i v e partners seem to d i f f e r only with respect to information a v a i l a b l e about them, 0 w i l l prefer the exchange r e l a t i o n s h i p where he knows P's resource base. In sum, i t seems that even i f we turn out to be wrong i n our assumption that people d e s i r e information i n order to compare  p r o f i t s , there may be a general preference, independent of comparison, that arises from the desire to assess accurately the p r o b a b i l i t i e s of d i f f e r e n t outcomes. This has clear implications f o r P i f he happens to be busy withold-ing information from 0. We are now i n a p o s i t i o n to describe some possible d i s t r i b u t i o n s of information across a l t e r n a t i v e exchange partners, and to make pre d i c -tions about the f a t e of informational t a c t i c s aimed at obtaining an advan-tage. While our main i n t e r e s t i s i n the case of asymmetric information, we w i l l o u t l i n e the cases of symmetric information and symmetric ignorance as well. The D i s t r i b u t i o n of Information Across A l t e r n a t i v e s To s i m p l i f y , l e t us assume that P i s able to conceal or reveal information about his resource base, and O's resource base i s always known 23 unambiguously to others. Then we w i l l conceptualize the two v a r i a b l e s of information and a l t e r n a t i v e rdationships as dichotomous: 0 may have i n f o r -mation about P, or he may have none (except for cues c o n t r o l l e d by P); and P and 0 may each have no a l t e r n a t i v e exchange r e l a t i o n s h i p s , or the same number of a l t e r n a t i v e s . To begin, we w i l l say that i f a l t e r n a t i v e s e x i s t , they vary only i n the amount and type of information a v a i l a b l e about them ( i . e . , they do not d i f f e r i n the amount of p r o f i t they o f f e r ) . Six cases can be distinguished. We w i l l l i s t them, making predictions about the pro-bable success of t a c t i c s . Where i t i s relevant, evidence for the p r e d i c -tions w i l l be noted. I The Isolated Dyad: No A l t e r n a t i v e s 1. Symmetric Ignorance In t h i s case, P and 0 have ro information about how the other pro-f i t s from a given transaction. On the basis of the posited preference for information about Other, we would expect the lack of information to create suspicion and caution, and that the bargainers would gradually f e e l t h e i r 50 way to a mutually acceptable agreement. This does not mean that P and 0 w i l l therefore decide to mutually reveal information about t h e i r p r o f i t s . Kelley, Beckman, and Fischer found that even i f the subjects i n a mixed motive game were given the opportunity to reveal t h e i r p r o f i t l e v e l s and costs, they did not do so, and that b i l a t e r a l bargaining was characterized by l y i n g and d i s t r u s t of the -information provided by the other person: 24 Open, honest communication affords one means by which the negotiators can a r r i v e at r e a l i s t i c and equitable goals, but each person's hope for a greater outcome than such communi-cations permit (and h i s expectations that the other person has the same hope) motivates communication that i s not open, honest, or trusted.^1 The statement by Kelley et a l . suggests that exchange has s t r u c t u r a l charac-t e r i s t i c s that lead to p a r t i c u l a r assumptions by P and 0 about the Other, i . e . , that he w i l l want to get more than a f a i r share, and that he w i l l not nec e s s a r i l y represent his p o s i t i o n honestly. 52 Fischer required subjects to negotiate for a share of a fixed number of points. P r i o r to negotiation, the experimenter provided each subject with a 'minimum necessary share', which he had to obtain before he could p r o f i t from an agreement. Subjects did not know what the other per-son's minimum necessary share was, and consequently, could not know for c e r t a i n whether a p a r t i c u l a r d i v i s i o n of the t o t a l payoff was more p r o f i t -able to other than to s e l f . In t h i s s i t u a t i o n , subjects would l i e about the minimum l e v e l of reward required, and managed to do better than oppon-ents who had a capacity to impose fi n e s f o r non-agreement. If l y i n g i s not pos s i b l e , i t seems that symmetric ignorance w i l l not advantage either party, 53 and w i l l lead to agreements that vary around an equal d i v i s i o n of p r o f i t s . So long as P i s l i m i t e d to exchange with only one other person, his s u s p i -cion and desire f o r information may disrupt or i n t e r f e r e with the exchange, and cause extended or d i f f i c u l t negotiation, but the partners w i l l probably continue to i n t e r a c t . If extended negotiation i s allowed, subjects would probably engage i n attempts to gain information about the opponent, for example, by observing his reaction to d i f f e r e n t i n i t i a t i o n s . 25 2. Symmetric Information In our previous discussion of how the terms of an exchange are arriv e d at i n a dyad, we noted the tendency to f a i r n e s s , and c i t e d evidence from b i l a t e r a l monopoly studies that actors would tend to a stalemate 54 around an even d i v i s i o n of p r o f i t s . Many of these studies show that sub-j e c t s w i l l remain i n dyads, and eventually reach agreement, i f they are not 55 given the a l t e r n a t i v e of q u i t t i n g , although Kahan found that subjects w i l l make use of a 'no-agreement' option when they cannot f i n d a s u i t a b l e 56 compromise, and there are no a l t e r n a t i v e partners. Studies of contrac-57 t u a l norm formation i n dyads show that, lacking a t t r a c t i v e a l t e r n a t i v e s , subjects w i l l even remain i n the r e l a t i o n s h i p where theLf . p r o f i t l e v e l i s below that of t h e i r partner. It appears on the basis of such fi n d i n g s , that people prefer some agreement to no agreement, so long as i t increases the l e v e l of t h e i r rewards. However, i t would not be wise to accept such a statement unconditionally. The u n a v a i l a b i l i t y of a l t e r n a t i v e partners, and the i m p l i c i t assumption that the task of bargaining experiments i s to reach agreement, probably increase the rates of agreement above what might be found i n a 'rea l world' bargaining task. Even labour negotiators, though they have no a l t e r n a t i v e bargaining partner, have the no-agreement a l t e r n a t i v e of s t r i k i n g , and management has the option of lock-outs. S t r a -tegies such as these probably prevent large deviation from f a i r n e s s when power of the pa r t i c i p a n t s i s equal. 3. Asymmetric Information Weinstein's d e s c r i p t i o n of t a c t i c s for advantage i s most applicable to the context i n which P knows how 0 p r o f i t s i n a transaction, but 0 i s 26, ignorant of P's gain. Under such circumstances, P i s most l i k e l y to be able to take advantage of one-sided information. He w i l l be able to 'play hard to get', and to ensure the best possible deal for himself within the l i m i t s set by Other's minimum p r o f i t . B i l a t e r a l monopoly bargaining experi-58 ments employing an informed stooge and an ignorant subject, i n d i c a t e that the former can use his information to formulate a hard bargaining strategy 59 that forces the ignorant person to lower his expectations of p r o f i t . I f , 60 i n addition, P could l i e about his p r o f i t , as i n Fischer's study, an advan-tage would very l i k e l y accrue to the more informed member of the dyad. Nevertheless, he w i l l s t i l l have to contend with suspicion on O's part. (It seems paradoxically true that the more 0 i s l i k e l y to t r u s t P, as i n the case of close f r i e n d s , or p o s i t i v e c o r r e l a t i o n of rewards, the l e s s l i k e l y P i s to want to get an advantage.) II A l t e r n a t i v e s to the Dyad 4. Symmetric Ignorance The case i n which neither P nor 0 has information about the other's p r o f i t s , and a l l a l t e r n a t i v e r e l a t i o n s h i p s are t y p i f i e d by lack of i n f o r -mation, i s l i k e l y to be an aggregate version of Case 1., except that d i f f i -c u l t y i n f i n d i n g a mutually acceptable agreement with one 0, would be more l i k e l y to lead P to engage i n sampling other p o t e n t i a l partners. The ex-tent to which he w i l l do t h i s w i l l depend on the time he has a v a i l a b l e , and the p o s s i b i l i t y of obtaining r e l i a b l e information about any Others. In a s i t u a t i o n such as t h i s , a reverse t a c t i c of giving out information might be 27 used, i . e . , P could d i s t i n g u i s h himself as the only person w i l l i n g to be honest, and thereby secure a transaction. This of course leaves him open to being taken advantage of. There i s l i t t l e empirical evidence concerning Case .4. , although i t does not seem t r i v i a l — the early stages of i n t e r -action, such as the sampling that goes on at p a r t i e s , i s probably of t h i s type. 5. Symmetric Information If P and 0 have perfect information about each other's resources and preferences, and where there are a v a i l a b l e substitutable a l t e r n a t i v e s 61 to the dyad, we approach the economist's model of pure competition. Here, the rates of exchange i n a l l transactions are u n i v e r s a l l y known, there i s no time pressure, a l t e r n a t i v e s are a v a i l a b l e and s u b s t i t u t a b l e , and there i s l i t t l e room for deviation from the f a i r or 'consensus' p r i c e . If P 62 demands more than a f a i r p r i c e , 0 simply moves to another a l t e r n a t i v e . Perfect or f u l l information implies not only i n i t i a l information about a l l P's resources, but also an updated i n t e l l i g e n c e of the terms of transaction 63 between s i m i l a r P's and O's. 6. Asymmetric Information If we accept the statement that P and 0 are normally not l i m i t e d to one exchange r e l a t i o n s h i p , the s i t u a t i o n i n which P witholds information from 0, but i n which 0 has other a l t e r n a t i v e s to P about whom he does have information, then t h i s s i x t h case i s a c r u c i a l one f o r examining the tac-t i c a l advantage of information c o n t r o l . I f P had d i s c r e t i o n over i n f o r -28 mation Others receive about him, he can follow three courses of ac t i o n : i ) he can withold information, giving o f f a minimum of cues, i i ) he can engage i n d e l i b e r a t e misrepresentation — simply, P can l i e about how he values the resources at stake, and how much he has of them. i i i ) he can be s e l e c t i v e about the release of information, giving out only those items that advance his aims. The success of a l l these t a c t i c s n e i i e s c n O's accepting at face value i n f o r -mation that originates from P. ( 0 may f i n d i t c o s t l y or impossible to test P's claims, or 0 may not have time to change to another r e l a t i o n s h i p . ) Because of the posited preference actors have for unambiguous i n -formation, and the fact that there often are a l t e r n a t i v e s a v a i l a b l e to both p a r t i e s i n an exchange, we argue that the success of informational t a c t i c s  i s not ensured i n simple exchange s i t u a t i o n s with asymmetric information. The constant sum nature of exchange also casts a competitive l i g h t on transactions, and interactants are more l i k e l y to be suspicious of a part-ner's information when that person's s e l f i n t e r e s t can be furthered by s k i l l f u l c o n t r o l of information. The theory and experimental design developed i n the rest of t h i s d i s s e r t a t i o n are addressed to Case 6. outlined above. Although i t can be argued that these persons would normally be able to choose occasions f o r concealment, and that these occasions w i l l depend on the means P has a v a i l -able for manipulating information, f o r purposes of beginning a study of a la r g e l y unexplored area, we w i l l s t a r t with a s i t u a t i o n where there i s either t o t a l or no d i r e c t information about P, and d i r e c t information about 29 0. In such a case, P's best hope for advantage rests on the presence of other P's whose resources are known to 0, and who ask 0 for terms in ex-change that are better than P would get in a f a i r transaction. (This would occur i f these other P's had small amounts of resource X (high costs) and/or relatively large amounts of Y (low rewards).) In this context, we w i l l investigate the extent to which Weinstein's claim that people w i l l u t i l i z e concealment to get advantageous exchanges holds, test the strength of a preference for exchanges about which the actor has unambiguous information, and observe the consequences that result from the interaction of the two processes. If we can find support for our conceptualization of processes involved in tactics of exchange for this case, i t should also shed light on the other cases. Of course, most ' r e a l - l i f e ' social exchange relationships are not clear-cut examples of any one of the combinations of alternatives and i n -formation we lis t e d . Alternatives to a dyad may exist, but i t w i l l involve some degree of effort and uncertainty to make the alternatives truly avail-able. People rarely have perfect information about other's values and re-sources, and even less about the transactions going on between members of other dyads — the actual rates of exchange may become known over time, as different pairs of actors in a group engage in transaction. We may expect, that in the early stages of interaction, an actor who can conceal his true resource position from others w i l l believe that i t i s to his ad-vantage to do so, even i f there are unconcealed potential partners for 0 in the group. To make such a prediction, i t is not necessary to assume that P is stupid — the costs to 0 of 'shopping' for and securing preferred 30 a l t e r n a t i v e s may lead him to accept whatever i n i t i a t i o n s he receives, es-p e c i a l l y i f 0 i s pressed f o r time, or has few i n i t i a t i o n s from which to choose. If the terms set i n transactions i n various dyads are not immed-i a t e l y known, P i s not u n r e a l i s t i c i f he thinks he may 'get away' with an advantageous exchange. The connectedness of the group and the a v a i l a b i l i t y of information about accumulating, as well as about i n i t i a l resources, w i l l c l e a r l y have 64 an e f f e c t on the success of P's t a c t i c s . Because of the preference f o r unambiguous information, we predict that a person who conceals h i s p o s i t i o n w i l l learn through experience that the perceived advantage i n asymmetric information i s not ne c e s s a r i l y an advantage at a l l . Over repeated i n t e r a c t i o n , as actors come to know more about each others' preferences through t h e i r overt behaviour, and as con-cealed partners get passed over for those who are open about t h e i r p o s i t i o n , we expect fewer attempts to use information c o n t r o l to gain an advantage. This i s what we mean when we claim that fairne s s i s a r e s u l t of i n t e r a c t i o n . The process w i l l proceed more quickly i f group members discuss the terms of transactions with each other, but we believe that the lack of d i r e c t com-munication does not prevent a gradual d i f f u s i o n of information through i n t e r a c t i o n , which i n the end w i l l have the same e f f e c t . The theory and experimental design developed i n the next two chapters provide a basis for t e s t i n g these predictions. 31 FOOTNOTES FOR CHAPTER 1 1. Major proponents of the exchange approach i n a s o c i a l psychological t r a d i t i o n are: R.M. Emerson, "Exchange theory: Parts I and I I ' , i n J. Berger, M. Z e l d i t c h , and B. Anderson, E d i t o r s , S o c i o l o g i c a l Theories  i n Progress, Vol. I I , Boston: Houghton M i f f l i n Co., 1972; J . Thibaut and H.H. Kelley, The S o c i a l Psychology of Groups, New York: John Wiley and Sons, 1959; P.M. Blau, Exchange and Power i n S o c i a l L i f e , New York: John Wiley and Sons, 1964; G.C. Homans, S o c i a l Behaviour;  Its Elementary Forms, New York: Harcourt, Brace and World, 1961, ( F i r s t e d i t i o n ) ; P. Secord and C. Backman, Soc i a l Psychology, New York: McGraw H i l l , 1974 (Third e d i t i o n ) . A more economic treatment i s given by A. Kuhn, The Study of Society: A Unified Approach; Lon-don, Tavistock Publications, 1966. 2. See, for example, Homans, op. c i t . , 1961; and Emerson, op. c i t . , 1972. 3. The formulation of rewards, costs and p r o f i t s used i n t h i s d i s s e r t a -t i o n most c l o s e l y follows that of Kuhn, op. c i t . , 1966, pp. 260-271. 4. Exchange t h e o r i s t s are aware of the tautologous nature of such a d e f i n i t i o n of rewards or reinforcement. (What does P prefer? Whatever he finds rewarding. What does he f i n d rewarding? What he prefers >' (chooses).) However, i t i s argued that the consequences deduced from a set of statements which include such a tautology are not themselves tautologous. Further, we assume independent p r i o r knowledge of what P finds rewarding. For a discussion of t h i s issue, see Kuhn, op. cit..; ', p. 275; Homans, op. c i t . , 1961, pp. 42-43. Also see R. Burgess and R. Akers, 'Are operant p r i n c i p l e s t a u t o l o g i c a l ? ' , The Psychological Record, 16, 1966, pp. 305-312. 5. Homans discusses the question of a generalized preference order that transcends momentary states of deprivation i n h i s S o c i a l Behaviour. He makes the point that a complete ordering of needs, and of the r e -sources to s a t i s f y them makes l i t t l e t h e o r e t i c a l sense i n some cases ( t h i r s t cannot be more important than hunger, water i s not preferred to food), and l i t t l e p r a c t i c a l sense i n others (P does not usually have the opportunity of s a t i s f y i n g a l l his needs at once, so that an ordering over everything i s i r r e l e v a n t ) . In addition, many resources are capable of s a t i s f y i n g several needs at once. Homans argues that since P's preference order i s r e l a t i v e l y stable over the short time periods i n -volved i n exchange transactions, and that P must only consider a small set of needs at once, that i t i s safe to take as given and constant h i s general preference order, and to examine the v a r i a t i o n i n value that d i f f e r e n t amounts of s p e c i f i e d resources X and Y bring to P. (See pp. 39-49 i n S o c i a l Behaviour for a f u l l e r discussion of these points.) 32 6. For example, see Burgess and Akers, op. c i t . , 1966. 7. In economics, t h i s i s referred to as the p r i n c i p l e of equimarginality. 8. Obviously, exchange i s not the only way to obtain resources. P may be able to produce valued goods, alone or i n cooperation with others. Such a process often involves investment of resources, as when group members put i n e f f o r t to produce a c t i v i t i e s or goods. In t h i s case, there may be an increase i n t o t a l resources, or what Kuhn c a l l s r e -source production. The resources produced can then be used i n further exchange, or resource t r a n s f e r . We are confining our i n t e r e s t to the l a t t e r case, which involves giving up one resource for another, while t o t a l resources remain constant. Kuhn argues that only resource trans-f e r q u a l i f i e s as exchange, but Homans and Emerson treat both forms. See: Kuhn, op. c i t . , pp. 268-69; Homans, op. c i t . , p. 75; and Emerson, op. c i t . , Part I I . 9. This i s a major d i f f e r e n c e between the exchange framework and decision theory. The l a t t e r , as Kuhn notes, does not take into account "whether opportunities l i e i n our own hands, i n the natural environment, or i n the hands of others". Kuhn, op. c i t . , 1966, p. 317. 10. See Emerson, op. c i t . , 1972, p. 59. 11. Kuhn, op. c i t . , 1966, p. 323. 12. Kelley and Thibaut r e f e r to t h i s as a 'stereotype u t i l i t y function'. See: H.H. Kelley and J.W. Thibaut, 'Group Problem.Solving'; i n G. Lindzey and E. Aronson, Edi t o r s , The Handbook of S o c i a l Psychology, Volume IV, Don M i l l s : Addison Wesley, 1969. 13. An example of t h i s i s a management representative who can a r r i v e at an i n i t i a l conception of how a union negotiator w i l l value a given agree-ment, because he can make c e r t a i n assumptions about the union man's value ordering, even though the union r o l e i s not symmetric to h i s own. Without a r e l i a b l e i n t e l l i g e n c e network, however, i t would be a more d i f -f i c u l t task to i n f e r the current l e v e l of s t r i k e funds. 14. Over extended i n t e r a c t i o n s , there w i l l frequently a r i s e a s o c i a l stan-dard that r e f l e c t s what s i m i l a r P's and O's have agreed on as a f a i r p r i c e , and P can use t h i s to help his value comparisons. S o c i a l stan-dards also make i t unnecessary to negotiate the terms for each exchange transaction. Thibaut and Kelley note that an important function of norms i s that they can substitute f o r the use of s o c i a l power to de-termine the terms i n s o c i a l exchange. See: Thibaut and K e l l e y , op.  c i t . , 1959. 33 15. The range solutions i s bounded by the zero p r o f i t points of P and 0, and the number of possible solutions i s r e s t r i c t e d by the d i v i s i b i l i t y of the units exchanged. 16. Kuhn, op. c i t . , 1966, pp. 325-332. Kuhn goes into more d e t a i l about comparability of units i n exchange than has been given here, but he does not discuss the importance of 'taking the other person's point of view' for the comparison of p r o f i t s . 17. See: R.M. Liebert, W.P. Smith, J.H. H i l l , and M. K e i f f e r , 'The e f f e c t s of information and magnitude of i n i t i a l o f f e r on interpersonal negotia-t i o n ' , Journal of Experimental S o c i a l Psychology 4_, 1968, pp. 431-444. In t h i s study of negotiation i n a b i l a t e r a l monopoly game, the authors concluded that "...where bargainers considered each other to be equals (in power and/or status) they f e e l that (equality of p r o f i t s ) i s the most they can hope f o r , but w i l l take more i f they can get i t . " 18. An i n t e r e s t i n g discussion of moral obligations as 'prudential maxims' and t h e i r place i n Hobbes' Leviathan i s given by T. Nagel, 'Hobbes on Obligation', P h i l o s o p h i c a l Review, 68, 1959, pp. 68-83. 19. Emerson, op. c i t . , 1972, p. 61. 20. E. Burnstein and S. Katz, 'Group decision involving equitable and o p t i -mal d i s t r i b u t i o n s of status', Chapter 12 i n C.G. McClintock, Experi- mental S o c i a l Psychology, Holt, Rinehart, and Winston, 1972; R.D. Prit c h a r d , 'Equity theory: A review and c r i t i q u e ' , Organizational Be-haviour and Human Performance, 4., 1969, pp. 176-211; M. Patchen, 'A conceptual framework and some empirical data regarding comparisons of s o c i a l rewards', Sociometry, 24^ 1961, pp. 136-156; R. Burgess and J.D. Gregory, 'Equity and inequity i n exchange r e l a t i o n s : an experi-mental re-examination of d i s t r i b u t i v e j u s t i c e ' , paper read at the annual meetings of the West Coast Conference for Small Groups Research, Hono-l u l u , Hawaii, 1971; and J.S. Adams, 'Inequity i n s o c i a l exchange', i n L. Berkowitz, ed. , Advances i n Experimental S o c i a l Psychology, 2., 1965, New York: Academic Press, pp. 267-299. 21. L.E. Fouraker, and S. Siegel, Bargaining Behaviour, McGraw H i l l , 1963. 22. W.H. Foddy, 'The formation of c l i q u e s i n c o l l e c t i v i t i e s as a consequence of i n i t i a l d i s t r i b u t i o n s of dimension of wealth', Unpublished Ph.D. d i s s e r t a t i o n , University of B r i t i s h Columbia, 1972. 23. There are several studies which show that comparison a f f e c t s a person's evaluation of outcomes, and h i s subsequent behaviour: K.E. Weick and B. Nesset,'Preferences among forms of e q u i t y 1 , Organizational Behaviour and Human Performance, _3, 1968, pp. 400-416; J.P. Sheposh and P.S. Gallo, 'Asymmetry of payoff structure and cooperative behaviour i n a Prisoner's Dilemma Game', Journal of C o n f l i c t Resolution, 17, 1973, 34 pp. 321-333; C.G. McClintock, Experimental S o c i a l Psychology, Holt, Rinehart and Winston, 1972. In addition, e x p l i c i t comparison has been shown to enhance competitiveness, i . e . , the d e s i r e to *do better than' someone else. This i s discussed i n the chapter by McClintock, and in M. Shubik, 'Games of St a t u s 1 , Behavioural Science, 16, 1971, pp. 117-129; and i n Fouraker and Siegel, op. c i t . , 1963. 24. It should be noted that a marginal u t i l i t y function i s not the only \aLuation function that may apply, and the general comments about f a i r -ness are not r e s t r i c t e d to such a function. So long as P and 0 know how each other values things, (whatever the function), then information about an actor's current p o s i t i o n with respect to t h i s function allows for assessment of f a i r n e s s . Fischer, f o r example, used a threshhold function, whare an actor's l e v e l of a s p i r a t i o n (induced by the experi-menter) determined the value to the subject of a given d i v i s i o n of payoffs. Actors would use information about a p a r t i c u l a r l e v e l of as-p i r a t i o n to assess the f a i r n e s s of agreements, and without such i n f o r -mation they were unable to see who p r o f i t e d most, even though the v a l -uation function was known to both p a r t i e s . See: C.S. Fischer, 'The e f f e c t s of threats on an incomplete information game', Sociometry, 32, 1969, pp. 301-314. 25. O.J. Bartos, Simple Models of Small Group Behaviour, New York: Col-umbia University Press, 1967, p. 268. 26. Pritchard, op. c i t . , 1969; and Burgess and Gregory, op. c i t . , 1971.G 27. See, for example, R.M. Emerson, 'Power dependence r e l a t i o n s ' , American S o c i o l o g i c a l Review, 27_, 1962, pp. 31-41; and H.A. Michener and R.W. Suchner, 'The t a c t i c a l use of s o c i a l power 1, i n J.T. Tedeschi, Ed., The S o c i a l Influence Processes, Chicago: Aldine, 1972. 28. For example, E.A. Weinstein, 'The development of interpersonal compe-tence', Chapter 17, i n D. Goslin, Ed., Handbook of S o c i a l i z a t i o n , Rand McNalley, 1968; E. Goffman, The Presentation of Self i n Everyday L i f e , Doubleday Anchor Books, 1959; and P.W. Blumstein, 'Audience, Machie-v e l l i a n i s m , and t a c t i c s of i d e n t i t y bargaining', Sociometry, 36, 1973, pp. 346-365. 29. See footnote 1. 30. Bartos notes that many s o c i a l i n t e r a c t i o n s have a dual nature — a 'getting to know you' phase, and a 'taking advantage of you' stage. If P can keep 0 from succeeding i n the f i r s t , he may have a decided advan-tage i n the second. Bartos, op. c i t . , 1967, p. 268. 31. Weinstein, op. c i t . , 1968, p. 767. 35 32. The supposed advantages of one-sided information are argued by many researchers, but are supported by anecdotal rather than systematic empirical t e s t s . Cf. Kuhn, op. c i t . , pp. 334-337 for a discussion of possible t a c t i c s i n bargaining; also 0. Bartps', 'Towards a r a t i o n a l emprical model of negotiations; S o c i o l o g i c a l Theories i n Progress, Vol.  I I , i n J. Berger, M. Z e l d i t c h and B. Anderson, Eds., HHoughton M i f f l i n , Co., 1972; pp. 239-286; and R.E. Walton and R.B. McKersie, A Behavioural  Theory of Labour Negotiations, New York, McGraw-Hill, 1965. 33. Bartos, op. c i t . , 1967, Chapter 12. 34. R.M. Emerson, 'Power and p o s i t i o n i n exchange networks', Paper presen-ted at national meetings" of American S o c i o l o g i c a l Association, 1971. Emerson has done some experiments concerning the e f f e c t s of conceal-ment on willingness to use a power advantage, and he notes that when r e l a t i v e p r o f i t s are not v i s i b l e , the more powerful person does tend to get more p r o f i t ( i . e . , to use his advantage) than when r e l a t i v e pro-f i t s are v i s i b l e . (Personal communication, December 1973). 35. L.L. Cummings, D.L. Harnett, and W.C. Hamner, 'Personality, bargaining s t y l e , and payoff i n b i l a t e r a l monopoly bargaining among European man-agers?, Sociometry, _36, 1973, pp. 325-344. G. Yukl, 'Effects of the opponents' '$nij'iaf"offer, concession magnitude, and concession frequen-cy on bargaining behahaviour', Journal of Personality and S o c i a l Psy-chology, 1974, 30, pp. 323-335; J.M. Chertkoff and M. Conley, 'Opening o f f e r and frequency of concessions as bargaining s t r a t e g i e s ' , Journal of Personality and S o c i a l Psychology, 1967, 1_, pp. 181^-85; 'H^HleKeliey, L.L. Beckman and C.S. Fischer, 'Negotiating the d i v i s i o n of a reward under incomplete information', Journal of Experimental S o c i a l Psy-chology, J3, 1967, pp. 361-98. S. Siegel and L.E. Fouraker, Bargaining  and Group Decision Making, New York: McGraw H i l l , 1960. 36. L i e b e r t , et a l . , op. c i t . , 1968, pp. 431-441. 37. Kuhn, op. c i t . , 1966, pp. 336-337. 38. H.H. Kelley, 'A classroom study of the dilemmas of interpersonal negoW t i a t i o n s ' , i n K. Archibald, Ed., Strategic I n t e r a c t i o n and C o n f l i c t , U niversity of C a l i f o r n i a Press, Berkeley, 1966, p. 60. 39. In the sense that a d o l l a r i s worth less to a m i l l i o n a i r e , than to a person on welfare. 40. Cummings and Harnett, op. c i t . , 1969, found that a subject with a steeply sloping u t i l i t y curve, which i s s i m i l a r to having a small resource base, conceded less during negotiations than did a person with a less-steep curve, implying that i t costs the former more to give up a given i n c r e -ment (absolute amount) of a reward. 36 41. For example, people may compare the worth of g i f t s exchanged;the value to P and 0 of hours spent helping another study (taking into account how much time each has to spare and how much he needs the help); peo-ple also seem to be able to assess whether someone has adequately r e -turned a favour, even though the objective units (lawn mowing for baby-s i t t i n g , for example), are not the same. A l t e r n a t e l y , d i s s i m i l a r commodities may each be equated to a common uni t , and these units compared. Money often serves the function of a 'common u n i t ' . 42. See, for example, the discussion i n Emerson, op. c i t . , 1972, pages 58-59; and H.A. Michener and R.W. Suchner, "The t a c t i c a l use of s o c i a l power", i n J.T. Tedeschi, Ed., The S o c i a l Influence Processes, Chicago: Aldine Publishing Company, 1972, pp. 239-286. 43. H.H. Kelley and D.P. Schenitzki, 'Bargaining', Chapter 10 i n C.G. McClintock, Experimental S o c i a l Psychology, Toronto: Holt, Reinhart and Winston, 1972. 44. There i s some evidence that i n competitive s i t u a t i o n s , people tend to assume that 0 i s trying to get as much as he can, and to l i m i t P's gain. (H.H. K e l l e y and A. Stahelski, 'Social i n t e r a c t i o n bases of coopera-' tors'' and competitors' b e l i e f s about others', Journal of Personality and S o c i a l Psychology, 16_, 1970, pp. 66-91.) Harsanyi suggests that the most common b e l i e f P w i l l entertain about 0 i s that 0 has the same motives as P, and w i l l employ s i m i l a r s t r a t e g i e s . See J.C.,Harsanyi, 'Bargaining i n ignorance of the opponent's u t i l i t y function', Journal of C o n f l i c t Resolution, 6_, 1962, pp. 29-38. 45. There could as e a s i l y be a bias toward assuming that P i s s i m i l a r to 0 i n terms of resource base as w e l l as value functions. 46. E l l s b e r g , D., 'Risk, ambiguity, and the Savage axioms', Quarterly Journal of Economics, 7^ 5, 1961, pp. 643-669; W. F e l l n e r , 'Distortion of subjective p r o b a b i l i t i e s as a reaction to uncertainty', Quarterly Jour-nal of Economics, 7_5, 1961, pp. 670-89; J.S. Chipman, 'Stochastic choice and subjective p r o b a b i l i t y ' , i n D. Willner, Ed., Decisions, Values and  Groups, New York: Pergamon Press, I960; and S.W. B e c k e r a n d F.O. Brownson, 'What p r i c e ambiguity? Or the r o l e of ambiguity i n decision making', Journal of P o l i t i c a l Economy, _72, 1964, pp. 62-73. 47. Introductions to decision theory abound. One good example: Wayne Lee, Decision Theory and Human Behaviour, New York: John Wiley and Sons, 1971. 48. S.W. Becker and F.O. Brownson, op. c i t . , 1964. 37 49. The issue of the c r e d i b i l i t y of the source of a communication when the source has a vested i n t e r e s t i s discussed i n : E.E. Jones and J.W. T h i -baut, 'Interaction goals as bases of inference i n interpersonal percep-t i o n ' , i n R. T a g i u r i and L. P e t r u l l o , Eds., Person Perception and Inter-personal Behaviour, Stanford, 1959; and D. Bramel, 'Determinants of b e l i e f s about other people', Chapter 4 i n J. M i l l s , Experimental S o c i a l  Psychology, Toronto: C o l l i e r MacMillan, 1969. There i s evidence that i f the r e c i p i e n t of information about P believes that P could p r o f i t from d i s t o r t i o n or s e l e c t i v e release of information about himself, the r e c i p i e n t has less confidence i n the information. Thus, information outside P's control i s more l i k e l y to be regarded by the r e c i p i e n t as y i e l d i n g a believable p i c t u r e of P's p o s i t i o n . Walton and McKersie, discussing the case of labour negotiations, point out that the use of spies and bugging devices i s directed at getting a 'true' p i c t u r e of the opponent's p o s i t i o n , i n d i c a t i n g that t h i s sort of information i s trusted over P's 'presentation' of i t . Although ' d i r e c t , v e r i f i a b l e ' cues are not immune to d i s t o r t i o n , we expect the r e c i p i e n t of such cues to have more f a i t h i n them than i f P i s sending the cues himself. See: R.E. Walton, and R.B. McKersie, A Behavioural Theory of Labour Negotia- tions , New York: McGraw-Hill, 1965, pp. 62-63. 50. H.H. K e l l e y , L.L. Beckman and C.S. Fischer, op. c i t . , 1967; H.H. Kelley and J.W. Thibaut, op. c i t . , 1969; H.H. Kelley and D.P. Schenitzki, op.  c i t . , 1972. 51. K e l l e y , Beckman and Fischer, op. c i t . , 1967, p. 363. 52. C.S. Fischer, 'The e f f e c t s of threats i n an incomplete information game', Sociometry, 32, 1969, pp. 301-314. 53. Fouraker and S i e g e l , op. c i t . , 1963. 54. See, for example, Fouraker and S i e g e l , op. c i t . , 1963; J.W. Thibaut and C L . Gruder, 'The formation of contractual agreements between par-t i e s of unequal power', Journal of Personality and S o c i a l Psychology, 11, 1969, pp. 59-65; R.M. Emerson, 'Power and p o s i t i o n i n exchange net-works', Paper presented at American S o c i o l o g i c a l Association n a t i o n a l meetings, 1971. 55. HTHK- Kel l e y and D. Schenitzki, 'Bargaining', Chapter 10 i n C.G. McClin-tock, Experimental S o c i a l Psychology, Toronto: Holt, Rinehart and Win-ston, 1972. 56. Kahan, J.P., 'Effect of l e v e l of a s p i r a t i o n i n an experimental bargaining s i t u a t i o n ' , Journal of ^Personality and S o c i a l ^Psychology, 8^, 1968, pp. 154-59. 38 57. For example, see J.W. Thibaut and C. Faucheux, 'The development of con-t r a c t u a l norms i n a bargaining s i t u a t i o n under two types of s t r e s s ' , Journal of Experimental S o c i a l Psychology, Vol. 1, 1965, pp. 89-102; J.W. Th'ibaut and C L . Gruder, op. c i t . , 1969; and P. Murdoch and D. Rosen, 'Norm formation i n an interdependent dyad', Sociometry, _33, 1970, pp. 264-276. Also R.L. Burgess and J.D. Gregory, op. c i t . , 1971. 58. Chertkoff and Conley, op. c i t . , 1967; Liebert et a l . , op. c i t . , 1967; YukeL, op. c i t . , 1974. 59. Some studies of the r o l e of information i n b i l a t e r a l monopoly bargaining focus on how ignorance or lack of information about what constitutes an equitable or f a i r agreement can work to a person's advantage by making him cautious about his own p r o f i t . Generally, these studies have found that i f one person knows what a f a i r agreement i s , and one does not, the knowledgeable person w i l l usually concede more than an ignorant one. While this may seem somewhat contradictory to the argu-ment being made here, the ignorant.person i n these cases does not seem to have been aware that h i s p r o f i t s were known to h i s partner, and for th i s reason, i s not subject to as great a pressure to be ' f a i r ' as i s ,j the knowledgeable player. However, Liebert et a l . used a s i m i l a r paradigm to the b i l a t e r a l monopoly experiments, and did not conclude that ignorance necessarily leads to b l i s s — a knowledgeable player could s u b s t a n t i a l l y a f f e c t the ignorant Other's expectations as to how high a p r o f i t he might achieve. L.L. Cummings and D.L. Harnett, 'Bargaining behaviour i n a symmetric t r i a d : the r o l e of information, communication, power-? and r i s k - t a k i n g propensity', Review of Economic Studies, 36_, 1969, pp. 484-499; D.L. Harnett and L.L. Cummings, 'Bargaining behaviour i n an asymmetric t r i a d ' , Chapter 2.5 i n B. Lieberman, Ed., S o c i a l Choice, Gordon and Breach Science Publishers, 1971; L i e b e r t , et a l . , op. c i t . , 1967. ' 60. Fischer, op. c i t . , 1969. 61. Kuhn, op. c i t . , 1966, pp. 375-378. 62. This does not deny the importance of an actor's resource base i n deter-mining the p r i c e he w i l l pay for Y; what varies when consensus prices p r e v a i l i s whether or not P w i l l choose to enter transactions. 63. I t may be that i f the s i t u a t i o n i s one of perfect of close to perfect competition, then a sample of a small number of O's w i l l provide stereo-t y p i c a l information about a l l , or almost a l l , p o t e n t i a l O's. 64. An example of how communication and connectedness of group members are necessary for norm enforcement i s given i n the study of 'revolutionary c o a l i t i o n s ' , by H.A. Michener and M. Lyons, 'Perceived support and up-ward mobility as determinants of revolutionary c o a l i t i o n behaviour', Unpublished paper, University of Wisconsin (undated). 39 CHAPTER 2 THEORY For purposes of constructing a theory based on the arguments presented i n Chapter 1, we w i l l conceptualize the s o c i a l actor, P, as a decision maker. P i s seen to choose between a l t e r n a t i v e actions, which, coupled with events i n the environment, w i l l lead to various outcomes, having subjective value to P. The events i n the environment occur with p r o b a b i l i t y between 0.0 and 1.0, and such p r o b a b i l i t i e s are estimated by P. To choose between a l t e r n a t i v e s , P i s described as comparing the sub-j e c t i v e l y expected value of each a l t e r n a t i v e , where t h i s SEV i s some function of the value of outcomes, and the p r o b a b i l i t i e s of events leading to d i f f e r e n t outcomes. Let us apply t h i s to our exchange s i t u a t i o n . In such a s i t u a t i o n , P i s faced with a set of decisions concern-ing: which of several a l t e r n a t i v e partners w i l l allow him to obtain as much reward as possible for as l i t t l e cost as possible — how much of O's resource P can ask f o r , without reducing the p r o b a b i l i t y  of a completed transaction — how much information P should reveal about h i s resources, and the value he puts on others' resources — how much P can t r u s t the information he has about p o t e n t i a l partners. P's a l t e r n a t i v e s thus involve a range of i n i t i a t i o n s of exchange that he can make ( i f he i s the i n i t i a t o r ) or decide to accept ( i f he i s the r e c i p i e n t ) , 40 and these exchange a l t e r n a t i v e s w i l l vary i n the s i z e and type of r e -sources involved, as well as with the c h a r a c t e r i s t i c s of p o t e n t i a l partners. The s i z e of an i n i t i a t i o n made by P determines h i s p r o f i t , but i n choosing the l e v e l of i n i t i a t i o n , P w i l l use information a v a i l a b l e to him about O's values and resources (such as d i r e c t knowledge, cues from 0, empathy, etc.) to assess how l i k e l y a given i n i t i a t i o n of exchange i s to succeed, and a transaction be completed. P's choice of a target for an i n i t i a t i o n w i l l r e f l e c t how P has assessed the value to him of transacting with a v a i l a b l e a l t e r n a t i v e partners (subject to the l i m i t a t i o n s of P's cognitive capacity — he cannot, of course, consider i n f i n i t e numbers of transactions). I f P has d i s c r e t i o n over information about hi s own p r e f e r -ences and resource base, he must also decide whether to withold or give out such information, and h i s decision w i l l depend on whether witholding information i s perceived to lead to increased expected value. Theepossible events i n t h i s s o c i a l context are acts emitted by Other — 0 may accept, or re;pct (or modify) P's i n i t i a t i o n of exchange, and 0 may i n i t i a t e exchange himself. Unlike a s i t u a t i o n i n which the decision maker is. faced with an i n d i f f e r e n t nature, i n a s o c i a l i n t e r a c t i o n context, the p r o b a b i l i t i e s of events are under the control of another actor, and are also subject to some degree of modification by P. Of p a r t i c u l a r i n -terest here are the r e l a t i v e p r o f i t s of P and 0, and how they a f f e c t the p r o b a b i l i t y that 0 w i l l accept P's o f f e r of exchange. The c r i t i c a l problem for P i s to f i n d ways of getting i n i t i a t i o n s accepted i n the face of the fact that as P's p r o f i t increases*,0's p r o f i t , and therefore the p r o b a b i l i t y of the transaction, w i l l decrease. 41 The outcomes of any exchange are the absolute and r e l a t i v e pro-f i t s to the actors. Theisoutcomes usually change the parameters of future exchanges — the p r o b a b i l i t i e s of i n i t i a t i o n and acceptance may increase or decrease as people s a t i s f y t h e i r needs f or one resource, and the d e s i r -a b i l i t y of a given exchange partner may change due to comparative or mar-ginal u t i l i t y reasons. The theory presented i n th i s chapter provides a basis f o r making predictions about the d i r e c t i o n and type of i n i t i a t i o n s of exchange that w i l l be made and accepted i n a group of in d i v i d u a l s i n which there are at least two valued resources d i s t r i b u t e d across the members of the group, and i n which the d i s t r i b u t i o n of resources i s not f u l l y known to a l l mem-bers. To provide t h i s b a s i s , the theory draws on three p r i n c i p l e s : 1) maximization of marginal u t i l i t y 2) comparative j u s t i c e or fairness 3) the a b i l i t y to take the point of view of others. There are four main sections i n th i s chapter: d e f i n i t i o n of concepts, scope conditions, assumptions, and a set of hypotheses. Operational hypo-theses are presented with the desc r i p t i o n of the research conducted to test the theory i n Chapter 3, and are not de t a i l e d i n the present chapter. D e f i n i t i o n of Concepts Several concepts are defined here which are used i n the develop-ment of the theory. The l i s t i s not exhaustive, but i s intended to f i x the meaning of the basic units of the theory. 42 Resource: a resource i s any behavioural, material or non-material com-modity that i s valued by i n d i v i d u a l s i n a s p e c i f i a b l e manner, and which 1 can be transferred from one i n d i v i d u a l to another. I n i t i a t i o n of exchange: an i n i t i a t i o n of exchange occurs when a per-son, P, off e r s an amount of some resource, say X, to another person, 0, in return for an amount of some other resource, say Y, from 0. Exchange i n t e r a c t i o n : an exchange i n t e r a c t i o n occurs when a person P makes an i n i t i a t i o n of exchange to another person 0 and t h i s i n i t i a -t i o n of exchange i s accepted by 0. (This i s also r e f e r r e d to as a transaction.) Exchange int e r a c t i o n s are generally observed when P i s short of, or desires, some resource Y, notes that 0 has some of resource Y, and that 0 i s short of some resource which P possesses. Resource base: the resource base of P i s the t o t a l amount of resources possessed by P at a given time. (P's wealth.) P's t o t a l resource base may include several d i f f e r e n t resources. Subjective value to P of a resource: the subjective value to P of a resource i s the worth of u t i l i t y that resource has for P. I t r e f l e c t s P's general preference ordering of d i f f e r e n t resources, and h i s current l e v e l of s a t i a t i o n f o r these resources. D i f f e r e n t i n d i v i d u a l s may have the same or d i f f e r e n t preference orderings over a given set of resources, and, of course, they may have d i f f e r e n t l e v e l s of s a t i a t i o n . Value function f o r a resource: the value function for a resource des-cribes how much subjective value d i f f e r e n t amounts of a resource provide to P. The subjective values of d i f f e r e n t resources as calculated by the 43 value function can be compared, i . e . , i n some common unit of value such as u t i l i t y . People may have s i m i l a r value functions even though they value d i f f e r e n t resources, that i s , even i f t h e i r preference 2 orders are not the same. 7. VaMeeposition of P on resource X: the value p o s i t i o n of P on r e -source X refers to the amount of X which P has at a given time. If the value p o s i t i o n i s known, the value function can be used to assess the subjective value of a further increment of X to P. 8. Net subjective p r o f i t to P,\ (in an exchange i n t e r a c t i o n with 6jf: the net subjective p r o f i t for P i n an exchange i n t e r a c t i o n with 0 i s the subjective value of what 0 gives him (reward), minus the subjective value of what he must give to 0 i n return (co s t ) . Net subjective pro-f i t s w i l l generally be referred to as P's p r o f i t , where p r o f i t = (sub--j e c t i v e reward-subjective c o s t ) . 9. F a i r exchange: a f a i r exchange between P and 0 i s one i n which the subjective p r o f i t to P i s perceived by P and 0 to be equal to the sub-j e c t i v e p r o f i t of 0. 10. Advantageous exchange for P-f (in an exchange with 0-) : P obtains an advantageous exchange with 0 when the subjective p r o f i t to P exceeds the subjective p r o f i t to 0. 11. P r o f i t overlap of P and 0: the p r o f i t overlap of P and 0 r e f e r s to the set of possible exchanges between P and 0, of resources X and Y, i n which P and 0 are able to gain a p o s i t i v e p r o f i t . An exchange of quantities such that P and/or 0 reach equality of marginal gain and 44 3 marginal cost ( i . e . , an optimum combination of resources) describes the upper l i m i t of the p r o f i t overlap, and the lower l i m i t i s the s t a -tus, quo (no exchange). I f P and 0 have no p r o f i t overlap, no exchange can occur; i f a p r o f i t overlap e x i s t s , any combination of p r o f i t s within the overlap gives each a p o s i t i v e p r o f i t , though not ne c e s s a r i l y an equal p r o f i t . D e f i n i t i o n 11 i s introduced mainly to provide a shorthand way of r e f e r r i n g to the fact that bothgparties must benefit for an ex-change to occur, but within this range of agreements, some may favour P more than 0, and v i c e versa. 12. Ambiguity of information: (about O's p r o f i t ) : ambiguity of informa-t i o n refers to a q u a l i t y of information that varies with the amount, type, r e l i a b i l i t y , source, and consistency of information, giving r i s e to P's degree of confidence i n h i s estimate of the expected value of a given transaction. The d e f i n i t i o n s of ambiguity and confidence are given together because, for purposes of t h i s theory, one i s the inverse of tbf;the other. Scope Scope Conditions Preliminary Statement We are interested i n simple exchange at a given point i n time (early stages of exchange r e l a t i o n s h i p s ) , i n which group members are moti-vated to gain more than others i n the group. The scope conditions must provide for a s i t u a t i o n i n which such motives can e x i s t , and where the means 45 to r e a l i z e a desire for advantage e x i s t s , but need not be employed. As advantage implies comparison, the opportunity to compare must also be present. We*argued i n Chapter 1 that c e r t a i n s t r u c t u r a l c h a r a c t e r i s t i c s of simple exchange enhance the desire for advantage. The 5constant sum nature of rewards i n simple exchange makes i t p r o f i t a b l e to P to do better than Other. The tendency f o r simple exchange to occur among people with 4 s i m i l a r wealth l e v e l s ( i . e . , peers), makes a desire to compare favourably with one's exchange partner l i k e l y , as we believe people are most concerned to guage t h e i r success with that of s i m i l a r others. In short, we do not l i m i t the theory to a set of i n d i v i d u a l s who have 'competitive' p e r s o n a l i t i e s . Scope condition 1: Two or more valued resources are d i s t r i b u t e d across  more than three members of a group. Each of the members has more of  one resource, X, than some other resource, Y, or v i c e versa-^and f o r  each member there e x i s t at l e a s t two other members who have complemen-tary value positions on X and Y. 'Valued resources' means that each member values each resource to some 5 extent. Scope condition 2: Resources are valued according to a marginal u t i l i t y  function. This w i l l generally imply that the resources are of a d i v i s i -b l e , material nature, and can be accumulated or stored by i n d i v i d u a l s , 6 but this is. not necessarily the case. Discussion: The p r i n c i p l e of diminishing marginal u t i l i t y i s entered as a scope condition, so that we can ensure the actors are motivated to exchange 46 some resource of which they have a l o t , f o r one of which they have l i t t l e or none. As noted i n Chapter 1, such a function makes i t possible for individuals- to gain subjective value v i a exchange, even though there may be no increase i n t o t a l resources d i s t r i b u t e d across a group. A diminishing marginal u t i l i t y function implies that successive equal absolute amounts of X have decreasing value to a given P. For pre-sent purposes, subjective p r o f i t i s treated as equivalent to net gain i n 7 marginal u t i l i t y . This implies that members perceive that exchanges of unequal absolute amounts of X and Y between P and 0 can s t i l l y i e l d equal 8 subjective p r o f i t (be f a i r ) . In the same way, two d i f f e r e n t exchanges (between P and 0^  and P and C^), invo l v i n g the same amount of X for d i f -ferent amounts of Y, can both be perceived as ' f a i r ' . Scope condition 3: Individuals know, or can reasonably i n f e r , the  value function describing the way i n which a l l others i n the group  value d i f f e r e n t resources, but know only some of the members' current  value positions i n given resources. A consequence of Scope condition 3 i s that i f the value functions of s e l f and other are known, members can in t e r p r e t what constitutes a f a i r exchange, assessed r e l a t i v e to the resource positions of these exchange partners. I t should be noted, however, that for exchanges to occur, whatever t h e i r f a i r n e s s , P and 0 must only be aware of some complementarity of resources... i t i s not necessary f o r eit h e r to know the other's value function or resource base. Discussion: Scope condition 3 l i m i t s the theory to cases where some com-parison of subjective p r o f i t i s possible. While the scope condition makes / 47 an existence statement about knowledge of others' value functions, i t i s admitted that the determination of such value functions i s often proble-matic, and the question i s worthy of study i n ' i t s own r i g h t . However, we assume that i t i s quite possible for members to obtain estimates of others' value functions, using such means as: — experience with s i m i l a r others i n s i m i l a r s i t u a t i o n s , where value functions are i n f e r r e d from behaviour. — i f a l l members have the same resons for valuing resources (e.g., X and Y are both needed to perform a task), P may assume that others w i l l have a c e r t a i n value function. — P learns to associate a given context and structure of rewards with a p a r t i c u l a r value function (e.g., elections t y p i c a l l y involve some sort of threshold function). On the basis of these considerations, we could a n t i c i p a t e that members with highly i d i o s y n c r a t i c value functions would be more dis r u p t i v e to smooth exchange r e l a t i o n s , at l e a s t u n t i l other members had had time to learn 9 t h e i r value function. Scope condition 4: P thinks that he may have the opportunity i n the future to compare His p r o f i t s with a l l members of the group, i n c l u d -ing those for whom P does not currently know value positions on the  resources i n the group. Scope condition;'5: The t o t a l absolute amount of resources i n the 10 group i s constant. 48 Discussion: This scope condition, together with the s t i p u l a t i o n of a mar-gin a l u t i l i t y function, creates the 'antagonistic cooperation 1 t y p i c a l of simple exchange. By marginal u t i l i t y , P can gain subjective u t i l i t y even as he loses some absolute net increment of a resource ( i . e . , i f he has a l o t of X and a l i t t l e of Y); he can gain even more u t i l i t y i f he both achieves an optimum balance of resources, and adds to h i s t o t a l absolute resource base. To do the l a t t e r under a condition of constant t o t a l absolute r e -sources means that 0 must lose some absolute amount. As there i s no over-a l l increase i n t o t a l resources, maximization means that i t i s 'better' to get mo IB than others i n the group. We thus increase the l i k e l i h o o d that per-sons w i l l be motivated to get a 'status edge', and can therefore observe the t a c t i c s they employ to obtain i t . Motivation i s imposed by the s t r u c -11 ture of rewards, and does not have to be measured beforehand. It can be argued that few s o c i a l s i t u a t i o n s e n t a i l a constant amount of resources — people are c o n t i n u a l l y producing new goods, s k i l l s , r e c e iving returns on investments and regular income, etc., and a l l of these can then be employed i n simple exchange. However, f o r any given ' s l i c e of time', the t o t a l amount of resources i s u n l i k e l y to vary d r a s t i c a l l y . We assume that the early stages of an exchange i n t e r a c t i o n , i n which we are 12 interested, can probably be sa f e l y approximated as constant sum. Scope condition 6: While some members of the group can withold i n f o r - mation about t h e i r resource l e v e l s , they cannot or do not choose to  give out f a l s e information concerning the exact s i z e of t h e i r resource bases. 49 Discussion; While t h i s may seem an excessively r e s t r i c t i v e condition, i t i s included to l i m i t the theory, temporarily, to the polar case involving d i s -c r e t i o n over information. When we have seen what happens i n t h i s case, i t should not be d i f f i c u l t to extend the theory to handle manipulation of i n -formation ( s e l e c t i v e l y witholding, d i s t o r t i n g , or f a l s i f y i n g ) . Assumptions Assumption 1: Individuals i n the group are motivated to maximize the  t o t a l subjective value of resources i n t h e i r possession. Discussion: Because scope condition 2 imposes a p r i n c i p l e of diminishing marginal u t i l i t y , each i n d i v i d u a l w i l l assign l e s s subjective value to suc-cessive, equal amounts of a given resource. This means that P w i l l often obtain more subjective value f o r a combination of two valued resources, than i f he has only one resource (e.g., a milkshake and a hamburger may be more valued than two milkshakes). As Kuhn says, "maximum s a t i s f a c t i o n i s achieved i f we continue to take a d d i t i o n a l units of a good u n t i l i t s marginal value j u s t equals i t s marginal cost," where cost i s the value of what P gives up. "This set of r e l a t i o n s h i p s i s known as the p r i n c i p l e of equimarginality, any v i o l a t i o n of which w i l l s a t i s f y a smaller want at the opportunity cost 15 of leaving a larger want u n s a t i s f i e d , and thus prevent maximum s a t i s f a c t i o n . " Marginal u t i l i t y increases for a given P i f : 1) he obtains an optimal combination of d i f f e r e n t valued e f f e c t s , holding the s i z e of P's resource base constant (in the sense of moving to a d i f f e r e n t p o s i t i o n on a u t i l i t y curve i n an Edgeworth box); 50 2) P increases the t o t a l amount of resources i n h i s resource base ( l i k e moving from one to another higher u t i l i t y curve i n an Edgeworth box). In a simple exchange s i t u a t i o n , Assumption 1 implies that P w i l l want to obtain as much marginal increase i n a desired resource (up to the point of equimarginality), for as l i t t l e cost as possible. However i t also assumes that P w i l l enter exchanges so long as he makes a p o s i t i v e p r o f i t . Assumption 2: Holding p r o f i t to himself constant, the subjective  value of an exchange to P i s affected by the subjective pirfit he per- ceives h i s partner receives. P p r e f e r s : 1) an advantageous exchange i n which P's p r o f i t i s greater than O's p r o f i t to 2) a f a i r exchange to 3) an inequitable exchange i n which P's p r o f i t i s l e s s than O's p r o f i t . Discussion: Assumption 2 provides a means for incorporating" the idea that P's u t i l i t y i s affected by what he sees 0 to be getting. By adding the proviso 'holding p r o f i t to himself constant', ( i . e . , for two a l t e r n a t i v e exchanges xn which P_ gains the same value), i t i s admitted that comparative gain i s a subsidiary concern to P's own p r o f i t . Together with the implica-t i o n of Assumption 1, that P w i l l transact so long as he obtains a p o s i t i v e p r o f i t , we admit that P may enter into u n f a i r transactions. Assumption 3: Holding p r o f i t to s e l f constant, the p r o b a b i l i t y that P w i l l make a given i n i t i a t i o n of exchange i s d i r e c t l y r e l a t e d to how l i k e l y 14 P perceives i t to be that h i s i n i t i a t i o n w i l l be accepted. 51 Discussion: From Assumption 1-3, we can t e n t a t i v e l y propose that the expec-ted value to P of an exchange with a given 0 can be represented as a l e x i c o -graphic function, which gives p r o f i t to P pre-eminence over perceived p r o f i t to 0y A lexicographic function r e f l e c t s the i n d i v i d u a l handling a m u l t i -dimensional decision problem one dimension at a time, putting the most im-portant dimension at the top of a l i s t of factors he w i l l consider; only i f he i s i n d i f f e r e n t between two or more a l t e r n a t i v e s on the most impor-tant dimension, does he compare them on the next most important, and so on. Here, a lexicographic function i s proposed to capture the idea that P w i l l f i r s t consider a l t e r n a t i v e exchanges to f i n d the cl a s s or set which o f f e r s the greatest p r o f i t to P; i f there are more than one i n t h i s c l a s s , he then sel e c t s on the basis of r e l a t i v e p r o f i t . As an approximation, l e t us repre-sent the s u b j e c t i v e l y expected value of an a l t e r n a t i v e (SEV) as: SEV to P, .= lex. f (perceived prob. of acceptance,. • prof i t to P; d) (.Oj > k) where (X r e f e r s to a given exchange partner, k to the units involved i n the exchange, and d = +1 i f 0 gets le s s than P = 0 i f 0 gets the same as P = -1 i f 0 gets more than P. The value of d as +1, 0 and -1 i s a crude approximation, and would have to be r e f i n e d to handle cases i n which P discriminates between how much more 15 of le s s p r o f i t 0 receives. Assumption 4: If P has information about the resource positions of  other members, each P can adopt each other member's point of view to 52 estimate how much of one resource 0. could be induced to give P i n r e -. : j : — a  turn f o r a given amount of another resource. Discussion: Assumption 4 provides a basis f o r comparison of subjective values. This i s necessary f o r the i n t e r p r e t a t i o n of ' f a i r exchange'. Be-cause the scope conditions s t i p u l a t e : 1) that a l l members do i n fact value the resources i n the group, and 2) that the resources are valued according to a marginal u t i l i t y function, P can assume that each member w i l l assign the same value to a given t o t a l amount of i t . This means too that members w i l l assign the same worth to an increment of X i f they are at the same  resource p o s i t i o n . The estimation of the marginal worth of an increment i s calculated against what Other already has; thus the proviso that P must 16 have some information about O's resource p o s i t i o n . While P can often assume that others have s i m i l a r value functions to h i s , or to i n f e r O's value function from the i n t e r a c t i o n context, he cannot make s i m i l a r assumptions about the current l e v e l of O's resources, as these are much more subject to v a r i a t i o n . By taking O's point of view, and u t i l i z i n g information about the current l e v e l of O's resources, P can locate exchange partners with whom he can obtain maximum gain at minimum cost to himself, i . e . , he can f i n d Others to whom a given increment of P's resource i s valued most, and to whom a given increment of O's resource i s 17 valued l e a s t (by 0). The converse of Assumption 4 i s that i f P has no information about O's resources, he cannot estimate O's subjective p r o f i t i n a given transac-t i o n , and cannot judge the l i m i t s of p r o f i t overlap between h i s own and Other's preferences. 53 An implication of Assumption 4 i s that each P knows that an ex-change transaction which 0 perceives to be f a i r i s more l i k e l y to be en-tered into by 0, than i s a transaction i n which P p r o f i t s more than 0. This serves to l i m i t the range of transactions.that P w i l l bother to propose to 0. In addition, i t informs P of what 0 would have to beli e v e about an exchange before 0 considered i t a t t r a c t i v e . A further extension of Assumption 4 i s that, so long as 0 obtains a p o s i t i v e p r o f i t i n a given i n i t i a t i o n of exchange from P, P perceives that there i s some p o s i t i v e probablity that the i n i t i a t i o n w i l l be accepted. This takes account of the f a c t ; t h a t , while 0 i s les s l i k e l y to accept i n i t i a - ' tions of exchange from P, the lower his p r o f i t r e l a t i v e to P, he w i l l never-theless have some l i k e l i h o o d of entering the transaction, so long as he ob-tains a p r o f i t . The importance of other p o t e n t i a l exchange partners i s ob-vious. Assumption 5: The greater the ambiguity of information that P has about O.'s resources, the les s confidence P has i n h i s estimate of the 3 8  expected u t i l i t y to P of an exchange with 0.. Discussion: It ispropojsed'jthat the function f o r P's expected u t i l i t y for a given exchange with 0 be weighted by a fac t o r , C, representing the confidence P f e e l s about his estimate of expected value (as i n Assumption 3), based on the information he has about 0 . As ambiguity of information increases (depending on the q u a l i t y , source, r e l i a b i l i t y , e t c . ) , P's confidence i n his assessment or estimate of the value of a given exchange decreases, and has the e f f e c t of reducing the estimated expected value. I f we represent 54 t o t a l c e r t a i n t y as C = 1.0, the estimate of expected value w i l l be l e f t b a s i -c a l l y undistorted; as ambiguity increases, the f r a c t i o n C decreases, and P gives le s s credence to h i s estimate of value. (This factor i s of i n t e r e s t only i f c e r t a i n t y varies across a l t e r n a t i v e s ; otherwise, a uniform reduc-t i o n of estimated u t i l i t y would not a l t e r P's ordering of a l t e r n a t i v e s , and would not a f f e c t our predictions of which a l t e r n a t i v e P would choose.) On these considerations, we would re v i s e the function giving P's ex-pected value for an a l t e r n a t i v e to the form: SEV to P^Q k ) = l e x « f (perceived prob. of acceptance ••; p r o f i t to P;d)) where C takes on values from 0,-to 1.0, and d i s defined as for Assumption 3. It would be more accurate to have: SEV to P . Q ,k.,= lex. f (C^3« perceived prob. .: prof i t to P; C 2d) j , of acceptance In t h i s way, could be equal to 1.0 i n cases where P i s the r e c i p i e n t of an i n i t i a t i o n of exchange ( i . e . , he i s c e r t a i n that, i f he agrees, the trans-action w i l l i n fact take place, and he w i l l make the stated p r o f i t ) . We do not wish to dwell on the construction of an appropriate formal model i n t h i s d i s s e r t a t i o n . A lexicographic representation i s mentioned as a possible means of l i n k i n g the process under study to the more rigorous body of theory 18 i n d ecision making studies. It also provides a shorthand notation for the decision s i t u a t i o n facing the actors. Let us consider b r i e f l y j u s t what 'ambiguity of information' means i n an exchange context. The information at issue i s that concerning O's r e -source l e v e l s , because t h i s , together with P's a b i l i t y to assume O's point of view, i s what allows P to estimate how much p r o f i t 0^  makes on a given transaction; t h i s i n turn l e t s P estimate: 55 1) how l i k e l y 0 i s to accept a given i n i t i a t i o n by P (pr o b a b i l i t y ) 2) how much 0 p r o f i t s r e l a t i v e to P,(d), and thus, the fai r n e s s of an ex-change. With respect to 1): p r i o r to a transaction, P's estimate of whether or not his i n i t i a t i o n w i l l be accepted i s ju s t an estimate — P cannot be c e r t a i n of being accepted, and he does not know the exact p r o b a b i l i t y of acceptance and r e j e c t i o n . He i s faced, i n essence, with an uncertain s i t u a t i o n (pro-19 b a b i l i t y d i s t r i b u t i o n f o r events unknown). As El l s b e r g argues, even i n uncertain s i t u a t i o n s , P may be able to assign subjective p r o b a b i l i t i e s to a l t e r n a t i v e events. The dimension of confidence of estimate i s a d d i t i o n a l  to the subjective estimate of p r o b a b i l i t y — a f t e r making use of a l l the information a v a i l a b l e to him to ca l c u l a t e subjective estimates of pr o b a b i l -i t y , P may s t i l l f e e l that he can put more or less f a i t h i n these estimates, on account of the q u a l i t y , source, r e l i a b i l i t y , and quantity of the i n f o r -mation on which i t was based. The studies by E l l s b e r g and others mentioned 20 i n Chapter 1 compared preferences for r i s k y and uncertain bets, and found that many subjects chose the former, even when the two bets had equal ex-21 pected value. Thus, i f (value x p r o b a b i l i t y = expected value) does not vary, some a d d i t i o n a l factor must be operating to create the preference for the r i s k y bet — t h i s E l l s b e r g a t t r i b u t e d to the lower confidence P has in an estimate of p r o b a b i l i t y based on ignorance of the p r o b a b i l i t y d i s t r i b u -22 t i o n . We argued i n the introduction that these findings could be ex-tended to a s o c i a l i n t e r a c t i o n context. There are many s o c i a l s i t u a t i o n s 56 p a r a l l e l to the uncertain bet — faced with a human opponent, with whom P has had l i t t l e or no p r i o r i n t e r a c t i o n , P must a r r i v e at subjective estimates of Oj's p r o b a b i l i t y of accepting P's i n i t i a t i o n . P i s l i k e l y to f e e l d i f -ferent degrees of confidence i n h i s estimates, depending on the ambiguity of information a v a i l a b l e to him. The question of what kinds of information are considered more or less ambiguous i s both a t h e o r e t i c a l and an empirical issue. We do not have a well established set of findings about what kinds 23 of cues P w i l l t r u s t most i n s o c i a l s i t u a t i o n s , although Jones et a l . have shown that information c o n t r o l l e d by 0 (concerning 0) i s suspect i f 0 could gain by manipulating such information. Studies of the inference of 24 Other's motives may eventually add to our understanding of the r e l a t i o n -ship between type of information and trust of such information. (Our argument that the structure of rewards a f f e c t s the motives of actors w i l l be extended to claim that the more competitive the structure of rewards, the more actors w i l l be suspicious of information that comes from 0.) In addition to factors that make P have l e s s confidence i n i n f o r -mation, we also know that people process only a li m i t e d amount of informa-25 t i o n , and wilj. not seek more so long as they have enough to proceed. I f t h i s l a t t e r process applies i n s o c i a l i n t e r a c t i o n , we can expect a further bias towards partners about whom the actor has information at hand (as opposed to those about whom he must f i r s t acquire information). The i m p l i -cation i s that P w i l l prefer to make use of unambiguous information i f i t i s a v a i l a b l e ; he w i l l , however, use more ambiguous information i f that i s a l l that i s a v a i l a b l e . 57 With respect to 2): Without information about 0 j ' s resources, P cannot r e l i a b l y estimate 0^'s p r o f i t i n a given exchange by means of taking O's point of view, as he does not know the base against which X i s evalu-ated. As with an uncertain bet, P may average the p o s s i b i l i t i e s — he can assume 0 gains more, l e s s , or the same as P, the 'average' guess being that 0 gains the same. Such an estimate would have a lower confidence weighting than i f P knew unambiguously what O's p r o f i t was. A lower weight-ing would also be indicated i f P i s conservative, and makes the 'worst possible' assumption. Assumption 6: P believes 0 w i l l u t i l i z e P's i n i t i a t i o n s of exchange as cues to O's value p o s i t i o n with respect to resources, when 0 does 26 not have v e r i f i a b l e , unambiguous information about P's resources. 27 Discussion: Kuhn argues that people measure the r e l a t i v e goodness or badness of a p a r t i c u l a r transaction by a) the consensus of what goes on around them, and b) the proposals of the other person. Lacking a consensus i n the early stages of i n t e r a c t i o n , or lacking information about the con-sensus, the o f f e r s of exchange from Other may be the only i n d i c a t i o n P has 28 of what 0 considers a reasonable p r i c e . L iebert, et a l . , provided e v i -dence that "the bids of Other are used as cues as to the reasonableness of one's own a s p i r a t i o n s " ; and indicated that an i n d i v i d u a l who was ignorant of h i s opponent's p r o f i t , was more influenced i n h i s own goal s e t t i n g by the informed bargainer's o f f e r s . Assumption 6 provides for the p o s s i b i l i t y that P w i l l t r y to i n -fluence O's perception of what meets the c r i t e r i o n of a f a i r exchange, i f 58 information is asymmetric. If P has information about 0, he can discern the limits of his and O's profit overlap; so long as P keeps his offers in this region, P can ask for more than a fai r exchange would allow, and main-tain a positive probability that 0 w i l l perceive i t to be f a i r . Assumption 7: Initiations of exchange are subject to principles of  reinforcement: 1) acceptance of an i n i t i a t i o n acts as a positive re-inforcement; 2) positive reinforcement leads to an increase (repeti-tion) of initiations of the same kind, and negative reinforcement w i l l lead to a decrease (failure to repeat, change of i n i t i a t i o n ) . Hypotheses The hypotheses w i l l be presented according to the argument made in the f i r s t chapter. There i t was argued that in exchange relationships, P and 0 make comparisons of subjective profits in exchange through a pro-cess of taking the other's point of view, and that this process depended on knowing the resource base of Other. Exchanges under f u l l information about resources were argued to approach fairness, or equality of profits assessed in relation to resource bases. P could get the best profit to himself by finding someone who valued highly what P could give him (re-source X), and who placed relatively low value on what P wanted (Y). The f i r s t hypothesis predicts how P w i l l act i f potential partners differ in their need for X and Y. The next two hypotheses make predictions about the possibility that when information is asymmetric, P may perceive an advan-tage in being able to withold information about his resources that 0 needs to assess the fairness of a proposed exchange transaction. P is then ex-5 9 pected to make initiations of exchange that would give him more profit than 0 (i.e., that are unfair to 0). It was argued that, in the presence of alternative P's about whom 0 had information, 0 could express his prefer-ence for unambiguous information about his partner by avoiding transactions with a partner whose resource base was not known. The fourth and f i f t h hy-potheses make predictions that 0 w i l l show his preference by a) i n i t i a t i n g to persons about whom he has information; and b) accepting or agreeing to i n i t i a t i o n from such persons. The last hypothesis claims that the net result of the process outlined above is that Persons who have tried to gain an advantageous exchange through information control w i l l f a i l to do so, and w i l l decide in time to make available information about their re-source bases. The situation in which the hypotheses apply i s one where some people have already revealed their resource base and value positions to other members of the group, and some other members have, by choice or c i r -cumstance, discretion over information about their value positions. We con-sider a situation where there are just two different commodities X and Y, valued in a known and specifiable way, and distributed non-uniformly across the members of the group. Individuals are seen to have an equal number of possible alternative exchange partners, so that no one possesses a power advantage by virtue of a greater number of alternatives. Hypothesis 1 Among those members about whose resource bases and value positions P  has unambiguous information, P is most l i k e l y to i n i t i a t e to a person,  0., who P perceives to have the greatest amount of resource Y, and the  smallest amount of resource X, relative to Y. 60 Hypothesis 1 i s based on the following argument. In Assumption 1 i t was argued that P w i l l t r y to maximize his t o t a l marginal u t i l i t y . In s i t u a t i o n s s p e c i f i e d by the scope conditions, when the actor has a lot. of one commodity X, and l i t t l e of another commodity Y, both of which he values, he can trade some of X for some of Y, to increase the t o t a l value to him of the resources he possesses. By d e f i n i t i o n , the marginal u t i l i t y of an increment w i l l be smaller for some 0 who has a l o t of Y-fy the costs to such an 0 w i l l be r e l a t i v e l y low. By Assumption 4, P can take the point of view of Others whose resource bases are known, and locate an 0 whose costs A Y are low, and whose marginal gain from the ./\ X which P i s o f f e r i n g are large. Such a person w i l l not only be w i l l i n g to give up more Y, but w i l l be anxious to obtain X, as i t increases h i s value. In conjunc-t i o n with Assumption 3, s t a t i n g that P w i l l be more l i k e l y to make an i n i -t i a t i o n i f i t has a good chance of being accepted, t h i s implies that P's subjectively expected p r o f i t w i l l be greater for an i n i t i a t i o n to an 0 who has a l o t of Y r e l a t i v e to X. Therefore, we predict that members are more l i k e l y to make i n i t i a t i o n s to these Others. The proviso "Among those about whose resource bases and value positions P has unambiguous information....", i s made to r e f l e c t the argument that without such information, P cannot en-gage i n the.role-taking process described by Assumption 4, to assess what he must give 0 to get an increment Y. 29 Discussion: W. Foddy provided evidence that P prefers to i n i t i a t e to others with a large resource base rather than to those with a small resource base of a given commodity, because the subjective costs for 'wealthy' Others 61 for giving up resource increments were smaller. Hypothesis 1 i s b a s i c a l l y a restatement of this p r i n c i p l e , except that: 1) i n the present context, we s h a l l go on to state a p r i o r preference for exchange partners about whom one has information; and 2) Hypothesis 1 generalizes W. Foddy's claim that P prefers Others with large amounts of resource Y, to include the case i n which a l l members have the same t o t a l wealth..., but d i f f e r e n t r a t i o s ^ of one resource to another. In such a case, P w i l l prefer to i n i t i a t e to Others with a r e l a t i v e l y greater imbalance of Y over X. Conversely, P w i l l avoid dealing with someone with a more even balance of resources than P, because t h i s O's prices w i l l be too high. A further point to note i s that i f we f i n d empirical support for Hypothesis 1, we w i l l have a basis for p r e d i c t i n g the rest of the process of concealing rewards and costs to gain advantage. That i s , i t w i l l give us confidence i n our claim that P does evaluate the r e l a t i v e p r o f i t s i n an exchange i n reference to how much he and 0 already have of the resources being exchanged. Hypothesis 2: If 0 has only ambiguous information about P's resource  base, and P has unambiguous information about O's resource base, P i s  more l i k e l y to make an i n i t i a t i o n of exchange that i s advantageous to  P, than when 0 does have information about P's resource base. Hypothesis 3: In the early stages of i n t e r a c t i o n , persons who have  the opportunity to conceal t h e i r resource bases and value positions  w i l l perceive such concealment to be an advantage. Hypotheses 2 and 3 are based on the following argument: We have assumed (Assumption 1) that members of the group want to obtain as much marginal p r o f i t as i s poss i b l e , and i n addition, to gain more 30 p r o f i t i n an exchange than the partner gains (Assumption 2). However, by 62 Assumption 4, when P has information about O's resources, P i s able to adopt the given Other's point of view to guage how much 0 i s w i l l i n g and able to give P. Taking O's point of view makes P r e a l i z e that 0 acts ac-cording to Assumptions 1 and 2, ju s t as P does. The net r e s u l t under con-d i t i o n s of f u l l information would tend to equality of subjective p r o f i t s i n exchange, as t h i s w i l l tend to be each actor's estimate of what 0 i s w i l l i n g and able to give up. By Assumption 3, P i s u n l i k e l y to i n i t i a t e an exchange that i s u n l i k e l y to be completed. Under f u l l symmetric i n -formation, exchanges u n f a i r to eit h e r party are u n l i k e l y to be completed. However, i f 0 has only ambiguous or no information about P's resource base, then by Assumption 6, Other i s l e f t i n the p o s i t i o n of i n f e r r i n g the p r o f i t P would make by the kinds of o f f e r s P makes. If 0 receives an o f f e r from P that, for example, asks more than i t o f f e r s , 0 cannot know whether P i s try i n g for an advantage, or whether P has a resource base that ' j u s t i f i e s ' such an o f f e r . We conclude that i f P does; i n f a c t , make an i n i t i a t i o n of exchange that i s advantageous to himself, 0 i s less l i k e l y to r e j e c t i t than i f 0 could see, on the basis of information about P's resource base, that the o f f e r was u n f a i r . This being the case, P i s more l i k e l y to make advantageous i n i t i a t i o n s when 0 has only ambiguous information about P's resource base and value p o s i t i o n s , as there i s a greater p r o b a b i l i t y of acceptance. By Assumption 3, the p r o b a b i l i t y of a given i n i t i a t i o n being made increases as the p r o b a b i l i t y of acceptance increases. Discussion: Hypothesis 2 proposes that t a c t i c s of concealment and decep-t i o n are used when others are uncertain about the true conditions of supply 63 and demand, and when i t i s c o s t l y or impossible to test the cues that P gives out. If one v i s u a l i z e s P's i n i t i a t i o n s of exchange being motivated by a desire to increase h i s gain, but subject to forces that prevent him from gaining as much as he wants, we may think of P's a b i l i t y to withold information about his resources as removing one force that constrains him from attempting an advantage. The p r e d i c t i o n i n Hypothesis 2 implies that a greater proportion of members whose resources are concealed w i l l attempt advantageous exchanges than w i l l those whose resource bases are known. It can be argued, perhaps, that a person would have enough foresight to r e a l i z e that an o f f e r from a person with concealed resources would be rejected, so that P w i l l not attempt an advantageous exchange. Aside from the counter-argument that the lower p r o b a b i l i t y of acceptance may be balanced by the increase i n p r o f i t possible, i t can be suggested that: — P i s not c e r t a i n that 0 w i l l receive any other i n i t i a t i o n s of exchange, and he may accept P's i n i t i a t i o n i f i t i s the only one 0 receives. — P also has a l t e r n a t i v e exchange partners, and i f he f a i l s with an ad-vantageous i n i t i a t i o n to G\ , he i s not excluded from future exchange. — P has l i m i t e d cognitive capacity, and w i l l focus more on h i s own in t e n -tions and acts, than on a n t i c i p a t i n g a l l the possible i n i t i a t i o n s to a l l the d i f f e r e n t a l t e r n a t i v e partners. Although the p r e d i c t i o n made i n Hypothesis 3 follows from the same argument as for Hypothesis 2, i t d i f f e r s i n making a p r e d i c t i o n about percep-tions , rather than about types of i n i t i a t i o n s . Members must f i r s t perceive an advantage i n concealment before they use i t , aid i t seems reasonable to argue that something which leads to an increased p r o b a b i l i t y of obtaining a 64 desired outcome i s an advantage. However, as we noted i n the previous d i s -cussion, a person w i l l not nec e s s a r i l y attempt an advantageous exchange when he sees that 0 i s unable to assess the fairness of such an exchange. P might f e e l that t a c t i c s of advantage work only when used sparingly, or P may see more value i n es t a b l i s h i n g a firm exchange r e l a t i o n s h i p i n the early stages of i n t e r a c t i o n . I f a member does not choose to make use of the p o t e n t i a l advantages -.of concealment, we are interested i n es t a b l i s h i n g whether i t i s because he does not perceive the advantage, or whether he recognizes the p o t e n t i a l but chooses not to u t i l i z e i t . We can separate the perception of an advantage and attempts to obtain one, by d i r e c t l y ask-ing group members for t h e i r opinions concerning the concealment of i n f o r -mation, and by assessing t h e i r preferences. Presumably, those who eit h e r choose to conceal, or continue to conceal t h e i r resource bases i f given the opportunity, are i n d i c a t i n g that they perceive i t to be i n t h e i r i n t e r e s t . There i s the p o s s i b i l i t y , too, that the perceived advantage of concealment does not l i e s o l e l y i n the increased chances for putting over an unf a i r o f f e r . Hypothesis 4: Members of a group are more l i k e l y to d i r e c t t h e i r i n i - t i a t i o n s of exchange to persons whose resource bases and value p o s i -tions are known (through unambiguous information) . ^  The argument leading to Hypothesis 4 i s as follows: We made the assumption (Assumption 4) that P w i l l be able to adopt Others' (0^ refelrr^'tDOasjSthfeereGipient'V'hffresf):)/points of view to estimate how much a given r e c i p i e n t w i l l give up of one resource Y, i n return f o r a given amount of X, from P. When 0 i s the i n i t i a t o r , then from O's point of view, he can make this kind of assessment only i f he has information about 65 the resource positions of d i f f e r e n t P's. Lacking such information, 0 can only guess as to what sort of p r o f i t P i s obtaining, and what he would f i n d i acceptable. When 0 compares two a l t e r n a t i v e exchange partners, P^, about whom 0 has unambiguous information, and P^, about whom he has none, 0 may enter-t a i n the hypothesis that P2 i s as good a partner, a better partner, or a worse partner than P^, i n that P^ may have more, l e s s , or the same t o t a l amount of resources as P^. However, by Assumption 5, i f 0 has no informa-t i o n about P 2 ' s resource base, he w i l l have less confidence i n whatever estimate of the expected value of an exchange with P2 he arri v e s at. This has the e f f e c t of depressing O's estimates of the expected value of exchanges with partners whose resource bases are not r e l i a b l y known to him. This does not exclude the p o s s i b i l i t y that the ambiguity-weighted estimate of SEV with P2 can exceed the SEV of an exchange with P^, i f the l a t t e r i s a very undesirable partner. However, i t seems reasonable to predict that 0 w i l l frame his expectations about these concealed others r e l a t i v e to what he knows about the p o t e n t i a l partners about whom he has r e l i a b l e informa-32 t i o n . (Information under a person's c o n t r o l , including the information contained i n the types of i n i t i a t i o n s that person makes, i s considered more ambiguous than d i r e c t , v e r i f i a b l e information.) We therefore p r e d i c t , that other things being equal, the weighted expected value ( p r o f i t ) i n an exchange with a partner about whom 0 has unambiguous information w i l l be greater than the ambiguity-weighted expected p r o f i t i n an exchange with "a partner who witholds information about his resources (as s p e c i f i e d i n scope 66 conditions 3 and 6). Together with Assumption 1 that members of the group seek to maximize subjective value of resources, the foregoing argument leads to the p r e d i c t i o n i n Hypothesis 4. Hypothesis 5: I f a member has a choice of i n i t i a t i o n s from and P^, and the i n i t i a t i o n s are i d e n t i c a l with respect to absolute amount of  resources offered and requested, and i f the member has unambiguous  information about the value p o s i t i o n of P , and ambiguous or no i n f o r -mation about the value p o s i t i o n of P 2 , then the member w i l l be. more  l i k e l y to accept the i n i t i a t i o n from P^. Hypothesis 5 i s based on the following argument: (P^ and P2 are i n i t i a t o r s ; 0 i s r e c i p i e n t ) If the group member (refer to him simply as 0) receives i n i t i a -tions of exchange that are i d e n t i c a l except for information a v a i l a b l e about the i n i t i a t o r (P^ or P^), then 0 cannot decide between the o f f e r s on the basis of p r o f i t to himself. This p r o f i t i s f i x e d once the o f f e r i s made, and i f he accepts an i n i t i a t i o n , the transaction goes through with probabi-l i t y 1.0. By the choice function given with Assumption 5, i f O's own pro-f i t i s the same for both o f f e r s , he w i l l then attend to how h i s p r o f i t compares to what the i n i t i a t o r receives i n the transaction. Because the only information 0 has about P 2 ' s value p o s i t i o n must be i n f e r r e d from P2's i n i t i a t i o n , 0 w i l l have less confidence i n h i s assessment of the comparison of his p r o f i t with that of P 2 , on the grounds that information under P2's control i s more ambiguous. The ambiguity of information about P2's r e -source base depresses the expected p r o f i t of an exchange with P 2 J i n res-pect to the second dimension of the lexicographic choice function; and when two a l t e r n a t i v e s are i n the same payoff class with respect to O's own p r o f i t , he w i l l s e l e c t the a l t e r n a t i v e with the greatest expected value on the second dimension — thus the p r e d i c t i o n that 0 w i l l s e l e c t the o f f e r from P 1. 67 Discussion: If a person were i n d i f f e r e n t between two such o f f e r s , there might w e l l be a sound basis for an actor with concealed resource l e v e l s to gain an advantage, e s p e c i a l l y i f h i s costs happened to be lower than those 33 of P^. The fact that P2* s concealment destroys O's i n d i f f e r e n c e between otherwise i d e n t i c a l i n i t i a t i o n s l i m i t s the advantage i n concealment, jThus, hypothesis 5 i s c r u c i a l to the theory,as i t holds constant a l l factors (size of t y p i c a l i n i t i a t i o n , number of o f f e r s to choose from), except that of information, and predicts that people w i l l opt for the a l t e r n a t i v e with better and more information. It should be noted that Hypothesis 5 does not deny the p o s s i b i l i t y that i f 0 gets a generous o f f e r from a person with concealed resources, such that O's net p r o f i t i s larger than for any other a l t e r n a t i v e , 0 w i l l accept i t , despite i t s unknown f a i r n e s s , because i t compares w e l l with at l e a s t one other a l t e r n a t i v e . However, because of the preference by a l l actors to seek exchanges that p r o f i t them most, such generous o f f e r s are u n l i k e l y . In addi-t i o n , i f an i n i t i a t i o n from a source whose resource base i s not known to 0 i s the only i n i t i a t i o n 0 ; Preceives, he i s l i k e l y to accept i t , so long as he makes a p o s i t i v e p r o f i t . Hypothesis 6: P's perception of the advantage of witholding informa-ti o n about his resource base and value p o s i t i o n w i l l decrease over  repeated exchange i n t e r a c t i o n , and P w i l l be more l i k e l y to choose to  reveal unambiguous information about h i s value positions :-and p r o f i t s ,  than during the i n i t i a l stages of i n t e r a c t i o n . The argument leading to Hypothesis 6 i s given below: In Assumption 7, i t was stated that group members' i n i t i a t i o n s are--isubject to p r i n c i p l e s of reinforcement, such that an acceptance of an i n i t i a t i o n acts as a p o s i t i v e r e i n f o r c e r , making a given P more l i k e l y to 68 to repeat an i n i t i a t i o n of a s i m i l a r sort to a s i m i l a r person; and such that r e j e c t i o n of an i n i t i a t i o n would lead the i n i t i a t o r , P, to change his o f f e r s , by going to d i f f e r e n t targets, and/or by changing the type of o f f e r . In ad d i t i o n , the receipt of i n i t i a t i o n s should act as a p o s i t i v e r e i n f o r c e r . In Hypothesis 3, i t was predicted that the persons most l i k e l y to receive i n i t i a t i o n s were those whose resource bases and value positions were unam-biguously known. In Hypothesis 5, i t was predicted that these same persons were more l i k e l y to have t h e i r i n i t i a t i o n s of exchange accepted, i f they were considered beside an i d e n t i c a l o f f e r from a member with concealed r e -sources. In addition, on the basis of Hypothesis 1, members with concealed resources are more l i k e l y to make o f f e r s unattractive to the r e c i p i e n t s , r e -l a t i v e to the o f f e r s made by a member with a s i m i l a r but unconcealed r e -source base. Thus, people who reveal honest information about t h e i r re-sources are more l i k e l y to complete transactions, i . e . , they w i l l 'suc-ceed' more. This sequence of predicted events, coupled with Assumption 7, implies that a Person who has concealed h i s resources w i l l notice the pat-tern of acceptances ( i f i t occurs), and by d i s c r i m i n a t i o n , then generaliza-t i o n , he w i l l probably conclude that h i s f a i l u r e to provide unambiguous information i s behind his low rate of success. P's choice of revealing information i s thus more l i k e l y a f t e r i n t e r a c t i o n has proceeded for some time than during the early stages. The general idea i s that people w i l l not use information control as a t a c t i c i f i t does not work, but that since a person w i l l i n i t i a l l y b elieve concealment to be an advantage, he must learn through experience. 69 In the early stages of i n t e r a c t i o n , people do not have v i a b l e , established a l t e r n a t i v e partners, and P i s l i k e l y to think there i s a chance of having an advantageous i n i t i a t i o n of exchange accepted. As i n t e r a c t i o n proceeds, information about the prices i n transactions gets d i s t r i b u t e d (even without d i r e c t communication), and even i f 0 does not know the l e v e l of P's resources, he can hold P to the l e v e l of established a l t e r n a t i v e transactions. Thus the person who i n i t i a l l y concealed h i s p r o f i t l e v e l i s forced to pay the going p r i c e , or to give information about h i s value p o s i -34 t i o n s , which has the e f f e c t of f o r c i n g h i s p r i c e down. No e x p l i c i t hypotheses are made here concerning the p r o b a b i l i t y that P w i l l make use of concealment to t r y advantageous exchange, although i t was implied i n Chapter 1 that P would be l e a s t l i k e l y to perceive and use information control as a t a c t i c i f he was i n a large, well-connected group of long standing; and most l i k e l y to try to gain an advantage i f he was exchanging with someone who had no r e a d i l y accessible a l t e r n a t i v e s , and where P could manipulate information without being suspected or checked. (So long as P can control O's a c c e s s i b i l i t y to a l t e r n a t i v e s , advantageous exchange can occur i n groups of long standing — t h i s moves us into the realm of r e a l power advantage, beyond the scope of the present theory.) The invocation of reinforcement p r i n c i p l e s might imply that once P has learned his lesson, he w i l l ne'er more wander from the paths of f a i r n e s s . In the early stages of i n t e r a c t i o n (with new persons, or new resources), the conditions that enhance advantageous exchange attempts (low connected-ness, low information, lack of established partners) are present to some 70 degree, and we therefore expect advantageous attempts, e s p e c i a l l y when we acknowledge that P cannot foresee the en t i r e course of i n t e r a c t i o n . While the theory could be expanded to follow the i n t e r a c t i o n process through to the point where s o c i a l prices reach a consensus l e v e l , such an extension would take us beyond the scope of the research paradigm, and so we leave t h i s task for future research. The next chapter presents the research design used to test the s i x hypotheses presented above. 71 FOOTNOTES FOR CHAPTER 2 The s p e c i f i c a t i o n of what constitutes a resource i s not a s e t t l e d issue i n exchange theory, and for present purposes, material resources are used for s i m p l i c i t y . The problem of defining resources i s compli-cated by the fact that d i f f e r e n t resources appear to have d i f f e r e n t properties. For example, some resources can be transferred only (e.g., money), while others can be kept and transferred at the same time (e.g., e x p e r t i s e ) . See S. Rosen, 'The comparative roles of informational and material commodities i n interpersonal transactions', Journal of Experi-mental S o c i a l Psychology, 2_, 1966, pp. 211-226., for a discussion of an i n v e s t i g a t i o n to test the thesis that the owner of valuable i n f o r -mation would engage i n d i f f e r e n t p r i c i n g behaviour than the owner of valuable material commodities. For purposes of t h i s study, the value function i s a marginal u t i l i t y function that states that as P gainsSmore of a given resource X, suc-cessive standard increments of X are less and less valuable. Other functions might also describe the manner i n which P values X: for example, a l i n e a r function would mean that each successive increment of X gave the same added subjective value, regardless of what P had; a threshold function would mean that P did not receive any subjective value for X u n t i l he had an amount k of X, a f t e r which further amounts of X were not any more valuable (e.g., i f k out of N votes were needed to win an election.) " The marginal reward/cost of an increment of Y r e f e r s to the subjective reward/cost to P of that increment, and i s assessed r e l a t i v e to how much Y P already has. See A. Kuhn, The Study of Society. London: Tavistock P u b l i c a t i o n s , 1966, pp. 285-86. See: W.H. Foddy; 'The formation of cliques i n c o l l e c t i v i t i e s as a consequence of i n i t i a l d i s t r i b u t i o n of dimensions of wealth', Unpublished Ph.D. D i s s e r t a t i o n , University of B r i t i s h Columbia, 1971; and R.K. L e i k , R.M. Emerson, and R.L. Burgess, 'The emergence of s t r a t i f i c a t i o n i n exchange networks: An experimental demonstration'; Paper presented at the West Coast Conference for Small Group Research, San Diego, 1968. I n s t i t u t e for S o c i o l o g i c a l Research, University of Washington, Seattle (Mimeograph). Two d i f f e r e n t resources are a l l that i s necessary to represent the abstract r e l a t i o n s h i p ; more than two complementary resources would not change the nature of the p r i n c i p l e s of the theory, but introduce com-p l i c a t i o n s due to the fact that P might need d i f f e r e n t exchange networks (groups of people) to meet his needs for d i f f e r e n t kinds of resources. 72 6. Blau gives an example of a non-material commodity which may be valued according to a decreasing marginal u t i l i t y p r i n c i p l e i n h i s discussion of the exchange of advice for deference among o f f i c e workers. In addi-t i o n , o b l i g a t i o n , though non-material, i s often thought to be cumulative. See P.M. Blau, Exchange and Power i n S o c i a l L i f e , New York: John Wiley and Sons, 1964. 7. If P has no X, but s t i l l says he does not value an increment of X at a l l , we may i n f e r e i t h e r that i t would take a great deal of X to pro-duce any . u t i l i t y f o r P (p o s i t i v e goods) , i . e . , h i s j u s t noticeable difference i s very large; or that P has a negative u t i l i t y for X (neg-a t i v e goods), and w i l l pay to be r i d of i t . The exchange of goods and behaviour to avoid undesired behaviour from Other i s s i m i l a r to coer-cion, but can be handled i n an exchange framework. See Kuhn, op. c i t . , pp. 361-370. 8. This condition eliminates s i t u a t i o n s where i n d i v i d u a l s f e e l constrained to give a fi x e d number of units of X for a fi x e d number.of units of Y, regardless of the s i t u a t i o n of the partners to the exchange. Commer-c i a l goods are usually sold f o r the same amount of money, regardless.-'--of the buyer's wealth, and are not open to manipulation through i n f o r -mational t a c t i c s . (Exceptions do e x i s t : Salvation Army stores reportedly charge according to the customers' means, a p o l i c y which has occasioned some i n t e r e s t i n g examples of well-to-do people getting 'dressed down' i n old clothes for a t r i p to the S a l l y Ann.) 9. An i l l u s t r a t i o n of such confusion was seen by the author during one Hallowe'en. In recent years, c h i l d r e n have begun c o l l e c t i n g money for char i t y instead of candy. As they are instructed not to take both money and candy, an impasse can r e s u l t when an adult t r i e s to give the c h i l d something which he knows a l l c h i l d r e n value (candy) together with money, and the c h i l d refuses. - 10. This condition could perhaps be replaced by a requirement that i f re-sources are increasing, they do so at a constant rate across members. For example, people r e g u l a r l y receive installments of income (ahdc.con-sume i t ) , but the comparative resource bases among group members does not change. That i s , the theory probably does not have to be l i m i t e d to e n t i r e l y s t a t i c s i t u a t i o n s i n which there i s no input of resources. Advantageous exchanges could s t i l l occur i n a given transaction, but would perhaps have a smaller e f f e c t on o v e r - a l l ranking, which i s c a l -culated over a longer time. 11. See H.H. Kelley and D. Schenitzki, 'Bargaining', Chapter 10 i n C.-G. McClintock, Ed., Experimental S o c i a l Psychology, Holt, Rinehart and Winston, 1972, for a discussion of the means of measuring and manipulat-ing motivational variables i n bargaining experiments. 73 12. In an experimental context, i t i s d i f f i c u l t to construct a task that could simultaneously involve resource production and t r a n s f e r , although there are probably elements of both i n enduring s o c i a l r e l a t i o n s h i p s . 13. Kuhn, op. c i t . , p. 286. 14. P w i l l not i n i t i a t e an exchange j u s t because i t would benefit him a l o t , unless he believes there i s some non-zero p r o b a b i l i t y of acceptance. The issue of the weight given to value gained as opposed to the p r o b a b i l -i t y of gain i s treated as an empirical issue here, though perhaps one can make two rough assumptions: 1) the more often P's i n i t i a t i o n s are rejected, the more s a l i e n t and therefore the more weight that i s given to the p r o b a b i l i t y of an outcome; 2) . i n the i n i t i a l stages of exchange, P has l i t t l e basis on which to i n f e r the p r o b a b i l i t y of acceptance of his o f f e r s , and w i l l tend to be o p t i m i s t i c , i . e . , i n f l a t e the perceived p r o b a b i l i t y of success. Psychological studies on b e t t i n g behaviour give evidence that peo-ple attend more to the p r o b a b i l i t y of gain under some conditions, and more to the s i z e of gain under others; i t also appears that there are i n d i v i d u a l differences with respect to how i n d i v i d u a l s weight the two f a c t o r s . See: P. S l o v i c and S. L i c h t e n s t e i n , 'The r e l a t i v e importance of pro-b a b i l i t i e s and payoffs i n r i s k taking', Journal of Experimental Psycho-logy, Monograph Supplement, 1968, No. 3, Part 2, pp. 1-18. 15. A weighted l i n e a r function (w^ P r o f i t to P + P r o f i t to 0) could also be used, but requires a more rigorous l e v e l of measurement of u t i l i t y and p r o b a b i l i t y , and independence of the f a c t o r s . I t would make i t possible to assume that P does i n fact consider both his own and Other's p r o f i t , and chooses an a l t e r n a t i v e giving the best balance of the two factors (so that P may prefer an exchange giving less p r o f i t to s e l f , i t i f l e t s him get more than some 0, than i n a transaction where P gains more himself, but 0 gets even more). 16. T h e o r e t i c a l l y , P should be able to take O's point of view even i f 0 does not have the same value function as P, so long as P has an understanding of that function ( i . e . , through past experience, extended exposure to others with that value function, e t c . ) . An i n t e r e s t i n g p o s s i b i l i t y i s that P w i l l f i n d i t much easier to take O's point of view when i t i s -the same as h i s own; t h i s may help to account for why people tend to compete and compare with s i m i l a r Others : the comparisons are more i n -terpretable to P. See P. Hoffman, L. Festinger, and D.H. Lawrence, 'Tendencies toward group comparability i n competitive bargaining', Human Relations, 7_, 1954, pp. 141-159; and B. Latane, Ed., 'Studies i n S o c i a l Comparison', Supplement 1, Journal of Experimental S o c i a l Psy-chology, 1966, for a discussion of conditions under which people tend to compare with s i m i l a r others. 74 17. While i t i s assumed that P adopts the point of view of others towards-himself, i t i s not assumed that he takes the point of view of every other to a l l possible others. While such a procedure might l e t P e s t i -mate to whom i n i t i a t i o n s w i l l be made i n the group and the nature of the i n i t i a t i o n s , we assume i t i s beyond the cognitive capacity of the actors to do so. 18. lexicographic function requires only that P be able to order his preferences, not assign numerical values to them. While t h i s i s an ad-vantage i n some ways, i t does not n e c e s s a r i l y y i e l d values which can be m u l t i p l i e d by a weighting f a c t o r . Consideration of the most s a t i s f a c -tory model are premature at t h i s point, as one must f i r s t e s t a b l i s h the manner i n which the process operates, before i t can be s u c c e s s f u l l y modelled. 19. D. E l l s b e r g , 'Risk, ambiguity and the savage axioms', Quarterly Journal of Economics, 7_5, 1961, pp. '643-669. 20. In addition to the E l l s b e r g a r t i c l e , see also J.S. Chipman, 'Stochas-t i c choice and subjective p r o b a b i l i t y ' , i n D. W i l l n e r , Ed., Decisions,  Values and Groups, New York: Pergamon Press, 1960; and S.W. Becker and F.O. Brownson, 'What p r i c e ambiguity? or the r o l e of ambiguity i n d e c i -sion making', Journal of P o l i t i c a l Economy, 7_2, 1964, pp. 62-73. These studies are reviewed by W. Lee i n h i s book Decision Theory and Human  Behaviour, New York: John Wiley and Sons,1971, pp. 119-123. 21. The claim that the r i s k y and the uncertain bets have equal expected value rests on the decision theoretic assumption that i f the p r o b a b i l i t y of events i n an uncertain bet are completely unknown, P's best estimate of p r o b a b i l i t y i s the midpoint of a l l possible and equally l i k e l y pro-b a b i l i t i e s . See Lee, op. c i t . , p. 121. 22. See E l l s b e r g , op. c i t . , p. 664, for h i s 'ambiguity corrected 1 payoff — b a s i c a l l y , P makes an estimate of the p r o b a b i l i t y d i s t r i b u t i o n over events, and then steps back and asks himself how confident he feels i n hi s estimate. I f his confidence i s low, he w i l l give more weight to a p e s s i m i s t i c or conservative estimate of p r o b a b i l i t i e s . Becker and Brownson, op. c i t . , questioned the appropriateness of E l l s b e r g ' s formu-l a t i o n s , but t h e i r findings support the claim that subjects w i l l pay to avoid an ambiguous course of action when that course of action has an expected value equal to an unambiguous a l t e r n a t i v e . One could also argue that i t i s not subjective p r o b a b i l i t y that i s revised by P, but rather his estimate of u t i l i t y — the main point being that somehow, expected u t i l i t y i s lowered i n uncertain bets. 23. E.E. Jones, K.J. Gergen, P. Gumpert, and J.W. Thibaut, 'Some conditions a f f e c t i n g the use of i n g r a t i a t i o n to influence personal ev&uation', Journal of Personality and S o c i a l Psychology, 1, 1965, pp. 613-623; 75 E.E.-.; Jones and J.W. Thibaut, 'Interaction goals as bases of inference i n interpersonal perception', i n R. T a g i u r i and L. P e t r u l l o , Eds.', .r Person Perception and Interpersonal Behaviour, StanfordrUnStanford. University Press, 1959; and D. Bramel, 'Determinants of b e l i e f s about other people', Chapter 4 i n J . M i l l s , ed., Experimental S o c i a l Psychology, Toronto: C o l l i e r MacMillan, 1969. 24. See footnote 23; also H.H-?.Kelley and A.J. S t a h e l s k i , 'The inference of intentions from moves i n the Prisoner's Dilemma Game', Journal of Experimental S o c i a l Psychology, 6_, 1970, pp. 401-419, and H.H. K e l l e y , ' A t t r i b u t i o n theory i n s o c i a l psychology', i n D. Levine, Nebraska Sym- posium on Motivation, Un i v e r s i t y of Nebraska Press, 1967, pp. 192-238. 25. J.T. Lanzetta and V.Tv.Kanareff, 'Information cost, amount of payoff, and l e v e l of a s p i r a t i o n as determinants of information seeking and de-c i s i o n making', Behavioural Science, 7_, 1962,. pp. 459-73; S. L i c h t e n s t e i n , 'Bases for preference among three-outcome bets', Journal of Experimental Psychology, 69_, pp. 162-169. 26. I t seems not unreasonable to argue that the i n i t i a t o r of an impression believes the impression w i l l be accepted at face value, even though when the i n i t i a t o r i s at the receiving end, he tends to be more susp i -cious of a discrepancy between the r e a l and the presented s e l f . One s e l -dom hears a person r e f e r to himself as a 'phoney', or as d e l i b e r a t e l y working a deception on others, though we commonly hear people v e r b a l i z e suspicions that others do t h i s . Support for the claim that P i s l i k e l y to b elieve 0 w i l l take his o f f e r s at face value comes from evidence about the existence of cognitive biases: people tend to think t h e i r attempts to exchange w i l l be reciprocated, even i f they do not intend to do the same; and they expect others to behave p o s i t i v e l y toward them even i f they f e e l or are acting negatively toward other. There i s a discussion of these findings i n E. Burnstein, 'Cognitive factors i n behavioural interdependence', i n J . M i l l s , Ed., Experimental S o c i a l  Psychology, Toronto: C o l l i e r MacMillan Co., 1969, pp. 309-340. 27. Kuhn, op. c i t . , p. 330. 28. R.M. L i e b e r t , W.P. Smith, J.H. H i l l , and M. K i e f f e r , 'The e f f e c t s of information and magnitude of i n i t i a l o f f e r on interpersonal negotia-t i o n ' , Journal of Experimental S o c i a l Psychology, 4_, 1968, pp. 432-441. i. 29. W.H. Foddy,'op. c i t . , 1971. 30. Emerson i s currently conducting research with 5 man groups, varying the degree of information about p r o f i t s i n bargaining dyads within a larger ,. exchange network. Results to date show that "Under conditions of symbolic; payoff and v i s i b l e equity, trade agreements other than (one for one) were extremely.rare, and when they did occur, they,/were l i k e l y to be r e c i p r o -76 cated." Emerson predicted that when "equity conditions are not v i s i b l e , (subjects with a power advantage) w i l l tend to use t h e i r power i n e f f e c -t i v e bargaining", i . e . , w i l l not f e e l held to equitable exchanges. See R.M. Emerson, 'Power and p o s i t i o n i n exchange networks', Paper presented at meetings of American S o c i o l o g i c a l Association, 1971. In a personal communication, Emerson stated that the trend i n the data was as pre-dicted, but that subjects frequently opted to make p r o f i t s v i s i b l e and to share (December, 1973). 31. Scope condition 6 ensures t h i s i s not tautological by requiring that P be able to initiate?'to 0 even i f he does not know O's value p o s i t i o n — a l l he needs to know i s what resource 0 wants. 32. See discussion i n Chapter 1 of the idea of a P with concealed resources making use of a 'model' transaction with lower costs and higher rewards to P. 33. See discussion of Case 6 — Asymmetric information;^alternatives i n Chapter 1. 34. Because the theory does not predict that a person who witholds i n f o r -mation about his resources w i l l never have an i n i t i a t i o n accepted (e.g., he may be accepted by a person who gets no other i n i t i a t i o n s ) , there w i l l be some people who 'succeed' with t a c t i c s of information c o n t r o l , and obtain an advantage — they are l i k e l y to repeat t h e i r behaviour. I t i s simply argued that t h e i r average rate of successful transactions w i l l be below that of persons whose resource positions and p r o f i t s are unambiguously known. Those people who f a i l i n t h e i r use of informational t a c t i c s w i l l be less l i k e l y to continue t r y i n g them. TThey need not a l l conclude that p r o v i s i o n of information i s necessa'E.yyt.— an a l t e r n a t i v e to t h i s might be to locate 'unpopular' others, (e. g. other concealed persons). However, i f a person i s unpopular, i t i s usually because he ttfakesmana undesirable exchange partners. 77 CHAPTER 3 RESEARCH DESIGN To test the hypotheses presented in Chapter 2, a modification was made to an experimental paradigm used i n i t i a l l y by W. Foddy in a study 1 of clique formation in exchange networks. A descriptionoqftthe experi-mental situation w i l l be presented f i r s t , including a discussion of the operationalization of variables central to the theory. The same basic de-sign was used for two sets of experiments, and a pilot study which is re-ported in Appendix IV. The two main sets of experiments are described in more detail at the end of this chapter, where attention is drawn to d i f -ferences in procedure between them. The variations in design were neces-sary, due to the fact that a satisfactory operationalization of a l l rele-vant variables could not be achieved in one design. Description of the Experimental Paradigm The research design centered on a trading game. Subjects were provided with supplies of red and green plastic bingo buttons, and they ex-changed one colour for another colour. In each experiment, eight subjects sat behind cardboardebooths which had been arranged on the top of an octagonal table. (See Figure A . l , Appendix I). The booths eliminated visual contact among subjects, while a mesh-covered window in the front of each booth allowed subjects to see the table in front of each booth. Each subject had a large pile of either red or green buttons on the table in front of him, and a smaller pile of the 78 other colour. Four of the booths were equipped with covers that concealed the buttons from the view of others; four booths had no covers (Figure A . l , Appendix I ) . Subjects could reach t h e i r buttons through a gap at the bottom of each booth. A f i x e d amount of buttons of the two colours were d i s t r i b u t e d before each experiment, so that four subjects had a large number of red and a smaller number of green. Two of these had covers over the buttons and two did not. Four others had complementary p r o f i l e s of many green and few red, again with covers on two of the p i l e s of buttons, and no covers on 2 the other two. Noofurther resources were added a f t e r the experiment be-gan. The two l e v e l s of v i s i b i l i t y of resources (covers and no covers), and the d i f f e r e n t resource p r o f i l e s were randomly assigned to the eight seating p o s i t i o n s , and the subjects were randomly assigned to booths i n 3 each experiment. A card pinned in s i d e each booth t o l d the occupant how much he had of each colour (Figure A.4, Appendix 1). During the experiment, the main l i g h t s i n the room were switched o f f , and a hanging lamp placed over the center of the table. This caused the subjects to s i t i n the shadows, but c l e a r l y illuminated any p i l e s of buttons that were out i n the open. Each person could i d e n t i f y the other 4 group members by means of l e t t e r s (G, H, ...N) attached to the outside of each booth; an i d e n t i c a l tag was pinned i n s i d e . Subjects with covers over their-resources also had a coloured tag on the outside of t h e i r booths, i n d i c a t i n g to others the colour of t h e i r large p i l e of buttons. Each booth contained a set of mimeographed i n i t i a t i o n forms (see Figures A.2 and A.3, Appendix I ) , and a bowl i n which to send buttons to 79 another person. A value chart used to demonstrate that the principle of diminishing marginal u t i l i t y applied to the buttons was taped to the inside of each Booth. CSee Figure A.5, Appendix I.) Subjects were provided with pencils and paper to make calculations. At the beginning of each experiment, subjects were asked to liste n 5 to a set of tape recorded instructions. (See Appendices II and III.) The instructions familiarized the subjects with the setting, indicated the man-ner in which they could make exchanges, and acquainted them with the value table. Subjects were told that they were in the f i r s t part of a two part game, and that they would need both colours of buttons for the second part, where the two colours would be used for completely different purposes. (This second part did not take place.) The instructions included extended examples showing how the worth of a given total of buttons was increased by having some of red and some of green, and how successive equal increments of one given colour were worth less and less, the larger the pile a person had of that colour. After the instructions were given, subjects could make an offer of buttons to another subject by f i l l i n g out an i n i t i a t i o n form (Figure A.3, Appendix I ) , indicating: — who was sending the offer (the subject's identification letter) — to whom the offer was directed (the letter on someone else's booth) — the amount of one colour that was being offered in return for a stated amount of the other colour. 80 Subjects were allowed to make only one i n i t i a t i o n at a time, although they were not required to make an i n i t i a t i o n i f they chose not to. In addition, they could only accept one o f f e r at a time ( i . e . , one per t r i a l ) , even i f they received more than one. (This r u l e was necessary to allow a tes t of subjects' preferences, given c e r t a i n combinations of off e r s . ) Subjects knew that they did not have to accept any of the o f f e r s they received, i f they did not want to. They were l i m i t e d to o f f e r i n g no more than 100 buttons at a time, to reduce extreme v a r i a b i l i t y i n the types of o f f e r s sent, and to make subjects believe the experiment would take several t r i a l s . No l i m i t s were placed on what they could ask f o r , or agree to give i n return for an o f f e r . The l i m i t of 100 on o f f e r s was uniform f o r a l l subjects. The subjects put the buttons being offered i n t h e i r bowls, along with a completed i n i t i a t i o n form, and the bowls and offers were then picked up by the experimenter. The of f e r s on t r i a l 1 were e i t h e r delivered to the person they were addressed to or prearranged o f f e r s were substituted for the r e a l offers and delivered. About three feet from the subjects' table, there was a long rec-tangular t a b l e , on which were placed large wooden screens (about 30 inches t a l l ) i n a haphazard fashion. Behind g the screens the experimenter kept bowls with completed i n i t i a t i o n forms, and bowls, of buttons made up to match the f a l s e i n i t i a t i o n forms. These were switched for the r e a l o f f e r s a f t e r the experimenter had c o l l e c t e d a l l the subjects' o f f e r s f o r a t r i a l . Subjects' at-tention was diverted from the deception by giving them a questionnaire to f i l l 81 out a f t e r the i n i t i a t i o n s werevcollected; and by t e l l i n g them that the experimenter was going to record t h e i r o f f e r s . Only two subjects i n a l l the experiments suspected that t h e i r o f f e r s had been intercepted; i n f a c t , during the debriefing, i t was sometimes d i f f i c u l t to convince a subject that h i s r e a l o f f e r had never been received by the person to whom he had addressed i t . When subjects received an o f f e r , they could reply to i t e i t h e r by c i r c l i n g the word 'rejected' on the i n i t i a t i o n form (Figure A.2, Appendix I ) , or, by c i r c l i n g the word 'accepted', and then replacing the buttons offered with the requested amount of another colour. One t r i a l , or opportunity for exchange, consisted of the sequence beginning with the sending of an o f f e r , and ending with the return of the bowls to t h e i r owners. No experi-6 ment was run for more than two t r i a l s . Questionnaires were given to the subjects: 1) a f t e r they had made i n i t i a l o f f e r s , but before any were delivered, and 2) a f t e r subjects had r e p l i e d to the f i r s t i n i t i a t i o n s but before t h e i r r e p l i e s had been delivered. (See Appendices II and I I I f o r the questionnaires that were used.) The questionnaires probed subjects' reasons for i n i t i a t i o n s and acceptances of o f f e r s , perception of the marginal value of buttons, and t h e i r preference for concealment of t h e i r own and others' resources. At the end of each experiment, subjects were debriefed and asked not to describe the experiment to anyone. Communication between subjects was r e s t r i c t e d to pre-written forms (Figures A.2 and A.3 i n Appendix I ) . Subjects were in s t r u c t e d that they could not write a d d i t i o n a l messages, beyond f i l l i n g i n the blanks on 82 the forms. The experimenter restricted her conversation with subjects to repetition of instructions i f c l a r i f i c a t i o n was requested, and reminders to the subjects to f i l l out their i n i t i a t i o n forms and questionnaires com-pletely. The subjects did not appear to be concerned that the experimenter could see the offers they made, and i t was clear that the experimenter would 7 make no announcements about the proposed terms of transactions. Operational Definitions and Scope Conditions The theory in Chapter 2 deals with two or more valued resources, X and Y, distributed unevenly in a group, such that for each member there are at least two others with complementary value positions on X and Y. (Scope condition 1.) The resources could be behavioural, material, or non-material, but had to be transferable. In addition, they were to be valued according to a marginal u t i l i t y function (Scope condition 2), and i t was specified that group members could know or infer that this value function applied to Others as well as themselves (Scope condition 3). To meet these requirements, two resources were operationally defined as red and green  bingo buttons (material commodities). The buttons were given value by instructing the subjects that both colours of buttons would be needed for the second part of the experiment, where the two colours would be used for 8 completely different purposes (see Appendices II and III for instructions). The buttons had the advantage of being free of individualistic associations- :-' which might cause each subject to value them according to an unknown func-tion that might vary across subjects. 8 3 The requirement that the buttons be valued according to a marginal  u t i l i t y function was met through i n s t r u c t i o n s to the subjects that the worth of d i f f e r e n t amounts of buttons was to be assessed by reference to the value  chart, on which equivalences of buttons and value units had been set accord-ing to the function a = 1/ b/2, where a i s value u n i t s , and b i s the number of buttons. The t o t a l number of buttons given to a subject constituted h i s resource base, and the number of each colour defined the value p o s i t i o n f o r each resource. The gain i n value units brought by an addition of a number of one colour of buttons could be calculated by comparing the equivalent amount's i n value units f o r the p i l e of buttons of that colour before and af t e r the addition was made. Costs i n value units could be calculated by comparing the worth i n value units of a p i l e before and a f t e r an increment of buttons was given up. Net p r o f i t was the r e s u l t of subtracting costs from rewards. These c a l c u l a t i o n s were made clear by means of examples. (Once p r o f i t i s operationally defined, the d e f i n i t i o n s of f a i r and advan-tageous exchanges follow.) Subjects were motivated to trade because an increase to t h e i r small p i l e of buttons brought a larger gain than the cost incurred by giving up an equivalent amount from the large p i l e . I t was emphasized i n the i n s t r u c -tions that a balance of two colours had more worth i n the second part of the game. I t was also made clear that large balanced p i l e s were worth more than small balanced p i l e s . Given that the t o t a l number of buttons i n the group was constant, this imposed a competitive o r i e n t a t i o n ; i n addition, 9 subjects were t o l d to do as we l l as they could. The expectation of a sec-ond part of the game met the condition that future comparison with others 10 would occur (Scope condition 4). 84 The value chart had the advantage of standardizing the worth of d i f f e r e n t numbers of d i f f e r e n t coloured buttons, without going into d e t a i l about the second part of the game. In addition, i t met Scope condition 3, s p e c i f y i n g that i n d i v i d u a l s could know the value function by which resources were valued, without having to know the resource base and value positions 11 of any p a r t i c u l a r Other. Scope condition 6 s t i p u l a t e d that some members of the group could withold information about resources, but would be unable to give out f a l s e information. The covers over four of the subjects' buttons s a t i s f i e d t h i s condition. These subjects were not allowed to communicate the s i z e of t h e i r resource>?bases to others. The coloured tags i n d i c a t i n g the predominant colour, and the types of o f f e r s made by these subjects were the only i n f o r -mation a v a i l a b l e about the p r o f i t s they were r e c e i v i n g , and t h i s information was ambiguous. Subjects were t o l d that persons with covers might have more, 12 l e s s , or the same, as those with v i s i b l e p i l e s of buttons; t h i s was i n -tended to create doubt about the s i z e and value p o s i t i o n of the concealed resources. On the other hand, the p i l e s of buttons of the four subjects that had no covers gave unambiguous information about the s i z e of the resource 13 bases, e s p e c i a l l y i n the set where the exact numbers of buttons possessed by v i s i b l e subjects was displayed on a card on the outside of the booths concerned. During the experiment, these v i s i b l e subjects had to place addi-tions to t h e i r resource bases made through trading out on the table where 85 14 others could see them. ;iNote: For e f f i c i e n c y , we w i l l henceforth r e f e r to subjects with covered buttons as n o n - v i s i b l e s ; and to those with no covers as v i s i b l e s . The task facing each subject, then, involved entering into ex-changes with other persons i n order to obtain roughly equal sized p i l e s of red and green buttons for use l a t e r on i n the experiment. The actor has, to s e l e c t from a set of four possible exchange partners, who ha^eresources complementary to h i s own. The s i t u a t i o n c a l l s f o r a degree of cooperation for anyone to b e n e f i t , but i n which the terms of the transaction are i n c o n f l i c t . In the experiment, P must decide which of several a l t e r n a t i v e others to whom he should i n i t i a t e . He has some information about the r e -sources of 3 or 4 other subjects, but i s i n the dark about 3 or 4 others. I n i t i a l l y , P has a choice i n terms of the number of buttons he w i l l o f f e r for a given return, and the l e v e l of p r o f i t he o f f e r s to 0 w i l l a f f e c t the chance his i n i t i a t i o n w i l l be accepted. Subjects would be generally aware of two c o n f l i c t i n g tendencies — as P's own p r o f i t increases (and P prefers t h i s ) , O's p r o f i t decreases (and 0 does not prefer t h i s ) . Thus's?' some b a l -ance must be struck such that P asks as much as p o s s i b l e , but does not expose himself to c e r t a i n r e j e c t i o n . Although P has no control over how many others i n i t i a t e to him, he does not have to take up an i n i t i a t i o n unless i t s u i t s him. Since t h i s i s true for the targets of h i s i n i t i a t i o n s too, the outcomes f o r a given t r i a l are uncertain, and P can complete a t r i a l with two, one, or no completed transactions. He can have an o f f e r accepted and accept one; ei t h e r of these events alone can occur; or he may receive no i n i t i a t i o n s 86 and have his own rejected. One subject can of course receive as many as four i n i t i a t i o n s from those with a complementary colour. Variations i n the Paradigm The next sections describe the d e t a i l s of two sets of experiments conducted to test the hypotheses developed i n Chapter 2. A de s c r i p t i o n of a p i l o t study i s included i n Appendix IV. There were fourteen experiments:-i n each set. The research paradigm described i n th i s chapter was used f o r a l l sets. The varia t i o n s resulted as attempts were made to improve the ope r a t i o n a l i z a t i o n of t h e o r e t i c a l v a r i a b l e s , and to eliminate confounding e f f e c t s . However, the changes cannot be viewed as a developmental sequence moving from bad to good, since i t turned out that changes made to strengthen c e r t a i n aspects of the experiment weakened others. We w i l l comment b r i e f l y on three of the problems created by the paradigm. To test Hypothesis 1, that P w i l l prefer to i n i t i a t e to an Other with a large amount of Y r e l a t i v e to X, i t was necessary to have at le a s t two l e v e l s of resource base, or two le v e l s of value p o s i t i o n ( t o t a l num-bers of buttons possessed by d i f f e r e n t v i s i b l e s ; balance of red and green for d i f f e r e n t v i s i b l e s ) . However, this confounded the test of Hypothesis 4, that v i s i b l e s were the most preferred partners, as i t would not be possible to t e l l i f subjects had i n i t i a t e d to a v i s i b l e Other because he was v i s i b l e , because he had a l o t of buttons, or both. Consequently, i t was necessary to have a set with differences i n resource bases among v i s i b l e s , and a set where there were none. 87 I t was necessary to intercept subjects' r e a l o f f e r s , and s u b s t i -tute two i d e n t i c a l i n i t i a t i o n s (one each from a v i s i b l e and n o n - v i s i b l e ) , 15 to check subjects' preferences as predicted i n Hypothesis 5. This had to be done on the f i r s t t r i a l , before subjects set up obligations with subjects on the basis of f i r s t i n i t i a t i o n s and acceptances. Such interven-t i o n , however, prevented us from seeing i f the 'natural' sequence outlined i n Chapter 1 ( i . e . , where non-visibles t r y for advantages?.'^ but receive few i n i t i a t i o n s , f a i l to have t h e i r o f f e r s accepted) did i n fact lead to the view that witholding information i s r e a l l y no advantage when Others have a l t e r n a t i v e s . Scope-Condition 3 s t i p u l a t e d that subjects be i n doubt about the resource l e v e l s of non-visible subjects. I t has been argued that the main strategy open to P to obtain advantage i n this case i s to act 'as i f he could not af f o r d to give up as many buttons as he would i n a f a i r trade. For t h i s to be a r e a l i s t i c strategy, other members must at le a s t e n tertain the p o s s i b i l i t y that some of the non-visibles were 'wealthier' or 'poorer' than themselves, so that f a i r exchanges would not be seen simply as one for one trades of buttons. A problem arose because subjects tended to assume symmetry i n the d i s t r i b u t i o n of resource p r o f i l e s among v i s i b l e s and non-v i s i b l e s . Several non-visible subjects mentioned that i t was a 'simple matter'.to guess that the non-visibles had the same p r o f i l e s as the visiblesy,, when a l l the v i s i b l e s had the same resource p r o f i l e (and t h i s was a neces-16 sary condition for the test of Hypotheses 4 and 5). Unless doubt could be created, Scope condition 3 would not be met. I f the experimenter t o l d subjects that non-visibles were d i f f e r e n t , and i t was patently obvious to 88 each non-visible that lie was not d i f f e r e n t , then the experimenter was l y i n g . Although i t would have been better to make a l l subjects a l i k e for the pur-pose of t e s t i n g Hypotheses 4 and 5, the simultaneous t e s t i n g of Hypothesis 2 made t h i s seem unworkable. It was decided to create differences of r e -sources between v i s i b l e s and no n - v i s i b l e s , and to e x p l i c i t l y inform the subjects that differences of some sort did e x i s t . I t was s t i l l p ossible to set the resource bases and value positions of v i s i b l e s a l l equal (excepting complementarity). With these preliminary considerations i n mind, the d e t a i l s of Sets A and B are given i n the following section. Set A Fourteen experiments were conducted f o r Set A, with the follow-ing d i s t r i b u t i o n of resources and information (see Figure 3.1). The t o t a l s of 1630 and 830 were used, because bingo buttons are quite small, and one needs a p i l e of over 600 to look noticeably large enough to trade from. Non-visible subjects had somewhat more information than v i s i b l e subjects, i n that t h e y . ; C o u l d see that they had an i d e n t i c a l p r o f i l e to one v i s i b l e subject, and the same s i z e of resource base as two others. Noteithat i n Set A, subjects did not know the exact s i z e of others' resource bases, but only that some players had more buttons than others. Thus, an advantageous exchange can be defined by the experimenter, but the subjects could not know the exact number of buttons that would s a t i s f y a Figure 3.1 D i s t r i b u t i o n of Resources and Information i n Set A Covers Red Buttons Green Buttons m JZL y / - I Hi red v i s i b l e 1600/30 Hi green v i s i b l e 1600/30 Lo red v i s i b l e 800/30 Lo green v i s i b l e 800/30 F u l l information about resource base and value positions -available to'^fbers. Hi red non-vis 1600/30 Hi green non-vis 1600/30 Lo red non-vis 800/30 Lo green non-vis 800/30 No information about resource base and value positions available to others. Note: Eaqh subject has 4 possible partners: two v i s i b l e and two non-visible, with complementary-resource p r o f i l e s . 90 c r i t e r i o n of f a i r n e s s . I t was expected the actor would consider other players to be of the same resource l e v e l , or as having more (or fewer) r e -sources. In Set A, the f i r s t o f f e r s were delivered to the booths indicated  by the i n i t i a t o r on the form. No o f f e r s were delivered u n t i l a l l had been c o l l e c t e d , to avoid a f f e c t i n g the o f f e r s of subjects who might receive an i n i t i a t i o n before sending one. Before d e l i v e r i n g the f i r s t set of o f f e r s , the experimenter had subjects complete the Questionnaire (see Appendix I I ) , holding the bowls at a side table u n t i l the subjects had completed the ques-tion s . After the t r i a l 1 o f f e r s had been processed by the subjects, the experimenter returned the bowls, with buttons and i n i t i a t i o n forms to t h e i r owners, and i n s t r u c t e d the subjects to make t h e i r second o f f e r . AAgain the experimenter gave them the same questionnaire to complete, and while the subjects were occupied with i t , switched the r e a l o f f e r s for f a l s e ones, a l l directed to v i s i b l e subjects. A l l f a l s e offers asked for 100 buttons of the r e c i p i e n t ' s predominant resource, i n return f o r 100 buttons of the i n i -t i a t o r ' s predominant resource. High v i s i b l e s received bogus o f f e r s from the other high v i s i b l e , and from a high n o n - v i s i b l e ; low v i s i b l e s received them from the other low v i s i b l e and from a low non-visible. They did not, of course, know the resource l e v e l of the non-visible source. The fake o f f e r s were made within resource l e v e l s to l i m i t the number of sources of v a r i a t i o n . Set B This set r e a l l y consisted of two sub-sets of experiments, with seven groups of subjects i n each. When i t i s necessary to d i f f e r e n t i a t e them, 91 the subsets w i l l be referred to as B l and BII. TheVresource p r o f i l e s f o r the two subsets are shown i n Figure 3.2. While i n Set A there were two high and two low non-visibles i n each experiment, i n Set B there were either four high or four low non-v i s i b l e s . Since no one knew t h i s , there i s no c r u c i a l d i f f e r e n c e between these, and experiments where highs and lows are both present. V i s i b l e sub-j e c t s had signs on the outsides of t h e i r booths showing the exact number of buttons. Non-visibles had coloured tags showing whether t h e i r predominant resource was red or green. Further differences introduced i n Set B were as follows: 1) A l i m i t of f i v e t r i a l s was stated i n the i n s t r u c t i o n s , to attach some cost i n time to exploratory o f f e r s . 2) Only one transaction (one t r i a l ) was a c t u a l l y completed (compared with two i n Set A). 3) Subjects were t o l d there would be four winners, and that these would be the four who had 'done best'. 4) False o f f e r s asked for 110 buttons i n return f o r 100 offered. While t h i s increases the l i k e l i h o o d of double r e j e c t i o n s , i t forces subjects to express a preference f o r the source of an o f f e r which does c e r t a i n l y (in the case of a v i s i b l e i n i t i a t o r ) , and may possibly (with the non-v i s i b l e i n i t i a t o r ) be giving the r e c i p i e n t a lower p r o f i t than the partner. 17 5) Non-visibles, as well as v i s i b l e s , received two f a l s e o f f e r s each. This interference with the f i r s t set of i n i t i a t i o n s made i t impossible Figure 3.2 D i s t r i b u t i o n of Resources and Information i n Set B Red Green t-Cover |_ \ J L J L A / Red v i s n Red v i s , Green v i s . Green v i s , Red NV Red NV„ Green NV, 1100R/100G 1100G/100R VISIBLES 1 • "'2 "'1 1250R/150G 1205G/150R HIGH NON-VISIBLES Green NV, / / /f / / / / / / / / / / / / / / n / / L J Red v i s . Red v i s , Green v i s . Green v i s . Red NV Red NV, Green NV, 1100R/100G 1100R/100G VISIBLES 1 *"2 — 950R/50G 950G/50R LOW NON-VISIBLES Green NV, 93 to continue the experiment beyond one t r i a l , without exposing the de-18 ception. 19 In other respects the procedure was b a s i c a l l y the same as for Set A. More extensive questions were asked concerning subjects' reasons for the s i z and target of i n i t i a t i o n s , and reasons for acceptance and r e j e c t i o n of o f f e r s . The data that r e s u l t from Set B consist of r e a l i n i t i a t i o n s by subjects on T r i a l 1; acceptance of pre-written o f f e r s (and i n Set BI, acceptance of r e a l o f f e r s to v i s i b l e s ) ; and the subjects' written comments about t h e i r perception of various aspects of the s i t u a t i o n . It should be emphasized that the strongest test of the p r e f e r -ence for a v i s i b l e exchange partner would come from the data for v i s i b l e subjects i n Set B. They were the only subjects who could see no d i f f e r e n -t i a l s i n resource bases among v i s i b l e players, and who were t r u l y uncer-t a i n about the resource p r o f i l e s of non-visible others. For both types of no n - v i s i b l e , high and low, a resource d i f f e r e n t i a l with the v i s i b l e s i s evident — non-visible highs have more than v i s i b l e s , and non-visible lows i n Set BII have l e s s . Predictions f or Experimental Paradigm A short d e s c r i p t i o n of the data generated from the experimental paradigm which w i l l be relevant to t e s t i n g each of the hypotheses i s given below. Actual r e s u l t s from the experiments w i l l be presented i n the next chapter. 94 Hypothesis 1: The p r e d i c t i o n that an actor w i l l prefer to i n i t i a t e to an Other who has the best prices implies that he w i l l o f f e r to the High resource players or to players with a large imbal-ance of one commodity over another. As t h i s hypothesis applies only to v i s i b l e targets Set B i s not relevant, since a l l the v i s i b l e s i n Set B had the same t o t a l resource base and value p o s i t i o n s . Thus we w i l l use data from Set A, and some r e s u l t s from the p i l o t study w i l l also be presented. In the l a t t e r , subjects a l l had the same t o t a l number of buttons (1200 each), and d i f f e r e d only i n the resource p o s i -tions on each colour. Some subjects were near to balance, some were very unbalanced, and r e l a t i v e to these v i s i b l e s , the non-visibles had an intermediate l e v e l of balance. In both sets, the hypothesis w i l l receive support i f the major-i t y of i n i t i a t i o n s ^ a r e addressed to the v i s i b l e s with the best p r i c e s . Hypothesis::2: To test the tendency f o r non-visibles to make advantageous exchanges, we s h a l l compare the proportions of non-visibles / and v i s i b l e s making advantageous i n i t i a t i o n s . In addition, the ' t y p i c a l ' o f f e r s of the two groups w i l l be compared, on the expectation that non-visibles w i l l make i n i t i a t i o n s that bring more p r o f i t to themselves, i f accepted. V i s i b l e s should make t y p i c a l o f f e r s that are ' f a i r ' . F a i r w i l l be defined r e l a t i v e to the resource base of the exchange part-9 5 ners. Data w i l l mainly be from Set B, where the confounding of wealth l e v e l and v i s i b i l i t y d i d not occur, and where sub-j e c t s knew the exact s i z e of the v i s i b l e Others' resource bases. Hypothesis 3 : Data relevant to whether or not the non-visibles perceived an advantage i n the covers w i l l be responses made by non-v i s i b l e s on the questionnaire given on T r i a l 1 i n both sets. In p a r t i c u l a r , questions about the perceived advantages and disadvantages of the covers, and about the desire to r e t a i n them, are relevant. Again, data from Set B w i l l be of greatest i n t e r e s t , because subjects' perception of the covers was more systematically probed i n t h i s set. Data from both sets w i l l be examined concerning the i n i t i a l de-s i r e to r e t a i n the covers. In addition, we can examine subjects' estimates of the p r o b a b i l i t y that t h e i r i n i t i a l offers w i l l be accepted, to see whether the non-visibles an-t i c i p a t e any reduced l i k e l i h o o d of acceptance due to the con-cealment of t h e i r resources. Hypothesis 4: Set B i s the best t e s t of the preference to i n i t i a t e to someone about whom one has r e l i a b l e information, as the v i s i -bles i n Set B are i d e n t i c a l i n t o t a l resource base s i z e , and i n balance of resources. In addition, t h e i r t o t a l resource base i s known exactly by cards on the outside of t h e i r booths. The relevant data are simply-proportions of i n i t i a t i o n s to 96 v i s i b l e s and non-vi s i b l e s, by both v i s i b l e s and no n - v i s i b l e s . The expectation i s that s i g n i f i c a n t l y more than h a l f of a l l o f f e r s w i l l go to the v i s i b l e s . Despite the confounding of wealth and v i s i b i l i t y i n Set A, data from these experi-ments w i l l also be examined, to see i f a s i g n i f i c a n t l y greater proportion of i n i t i a t i o n s continue to be addressed to v i s i b l e s on T r i a l 2 . Subjects' written reasons for i n i t i a -t ions w i l l give supplementary evidence that the subjects responded to v i s i b l e resources i n the manner predicted. Hypothesis 5 : The f a l s e o f f e r s prepared by the Experimenter to make two i d e n t i c a l i n i t i a t i o n s , one from a v i s i b l e and one from a non-v i s i b l e , allow a test of the f i f t h hypothesis. We expect a preponderance of acceptances of the o f f e r s from a source with  v i s i b l e resources.. Set B provides the best test of t h i s p r e d i c t i o n , as the f a l s e o f f e r s were" given both to v i s i b l e s and non- v i s i b l e s , and they were received by players before they had completed any other transactions. I f , i n addition to a simple preference for v i s i b l e s , non-visibles a c t u a l l y made worse o f f e r s ( i . e . , o f f e r s to P that gave P less p r o f i t ) , then the rate of acceptance of i n i t i a t i o n s from a non - v i s i b l e source would be even lower than i f a non-visible and a v i s i b l e made i d e n t i c a l o f f e r s to a given P. To tes t t h i s conjecture, o f f e r s from Set A, which were delivered on the f i r s t t r i a l without interference, w i l l be examined. 97 Hypothesis 6: If the n o n - v i s i b l e s ' perception of the advantages i n covers decreases over exchange transactions, we expect one possible reaction to t h e i r predicted f a i l u r e to be a d e s i r e to reveal information about t h e i r resources. Thus, the data relevant to t h i s hypothesis are the subjects' stated preferences (on the Questionnaire) about r e t a i n i n g or removing the covers. Set A must provide the data here, '. as the non-visibles were questioned 'before' and ' a f t e r ' engaging i n transactions. An increase i n the frequency of subj ects d e s i r i n g the covers o f f can be read as support for the p r e d i c t i o n . Sample Subjects for the experiments were male college students at the University of Alberta, Canada, between the ages of 17 and 39 years. The students were a l l undergraduates, who had volunteered at the request of the experimenter. Over 500 students from introductory chemistry, physics, engi-neering, a g r i c u l t u r e , commerce and business administration, and sociology courses volunteered, and of these, 352 subjects were used. Subjects were assigned to experiments on the basis of free time i n t h e i r timetables. No assumptions are made about whether the sample i s representative of a given population: the p o s i t i o n taken here i s that the behavioural p r i n c i p l e s under i n v e s t i g a t i o n apply to a l l people, and that demographic v a r i a b l e s are not systematically related to the dependent variables of t h e o r e t i c a l i n t e r e s t i n t h i s study. No personality measures were used as independent v a r i a b l e s , 98 because the theory presented here i s concerned with variables which are assumed to operate i n a s i m i l a r way for a l l subjects. It i s conceivable that personality may influence the development of an i n d i v i d u a l ' s p r e f e r -ence ordering for resources — once t h i s i s known, however, the p r i n c i p l e s governing t h e i r exchange behaviour are not assumed to be i d i o s y n c r a t i c . Females were not used as subjects, on the expedient grounds that, although females volunteered at a higher rate than males, past exper-ience with female subjects indicated that they do not a c t u a l l y show up as r e l i a b l y as males. Since the experiments required eight subjects at once, i t was decided to use males to minimize the number of experiments that could not be run for lack of subjects. (As i t was, an extra subject was occasionally s o l i c i t e d i n the c a f e t e r i a , i f only seven subjects appeared.) 20 Sex of subjects i s thus held constant. Because both l e v e l s of the information v a r i a b l e were run simul-taneously, no precautions were necessary to avoid unequal maturation, or fatigue, of the subject pool. Subjects were, requested not to describe the experiment to t h e i r f r i e n d s , as knowledge of the hypotheses or the experi-mental manipulation of o f f e r s received during the game would i n v a l i d a t e the r e s u l t s . Subjects appeared to have honoured t h i s request i n the main — only two subjects (in the same experiment) suspected the deception, and the data from t h i s experiment were discarded. A f i n a l comment should bamade concerning the comparison of t h i s 21 paradigm with those frequently used i n studies of bargaining. To date, most experiments have provided neither a l t e r n a t i v e exchange opportunities, 9 9 nor a resource base against which fairnes s or equity can be defined i n terms other than a 50-50 s p l i t of money p r o f i t s (which r e a l l y represents equality, not equity). The present experiment attempts to operationalize such fac-tors, which are generally agreed to be important. Because each subject has several a l t e r n a t i v e s , there i s also no need to require subjects to reach 2 2 a settlement within dyad before they can leave the experiment. Studies of b i l a t e r a l monopoly bargaining have been able to focus on the sequential adjustment of i n i t i a t i o n s and expectancies, and on the e f f e c t s of symmetric 2 3 and assymetric information for the bargaining outcomes, but have been unable to study the factors of a l t e r n a t i v e s , and preferences for c e r t a i n kinds of partners. Thus, while the present study arose i n the context of the s o c i a l psychology of bargaining and exchange, the method of i n v e s t i g a -t i o n of the phenomena i s d i f f e r e n t , with the t h e o r e t i c a l emphasis on the choice of exchange r e l a t i o n s h i p s , and i n i t i a l t a c t i c s , rather than on the i n t e r n a l bargaining i n a p a r t i c u l a r dyad that leads to a si n g l e s o l u t i o n and d i v i s i o n of a payoff. Figure 3.3 Summary of Features of the Two Experimental Sets SET A 14 experiments Each experiment had: - v i s i b l e s - 2 highs(1630 buttons) -2 lows ( 830 buttons) -non-visibles -2 highs(163;0 buttons) _ _ „-? lows ( 830 buttons) Exact t o t a l s of v i s i b l e s r e -sources not displayed. F i r s t i n i t i a t i o n not intercep-ted, delivered to r e a l target, and returned to i n i t i a t o r ( T r i a l 1). T r i a l 2 i n i t i a t i o n s intercepted, f a l s e o f f e r s of 100/100 delivered to v i s i b l e s  only, r e p l i e s c o l l e c t e d . Questionnaires: a f t e r i n i t i a -tions made and picked up but p r i o r to d e l i v e r y on T r i a l 1; during T r i a l 2, non-visibles r e p l i e d to second question-naire. Experiment terminated a f t e r T r i a l 2. SET B Set Bl — 7 experiments Each experiment had: - v i s i b l e s - 4 mediums(1200 buttons) -non-visibles -4 highs (1400 buttons) Exact t o t a l s of v i s i b l e s ' r e -source p i l e s written on cards on subjects' booths. Real offers delivered to v i s i b l e s only; r e p l i e s picked up; off e r s of non-visibles intercepted, sub-s t i t u t e d two off e r s of 100/110 from v i s i b l e and non-visible - . T r i a l 1. Questionnaires: a f t e r T r i a l 1 i n i t i a t i o n made; second question-naire a f t e r r e p l i e s to T r i a l 1 of f e r s c o l l e c t e d . Set BII — 7 experiments Each experiment had: -v i s i b l e s - 4 mediums(1200 buttons) -non-visibles -4 lows (1000 buttons) Exact t o t a l s of v i s i b l e s ' r e -source p i l e s displayed. T r i a l 1: r e a l offers of a l l sub-j e c t s intercepted; two f a l s e offers to a l l subjects of 100/110. Questionnaires: after T r i a l 1 i n i t i a t i o n made; second question-naire a f t e r r e p l i e s to T r i a l 1 offers c o l l e c t e d . Experiment terminated at end Experiment terminated at end of T r i a l 1. of T r i a l 1. 101 FOOTNOTES FOR CHAPTER 3 1. W.H. Foddy, 'The formation of cliques i n c o l l e c t i v i t i e s as a consequence of i n i t i a l d i s t r i b u t i o n s of dimensions of wealth', Unpublished Ph.D. D i s s e r t a t i o n , University of B r i t i s h Columbia, 1971. 2. S. Siegel, A.E. Siegel, and J.M. Andrews, Choice, Strategy and U t i l i t y , New York: McGraw H i l l , 1964. Siegel.postulates that the choice of a response i n p r o b a b i l i t y learning s i t u a t i o n s i s determined by the u t i l i t y of a correct response and the u t i l i t y of varying choice, which Siegel claimed arose out of boredom. In s o c i a l exchange, more than j u s t the escape from boredom i s achieved by choice v a r i a b i l i t y — people do not l i k e to incur too many obligations with one i n d i v i d u a l , as i t increases dependence on that person. 3. This was to eliminate systematic bias due to greater v i s u a l c e n t r a l i t y of the person opposite. In W. Foddy's study, there was a tendency for subjects to d i r e c t a greater proportion of t h e i r i n i t i a t i o n s to persons d i r e c t l y opposite. See W. Foddy, op. c i t . , 1971. 4. To avoid alpha preference. 5. For each set of experiments, subjects l i s t e n e d to a tape recording of i n s t r u c t i o n s , and i n Set B but not Set A, read a t r a n s c r i p t of the tape while i t was playing. Subsequent to the f i r s t set of experiments, a study conducted by W. Foddy indicated that subjects could r e c a l l more information from the dual presentation of i n s t r u c t i o n s , than they could i f tape only or t r a n s c r i p t only was used. Because the i n s t r u c t i o n s i n -duce some of the main scope conditions and assumptions, and because i t was important that subjects understand them i n the most uniform and com-plete manner possible, i t was appropriate that Foddy's findings should be applied. .Our experiments were a l l conducted by the author, so that bias due to the sex of the experimenter i s held constant, though i t i s unknown. See: W.H. Foddy, 'On getting through to some of the people some of the time', Unpublished Manuscript, Edmonton: University of Alberta, 1972. 6. Subjects believed the experiment would continue for more than two t r i a l s See d e s c r i p t i o n of the two experimental sets f o r d e t a i l s . 7. It could be objected that n o n - v i s i b l e subjects were v i s i b l e to the exper imenter, and may have f e l t constrained to be ' f a i r ' . Such an objection has l e s s force, i f one accepts the conceptualization of norms as 'pru-d e n t i a l maxims' rather than moral rules that make one f e e l g u i l t y i f one benefits by breaking them. That i s , since n o n - v i s i b l e subjects were not trading with the experimenter, the f a c t that she could see whether or not t h e i r o f f e r s were equitable, would not have affected the probabil i t y that the o f f e r would be accepted by the subject to whom i t was d i r e c ted. 102 8. The provision of a value chart and the deception regarding a second part were deemed necessary, because previous work with the paradigm suggested that a p r i n c i p l e of diminishing marginal u t i l i t y would only operate when subjects knew why they wanted a resource, and what i t was worth to them. See W. Foddy, op. c i t . , 1972. 9. In the second set, subjects were t o l d there would be four winners. The objection could be made that having only one winner would induce more competition. It was decided to have four winners, so that subjects s t a r t i n g at a less advantaged p o s i t i o n i n terms of t o t a l resource base would s t i l l believe that they could 'catch up', and so that these people would not give up and lose i n t e r e s t i n comparing with others. Some comments from subjects i n Set A, where there were large differences i n wealth, and no winners, indicated that some subjects did not tr y to make comparisons with players at a d i f f e r e n t wealth l e v e l . Hoffman, et a l . , give ' evidence that people cease comparison with a person who i s doing much better i n a game, and cannot be 'caught up'. See: P. Hoff-man, L. Festinger, and D.H. Lawrence, 'Tendencies toward group compara-b i l i t y i n competitive bargaining', Human Relations, 7_, 1954, pp. 141-159; see also B. Latane, Editor, 'Studies i n s o c i a l comparison', Journal of Experimental S o c i a l Psychology, Supplement 1, 1966. 10. There i s a p o s s i b i l i t y that subjects w i l l trade with others with whom they cannot compare, i n order to get ahead of someone they can see, some-one with whom they, are comparing p r o f i t s . In t h i s sense, comparability refers to s i m i l a r i t y of progress towards maximizing t o t a l value units (balancing the p i l e s of buttons, rather than to comparison of subjec-t i v e p r o f i t s f or a given exchange)". Hoffman, Festinger and Lawrence (op. c i t . , 1954), found that subjects preferred to share payoffs with a non-comparison other, as a means of getting ahead of another subject with whom they were competing. It may be the case that v i s i b l e others were more s a l i e n t competitors, since t h e i r progress toward balanced resource p i l e s could be watched. There might then be an 'end-game' e f f e c t , with v i s i b l e subjects avoiding exchanges with one another, but paying no atten-t i o n to the p o s s i b i l i t y that non-visible players might also be nearing optimum balance of resources. The consequence of t h i s would be that the al t e r n a t i v e s about which P has r e l i a b l e information seem so undesirable, that the expected value of the uncertain exchange (with a non-visible) i s greater i n s p i t e of i t s greater ambiguity. Such an ef f e c t i s more l i k e l y i n the experimental s i t u a t i o n with fixed t o t a l resources and a clear-cut f i n a l t a l l y i n g of winners and l o s e r s . These conditions may give an u n r e a l i s t i c representation of conditions found i n ongoing ex-change r e l a t i o n s h i p s , where people probably attend more to t h e i r current exchanges and short-term comparisons. Where t o t a l resources i n the group are s t a t i c , end-game e f f e c t s w i l l occur because i t i s possible f o r a l l to reach equimarginality of resources, and have no further motivation to exchange; i f resources are consumed and renewed (as with s a l a r i e s ) , 103 or changing (as with a change i n task), people probably attend more to immediate balance of p r o f i t s . Since experimental groups are short-term, i t i s d i f f i c u l t to design an experiment where f i n a l rankings are not ex-tremely s a l i e n t . 11. In the end, i n d i v i d u a l i s t i c associations and values would have to be measured, and provided for i n a theory. At t h i s stage of t e s t i n g general p r i n c i p l e s , however, i t i s simpler to c o n t r o l f o r such v a r i a b i l i t y , than to i d e n t i f y a range of value functions, and t r y to include a l l combina-tions i n a study. 12. In the second set, they were t o l d that subjects with covers might have more, or l e s s , but did not have the same numbers as v i s i b l e s . 13. Bartos says that P has ' r e l i a b l e knowledge' i f there i s an unimpeachable authority which guarantees information ( i . e . , implying that 0 i s not r e l i a b l e , and a d i s i n t e r e s t e d party w i l l be looked on as a source of trusted information). See O.J. Bartos, 'Towards a r a t i o n a l empirical model of negotiations', i n J . Berger, M. Z e l d i t c h and B. Anderson, E d i -t o r s , jSoj^icd^j^Lcji^ Houghton, M i f f l i n , Co., 1972, Vol. I I . 14. Some subjects t r i e d to conceal additions to t h e i r p i l e s of buttons by keeping them insi d e t h e i r booths. 15. An o f f e r of 100 buttons of one colour f o r 100 buttons of the other colour was used i n Set A. In Set B, the f a l s e o f f e r s asked 110 i n return for 100. To avoid suspicion, the experimenter varied the handwriting on o f f e r s going to the same person. An o f f e r of 100 for 100 was chosen as t y p i c a l , though perhaps an o f f e r asking more than i t offered would have forced the subject to express his preference more strongly, as a one for one o f f e r i s 'prominent', and may have aroused l e s s suspicion. Be-cause the rules permitted subjects to accept only one o f f e r at a time, t h e i r choice between two i d e n t i c a l o f f e r s from a v i s i b l e and a non-v i s i b l e source could be taken as an indica t o r of t h e i r preference. See: T.C. S c h e l l i n g , The Strategy of C o n f l i c t , Cambridge: Harvard Univer-s i t y Press, 1960. S c h e l l i n g discusses the prominence of some solutions i n games of coordination, suggesting that i f communication i s prohibited, subjects choose obvious solutions i n the hope that t h e i r partners w i l l do likewise (e.g., meeting at a crossroads; d i v i d i n g a d o l l a r 50-50, etc.) 16. The bias may also exist due to the expectation that much of normal ex-change does i n fa c t take place among peers, despite the desire to ex-change with those who have a l o t of resources. In an experimental para-digm s i m i l a r to the one used here, W. Foddy demonstrated that although both high and low resource persons made f i r s t i n i t i a t i o n s to high r e -source persons, there was a tendency for exchange to s t r a t i f y , with high resource persons trading with each other, and low resource persons by 104 default becoming trading partners with each other. This was i n t e r p r e -ted as the r e s u l t of high resource persons being able to o f f e r one ano-ther a higher p r o f i t compared with that offered i n i n i t i a t i o n s from low resource persons. See W.H. Foddy, op. c i t . , 1972. 17. False o f f e r s were d i s t r i b u t e d on T r i a l 1. In Set B l , only the non-v i s i b l e s received f a l s e o f f e r s . Real o f f e r s were delivered to v i s i b l e s . In Set BII, a l l subjects received two f a l s e o f f e r s . It was i n i t i a l l y believed to be impossible to t r o t out 16 bowls when only 8 were supposed to e x i s t . Experience indicated that subjects did not expect to be de-ceived, and simply d i d not notice how many bowls the experimenter car-r i e d around. 18. The only return possible would have been to r e j e c t a l l the n o n - v i s i b l e s ' o f f e r s i n Set B l , and a l l o f f e r s i n BII; because the o f f e r s were i n t e r -cepted, the o f f e r s subjects responded to could not be given back to the supposed sender. The major purpose of Set B was to separate the ef-fe c t s of d i f f e r e n t i a l resource p r o f i l e s and v i s i b i l i t y , and to t e s t Hypothesis 5 concerning acceptances for both v i s i b l e s and n o n - v i s i b l e s . 19. See Figure 3.3 at end of t h i s Chapter. 20. The l i t e r a t u r e on experimental games provides contradictory evidence about sex differences i n game behaviour, e s p e c i a l l y concerning the issue of whether females are kinder and more cooperative, or more competitive and suspicious. Vinacke found more cooperative behaviour among females i n a c o a l i t i o n game, except when cumulative scores or rank considerations were introduced, at which point sex differences i n behaviour vanished. Bixenstein, Chambers and Wilson (1964) found males played more coopera-t i v e l y than females i n a Prisoner's Dilemma game; Komorita found females more cooperative than males; and Bixenstein and Wilson (1963), Bixen-s t e i n , Potash and Wilson (1963), and Tedeschi, et a l . (1969) found no s i g n i f i c a n t sex/choice behaviour c o r r e l a t i o n . The findings about sex differences are equivocal because sex, which i s a global, many-faceted category, has not been linked i n a systematic t h e o r e t i c a l way to a p a r t i c u l a r sort of behaviour or behaviours i n experimental games. Often, researchers s t a r t with a commonsense idea of how boys d i f f e r from g i r l s , and see i f t h i s d i f f e r e n c e predicts aspects of game behaviour. In addition, personality and sex differences are not usually good predictors of behaviour unless the statements employing them are conditioned by s i t u a t i o n a l v a r i a b l e s . In terms of t h i s research, there i s no p a r t i c u l a r reason to suspect women are governed by d i f f e r e n t p r i n c i p l e s i n exchange than are men, though they may well assign d i f f e r e n t values to various resources. For studies that attempt to r e l a t e sex r o l e to game behaviour (mainly i n Prisoner's Dilemma), see: D.W. Conrath, 'Sex r o l e and cooperation i n the game of Chicken', Journal of C o n f l i c t Resolution') _16, 1972, pp. 433-442; B. Jones, M. Steele, J. Gahagan, and J. Tedeschi, 'Matrix values 105 and cooperative behaviour i n the Prisoner's Dilemma game', Journal of Personality and S o c i a l Psychology, 8^, 1968, pp. 148-153; S.S. Komorita, 'Cooperative choices i n a Prisoner's Dilemma game', Journal of Person-a l i t y and S o c i a l Psychology, 2^ , 1965, pp. 741-45; V.E. Bixenstein, N. Chambers, and K.V. Wilson, "Effect of asymmetry i n payoff on behaviour i n a two-person, non-zero-sum game', Journal of C o n f l i c t Resolution, 8_, 1964, pp. 151-59; Bixenstein, Potash and Wilson, 'Effects of l e v e l of cooperative choice by the other player on choices i n a Prisoner's Dilemma Game, Part 1', Journal of Abnormal and S o c i a l Psychology, 66, 1963, pp. 308-13; Bixenstein and Wilson, 'Effects of l e v e l of cooperative choice by the other player on choices i n a Prisoner's Dilemma game, Part 2', same jour n a l , 67, 1963, pp. 139-47; W.E. Vinacke, 'Sex roles i n a three person game', Sociometry, 22, 1959, pp. 343-59; W.E. Vinacke, 'Negotia-tions and decisions i n a p o l i t i c s game', i n B. Lieberman, S o c i a l Choice, Gordon and Breach Science Publishers, 1971, pp. 51-81; Tedeschi, J.T., et a l . , 'Start e f f e c t and response bias i n the Prisoner's Dilemma game', Psychonomic Science, 11, 1968, pp. 149-50; S. Oskamp and D. Perlman, 'Factors a f f e c t i n g cooperation i n a Prisoner's Dilemma game', Journal, of C o n f l i c t Resolution, 9_, pp. 359-74. Gallo and McClintock sum i t up by stating- "A large number of studies have f a i l e d to demonstrate any r e l a t i o n s h i p between sex of the players and choice behaviour," and they l i s t several other studies that lead them to t h i s conclusion. See P.S. Gallo and C.G. McClintock, 'Cooperative and competitive behaviour i n mixed motive games, ' Journal of C o n f l i c t Resolution, 9_, 1965, pp. 68-79. 21. L^E. Fouraker and S. Si e g e l , Bargaining Behaviour, New York: McGraw H i l l Book Co., 1963; C.S. Fischer, 'The e f f e c t s of threats i n an incom-plete information game', Sociometry, 32, 1969, pp. 301-314; H.H. Kelley, 'A classroom study i n the dilemmas of interpersonal negotiations', i n K. Archibald, Ed., Strategic Interaction and C o n f l i c t , University of C a l i f o r n i a Press, 1966, pp. 49-73; R.M. Liebert, et a l . , 'The e f f e c t s of information and magnitude of i n i t i a l o f f e r on interpersonal negotiation', Journal of Experimental S o c i a l Psychology, 4_, 1968, pp. 431-444. For a review of the paradigm most frequently used to study dyadic bar-gaining, see H.H. Kelley and D. Schenitzki, 'Bargaining 1, Chapter 10 i n C.G. McClintock, Ed., Experimental S o c i a l Psychology, Holt, Rinehart and Winston, 1972. 22. For a discussion of the implications of req u i r i n g subjects to reach agree-ment before the experiment-.can terminate, see Kelley and Schenitzki, i b i d . , pp. 306-307. 23. L.L. Cummings and D.L. Harnett, 'Bargaining behaviour i n a symmetric t r i a d : the r o l e of information, communication, power and r i s k - t a k i n g propensity', Review of Economic Studies, 36, 1969, pp. 484-499; and D.L. Harnett and L.L. Cummings, 'Bargaining behaviour i n an assymetric t r i a d ' , Chapter 2.5 i n B. Lieberman, S o c i a l Choice, Gordon and Breach Science Publishers, 1971. 106 CHAPTER 4 RESULTS AND EVALUATION OF RESULTS Introduction In the f i r s t two chapters, a process was outlined i n which P-'sought to maximize value by obtaining an optimum balance of resources which were valued according to a p r i n c i p l e of diminishing marginal u t i l i t y . In f a i r exchanges, he could get the best terms i n a transaction by l o c a t i n g a part-ner who placed r e l a t i v e l y high value on what P could give him (X), and r e l a t i v e l y low value on what P wanted (Y). Beyond seeking terms i n f a i r exchanges which brought him the best possible p r o f i t , P was motivated to gain a d d i t i o n a l p r o f i t through advantageous exchanges, by c o n t r o l l i n g i n -formation required by Other to assess the fairness of a transaction. P's success i n obtaining advantage was seen to be l i m i t e d by a preference on the part of P's potential, partners for r e l i a b l e information about the resources of anyone with whom they might agree to trade. As a r e s u l t , P was expected to learn that witholding information was i n fact a disadvantage. To test the p r e d i c t i o n s , two sets of experiments were developed as described i n Chapter 3. The presentation of r e s u l t s from these experiments w i l l follow the same order as the presentation of hypotheses i n Chapter 2. The evaluation of each hypothesis w i l l include a restatement of the formal hypothesis, followed by s p e c i f i c experimental p r e d i c t i o n s ; the relevant data w i l l then be given, and conclusions drawn on the basis of the r e s u l t s . An argument f o r the operational i n t e r p r e t a t i o n of the variables has been 1 made i n Chapter 3, and w i l l not be repeated here. 107 Many of the predictions from the theory depend on the p r i o r con-d i t i o n that subjects value the resources according to a marginal u t i l i t y function. Because the success of the manipulation to b u i l d i n t h i s value function i s so important to the i n t e r p r e t a t i o n of r e s u l t s , at appropriate points : i n the discussion of the r e s u l t s , we w i l l draw attention to evidence that t h i s condition was s a t i s f i e d . This evidence w i l l be summarized at the end of the chapter. I n i t i a t i o n s to Persons Perceived to be W i l l i n g to Pay the Best Prices Hypothesis 1; Among those members of the group about whose resource bases and value positions P has unambiguous information, P i s most l i k e l y to i n i t i a t e to a person, 0, whom P perceives to have the greatest, amount of resource Y, and the smallest amount of resource X r e l a t i v e to Y. In the experiments, d i f f e r e n t i a l amounts of resources X and Y (red and green buttons) were operationally defined i n two ways: 1) In Set A, some subjects had e i t h e r a t o t a l resource base with 1630 of one colour and 30 of the other (high resource subjects) , or 800';-of one colour and 30 of the other (low resource subjects). 2) In the P i l o t study, a l l subjects had the same t o t a l resource base of 1200 buttons, i n d i f f e r e n t combinations of red and green ( d i f f e r e n t value p o s i t i o n s ) . These resource p r o f i l e s are shown i n Appendix IV. Since the experimental predictions for these two sets are d i f f e r e n t , we w i l l discuss each case separately. Experimental p r e d i c t i o n (Set A): A s i g n i f i c a n t l y larger proportion of the i n i t i a t i o n s made to v i s i b l e s w i l l be directed to high resource 2 v i s i b l e s . 108 On the f i r s t t r i a l , 80/112 i n i t i a t i o n s were directed to v i s i b l e 3 subjects. The d i r e c t i o n of the i n i t i a t i o n s by both v i s i b l e s and non-v i s i b l e s are given i n Table 4.1. The test of the hypothesis involves only the i n i t i a t i o n s sent to v i s i b l e subjects. Table 4.1 Frequencies of I n i t i a t i o n s by Resource Level and V i s i b i l i t y of I n i t i a t o r , and Resource Level of V i s i b l e Other (Set A, T r i a l 1) Recipient of I n i t i a t i o n I n i t i a t o r . High v i s i b l e Low v i s i b l e Non-visible* T o t a l High v i s i b l e 18 3 7 28 Low v i s i b l e 13 8 7 28 High NV 16 3 9 28 Low NV 16 3 9 28 To t a l 63 17 32 112 H q : Of t o t a l i n i t i a t i o n s to v i s i b l e s (n = 80), proportion of i n i t i a t i o n s to high v i s i b l e s = proportion of i n i t i a t i o n s to low v i s i b l e s . Obtained proportion of i n i t i a t i o n s to highs was .79. Binomial t e s t , one t a i l e d , Z = 6.1 (Z . = 1.65, p = .05). Reject H . e n t ' r J o * High and low non-visibles looked the same to i n i t i a t o r s . Table 4.1 shows that a s i g n i f i c a n t l y larger proportion of i n i -t i a t i o n s to v i s i b l e s were directed to the high resource v i s i b l e s as pre-dicted. Only i n the row for low v i s i b l e i n i t i a t o r s i s the difference not s i g n i f i c a n t . Each category of non-visible subject received almost the same number of o f f e r s as the low v i s i b l e s . The fact that some subjects did not i n i t i a t e to the high resource v i s i b l e s as expected may simB'ly'b_'lj;e4'due>to' error, r e s u l t i n g from the f a i l u r e of the subject to understand the function 109 4 expressed i n the value charts. Another p o s s i b i l i t y i s that some subjects may have anticipated the popularity of the high v i s i b l e s , and preferred to i n i t i a t e where the unpopularity of low v i s i b l e s and non-visibles was expected to guarantee a high p r o b a b i l i t y of acceptance. We do not have su i t a b l e data to test either of these i n t e r p r e t a t i o n s . However, the main trend of the r e s u l t s does support Hypothesis 1. Because the derivation of Hypothesis 1 rested on P's assessment of the value to 0 of an increment of a resource r e l a t i v e to O's resource base, the support f o r the f i r s t hypothesis gives i n d i r e c t evidence that the marginal u t i l i t y manipulation was on the whole e f f e c t i v e . Without such a function, there was no reason f o r subjects to think a high resource person would be any more w i l l i n g to trade than a low resource person. In addition, 37% of i n i t i a t i o n s to high v i s i b l e s asked for more than they offered, compared with only 23% of o f f e r s to lows. Experimental p r e d i c t i o n ( P i l o t set) : A s i g n i f i c a n t l y l a rger propor-t i o n of the i n i t i a t i o n s made to v i s i b l e s w i l l be directed to those wi 5 the greatest imbalance of red and green buttons. On T r i a l 1, 84/112 off e r s were directed to v i s i b l e s . Table 4.2 shows the d i r e c t i o n of i n i t i a t i o n s made to both v i s i b l e s and non-v i s i b l e s , but again, the test of the hypothesis involves only the former. 110 Table 4.2. Frequencies of I n i t i a t i o n s by Value P o s i t i o n and V i s i b i l i t y of I n i t i a t o r , and Value P o s i t i o n of Recipient ( T r i a l 1, P i l o t Set) Recipient of I n i t i a t i o n I n i t i a t o r Unbalanced V i s i b l e Balanced V i s i b l e Non-Visible T o t a l Unbalanced v i s i b l e 18 6 4 28 Balanced v i s i b l e 15 4 9 28 Non-visible* 29 12 15 56 Total 63 22 28 112 H q : Of t o t a l i n i t i a t i o n s to v i s i b l e s (n = 84), proportion of i n i t i a t i o n s to unbalanced v i s i b l e s = proportion of i n i t i a t i o n s to balanced v i s i b l e s . Obtained proportion of i n i t i a t i o n s to unbalanced v i s i b l e s was .73. Bino-mial t e s t , one-tailed, Z = 4.8 (Z . = 1.65, p = .05). Reject H . ' ' c r i t o * Non-visibles combined as no differences i n value p o s i t i o n s . The s i g n i f i c a n t l y greater proportion of o f f e r s to v i s i b l e s with a high imbalance of the two colours of buttons extends the support f o r Hypo-thesis 1, to the case where subjects a l l have the same t o t a l resource base, but d i f f e r e n t r e l a t i v e amounts of X and Y. Those o f f e r s which were not made i n the predicted d i r e c t i o n may again i n d i c a t e v a r i a b l e success of the marginal u t i l i t y manipulation, or an a n t i c i p a t i o n that the popularity of the unbalanced v i s i b l e s would make the other subjects very anxious to trade. In addition to the preference shown for the unbalanced v i s i b l e s , the average number of buttons requested per 100 offered v a r i e d with the 6 value p o s i t i o n of the re c i p i e n t and i n i t i a t o r . Subjects who began nearest I l l to the optimum of equal numbers of each colour (those with 850 red and 350 green), made i n i t i a t i o n s that on average offered the fewest buttons of one colour (67) for a return of 100 buttons of the other colour, when they made i n i t i a t i o n s to players with the least balanced resource bases, (those with 1100 green and 100 red). Subjects with a combination of 900 green and 300 red made an average o f f e r of 82 i n return for 100 when they addressed i n i t i a t i o n s -^to the complementary v i s i b l e , who had 1000 red and 250 green. O v e r a l l , the ordering of the average numbereqffered per 100 asked (balanced v i s i b l e <^  n on-visible <^  unbalanced v i s i b l e ) was consistent with the assump-ti o n that equal amounts of buttons were valued d i f f e r e n t l y depending on the amounts of each colour the i n i t i a t o r possessed. In the comments on the questionnaire, players seemed to a n t i c i p a t e that subjects close to a balance of colours would have higher p r i c e s . This was a reasonable expec-t a t i o n i n f a c t , since the players with more balanced resources acted as i f they did indeed need more buttons to balance the cost to them of giving 7 some away. (See Appendix IV for other r e s u l t s from the P i l o t set.) Advantageous Exchanges To gain more than would be possible on f a i r transactions, a non-v i s i b l e could o f f e r unfair terms, i h the hope that the r e c i p i e n t would as-sume they were f a i r . A f a i r exchange was defined i n Chapter 3 as one i n which both p a r t i e s gained equal p r o f i t s i n value u n i t s . P obtained an ad-vantageous exchange i f his p r o f i t exceeded O's, and the reverse case gave a generous exchange to 0. The means by which subjects could use the value 112 chart, together with information about the s i z e of resource bases, to c a l -culate p r o f i t s i n value u n i t s , was described i n Chapter 3. For purposes of presenting * the r e s u l t s , we define an exchange 8 r a t i o as the number of buttons offered per 100 requested. The f a i r ex-change r a t i o t e l l s us how many buttons of one colour P had to o f f e r 0 f or t h e i r p r o f i t s to be equal. A f a i r exchange was equal to a one for one trans-fer of buttons only i f the partners had i d e n t i c a l but complementary resource p r o f i l e s . Hypothesis 2: I f 0 has only ambiguous information about P's resource base, and P has unambiguous information about O's resource base, then P i s more l i k e l y to make an i n i t i a t i o n of exchange that i s advanta-geous to P, than when 0 does not have unambiguous information. F a i r Exchange Ratios f o r Set B At the s t a r t of the game, low non-visibles had a small p i l e of 50 buttons; v i s i b l e s had 100 i n t h e i r small p i l e s , and high non-visibles had 150. The marginal value increase of an increment of 100 buttons on the f i r s t opportunity for exchange would thus give the low non-visibles more reward than the others. At the same time, the low non-visibles' marginal costs were only s l i g h t l y higher than for the v i s i b l e s and non- v i s i b l e s . Thus, to be f a i r i n an exchange with a v i s i b l e , a low non-visible had to give up more than he requested (125 for 100); on the other hand, the f a i r r a t i o between high non-visibles and v i s i b l e s was 90 for 100. Between v i s i -9 ble partners, an exchange of 100 for 100 was f a i r . Any r a t i o with a num-erator smaller than that i n the f a i r r a t i o was advantageous to the i n i t i a t o r . 113 Experimental p r e d i c t i o n : A s i g n i f i c a n t l y greater proportion of non-v i s i b l e s (who present ambiguous information to Other) , w i l l attempt advantageous exchanges than w i l l v i s i b l e subjects.' To see whether t h i s p r e d i c t i o n was true i n the experiment, sub-j e c t s ' exchange r a t i o s were c l a s s i f i e d according to whether they were above (generous), equal to, or below (advantageous) the f a i r r a t i o . The f a i r and generous o f f e r s were grouped together as non-advantageous. Table 4.3 shows 10 the proportion of advantageous i n i t i a t i o n s to v i s i b l e subjects. Table 4.3 Proportions of Advantageous I n i t i a t i o n s to V i s i b l e s by V i s i b l e and Non-visible I n i t i a t o r s ( T r i a l 1, Set B) I n i t i a t o r (n) Nonsyisible .87 (31) V i s i b l e .26 (39) H q : Proportion of advantageous i n i t i a t i o n s made by v i s i b l e s = proportion of advantageous i n i t i a t i o n s made by non-v i s i b l e s . Binomial t e s t , one-t a i l e d , Z = 5.1, (Z . = 1.65, p = .05). Reject H . / c r i t J o On the basis of these data, we would claim strong support f or Hypothesis 2. Because the p o t e n t i a l for having an advantageous o f f e r perceived by the r e c i p i e n t as ' f a i r ' d i f f e r e d f o r the low and high n o n - v i s i b l e s , we w i l l separate the re s u l t s for Sets B l and BII. We have noted that the f a i r exchange r a t i o for a transaction with a v i s i b l e for the low non-visibles and high non-visibles was 125/100, and 90/100, r e s p e c t i v e l y . Did they d i f f e r 114 i n t h e i r tendency to make i n i t i a t i o n s which offered less than the f a i r amount? Table 4.4 gives the r e s u l t s for the high^non-visibles. Table 4.4 Proportions of High Non-visibles and V i s i b l e s Attempting Advantageous Exchanges with V i s i b l e s ( T r i a l \, Set BI) I n i t i a t o r (n) High non-visible .67 (12) V i s i b l e .21 (19) H q : proportion of advantageous i n i t i a t i o n s made by v i s i b l e s = proportion of advantageous i n i t i a t i o n s made by high n o n - v i s i b l e s . Binomial t e s t , one-t a i l e d , Z = 2.55; (Z . = 1.65, p = .05). Reject H . crxt r J o Hypothesis 2 i s supported f o r high n o n - v i s i b l e s . Note, however, that a majority (16/28) of high non-visibles i n i t i a t e d to other n o n - v i s i b l e s . Given that these non-visibles had to ask 100 i n return for an o f f e r of only 90 to a v i s i b l e , for the trade to be f a i r , i t " i s perhaps not s u r p r i s i n g that so many offered to other non-visibles. The p r o b a b i l i t y of having an advantageous, or even a f a i r o f f e r accepted by a v i s i b l e may not have seemed high enough, e s p e c i a l l y i f the v i s i b l e s were expected to o f f e r to one ano-ther even trades of buttons (the f a i r r a t i o i n t h e i r case) . There was no v i s i b l e model for the high non-visibles to imitate which would y i e l d them more p r o f i t than t h e i r partner, and t h i s aspect of the design may have de-l l pressed the rate of i n i t i a t i o n to v i s i b l e s . Nevertheless, those non-v i s i b l e s who did di r e c t o f f e r s to v i s i b l e s lend support to the theory, as th e i r exchange r a t i o s went below the comparatively unattractive f a i r ex-change r a t i o of 90/100. 115 I t i s more d i f f i c u l t to draw conclusions on the basis of these res u l t s f o r low non-visibles (see Table 4.5). Table 4.5 Proportions of Low Non-visibles and V i s i b l e s Attempting Advan-tageous Exchanges with V i s i b l e s s ( T r i a l 1, Set BII) I n i t i a t o r (n) Low non-visible 1.00 (19)* V i s i b l e .30 (20) Ho3? Proportion of advantageous i n i t i a t i o n s made by v i s i b l e s = proportion of advantageous i n i t i a t i o n s made by low n o n - v i s i b l e s . Binomial t e s t , one-t a i l e d , Z = 4.8, (Z . = 1.65, p = .05). Reject H . c r i t o * Includes 10 o f f e r s of 100;-: buttons for 100 buttons. On the one hand, they were i n an i d e a l p o s i t i o n to gain advantage, because the f a i r rate between v i s i b l e s was 100 for 100, which provided a 12 model for a trade that was to the low non-visibles' advantage. On the other hand, i f the low non-visibles showed an average exchange r a t i o below 125/100, one cannot u n c r i t i c a l l y accept t h i s as support for the hypothesis. In a game with 'winners' such as t h i s one, i t was very u n l i k e l y that any players would have a r a t i o of 1.25, even i f they were v i s i b l e . I t seems there may have been a c e i l i n g on the marginal u t i l i t y e f f e c t , with subjects unwilling to act ' f a i r l y ' i f i t was not i n t h e i r own favour. This i s con-s i s t e n t with the p r i n c i p l e that we most d i s l i k e inequity i f i t favours some-one e l s e . 116 T y p i c a l o f f e r s by V i s i b i e s s and Non-Visibles In addition to the p r e d i c t i o n of a greater frequency of advan-tageous i n i t i a t i o n s by the n o n - v i s i b l e s , Hypothesis 2 also implies that the ' t y p i c a l ' or average i n i t i a t i o n made by them would propose terms of ex-change y i e l d i n g more net p r o f i t i n value units to the i n i t i a t o r . The t y p i c a l i n i t i a t i o n of the v i s i b l e s , on the other hand, should have proposed f a i r terms. This measure w i l l give us an i n d i c a t i o n of the degree to which non-visibles departed from f a i r n e s s . Table 4.6 shows the difference between the number that should have been offe r e d i n return f o r 100 buttons on a 13 f a i r trade, and the number a c t u a l l y offered. (The data for v i s i b l e s i n the two sub setss are combined.) Table 4.6 F a i r Exchange Ratios and Obtained Average Exchange Ratios for I n i t i a t i o n s to V i s i b l e s , b y - A l l Subjects ( T r i a l 1, Set B) Mean Number I n t e r q u a r t i l e F a i r Difference I n i t i a t o r (n) offered/100 Range Number (Obtained-f a i r ) High NV (12) 76 (50-100) 90 -14 V i s i b l e (39) 96 (83-100) 100 - 4 Low NV (19) 85 (67-100) 125 -40 The largest deviations from fairness were made by the n o n - v i s i b l e s , and t h e i r o f f e r s showed more v a r i a b i l i t y than those made by v i s i b l e s . These figures cannot p'roperlyrbe given a r a t i o or i n t e r v a l i n t e r p r e t a t i o n , but they provide an i n d i c a t i o n of the magnitude of the deviations involved. The low average exchange r a t i o of the low non-visibles seems consistent with the argument many were attempting to take advantage of the covers, and 117 making of f e r s that would be assumed to be f a i r by the r e c i p i e n t . However, i t i s not possible to know whether the low non-visibles would have offered more than they requested, had t h e i r resources been v i s i b l e , when to do so 14 would have made winning i n the game as a whole impossible. Subjects could not t e l l what constituted a f a i r exchange i f they i n i t i a t e d to a non-visible. Table 4.7 shows that non-visibles were on average more generous i n t h e i r i n i t i a t i o n s to other n o n - v i s i b l e s , while the exchange r a t i o s of the v i s i b l e s were e s s e n t i a l l y the same as when they i n i t i a t e d to v i s i b l e s (94/100 to non-v i s i b l e s , compared with --96/100 to v i s i b l e s ) . Table 4.7 Average Exchange Ratios for I n i t i a t i o n s to Non-Visibles ( T r i a l 1, Set B) Mean Number I n t e r q u a r t i l e I n i t i a t o r (n) Offered/100 Range  High NV (16) 99 ( 91-100) V i s i b l e (17) 94 (100-105) Low NV ( 9) 91 ( 74-100) Note: Results from Set A relevant to Hypothesis 2 are given i n Table A.6, 15 Appendix V. Because of the ambiguity of i n t e r p r e t a t i o n f o r the i n i t i a t i o n s from low non-visibles noted above ( i . e . , a l l of them sent t e c h n i c a l l y ad-vantageous o f f e r s ) , i t i s of i n t e r e s t to know whether they perceived there to be an advantage i n having covers over t h e i r buttons. 118 Perceived Advantage i n Concealing Resources Hypothesis 3: In the early stages of i n t e r a c t i o n , persons who have the opportunity to conceal t h e i r resource bases and value positions w i l l perceive such concealment to be an advantage. Experimental p r e d i c t i o n s : We make three experimental predictions f o r Hypothesis 3: 1) More non-visibles w i l l state on questionnaires on T r i a l 1 that they perceive there to be an advantage i n having a cover over t h e i r buttons than w i l l perceive covers to be a disadvantage. 2) Non-visibles w i l l choose to r e t a i n the covers i f given the oppor-tunity to remove them. 3) Non-visibles w i l l make estimates of the p r o b a b i l i t y of acceptance for t h e i r i n i t i a t i o n s that are not lower than estimates made by v i s i -b l e s . This w i l l indicate that they do not think the covers w i l l i n t e r -fere with t h e i r a b i l i t y to make exchanges. In a sense, Hypothesis 3 i s a checkson the scope conditions of the theory. In addition i t does make the p r e d i c t i o n that non-visibles w i l l synthesize t h e i r desire to maximise again, with the other's i n a b i l i t y to guage no n - v i s i b l e s ' p r i c e s , and a r r i v e at the conclusion that concealment could be used to advantage. Hypothesis 3 i s not ' b u i l t i n ' to the experiment i n the same way that the scope conditions and assumptions are. Subjects could j u s t as e a s i l y have perceived the covers as a hindrance (as some d i d ) . 1. Questionnaire responses describing advantages of covers when  subjects had differences i n resource base (Set A): In these experiments, only 48 of the 56 subjects were asked i f they considered the covers to be an advantage or disadvantage. Of these, 63% (or 30/48) said e i t h e r that the 119 covers were an advantage, or that they were more of an advantage than a disadvantage. This proportion i s s i g n i f i c a n t l y greater than 0.5 (Z = 3.22). Comments on the questionnaire mentioned advantages such as: "I can manipulate others by the s i z e of my o f f e r s ; no one can see that I am hoarding up or getting ahead;' other non-visibles are easy to trade with as they get fewer o f f e r s ; I can o f f e r less f or more; I can compete better i f the other guy does not know my p r o f i t s . " A s u b s t a n t i a l minority saw the covers to be a disadvantage. Some of the reasons c i t e d were: "I have to make good o f f e r s to get accepted; others are a f r a i d of covers; people prefer to deal with those they know; others don't trus t players with covers; others are attracted to a large p i l e . " Of the 30 subjects who saw the covers as an advantage, 47% (or 14) made i n i t i a t i o n s asking for more buttons than they offered. In con-t r a s t , only 11% (2/18) of the subjects who saw the covers as a disadvantage made such o f f e r s . I t i s not possible to decide from the data whether subjects who claimed advantage i n covered resources M l e d to see the disadvantages, or did not f e e l they were large enough to outweigh the advantages. When v i s i b l e s had same resource p r o f i l e s (Set B): Subjects i n these experiments were asked to describe both advantages and disadvantages of covers, so that i t i s not possible to compare frequencies of subjects who 17 saw mainly one or the other. However, we can summarize the most frequent comments made concerning the covers. (These are l i s t e d i n Tables A.7 and A.8 i n Appendix V.) 120 Although i t was d i f f i c u l t to make an unambiguous c l a s s i f i c a t i o n of the responses to the open-ended question about the covers, there did appear to be two main advantages perceived: a) Others were unable to assess how 'hard up' the non-visibles were. Others were therefore expected to ask for l e s s , and be forced into giving more to a non-visible than they should i n a f a i r trade. (Subjects used words, l i k e b l u f f i n g , f o o l i n g others, playing hard to get, f o r c i n g up bids.) This advantage focussed on the comparison of p r o f i t s on a par-t i c u l a r exchange. b) Others were unable to t e l l how non-visibles stood i n terms of progress toward balance. (Subjects referred to t h e i r rank i n the game, and to the fact that Others could not see a non-visible balancing, and could not stop him.) While both of these reasons are re l a t e d to the s i z e of the r e -source base, some subjects seemed to take a view of the whole experiment i n assessing who they could 'beat' (reason b). This focused on end-game as-pects of the s i t u a t i o n , and to a c e r t a i n extent, worked i n a d i f f e r e n t way than the advantages c i t e d i n reason a. That i s , over the e n t i r e game, low non-visibles could a f f o r d to give up l e a s t , being furthest behind i n t o t a l resources, and high non-visibles could a f f o r d more, having a lead i n t o t a l resources of 200 on t h e i r nearest v i s i b l e opponent. (Only 5/56 non-v i s i b l e s l i s t e d no advantages.) Of the non-visibles l i s t i n g disadvantages i n covers (approximately 75%) a great many stated that others would 'fear the unknown' and be r e l u c -tant to trade with non-visibles. The preference to r e t a i n the covers (Table 121 4.8) would then seem to ind i c a t e that the advantages outweighed the d i s -advantages, at l e a s t at t h i s point i n the game. V i s i b l e s ' views of having t h e i r buttons out i n the open comple-mented those of non-visibles. They focused on the negative aspect that others could take advantage of knowing the v i s i b l e s ' need, and that being v i s i b l e prevented them from b l u f f i n g , or 'getting away' with a good deal. The-; most frequently c i t e d advantage for the v i s i b l e s was that they were ensured of i n i t i a t i o n s from others. In sum, the predicted advantage of covers was r e f l e c t e d i n comments of both v i s i b l e and non-visible subjects. This gives us further confidence that p r o f i t was being assessed r e l a t i v e to resource base. Comments to the e f f e c t that Others could not assess a f a i r p r i c e with a n o n - v i s i b l e , and reference to the a b i l i t y of those with a l o t of buttons to pay more, would not make sense outside the context of a marginal u t i l i t y function. 2. Preference for r e t a i n i n g the covers: The next measure r e l e -vant to the t e s t of Hypothesis 3 i s the response to a question asking each non-visible subject whether he would l i k e to have the cover removed from 18 his buttons on T r i a l 3. If the covers were not seen as an advantage, one would expect a large proportion of subjects to want the covers removed. Table 4.8 shows that t h i s was not the case. V i s i b l e s were asked i f they would l i k e the covers removed from a l l the n o n - v i s i b l e s . 122 Table 4.8 Proportion of V i s i b l e s and ^Non-visibles who Preferred to Re-t a i n Covers on Resource P i l e s of Non-visibles ( T r i a l 1, Set B) Subject Proportion Wanting Covers L e f t On (n) High NV .93 (28) Low NV .79 (28) V i s i b l e .41 (56) Q@ = 25.26, d.f. = 2,- p < .001) It should be noted that the r e l a t i v e l y lower percentage of low non-visibles who wanted to r e t a i n t h e i r covers (79% versus 93% for high non-visibles) i s not consistent with the view that the lows were i n the best p o s i t i o n to have advantageous o f f e r s 'read as f a i r ' . However, the 19 proportion i n favour of covers is s t i l l very high. The r e s u l t s from Set A were very s i m i l a r to those above, except that more non-visibles wanted the covers o f f . On T r i a l 1, 8/28 of the high n o n - v i s i b l e s , and 9/28 of the low non-visibles said they would choose to have t h e i r covers removed i f given the opportunity. At that time, no 20 o f f e r s had been delivered to anyone. 3. Perceived l i k e l i h o o d of acceptance: Although many non-v i s i b l e s were aware that v i s i b l e s would be considered more desirable ex-change partners, the non-visibles did not seem to b e l i e v e they would f a i l i n t h e i r i n i t i a t i o n s . On the f i r s t questionnaire, subjects indicated on a f i v e point s c a l e , from 'extremely l i k e l y ' to 'not at a l l l i k e l y ' , the per-21 ceived p r o b a b i l i t y of acceptance of the o f f e r they had j u s t made. Table 4.9 shows the proportions of subjects who checked each of the f i v e categories. 123 Table 4.8 ^ I n i t i a t o r s ' Evaluation of Likelihood that Their Offers Would be Accepted ( T r i a l 1, Set B) Likelihood of Offer A l l Types of Offers Advantageous Of f e r s * * Being Accepted V i s i b l e * High NV Low NV V i s i b l e High NV Low NV Extremely l i k e l y .07 .07 .07 .07 Or : .09 Quite l i k e l y .52 .61 •57 .23 .69 .50 50-50 chance .39 .29 .32 .62 .23 .36 Not very l i k e l y .02 .04 .04 .08 .08 .05 Not at a l l ' 1 i \ Not at a l l l i k e l y 0 0 0 0 Or' ' 0 Total (n) 56 28 28 15 13 22 Median category 2 2 2 3 2 2 * Visibles.S' from sets BI and BII are combined. ** Advantageous i n i t i a t i o n s include those defined as advantageous to v i s i b l e s , and i n i t i a t i o n s to non-visibles which simply ask for more than they offered. We can see i n the table above that non-visibles were j u s t as o p t i m i s t i c as the v i s i b l e s that t h e i r o f f e r s would be accepted. Non-v i s i b l e s also estimated higher p r o b a b i l i t i e s of acceptance for i n i t i a t i o n s which asked f o r more than they offered (median category 2, versus 3 for v i s i b l e s ) . There were few differences by target of the i n i t i a t i o n (not included i n t a b l e ) , except that the high non-visibles tended to give higher p r o b a b i l i t i e s of acceptance by a v i s i b l e target, than was estimated by other i n i t i a t o r s . We do not have for comparison the subjects' estimates for targets they did not choose. P i l o t work showed that subjects were un-w i l l i n g or unable to make serious estimates of the l i k e l i h o o d of acceptance by an Other, unless they chose to i n i t i a t e to him. O v e r a l l , the evidence from the two sets of experiments i s con-s i s t e n t with Hypothesis 3. Although i t could be claimed that f o r c i n g the 124 subjects to have covers did not mean they wanted them, t h e i r comments, and 22 desire to r e t a i n the covers, gives l i t t l e support for such a claim. Preference f o r Information about Exchange Partner While the non-visibles did not a n t i c i p a t e a smaller chance of having t h e i r i n i t i a t i o n s accepted, the theory predicts that there would be a preference by a l l subjects for v i s i b l e s . Hypothesis 4: Members of a group are more l i k e l y to d i r e c t t h e i r i n i t i a t i o n s of exchange to persons whose resource bases and value positions are known through unambiguous information. Experimental p r e d i c t i o n s : The op e r a t i o n a l i z a t i o n of the fourth hypo-thesis simply implies that v i s i b l e subjects w i l l receive a s i g n i f i c a n t l y greater proportion of i n i t i a t i o n s of exchange than w i l l the non-23 v i s i b l e s . Results when v i s i b l e s a l l have same sized resource base (Set B): In Set B, a l l v i s i b l e subjects had the same resource bases and value p o s i -t i o n s , two with predominantly red p i l e s , and two with a predominance of green buttons. The proportion of i n i t i a t i o n s to v i s i b l e s i n these experi-ments, e s p e c i a l l y those made by other v i s i b l e s , allows a test of Hypothe-24 s i s 4 unconfounded by a preference for high resource players. The pro-portions of o f f e r s to v i s i b l e and non-visible subjects are shown i n Table 4.10. 125 Table 4.10 • Proportions of I n i t i a t i o n s to V i s i b l e s ( T r i a l 1, Set B) SET BI I n i t i a t o r ( n) V i s i b l e .68 (28) High NV .43 (28) 1H q: Proportion of i n i t i a t i o n s received by v i s i b l e s = proportion of i n i -t i a t i o n s received by high n o n - v i s i b l e s . Binomial t e s t , one-tailed: for a l l i n i t i a t o r s , Z = .9, (Z . =1.65, p = .05). Retain H . For i n i t i a -c r i t o tions by v i s i b l e s , Z = 1.89, p <^  .05; for high n o n - v i s i b l e s , Z = .56. n.s. SET BII I n i t i a t o r (n) V i s i b l e .71 (28),,. Non-visible .68 (28X.. 2H q: Proportion of i n i t i a t i o n s received by v i s i b l e s = proportion of i n i -t i a t i o n s received by low non-visibles. Binomial t e s t , one-tailed: f o r a l l i n i t i a t o r s , Z = 2.9, (Z . =1.65). Reject 2H . c r i t o Hypothesis 4 i s supported for a l l v i s i b l e i n i t i a t o r s , and for low non-visible i n i t i a t o r s . High non- v i s i b l e s , however, did not prefer v i s i b l e s , and showed a non-significant tendency to choose non- v i s i b l e s . I t has been noted that the c l e a r e s t test of Hypothesis w a s provided by the data for v i s i b l e i n i t i a t o r s . ;TThe s l i g h t v a r i a t i o n s i n the resource bases and value positions of the non-visibles leads us to sus-pect that the experimental design was at least p a r t l y responsible f o r the negative results for the high n o n - v i s i b l e s . 126 R e c a l l that the high non-visibles had a small resource p i l e of 150 buttons. In a one for one trade of buttons (which was the f a i r p r i c e i n a trade between two v i s i b l e s ) , the high non-visibles received a lower p r o f i t than t h e i r partners, due to t h e i r larger marginal reward. In terms of the choice model, the value of an o f f e r acceptable to a v i s i b l e would have had lower expected p r o f i t for the non-visible than he would l i k e ; but for the non-visible to get a reasonable p r o f i t , he would have had to ask for more than he offered, and r i s k a higher p r o b a b i l i t y of r e j e c t i o n . E i t h e r way, the expected value of such an exchange would compare poorly with the uncertain non-visible partners, who were not known c e r t a i n l y to 25 be undesirable. Although unexpected, the r e s u l t s f or the i n i t i a t i o n s of these non-visibles lend support to the claim that subjects calculated pro-f i t s r e l a t i v e to t h e i r resource bases. Low non-visibles i n these experiments were i n a d i f f e r e n t p o s i t i o n due to t h e i r very small non-predominant resource p i l e . From the point of view of the theory, they were i n an i d e a l p o s i t i o n to gain advantage by trading with v i s i b l e s . A one for one swap of buttons with a v i s i b l e gave the low non-visible a higher p r o f i t than h i s partner, and yet would seem l i k e l y to be accepted, because i t would compare well with o f f e r s between v i s i b l e s . With this a t t r a c t i v e p o t e n t i a l transaction a v a i l a b l e , the low non-visibles would have l i t t l e reason to r i s k the uncertainty of a trade with a n o n - v i s i b l e . The preference of v i s i b l e s for other v i s i b l e s i s encouraging support for the theory, because, i n t h e i r case, v i s i b l e and non-visible 127 targets were most l i k e l y to be considered equivalent, except for the am-b i g u i t y of information a v a i l a b l e about t h e i r resources. (The v i s i b l e s were t o l d some v i s i b l e s had more, and some l e s s , but there was no way of t e l l i n g which players had which.) O v e r a l l , i n experiments with this paradigm (Sets A and B, the 26 P i l o t work, and more recent work conducted i n A u s t r a l i a ) , i t has been found that the proportions of i n i t i a t i o n s to v i s i b l e s on the i n i t i a l ex-change opportunity f a l l c o nsistently between 68% and 75%, with the sole exception of the high non-visibles i n Set B l . I n i t i a t i o n s to V i s i b l e s when they have D i f f e r e n t Sized Resource Bases The r e s u l t s from Set A relevant to Hypothesis 4 were very s i m i -l a r to those from Set B. Although preference for v i s i b l e s , was confounded with the preference for high resource persons, we give the r e s u l t s from Set A here, to show that the v i s i b l e i n i t i a t o r s maintained a preference for v i s i b l e s from the f i r s t to the second t r i a l , even though they tended to change from partners of one resource l e v e l to another. The gross categories of i n i t i a t i o n s on T r i a l s 1 and 2 are given i n Table 4.11 below. Table 4.11 Proportions of Offers to V i s i b l e s Made by V i s i b l e s and Non-v i s i b l e s ( T r i a l s 1 and 2, Set A) I n i t i a t o r T r i a l 1 (n) T r i a l 2 (n) V i s i b l e .75 ( 56) .75 ( 56) Non-visible .68 ( 56) .45 ( 57)* Total .71 (112) .60 (111) 128 Table 4.11 (Continued) * One non-visible made no o f f e r on T r i a l 2. H q : Proportion of i n i t i a t i o n s to v i s i b l e s = proportions of i n i t i a t i o n s to non-visibles. One-tailed test for difference of proportions, T r i a l 1: Z = 4.54, (Z . = 1.65, p = .05). Reject H . crxt - J o The i n i t i a t i o n s made on the second t r i a l were not independent of the o f f e r s and acceptances on T r i a l 1. I t i s c l e a r , however, that while the v i s i b l e s continued to d i r e c t the majority of t h e i r o f f e r s to the v i s i -28 ble s , non-visibles made a change to other n o n - v i s i b l e s . We noted i n the discussion of Hypothesis 1 that the majority of i n i t i a t i o n s to v i s i b l e s on T r i a l 1 i n Set A were sent to high resource v i s i -bles. Due to the rule that each person could accept only one o f f e r / t r i a l , some subjects were bound to have t h e i r o f f e r s turned down. Though these data are more relevant to a consideration of Hypothesis 5, we w i l l show the pattern of acceptances on T r i a l 1, as they help to explain the d i r e c -t i o n of T r i a l 2 i n i t i a t i o n s . Table 4.12 shows that i n i t i a t i o n s made by v i s i b l e s had a higher rate of acceptance. Table 4.12 Acceptance of Real Offers ( T r i a l 1, Set A) Recipient (Proportion of I n i t i a -tions Accepted of Number Received) I n i t i a t o r il V 62. " T o t a l V i s i b l e .62 (26/42) .86 (12/14) .68 (38/56) Non-visible .26 (10/38) .77 (14/18) .43 (24/56) 129 Table 4.12 (Continued) l H o : Proportions of i n i t i a t i o n s from v i s i b l e s accepted by v i s i b l e s = proportion of i n i t i a t i o n s from non-visibles accepted by v i s i b l e s . One-t a i l e d test for difference of proportions, Z = 3.3, (Z . = 1.65, p = .05) c r i t Reject IH . J o 2H Q: Proportions of i n i t i a t i o n s from v i s i b l e s accepted by non-visibles = proportions of i n i t i a t i o n s from non-visibles accepted by non - v i s i b l e s . One-t a i l e d test f o r difference of proportions gives non-significant di ff er en ce . How did the more frequent reijectionn of non - v i s i b l e s ' i n i t i a t i o n s a f f e c t t h e i r next offers? Table 4.13 shows the d i f f e r e n t pattern of i n i -t i a t i o n s on T r i a l 2 i n more d e t a i l . Table 4.13 Frequency of I n i t i a t i o n s to V i s i b l e s and Non-visibles, by V i s i b l e and Non-visible I n i t i a t o r s ( T r i a l 2, Set A) Recipient  I n i t i a t o r High V i s i b l e Low V i s i b l e Non-visible* High v i s i b l e 8 12 8 Low v i s i b l e 15 7 6 High NV 6 7 15 Low NV** _8 4 15 Tota l 37 30 44 * Non-visible r e c i p i e n t s could not be distinguished i n terms of resource l e v e l . ** One low non-visible made no o f f e r on T r i a l 2. 130 The fact that the v i s i b l e s did not r e s t r i c t themselves to exchange within t h e i r own resource l e v e l s by T r i a l 2 would i n d i c a t e that v i s i b i l i t y was a more important factor to them than resource l e v e l . Of those v i s i b l e s whose o f f e r s to v i s i b l e s were accepted on T r i a l 1, and who continued to i n i t i a t e to v i s i b l e s , twice as many switched to a d i f f e r e n t v i s i b l e , as stayed with the same person. Most of those who were rejected by v i s i b l e s on T r i a l 1, but s t i l l continued to i n i t i a t e to a v i s i b l e , changed to the other v i s i b l e . The data from the f i r s t t r i a l , and for the v i s i b l e s on T r i a l 2, support Hypothesis 4. The f a i l u r e of non-visibles to i n i t i a t e as predic-ted to v i s i b l e s on the second t r i a l seemed to be due to t h e i r f a i l u r e to complete transactions with v i s i b l e s on t r i a l 1. We w i l l return to t h i s point below. Did the preference to accept i n i t i a t i o n s from v i s i b l e s hold when a subject had two i d e n t i c a l i n i t i a t i o n s , one from a v i s i b l e , and one from a non-visible source? Hypothesis 5: If a member has a choice of i n i t i a t i o n s from two others, P^ and P2> and the i n i t i a t i o n s are i d e n t i c a l with respect to the abso-lut e amount of resources offered and requested, then i f the r e c i p i e n t has unambiguous information about P^'s resource base, and ambiguous information about P2*s resource base, he w i l l be more l i k e l y to ac-cept the i n i t i a t i o n from P^. Experimental p r e d i c t i o n : I f Hypothesis 5 i s v a l i d , then a s i g n i f i -cantly greater proportion of the f a l s e o f f e r s which are believed to come from v i s i b l e s w i l l be accepted, i n preference to the i d e n t i c a l o f f e r which the r e c i p i e n t thinks has come from a n o n - v i s i b l e . 131 Acceptance of of f e r s when s i z e and source of o f f e r c o n t r o l l e d : (Set B): In Set B l , only non-visible subjects received two bogus i n i t i a -29 tions on T r i a l 1. In Set BII, a l l subjects received two f a l s e o f f e r s . For c l a r i t y of discussion, the two subjects w i l l . b e dealt with separately. A l l f a l s e i n i t i a t i o n s asked for 110 buttons, i n return f o r 100 offered. Table 4.14 shows the proportions of o f f e r s from v i s i b l e s and non-visibles accepted by the high n o n - v i s i b l e s . To c a l c u l a t e s i g n i f i c a n c e , subjects who rejected both o f f e r s were divided equally between the v i s i b l e and non-visible categories, as there was no evidence of a preference be-tween the two i n the case of double r e j e c t i o n s . i Table 4.14 Acceptance of False Offers by High Non-visibles (Set Bl) Offer Accepted From Proportion Accepted V i s i b l e .46 Non-visible .39 Neither .15 Tot a l (n) (28) H Q: Proportion accepting v i s i b l e s = proportion accepting non-visibles. Binomial t e s t , Z = .19, (Z • . = 1.65, p = .05) . Retain H . ' c r i t o The negative r e s u l t s for the high non-visibles could be seen i n the same l i g h t as the data f or Hypothesis 4, i n which a majority of high non-visibles i n i t i a t e d to non-visibles. To obtain a f a i r deal with^a v i s i -b l e , non-visibles had to receive more from the v i s i b l e than they gave up, 132 and the f a l s e i n i t i a t i o n s asked them to do the reverse of t h i s . 6/11 high non-visibles who rejected v i s i b l e s r eferred to a desire not to l e t the v i s i b l e s have an advantage. However, the fact that the proportions of acceptance of both v i s i b l e and non-visibles i s so s i m i l a r , does not allow a c l e a r i n t e r p r e t a t i o n . In Table 4.15, we see that the pattern of acceptances of the f a l s e i n i t i a t i o n s by low non-visible r e c i p i e n t s was d i f f e r e n t . Again, the double rejections were treated as giving no information for or against 30 the hypothesis. Table 4.15 Acceptance of False Offers by V i s i b l e s and Low Non-visibles (Set BII) Proportion Accepted By Offer Accepted From V i s i b l e Low Non-visible V i s i b l e .54 .61 Low non-visible .18 . .14 Neither .28 .25 1H Q: Proportion of v i s i b l e s accepting v i s i b l e s = proportion accepting non-v i s i b l e s . 2H q: Proportion of low non-visibles accepting v i s i b l e s = proportion accept-ing non-visibles. One-tailed binomial t e s t ; for 1H , Z = 2.05; for 2H , o o' Z = 2.17, (^ C R^ T = 1-65, p = .05). Reject both n u l l hypotheses. The high rates of double rejections shown i n Table 4.15 gives an i n d i c a t i o n that there i s a l i m i t to the assumption that P w i l l enter into an exchange so long as he makes a p o s i t i v e p r o f i t . (In t h i s case, the l i m i t arises both from the a n t i c i p a t i o n that there w i l l be winners, and that i t i s 133 not necessary to s e t t l e for transactions that offered fewer buttons than were requested.) Nevertheless, a majority of subjects did accept one of the f a l s e i n i t i a t i o n s . Because an o f f e r of 100 for 110 from a v i s i b l e was reasonable i n terms of the f a i r exchange r a t i o between v i s i b l e s and low non-vi s i b l e s , the l a t t e r had no reason to r i s k accepting an o f f e r from a non-v i s i b l e source, which might be giving them less p r o f i t than t h e i r partner. For these subjects, the proportiohof o f f e r s accepted from v i s i b l e s was the highest i n Set B. Acceptance of of f e r s when s i z e and source of i n i t i a t i o n not con- t r o l l e d : (Set A and Set B l ) ; Real o f f e r s were given to the intended r e -ci p i e n t i n Set A on T r i a l 1, and to the v i s i b l e s i n Set B l . Hypothesis 2 claimed that non-visibles would make more o f f e r s that asked f o r greater p r o f i t to themselves; i n conjunction with the preference fo r v i s i b l e s , i t was expected that both v i s i b l e and non - v i s i b l e subjects would accept more r e a l o f f e r s from v i s i b l e i n i t i a t o r s than from non - v i s i b l e s . 1. In Set A, where v i s i b l e subjects received 71% of i n i t i a t i o n s , a s i g n i f i c a n t l y l a rger proportion of v i s i b l e i n i t i a t o r s were accepted by 31 v i s i b l e s , but not by no n - v i s i b l e s , who accepted most of the o f f e r s they received (26/32). This was shown previously i n Table 4.12. The pattern of acceptance i s shown i n more d e t a i l i n the table below. 134 Table 4.16 Number of Real Offers Accepted Per Number Received by V i s i -b i l i t y and Resource Level of Recipient ( T r i a l 1, Set A) Recipient  Offer Accepted From High v i s i b l e Low v i s i b l e High NV Low NV High v i s i b l e 11/16 2/3 3/3 3/4 Low v i s i b l e 7/13 6/8 3/3 3/4 Non-visible 7/32 3/6 4/5 10/13 T o t a l 25/63 11/17 10/11 , 16/21 In general, i t seemed that only i f one had a range of choice, did 32 the preference for v i s i b l e s exhibit i t s e l f . This i s consistent with the choice model presented i n Chapter 2, i n which i t was claimed that P would attend to O's p r o f i t only i f he could not decide between two or more a l t e r -33 natives on the basis of p r o f i t to himself. Non-visible subjects accepted 87% of the o f f e r s they got; of the 6 o f f e r s they rejected, 5 cases involved the receipt of two off e r s at once, and i n one case, the non-visible was asked for greatly more than he was offered. Over a l l subjects who received only one o f f e r on the f i r s t exchange opportunity, 91% accepted that o f f e r . 2. V i s i b l e s i n S e t B I also received i n i t i a t i o n s as they were t r u l y directed. I t can be seen i n Table 4.17 that they accepted a larger proportion of i n i t i a t i o n s from v i s i b l e s . Table 4.17 Acceptance of Real Offers to V i s i b l e s ( T r i a l 1, Set BI) I n i t i a t o r Number of Offers Received Proportion Accepted V i s i b l e 19 .63 Non-visible 12 .41 135 Table 4.17 (Continued) Ho-. Proportion accepting v i s i b l e s = proportion accepting n o n - v i s i b l e s . One-tailed t e s t for difference of proportions: Z = 1.22, (Z . =1.65, r f 5 c r i t p = .05) . Retain H . r o The higher rate of acceptance of non-visibles i n t h i s set r e -f l e c t s the smaller number of o f f e r s they had to choose from, as a r e s u l t of the tendency for non-visibles to i n i t i a t e to other n o n - v i s i b l e s . The proportion of o f f e r s from v i s i b l e s that were accepted i s s i m i l a r to that i n Set A. Here again, the acceptances seemed to be a function of type of o f f e r , preference for v i s i b i l i t y , and the number of a l t e r n a t i v e o f f e r s from which the r e c i p i e n t could choose. For the seven rej e c t i o n s of non-v i s i b l e s , the v i s i b l e always had another o f f e r from a v i s i b l e , and usually i t was a better one i n p r o f i t to the r e c i p i e n t . Acceptance of c o n t r o l l e d o f f e r s a f t e r one set of transactions has  been completed:(Set A): False o f f e r s were dire c t e d to v i s i b l e subjects i n Set A on the second t r i a l . The pattern of acceptances i s shown i n Table 4.18.-Table 4.18 Acceptance of f a l s e Offers by V i s i b l e s ( T r i a l 2, Set A) Recipient I n i t i a t o r High v i s i b l e Low v i s i b l e Total V i s i b l e .64 (18/28) .60 (17/28) (35/56) Non-visible .36 (10.28) .36 (10/28) (20/56) Reject both .00 ( 0/28) .04 ( 1/28) ( 1/56) 136 H q: Proportion accepting v i s i b l e = proportion accepting n o n - v i s i b l e . One-t a i l e d binomial t e s t for differences of proportions gives Z = 1.87, (Z . = 1.65, p = .05). Reject H . c r i t ' v J o The table shows that when the r e c i p i e n t could choose between iden-t i c a l o f f e r s from a v i s i b l e and a n o n - v i s i b l e , the proportion of o f f e r s accepted from v i s i b l e s was again s i g n i f i c a n t l y greater than the proportion accepted from non-visibles. Since the f a l s e o f f e r s c o n t r o l l e d for the greater tendency for the non-visibles to make o f f e r s more favourable to the i n i t i a t o r , t h i s i s encouraging support for Hypothesis 5. I t must be kept i n mind, however, that the T r i a l 2 acceptances, l i k e the T r i a l 2 r e a l offers-?,;, would have been aff e c t e d by the pattern of o f f e r s and acceptances on the f i r s t opportunity for exchange. For example, v i s i b l e s who were accepted on T r i a l 1 by the same v i s i b l e who supposedly 'offer s ' to them on the second t r i a l , may have accepted that person as an act of r e c i p r o c i t y . A t o t a l of 11/35 acceptance on T r i a l 2 involved t h i s sort of r e c i p r o c i t y . Of the subjects who accepted non-visibles on the second t r i a l (20 i n a l l ) , only one case involved acceptance of someone who had accepted the subject on T r i a l 1. There was no observable tendency for subjects who ac-cepted non-visibles on T r i a l 2 to be those who were rejected by a v i s i b l e on the f i r s t t r i a l , and no clear tendency for the people who accepted non-v i s i b l e s to be those who had offered to a non-visible previously (12/20 had i n i t i a t e d to non-visibles on T r i a l 1, T r i a l 2, or both.) 137 In sum, over the two sets of experiments, Hypothesis 5 receives support, except i n the case of the high non-visibles i n Set BI'. Non-visibles' Reaction to the Preference for V i s i b l e s The p r e d i c t i o n i n the l a s t hypothesis depends on the v a l i d i t y of the previous hypotheses — i f non-visibles did not receive many i n i t i a -tions , and had few accepted, they should have reached a point when they wanted the covers removed. Hypothesis 6; P's perception of the advantage of witholding informa-t i o n about h i s resource base and value p o s i t i o n w i l l decrease over repeated exchange transactions, and P w i l l be more l i k e l y to choose to reveal unambiguous information about h i s value p o s i t i o n and p r o f i t s , than during the i n i t i a l stages of i n t e r a c t i o n . Experimental p r e d i c t i o n : By the end of T r i a l 2, a s i g n i f i c a n t l y greater proportion of non-visibles w i l l state that they would l i k e the covers removed, than the proportion on T r i a l 1. Results for subjects who completed two exchange transactions: We know that non-visible subjects were not very successful i n Set A — they were unpopular on T r i a l 1, and the experimenter intercepted o f f e r s to them on T r i a l 2. Table 4.19 compares the no n - v i s i b l e s ' wishes about the covers on T r i a l s 1 and 2. Table 4.19 Non-visibles' Preference f o r Removing Covers on F i r s t and Second T r i a l s (Set A) Number Who Want Cover OFF Subject T r i a l 1 T r i a l 2 High non-visible 8/28 14/28 Low non-visible 9/28 14/28 138 Table 4.19 (Continued) H : Proportion of non-visibles wanting covers o f f on T r i a l 1 = proportion wanting covers o f f on T r i a l 2. One-tailed t e s t f o r difference of propor-tion s , Z = 1.54, (Z . = 1.65, p = .05). Retain H (p = .06). c r i t o ^ Although the change i n proportions i s not s t a t i s t i c a l l y s i g n i f i -cant, the proportion wanting the covers removed does increase from the f i r s t to the second t r i a l . An examination of the past exchanges of those subjects wanting the covers o f f showed they were only s l i g h t l y more l i k e l y to have been rejected, or to have received noi.offers, than were those voting to r e t a i n the covers. Two t r i a l s may have been too short a time f o r subjects to learn that the covers were not to t h e i r advantage. Removal of the covers i s , of course, not the only response to f a i l u r e to enter into transactions on the f i r s t opportunity. Although no s p e c i f i c predictions were made concerning the adjustment of subjects to r e j e c t i o n on T r i a l 1, the propensity to repeat an o f f e r to the same sort of target was contingent on the response to the f i r s t i n i t i a t i o n . This i s shown i n the table below. Table 4.20 Proportions of Subjects Making the Same Type of Choice ( V i s i b l e s versus Non-visibles) on T r i a l s 1 and 2 (Set A) Response to I n i t i a t i o n on T r i a l 1  Accepted Rej ected To V i s i b l e on T r i a l 1 (n) (n) V i s i b l e i n i t i a t o r .92 (26) .56 (16) NV i n i t i a t o r .60 (10) .36 (28) 139 Table 4.20 (Continued) Response to I n i t i a t i o n on T r i a l 1  Accepted Rej ected To NV on T r i a l 1 (n) (n) V i s i b l e i n i t i a t o r .17 (14) .50 ( 2) NV i n i t i a t o r .50 (14) .50 ( 4) * Theanumbers indicate the proportion of subjects repeating a given sort of i n i t i a t i o n . These r e s u l t s show that v i s i b l e subjects did indeed appear to respond to acceptance by repeating the reinforced behaviour, but only when the reinforcement was acceptance by a v i s i b l e . V i s i b l e s who had been ac-cepted by a non-visible on T r i a l 1 did not repeat an o f f e r to a n o n - v i s i b l e : .83 of them switched to a v i s i b l e . They may have been encouraged by t h e i r success with a non-visible. Acceptance of o f f e r s to v i s i b l e s on the f i r s t t r i a l did not have such a pronounced e f f e c t on the non-visibles (only .60 repeated to a v i s i b l e ) . Rejection on the f i r s t t r i a l l e d to more frequent changes from v i s i b l e to non-visible targets, and v i c e versa. In general, i t seemed that the NV subjects were more l i k e l y to respond to f a i l u r e of an i n i t i a -t i o n by changing the v i s i b i l i t y of the target. I t seems p l a u s i b l e that the non-visibles. would generalize about the cause of t h e i r unpopularity, and adopt remedies that r e l a t e d to the covers. V i s i b l e s would not have i n -ferred that r e j e c t i o n was due to anything but how well t h e i r o f f e r s com-pared with others received by the person they had i n i t i a t e d to. 140 Evaluation of the Effectiveness of the Manipulation of the Value Function We have drawn attention to several pieces of i n d i r e c t evidence that subjects i n the experiments assessed p r o f i t i n value u n i t s , i . e . , that they took into account t h e i r own and other's resource bases i n c a l c u -l a t i n g the value of additions and losses of buttons. The main points were that: 1) A majority of i n i t i a t i o n s were made to subjects with r e l a t i v e l y large amounts of the colour desired by the i n i t i a t o r ; 2) exchange r a t i o s i n the P i l o t set varied according to resource p r o f i l e of i n i t i a t o r and r e c i p i e n t ; 3) perceived advantages of the covers were described i n terms of deter-mining prices i n a transaction r e l a t i v e to need, and to a b i l i t y to give up buttons ( i . e . , s i z e of large resource p i l e ) ; 4) high non-visibles i n ,SSet B, whose f a i r exchange r a t i o with v i s i b l e s was below the f a i r r a t i o for two v i s i b l e partners, were the only sub-j e c t s i n a l l the experiments who did not show an i n i t i a l preference for v i s i b l e s . The most p l a u s i b l e explanation for such a r e s u l t i s that non-visibles would have had to ask for much much more than they offered * i n order to get an advantageous exchange. The p r o b a b i l i t y of acceptance of such o f f e r s would probably be anticipated to be low, and consequently, i n i t i a t i o n s to non-visibles may have had higher expected value. In sum, i t seems there i s s u f f i c i e n t evidence that the marginal u t i l i t y manipulation was e f f e c t i v e . There were undoubtedly exceptions, and the most reasonable explanation for data that did not conform to the hypo-141 theses may be that they can be considered as error r e s u l t i n g from the var-i a b l e success i n s a t i s f y i n g the scope conditions about the subjects' value function. Summary Let us b r i e f l y summarize the findings reported i n th i s chapter. The predictions i n the hypotheses were made l a r g e l y i n terms of expected departures from a chance d i s t r i b u t i o n of the types and di r e c t i o n s of i n i -t i a t i o n s made and accepted. Support for an hypothesis has been claimed when the data depart s i g n i f i c a n t l y from a chance r e s u l t , but no predictions were made concerning the strength of the rel a t i o n s h i p s expected, and i t i s recognized that a good deal of v a r i a b i l i t y remains unaccounted f o r . Never-theless, the pattern of results i s generally i n l i n e with p r e d i c t i o n s . The most s t r i k i n g r e s u l t i s the cons i s t e n t l y high proportion of subjects who both i n i t i a t e d , to, and accepted from, players with v i s i b l e re-sources. With the si n g l e exception of high non-visibles i n Set BI, be-tween 2/3 and 3/4 of f i r s t i n i t i a t i o n s were addressed to v i s i b l e s , and approximately 2/3 of of f e r s accepted were from v i s i b l e s , i f a subject r e -ceived two i d e n t i c a l o f f e r s from a v i s i b l e and non-visible source. This preference held whether v i s i b l e subjects a l l had the same size d resource bases (Set B), d i f f e r e n t r a t i o s of resources but equal t o t a l s ( P i l o t set) , or d i f f e r e n t t o t a l resource bases (Set A). I t must be kept i n mind, how-ever, that there was also a consistent minority who we i n f e r had a p r e f e r -ence for n o n - v i s i b l e s , by v i r t u e of t h e i r choice of no n - v i s i b l e s . While a desire for information about the exchange partner appears, as the theory 142 pr e d i c t s , to be an important determinant of i n i t i a l choice of partner, i t i s c l e a r that other factors were operating, such as d i f f e r e n t i a l a b i l i t i e s of the subjects to s t r a t e g i z e , and to ant i c i p a t e where the majority of i n i t i a t i o n s would go i n the group; d i f f e r e n t i a l i n t e r e s t i n comparison with others; and to some extent, d i f f e r e n t i a l f a s c i n a t i o n with the 'mystery' of an uncertain a l t e r n a t i v e . (With this l a t t e r f a c t o r , one suspects that i f the stakes were higher, fewer people would exhibit a preference for mystery.) As i n many s o c i a l s i t u a t i o n s , the experiment was not a perfect abstraction, and contained c o n f l i c t i n g forces which made i t impossible to predict accurately f o r a l l i n d i v i d u a l s which factors w i l l be overriding. Nevertheless, the behaviour and comments of the subjects gave support to the conviction that the theory was addressing an aspect of i n t e r a c t i o n that was meaningful to the subjects. They tended to see the covers i n terms of providing a means of b l u f f i n g , f o r c i n g p r i c e s , and for getting ahead of others. The s t a t i s t i c a l l y s i g n i f i c a n t tendency of non-visibles to make more advantageous i n i t i a t i o n s than v i s i b l e s i s consistent with the r e s u l t s concerning perceived advantage i n covers. A conservative evaluation of these r e s u l t s seems warranted, however, i n l i g h t of: the lower proportion of high non-visibles i n Set B who offered to v i s i b l e s ; the constraints against subjects o f f e r i n g more than they requested because they were i n a game; and the marginal support for Hypothesis 2 i n Set A (see Appendix V). 143 FOOTNOTES FOR CHAPTER 4 1. In Chapter 3, i t was noted that the two sets were not equally good tests of a l l s i x hypotheses. For this reason, we w i l l present data for each hypothesis from the set which bears most d i r e c t l y on that hypothesis. We w i l l discuss relevant data from the other set where i t gives further information, or where the r e s u l t s of one set c o n f l i c t with those of the ether. Some res u l t s from the P i l o t set described i n Appendix IV are given, when the problems of design for that set do not a f f e c t the r e s u l t . 2. Hypothesis 1 refers to i n i t i a t i o n s to v i s i b l e s only; subjects could not see i f non-visibles had a large or small p i l e of buttons. 3. The proportions of i n i t i a t i o n s to high v i s i b l e s on T r i a l 2 w i l l be given with the discussion of Hypothesis 4, as they shed l i g h t on the r e l a t i v e importance of v i s i b i l i t y of partner and the s i z e of h i s resource base. 4. A few subjects noted during debriefing that they had not noticed the two d i f f e r e n t sizes of resource bases. 5. There were two unbalanced v i s i b l e s i n each experiment: one had 1100 green, 100 red; the other 1000 red, 200 green. Balanced v i s i b l e s had 900 green, 300 red; or 850 red, 350 green. 6. Not a l l subjects framed t h e i r o f f e r s i n terms of 100 buttons. Proposed terms ranged fromJ10 for 20, to a hopeful! 100 for 780 i n return (Set A). 7. Note: Results from the P i l o t Set are given here, because Sets A-and B did not provide an opportunity to test whether subjects with an im-balance i n value p o s i t i o n were as popular as those with large resource bases. The P i l o t Set was also the only one i n which i t was possible to observe whether the terms of proposed exchanges varied with the value positions of partners. 8. The rules did not permit subjects to o f f e r more than 100 buttons at a time. However, the numerator of the exchange r a t i o w i l l sometimes ex-ceed 100, and t h i s indicates an i n i t i a t i o n o f f e r i n g more buttons than were requested, standardized on the denominator of 100. 9. The p r o f i t i n value units to each type of subject i n a trade of 100 buttons of one colour for 100 of the other i s shown i n the Table below: 144 Subjective Gain and Costs* to P for a 100/100 Exchange on T r i a l 1, Set B Subject #• of buttons i n small and large p i l e s increase m value units costs i n value units net p r o f i t i n value units Hi NV V i s i b l e Low NV 150/1250 100/1100 50/9-500 250 290 370 87. 100 113 162. 190 257 * Calculated from the value chart, Appendix 1. It can be seen from t h i s table that the low non-visibles made consid-erably more p r o f i t i n value units from a one for one trade of buttons, than did the v i s i b l e s or the high non-visibles (257 versus 190 and 162.5, r e s p e c t i v e l y ) . Thus, to be f a i r i n an exchange with a v i s i b l e , a low n o n - v i s i b l e would have to give up more than he requested. 10. Only i n i t i a t i o n s to v i s i b l e s were considered, because fairness was not defined from the i n i t i a t o r ' s point of view, i n an o f f e r to a non-visible. 11. The advantage to high non-visibles may have l a i n elsewhere than i n advan-tageous i n i t i a t i o n s i n v i s i b l e s . In Set B l , the high non-visibles could i n f e r from the i n s t r u c t i o n s that so long as they^kept ahead of the v i -s i b l e s , they would most l i k e l y be among the four winners at the end. As they were t o l d that some non-visibles had more than v i s i b l e s , and some l e s s , they might, on symmetry grounds, think only one other non-v i s i b l e was a high. To win, they could e i t h e r make the sort of deals with v i s i b l e s that would not l e t the v i s i b l e improve his t o t a l wealth, or they could deal with a non-visible who might be the same, or greatly behind in t o t a l resources. Since they had no information about non-v i s i b l e s , they could not make use of i t to decide what p r i c e was rea-sonable; however, so long as they traded even amounts of buttons with non-v i s i b l e s , they were l i k e l y to maintain t h e i r p o s i t i o n v i s - a - v i s the v i s i b l e s , sinply because they started with more buttons. In a way, high non-visibles had a l l the information they needed to play a winning strategy — they believed only one other v i s i b l e had as much as they did. I f a person had enough information to f e e l secure that no one would do better than he, he would also f e e l f r e e r to take r i s k s , e.g., by trading with the non-visibles, or by t r y i n g to trade with the v i s i -bles i n such a way that i t gave an advantage to the non-visibles. 12. See Footnote 9. 13. Offers were standardized to a r a t i o of X/100, and the mean over a l l r a t i o s for a group of N subjects was c a l c u l a t e d : n 'fX/10.6V , :>n' where i i s the i n i t i a t o r i = l , 2 ... n. ^ (X/10Q)-••./,£ 145 14. In the P i l o t work described i n Appendix IV, where some v i s i b l e s had ' f a i r ' exchange r a t i o s greater than 100/100, the actual ratios were lower than the f a i r r a t i o . The average departure from the f a i r r a t i o i n these cases was -.30 (below the f a i r r a t i o ) , contrasted with an average departure of +.066 (above the f a i r r a t i o ) , when fairness c a l l e d on sub-j e c t s to o f f e r less than they requested. I f low non-visibles i n Set B had followed the pattern of the subjects i n the P i l o t set, t h e i r average exchange r a t i o would have been 95/100, compared with t h e i r actual ob-tained average r a t i o of 85/100. 15. In Set A, subjects did not know the exact number of buttons the v i s i b l e s had, although the high resource v i s i b l e s had p h y s i c a l l y larger p i l e s of buttons. The r e s u l t s from these subjects did not support Hypothesis 2. However, only 63% of the ron-visibles stated that they perceived the covers to.be an advantage: of these, 47% made i n i t i a t i o n s which asked for more than they offered. (See discussion of Hypothesis 3.) 16. See Appendix I I , Questionnaire for non-visibles. 17. See Appendix I I I , Questionnaire 3. 18. See Appendix I I I , Questionnaire 3. Note that although non-visibles had received f a l s e o f f e r s by the time the questionnaire was administered, they had not received r e p l i e s to t h e i r own o f f e r s . 19. Non-visibles' reasons for wanting the covers l e f t on were b a s i c a l l y the same as t h e i r reasons for perceiving the covers to be an advantage. They-,focussed on the a b i l i t y to keep one's progress i n the game a sec-r e t , so that no one would try to 'thwart' them, and on being able to force more out of another player whose need was apparent. The importance of the two senses of advantage — over the game as a whole, and i n s e t t i n g the terms for i n d i v i d u a l transactions — i s again i n evidence. Both high and low non-visibles gave s i m i l a r reasons for wanting to r e t a i n the covers. Of the lows who wanted the covers o f f , the main reason appeared to be the b e l i e f that i f others saw how poor they were, they would be perceived to be j u s t i f i e d i n asking for a larger return. 20. Note: A l l 56 non-visibles i n Set A were asked about r e t a i n i n g the covers on T r i a l 1 , and 17 of these wanted them removed i f they were given the choice on T r i a l 3. For the 48 subjects asked about what they saw the advantages to be, 2/30 claiming advantages i n covers said they wanted them o f f ; 12/16 who saw covers as a disadvantage wanted them o f f . 21. See Appendix I I I , Questionnaire 1. 22. The data from Set A followed e s s e n t i a l l y the same pattern, and are given i n Appendix V, Table A.10. 146 23. Players were not completely c e r t a i n that any i n i t i a t i o n they made would lead to a completed transaction. However, expected values of exchanges with v i s i b l e s could be calculated on the baiss of the l e a s t ambiguous sort of information a v a i l a b l e . 24. Some res u l t s from Set A w i l l also be given, as they demonstrate that the preference for v i s i b l e s continued beyond the f i r s t t r i a l , even when the strong preference for the high v i s i b l e s had declined. 25. In addition, i f non-visibles anticipated that other non-visibles might receive fewer o f f e r s than v i s i b l e s , then an i n i t i a t i o n to a non-visible might have seemed more l i k e l y to be accepted. Subjects' reasons for i n i t i a t i n g to v i s i b l e s and non-visibles are l i s t e d i n Appendix V, Table A.11 through A114. Approximately one-third of subjects who sent o f f e r s to non-visibles gave the anticipated unpopularity of non-visibles as the reasons. 26. A set of ten experiments with M> :subjects i n each ( i . e . , s i x v i s i b l e s and s i x non-visibles) a l l with the same t o t a l resource bases and,, r a t i o s of red to green or green to red, 0was conducted for purposes outside the scope of t h i s d i s s e r t a t i o n . In this set, 71% of T r i a l 1 i n i t i a t i o n s went to v i s i b l e s . I t i s i n t e r e s t i n g to note that females directed 77% of t h e i r T r i a l 1 i n i t i a t i o n s to v i s i b l e s , compared to 65% for males. 27. Some p l a u s i b l e explanations of these r e s u l t s w i l l be discussed i n the next Chapter. 28. Binomial test gives Z = 3.63 for i n i t i a t i o n s made by v i s i b l e s . The non-visibles did not d i r e c t s i g n i f i c a n t l y more i n i t i a t i o n s to v i s i b l e s on T r i a l 2. 29. Real o f f e r s were delivered to v i s i b l e s . 2 2 30. An9(>, test for the table gives X\ = 9,6, d.f. = 1, p .01. 31. The r e j e c t i o n of non-visibles by v i s i b l e s , by type of o f f e r , i s shown i n the table below: Number Rejected per Number Received  Person Rejecting Offer asks same asks more asks le s s as o f f e r than o f f e r than o f f e r High v i s i b l e 10/13 13/15 2/3 Low v i s i b l e 0/2 2/2 1/2 From this table, i t i s not clear that non-visibles could have entered into trades i f they gave up the idea of advantageous exchanges, and made f a i r or generous o f f e r s , as even f a i r o f f e r s were frequently rejected. 147 32. See H.H. Kelley and D.P. Schenitzki, 'Bargaining', Chapter 10 i n C.G. McClintock, "Edi t o r , Experimental S o c i a l Psychology^. Holt, Rinehart and Winston, 1972, pp. 299sr307, for a discussion of the experimental paradigm most frequently used for studies of b i l a t e r a l monopoly bar-gaining. 33. Of the 7 o f f e r s from non-visibles accepted by the v i s i b l e s , e i t h e r no o f f e r was received from a v i s i b l e , or the o f f e r from the v i s i b l e asked for more than i t offered, i . e . , was worse i n terms of absolute p r o f i t to P than the one accepted. One person said he had made a mistake i n accepting a non-visible (he rejected two o f f e r s of 100 f o r 1 0 0 , and gave up 150 for 100). Two out of three low v i s i b l e s accepting a non-v i s i b l e received only that one o f f e r . 148 CHAPTER 5 EVALUATION AND SUGGESTIONS FOR FUTURE RESEARCH This d i s s e r t a t i o n has been concerned with the use of information control as a t a c t i c for obtaining advantage i n exchange r e l a t i o n s h i p s , and the l i m i t a t i o n s on the successful use of such a t a c t i c imposed by a p o s i -ted preference for r e l i a b l e information about exchange partners. A theory was constructed to explain the processes operating i n such a s i t u a t i o n , and an experiment was designed to test a set of s i x hypotheses derived from the theory. As the test and r e s u l t s have been presented i n some d e t a i l , there i s no need to repeat them here. Given that the r e s u l t s on the whole gave support for the theory i n the experimental context, i t i s appropriate at t h i s point to make a d i f f e r e n t evaluation of the theory, based on other c r i t e r i a . At the conclusion of an experimental study, someone usually asks: "Well, what does this t e l l us about the r e a l world?" This question essen-t i a l l y refers to the generalization from the study to s i m i l a r s i t u a t i o n s outside the laboratory. I f the interrogator wwishes to know where e l s e we can f i n d groups of people s i t t i n g around dimly l i t tables trading coloured tokens, our answer must be not "In poker h a l l s " , nor "At children's p a r t i e s " , but "Nowhere". I t should be clear from the way t h i s study has been con-structed that we do not expect to generalize from the concrete experimental s e t t i n g d i r e c t l y to another concrete singular s i t u a t i o n with s i m i l a r char-a c t e r i s t i c s . The view taken here i s s i m i l a r to that expressed by Webster 1 and Kervin: 149 The connection between the laboratory and the natural s e t t i n g , we claim, i s the theory. Without the theory there i s , and i n fac t there can be, no l i n k . S p e c i f i c a l l y , the l i n k i s the set of ab-st r a c t scope conditions which t e l l whether the theory can be used to make predictions for a p a r t i c u l a r s e t t i n g . I f there i s no theory, and i f there are, therefore, no e x p l i c i t scope conditions, then no genera l i z a t i o n of the res u l t s of a laboratory study i s permissible. S t r i c t l y speaking, t h i s i s also true of any study, laboratory or other. We construct theories with general statements in v o l v i n g abstract v a r i a b l e s . As indicated above, these are made p r e d i c t i v e i n p a r t i c u l a r instances by a set of i n i t i a l conditions which, together with the abstract assumptions, are used to deduce hypotheses. In any concrete s i t u a t i o n , experiment or other-wise, where i t can be shown that the relevant conditions are met, and that 2 the assumptions of the theory apply, predictions can be made. TThus, the findings of a p a r t i c u l a r experiment are generalizable through the theory to other s e t t i n g s , not by d i r e c t extrapolation from an experiment to the r e a l world. In this\''view, the scope conditions both provide generality and impose l i m i t a t i o n s . LLet us look at the implications of such a view f o r the work reported i n th i s study. Any empirical s c i e n t i s t , of course, wants to devise theories with antecedent conditions that have a va r i e t y of concrete instances, but no theory can encompass a l l the world, or for that matter, a l l of s o c i a l be-haviour. Our concern has been what part of s o c i a l behaviour that can be conceptualized as s o c i a l exchange. We further l i m i t e d our i n t e r e s t to ex-change s i t u a t i o n s i n which: 150 1) People could obtain complementary commodities from at l e a s t two others (Scope condition 1). 2) The resources were valued according to a marginal u t i l i t y function (Scope condition 2). 3) The value functions of i n d i v i d u a l s could be known or i n f e r r e d , though information about current resource holdings of others might be incom-plete (Scope condition 3). 4) People could compare p r o f i t s at some stage (Scope condition 4). 5) , Total resources i n the group did not increase (Scope condition 5) . 6) Some people could control but not f a l s i f y information about t h e i r cur-rent resource holdings (Scope condition 6). This admittedly t y p i f i e s only a part of s o c i a l i n t e r a c t i o n , but c e r t a i n l y not so small a part as groups of students trading red and green buttons according to a value chart. Support for the experimental predictions based on the theory gives us more confidence i n predictions to other s i t u a t i o n s that meet the i n i t i a l conditions. We w i l l b r i e f l y consider some possible s i t u a t i o n s to which the theory could be shown to be relevant, and at the same time, suggest where further work i s needed to s p e l l out the a p p l i c a -b i l i t y of the theory. 1. Valued complementary resources d i s t r i b u t e d i n groups of four or more. This condition requires l i t t l e comment, i f one accepts a model of s o c i a l behaviour as exchange, because almost any s e r v i c e , good, or sentiment 3 can be seen as a resource. Thus, a father who takes h i s son s k i i n g , i f the son washes the car, can be seen as s i m i l a r i n relevant respects -to stu-dents who help one another study, or c h i l d r e n who trade hockey cards. Pro-151 bably the most d i f f i c u l t problem to resolve concerning the a p p l i c a b i l i t y of this condition i s to work out how many d i f f e r e n t resources are involved i n a r e l a t i o n s h i p — for example, do students exchange only help, or i s companionship also a reason for the relationship? In addition, i t i s not s u f f i c i e n t to note that a group has four or more members, but that the a l -ternative exchange partners are i n fact a v a i l a b l e to enter exchange transac-tio n s . 2. Marginal u t i l i t y function I t i s a simple matter to note that the more we get of a resource, the less we value even more of i t . 1-Intuitively, this p r i n c i p l e seems to apply to help, deference, approval, and i n reverse, to time and e f f o r t 4 given up. I t i s much les s simple to specify the periods of time over which s a t i a t i o n and deprivation w i l l take place, and the units i n which resources w i l l be obtained or given up. Without being able to specify the u n i t s , even a r b i t r a r i l y , i t i s d i f f i c u l t to know whether t h i s i n i t i a l condition i s met or not. In addition, some resources, such as money, or votes, that can be used as resources i n exchange, are not always valued according to a marginal u t i l i t y function. This seems to imply that the second i n i t i a l condition could be generalized to specify only that there be a d i f f e r e n t i a l valuation of resources, to provide a basis f o r exchange. 3. The condition that i n d i v i d u a l s know or be able to i n f e r the value functions of others seems at once a very l i m i t e d and a widely applicable statement. There i s a vast array of people i n the world whose value functions we do not know, but at the same time, we are not so l i k e l y to enter into 152 exchanges with them, fContinuing i n t e r a c t i o n i n exchange r e l a t i o n s h i p s builds up a store of information about the outcomes that Other has pursued most e n t h u s i a s t i c a l l y , the transactions that have given him the most s a t i s -f a c t i o n , and the^ terms of transactions he has agreed to with oneself and with others. In addition, people can generalize t h i s kind of information to s i m i l a r others i n s i m i l a r s i t u a t i o n s . The second part of Scope condition 3, that people do not know how much of a resource some others have,limits the a p p l i c a b i l i t y of the theory to a narrower range of s i t u a t i o n s . People, i n close contact are l i k e l y to know what each other has; some relationships probably demand f u l l information, such as i s found i n f a m i l i e s . Neverthe-l e s s , actors often have d i s c r e t i o n over the information a v a i l a b l e about them, and p a r t i c u l a r l y i n the early stages of i n t e r a c t i o n i n v o l v i n g new resources, people w i l l often not be c e r t a i n about the resource bases of p o t e n t i a l partners. (Even i n extended interactions such as the family, husbands have been known to conceal from other members the size of t h e i r incomes.) 4. Opportunity to compare p r o f i t s The opportunity to compare p r o f i t s i n exchanges i s a consequence of the d i s t r i b u t i o n of information, which can change.over time. I t may often occur that exchange partners do not know how the other gains at the time of the transaction, and that comparison l a t e r becomes possible when one i s able to observe the other's subsequent reaction (e.g., his s a t i s f a c -t i o n , what he does with the p r o f i t , or what he t e l l s others). Again, t h i s points to the early stages of exchange r e l a t i o n s h i p s that w i l l continue, or to established ones i n which d i f f e r e n t commodities are introduced. I f ex-153 change networks are well connected, and have existed over a long period of time, i t i s less l i k e l y that members would be unable to compare p r o f i t s at the time of the exchange. TT-here are exceptions, of course, such as labour-union negotiations, i n which comparison, or at l e a s t , honest comparison of outcomes at the time of bargaining i s d e l i b e r a t e l y avoided by both p a r t i e s . The desire to compare p r o f i t s seems to t y p i f y a great many r e l a t i o n s h i p s , and may lead to information-gathering where there i s inadequate information to allow comparison. 5. T o t a l resources i n the group constant For short periods of time, most simple exchange re l a t i o n s h i p s can be t y p i f i e d as constant-sum s i t u a t i o n s . Even i f resources increase at regu-l a r i n t e r v a l s , as i n the case of income, consumption usually ensures that the resource bases of members do not a l t e r dramatically, and i f they do, i t occurs i n a predictable fashion. However, the f i f t h scope condition rules out the many i n t e r a c t i o n s — i n which people j o i n t l y produce new r e -sources, or when ad d i t i o n a l amounts of resources are i n j e c t e d into the group from an external source. In such cases, there are often enough rewards to s a t i s f y everyone, and cooperation and trust are more l i k e l y to p r e v a i l . Where rewards of P^and 0 are p o s i t i v e l y c orrelated, the sharing of informa-t i o n i s more t y p i c a l , since i t increases both p a r t i e s ' a b i l i t y to coordinate, 5 and obtain further rewards. In such cases, problems of fairness tend to involve questions of proportional returns on investments, and these ques-tions are not handled by the assumptions of the theory presented i n t h i s study 154 6. Some members can withold information, but not give out f a l s e informa-t i o n I t might at f i r s t seem that i n any s i t u a t i o n where P can con t r o l information, he w i l l , i n add i t i o n , give out f a l s e information that construes the s i t u a t i o n to h i s advantage, rather than r i s k the chance that Other w i l l make assumptions unfavourable to P. However, there are many con-s t r a i n t s against f a l s i f i c a t i o n , not the le a s t of which are sanctions meted out f o r l y i n g , but which are not applied for saying nothing. One can also be s e l e c t i v e i n the release of information, so that only that serving one's 7 i n t e r e s t i s made a v a i l a b l e . I f P simply witholds information, the assump-tions made by 0 w i l l often be favourable to P, as when 0 believes everyone w i l l act f a i r l y , or i f there are norms against suspecting other people's motives. The key to obtaining an.advantage probably l i e s i n s e l e c t i n g 8 occasions where the ri g h t assumptions w i l l be made. Once again, Scope condition 6 i s more l i k e l y to be applicable to the early stages of i n t e r -action, such as the fencing that occurs at c o c k t a i l p a r t i e s , the early stages of dating, or i n re l a t i o n s h i p s where the members cannot obtain an immediate independent v e r i f i c a t i o n of the cues given by P. (For examplej l i t t l e brothers and s i s t e r s who t e l l secrets undermine the control of cru-c i a l information i n rel a t i o n s h i p s as diverse as the.swapping of hockey cards, and impressing a new boyfriend.) Our interrogator may now object that to provide instances f o r each of the scope conditions separately does not ensure they w i l l a l l hold i n any given s i t u a t i o n . There are two r e p l i e s that may be made to such an objection. 155 1. F i r s t , we would argue that the assessment of the relevance of an abstract t h e o r e t i c a l formulation such as t h i s one does not rest s o l e l y on the a b i l i t y to go out and count up a large number of concrete cases that meet the assumptions and scope conditions of the theory. 1'The h i s t o r y of the natural sciences contains many instances of seemingly i r r e l e v a n t the-o r e t i c a l formulations which l a t e r proved to have important a p p l i c a t i o n s . One reason for th i s seems to be that u n t i l a p a r t i c u l a r i n t e r p r e t i v e frame-work i s developed for a set of events, we do not characterize or recognize the events as instances that f i t that framework. In addition, processes that occur infrequently (or not at a l l ) i n the 'real world' are not necessarily unimportant by v i r t u e of being infrequent for they may provide the occasion f o r t e s t i n g predictions about underlying p r i n c i p l e s which are usually confounded with other processes. One of the strongest arguments i n favour of a r t i f i c i a l i t y i n laboratory experimentation i s that i t allows us to eliminate factors not s p e c i f i e d i n the theory. Then the r e s u l t s w i l l bear unequivocally on the soundness of the propositions, making d i f f i c u l t the pr e v a r i c a t i o n that 'other f a c t o r s ' may have prevented us f i n d i n g support for our pre d i c t i o n s . The r e s u l t s i n the present study, for example, make i t clear that our propositions are not adequate to account for the behaviour of a l l our subjects, and that further refinement of the 9 assumptions and o p e r a t i o n a l i z a t i o n are necessary. A further j u s t i f i c a t i o n f o r studying processes which may occur only infrequently i s that these processes can have important and long term consequences. In our case, for example, the process of e s t a b l i s h i n g the terms of transactions to favour oneself may occur r a r e l y , because we tend 156 to l e t most of exchange proceed by habit, according to standard terms. Nevertheless, i f an actor can influence the i n i t i a l expectations about acceptable terms i n transactions, i t can exercise an influence on routine i n t e r a c t i o n w e l l into the future. 2) The second reply to the question of whether a l l our scope conditions are ever met together i n concrete s i t u a t i o n s i s that the set of i n i t i a l conditions are not mutually exclusive, and there i s l o g i c a l l y no reason to expect they w i l l not occur together. We have suggested that sev-e r a l of the scope conditions t y p i f y the early stages of exchange i n t e r a c -t i o n s , e i t h e r when a new set of actors are e s t a b l i s h i n g r e l a t i o n s h i p s , or are negotiating the prices of a new set of resources for which there are no standard valuations. We believe the scope conditions may be shown to be t y p i c a l of several d i f f e r e n t kinds of exchange contexts. For example, the theory might be used to predict that employees would be more w i l l i n g to accept wage r e s t r a i n t (lower p r o f i t ) i n a company i n which the employees share i n decision making and have access to information about the true costs borne by the organization, than would workers who must i n f e r management's true costs on the ambiguous basis.of wage o f f e r s presented to the employees. Again,in organizations, the theory would predict that the success of 'games-manship' ( i n which people.engage i n the s e l e c t i v e release of information that puts them i n a favourable light).would be severely constrained i f there were also 'honest players' i n the group, who t y p i c a l l y released a l l relevant information. Unpleasant feelingsmight a r i s e toward a person who had benefitted from the s e l e c t i v e release of information, i f he did not also continue to conceal the s a t i s f a c t i o n he had obtained from the advantage. ','i'The r e c i p r o c a l 157 trading of votes on issues of d i f f e r e n t importance to d i f f e r e n t people i n p o l i t i c a l spheres could also be an area i n which the theory could be made p r e d i c t i v e . For example,muerit e f f o r t i s probably spent i n concealing true cost and reward l e v e l s i n such cases, while the desire f o r r e l i a b l e information about the Other i s very high. Other examples i n informal r e -l a t i o n s h i p s such as dating networks,people i n cooperative houses where costs are assessed r e l a t i v e to wealth, students engaging i n r e c i p r o c a l help, could probably~be shown to s a t i s f y the complete set of scope condi-.' 10 tions of the theory. We w i l l not continue with a l i s t of possible instances of s i t u a -tions that could possibly be accounted for by the theory, but w i l l now mention some issues which t h i s study suggests would merit further research. Suggestions for Future Research The support f o r the hypotheses described i n Chapter 4 gives us a degree of confidence i n the assumptions on which they were based, but there was s t i l l a good deal of e r r o r i n the p r e d i c t i o n s . We cannot be c e r t a i n the assumptions, the hypotheses, or the o p e r a t i o n a l i z a t i o n are responsible 11 for the f a i l u r e , but i t i s possible to make some reasonable guesses. I t was assumed that people are not only concerned with t h e i r own p r o f i t s i n exchange, but also with a comparison of p r o f i t s . We argued that people preferred to gain more than others, p a r t i c u l a r l y where the t o t a l r e -sources are constant, since a gain to 0 i s a loss to P. I t seemed clear i n the experiments, however, that not a l l subjects were interested i n compari-son of p r o f i t s , and that i n d i v i d u a l differences i n o r i e n t a t i o n to others may 158 be an important factor needing i n v e s t i g a t i o n . People are a l l l i k e l y to be motivated to compare with others i n some s i t u a t i o n s , but the motivation to do so varies across i n d i v i d u a l s and across s i t u a t i o n s , i n a manner that needs much more s p e c i f i c a t i o n . We argued that comparison i s s a l i e n t f or peers who can observe others' gains and losses, but i t i s clear that no one can systematically compare how well he does i n r e l a t i o n to a l l possible partners, due to l i m i -ted cognitive capacity, and the increasing costs of doing so. I t would be of great i n t e r e s t to be able to predict the subset of others with whom P w i l l choose to compare. Recent studies i n s o c i a l comparison theory could 12 possibly provide p r i n c i p l e s that might be integrated into our theory. A further question a r i s i n g from the comparison of p r o f i t s i s the soundness of the assumption that P can take the point of view of others, i f he knows t h e i r resource bases, to assess how others value outcomes. In the experiments, t h i s process was made very easy for subjects by the pro-v i s i o n of the value chart. Even so, some subjects did not appear to pro-cess information i n the manner predicted, and some commented that i t was 'too much trouble'. I t must be admitted that people often f i n d i t simpler to proceed through a process of t r i a l and e r r o r , making adjustments i f something goes wrong, or i f P does not manage to obtain a l e v e l of reward that i s adequate. In addition, terms of exchange are often a matter of habit and h i s t o r y , and do not involve (regular) and continuous negotiation. One of the functions of s o c i a l standards of fairness i s to r e l i e v e the actor of the chore of negotiating each encounter afresh, and any decisions he makes about 159 entering transactions may r e l a t e to s e l e c t i n g those whose standard terms are a t t r a c t i v e . In s p i t e of the fact that role-taking may not occur i n a routine way, i n most i n t e r a c t i o n , we would argue that: l ) i f the rewards are impor-tant to P, or i f 'something goes wrong', and his normal l e v e l of rewards i s disrupted, i t w i l l be worth h i s while to take the point of view of other to re-assess what terms he i s able to get i n an exchange, and 2) when new or unique re l a t i o n s h i p s are f i r s t established, some negotiation of terms 13 for the r e l a t i o n s h i p i s required. ' The widespread perception of advantage i n concealed resources i n the experiment leads us to believe that we are addressing a meaningful aspect of exchange r e l a t i o n s h i p s . Apart from a vague sense that i t i s 'good to keep one's a f f a i r s p r i v a t e ' , many subjects were very a r t i c u l a t e about the manner i n which asymmetric information could be used to give 0 the impres-sion that P required good terms to agree to a transaction, at the same time that 0 was kept to terms r e f l e c t i n g his true needs. TT-he f a c t that more people perceived an advantage, than made advantageous o f f e r s , may mean that the advantage can operate i n d i f f e r e n t ways. (For example, i n the ex-periment, one advantage lay i n preventing others from knowing that P had made several transactions.) I t i s also l i k e l y that t a c t i c s work only i f used sparingly: Tactics^are highly, personal, subtle, and evanescent; t h e i r outcome depends on the correct (or incorrect) i n t e r p r e t a t i o n of any one or more of many behavioural cues, which may themselves be genuine or pretended; and the net e f f e c t often involves such complex i n t e r -actions as what A thinks B thinks about what A i s thinking.-*-^ 160 While i t may be true that t a c t i c s are not used r e g u l a r l y , our f o r -mulation of how P takes O's point of view provides a r a t i o n a l e for t h i s process that goes beyond the simple assumption by many j-researchers that asym-metric information i s somehow a t a c t i c a l advantage. I f we agree that tactics are 'evanescent', i t would be desirable to devise indices of the use of t a c t i c s which are more s e n s i t i v e than the simple proportions of advantageous i n i t i a t i o n s that were used i n the present research. In addition, we would expect that t a c t i c s of advantage w i l l only be used u n t i l the advantage i s obtained — i f i t i s only necessary to do better than Other once i n order for the advantage to continue, then tactic's w i l l occur infrequently. I f the s i t u a t i o n requires that the advantage be renewed, we expect t a c t i c s to be used with greater frequency. The process of taking the point of view of Other also helps us to understand why information about P's outcomes i s so important to 0 — without information, he cannot make comparisons. The support for the exper-imental predictions concerning the preference for others about whom P has information gives us confidence i n the assumption that ambiguity of i n f o r -mation about an a l t e r n a t i v e leads to uncertainty, and that t h i s reduces the expected value of an a l t e r n a t i v e . We argued that subjects would prefer to have unambiguous informa-t i o n about a partner's resources, and that i f such information was a v a i l a b l e about some Others, t h i s information would be processed to make inferences about the p r o b a b i l i t y of acceptance of o f f e r s , and the fairness of d i f f e r e n t exchanges. Subjects i n the experiments seemed to have believed that they could discover the n o n - v i s i b l e s ' true p r o f i t s by means of the types of o f f e r s the l a t t e r made and accepted. However, given that such inferences might be unreiliabre;, they preferred a l t e r n a t i v e s for which more c e r t a i n 161 information was already a v a i l a b l e . The extension of findings concerning a preference for r i s k over uncertainty, to the case of degrees of uncertainty i n s o c i a l s i t u a t i o n s , seems a promising area of inquiry. I f we can even-t u a l l y delineate the factors i n a s i t u a t i o n which lead to a greater or l e s s e r degree of uncertainty, then a s o c i a l choice model incorporating uncertainty as suggested i n t h i s t h e s i s , would be a contribution. Factors that could be investigated include P's b e l i e f s about O's s e l f - i n t e r e s t , (for example, the tendency of powerless people to a t t r i b u t e negative i n t e n -tions to others); the v a r i a b i l i t y of information, (for example, i f i t comes from several d i f f e r e n t sources); the c r e d i b i l i t y of d i f f e r e n t people, (for example, i s information from high status persons more cr e d i b l e than from low 15 status persons?). A model which s p e c i f i e d the antecedent factors i n f l u -encing uncertainty would also provide a framework for p r e d i c t i n g the con-di t i o n s under which witholding information could succeed as a t a c t i c , (for example, when no alt e r n a t i v e s provide r e l i a b l e information, or when a l t e r -natives about whom there i s r e l i a b l e information are undesirable on other grounds). The s i x cases of.information and a l t e r n a t i v e s described at the end of Chapter 1 could be assimilated into the theory by making a l t e r a t i o n s i n the scope conditions concerning the number of a l t e r n a t i v e partners, and the amount of information available about them. I t would then be possible to use the assumptions i n the theory to make predictions for the s i x cases. This would have the advantage of i n t e g r a t i n g a somewhat a t h e o r e t i c a l body of work concerning information i n negotiation into a s i n g l e framework and 162 extend the conceptualization of bargaining to include the e f f e c t s of a r e -source base against which p r o f i t s are calculated. The consistent minority of subjects who,did not display a pre-ference for v i s i b l e s i n the experiments requires some comment. The data do not permit an unambiguous ad hoc explanation of i n i t i a t i o n s to players with hidden resources, and at least three ways of accounting for the negative cases could be pursued: 1) There may be i n d i v i d u a l differences i n the avoidance of uncertainty. Ellsberg,and MacCrimmon,, note that only some subjects display a p r e f e r -ence for r i s k y over uncertain bets, when the bets have equal expected 16 value. Individual differences i n the a b i l i t y to a n t i c i p a t e the l i k e -lihood of having an o f f e r accepted might have led some subjects to d i r e c t o f f e r s to the less popular non-visibles; and there may well be i n d i v i d u a l differences i n the desire to compare p r o f i t s with others. These factors r e l a t e to d i f f e r e n t components of the decision function given i n Assump-t i o n 5. I n i t i a t i o n s to non-visibles were taken to i n d i c a t e that the s u b j e c t i v e l y expected value of an exchange with a non-visible was greater than that with a v i s i b l e , but i t i s not c l e a r which part of the function was affected, i . e . , the p r o b a b i l i t y of acceptance, the weight-ing of uncertainty, or the value to P of comparison with 0. 2) Future work could explore the p o s s i b i l i t y of developing a model that conceives of the actor as randomly considering one of three p o s s i b i l i -t i e s — that a non-visible i s b e t t e r , as good, or worse than a v a i l a b l e v i s i b l e partners. This might then lead to the two-thirds of the subjects 163 who on average made off e r s to v i s i b l e s , i . e . , t h i s proportion would represent those who believed the v i s i b l e s were as good or better than the ron-visibles. This would be i n accordance with Ellsberg's sugges-t i o n that subjects w i l l p r e f e r r i s k to uncertainty, i f the r i s k y a l t e r -native i s estimated to be as good or better than the uncertain one. The remaining one-third of the subjects who hold the b e l i e f that the non-v i s i b l e s would make better partners would then account for the proportion of subjects who did not act according to the theory as i t now stands. Such a model would require a more d e t a i l e d d e s c r i p t i o n of the factors the actor takes into account i n deciding that one a l t e r n a t i v e i s 'as 'good' as another, and more s e n s i t i v e methods of measuring the subjects' decision processes during i n t e r a c t i o n . 3) We can also conceive of the subjects as assigning p o s i t i v e expected value to exchanges with both v i s i b l e s and n o n - v i s i b l e s , but a higher value to the former. Offers may be directed i n proportion to the r e l a - t i v e expected value, i . e . , frequency of i n i t i a t i o n / n = expected value ( v i s i b l e partner) EV ( v i s i b l e ) + EV (non-visible) Ins such a model, a c e r t a i n smaller proportion of o f f e r s would be expec-ted to go to n o n - v i s i b l e s , and some of these might be sent on the f i r s t opportunity for exchange. Studies of choice i n p r o b a b i l i t y matching tasks never show a 100% choice of the a l t e r n a t i v e s with the highest ex-pected value, i f subjects are given a s e r i e s of choices. In a l l . t h e s e conceptualizations, the d i f f i c u l t problem of assessing expected value i s compounded by the fact that p r o b a b i l i t y and u t i l i t y i n t e r a c t — 164 the more desirable a transaction to P, the less l i k e l y i t i s to occur, be-cause i t i s undesirable to 0. The lexicographic representation of the choice process, though imprecise, seems p l a u s i b l e i n l i g h t of the data. I t seems clear that subjects did i n fact consider p r o f i t to s e l f as the most important f a c t o r , indicated i n the acceptance of o f f e r s giving the largest p r o f i t , regardless of source (e.g., Set A, T r i a l 1). When p r o f i t to P was held constant, by d e l i v e r i n g f a l s i f i e d o f f e r s , the a b i l i t y to compare p r o f i t s with the partner seemed to lead a majority of subjects to accept the v i s i -b l e s . In s p i t e of the negative, cases, the evidence for the predicted preference for partners with known resource l e v e l s gives up more confidence i n our explanation of the process by which concealment of information does not turn out to be a s u f f i c i e n t t a c t i c to ensure advantageous exchanges. The preference f o r information and the inferences made about those who withold information, appear to l i m i t the success of such a t a c t i c when a l t e r n a t i v e partners are av a i l a b l e . I t has already been noted that the theory could be extended .to make predictions - for cases when there are no a l t e r n a t i v e s , and information i s d i f f e r e n t l y d i s t r i b u t e d , and future work could involve a systematic test of the d i f f e r e n t cases. The paradigm used i n th i s study lends i t s e l f to further research i n the use of informational t a c t i c s . In p p a r t i c u l a r , d i s c r e t i o n over the release of information about resources could be simply v a r i e d , by equipping the booths with i n d i c a t o r s of the numbers of d i f f e r e n t coloured buttons. These indicators could be co n t r o l l e d by the experimenter or by the subjects. 165 I t i s not necessary that buttons change hands i n the experiment, and many more t r i a l s could be run i f i n d i c a t o r s were used to record, honestly or d i s -honestly, successive changes i n resource bases through exchange. I t would also be possible to vary the type of value function across subjects, so that deception could occur concerning both the s i z e of the base, and the value function. The foregoing discussion should make i t clear that there i s scope f o r further research based on the theory, even within the present experimen-t a l paradigm. I t should also be possible to design d i f f e r e n t experiments that overcome some of the weaknesses i n the paradigm used i n t h i s study, such as the confounding of long- and short-term gain, and the d i f f i c u l t y of creating doubt about the resource bases of concealed subjects. Computer terminals o f f e r a l o t of p o t e n t i a l for both recording the subjects' reac-tions throughout an experimental exchange s i t u a t i o n , and for simulating the other members of a 'group'. Because attempts at advantageous exchange seem more l i k e l y to succeed i f t r i e d only o c c a s i o n a l l y , i t may be that f i e l d studies would further aid i n the d e l i n e a t i o n of conditions antecedent to attempts by s o c i a l actors to obtain advantage, and the mechanisms involved i n carrying i t o f f . 166 FOOTNOTES FOR CHAPTER 5 1. M. Webster, J r . , and J.B. Kervin, ' A r t i f i c i a l i t y i n experimental s o c i o -logy', Canadian Review of Sociology and Anthropology, 8_, 1971, p. 268. 2. The process of p r e d i c t i o n i s not so simple, of course, because one theory does not include propositions about a l l the factors i n a par-t i c u l a r concrete s i t u a t i o n that may contribute to and change the values that the variables assume i n that s i t u a t i o n . 3. See: A. Kuhn, The Study of Society: A U n i f i e d Approach, London: Tavistock P u b l i c a t i o n s , 1966, pp. 260-61. 4. For example, the v a r i e t y of occasions on which the expression 'carrying coals to Newcastle' i s used, i s testimony to t h i s f a c t . 5. Friends probably s e l e c t a c t i v i t i e s that are characterized by p o s i t i v e l y correlated rewards,, or i n which those aspects can be emphasized. In addition, such - a c t i v i t i e s also would generate p o s i t i v e sentiment and f e e l i n g s of friendship, 6. The issue of proportional return for investments i s usually r e f e r r e d to as d i s t r i b u t i v e j u s t i c e or equity. See, for example, G.C. Homans, Social!. Behaviour: I t s Elementary Forms, New York: Harcourt, Brace and World, 1961, pp. 232-264. 7. Press censorship i s based on the s e l e c t i v e release of information that supposedly reinforces a p a r t i c u l a r d e f i n i t i o n of the s i t u a t i o n . Its effectiveness i s probably dependent on whether readers suspect they are being given only part of the f a c t s . 8. Paradoxically, the desire for advantage i s l i k e l y to be lower when con-diti o n s are most favourable for i t s success, as when trust e x i s t s be-tween the p a r t i e s . P would r i s k a loss over the long run i f he v i o l a t e d t r u s t , because i t would disrupt the r e l a t i o n s h i p . 9. A further advantage of making a c o n t r o l l e d t e s t of predictions from ex-p l i c i t assumptions and scope conditions i s that we then have a better idea of where the weaknesses l i e . Factors which have been c o n t r o l l e d or eliminated cannot be blamed f o r negative f i n d i n g s , r and this narrows the range of d i r e c t i o n s to take i n r e v i s i n g p r e d i c t i o n s . 10. Z e l d i t c h makes the point that disputes about the a p p l i c a b i l i t y of a theory depend for t h e i r ultimate r e s o l u t i o n on d e s c r i p t i v e knowledge of a p a r t i c u l a r s i t u a t i o n . However, experimental i n v e s t i g a t i o n can be useful for studying the e f f e c t of processes that were neglected or held constant i n the e a r l i e r tests of the theory, but which seem to be impor-tant i n a given a p p l i c a t i o n . M. Z e l d i t c h , J r . , 'Can you r e a l l y study an army i n a laboratory?', i n A. E t z i o n i , Ed., A ^ S o c i o l o g i c a l Reader  in Complex Organizations, New York: Holt, Rinehart and Winston, 1969, Second e d i t i o n , pp. 528-539. 167 11. In CChapter 3, we addressed the issue of whether we had su c c e s s f u l l y i n -duced a marginal u t i l i t y function f o r the buttons, and suggested that some errors i n p r e d i c t i o n could probably be a t t r i b u t e d to a f a i l u r e of the manipulation f o r some subjects. Our design did not allow us to locate pi6:sS"'S!iib!j:eGfc§ '<&§&; were ..unaffected by the manipulation. ' 12. See, for example, Bibb Latane', E d i t o r , 'Studies i n S o c i a l Comparison', Supplement 1, Journal of Experimental S o c i a l ^Psychology, 1966. 13. A mundane example of t h i s i s the agreement required about payment when one engages a new babysitter. While there appears to be a s o c i a l stan-dard for the range- of pay acceptable to both p a r t i e s , terms can vary according to a b i l i t y to pay, perceived need of the s i t t e r , and a l t e r n a -tives a v a i l a b l e . 14. Kuhn, op. c i t . , 1966, p. 337. 15. See, for example, E.E. Jones, K.J. Gergen, P. Gumpert and J.W. Thibaut, 'Some conditions a f f e c t i n g the use of i n g r a t i a t i o n to influence per-sonal evaluation', Journal of Personality and S o c i a l Psychology, 1, 1965, pp. 613-623; and D. Bramel, 'Determinants of b e l i e f s about other people', Chapter 4 i n J . M i l l s , E d i t o r , Experimental S o c i a l Psychology, Toronto: Collier-MacMillan, 1969; H.H. Kelley and A. St a h e l s k i , 'Social i n t e r a c t i o n bases of cooperators' and competitors' b e l i e f s about others', Journal of Personality and S o c i a l Psychology, jL6_, 1970, pp. 66r91; and W. '.! Thorngate, 'Predictions, Attributions, and Evaluations of Behaviour i n Decomposed Games', Unpublished manuscript, Un i v e r s i t y of Alberta, Department of Psychology,1974 . 16. D. E l l s b e r g , 'Risk, ambiguity, and the Savage axioms', Quarterly Journal of Economics, _6, 1961, pp. 643-669; K.R. MacCrimmon, 'Descriptive and normative implications of the decision-theory postulates', i n K. Borch and J . Mossin, E d i t o r s , Risk and Uncertainty, London: Macmillan, 1968. 17. Examples of p r o b a b i l i s t i c choice experiments are: those of S i e g e l , and of Ofshe and Ofshe. See: S. S i e g e l , A.E. S i e g e l , and J . J . Andrews, Choice, Strategy and U t i l i t y , New York: McGraw H i l l , 1964. L. Ofshe, and R. Ofshe, U t i l i t y and Choice i n S o c i a l I nteraction, Englewood C l i f f s : P r e n t i c e - H a l l , 1970. 168 BIBLIOGRAPHY Books Adams, J.S. Inequity i n s o c i a l exchange. In: Advances i n Experimental  S o c i a l Psychology, Vol. 2, New York: Academic Press, 1965. Archibald, K. Strategic Interaction and C o n f l i c t . Berkeley: University of C a l i f o r n i a Press, 1966. Bartos, O.J. Towards a r a t i o n a l - e m p i r i c a l model of negotiations. In: S o c i o l o g i c a l Theories i n Progress, Edited by J. Berger, M. Z e l d i t c h , J r . and B. Anderson. Boston: Houghton M i f f l i n Co., 1972. Bartos, O.J. Simple Models of Small Group Behaviour. New York: Columbia University Press, 1967. Blalock, H.M. S o c i a l S t a t i s t i c s . McGraw-Hill Book Company, 1960. Blau, P.M. Exchange and Power i n S o c i a l L i f e . New York: John Wiley, 1964. Bramel, D. Determinants of b e l i e f s about other people. In: Experimental  Soc i a l Psychology, Edited by J. M i l l s . Toronto: Collier-MacMillan, 1969. Bruner, J.S., J.J. Goodnow and G.A. Austin. A Study of Thinking. New York: John Wiley and Sons, 1956. Burnstein, E. Cognitive factors i n behavioural interdependence. In: Experimental S o c i a l Psychology, Edited by J. M i l l s . Toronto: C o l l i e r MacMillan, 1969. Burnstein, E. and S. Katz. Group decisions involving equitable and optimal d i s t r i b u t i o n s of status. In: Experimental S o c i a l Psychology, Edited by C.G. McClintock. Toronto: Holt, Reinhart, 1972. Chipman, J.S. Stochastic choice and subjective p r o b a b i l i t y . In: Decisions,  Values and Groups, Edited by D. Willner, New York: Pergamon Press, 1960. Crano, W.D. and M.B. Brewer. P r i n c i p l e s of Research i n S o c i a l Psychology. New York: McGraw H i l l , 1973. Emerson, R.M. Exchange theory: Parts I and I I . In: S o c i o l o g i c a l Theories  i n Progress, Volume I I , Edited by J. Berger, M. Z e l d i t c h , J r . , and B. Anderson. Boston: Houghton-Mifflin, 1972. 169 Fouraker, L.E. and S. Siegel . Bargaining Behaviour. New York: McGraw H i l l Book Co., 1963. Goffman, E. The Presentation of Self i n Everyday L i f e . Doubleday Anchor Books, 1959. Harnett, D.L. and L.L. Cummings. Bargaining behaviour i n an asymmetric t r i a d . In: Soci a l Choice, Edited by B. Lieberman. Gordon and Breach Science Publishers, 1971. Hays, W.L. S t a t i s t i c s . New York: Holt, Rinehart and Winston, 1969. Homans, G.C. So c i a l Behaviour: Its Elementary Forms. New York: Harcourt, Brace and World, 1961. Jones, E.E. and J.W. Thibaut. Interaction goals as bases of inference i n interpersonal perception. In: Person Perception and Interpersonal Be- haviour , Edited by R. T a g i u r i and L. P e t r u l l o . Stanford: Stanford Uni-v e r s i t y Press, 1959. Kelley, H.H. A t t r i b u t i o n theory i n s o c i a l psychology. In: Nebraska Sym- posium on Motivation, Edited by D. Levine. University of Nebraska Press, 1967. Kelley, H.H. A classroom study of the dilemmas of interpersonal negotiations. In: Strategic Interaction and C o n f l i c t , Edited by K. Archibald. Berke-l e y : University of C a l i f o r n i a Press, 1966. Kelley, H.H. and D.P. Schenitzki. Bargaining. In: Experimental S o c i a l Psy- chology, Edited by C.G. McClintock, Toronto: Holt, Rinehart and Winston, 1972. Kelley, H.H. and J.W. Thibaut. Group Problem Solving. In: The Handbook  of S o c i a l Psychology, Volume IV, Edited by G. Lindzey and E. Aronson. Addison Wesley, 1969. Kuhn, A. The Study of Society. London: Tavistock Publications, 1966. Lee, W. Decision Theory and Human Behaviour. New York: John Wiley and Sons, 1971. McClintock, C.G. Game behaviour and s o c i a l motivation i n interpersonal set-t i n g s . In: Experimental S o c i a l Psychology. Edited by C.G.' McLintock^ Toronto:' Holt, Rinehart and Winston, 1972. ; MacCrimmon, K.R. Decision making among m u l t i p l e - a t t r i b u t e a l t e r n a t i v e s : a survey and consolidated approach. Memorandum, RM-4823-ARPA, Rand Corporation, 1968. 170 MacCrimmon, K.R. Descriptive and normative implications of the d e c i s i o n -theory postulates. In: Risk and Uncertainty, Edited by K. Borch and J. Mossin. MacMillan, 1968. Michener, H.A. and R.W. Suchner. The t a c t i c a l use of s o c i a l power. In: The S o c i a l Influence Processes, Edited by J.T. Tedeschi. Chicago: Aldine Publishing Company, 1972. Schelling, T.C. The Strategy of C o n f l i c t . Cambridge: Harvard University Press, 1963. Siegel, S., A.E. Siegel and J.M. Andrews. Choice, Strategy and U t i l i t y . New York: McGraw-Hill, 1964. Thibaut, J.W. and H.H. Kelley. The S o c i a l Psychology of Groups. New York: John Wiley, 1959. Vinacke, W.E. Negotiations and decisions i n a p o l i t i c s game. In: So c i a l Choice, Edited by B. Lieberman. Gordon and Breach Science P u b l i c a t i o n s , 1971. Walton, R.E. and R.B. McKersie. A Behavioural Theory of Labour Negotiations. New York: McGraw-Hill, 1965. Weinstein, E.A. The development of interpersonal competence. In: Handbook  of S o c i a l i z a t i o n , Edited by D. Goslin. Rand McNalley, 1968. A r t i c l e s Becker, S.W. and F.O. Brownson. What p r i c e ambiguity? or the r o l e of ambi-guity i n dec i s i o n making. Journal of P o l i t i c a l Economy, 12_ (1964), pp. 62-73. Bixenstein, V.E., L. Potash and K.V. Wilson. E f f e c t s of l e v e l of coopera-t i v e choice by the other player on choices i n a Prisoner's Dilemma game. Part I. Journal of Abnormal and S o c i a l Psychology, 66^ (1963), pp. 308-313. Bixenstein, V.E., L. Potash and K.V. Wilson. E f f e c t s of l e v e l of coopera-t i v e choice by the other player on choices i n a Prisoner's Dilemma game. Part I I . Journal of Abnormal and So c i a l Psychology, 6i7_ (1963), pp. 139-147. Bixenstein, V.E., N. Chambers and K.V. Wilson. E f f e c t of asymmetry i n pay-of f on behaviour i n a two-person, non-zero-sum game. Journal of Con-f l i c t Resolution, £ (1964), pp. 151-159. 171 Blau, P.M. J u s t i c e i n s o c i a l exchange. S o c i o l o g i c a l Inquiry, 3_4 (1964), pp. 193-206. Blumstein, P.W. Audience,Machiavellianism, and t a c t i c s of i d e n t i t y bargain-ing. Sociometry 36 (1973), pp. 346-365. Chertkoff, J.M., and M. Conley. Opening o f f e r and frequency of concessions as bargaining s t r a t e g i e s . Journal of Personality and S o c i a l Psycholo-gy, 7_ (1967), pp. 181-185. Conrath, D.W. Sex r o l e and cooperation i n the game of Chicken. Journal of C o n f l i c t Resolution JJ> (1972), pp. 433-442. Cummings, L.L., D.L. Harnett and W.C. Hamner. Personality, bargaining s t y l e , and payoff i n b i l a t e r a l monopoly bargaining among European managers. Sociometry 36 (1973), pp. 325-344. Cummings, L.L. and D.L. Harnett. Bargaining behaviour i n a symmetric t r i a d : the r o l e of information, communication, power and r i s k - t a k i n g propen-s i t y . Review of Economic Studies 36 (1969), pp. 484-499. E l l s b e r g , D. Risk, ambiguity and the Savage axioms. Quarterly Journal of Economics 7_5 (1961), pp. 643-669. Emerson, R.M. Power-dependence r e l a t i o n s . American S o c i o l o g i c a l Review 17 (1962), pp. 31-41. F e l l n e r , W. D i s t o r t i o n of subjective p r o b a b i l i t i e s as a reaction to uncer-t a i n t y . Quarterly Journal of Economics 75 (1961), pp. 670-689. Fischer, C.S. The e f f e c t s of threats on an incomplete information game. Sociometry 32 (1969), pp. 301-314. Gallo, P.S. and C.G. McClintock. Cooperative and competitive behaviour i n mixed-motive games. Journal of C o n f l i c t Resolution 9_ (1965), pp. 68-79. Harsanyi, J.C. Bargaining i n ignorance of the opponent's u t i l i t y function. Journal of C o n f l i c t Resolution 6^  (1962), pp. 29-38. Hoffman, P., L. Festinger and D.H. Lawrence. Tendencies toward group com-p a r a b i l i t y i n competitive bargaining. Human Relations 7 (1954), pp. 141-159. Homans, G.C. S o c i a l behaviour as exchange. American Journal of Sociology 63 (1958), pp. 597-606. 172 Jones, E.E., K.J. Gergen, P. Gumpert and J.W. Thibaut. Some conditions a f f e c t i n g the use of i n g r a t i a t i o n to influence personal evaluation. Journal of Personal-ity and S o c i a l Psychology 1_ (1965), pp. 613-626. Jones, B., M. Steele, S. Gahagan and J. Tedeschi. Matrix values and cooper-ative behaviour i n the Prisoner's Dilemma game. Journal of Personal-i t y and Soc i a l Psychology j5 (1968), pp. 148-153. Kahan, J.P. E f f e c t of l e v e l of a s p i r a t i o n i n an experimental bargaining s i t -uation. Journal of Personality and S o c i a l Psychology 8^  (1968), pp. 154-159. Kelley, H.H., L.L. Beckman and C.S. Fischer. Negotiating the d i v i s i o n of a reward under incomplete information. Journal of Experimental So c i a l Psychology _3 (1967), pp. 361-398. Kelley, H.H. and A. Stahels k i . S o c i a l i n t e r a c t i o n bases of cooperators' and competitors' b e l i e f s about others. Journal of Personality and Soci a l Psychology 16. (1970), pp. 66-91. Kelley, H.H. and A.J. Stahelski. The inference of intentions from moves i n the Prisoner's Dilemma game. Journal of Experimental S o c i a l Psychol-ogy 6 (1970), pp. 401-419. Komorita, S.S. Cooperative choices i n a Prisoner's Dilemma game. Journal of Personality and S o c i a l Psychology 2_ (1965), pp. 741-745. Laing, J.D. and R.J. Morrison. Sequential games of status. Behavioural Science 19 (1974), pp. 177-196. Latanef, B. ( E d i t o r ) . Studies i n S o c i a l Comparison. Supplement 1, Journal of Experimental S o c i a l Psychology (1966). Lanzetta, J.T. and V.T. Kanareff. Information cost, amount of payoff, and l e v e l of a s p i r a t i o n as determinants of information seeking and decision making. Behavioural Science ]_ (1962), pp. 459-73. Lich t e n s t e i n , S. Bases for preference among three-outcome bets. Journal of Experimental Psychology 69 (1965), pp. 162-169. Liebert, R.M., W.P. Smith, J.H. H i l l and M. K i e f f e r . The e f f e c t s of i n f o r -mation and magnitude of i n i t i a l o f f e r on interpersonal negotiation. Journal of Experimental S o c i a l Psychology k_ (1968), pp. 432-441. McClintock, C.G. and S.P. McNeel. P r i o r dyadic experience and monetary r e -wards as determinants of cooperative and competitive game behaviour. Journal of Personality and Soc i a l Psychology J> (1967), pp. 282-294. 173 McClintock, C.G. and S.P. McNeel. Reward and score feedback as determinants of cooperative and competitive game behaviour. Journal of Personality and So c i a l Psychology 4_ (1966), pp. 606-613. Murdoch, P. and D. Rosen. Norm formation i n an interdependent dyad. Socio-metry 33 (1970), pp. 264-276. Nagel, T. Hobbes on Obligation. Philosophical Review 68 (1959), pp. 68-83. Oskamp, S. and D. Perlman. Factors a f f e c t i n g cooperation i n a Prisoner's Dilemma game. Journal of C o n f l i c t Resolution 9_ (1965), pp. 359-374. Patchen, M. A conceptual framework and some empirical data regarding com-parisons of s o c i a l rewards. Sociometry 24_ (1961), pp. 136-156. Pritc h a r d , R.D. Equity theory: A review and c r i t i q u e . Organizational Be-haviour and Human Performance h_ (1969), pp. 176-211. Rosen, S. The comparative role s of informational and material commodities i n interpersonal transactions. Journal of Experimental S o c i a l Psycho-logy 2 (1966), pp. 211-226. Sheposh, J.P. and P.S. Gallo. Asymmetry of payoff structure and cooperative behaviour i n a Prisoner's Dilemma game. Journal of C o n f l i c t Resolution' 17 (1973), pp. 321-333. Shubik, M. Games of status. Behavioural Science 1J6 (1971), pp. 117-129. Sl o v i c , P. and S. Lichte n s t e i n . The r e l a t i v e importance of p r o b a b i l i t i e s and payoffs i n r i s k taking. Journal of Experimental Psychology, Mono- graph Supplement No. 3, Part 2 (1968), pp. 1-18. Tedeschi, J.T. Start e f f e c t and response bias i n the Prisoner's Dilemma game. Psychonomic Science 11_ (1968), pp. 149-50. Thibaut, J. and C. Faucheux. The development of contractual norms i n a bargaining s i t u a t i o n under two types of s t r e s s . Journal of Experimental S o c i a l Psychology 3. (1965), pp. 89-102. Thibaut, J.W. and C L . Gruder. The formation of contractual agreements between p a r t i e s of unequal power. Journal of Personality and S o c i a l Psychology 11 (1969), pp. 59-65. Vinacke, E. Sex ro l e s i n 3 person games. Sociometry 22^  (1959), pp. 343-360. Weick, K.E. and B. Nesset. Preferences among forms of equity. Organizational Behaviour and Human Performance 3_ (1968), pp. 400-409. 174 Webster, M. and J . Kervin, ' A r t i f i c i a l i t y i n Experimental Sociology', Cana-dian Review of Anthropology and Sociology, 8, 1971, pp. 268-276. Weinstein, E.A—,: and P. Deutschberger. Tasks, bargains and i d e n t i t i e s i n s o c i a l i n t e r a c t i o n . S o c i a l Forces 42_ (1964), pp. 451-456. Whittemore, I.C. The competitive consciousness. Journal of Abnormal and S o c i a l Psychology 20 (1925-26), pp. 17-33? Yukl, G. E f f e c t s of the Opponent's i n i t i a l o f f e r , concession magnitude, and concession frequency i n bargaining behaviour. Journal of Personality and S o c i a l Psychology 30 (1974), pp. 232-335. Unpublished Material ' Burgess, R. and J.D. Gregory. Equity and inequity i n exchange r e l a t i o n s : an experimental re-examination of d i s t r i b u t i v e j u s t i c e . Paper pre-sented at the annual meetings of the West Coast Conference for Small Groups Research, Honolulu, Hawaii, 1971. Emerson, R.M. Power and p o s i t i o n i n exchange networks. Paper presented at national meetings, American S o c i o l o g i c a l Association, 1971. Foddy, W.H. The formation of cliques i n c o l l e c t i v i t i e s as a consequence of i n i t i a l d i s t r i b u t i o n s of dimensions of wealth. Unpublished Ph.D. D i s s e r t a t i o n , University of B r i t i s h Columbia, Canada, 1971. Foddy, W.H. On getting through to some of the people some of the time. Unpublished Manuscript. University of Alberta, Canada, 1972. Leik, R.K., R.M. Emerson and R.L. Burgess. The emergence of s t r a t i f i c a t i o n i n exchange networks: An experimental demonstration. Paper presented at the West Coast Conference f o r Small Group Research, San Diego, 1968. I n s t i t u t e for S o c i o l o g i c a l Research: University ;of Washington, Seattle. Michener, H.A. and M. Lyons. Perceived support and upward mobility as deter-minants of revolutionary c o a l i t i o n behaviour. Unpublished paper, Uni-v e r s i t y of Wisconsin (undated). 175 APPENDIX I LABORATORY SET-UP A diagram of the booths i s shown i n Fugure A . l Figure. A. 1 Main features of booths used i n experiments View from ins i d e a v i s i b l e or non-visible subject's booth View from outside a v i s i b l e ' s booth View from outside a non-visible's booth 176 1. Window covered with loudspeaker mesh. When room was illuminated from the center, subjects could see through t h e i r own screen to the center of the tables, but could not see through both t h e i r own and another booth's windows. 2. Card t e l l i n g subject what he had to begin with (numbers of each colour of buttons). 3. Subject's i d e n t i f i c a t i o n l e t t e r (also on front of booth). 4. A four inch gap allowed subjects to reach h i s buttons p i l e d i n front of the booth. 5. Card t e l l i n g subject he could not o f f e r more than 100 buttons at a time, although he could ask for more than 100, and give more than 100. 6. Table i n d i c a t i n g the worth of d i f f e r e n t numbers of buttons f o r the second part of the game. The i n s t r u c t i o n s drew attention to the fact that the table implies a p r i n c i p l e of diminishing marginal u t i l i t y . 7. Card on outside of v i s i b l e subjects' booth, t e l l i n g the exact number of buttons of each colour that subject started with ( t h i s could be seen by a l l other subjects around the t a b l e ) . 8. Coloured tag on the outside of non-visibles booths; the colour was the same as the colour of buttons i n that person's predominant resource p i l e . 9. An i l l u s t r a t i o n of the covers that were placed over the resources of non- v i s i b l e subjects. Figure A.2 I n i t i a t i o n forms used i n Set A I n i t i a t o r ' s l e t t e r Offer directed t o _ _ ( l e t t e r ) ( c i r c l e one) W i l l give (No.) red green buttons for (No.) red green buttons ( c i r c l e one) Receiver c i r c l e s one accepted rej ected j 17.7 Figure A, 3 I n i t i a t i o n forms used i n Set B and i n p i l o t study I n i t i a t o r ' s l e t t e r Offer directed to ( c i r c l e one) W i l l give (No.) red green buttons for (No.) red . green buttons ( c i r c l e one) RECEIVER CIRCLES ONE Accepted Rej ected o f f e r not good enough ; don't want to deal with you JFigure A.4 i l l u s t r a t e s the card pinned in s i d e a subject's booth, i n d i c a t i n g ithe number of buttons he had to begin with. jpjgure A.4 Card showing to subject his resource base. 1600 Red 30 Green The following page shows the value chart that was pinned in s i d e each booth. A somewhat d i f f e r e n t chart was used i n Set A, i n which the exact values for the function Y = 100 . X/2 (Y i s value, X i s number of buttons) was used. These values were rounded off to the nearest number for use i n Set A. In Set B and i n the p i l o t study, the numbers i n the l e f t column were further rounded, so that the smallest increment would be 5 value u n i t s . 178 Figure A.5 Value chart used for Set B and P i l o t 1400 2590 1380 2575 1360 2560 1340 2545 1320 2530 1300 2515 1280 2500 1260 2480 1240 2460 1220 2440 1200 2420 1180 2400 1160 2380 1140 2360 1120 2340 1100 2320 1080 2300 1060 2280 1040 2260 1020 2240 1000 2220 980 2200 Total number of 960 2180 buttons of a 940 2160 given colour. 920 2140 (Notice that 900 2120 the increments 880 2095 on t h i s side 860 2070 are a l l equal) 840 2045 820 2020 800 1995 780 1970 760 1945 740 1920 720 1895 700 . 1870 680 1845 660 1820 640 1790 620 1760 600 1730 580 1700 560 1670 540 1640 520 1610 500 1580 480 1550 Net worth of t o t a l number of buttons of a given colour i n value units f o r the second phase of the experiment. (Notice that the increments on t h i s side are smaller at the top than at the bottom) 179 Figure A.5 (Continued) 460 1520 440 1485 420 1450 400 1415 380 1380 360 1340 Total number of 340 1300 Net worth of t o t a l buttons of a 320 1260 number of buttons of given colour. 300 1220 a given colour i n (Notice that 280 1180 value units for the the increments 260 1140 second phase of the on t h i s side 240 1100 experiment. are a l l equal) 220 1050 (Notice that the 200 1000 increments on t h i s side 180 950 are smaller at the top 160 900 than at the bottom) 140 840 120 775 100 710 80 630 60 550 40 450 20 320 Base l i n e zero 0 - Base l i n e zero APPENDIX II EXPERIMENTAL INSTRUCTIONS AND . QUESTIONNAIRES, SET A The i n s t r u c t i o n s f o r Set A were administred by means of cassette tape recorder. The following i s a t r a n s c r i p t of the in s t r u c t i o n s employed. "Hi there. Thanks f or turning up to take part i n the experiment. You're going to play a game c a l l e d "Exchange and Bu i l d " , and as the name suggests, there w i l l be two parts to i t . The i n s t r u c -tions I ' l l give you now w i l l only be concerned with the f i r s t part, and we'll forget about the second part u n t i l l a t e r . During t h i s part of the game, you are going to be trading, or exchanging buttons with one another, and the object of the f i r s t part of the game i s to b u i l d up the small p i l e of buttons i n front of you, without l o s i n g too many buttons from the large p i l e you have. In other words, the object i s to increase the number of buttons of which you have the least at the moment without l o s -ing too many buttons of the colour of which you have the most. Now, y o u ' l l need to do t h i s because i n the next part of the game the two colours are used f o r completely d i f f e r e n t purposes. You w i l l need buttons of both colours, red and green, i n the next part; that i s , both colours are valuable. You w i l l probably have noticed that h a l f of you have your p i l e s of buttons out where everyone can see them, while h a l f of you have covers over the buttons. Now the covers have been placed there so that some of you w i l l not know how many buttons some others have. This does not mean that the people with covers have no buttons — they do. And you can t e l l which colour they have most of by the t i c k e t on the upper r i g h t hand corner of t h e i r booths — for example, a green t i c k e t means that person has a predominant p i l e of green buttons, and a smaller p i l e of red ones. If y o u ' l l look at the table on the side of your screen, you w i l l n o tice that there are two columns of figures there. The column of figures on the l e f t r e f e r s to d i f f e r e n t sized p i l e s of buttons of a given colour. The column of figures on the ri g h t t e l l s you how much these d i f f e r e n t sized p i l e s of buttons of a given colour w i l l be worth i n the next part of the game. So the column on the l e f t t e l l s you how many buttons, and the column on the ri g h t indicates values. O.K.? 181 Now i f you look c l o s e l y at the figures i n the columns, you w i l l n o tice that the figures on the l e f t increase by 20 at a time — so they go: 20, 40, 60, 80, and so on r i g h t up to 1,600. The f i g -ures on the r i g h t , however, increase i n b i g jumps to begin with, and the jumps get smaller and smaller as you go from the bottom to the top of t h i s column. Because the figures i n the two columns increase ±i d i f f e r e n t ways, the table t e l l s us two very important things. The f i r s t thing i t t e l l s you i s that i f you've already got a l o t of buttons of a given colour, 20 more would be worth less to you than i f you only had a few buttons of that given colour. Let me show you how th i s works. Say you had a p i l e of 1,580 green buttons. You can see that a p i l e of 1,580 green buttons would be worth 2,806 value units i n the next part of the game. Now, i f you got 20 more buttons (green ones), t h i s would bring you up to 1,600, and a p i l e of 1,600 buttons i s worth 2,826; so that you would have gained 20 value u n i t s . In other words, 20 more green buttons when you've already got 1,580 would be worth 20 value u n i t s . If you only had a p i l e of 200 green buttons to begin with, though, and you got 20 more, you'd f i n d that 200 green buttons (what you started with) would be worth 1,000 value units i n the next part of the game, and a p i l e of 220, that i s , the 200 plus 20 more, i s worth 1,049. So the 20 extra i n t h i s case would be worth 49 value u n i t s . Remem-ber, when you had 1,580, 20 extra are worth 20, but when you've only got 200, 20 extra are worth 49. Once again, the f i r s t point i s that the more buttons you have of a given colour, the les s worth 20 extra would be. This i s l i k e saying that $20 i s worth l e s s to a m i l l i o n a i r e than, say, to a person on welfare. The second thing the table t e l l s you i s that i f you have a l o t of buttons of one colour, and only a few of the other colour, you w i l l a c t u a l l y increase the t o t a l worth of your buttons every time you exchange some of the buttons of which you have most, for some of the buttons of which you have the l e a s t . Let me show you how that works. I f , say, you had 1,600 green buttons; you f i n d they are worth 2,826 value units i n the next part of the game. O.K.? And i f that was a l l you had, you decide to exchange some of the green buttons f o r some of the red buttons, so that you would end up with 800 green buttons, and perhaps 800 red buttons. And you f i n d that a p i l e of 800 green buttons would be worth — we l l , have a look at i t on your table — 2,000 value u n i t s . So your buttons 1 8 2 would now be worth 2,000 value units f or the green p i l e , and 2,000 value units for the red p i l e , so that the t o t a l worth of your red and green buttons would be worth 4,000 value u n i t s . Whereas the p i l e of 1,600 green buttons alone was worth 2,826 value u n i t s — two p i l e s — one of green and one of red, 800 each, would be worth 4,000 value u n i t s . To emphasize t h i s second point then, you a c t u a l l y increase the value of your buttons by exchanging. In addition, since b i g p i l e s are of course better than small p i l e s , y o u ' l l even be better o f f i f you can pick up a few buttons while you are exchanging — that i s , i f you can get the others to give you a few more i n return than you have given them. Of course, you may f i n d t h i s d i f f i c u l t to do as the others might not l i k e the idea. The table doesn't contain enough d e t a i l f o r you to make precise c a l c u l a t i o n s . I t ' s rather intended to give you an idea of how the buttons are valued. If you look through your screen you w i l l notice a l e t t e r printed on the ins i d e of your own screen. This i s to i d e n t i f y you. I ' l l j u s t run through the steps involved i n the s i n g l e opportun-i t y f o r exchange now, to give you a better idea of what you're going to do during each opportunity for exchange. Remember, you are going to have a number of these opportunities for exchange. F i r s t of a l l , y o u ' l l look through your screens to see what the others have, or what t h e i r predominant colour i s , and decide whether you want to send an o f f e r to.one of the others. Now, you do not have to send an o f f e r i f you don't want to, O.K.? So that i f you decide that you want to send an o f f e r , then you'd f i l l out one of the forms i n front of you, remembering one thing, that you cannot send an o f f e r of more than the l i m i t that i s written on the card on the lower bar of your screen. While you are not allowed to send an o f f e r of more than 100 buttons, you may request more than t h i s from others, and should you be asked for more than 100 buttons, you may agree to do so. You simply cannot i n i t i a t e , or begin by o f f e r i n g more than the 100 button l i m i t . You can of course ask for le s s than or up to the l i m i t from the person you send your o f f e r to, and give l e s s than or more than the l i m i t i n return. Once you've done t h i s , f i l l out the forms, count out the buttons, and put both the form and the buttons i n the bowl i n front of you. When everyone has done t h i s who wants to, I d e l i v e r a l l the bowls to the people they are addressed to — that i s , to the booth they are addressed to. It i s c l e a r that while your bowl i s around at someone else's booth, either one or more bowls may come around to you. You can accept one but only one. If you accept an o f f e r , c i r c l e 'Accepted' on - 183 the form that came with the bowl. Any other o f f e r s that you r e -ceive and decide not to accept, c i r c l e 'Rejected' on the form that came with them. When everyone has done t h i s , I ' l l ask those who have c i r c l e d 'Accepted' on an o f f e r to take the buttons that came with that o f f e r , and to count up the buttons that they were requested to give i n return. Put these buttons i n the bowl. I w i l l then return the bowls to t h e i r owners. And we w i l l then be ready to begin the next opportunity for exchange. I would j u s t l i k e to be clear on one point: during each oppor-tunity for exchange, two things are happening. Somebody might be accepting or r e j e c t i n g an o f f e r from you, and at the same time you may be accepting an o f f e r from someone else, or r e j e c t -ing o f f e r s . O.K.? Now, those people whose buttons are out i n the open should leave them there — do not t r y to hide them, or haul them o f f behind your screens. There i s some yellow scratch paper i n front of you, i f you want to keep track of how many buttons you have. The numbers you are beginning with, that i s , the s i z e of your p i l e s , are written on a small card on the lower part of your screen. The f i r s t part of the game w i l l take more time than the second, and you w i l l have plenty of time to make a l l the exchanges you want. I'd l i k e to ask you too, please not to cheat. Count out any buttons you are o f f e r i n g accurately and observe the l i m i t s i n making your o f f e r s . O.K.? If you would j u s t l i k e to look through your screens now, decide i f you want to send an o f f e r to any of the otheis during the f i r s t opportunity to exchange, we can begin. Verbal addition, not included i n tape: You should note that some of you have large p i l e s of buttons, and some of you have very large p i l e s . If you cannot see t h i s , per-haps i f you lean out closer to your screens, you w i l l get a bet-ter view of the p i l e s i n front of the others' booth. Be c a r e f u l that you do not lean back and look around at the person on either side of you. Any further questions?" End of Instructions. 184 Appendix I I : Set A Questionnaire given twice to subjects with v i s i b l e resource p i l e s , a f t e r S/s had made o f f e r s on T r i a l 1 and on T r i a l 2, but p r i o r to d e l i v e r y of o f f e r s . Your l e t t e r 1. Who d i d you j u s t make an o f f e r to? G H I J K L M N ( c i r c l e one) red red 2. What was the of f e r ? buttons offered f o r buttons green green 3. What are your reasons f o r the s i z e of o f f e r you made? 4. How l i k e l y do you think he i s to accept your offer? ( C i r c l e the p o s i -t i o n on the l i n e below that shows how l i k e l y i t i s your o f f e r w i l l be accepted) / / / / ; / extremely quite 50-50 not very not at a l l l i k e l y l i k e l y chance l i k e l y l i k e l y What are your reasons f o r thinking t h i s ? Do you think the person you made the o f f e r to i s more l i k e l y than any of the others to accept your o f f e r ? Yes No ( c i r c l e one) If you said NO, who do you think i s more l i k e l y to accept? Why do you think t h i s ? Questionnaire given twice to subjects with non-visible .resource p i l e s , a f t e r S's had made of f e r s on T r i a l 1 and on T r i a l 2: p r i o r to d e l i v e r y of o f f e r s . Your l e t t e r 1. Who did you ju s t make an o f f e r to? G H I J K L M N ( c i r c l e one) red red 2. What was the of f e r ? buttons offered f o r buttons green green 185 3. What are your reasons f o r the s i z e of o f f e r you made? 4. How l i k e l y do you think he i s to accept your o f f e r ? ( C i r c l e the p o s i -t i o n on the l i n e below that shows how l i k e l y i t i s your o f f e r w i l l be accepted) / / / / / extremely quite 50-50 not very not at a l l l i k e l y l i k e l y chance l i k e l y l i k e l y What are your reasons f o r thinking thi s ? Do you think the person you made the o f f e r to i s more l i k e l y than any of the others to accept your offer? Yes No ( c i r c l e one) If you said NO, who do you think i s more l i k e l y to accept? Why do you think t h i s ? For those of you who have covers over your buttons: If you are given the opportunity on the t h i r d t r i a l to remove the covers from your buttons, would you choose to do so? Yes No ( c i r c l e one) Questionnaire given to subjects with n o n - v i s i b l e resources, a f t e r f a l s e o f f e r s had been given to v i s i b l e s , T r i a l 2. 1. Did you see any advantage or disadvantage ( c i r c l e one) i n having covers over your buttons? What kind of advantage or disadvantage? 2. What do you think would be the long term e f f e c t s of having covers on your buttons, i f you continued to play f o r several t r i a l s ? No.te: On the questionnaires given to subjects, there were no references to i d e n t i f y the questionnaires as being for v i s i b l e s only, or for n o n r-visibles only. S u f f i c i e n t space was provided for r e p l i e s . 186 APPENDIX I I I EXPERIMENTAL INSTRUCTIONS AND QUESTIONNAIRES USED IN SET B The Instructions used for Set B were administered by means of a cassette tape recorder, while subjects read a t r a n s c r i p t of the tape. Below i s a copy of the t r a n s c r i p t given to each subject i n Set B. Thanks f o r turning up to take part i n the experiment. You are going to play a game c a l l e d "Exchange and Build", and as the name suggests, there w i l l be two parts to i t . The i n s t r u c -tions I ' l l give you now w i l l only be concerned with the f i r s t part, and we'll forget about the second part u n t i l l a t e r . Exchange and Build i s the sort of game i n which some of you may do better than others. You should t r y to do as well as you can. At the end of the second part of the game, the four players who have done the best w i l l be declared the winners. While you may fin d i t a b i t d i f f i c u l t to see how well you are doing during t h i s f i r s t part of the game, you w i l l be able to see t h i s more c l e a r l y during the second part. During t h i s part of the game, you are going to be trading, or exchanging buttons with one another, and the objects of the f i r s t part i s to b u i l d up the small p i l e of buttons i n front of you, without l o s i n g too many buttons from the large p i l e you have. In other words, the object i s to increase the number of buttons of which you have the least at the moment, but not lose too many of the colour of which you have the most. Now, y o u ' l l need to do t h i s because i n the next part of the game the two colours are used f o r completely d i f f e r e n t purposes. You w i l l need buttons of both colours, red and green, i n the next part; that i s , both colours are valuable. You are a l l beginning with some red and some green buttons. How many you have of.each colour i s written on a small card on the lower bar of your screen ( i n s i d e ) . As you can see, h a l f of you have your buttons out where everyone can see them, while h a l f of you have covers over your buttons. I ' l l stop for a few seconds while you look through your screens. The players out i n the open have a sign on the outside top of t h e i r screens, that shows you exactly how many red and how many green buttons they have. The covers have been placed on four other booths so that you w i l l not know how many buttons these players have. I can t e l l you that none of these players has a t o t a l wealth of 1,200 buttons. That i s , one of them has the same number of buttons as the players without covers — some of them have a larger t o t a l number of buttons (more than 1,200), and some have a smaller t o t a l number (less than 1,200). You can, however, t e l l which colour players with covers have most of, by the t i c k e t on the upper r i g h t hand corner of t h e i r booths —- for example, a red t i c k e t means that player has more red buttons than green ones. If y o u ' l l look at the table on the side of your screen, you w i l l notice that there are two columns of figures there. The column of figures on the l e f t r e f e r s to d i f f e r e n t sized p i l e s of buttons of a given colour. The column of figures on the r i g h t t e l l s you how much these d i f f e r e n t sized p i l e s of buttons of a given colour w i l l be worth i n the next part of the game. So the column of f i g -ures on the l e f t t e l l s you how many buttons, and the column on the r i g h t indicates values. O.K.? Now i f you look c l o s e l y at the figures i n the columns, you w i l l notice that the figures on the l e f t increase by 20 at a time — so they go 20, 40, 60, 80, and so on r i g h t up to 1,_600. The figures on the r i g h t , however, increase i n big jumps to begin with and the jumps get smaller and smaller as you go from the bottom to the top of t h i s column. Because the figures i n the two columns increase i n d i f f e r e n t ways, the table t e l l s us two very important things. The f i r s t thing i t t e l l s you i s that i f you've already got a l o t of buttons of a given colour, 20 more would be worth less to you than i f ' y o u only had a few buttons of that given colour. Let me show you how t h i s works. Say you had a p i l e of 1,080 green buttons. You can see that a p i l e of 1,080 green buttons would be worth 2,300 value units i n the next part of the game. Now, i f you got 20 more buttons (green ones), t h i s would bring you up to 1,100, and a p i l e of 1,100 but-tons i s worth 2,320; so that you would have gained 20 value u n i t s . If you only had a p i l e of 200 green buttons to begin with though, and you got 20 more, you'd f i n d that 200 green buttons (what you started with) i s worth 1,000 value units i n the next part of the game, and a p i l e of 220, that i s , the 200 plus 20 more, i s worth 1,050. So the 20 extra i n t h i s case would be worth 50 value units Once again, the f i r s t point i s that the more buttons you have of a given colour, the le s s worth 20 extra would be. This i s l i k e saying that $20 i s worth less to a m i l l i o n a i r e than, say, to a person on welfare. 188 The second thing the table t e l l s you i s that i f you have a l o t of buttons of one colour, and only a few of the other colour, you w i l l a c t u a l l y increase the t o t a l worth of your buttons every time you exchange some of the buttons of which you have most, for some of the buttons of which you have the l e a s t . Let me show you how that works. I f , say, you had 1,200 green buttons:, you f i n d they are worth 2,420 value units i n the next part of the game. O.K.? And i f that was a l l you had, you decide to exchange some of the green buttons for some of the red ones, so you might end up with perhaps, 600 green buttons, and 600 red ones. And you f i n d that a p i l e of 600 buttons would be worth — well, have a look on your table — 1,730 value u n i t s . So your buttons would be worth 1,730 for^ the green p i l e , and 1,7.30 for the. red, so that the t o t a l worth of your red and green buttons would be 3,460 value u n i t s . Whereas the p i l e of 1,200 green buttons alone was worth 2,420, two p i l e s , one of green and one of red (600 each) would be worth 3,460. To emphasize t h i s second point then, you a c t u a l l y increase the  t o t a l worth of your buttons by exchanging. As I w i l l explain i n a minute, there i s nothing i n the rules that says you have to trade an equal number of buttons of one colour for an equal num-ber of the other colour. Depending on who you are t r y i n g to trade with, you may choose to o f f e r more, or le s s than you want in return. I ' l l j u s t run through the steps involved i n the s i n g l e opportun-i t y f o r exchange now, to give you a better idea of what you are going to be doing during each opportunity for exchange. You w i l l have j u s t f i v e of these opportunities — I ' l l repeat that — f i v e opportunities, to make trades. F i r s t of a l l , y o u ' l l look through your screens to remind your-selves what others have, or what t h e i r predominant colour i s , and decide who you want to send an o f f e r to. You do not have to send an o f f e r during each opportunity for exchange i f you do not want to. The l e t t e r s printed at the top of your screens are to iden-t i f y you. If you decide that you want to send an o f f e r , then f i l l out one of the forms i n front of you, remembering one thing: you cannot send an o f f e r of more than the l i m i t that i s written on the card on the lower bar of your screen. This means that you can o f f e r anything up to but not over ,100 at a time. There i s , however, no l i m i t on what you can 'give '/in return f or an o f f e r someone makes XOlU€ to you. There i s only a l i m i t on how many you can o f f e r at a time. 189 Once you have decided what you want to do, f i l l out a form, count out the buttons, and put both the form and the buttons i n the bowl i n front of you. When everyone has done t h i s who wants to, I w i l l d e l i v e r a l l the bowls to the players they are addressed to — that i s , to the booth they are addressed to. It i s clear that while your bowl i s around at someone else's booth, either one or more bowls can come around to you, or perhaps none w i l l . You can accept one but only one. Let me emphasize that — i f you receive more than one o f f e r , at a time, you cannot accept them a l l — you must choose one. If you accept an o f f e r , c i r c l e 'Accepted' on the form that came with the bowl. Any other o f f e r s that you receive and decide not to accept, c i r c l e 'Rejected' on the forms that came with them. There i s also a place for you to check a reason for r e j e c t i n g an o f f e r . This l e t s you t e l l the person who sent the o f f e r whether you would accept a better o f f e r , or whether you have decided not to deal with him. You may check one of these reasons i f you wish. Please do not write any other messages on the forms. You may r e j e c t a l l the o f f e r s you get i f you choose to do so. If you happen to get an o f f e r from the same person you sent one to, remember that they are separate and independent — reply to the o f f e r you received, and don't worry about the one you sent. When everyone has checked t h e i r forms, I ' l l ask those who have c i r c l e d 'Accepted' on an o f f e r to take the buttons that came with that o f f e r , and to count out the buttons that they were requested to give i n return. Put these buttons i n the bowl. I w i l l then return the bowls to t h e i r owners. And we w i l l then be ready to begin the next opportunity for exchange. This f i r s t part of the game w i l l take more time than the second, and you w i l l have enough time to make up to f i v e exchanges. I'd l i k e to ask you too, please not to cheat when you are counting out the buttons, and to observe the l i m i t of 100 on what you can o f f e r . If you would l i k e to look through your screens now, decide i f you want to send an o f f e r to any of the others during the f i r s t oppor-tunit y to exchange, we can begin. End of i n s t r u c t i o n s , Set B. 190 Appendix I I I : Set B Questionnaire given to a l l subjects a f t e r they had made T r i a l 1 o f f e r s . Questionnaire No. 1 Your l e t t e r • 1. Who did you j u s t make an o f f e r to? G H I J K L M N ( c i r c l e one) red red 2. What was the offer? buttons offered f o r buttons green green 3. Did you o f f e r to a person with a cover or without a cover on h i s buttons? no cover cover ( c i r c l e one) 4. Why did you choose the person you did? 5. What i s the main reason for the s i z e of o f f e r that you made? 6. How l i k e l y do you think he i s to accept your offer? ( C i r c l e the p o s i -t i o n on the l i n e below that shows how l i k e l y you f e e l i t i s that your o f f e r w i l l be accepted) / / / / / Extremely quite 50-50 not very not at a l l l i k e l y l i k e l y chance . l i k e l y l i k e l y What are your reasons f o r thinking t h i s ? 7. Do you know anyone e l s e who would be more l i k e l y to accept your o f f e r ? YES NO If you said YES, who do you think i s more l i k e l y to accept? Why do you think t h i s ? Questionnaire for subjects with v i s i b l e resource p i l e s , delivered a f t e r r e p l i e s to f a l s e o f f e r s on T r i a l 1 were c o l l e c t e d . Questionnaire No. 2: V i s i b l e s Your l e t t e r 1. If you accepted an o f f e r on the f i r s t trading opportunity: What was your main reason f o r accepting the o f f e r you did? 191 2. If you rej ected an o f f e r ( s ) on the f i r s t opportunity, what was your main reason f o r doing so? 3. There are probably advantages and disadvantages to having your buttons out i n the open. What do you see to be the advantages? What do you see to be the disadvantages? 5. If you are given a choice a f t e r the second exchange opportunity, of having the covers removed from those players who have them, would you choose to have t h i s done? YES NO ( c i r c l e one) Why? Questionnaire for subjects with n o n - v i s i b l e p i l e s , given a f t e r r e p l i e s to f a l s e o f f e r s on T r i a l 1 were c o l l e c t e d . Questionnaire No. 2: Non-Visibles Your l e t t e r 1. If you accepted an o f f e r on the f i r s t exchange opportunity: What i s the main reason for accepting the o f f e r you did? 2. If you rejected an o f f e r ( s ) on the f i r s t opportunity, what was your main reason f o r doing so. 3. There are probably advantages and disadvantages to having covers over your buttons. What do you see to be the advantages? What do you see to be the disadvantages? 4. If you are given a choice a f t e r the second exchange opportunity i n t h i s game to remove the covers from your buttons, would you choose to do so? YES NO ( c i r c l e one) Why? Note: On the questionnaires given to subjects, there were no references to i d e n t i f y the questionnaires as being for v i s i b l e s only, or for non^ v i s i b l e s only. S u f f i c i e n t space was provided for r e p l i e s . 192 APPENDIX IV PILOT WORK The r e s u l t s and analysis of Set A indicated that changes i n the paradigm were necessary to eliminate the confounding of wealth and v i s i b i l i t y , so that the e f f e c t s of v i s i b i l i t y alone could be examined. It also seemed desir a b l e to ensure more consistent motivation i n subjects concerning advantageous exchanges. The v a r i a b i l i t y i n motives i n Set A seemed at lea s t p a r t l y due to the complexity of the i n s t r u c t i o n s , which were impor-tant for the induction of scope conditions, assumptions and independent v a r i a b l e s . A f i r s t attempt to remedy the problems of Set A w i l l be described b r i e f l y i n t h i s Appendix. A second and more successful attempt was presented as Set B. The r e v i s i o n of the paradigm i n t h i s P i l o t work unfortunately introduced added factors that made i t an inadequate test f o r the theory. These factors w i l l be discussed i n the context of the r e s u l t s . The data from the P i l o t study are given here because they were informative i n terms of the l i m i t s of the theory, and suggested further r e v i s i o n s that led to Set B. Description of the Experiments The same physical set up was used as i n Set A. This time, instead of d i f -ferent resource bases, a l l subjects had equal sized t o t a l resource bases (1,200 buttons). The r a t i o s of red to green varied across subjects, so that t h e i r value positions on the two resources were not the same, as shown i n Figure A.6. Note that the value positions of non-visibles were d i f f e r e n t from those of v i s i b l e s . If the non-visibles knew they were unlike any v i s i b l e , i t was expected .they would also be uncertain about the numbers and r a t i o s of red and green possessed by other non - v i s i b l e s . Their value posi t i o n s on the red and green buttons were set between those of the unbalanced and the b a l -anced v i s i b l e s . If non-visibles d i d use t h e i r concealment to 'act l i k e ' the v i s i b l e s , they would have a model, i n the balanced v i s i b l e s , of a f a i r o f f e r that asked f o r more buttons i n return f o r l e s s . In addition, the experi-menter could compare o f f e r s of non-visibles with those of the balanced v i s i b l e s , to see i f the concealed players acted 'as i f they were balanced v i s i b l e s . A value chart s i m i l a r to that used i n Set A was employed (see Appendix I ) . Figure A. 6 D i s t r i b u t i o n of resources and information i n P i l o t Set Red - F ^ l Green I _ _ _ Cover I J J L V. 7j / i I / I I Unbalanced Green V i s i b l e Unbalanced Red V i s i b l e Balanced Green V i s i b l e Balanced. <". Red V i s i b l e NV Red, NV Red, NV Green, NV Green, 110G/100R 1000R/200G 900G/300R 850W350G 950R/250G 950G/250R VISIBLES NON-VISIBLES "194 V i s i b l e subjects had cards on the outside of t h e i r booths, showing the exact amount of each colour i n t h e i r possession; subjects were t o l d that the coloured tags on the non-visibles' booths indicated those persons' pre-dominant colour, and that the non-visible players might have more, or l e s s , or the same, as v i s i b l e s . Instructions were tape recorded and subjects were given a t r a n s c r i p t of the tape to read as the tape was playing. The i n s t r u c t i o n s were essen-t i a l l y the same as i n Set B. In addition to the elimination of differences i n t o t a l resource bases, described above, the major differences of t h i s Set from Set A were: 1) A time l i m i t was imposed. Subjects were t o l d they had only four oppor-t u n i t i e s ( t r i a l s ) on which to make trades, so that they would not f e e l free to spend too much time 'exploring',(to r e f l e c t the fact that i n the ' r e a l world', wasted i n i t i a t i o n s are c o s t l y ) . In f a c t , the time l i m i t gave subjects close to balance (balanced red v i s i b l e s and balanced green v i s i b l e s ) a game advantage. 2) The addition of a paragraph i n the i n s t r u c t i o n s that t o l d subjects that four winners would be declared at the end of the second part of the game, and these winners would be the four 'who had done best'. These i n s t r u c t i o n s were intended to increase the importance of outrank-ing others, and to make the d e f i n i t i o n of the s i t u a t i o n more uniform. 3) False o f f e r s were switched f o r the r e a l ones a f t e r subjects had f i l l e d out t h e i r f i r s t i n i t i a t i o n s on T r i a l 1. Only v i s i b l e s received phoney o f f e r s , i n the following form: — each balanced v i s i b l e received an o f f e r of 100 for 100 from the other balanced . ' v i s i b l e , and from a no n - v i s i b l e with the complemen-tary resource. — each unbalanced v i s i b l e received an o f f e r of 100 f o r 100 from the other unbalanced v i s i b l e , and from a no n - v i s i b l e with the comple-mentary resource. Thus, each v i s i b l e received two f a l s e o f f e r s that d i f f e r e d only i n the v i s i b i l i t y of the sender. 4) While subjects were responding to these f a l s e o f f e r s , the experimenter checked the word 'rejected' on a l l o f f e r s made on T r i a l 1. In addi-t i o n , the phrase 'offer not good enough' was ticked on i n i t i a t i o n forms of the v i s i b l e subjects, and the phrase 'don't want to deal with you' was checked on the no n - v i s i b l e s ' o f f e r s (see Appendix I, Figure A.3 for an i l l u s t r a t i o n of the i n i t i a t i o n form). It was intended that non-v i s i b l e s should be encouraged to interpret t h e i r r e j e c t i o n as the r e -s u l t of t h e i r concealment, to see what behavioural adjustments they 195 would make i n l a t e r i n i t i a t i o n s , and to compare t h e i r reaction to such a r e j e c t i o n with the reaction made by v i s i b l e s . (A more informa-t i v e procedure would have been to make some non-visibles believe they were being rejected on the basis of concealment, and some because of the types of o f f e r s they made.) Afte r the responses to the phoney o f f e r s were c o l l e c t e d , the experimenter switched back to the r e a l o f f e r s , d e l i v e r i n g r e j e c t i o n s l i p s to a l l subjects. Players then made a second i n i t i a t i o n , which, was delivered as addressed, the r e c i p i e n t s r e p l i e d to them, and the r e p l i e s were c o l l e c t e d . The experiment then ended with a d e b r i e f i n g . During the experiment, three questionnaires were administered (see the end of t h i s Appendix), asking why subjects chose a p a r t i c u l a r target, why they accepted the o f f e r s they d i d , t h e i r perception of the advantages of con-cealment and lack of concealment, and t h e i r preference for r e t a i n i n g covers over the buttons. In retrospect, i t seems that the written questions were excessive, and interrupted the flow of the exchange transactions. Results and Discussion of Results The r e s u l t s f or Hypothesis 1 were presented i n the main text, and w i l l not be repeated. 1. Tendency to make advantageous exchanges: Hypothesis 2. As with Set B, i t i s f i r s t necessary to define what constituted a f a i r ex-change i n terms of value u n i t s . In t h i s set, those subjects with highly unbalanced resource p i l e s gained more net value units i n an exchange of k for k buttons on the f i r s t one or two exchanges, because an addition to t h e i r very small resource p i l e r a p i d l y increased t h e i r marginal u t i l i t y ; the marginal cost of a decrease i n t h e i r large p i l e was r e l a t i v e l y much smaller than f o r players with more equal value positions on each resource. The unbalanced players should therefore have been w i l l i n g to give more for l e s s i n the i n i t i a l exchanges. For example, because the unbalanced green p r o f i l e was the most uneven, and the balanced red p r o f i l e was most even, a trade between these two would be f a i r only i f the balanced red player received many more buttons than did the unbalanced green player. A f a i r exchange was, of course, only definable i f the target person had v i s i b l e resource p i l e s . In the table below, we w i l l l i s t the exchange r a t -ios (number of buttons asked per 100 i n return)that would have l e d to equality of p r o f i t i n value u n i t s , and the obtained r a t i o s f o r t h i s set of experiments. The d i f f e r e n c e between the two r a t i o s i s included to give an i n d i c a t i o n of the s i z e of deviations. Table A . l F a i r Exchange Ratios and Average Exchange Ratios, for I n i t i a t i o n s to V i s i b l e s ( P i l o t Set, T r i a l 1) K . . ........ ,:> >. (Figures give number offered per 100 asked); I n i t i a t o r (n) Recipient Fair Ratio Average Obtained Ratio Difference (Obtained-Fair)  1. V i s i b l e s Unbalanced Green 10 2 Unbalanced Red Balanced Red 125 200 92 103 -33 -97 Unbalanced Red 8 4 Unbalanced Green Balanced Green 80 125 78 97 - 2 -28-Balanced Green 7 2 Unbalanced Green Balanced Red 80 105 82 105 + 2 0 Balanced Red 8 2 Unbalanced Green Balanced Green 50 95 74 83 +24 -12 2. Non-visibles Green 15 5 Unbalanced Red Balanced Red 85 117 87 96 + 2 -21 Red 14 7 Unbalanced Green Balanced Green 60 110 83 88 +23 -22 197 While the subjects' o f f e r s did not uniformly meet the d e f i n i t i o n of f a i r n e s s , i n terms of equal p r o f i t i n value units for trading partners, neither were they uniformly following a prominent r u l e of 'one button for one button' as indicated i n the deviations of the r a t i o s from the r a t i o of 100/100. In the case of the balanced v i s i b l e s , i n p a r t i c u l a r , the r a t i o s of the num-ber offered to number asked seems to r e f l e c t t h e i r lower marginal gain for a given increment of buttons, (due to t h e i r more balanced resource p i l e s ) . The nearness of the unbalanced green v i s i b l e s ' r a t i o s to .100/100 probably r e f l e c t s a reluctance to give up more than one gives away, given the game urges them to 'do w e l l ' . In addition, such a r e s u l t i s not inconsistent with studies of equity, i n d i c a t i n g that people are less l i k e l y to restore an exchange r e l a t i o n s h i p to balance when Other should get more, and P must give up rewards, than i f P i s the one who w i l l benefit from equity. The unbalanced v i s i b l e s may have perceived that so generous an o f f e r as two for one was not necessary to ensure a transaction. Conversely, the balanced red v i s i b l e s undoubtedly r e a l i z e d they would be u n l i k e l y to obtain two f o r one from the unbalanced green v i s i b l e s . Their exchange r a t i o i s lower than 100/100, which seems to indi c a t e a tendency for P to consider what 0 had when decid-ing how much to o f f e r and request i n return. I n i t i a t i o n s jto^ the balanced subjects showed a higher exchange r a t i o (were more generous) than to most others, i n d i c a t i n g that other players recognized the 'rig h t ' of the balanced players to demand more. (Both unbalanced v i s i b l e s and other subjects men-tioned i n written comments that the unbalanced players were the most 'des-perate'.) . In Table A . l , the o v e r a l l exchange r a t i o s are ordered according to the order-ing i f a l l were ' f a i r ' : balanced v i s i b l e -\ n o n - v i s i b l e C unbalanced v i s i b l e (81 < 87 < 89). However, i t i s also clea r from Table A . l that n o n - v i s i b l e subjects did not t r y to ask for an advantageous exchange more than d i d the v i s i b l e s . They were expected to send o f f e r s that would lead the r e c i p i e n t to think they were s i m i l a r i n resource p r o f i l e s to the balanced v i s i b l e s . Only the r a t i o of o f f e r s to balanced v i s i b l e s looks advantageous (less than the ' f a i r ' r a -t i o ) , and t h i s i s probably due in part to the low frequency of any r a t i o s grea-ter than 1.0. The r e s u l t s do not support Hypothesis 1. Possible reasons for t h i s w i l l be discussed a f t e r the r e s u l t s f o r the other hypotheses have been given. It should be noted that while subjects seemed to pay attention to marginal value of buttons i n assessing the f a i r n e s s of an exchange, there i s a con-founding of fairness i n a s i n g l e exchange, and i n the en t i r e game. Subjects i n the experiment, being a l l of equal t o t a l wealth, may have seen t h e i r ex-changes i n terms of the f i n a l Outcome when everyone had balanced t h e i r p i l e s of buttons. Recall that balancing, or trading u n t i l a t o t a l resource base of k units had k/2 red and k/2 green buttons ( i . e . , equal value positions on each resource) would maximize the value of a f i x e d resource base. In addition, an absolute increase i n the s i z e of the t o t a l resource base ( i . e . , 198 making i t more than 1200) also increased t o t a l u t i l i t y , while a loss of but-tons (leading to les s than 1200) decreased t o t a l value u n i t s . Thus, to give up more than one received reduced one's chance of being a 'winner' at the end. Consequently, any o f f e r asking f o r more than i t offered could, i n a game sense, but not i n a s i n g l e exchange sense, be seen as an attempt at advantageous exchange. While t h i s tension between a strategy appropriate f o r a s i n g l e exchange and for the ent i r e game existed i n a l l the sets, i t was p a r t i c u l a r l y acute i n t h i s set, due to the small number of opportunities for exchange, the c l o s e -ness of some subjects to balance, and the length of the experiment. (The added time spent on the lengthy questionnaires may have made subjects per-ceive i t u n l i k e l y they would even have time for four t r i a l s . ) The presence of the balanced v i s i b l e s focussed attention on the end-game, and probably i n t e r f e r e d with comparisons of p r o f i t . 2. D i r e c t i o n of I n i t i a t i o n s : Hypothesis 4. Hypothesis 4 predicts that more i n i t i a t i o n s w i l l be directed to v i s i b l e sub-j e c t s than to non - v i s i b l e subjects. The data relevant to the fourth hypo-thesis are given i n the table below. Table A.2 D i r e c t i o n of I n i t i a t i o n s on T r i a l 1, by V i s i b i l i t y of I n i t i a t o r and Target ( P i l o t Set) I n i t i a t o r Recipient Total V i s i b l e Non-Visible V i s i b l e 43 (.77) 13 (.23) 56 Non-visible 41 (.73) 15 (.27) 56 Total 84 (.75) 28 (.25) 112 Ho$:w 5 i s i 6i e . ^ a n 4.non-i-visib-if receive equal proportions of offers. Binomial t e s t , on6-taile 'dV 7Z ="5,1,-(Z*^. = 1.65, p = .05) . Reject H ... c r i t • o On the basis of these data, we may r e j e c t the n u l l hypothesis that non-v i s i b l e s are as l i k e l y as v i s i b l e s to receive o f f e r s . The T r i a l 1 o f f e r s met with a r t i f i c i a l but un i v e r s a l r e j e c t i o n , and by the second T r i a l , the d i r e c t i o n of o f f e r s changed dramatically. This r e s u l t i s shown i n Table A.3 which gives the o v e r a l l d i r e c t i o n of i n i t i a t i o n s by v i s i -b i l i t y . 199 Table A.3 D i r e c t i o n of I n i t i a t i o n s on T r i a l 2, by V i s i b i l i t y of I n i t i a t o r and Target ( P i l o t Set) I n i t i a t o r Recipient T o t a l V i s i b l e Non-visible V i s i b l e 24 (.43) 31 (.57) 55* Non-visible 21 (.37) 35 (.63) 56 Total 45 (.40) 66 (.60) 111 * One balanced v i s i b l e made no o f f e r on T r i a l 2. To test whether .60 i s s i g n i f i c a n t l y greater than .50, binomial test gives Z = 2.17,CZ = 1.96, p = .05, two t a i l e d . ) ^ c r i t ' r ' • (Note: S t r i c t l y speaking, the events i n T r i a l 2 are not independent of those on T r i a l 1, and i f one argues that i n i t i a t i o n s to v i s i b l e s are affected by the expectation of many o f f e r s going to v i s i b l e s on the second as on the past t r i a l , then a binomial test i s not j u s t i f i e d here.) More v i s i b l e s changed to a n o n - v i s i b l e a f t e r r e j e c t i o n than i n Set A, but t h i s i s probably p a r t l y due to the f a c t that the v i s i b l e who had not been i n i t i a t e d to on T r i a l 1 was usually the balanced v i s i b l e , who was undesir-able because of h i s game advantage. The change from T r i a l 1 to T r i a l 2 was dramatic — a drop of 35% i n the i n i t i a t i o n s to v i s i b l e s . The preference f o r dealing with a player whose resources could be seen, seemed i n large degree to depend on the success of such i n i t i a t i o n s . In Set A, we noticed a s i m i l a r s h i f t by the non-visibles away from v i s i b l e s when they were r e -jected, while the v i s i b l e s maintained a high l e v e l of i n i t i a t i o n s to v i s i -b l es, when the r e a l o f f e r s had been delivered on T r i a l 1, and the majority of acceptances went to v i s i b l e s . Rejection, e s p e c i a l l y of a f a i r o f f e r , probably lowered the perceived p r o b a b i l i t y of acceptance enough to make the previously uncertain deal with a n o n - v i s i b l e seem more desi r a b l e i n t h i s Set. In the experimental context, arranging an exchange with a partner about whom one has no information was preferred to no exchange at a l l . This i s consistent with the theory. The r a p i d i t y of the s h i f t i s somewhat s u r p r i s i n g . 3. Acceptance of False Offers Hypothesis 5 predicts that, a l l other things equal, subjects w i l l prefer to accept an o f f e r from a v i s i b l e , rather than an'offer from a n o n - v i s i b l e . 200 The f a l s e i n i t i a t i o n s of 100 for 100 were delivered a f t e r subjects had made th e i r o f f e r s on T r i a l 1. Table A.4 shows that, unlike the r e s u l t s f o r Sets A and B, there i s not a strong preference for o f f e r s from v i s i b l e s i n t h i s case. Table A.4 Acceptance of False Offers by V i s i b l e s , T r i a l 1 (P i l o t Set N=56) Accepts Recipient  Offer From Unbalanced V i s i b l e Balanced V i s i b l e V i s i b l e 15 (.53) 10 (.36) Non-Visible 13 (.47) 15 (.53) Neither 0 3 (.11) The most frequently mentioned reasons f o r d e l i b e r a t e l y r e j e c t i n g v i s i b l e s were that the balanced v i s i b l e s wanted to hinder the progress of the other balanced v i s i b l e (6/15). Unbalanced v i s i b l e s seemed less concerned with how the person they accepted was doing, except that when unbalanced v i s i b l e s accepted v i s i b l e s , then they referred to a preference for knowing how Other was doing (6/15). Some unbalanced resource players also mentioned that they di d not accept the same person they had i n i t i a t e d to (usually a v i s i b l e ) , because they wanted to set up as many partners as they could — that i s , i f they offered to one and accepted another, they had two instead of one part-nership to draw on for future exchanges. The data for acceptance of f a l s e o f f e r s do not lead to an unambiguous i n t e r -pretation. The s i m i l a r proportions of acceptances of v i s i b l e s and non-visibles may r e f l e c t i n d i f f e r e n c e on the part of subjects; or i t may in d i c a t e i n d i v i -dual differences i n a t t i t u d e to uncertainty. Subjects' comments never men- tioned i n d i f f e r e n c e , as they sometimes did i n Set B. However, for the 26 v i s i b l e s whose comments referred only to t h e i r own gain or benefit as the reason f o r accepting one of the phoney o f f e r s , 19 had.accepted n o n - v i s i b l e s . Acceptance of a v i s i b l e o f f e r tended to be accompanied by more references to Other's p r o f i t . On T r i a l 2, acceptances favoured n o n - v i s i b l e s , both because they made more generous o f f e r s (these were not intercepted), and because subjects were un-w i l l i n g to accept subjects who had rejected them on T r i a l 1 (mainly v i s i b l e s ) . The average exchange r a t i o f o r non-visibles on T r i a l 2 was 105 (111 for of-fers to v i s i b l e s ) , compared with a r a t i o of 87 for o f f e r s made by balanced v i s i b l e s , and 92 for those made by unbalanced v i s i b l e s . By the beginning of T r i a l 2, no. non-visible had either received an o f f e r or had one accepted, and the improved r a t i o s probably showed an increased concern with getting 201 into exchanges during the remaining three t r i a l s . Table A.5 shows the pat-tern of acceptances on T r i a l 2. Table A.5 Relative Frequencies of Acceptance of Real Offers on T r i a l 2, by Value P o s i t i o n and V i s i b i l i t y of I n i t i a t o r and Target ( P i l o t S e f N = 111*) Recipient I n i t i a t o r ' .. Unbalanced Vis Balanced Vis Non-Visible Unbalanced V i s i b l e 4/8 2/9 8/12 Balanced V i s i b l e 1/4 1/3 6/9 Non-Visible 10/16 7/15 24/35 Total 15/28 (.54) 10/27*(.37) 38/56 (.68) * One balanced v i s i b l e made no o f f e r on T r i a l 2. If these acceptances are compared with T r i a l 1 i n Set A (which also ran n a t u r a l l y ) , we see that the balanced v i s i b l e s have about as large a propor-t i o n of o f f e r s accepted as the non-visibles i n Set A. 4. Discussion Why did the changes i n the paradigm produce r e s u l t s so d i f f e r e n t from those of Sets A and B? Let us t r y to imagine how the players saw the game i n t h i s set. Since a l l had equal resource bases (1200 buttons), subjects were bas-i c a l l y a l l peers — a l l comparable, with an equal ' r i g h t ' and equal chance at winning. However, two players, the balanced v i s i b l e s , had been given a head s t a r t i n reaching the stated game goal of balanced resource p i l e s . In two respects, they were undesirable exchange partners: because t h e i r i n i -t i a l p r i ces were high r e l a t i v e to what the unbalanced players could demand; and because any trade they arranged would move them even further ahead of the others. (Balanced v i s i b l e s did i n fact ask for more than they offered, as shown i n Table A.1.) While there i s nothing i n p r i n c i p l e wrong with having some v i s i b l e s who were r e l a t i v e l y l e s s d e s i r a b l e partners (see Hypo-the s i s 1), the p o s i t i o n of the balanced v i s i b l e s made them not second most preferred, but least preferred, e s p e c i a l l y i n l i g h t of the sorts of o f f e r s they made. As a r e s u l t , though the experimental design was intended to make subjects d i f f e r only i n the ambiguity of the information a v a i l a b l e about them, the nearness of some players to balance tipped the assessment of expected value of an exchange i n favour of the non-visibles over the balanced v i s i b l e s . 202 Balanced v i s i b l e s could, by trading at a f a i r rate i n terms of value u n i t s , quickly balance t h e i r resource p i l e s , and f i n i s h with more than the other players. With the small number of trading opportunities, the only way the balanced v i s i b l e s would not win was i f no one traded with them. While un-balanced and non - v i s i b l e players could refuse to go along with the advanta-geous ( i n a game sense) o f f e r s made by balanced v i s i b l e s , and to make o f f e r s to them that did not give them an increase i n t o t a l resources, t h i s was to some extent a v i o l a t i o n of how they saw the s o c i a l p rices i n the game. An even more e f f e c t i v e strategy was to refuse to i n i t i a t e to or to accept from a balanced v i s i b l e at a l l . In addition, fellow balanced v i s i b l e s seemed to s i n g l e each other out as the man to beat, and they competed not only by r e j e c t i n g each others' o f f e r s , but also by refusing to i n i t i a t e to each other. As a r e s u l t , i n i t i a t i o n s to non-visibles increased, contrary to pre d i c t i o n . If trading continued for several t r i a l s , one would have expected that the less advantaged players would trade u n t i l they were i n a s i m i l a r p o s i t i o n (value position) to the balanced v i s i b l e s ; only then would the o r i g i n a l balanced v i s i b l e s be acceptable partners. At t h i s point, the theory would probably make better p r e d i c t i o n s , because then no one could see exactly where the non-visibles stood, and would want to avoid trading with them. Non-visibles might, at t h i s stage, be avoided 1) to avoid u n f a i r exchanges ( i . e . , one would be aware that resource positions had been moving toward balance, and be even more uncertain as to whether a non-visible was demand-ing a f a i r p r i c e ; and 2) to avoid s i t u a t i o n s i n which a trade gives a non-v i s i b l e a further lead. The nearness of a v i s i b l e to balancing could be monitered, and even a generous o f f e r refused i f i t seemed the i n i t i a t o r might win as a r e s u l t of the transaction. But why did subjects not fear that the non-visibles were balanced to begin with? The main cue was i n the type of o f f e r received from n o n - v i s i b l e s : — T r i a l 1: a 100/100 o f f e r (phoney) was probably not read by sub-j e c t s as i n d i c a t i v e of a player with balanced resources, e s p e c i a l l y i f subjects compared i t with o f f e r s they themselves had sent. — T r i a l 2: non-visibles a c t u a l l y did make the most generous o f f e r s (average r a t i o of advantage was 105 o v e r a l l , versus 92 f o r unbal-anced v i s i b l e s , and 87 for balanced v i s i b l e s ) . Given the time l i m i t , unbalanced v i s i b l e s and non-visibles had to concern themselves i n i t i a l l y with 'getting going'; that i s , they may have been more concerned with the p r o b a b i l i t y of acceptance, which was evidenced by t h e i r more generous o f f e r s . This concern, and the tendency to focus on the b a l -anced S's as competition, drew attention away from comparison with others. As a r e s u l t , scope condition 4 may not have been met. 203 The theory predicted that when the non-visibles were rejected because Other 'did not want to deal with them', they would want to reveal information about t h e i r resources. Before they could get to thinking t h i s , they received a deluge of o f f e r s i n reaction to the T r i a l 1 r e j e c t i o n s , so that they had fewer reasons to reveal information, e s p e c i a l l y since they perceived the covers would hide t h e i r progress toward balancing. Given equal t o t a l wealth, and caution on the part of the players, no s t r a -tegy was l i k e l y to advance anyone very f a r ahead of others i n t h i s game. In f a c t , the s i t u a t i o n was a v i v i d i l l u s t r a t i o n of norms as a ' c o a l i t i o n of actors', as Emerson argues.^ Fairness w i l l be maintained because no one w i l l put up with the loss involved i n y i e l d i n g to an attempt to gain advantage, as long as there are a l t e r n a t i v e s a v a i l a b l e . While the theory argues along these l i n e s , the experiment f a i l e d to a n t i c i p a t e who would be perceived to have the advantage. In terms of the predictions made about attempts at advantageous exchange, i t i s probably reasonable to say that the present theory i s most applicable to the i n i t i a l phases of exchange int e r a c t i o n s — to the i n i t i a l attempts made at impression management, when P attempts to l i m i t the range of a l t e r n a t i v e responses open (to 'Other. The more accurate the information P has about 0, but not v i c e versa, the more P w i l l believe he can have an e f f e c t on the def-i n i t i o n of the s i t u a t i o n . As i n t e r a c t i o n continues, or i f we 'cut i n ' on i n t e r a c t i o n at a l a t e r stage (as i t seems the P i l o t set design does), the r e a l bargaining and power advantages, i f they e x i s t , w i l l have become more apparent, and have a greater e f f e c t on the course of i n t e r a c t i o n . To summarize: The P i l o t set was considered an inadequate test because: the short time l i m i t made players overly conscious of end-game e f f e c t s . This i n t e r f e r e d with and operated i n opposition to a comparison of r e l a t i v e mar-gin a l gain. Focus on the balanced v i s i b l e as the man nearest winning may have led to a greater frequency of i n i t i a t i o n s to, and acceptance of, non-v i s i b l e s , as a strategy for thwarting the leader u n t i l he could be caught up. It probably led the non-visibles to be more concerned with ensuring transactions, as they could not r i s k missing transactions, with the balanced v i s i b l e nearing equalization of h i s resources. Thus, the opportunity to choose between more and l e s s uncertain a l t e r n a t i v e s with equal expected u t i l i t y was not provided adequately, and the predictions about preference between such a l t e r n a t i v e s could not be properly tested. "204 FOOTNOTES: APPENDIX IV E l l s b e r g , and MacCrimmon found that only a minority of subjects seem to prefer the l e s s ambiguous a l t e r n a t i v e when r a t i o n a l i t y postulates i n d i -cate they should be i n d i f f e r e n t between a r i s k y and an uncertain a l t e r -native. The proportion avoiding uncertainty seems much larger i n t h i s experiment, perhaps because the source of the uncertain information i s l e s s r e l i a b l e i n the present study. See: D. E l l s b e r g , 'Risk, Ambiguity, and the Savage Axioms', Quarterly Journal of Economics, 75, 1961, pp. 643-69; and K.R. MacCrimmon, 'Descriptive and Normative Implications of the Decision-Theory Postulates', i n K. Borch and J. Mossin, E d i t o r s , Risk and Uncertainty, London: MacMillan, 1968. R.M. Emerson, 'Exchange Theory: Part I I ' , i n J. Berger, M. Z e l d i t c h , and B. Anderson, Edi t o r s , S o c i o l o g i c a l Theories i n Progress, Volume I I , Houghton-Mifflin, 1972, pp. 85-86. 205 Appendix IV: P i l o t Set Questionnaire given to a l l subjects a f t e r they had made T r i a l 1 i n i t i a t i o n s . Questionnaire #1 Your l e t t e r _____ 1. Who did you ju s t make an o f f e r to? G H I J K L M N ( c i r c l e one) red red 2. What was the of f e r ? buttons offered f o r buttons. green green 3. What are your reasons for the s i z e of o f f e r that you made? 4. How l i k e l y do you think he i s to accept your offer? ( C i r c l e the p o s i -t i o n on the l i n e below that shows how l i k e l y you f e e l ' i t i s that your o f f e r w i l l be accepted. / / / _ / ; / extremely quite 50-50 not very not at a l l l i k e l y l i k e l y chance l i k e l y l i k e l y What are your reasons for thinking t h i s ? 5. Do you think the person you made the o f f e r to i s more l i k e l y than any of the others to accept your offer? Yes No ( c i r c l e one) If you said NO, who do you think i s more l i k e l y to accept? Why do you think t h i s ? 6. Did you t r y to see your o f f e r from the other person's point of view? Questionnaire given to v i s i b l e s * a f t e r f a l s e o f f e r s were c o l l e c t e d - T r i a l 1. Questionnaire #2 Your l e t t e r 1. I f you accepted an o f f e r on the f i r s t opportunity: What was your main reason f o r accepting the o f f e r you did? 206 2. There are probably advantages and disadvantages to having your buttons out i n the open. What do you see to be the advantages? What do you see to be the disadvantages? 3. If you could begin the game again, would you prefer a game i n which everyone had t h e i r buttons out i n the open, or where everyone had covers over t h e i r buttons? open covers Why? 4. Which players do you think have the best chance of doing well i n t h i s part of the game? G H I J K L M N don't know *Note: Questionnaire #2 for non-visibles was the same, except for one extra question asking the non-visibles i f they would choose to r e -move t h e i r covers. Questionnaire given to v i s i b l e s a f t e r reply to T r i a l 2 o f f e r s . Questionnaire #3 Your l e t t e r 1. If you accepted an o f f e r on the second trading opportunity: What was your main reason f o r accepting the o f f e r you did? 2. Do you s t i l l see the same advantages i n having your buttons out i n the open? YES NO What other advantages do you see? Do you see the same disadvantages s t i l l ? YES NO What other disadvantages do you see? 3. I f , a f t e r the second exchange opportunity,, you were given the choice of having the covers removed from those players who have them, would you choose to have t h i s done? YES NO Why? 207 Questionnaire given to non-visibles a f t e r reply to T r i a l 2 o f f e r s . Questionnaire #3 Your l e t t e r • 1. If you accepted an o f f e r on the second trading opportunity: What was your main reason f o r accepting the o f f e r you did? 2. Do you s t i l l see the same advantages i n having covers over your but-tons? YES NO Do you see any other advantages? What are they? Do you s t i l l see the same disadvantages i n having covers? YES NO What other disadvantages do you see? 3. I f you are given a choice a f t e r the second exchange opportunity i n t h i s game to remove the covers from your buttons, would you choose to do so? YES NO Why? Note: On the questionnaires given to subjects, there were no references to i d e n t i f y the questionnaires as being f o r v i s i b l e s only, or for non-v i s i b l e s only. S u f f i c i e n t space was provided f o r r e p l i e s . 208 APPENDIX V DATA REFERRED TO BUT NOT INCLUDED IN MAIN TEXT Table A.6 Types of Offers by Resource Level and V i s i b i l i t y of I n i t i a t o r , Set A (reference: Hypothesis 1) TRIAL I N = 112 for a l l resource l e v e l s N f o r highs = 56 I n i t i a t o r High v i s i b l e High NV Tota l 2 Asks same as o f f e r s 18 14 32 Asks more than o f f e r s 13 Asks less than o f f e r s 5 _6 11 1.28 df = 2, n.s. Collapsing f i r s t and^third categories (equal and l e s s ) , f o r comparison with of f e r s asking more than i s offered, Z for differences of proportions i s .21, n.s. I n i t i a t o r Low v i s i b l e Low NV Tota l 2 = 1.35 df = 2, n.s. Asks same as o f f e r s 14 12 26 Asks more than o f f e r s 12 20 N f o r lows =56 Asks less than o f f e r s 6 _4 10 Collapsing f i r s t and t h i r d categories f o r comparison with o f f e r s asking more than i s offered, test for difference of proportions, Z = 1.2, s i g n i -f i c a n t at .11 l e v e l . 209 Comments on Table A.6 As noted i n Chapter 3, subjects did not know the exact s i z e of the v i s i b l e s ' resource bases, so could not p r e c i s e l y define a ' f a i r ' exchange. S t i l l , we expect low resource persons to have offered l e s s f o r a given return than did highs. Because low non-visibles are e n t i t l e d by v i r t u e of a small r e -source base to ask for more buttons than they offered, i t i s of more i n t e r -est to Hypothesis 1 i f the o f f e r s of the high non-visibles resembled most those of the low v i s i b l e s , or the high v i s i b l e s . By modeling t h e i r o f f e r s on what would be a f a i r transaction between a low and a high, the high non-v i s i b l e s could p o t e n t i a l l y gain much more than a high v i s i b l e partner. Although the r e s u l t s were i n the predicted d i r e c t i o n , fewer subjects made advantageous o f f e r s than did subjects i n W. Foddy's experiments, i n which a l l subjects were v i s i b l e s . Because the r e s u l t s i n Set B generally sup-ported the theory, i t seems l i k e l y that part of the f a i l u r e of Hypothesis 1 i s due to o p e r a t i o n a l i z a t i o n f a i l u r e . More s p e c i f i c a l l y , the uneven effectiveness of the marginal u t i l i t y manipulation, an i n a b i l i t y to make exact comparison of p r o f i t s , and a v a r i a b l e desire to do better than others, probably contributed to an incomplete f u l f i l l m e n t of the scope conditions, making Set A an inadequate test f o r Hypothesis 2. Table A.7 Raw Data from Questionnaire Given a f t e r Subjects had Replied to False Offers (See Appendix III Questionnaire 2 for Questions) (Reference: Hypothesis 3) SET B l - HIGH NON-VISIBLES* PERCEPTION OF COVERS AFTER TRIAL 1 I. ADVANTAGES (n = 27)* Subjects' comments: - others can't see i f I ' l l give more f o r less - can b l u f f and draw up a bid f o o l others into bidding higher - others make better o f f e r s , can't see how l i k e l y I am to accept - bargaining power - others can't see I'm r i c h and t r y i n g to get an advantageous trade - people take chance on you, don't ask too much - no one knows how badly I want buttons - keeps others guessing - see v i s i b l e s ' weaknesses, hide my strength - others can't see how badly I want to get r i d of red so I don't have to give up so much - I know more,easier to decide i f other unsure, he may o f f e r more than he asks, I have the edge of uncertainty - no one knows how many you have - see where v i s i b l e s stand, they can't see me - no one sees your hand, gain s e c u r i t y - no one sure I have more than 1,200 - no one sees my t o t a l - l i k e to keep i t secret - others can't t e l l how large a surplus I'm main-t a i n i n g - no one can stop non-visibles from obtaining a high score - others don't know actual numbers I have - others don't know my amounts - others can't see me balance - no one can see how many chips I have - others can't see what I have and w i l l trade as v i s i b l e s balance - others may t r y us i f they think we have a l o t - people who are hard up w i l l o f f e r . * Note: One subject gave no answer. Table A.7 (Continued) I I . DISADVANTAGES (n = 21)* Subj ects' Comments - people l i k e to make o f f e r s knowing r e l a t i v e amounts of buttons - others are suspicious - others are a f r a i d to exchange - others may not trade - others reluctant to trade i f they can't see - not enough in t e r e s t i n trading with me - may get fewer o f f e r s - others more l i k e l y to deal with v i s i b l e s - others won't o f f e r to covereds others unsure about trading with you - others may be unwilling to o f f e r - others may be less w i l l i n g to deal - may get fewer o f f e r s due to uncertainty with respect to what I ' l l accept - can't see how f a r other covered ones w i l l go in trades - may get fooled by a non - v i s i b l e - can't see non-visibles - non-visibles may have les s than 1,200 and won' be seen - can't t e l l i f o f f e r w i l l work i f don't know what opposition has - can't make sure o f f e r s confusion over covert, undefined deals - others can't see how much I want to trade. * Note: 7 subjects gave no answer, or said "none". 212 Table A.8 Raw Data from Questionnaire Given a f t e r Subjects had Replied to False Offers (See Appendix III Questionnaire 2, for Questions) (Reference: Hypothesis 3) SET BII - LOW NON-VISIBLES' PERCEPTION OF COVERS AFTER TRIAL 1 I. ADVANTAGES (n = 2>4)* Subjects' Comments - others w i l l o f f e r better deals i f they can't see me - others can't see what I gain, and they may be curious - no one knows what would be f a i r to get from me - no one knows how desperate you are - v i s i b l e players may lose i n speculation with ones they can't see - can see what v i s i b l e players w i l l trade, they can't see me - people can't see how you p r o f i t each round - others think you are anxious to trade - no one i s sure about me - others don't know what buttons are worth to me others can't t e l l my v u l n e r a b i l i t y - others may make better o f f e r s - no one sees what I have, so w i l l make better o f f e r s see how others do; they don't know my rank - others can't see how I do; don't know my wealth - people can't see how much I have - no one knows my p o s i t i o n - no one knows how close you are to them - others can't see how you're doing others can't see how I've traded; I l i k e to be independent of others' knowledge - no one knows my true p o s i t i o n , I can gain i n secrecy - others can't see i f I balance - people may trade i f competing with v i s i b l e s - get more o f f e r s , can pick and choose. * Note: 4 subjects gave no answer, or said "none". Table A.8 (Continued) II . DISADVANTAGES (n =20)* Subjects' Comments - others reluctant to deal - others may think I'm doing too well - others don't know what to ask and offer - others may not deal; visibles have more choice - uncertainty - others scared to deal with me - people reluctant to deal - others may think I'm doing too well - no offers - others less l i k e l y to deal with covered person - others may suspect I'm a threat - others unwilling to deal - get fewer offers - don't know how covered players are doing - can't see bargaining power of non-visible - I can't see how covered players are doing - unsure about hidden players - can't see what hidden players have and so I must be aggressive - can't t e l l what invisible ones have - i f you have less than 1,200, have nothing to hide — people think I'm cheap, not poor - others may feel I'm not anxious to trade. * Note: 8 subjects gave no answer, or said "none". 214 Table A.9 Raw Data from Questionnaire Given a f t e r Subjects had Replied to Offers (See Appendix I I I , Questionnaire 2, for Questions) (Reference: Hypothesis 3) SET B (I AND I I ) : VISIBLES' PERCEPTION OF HAVING NO COVERS AFTER TRIAL 1 I. ADVANTAGES (n = 52)* Subj ects' Comment s - get more o f f e r s i f known - people trade more w i l l i n g l y - more l i k e l y to get o f f e r s - everyone w i l l deal - people more l i k e l y to trade should get more o f f e r s - others see what you have and may be more w i l l i n g to trade - others trade i f they know where you stand - v i s u a l enticement to trade - others w i l l i n g to o f f e r i f they see me - others are freer to deal with me - get o f f e r s - people w i l l trade - more people w i l l trade with uncovered ones people prefer known partners, a f r a i d of covers - get more o f f e r s - others trade, see I'm w i l l i n g - more o f f e r s - others more w i l l i n g to trade - get more o f f e r s - people more w i l l i n g to trade i f they see what you have - others are l i k e l y to o f f e r - others trade i f they can see how I compare - others w i l l trade - others prefer v i s i b l e s - others can see what to o f f e r - others know how much to ask for - get more reasonable o f f e r s - others have an idea of what i s a mutually b e n e f i c i a l o f f e r - others see what you have to trade - others know there i s no r i s k i n dealing with open ones — prefer an open market 215 Table A.9 (Continued) I. ADVANTAGES (Continued) Subjects' Comments others can t e l l what to o f f e r others know what to ask others see my r a t i o s others see I'm a p o t e n t i a l partner others w i l l trade as know what I'm doing people know what to o f f e r people see what you have to trade others know my p o s i t i o n , can assess mutual gain people know what trades I ' l l make I can see who i s l i k e l y to trade i f you watch, can see how other v i s i b l e s have traded see what others i n the open value, and when trade more profitable„ see trading p o s s i b i l i t i e s and who needs buttons most o f f e r from a p o s i t i o n of knowledge; can be more ca l c u -l a t i n g know who to o f f e r to i n the open can see what others have can see how v i s i b l e players do can see who w i l l mutually benefit see open ones', need f o r buttons can see who i s l i k e l y . t o trade can see who i s i n same p o s i t i o n as I * Note: 4 subjects gave no answer, or said "none". I I . DISADVANTAGES (n = 49)* Subj ec t s' Comment s - others see my need, what I ' l l accept - others w i l l r e j e c t me, thinking I ' l l pay more - others know what I'm doing and thinking - others see how badly I need a colour - others see how much I have of everything - partners know what they can expect - everyone knows what you have - shows my wealth so I can't bid 216 Table A.9 (Continued) I I . DISADVANTAGES (Continued) Subjects' Comments - others see what kind of deal I ' l l make - covered ones can take advantage of me - everyone sees what I need, no secrecy - hidden ones can press v i s i b l e s into a poor deal - covered players can be more c a l c u l a t i n g - others w i l l ask too much and I ' l l accept - can't conceal a trading advantage - others know I have l o t s , can a f f o r d to pay - others - may get more than they o f f e r - others can adjust t h e i r bids i f they see me - can't b l u f f , have to trade even - others know what I have - covered players know more - others ask too much - others know how l i k e l y I am to trade - hidden others can take advantage of my blindness - can't make advantageous trade - others w i l l refuse - others t r y to take advantage of my need - others see my need, o f f e r l e s s - others know what I ' l l accept and take advantage - others may expl o i t you - other people know what you have - others can see my rank and l i m i t my chances - can't see where the competition stands - others see where I stand - others won't deal as I balance fewer o f f e r s as I balance people see when I balance - others see how you are doing - others can block you i f they see you do well - covered players can keep open ones from winning - others see my rank - unwilling to lower t h e i r p o s i t i o n - others know when to stop trading with you - hidden ones can see to compete with me - others avoid you as you balance - competition can see you - others see my success and t r y to hinder i t - no sense of mystery - unknown i s i n t e r e s t i n g - no guesswork * Note: 7 subjects gave no answer, or said "none". 217 Table A.10 Set A - Perceived P r o b a b i l i t y That T r i a l 1 Offer w i l l be Accepted Reference: Hypothesis 3) N = 112 I n i t i a t o r V i s i b l e Non-Visible P r o b a b i l i t y of advantageous advantageous Acceptance a l l o f f e r s o f f e r s only a l l o f f e r s o f f e r s only Extremely l i k e l y 11 3 12 U Quite l i k e l y 31 4 27 -6 50-50 chance 12 6 12 8 Not very l i k e l y 2 0 4 3 Not a l a l l l i k e l y 0 0 1 0 For i n i t i a t i o n s asking for more than i s offered, the modal category f o r both v i s i b l e s and non-visibles i s '50-50 chance'; the mode i s one step higher over a l l o f f e r s . I t would seem, then, that non-visibles are not more l i k e l y to a n t i c i p a t e a low p r o b a b i l i t y of acceptance on the f i r s t t r i a l , as predicted. The non-visibles have a s l i g h t l y more pessimistic view about the possible success of an o f f e r that asks more than i t o f f e r s , but they make more of these o f f e r s , and when they ask for more, they tend to ask for a greater number more than do vis i b l e s . , . 218 Table A.11 Written Reasons Given by High Non-Visibles, Set B l , for I n i t i a -tions to V i s i b l e s and Non-Visibles (See Questionnaire 1, Ques-t i o n 4, Appendix III) (Reference: Hypothesis 4) I. High n o n - v i s i b l e s ' reasons for i n i t i a t i n g to v i s i b l e other (n = 12) can see his r e l a t i v e amounts of red and green - h i s l o t i s uncovered'- any v i s i b l e would do - f e l t no one else would choose him he w i l l be w i l l i n g to deal cheaply - he wants red - can see what he w i l l consider reasonable - he w i l l probably accept - he has surplus of what I need - i t benefits both - he has excess - don't want to deal with covered - he w i l l be w i l l i n g to trade prefer v i s i b l e partner - know he wants to trade some - he has excess of red II. Reasons f o r i n i t i a t i n g to a non-visible other (n = 16) - covered ones won't get so many o f f e r s so w i l l accept mine - he won't get o f f e r s but he can see what's going on - might get a partner - he may deal - he won't get bids so he i s l i k e l y to accept - he may have more than 1,200 and be w i l l i n g to trade - gamble: he may have more than v i s i b l e s do - f e l t he wouldn't get another o f f e r - everyone else w i l l o f f e r to a person i n the open - guessing what he has, and he w i l l give me l o t s - s t r a i g h t across from me - f e e l he w i l l get fewer o f f e r s - chance he has more than 1,200 and w i l l be receptive - he has a red tag - we each have what the other needs - won't be on the defensive and w i l l be w i l l i n g to trade - he has excess of red 219 Table A.12 Written Reasons Given by Low Non-Visibles, Set BII, for I n i t i a -tions to V i s i b l e s and Non-Visibles (See Questionnaire 1, Ques-t i o n 3, Appendix III) (Reference: Hypothesis 4) I. Low non-visibles' reasons f or i n i t i a t i n g to v i s i b l e other (n = 19;) - he has a l o t , i s l i k e l y to accept - he has a surplus, may accept uneven o f f e r - I know he has red and how much - don't want to deal with covered - open ones probably have more than covered; he doesn't know what i s f a i r - I can see how he gains, he has more - no answer - prefer to deal with v i s i b l e s , he has l o t s ; won't deal with covered - can see h i s excess only one I can c l e a r l y see the wealth of - prefer to deal with open one - can see how he's doing - he has l o t s , and prefer to deal with v i s i b l e - he can afford i t - can't see how covered one does - impulse - can t e l l what he has - he has a l o t ; unsure what i n v i s i b l e s have - he won't get many o f f e r s - prefer to deal with person i n the open - l e s s r i s k , and he has an excess I I . Low n o n - v i s i b l e s ' reasons f o r i n i t i a t i n g to non-visible (n = 9) - f e e l i n g him out - he may have the same as me and be w i l l i n g to trade - he won't get o f f e r s , w i l l trade - taking a chance he has a large number - d i r e c t l y across, i n same spot as I am - t e s t i n g to see i f he i s one of the players with a l o t - he may have more than the v i s i b l e s and be w i l l i n g to trade - he w i l l get few o f f e r s ; high chance of accepting - hoping no one else w i l l o f f e r to him 220 Table A.13 Written Reasons Given by V i s i b l e Subjects, Set BI, for I n i t i a -tions to V i s i b l e s and Non-Visibles (See Questionnaire 1, Ques-ti o n 4, Appendix III) (Reference: Hypothesis 4) I. V i s i b l e s ' reasons for i n i t i a t i n g to v i s i b l e other (ri = 19) - he has l o t s of green - complementary p i l e s , know what to o f f e r - we have what each other needs - he has a l o t of wh£ I want - we have what each other needs - he has a large number of green - he w i l l gain as much as me; l i k e l y to accept - prefer open one - covered players may do too w e l l - we both be n e f i t ; he has surplus - he has what I want - no r i s k ; complementary p r o f i l e s he has an excess of what I need - large number of exposed buttons - we have what each other wants; prefer to act from knowledge - his numbers of buttons was known - closest to me, and v i s u a l enticement - he has a surplus and w i l l part with some - he has red, needs green, w i l l accept complementary p i l e s I I . V i s i b l e s ' reasons for i n i t i a t i n g to non-visible other (n = 9) - he may have too many - covered ones won't get o f f e r s , very l i k e l y to accept - he has an excess, and we both gain - can take a chance on f i r s t round to see i f he has more than 1,100 - he w i l l get few o f f e r s see what he has; i f he refuses, he has less f e l t others would not o f f e r to a non-visible he has mostly green, l i k e l y to accept 2-21 Table A.14 Written Reasons Given by V i s i b l e s , Set BII, for I n i t i a t i o n s to V i s i b l e s and Non-Visibles (See Questionnaire 1, Question 4, Appendix III) (Reference: Hypothesis 4) I. V i s i b l e s ' reasons for i n i t i a t i n g to v i s i b l e other (n = 20) - he has excess and w i l l trade - he has less of what I have - his p o s i t i o n i s l i k e mine - I can see what w i l l appeal to him - could see h i s buttons see what he has, know how he benefits - he needs what I have - no r i s k - can see what he has - can see.he has excess c e r t a i n of his amount - p r o b a b i l i t y of acceptance high - most v i s i b l e to me; he has a l o t - prefer ones i n open; mutually b e n e f i c i a l - deal benefits us both - v i s i b l e s are more l i k e l y to trade - same p r o f i l e as me - he gains same as I do; w i l l accept - same p r o f i l e as me - I know what he has and v i c e versa I l i k e him, and he has green - less r i s k ; he needs red and w i l l accept ones i n the open w i l l trade together; don't l i k e covered ones I I . V i s i b l e s ' reasons f or i n i t i a t i n g to non-visible other (n = 8) f i n d out how even h i s p i l e s are f i n d out h i s p o s i t i o n ; see what he i s l i k e covered ones won't get o f f e r s , w i l l accept - taking chance; l i k e l y to be accepted - he i s l i k e l y to accept too much competition for v i s i b l e s ; see i f covered one i s good trader - he may have a large number - he may have more than v i s i b l e s . 222' APPENDIX VI RAW DATA I. Set A. I n i t i a t i o n s -and Transactions Key to notation: hrv: H-ighb red v i s i b l e (1,600 red, 30 green) hgv: High green v i s i b l e (1,600 green, 30 red) l r v : Low red v i s i b l e (800 red, 30 green) lgv: Low green v i s i b l e (800 green, 30 red) Corresponding non-visibles are denoted by hrnv, hgnv, lr n v , lgnv Example lhgv-lrnvl00gl20rE = In experiment 1, the high green v i s i b l e made^an of f e r of 100 green buttons for 120 red buttons which was rejected. A. SET A: FIRST EXCHANGE OPPORTUNITY (TRIAL 1) Ihrv-hgvl02rl00ga lhgv-hrvlOOglOOraq Hrv-lgnvlOOrlOOga Ilgv-lrnvl00gl50ra Ihrnv-hgv80rll0gr lhgnv-lrnvlOOglOOrr Urnv-hgvlOOrlOOgr Ilgnv-hrv70g90rr 3hrv-lgv50r50ga 3hgv-hrvl00gl00rr 31rv-lgnvl00rl20gr 31gv-hrvl00gl00ra 3hrnv-lgnv99r99ga 3hgnv-lrvl00g90ra 31rnv-lgnv50r50gr 31gnv-hrvl00gl00rr 5hrv-hgv60rl00gr 5hgv-hrvl00gll0rr 51rv-hgvl00r99ga 51gv-hrv60gl00rr 5hrnv-hgv80r80gr 5hgnv-lrnvl00gl00ra 51rnv-lgnvl00rl25gr 51gnv-hrvl00gl00ra 2hrv-hgv90r90ga 2hgv-hrvl00gl00ra 21rv-lgvl00rl00ga 21gv-lrv50g60rr 2hrnv-lgnvl00rl00ga 2hgnv-hrvl00g200rr 21rnv-lgv60r60gr 21gnv-lrv90g65rr 4hrv-lgvl00rl00ga 4hgv-hrnv25g30ra 41rv-hgnv75r65ga 41gv-hrvl00g90ra 4hrnv-hgvlOOr75ga 4hgnv-hrvl00g780rr 41rnv-hgv50r70gr 41gnv-hrvl00g200rr 6hrv-lgnv20rl5ga 61hgv-lrnvl00gl00rr 61rv-lgv90rl00ga 61gv-lrvl00gl00rr 6hrnv-hgvl00rl00ga 6hgnv-lrnvl00g98ra 61rnv-hgvl00rl00gr 61gnv-lrvl00gl00ra 7hrv-hgvl00rl00gr 7hgv-hrnv70g60ra 71rv-hgv90rl00ga 71gv-hrvl00gl00ra 7hrnv-hgvl00rl00gr 7hgnv-hrvl00gl00rr 71rnv-lgnv75r75ga 71gnv-hrv50g60rr 9hrv-hgvl00r95ga 9hgv-hrnvl00gl00ra 91rv-hgvl00ri00gr 91gv-hrv30g20rr 9hrnv-hgv60r70gr 9hgnv-hrv80|55ra 91rnv-lgnvi00rl00ga 91gnv-hrv80g60rr Hhrv-lgnvlOOrlOOga Hhgv-hrvlOOglOOra Ulrv-hgvlOOrlOOga Illgv-hrnv40g20ra Ilhrnv-hgv60rl00gr Ilhgnv-hrnvl00g80rr Illmv-hgvl00rl50gr Illgnv-hrvl00g75rr 13hrv-lgnv60r60ga 13hgv-hrvl00gl00ra 131rv-hgv50r75gr 131gv-hrvl00gl00rr 13hrnv-lgv50r50ga 13hgnv-hrvl00gl50rr 131rnv-hgvl00rl00ga 131gnv-hrvl00g200rr 8hrv-hgv60r60ga 8hgv-lrvl00gl00rr 81rv-lgv50r50ga 81gv-hrvl00gl50ra 8hrnv-lgv50rl00gr 8hgnv-hrv60g60rr 81mv-hgnv80r70ga 81gnv-hrnvl00gl00ra 10hrv-hgv75r50ga lOhgv-hrvlOOglOOra lOlrv-hgvlOOrlOOgr lOlgv-lrvlOOglOOra lOhrnv-lgnvlOOrlOOga lOhgnv-hrvlOOglOOrr 101rnv-hgv50r50gr 101gnv-hrnvl00gl25ra 12hrv-hgvl00*100gr 12hgv-hrvl00g200rr 121rv-lgnvl00rl00ga 121gv-lrvl00gl00ra 12hrnv-hgvlOOrlOOgr 12hgnv-lrnvl00g80ra 121rnv-hgvl00rl50ga 121gnv-hrv20g30ra 14hrv-hgv50rl00gr 14hgv-hrvl00gl00ra 141rv-lgvl00rl00ga 141gv-hrnvl00g95ra 14hrnv-lgnvl00rl00ga 14hgnv-hrvl00gll0rr 141rnv-hgnvl004100ga 141gnv-lrnv90gl00ra 224, B. SET A: SECOND EXCHANGE OPPORTUNITY (TRIAL 2) (Note: I n i t i a t i o n s from T r i a l 2 were intercepted and replaced with f a l s e o f f e r s ; therefore, the data have no entry f o r 'ac-cepted' or 'rejected'.) Ihrv-lgv90r90g lhgv-hrnvlO 2 g10 Or Ilrv-hgv50rl00g Ilgv-hrnvl00gl50r lhrnv-hgnvlOOrlOOg Ihgnv-hrnvl00gll9r Ilrnv-lgnv50r75g Ilgnv-lrnv60g60r 3hrv-hgv90rl00g . 3hgv-lrvl00g80r ' 31rv-lgvl00rl00g 31gv-lrvl00gl00r 3hrnv-lgnv90r99g 3hgnv-hrv80g80r 31rnv-hgv50r50g 31gnv-hrnvl00gl00r 5hrv-lgnv70r80g 5hgv-hrvl00g98r 51rv-lgvl00r99g 51gv-hrnv80gl00r 5hrnv-lgnvl00rl00g 5hgnv-lrnvl00gl00r 51rnv-hgnvl00rl00g 51gnv-hrvl00gl00r 7hrv-lgvl00rl00g 7hgv-lrv90gl00r 71rv-hgnv90rl00g 71gv-hrvl00g95r 7hrnv-hgnvl00rl00g 7hgnv-lrvl00gl00r 71rnv-lgvl00rl00g 71gnv-lrv75g7Sr 9hrv-lgvl00r90g 9hgv-lrvl00g95r 91rv-lgvl00rl00g 91gv-hrvl00gll0r 9hrnv-lgv70r70g 9hgnv-lrnvl00gl00r 91rnv-lgnvl00rl00g 91gnv-lrnvl00gl00r 2hrv-lgvl00rl00g-2hgv-hrvl00gl00r 21rv-hgv90rl00g 21gv-hrv90gl00r ' 2hrnv-hgnvl00rl00g 2hgnv-hrnvl00g200r 21rnv-hgvl00rl00g 21gnv-hrnvl00gl00r 4hrv-hgvl00rll0g 4hgv-lrnv75glOOr 41rv-hgv50r60g 41gv-lrvl00gl00r 4hrnv-lgvl00r90g 4hgnv-lrnvl00gl00r 4lrnv-lgnv50r70g 4hgnv-hrvl00g200r 6hrv-hgv45r45g 6hgv-hrnvl00gl00r 61rv-hgv50r70g 61gv-hrnvl00gl00r 6hrnv-hgvl00rl00g 6hgnv-hrvl00g98r 61rnv-lgvl00r98g 61gnv-hrvl00gl25r. 8 h r v - l g v l 0 0 r l l 0 g 8hgv-hrvl00gl00r 81rv-hgvl00rl50g 81gv-lrnv50g60r •8hrnv-hgv80rl00g 8hgnvlrvl00g90r 81rnv-hgvl00r80g 81gnv-lrnvl00gl00r 10hrv-hgvl00rl00g. 10hgv-lrvl00gl00r ; : lOlrv.-lgvlOOrlOOg lOlgv-hrvlOOglOOr 10hrnv-lrvl00r200g. v lOhgnv-lrnvlOOglOQf," 101rnv-hgnvl00r90g~ 101gnv-lrnvl00gl50r 225-Utirv-lgvlOOrlOOg Hhgv-hrnvlOOglOOr Ulrv-hgvlOOrlOOg Illgv-hrv40g30r Ilhrnv-lgv50rl00g Ilhgnv-hrvl00g300r Ulrnv-hgnvlOOrlOOg Hlgnv-hrnvlOOglOOr 13hrv-lgv60r60g .. 13hgv-hrvlO0gl00r ; 131rv-hgnvlOOrld6g 131gv-hrnvl00gl00r 13hrnv-hgnvl00rl00g 13hgnv-hrnvl00gl30r 131rnv 131gnv-hrvl00gll0r 12hrv-hgnv60r50g 12hgv-lrnvl00gl00r 121rv-hgvl00rl00g 121gv-hrv50g50r 12hrnv-hgnvl00r80g 12hgnv-lrv80g90r 12hrnv-hgnvl00rl50g 121gnv-lrv40g50r 14hrv-lgnv50r75g 14hgv-lrv65g50r 141rv-hgvl00rl50g 141gv-hrvl00g95r 14hrnv-hgvl00rl00g 14hgnv-lrnvl00gl00r 141rnv-hgnvl00rl00g 141gnv-hrv80gl00r \ 226 I I . Set B. I n i t i a t i o n s on F i r s t Exchange Opportunity ( T r i a l 1) Key to notation: rv: Red v i s i b l e 1 gv: Green v i s i b l e 1 hrnv: High red non-visible 1 ) hgnv: High green non-visible 1 ) l r n v : Low red non-visible 1 ) lgnv: Low green non-visible 1 ) Set B l Set BII As there were two i d e n t i c a l subjects i n each experiment, they w i l l be d i f f e r e n t i a t e d i n the following manner: r v l rv2 etc. F i r s t d i g i t i n entry r e f e r s to number of experi-ment, 1-14. Note: The in i t a t i o n s o o n T r i a l 1 from high non-visibles i n Set B l , and from a l l subjects i n Set BII were intercepted, and f a l s e o f f e r s substituted. There-fore, there i s no entry f o r 'Accepted' or 'Rejected', except for v i s i b l e s i n Set B l , who received r e a l o f f e r s . ' Example: l h r n v l - g v l lOOrllOga: In Experiment I, one of the high red non-visibles made an o f f e r of 100 red i n return f o r 110 green to one of the green v i s i b l e s , and t h i s o f f e r was accepted ( v i s i b l e s i n Set B l received r e a l o f f e r s ) . A. SET B l : FIRST EXCHANGE OPPORTUNITY Irvl-gvl50rl00gr Igvl-rv2100gl00ra Irv2-hgnv210r20g Igv2-rvll00gl00ra Ihrnvl-gv2100rl00ga Ihgnvl-rvll00g200rr Ihrnv2-hgnvll00rl00g Ihgnv2-rv270gl00rr 2rvl-gvl50r50ga 2gvl-rv240gl00rr 2rv2-gvll00rl00gr 2gv2-hrnv250g50r 2hrnvl-hgnv2100r60g 2hgnvl-rvll00g300rr 2hrnv2-hgnvll00rl00g 2hgnv2-rvl75gl00ra 3rvl-gvll00rl00ga 3gvl-rvll00gl00ra 3rv2-hgnvll00rl00g 3gv2-hrnv2100gl75r 3hrnvl-hgnvl75r80g 3hgnvl-hrnvll00gl00r 3hrnv2-gvl50rl00ga 3hgnv2-hrnv2100gll0r 5rvl-hgnvl50rl00g 5gvl-hrnvl30g30r 5rv2-gv2100rl00ga 5gv2-rv2100gl00ra 5hrnvl-hgnvll00r300g 5hgnvl-hrnvll00g75r 5hrnv2-hgnv2100rl00g 5hgnv2-hrnv2100gl20r 7rvl-9v2100rl00gr 7gvl-rvll00gl00rr 7rv2-hgnv250r50g 7gv2-rvll00gl00ra 7hrnvl-rv2100rl25gr 7hgnvl-hrnvl50g50r 7hrnv2-gv2100r80ga 7hgnv2-rv250g75ra 4rv l - g v l l 0 0 r l 0 0 g r 4gvl-rvl80g75ra 4rv2-hgnvll00r95gs. 4gv2-rv2100gl25ra 4hrnvl-hgnv2100rl20g 4hgnvl-rvll00gl00rr 4hrnv2-gvl50r75gr 4hgnv2-h4nvll00gl00r 6rvl-gv2100rl00ga 6gvl-rvll40gl40ra 6rv2-hgnvll00r90g 6gv2-rv2100gl20rr 6hrnvl-gv2100rl00gr 6hgnvl-hrnv250g50r 6hrnv2-hgnv2100r90g 6hgnv2-hrnvl50g50r Sent to wrong colour. 228 B. SET BLI: FIRST EXCHANGE OPPORTUNITY (No r e a l o f f e r s delivered) 8rvl-lgnvl90r90g 8gvl-rvll00g200r 8rv2<-lgnv2100rl00g 8gv2^rv250g20r 81rnvl-gvll00rl00g 81gnvl-rvll00gl50r 81mv2-gv2100rl00g 81gnv2-rvll00gl00r 10rvl-lgnvll00r75g 10gvl-rvl80g80r 10rv2-gv2100rl00g 10gv2-rvll00gl50r 101rnvl-gvl60rl00g lOlgnvl-lrnvllOOglOOr 101rnv2-gv2100rl00g 101gnv2-lrnvll00gl80r 12rvl-lgnv280rll0g 12gvl-rvll00g75r 12rv2-gvll00r200g 12gv2-rv2100gl00r 121rnvl-lgnv2100rl00g 121gnvl-rvll00gl00r 121rnv2-gvll00rl00g 121gnv2-rv280gl00r 14rvl-gv2100r200g 14gvl-lrnvll00gl00r 14rv2-gv250r60g 14gv2-rvll00gl00r 141rnvl-gv2100rl00g 141gnvl-rv280g70r 141rnv2-gvl90r90g 141gnvl-lrnvll00gl00r 9rvl-gv2100rl80g 9 g v l - l r n v l l 0 0 g l 0 0 r 9rv2-lgnv2100r80g 9gv2-lrnv2100gl00r 91rnvl-gvll00r200g 91gnvl-rvll00gl00r 91rnv2-lgnvl80r75g 91gnv2-lrnv2100gl00r Ilrvl-gv2100rl00g Ilgvl-rv280g60r I l r v 2 - g v l l 0 0 r l 0 0 g Ilgv2-rvll00gl00r I l l r n v l - g v l l 0 0 r 3 0 0 g Illgnvl-lrnv2100gl00r Illrnv2-lgnv250r70g Illgnv2-rv2100gl25r 13rvl-gvll00rl00g 13gvl-rvl60g60r 13rv2-gv2100rl00g 13gv2-rv2110gl00r 131mvl-gvll00rl00g 131gnvl-rv2100gl00r 131rnv2-lgnv2100rll0g 131gnv2-rv2100g250r 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100036/manifest

Comment

Related Items