UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Essays on new venture survival and growth Thornhill, Stewart 1999

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

Item Metadata

Download

Media
831-ubc_1999-464350.pdf [ 8.95MB ]
Metadata
JSON: 831-1.0099468.json
JSON-LD: 831-1.0099468-ld.json
RDF/XML (Pretty): 831-1.0099468-rdf.xml
RDF/JSON: 831-1.0099468-rdf.json
Turtle: 831-1.0099468-turtle.txt
N-Triples: 831-1.0099468-rdf-ntriples.txt
Original Record: 831-1.0099468-source.json
Full Text
831-1.0099468-fulltext.txt
Citation
831-1.0099468.ris

Full Text

E S S A Y S O N N E W V E N T U R E S U R V I V A L A N D G R O W T H by STEWART THORNHILL B.Sc. (Eng.), The University of New Brunswick, 1987 M.B.A., The University of British Columbia, 1993 A thesis submitted in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Commerce and Business Administration) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA June 1999 © Stewart Thornhill, 1999 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 or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of (0mmefP-C^T The University of British Columbia Vancouver, Canada DE-6 (2/88) A B S T R A C T This thesis is comprised of three essays dealing with the survival and growth of business enterprises. The first paper (Chapter 2) explores a long-standing question in corporate venture management: How closely should a corporate parent link itself with its own venture? We challenge the conventional view that autonomy is best for venture growth by arguing that access to the parent's resources and capabilities (i.e., a "tight fit") is essential if a venture is to demonstrate competitive advantage. Data from 97 Canadian corporate ventures generally support the "tight-fit" hypothesis. We also find empirical support for the proposition that the relationship between a corporate parent and its venture(s) evolves over time; economic ties diminish with venture maturity, relational ties remain intact. The next paper (Chapter 3) models the growth and decline of young firms as a function of their initial asset stocks, initial capabilities, rate of capability development, rate of asset depletion, and failure threshold. Data from 246 Canadian corporate bankruptcies confirm that young firms fail due to insufficient organizational capital at start-up and inadequacies in managerial knowledge, financial management skills, and marketing abilities. Older firms, on the other hand, are more prone to failure due to environmental change. The final paper (Chapter 4) utilizes detailed survey data from a proportionally stratified, representative sample of 3,000 Canadian firms to evaluate industry- and firm-level determinants of young firm growth. The competitive environment is found to be a poor predictor of the growth of young firms. In general, growth of the seven to ten year old firms in our study did not follow the growth trends of the industries in which they operated. Among firm strategies, innovation was the strongest predictor of revenue growth. Also of note was the finding that different types of managerial experience were significant in different sectors. For service firms, general ii management experience was positively associated with growth, while for goods-producing firms industry experience was a more important factor. iii TABLE OF CONTENTS Page ABSTRACT ii TABLE OF CONTENTS iv LIST OF TABLES vi LIST OF FIGURES vii PREFACE viii ACKNOWLEDGEMENTS ix CHAPTER 1: INTRODUCTION AND OVERVIEW 1 REFERENCES 10 CHAPTER 2: 11 A DYNAMIC PERSPECTIVE OF INTERNAL FIT IN CORPORATE VENTURING.... 11 A B S T R A C T 11 INTRODUCTION 11 CONCEPTUAL DEVELOPMENT 13 Corporate Venturing 13 Internal Strategic Fit 15 Economic Dimension 17 Relational Dimension 18 Parent-Venture Dynamics 20 METHODS 22 Sample 22 Measures 24 Performance Criteria 24 Economic Measures 25 Relational Measures 26 Analysis 26 RESULTS 28 DISCUSSION 33 CONCLUSION 36 REFERENCES 38 APPENDIX 43 CHAPTER 3: WHY DO YOUNG FIRMS FAIL? MANAGERIAL CAPABILITIES, ORGANIZATIONAL ASSETS, AND THE LIABILITY OF NEWNESS 49 ABSTRACT ; 49 INTRODUCTION 49 LITERATURE REVIEW 51 The Liability of Newness 51 Contrasts between Organizational Ecology and the RBV 54 Definition of Failure 54 Causes 56 THEORETICAL DEVELOPMENT 57 iv Organizational Mortality and the Resource-Based View 57 A Model of Young Firm Survival and Growth 58 Hypotheses 64 Newness 64 Age 65 M E T H O D O L O G Y 66 Sample 66 Analysis 69 RESULTS 72 DISCUSSION 76 REFERENCES 80 APPENDIX 85 CHAPTER 4: DETERMINANTS OF YOUNG FIRM GROWTH: EVIDENCE FROM CANADA 98 A B S T R A C T 98 INTRODUCTION 98 THEORETICAL FOUNDATIONS 101 Firm Growth 101 Growth Measurement 103 Competitive Environment 106 Competitive Strategy 109 Differentiation Strategies 109 Innovation I l l Firm Resources 112 Configuration 113 METHOD 115 Sample 115 Weighted OLS Regression Analysis 117 Structural Equation Modeling 119 Measures 120 Growth 120 Determinants of Growth 124 RESULTS 126 Hypothesis Tests 126 The Competitive Environment 126 Competitive Strategy 131 Firm Resources 133 Summary of Hypothesis Tests 135 Exploratory Analysis 138 DISCUSSION 142 REFERENCES 149 APPENDIX 154 COMPREHENSIVE REFERENCES 160 V LIST OF TABLES Page Table 1.1 Chapter Summaries 2 Table 2.1 Sample and Population Frame Characteristics 23 Table 2.2 Mean Scores of Survey Items by Venture Stage and Performance Category 29 Table 2.3 Composite Mean Score Comparisons between High and Low Performing Ventures 29 Table 2.4 Logistic Regression of Factor Scores 30 Table 2.5 Dynamic Evolution of Parent-Venture Relationship 33 Table 2.6 Summary of Hypotheses and Results 34 Table 3.1 Pearson Correlation Matrix 69 Table 3.2 Means Tests and Logit Analysis of Hypotheses 1-8 70 Table 3.3 Combined Logit Analysis 71 Table 4.1 Growth Rates by Industry 106 Table 4.2 SOFP Industry Classifications 117 Table 4.3 Population and Respondent Counts by Strata 118 Table 4.4 Weighted Summary Statistics and Correlation Matrix 123 Table 4.5 Weighted Means by High, Medium, and Low Growth Categories 122 Table 4.6 Weighted OLS Regression of Environmental Effects 126 Table 4.7 Structural Equation Model of Environmental Effects 127 Table 4.8 Weighted OLS Regression of Derived Principal Components Factors 129 Table 4.9 Weighted Logistic Regression of Derived Principal Components Factors 130 Table 4.10 Weighted OLS Regression of Strategic Effects 131 Table 4.11 Structural Equation Model of Firm Strategy Effects 132 Table 4.12 Weighted OLS Regression of Firm Resources 133 Table 4.13 Structural Equation Model of Firm Resource Effects 134 Table 4.14 Weighted OLS Regression of Environment, Strategy, and Firm Resources 136 Table 4.15 Summary of Hypotheses and Results 137 Table 4.16 Structural Equation Model of Firm Strategy and Resources 143 Table 4.17 Weighted OLS Regression of Firm Resources on Firm Strategies 144 vi LIST O F FIGURES Page Figure 2.1 Dynamic Model of Parent-Venture Fit 21 Figure 3.1 Hazard rates for Canadian firms 54 Figure 3.2 Illustrative growth profiles 63 Figure 3.3 Age distribution of bankrupt firms 68 Figure 4.1 Multi-level Structural Model of Young Firm Growth 99 Figure 4.2 Schematic Diagram of Environmental Effects Model 127 Figure 4.3 Schematic Diagram of Strategic Effects Model 132 Figure 4.4 Schematic Diagram of Firm Resources Model 134 Figure 4.5 Simplified Schematic Diagram of Firm Resources and Strategic Effects Model... 139 vii PREFACE Chapter 2 of this thesis has been accepted for publication in a forthcoming edition of the Journal of Business Venturing. The article is entitled "A dynamic perspective of internal fit in corporate venturing." It is jointly authored by Raphael Amit. The first version of the paper was solely authored by Stewart Thornhill. Subsequent revisions included valuable contributions from Professor Amit, who is also the chair of my dissertation committee. He has acknowledged that I contributed a majority of the substance of the paper. This is reflected in the ordering of the authors' name, which is not alphabetical. viii A C K N O W L E D G E M E N T S I am deeply indebted to Raffi Amit for the guidance, direction, and intellectual support that he has provided during my program of study at UBC. I am also very grateful to Dev Jennings, Marty Puterman, Monica Belcourt, and Craig Pinder for keeping me focused on my studies and for nurturing my development as a researcher. Janet Gannon, Janet Moore, and Charlene Zietsma provided invaluable assistance as readers and editors of the many drafts through which this thesis has evolved. I am grateful for the generous financial support of the Entrepreneurship Research Alliance and the Social Sciences and Humanities Research Council of Canada (Grant # 412 93 0005). I am solely responsible for any errors or omissions contained herein. ix CHAPTER 1 INTRODUCTION AND OVERVIEW A new firm in Canada has only a 50 percent chance of staying in business for five years, and only a 20 percent chance of remaining in operation for at least a decade. The prospects for growth are even more daunting; only one survivor in ten grows by 15 employees or more. Survival in the Canadian economy is a significant challenge. Growing a business is harder still. For new firms, growth and survival are inextricably linked with one another. Obviously, a firm must first survive if it is to grow. But almost as important is the dependence of survival on growth. Firms that enter an industry need to develop relationships with stakeholders, to establish a client base, to generate the all-important cash flows that mean solvency and continued operations. Without growth, the process of economic natural selection will cull more new entrants than it will allow to reach maturity. Even firms that demonstrate the ability to prosper, to survive and grow, cannot allow complacency to settle in. Managers must be constantly vigilant for growth opportunities, whether internal to the firm or external, such as joint ventures or merger and acquisition. This thesis is made up of three distinct essays, which addresses these two critical issues: survival and growth. A summary of important themes from each chapter, including research questions, methodologies, and key findings is presented in Table 1.1. Theoretically, each chapter is anchored in the Resource-based View (RBV) of the firm, within which firms are viewed as heterogeneous bundles of resources and capabilities that may confer competitive advantage (Barney, 1991; Conner, 1991; Rumelt, 1984, 1991; Wernerfelt, 1984). Yet, from this fundamental perspective, each chapter extends the RBV in different ways and approaches the issues of firm survival and growth from different points of view. Each 1 © lo u cu x cu s o •S CD I -8 c 00 — C + H « ° s M ,5 -+-» tn « c c .2 '5 JJ § •§ U in 6 0 1 S •a ci J S 2 .. ? a o a u cn a -a ss o u. ' f e CS cu cu fe 60 cu e -S co +2 -S J> cu Q- 3 cu c S o o o o g g H E E ' rt ' X> 'o' > CD to » J S <u ;r| co co ^ -5 -3 E-C O C O -in 5 cu c -° u cu U 60 3 O 3 « 60 "~> cu t. _ p « = co JS o .is o c C S e co ~ 3 CD rt ^ J S N « E § ss o go < o ca E 3 o a JO 3 o -a ft <» cu -! O 3 co fe C co ^ g ^ O CO tH ON c T rt C"l oo ON -a CCS ss S3 CJ CD W CO I D ON ON o-> .£ £ oo t> g tN J> -a 3 a cu > rt fe> "3 t50 co o C O 1—4 fi 2 60 3 cu o rt O W T3 CU CCS _ 3 00 O 'I s co e u IE ex s s iS fe. JS CJ 'co S3 O a. CU T3 rt o CU T3 .2J CU CU .S3 - M » co rt cu I £ C CS , * i , S ? C O C O CD co l -C O Q) C U rt r - D H CU rt •8 o .Si cU ^ J S o cS oi rt J S -*-» C O r^. is .2 CU o Is o o c CD J3 rt ^ £ U—( o O J J * C +J sS .2 J2 O co "a> C 2 o E X £ o <u ~ c -4 C rt J3 T3 cu rt J S C rt rt S c^ (N .5 rt CU 60 < C co CU " T3 60 3 C in w cU "5  cu hn H O c co « .2 S tn « rt cu cu CD CU 60 a <U ^ rt rt rt CD 0 'S 2 '-c D £ 3 CD s— =S •4—> aj ts ts « o 5b 6JD B B . w l> CU OS 1. © a u o U i co T3 — C rt rt c CD o o rt a. c CD a. CD CD o ts S rt 8 £ CD _fe "3 fe CU CU CD i l J3 > ^ . & ^ J S - ~ co -o o J S CD > rt i> .2 C O -4^ rt rt * fi CD u CD -O 3 C o rt C O CD ts cs J-r ^ co ^ cu w S g 8 •B 2 B co « S CD J i CU -H a. > J S *G c o a o _ cd co ' B +j Cn rt C cu cu CU O H wH CU T3 CU rt - § o w _ C O 5 CD SS 60 O -CD fi --« * .2 c rt rt 3 «> CU J2 fe +3 CU CD co rt CD i -~ O *— <+3 O CN E" CD ° & CD O .y w W CD >< 3 CD ' r t "CD 3 tZ cu „ cu rt ! « J S cx cu ts 3 g tJ ^ O '35 ° c fi C J CD b y -*-» r- 3 £-1* E -5 t -. - rt O •O C CD C CD > CD ts O & o U T3 rt 3 rt U ON co ts „«» •S 3° 3 ;= —^' +-* *- o C O H P -CD M H CD C O + H C co § c^ CU t-H . S rt rt • S ij o ^ 5 .2 • g C 60 1.1 1! 6< O) m ft, I C5 ft o IO i n bO C ca o cs . E CS co •a C+H _H o cs a o CD 3 <o « 5? i o cn 1H cn a) s z o c j co — 3 2 « E > « o .2 E a, r -" I o p hS 00 o T3 3 cS cn <u o 3 O . co co a o >, . 3 w fe * 1 «» , 3 3 cS cS ' ' 3 cS o _s bD s -a = 3 cd bp S C tj_, _2 o &.S O "O o. 3 C , 3 HH UH o c CU cd O fa S s i -£ U -a 3 CS l£3 3 T3 3 C3 £ J? co CS OJ E .8 £ +- X) CS — cS O ^  cS . 3 H J t .  •O "5 o cu Q , CS •3 o. o -== 1 £ c j O E tab OJ ex s . « l> CU -HJ C3 S-o a. s-o U E o £ o o u « u . cS .3 s 3 J S s £ S2 cs 3^ •S is H J 'J-J < c 3 O •3 J3 J2 cs ti > '*3 m O -a c OJ HO WH OJ cn _o o 3 3 OJ OJ OJ > 3 3 E — « OJ >> cS O -co OJ C OJ OJ CS 3 3 •M a j b ^ a .2 C 3 o M CS CN T3 cS OJ • H-cS '5 2 2 '« p. o S cn O cn O cS « OJ -4—» cS OJ JS cn OJ W M g cS O +H CJ . A 2 > a — « cn O cs > C OJ to OJ 43 3 ta 'fe 3 J3 OJ T3 > T3 TD 3 O o J3 HJ OJ cn H J OJ cn • — cS 2 3 CS cS cs .2 cn 3 O o OJ 3 . = U OJ m cs 3^ > 3 T3 U. 0> O o E « OJ So go c 3 cS O -3 j3 O cn O O 0> 1=0 a 'cn cS OJ LJ o OJ -a TD 3 cS - 3 >> o E « 2 3 3 O CS o 3 T3 CJ 3 2 E <3 OJ CS _ N _3 i n cS -2 85 3 © E S H ^ cS X> E g E o c j O H J OJ 3 -o OJ cS T 3 Is 3 O 'HJ O OJ DJ cn e i H J 3 co O cS cn ^ ^ n oj o CJ T3 3 cS 3 OJ c j 3 > .5 cS OJ 2 o o o o = - i D O CJ > m OJ cS u 3 -3 •I CJ _ 3 § 3 -3 s .s O J3 E ^ cS O 3 *->> M - O T3 s 3 S O — — .2f 3 bO o > H J OJ OJ T3 5 chapter asks different research questions and illuminates different scenes from within a complex landscape. The second chapter, "Performance Implications of Internal Fit in Corporate Venturing," seeks to understand the processes that occur when firms attempt to grow through internal development. This line of research centers on large organizations that are attempting to renew themselves, to recapture the entrepreneurial spirit that they needed when they were young and small - a spirit that is often so elusive to large, established firms. In essence, the large corporations that choose to engage in corporate venturing are attempting to grow new firms, similar to the organizations under study in Chapter 4, while also working hard to avoid the liabilities of newness that are discussed in some detail in Chapter 3. These corporate ventures have advantages that independent new ventures lack - the resources and capabilities of their corporate parents. In Chapter 2, we hypothesize that access to these resources and capabilities will be critical to the success of these corporate ventures, a viewpoint that is new to extant theory in the corporate venturing literature, and one which is supported by our survey of Canadian corporate ventures. Thus, the RBV anchors our model of performance, and we extend its breadth of application into the domain of firm renewal through corporate venturing. Chapter 3, "Why Do Young Firms Fail?" also develops and extends the RBV, but in a rather different direction. This chapter seeks to bridge the research streams of strategic management and organization theory. More specifically, the paper develops a firm-level model of young firm growth and survival with the intention of explaining and understanding observed population-level phenomena known as the liabilities of newness and adolescence (Stinchcombe, 1965; Fichman & Levinthal, 1991). It has been repeatedly established that young firms fail more often than do old firms, and that small firms fail with greater frequency than do large ones (Carroll, 1983; Freeman, Carroll & Hannan, 1983). It is also known that firm mortality risk is 4 greatest not at start-up, but at some later point in time, following a so-called honeymoon period (Bruderl & Schussler, 1990; Fichman & Levinthal, 1991). We suggest that a simple model, using firm age, managerial capabilities, costs of staying in business, and a learning parameter, can portray the dynamics of resource acquisition and depletion that, in turn, determine whether a firm will grow and survive. This theoretical extension directly addresses the twin issues of survival and growth that are the essence of this thesis. In the final chapter, "Determinants of Young Firm Growth," the theoretical approach is less one of extending the RBV than on focusing it to examine specific issues pertaining to the growth of young enterprises, firms that have survived past the stage of adolescence and into young adulthood. Much of the extant research in strategic management has concentrated on large, well-established firms and it is not clear whether the paradigms that have emerged in this context necessarily apply to younger, smaller business establishments. This chapter draws on a representative sample of nearly 3,000 young Canadian firms to test a series of hypotheses that are developed from the existing theoretical frameworks of entrepreneurship and strategic management. In general, while innovation and certain types of prior managerial experience are found to be positively associated with young firm growth, other anticipated relationships (e.g., the influence of the competitive environment) were not borne out by the data. Each chapter draws on a different database for empirical analysis. The nature of the respective data sets, along with the issues addressed in each chapter's research questions, jointly determined the methodologies that were employed. Both Chapter 2 and Chapter 3 rely on dichotomous dependent variables - milestone attainment in the case of the corporate ventures (Ch. 2) and young versus old failure in the case of the bankruptcies (Ch. 3). So, although very different issues are under study, the same analytical tools are employed in both chapters: univariate means tests, joint means tests, and logistic regression. 5 Chapter 4 differs from 2 and 3 methodologically in two significant ways. First, the dependent variable is continuous, not dichotomous. In Chapter 4, we are interested in understanding the phenomenon of young firm growth as operationalized by change in sales revenue (net of industry growth) over the life of the firms in our sample. Second, the samples utilized in Chapters 2 and 3 were obtained by simple random sampling. The final chapter, in contrast, draws on a complex, stratified sample that can be proportionally weighted to reflect the characteristics of the population frame from which it was drawn. This adds to the complexity of the analytical techniques, but the reward is much greater confidence in the veracity of the results. The main analytical tool of Chapter 4 is weighted OLS regression, although we also employ structural equation modeling as a complementary approach to understanding why some firms grow more than others. By virtue of the different issues of survival and growth addressed by each chapter, the contribution of each to the field of entrepreneurship also varies considerably. As noted above, Chapter 2 extends the RBV of the firm to account for the relationship between a parent and venture in the corporate venturing process. This paper also lends a dynamic flavor to the topic, postulating that the relationship between parent and venture evolves as the venture passes through various stages of maturity. Perhaps the most noteworthy finding of the corporate venturing research was that which confirmed our expectation of a positive relationship between close parent-venture "fit" and successful attainment of milestones by the ventures. This result contradicts some prescriptive writings on corporate venturing in which it has been recommended that ventures be given plenty of freedom and autonomy, rather than risking interference from a large, bureaucratic parent. Instead, our study indicates that close connection is to the venture's advantage, a conclusion that is consistent with the RBV's contention that rare, inimitable resources developed over time are critical to gaining and sustaining competitive advantage. 6 Chapter 3 represents an attempt to bridge two academic disciplines that have long studied similar phenomena from different vantage points. Our empirical analysis of Canadian bankruptcies enabled us to determine that there are differences between the causes of failure among young firms and the causes of bankruptcy among older organizations. Consistent with the predictions of our theoretical model, young firms typically were prone to failure if there were deficiencies in the levels of capitalization or managerial abilities. Among older bankrupts, environmental change was implicated as a cause of failure, although the findings for the cohort of older firms were not as strong as were the attributions among the younger firms in our sample. Past research at the population level of analysis has indicated that size and age were strongly linked with a firm's mortality risk, however there has been little prior research into firm-level causes of firm death. The present study represents a step toward understanding some of the firm-level dynamics that, in aggregate, give rise to the liabilities of newness and adolescence that are so commonly observed among entire populations of organizations. The final paper in the series provides some confirmation of prior research and also reveals some interesting aspects of the differences in growth prospects for firms in different sectors of the Canadian economy. Our empirical analysis of nearly 3,000 young firms, proportionally weighted to represent a sample frame of more than 30,000, strongly indicated that innovation was a distinguishing characteristic of the highest growth young firms in Canada. The data also confirmed an inverse relationship between firm growth and initial firm size. However, the less obvious and arguably more interesting results were those concerning the connection between firm growth and prior managerial experience. We hypothesized that growth and experience would exhibit a positive relationship, however we discovered that different types of experience were related to growth differently in different sectors. Among firms in goods-producing industries, industry experience was positively and significantly associated with 7 growth, while in the service sector, general management experience was linked to positive growth prospects. This speaks to the relative portability of prior experience, and suggests that mobility between industries may be easier for managers of service firms. One other somewhat surprising finding was the relatively low impact of competitive conditions on young firm growth rates. Young firms generally do not follow the growth trends of the industries in which they do business. And subjective measures of competitive conditions, including the levels of hostility and dynamism in the environment, had little predictive power with respect to firm growth. Like any research, each of the three empirical chapters of this thesis has limitations that, in turn, suggest directions for future study. Sample selection, and the match between the actual and ideal data sets, are common issues in academic research. Chapters 2 and 4 each attempt to understand longitudinal growth processes, but they are limited by the cross-sectional nature of their respective data. Chapter 3 suffers less from the issue of temporal compatibility, but it is limited by the absence of a control group of successful firms with which the bankrupts might be compared. Despite their respective shortcomings, each chapter is of value both through the findings that each contains, as well as through the areas for future research that they bring into focus. The corporate venturing study has shown that the relationship between a corporate parent and its venture(s) is a dynamic one, which changes as a function of time and venture maturity. This phenomenon has important implications for venture managers, and for any organization that elects to pursue growth through the internal development of new businesses. Future research into this aspect of corporate venturing would benefit greatly from longitudinal tracking of ventures from the nascent conceptual phase through initiation, and subsequent maturity or failure. As noted above, the study into firm failure lacks a control group of survivors with which the bankrupt firms can be compared. It may be possible to extract comparable firms from the 8 large database in Chapter 4, however at the present time such an endeavor is on hold pending approval from Statistics Canada. The bankruptcy study may also benefit from simulation modeling. The model that is developed in Chapter 3 has a functional form that can be utilized to identify parameters that most closely model the actual population-level dynamics of the Canadian economy. The database introduced in Chapter 4 is by far the most comprehensive of the three, and the one that has the greatest potential for future empirical research. The present study has identified trends that differ among high and low growth firms across the broad classifications of goods and services industries. Future research into the phenomenon of young firm growth will attempt to discover more fine-grained distinctions between industries. As well, the incorporation of financial data from Statistics Canada will enable the analysis to add an important missing dimension to the present study - the relationship between firm growth, strategy, and financial structure. In sum, the three papers presented below each contribute to different aspects of our understanding of firm survival and growth. Each essay is anchored in extant theory, and each draws on empirical data to advance our knowledge about the challenges and opportunities facing young firms in Canada. We have learned that managerial capabilities are critical to the survival of new ventures. This is consistent with our finding that corporate ventures are successful when they have access to the resources and capabilities (including managerial) of the corporate parent. We have further discovered that the nature of managerial experience is relevant to the growth outcomes of firms in different sectors. The quality of management is thus essential to both survival and growth, for independent enterprises and corporate ventures alike. There may not be a magic formula that ensures success in today's economy, but it appears that the managerial skills of the entrepreneur constitute a worthy foundation on which to build a business. 9 R E F E R E N C E S Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management. 17(11. 99-120. Carroll, G. R. (1983). A stochastic model of organizational mortality: Review and reanalysis. Social Science Research. 12(41. 303-329. Conner, K. R. (1991). A historical comparison of resource-based theory and five schools of thought within industrial organization economics: Do we have a new theory of the firm? Journal of Management, 17 (11. 121-154. Fichman, M . , & Levinthal, D. A. (1991). Honeymoons and the liability of adolescence: A new perspective on duration dependence in social and organizational relationships. Academy of Management Review. 16(21. 442-468. Freeman, J., Carrol, G. R., & Hannan, M . T. (1983). The liability of newness: Age dependence in organization death rates. American Sociological Review, 48(41. 692-710. Rumelt, R. P. (1984). Towards a strategic theory of the firm. In R. Lamb (Ed.) Competitive Strategic Management, (pp. 556-570). Englewood Cliffs, NJ: Prentice-Hall. Rumelt, R. P. (1991). How much does industry matter? Strategic Management Journal. 12(31. 167-185. Stinchcombe, A. L. (1965). Organizations and social structure. In J.G. March (ed.), Handbook of Organizations, (pp. 142-193). Chicago: Rand-McNally. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal. 5(21. 171-180. 10 C H A P T E R 2: A D Y N A M I C P E R S P E C T I V E O F I N T E R N A L F I T I N C O R P O R A T E V E N T U R I N G A B S T R A C T This paper explores a long-standing question in corporate venture management: How closely should a corporate parent link itself with its own venture? We challenge the conventional view that autonomy is best for venture growth by arguing that access to the parent's resources and capabilities (i.e., a "tight fit") is essential if a venture is to demonstrate competitive advantage. Data from 97 Canadian corporate ventures generally support the "tight-fit" hypothesis. We also find empirical support for the proposition that the relationship between a corporate parent and its venture(s) evolves over time; economic ties diminish with venture maturity, relational ties remain intact. I N T R O D U C T I O N Firms with growth aspirations have several ways of reaching their goals. Mergers, acquisitions, and joint ventures are a few of the better-known approaches to firm growth. Another route, which is of interest to both managers and researchers, is corporate venturing -growing a business from the inside out. The motives for launching a corporate venture include improving corporate profitability, (Zahra, 1991), generating strategic renewal (Guth and Ginsberg, 1990), fostering innovation (Baden-Fuller, 1995) and gaining knowledge that may be parlayed into future revenue streams (McGrath, Venkataraman, and MacMillan, 1994). Researchers have acknowledged the importance of the corporate venture (CV) as a vehicle for firm growth (Arrow, 1982; Burgelman, 1983) and have addressed several issues unique to this growth mechanism. The literature addresses the performance implications of corporate ventures (Biggadike, 1979), the relationship between CV performance and 11 environmental context (Covin and Slevin, 1994; Tsai, MacMillan, and Low, 1991; Zahra, 1993), the role of compensation practices within corporate ventures (Block and Ornati, 1987), and the influence of CV champions (Day, 1994). The relationship between a corporate parent (CP) and its corporate venture has also been studied (Miller, Spann, and Lerner, 1991; Sorrentino and Williams, 1995). Little has been done, however, to empirically test whether the connection, or fit, between parent and venture influences CV performance. While some authors have argued that high levels of relatedness between CP and CV are desirable (Dougherty, 1995; MacMillan, Block, and Narasimha, 1986), others have contended that tight coupling is antithetical to venture success (Burgelman, 1983; Ginsberg and Hay, 1994; Sykes and Block, 1989). In his report on 37 ventures at Exxon Enterprises, Sykes (1986) identified reasons both for and against the practice of allowing venture autonomy. The lack of consensus on this issue leads us to the first of two complementary research questions that we address in this paper: "What is the effect of internal strategic fit between a corporate parent and its venture on venture performance?" Following the resource-based view of the firm (Penrose, 1959; Wernerfelt, 1984), which argues that competitive advantage derives from idiosyncratic capabilities that firms develop internally, we suggest that a tight fit is positively associated with venture performance because of the venture's access to its parent's resources. This position, along with several specific hypotheses, is developed in the sections that follow. A number of studies have argued that corporate venturing is a dynamic process, that is, one in which the relationship between parent and venture evolves as the venture matures (Burgelman, 1983; Garud and Van de Ven, 1992; Schrader and Simon, 1997; Sykes, 1986). Our second research question addresses this issue by asking: "Does the relationship between a corporate parent and its venture(s) evolve over time, and if so, how? " 12 In the following sections, we develop a model of the CP-CV relationship that incorporates the economic and relational dimensions of firm growth within a dynamic, evolutionary framework. This perspective is anchored in both the extant literature and a series of interviews with corporate venture managers. We formulate specific hypotheses from the model and test them with survey data from a sample of 97 Canadian corporate ventures. Finally, we present and discuss the results of the empirical tests. C O N C E P T U A L D E V E L O P M E N T Corporate Venturing Our review of the literature on the processes and outcomes of corporate venturing (summarized in Appendix, Table A2.1) reveals a few points of general agreement. First, it is generally agreed that corporate venturing has a positive effect on firm performance (Biggadike, 1979; Zahra, 1991, 1993; Zahra and Covin, 1995), although such benefits are not guaranteed and ventures may take several years to become profitable. Second, ventures go through a series of stages as they mature (Garud and Van de Ven, 1992; McGrath et al., 1994; Schrader and Simon, 1997). Though a number of classification schemes have been suggested in the literature, there is general agreement that the nature of CVs is dynamic, not static. Third, and in keeping with the evolutionary nature of the ventures themselves, there is almost unanimous agreement that milestones are the best method for evaluating CV performance (Block and MacMillan, 1993; Block and Ornati, 1987). While it is reassuring that there are some areas of convergence in the literature, there are also several areas on which there is little or no agreement. Researchers disagree, for example, about the desired tightness of coupling or fit between parent and venture. The degree of fit may be thought of as a continuum, anchored at one end by what Sykes (1986) refers to as "total 13 congruence." In this case, "the 'venture' is no more than a new product extension by an existing operating division, and, even if innovative, would probably not qualify as 'internal venturing'" (Sykes, 1986: 281). At the other end of the continuum is an independent entrepreneurial enterprise (Miller et al., 1991; Schrader and Simon, 1997). The debate revolves around which point on this spectrum is optimal for corporate venture performance. The advantages of a close fit between parent and venture include resource sharing (e.g., access by the venture to the parent's suppliers and distributors) and the availability of internal corporate capital. On the other hand, ventures with greater autonomy may be free from the entrenched bureaucratic processes of the corporate parent and more flexible in their response to changing internal and external demands. Effective corporate venturing has been described as a balancing act with needs for creativity and change on one side and demands for cohesiveness and complementarity on the other (Lengnick-Hall, 1992; Tushman and Nadler, 1986). The few studies that have directly addressed internal fit have yielded mixed results. Ginsberg and Hay (1994), for example, argued that the flexibility associated with autonomy facilitates C V success. Similar conclusions were presented by Dougherty (1995) and Block (1989), based on the premise that the pressures and rigidities emanating from a corporate parent adversely affect venture performance. A similar argument contends that management practices that work for large corporations are inappropriate for ventures (Block, 1983; Kanter, 1985; Sykes and Block, 1989). A study of 88 industrial product corporate ventures from the PIMS STR4 database found that relatedness between corporate parents and ventures does not affect venture performance (Sorrentino and Williams, 1995). This finding echoes the results of a similar study, also using PIMS data, in which the reporting level for CVs was found to have no main effects on venture performance (Miller et al., 1991). However, Lengnick-Hall (1992) presented evidence in favor of 14 a close fit in her study of 86 firms sampled from the Business Week 1000. Based on the results of discriminant analysis, she concluded that "the price of neglecting organizational consistency is increased organizational problems" (1992: 147). Internal Strategic Fit Our model of the relationship between a firm and its corporate venture(s) borrows from the grounded theory methodology of Glaser and Straus (1967). Under this methodology, theory is developed from and thereby grounded in data. This approach explicitly recognizes that data collection and analysis precede hypothesis building. In fact, the theory should emerge from the data as it is coded, categorized, and analyzed. In Glaser's words: When the grounded theory approach is used the researcher constructs his theoretical framework out of the data. Through comparing the data as it is collected, the researcher creates more abstract levels of theoretical connections. In short, theory is gradually built up inductively from the progressive stages of analysis of the data. (1978: 39) We conducted seventeen semi-structured interviews, one to two hours in length, with senior executives and venture managers of nine large Canadian corporations that had engaged in a wide range of corporate ventures. The issue of growing businesses in businesses was approached with some initial concepts in mind and a general idea of a research strategy. This perspective is amenable to grounded theory research provided that data is gathered with an open agenda and, during analysis, that theoretical constructs are allowed to emerge rather than being forcibly fit into preconceived categories. The interview structure was based on a model in which growth factors for new ventures were expected to align themselves according to their relevance to either the venture or the parent company. An internal versus external alignment was also anticipated. The responses of the interview participants led to a revision of this framework. A second model was developed in which centralization versus decentralization of control was emphasized. This also proved to be unworkable. Our subsequent analysis of the interview transcripts guided our construction of a 15 model in which the parent and venture interact on the basis of economic drivers (based on the resources of the parents and needs of the venture) and relational drivers (flowing from the structures and cultures of the parent and venture). The economic aspect of the CP-CV relationship was expressed by one venture manager in his observation that "Certainly the deep pockets of (the parent company) helped because I had lost a lot of money." The nature of the economic ties is complex and often difficult to manage, a sentiment expressed by another senior executive who stated that "We're probably under-funding ... We're hobbling our young entrepreneurs too much." Yet, the same executive was quick to qualify his remark: "... but, then again, I'm not so sure that that doesn't create innovation." Our interviews also revealed a relational dimension to the CP-CV structure. In the words of one CEO, "I think that building a new business is very much managing relationships." Two of our interviewees also drew analogies between child-rearing and corporate venturing. This metaphor not only captures the relational issue; it also encompasses the evolving, dynamic nature of the relationship. The economic dimension pertains to issues such as investment and compensation, while the relational dimension involves issues such as the levels of support and trust that exist between a venture and its corporate parent. These categories are similar to the intrinsic and extrinsic dimensions proposed by Sykes (1986). Under his typology, the extrinsic dimension captures the relationship between a venture and its corporate sponsor and includes structural and procedural sub-dimensions. The intrinsic dimension pertains to the characteristics of the venture itself and is divided into product-related and managerial facets. While Sykes's dimensions are respectively parent-focused and venture-focused, our dimensions portray the parent-venture relationship in distinct economic and relational terms. 16 Economic Dimension Our interviews revealed four aspects of the economic dimension of the parent-venture relationship. The first stems from the parent's reasons for launching a venture. Ventures launched with the objective of earning a target return on investment may be run very differently from ventures launched for defensive (responding to competitors' initiatives) or developmental purposes. McGrath (1995) argued that ventures must be able to demonstrate "market worth," (i.e., economic viability) without which they will be unlikely to survive. The prospect of turning a profit should also enhance support within corporate top management, further enhancing the likelihood of venture success. We anticipated that ventures that are anchored in well-developed business plans with articulated, profit-based objectives would experience greater success than those that are not. This raises the issues of performance evaluation and accountability. There are many ways to evaluate the success of a venture. For the purposes of this study, we define success according to a venture's ability to meet milestones on schedule (see Method section below). While there are as many different types of milestones as there are new ventures (e.g., target shipping dates, market share, ROI), profitability eventually enters the discussion. Block (1993) observed that lax financial controls are among the more common causes of corporate venture failure. We anticipated that firms that evaluate their ventures on the basis of financial targets would be successful more often than those that eschew financial benchmarks. Another dimension of corporate venturing is the capital stake that the parent commits to a venture. How much the parent invests, whether the funds are sunk and/or restrict redeployment, and whether funds are delivered as promised all send signals to the venture team and to external stakeholders such as clients, competitors, and suppliers about the level of commitment of the parent (Ghemawat, 1991). We expected that economic commitments in the form of large, specialized, non-recoverable investments would be associated with venture success. The fourth facet of the economic relationship concerns the degree of congruence between the practices of a parent and its venture. This dimension speaks directly to the issue of tight fit versus autonomy. Are venture managers compensated differently than are the managers of the parent company? Are training budgets larger? Is the budgeting process more flexible? Our preliminary interviews and subsequent theory development led us to predict that firms that maintain consistency in the administration of compensation and budgeting practices between parent and venture would be more successful than firms allowing autonomy among their ventures' financial practices. In other words, we expected to see greater success among firms that treat their ventures more like divisions of the parent than like stand-alone entities. This reflects our belief that corporate ventures that do not maintain close connection with the parent forego the resource-based competitive advantages of the parent, and thus hinder their own prospects for success. We thus have the following hypotheses: HI: Ventures selected on the basis of economic decision making (e.g. rate of return) are more successful than those that are not. H2: Venture success is associated with the use of financial targets. H3: Venture success is associated with large, specialized investments of capital by the parent company. H4: Venture success is associated with uniform financial practices between parent and venture. Relational Dimension Ventures often have to compete with other CVs or with other corporate divisions for a limited pool of resources. However, ventures can diminish the effects of competition by 18 operating under the meritorship of a chief executive. Champions are often critical to the survival and success of internal ventures (Day, 1994; Frost and Egri, 1990). We predict that ventures that obtain top management sponsorship, in the form of active support from the parent CEO and the CEO running interference for the venture, experience greater success than those that lack such support. Another element of the relational dimension involves the visibility or preeminence of the venture within the parent company. Hornsby, Naffziger, Kuratko, and Montagno (1993) identified both management support and time availability as factors that contribute to the success of a corporate venture. Venkataraman, MacMillan, and McGrath (1992) emphasized the need to manage both the hierarchical processes and the institutional context within which corporate venturing activities take place. Preeminence may flow from efforts on the part of the venture manager to secure buy-in at the senior management levels of the parent company. Preeminence will also be indicated by the position of the venture on the parent's business agenda. A third element of the relational dimension is confidence, or trust. Barney and Hansen (1994) have identified trustworthiness as a potential source of competitive advantage. Ventures that have faith that the parent will not abandon them when the going gets tough, and whose parents have solid track records of meeting commitments, should be more likely to succeed than ventures without such qualities. A final aspect of the parent-venture relationship mirrors the issue of economic connection and consistency. We predict that autonomy, as indicated by empowerment of venture employees and managers and a strong culture within the venture, is negatively related to venture success. We thus have the following hypotheses: H5: Venture success is associated with active protection and support by the parent CEO. 19 H6: Venture success is associated with preeminence in the eyes of the parent. H7: Venture success is associated with high levels of commitment and trust between parent and venture. H8: Venture success is associated with low levels of venture autonomy. Parent-Venture Dynamics Because of the dynamic nature of the parent-venture relationship, longitudinal studies of venture performance are highly valued and in high demand. Yet they tend to be the exception rather than the norm in entrepreneurship and strategy research. Those studies that have taken a longitudinal approach tend to be characterized by small sample sizes that limit their scope and generalizability. Although our survey is fundamentally cross-sectional, we capture some of the dynamic nature of the CP-CV relationship by asking respondents about parent and venture practices in progressive stages of venture growth and maturity. Our approach is exploratory and the following propositions are tentative, yet we believe that our research is a step in the direction of capturing the dynamic processes underlying the growth of corporate ventures. We propose that the relationship between a CP and CV evolves as a venture matures but that the economic and relational dimensions of the CP-CV fit evolve in different ways. When a corporate parent launches a venture, the venture is critically dependent on the economic resources of the parent. As the venture begins to grow, this dimension may become less important. The relational ties between parent and venture, on the other hand, do not necessarily diminish in importance as a function of venture maturity. Returning to our original position that the connection between CP and CV enables a venture to capitalize on the idiosyncratic, distinctive competencies of the parent, we contend that maintaining close relational fit serves a venture well, regardless of its stage of development. This view of the evolution of the economic 20 and relational ties between parent and venture is captured in the following propositions and depicted schematically in Figure 2.1. Pl: Economic ties tend to diminish between parent and venture as a venture matures. P2: Relational ties tend to remain consistently strong between parent and venture as a venture matures. Figure 2.1 Dynamic Model of Parent-Venture Fit Early Stage Middle Stage Established Stage E: Economic Ties R : Relational Ties Note: The strength of the relationship is indicated by the relative thickness of the arrows 21 M E T H O D S Sample Our initial sample frame comprised 2614 of Canada's largest companies. We sent a screening letter to these firms asking if they had developed any new business units as part of their growth strategy. A business was considered "new" if it had developed any three of the following: new markets, new methods of distribution, new products/services, and/or new technology. A total of 448 firms responded, 261 in the affirmative. The interview phase of our research project, supported by an extensive review of the corporate venturing and strategic management literature, served as the foundation for a detailed survey of practices and processes. The survey, which we pilot tested with senior managers of parent corporations and their ventures, was sent to the CEOs of the 261 firms that had responded positively to our initial mailing. One follow up letter, and a phone call to initial non-respondents, yielded a total of 102 completed surveys. Our rationale for using a mail survey as our method of data collection is consistent with Schrader and Simon (1997), who noted that A survey was the most appropriate means of collecting data, because secondary sources did not contain detailed information regarding companies' resources, strategies, and performance Privately owned IVs (independent ventures) do not publish annual reports, and data on CVs are often subsumed into the sponsors' reports. (1997: 54) Of the 102 responses that we received, one firm was excluded because we felt that its reported venture age (38 years) made it inappropriate to include that firm with a group of relatively young corporate ventures. Four other firms were excluded due to non-response to an item on milestone attainment which we use as the dependent variable of our study (see Measures below). The 97 remaining ventures have a mean age of 3.4 years (with a standard deviation of 3.4 years). The majority of the ventures (82%) are 5 years old or younger. Of the 448 firms that 22 responded to our initial screening letter, we were able to obtain information about industry membership, revenue, and assets for 312 firms. Within the smaller sample of 97 firms that responded to our survey and were retained for analysis, this data is available for 56 firms. Table 2.1 contains summary information about the population frame, the initial response subset, and the final sample. There appears to be some over-representation in our sample in construction, manufacturing, and trade, and under-representation in agriculture and natural resources. However, it is not clear whether this is an artifact of our sample or representative of less corporate venturing in agriculture and natural resources. The mean values for assets and revenues of firms in the population frame, the initial response set, and the final sample do not differ significantly when compared by t-test. Thus, while not perfectly representative in terms of industry affiliation, our sample appears to be representative for size and revenue characteristics. Table 2.1 Sample and Population Frame Characteristics Sector Final Sample Initial Population Respondents Frame # of firms in total 97 448 2614 # of firms with asset & revenue data 56 312 2376 Assets ($M) 1268(2930) 1127 (6418) 1198 (9136) Revenues ($M) 546(1419) 430(1090) 391 (1407) Agriculture and natural resources 19% 34% 36% Construction, manufacturing and trade 27% 20% 18 % T. C. U. and F. I. R. E. 28% 31 % 29% Accommodation and consumer goods 7% 9% 12% Other 19% 6% 4% Most of the firms (80%) had prior venturing experience. Of those, 77% described their previous ventures as positive experiences. The reasons most often cited for launching corporate ventures were to complement existing products/services and to develop new competencies. Other, less common reasons cited for venturing include the utilization of idle resources, and 23 offensive or defensive moves relative to competitors' actions. The firms in our sample all chose to launch ventures in the same industry as the parent's primary business. We were thus unable to evaluate the potential influence of line-of-business similarity with this data. Measures Performance Criteria Measuring the performance of corporate ventures shares many of the difficulties associated with evaluating the performance of small, entrepreneurial firms. The complexity of the issue has been addressed in the literature (Covin and Slevin, 1989; Naman and Slevin, 1993; Sandberg and Hofer, 1987; Sapienza, Smith, and Gannon, 1988) although it remains far from being resolved. Covin and Slevin (1989) identified three reasons for using subjective performance measures of small-firm performance over more objective, 'hard' numerical data: (1) the inability and/or unwillingness of firms to provide financial data (Fiorito and LaForge, 1986), (2) the difficulty of interpretation and comparison of data due to differing firm objectives (Cooper, 1979), and (3) the influence of industry effects (Miller and Toulouse, 1986). Their solution to the problem of performance evaluation was to create a weighted average performance index for firms based upon the product of 'importance' scores and 'satisfaction' scores on a series of questions about various financial criteria (e.g., sales, cash flow, profit margin). A similar approach was used by Venkatraman (1990) who operationalized performance with three indicators, two of which reflect managerial satisfaction and a third that evaluates the performance of the competition. He argued that such measures are "reasonable proxies for often unobtainable secondary-source data (Venkatraman and Ramanujam, 1987)" (1990: 27). Respondents to our survey were asked to indicate on a scale of one to seven the degree to which they agreed or disagreed (1 = strongly disagree, 7 = strongly agree) with the statement that 24 their venture had been able to meet milestones on schedule. We then categorized firms as either High or Low performers on the basis of their responses to this question. Fifty-two ventures, which gave responses of 5 or higher, were classified as High performers while 45 ventures (with responses of 4 or less) were Low performers. Milestones included such measures as profit, revenue, market share, customer satisfaction, and technical objectives. Roughly half of the ventures (52%) relied primarily on measures of profitability such as ROI and ROA. 1 The milestones were typically established with input from both the parent company and the venture. Economic Measures Eleven items on the survey addressed the economic interaction between parent and venture. The first of these asked what proportion of the parent's capital budget had been invested in the venture to date. This item was designed to capture the level of financial commitment of the parent. Another item asked if the venture offered the best potential rate of return among alternative growth opportunities. The remaining nine items were designed to capture the dynamic processes of a growing business unit. Respondents were asked to answer questions categorized by the stage of development of the venture: Early, Middle, and Established. The Early stage was defined in the survey as commencing with financial investment in the venture and continuing until the venture began to generate revenue. The Middle stage begins with the beginning of a revenue stream and continues until the venture realizes a profit. At this point, the venture has become Established. Nineteen of the ventures in our sample described themselves as early stage entities and, consequently, responded only to the early stage portion of multi-stage survey items. Forty-nine firms were in the middle stage and twenty-nine classified themselves as established. Samples of the multi-stage question format are included in the Appendix (Table A2.2). 1 Many of the respondents indicated that they used multiple milestones, including both financial and non-financial measures of performance. As such, we were not able to control for the specific type of milestone(s) used. 25 Our decision to adopt this unique format in our questionnaire stemmed from the preliminary interviews in which managers consistently noted that many aspects of the parent-venture relationship evolve as the venture itself matures. Pilot testing of our survey indicated that managers easily grasped the intent of the multi-stage questions and were receptive to describing the dynamic process of venture management. The multi-stage, economic survey items asked whether funds promised to the venture are ever diverted, whether they are sunk, whether the investment restricts alternative venturing activities, and whether, the investments in the venture are highly specialized. They also inquired about the relative budgets and compensation systems of parent and venture, the use by the venture of the parent's systems, and whether the venture is accountable to financial targets. Relational Measures The relational measures included one single-stage question that asked if the parent would withdraw support if the venture were to experience adverse conditions. The remaining eight items in this category were of the multi-stage format described above. Items included the level of support provided by the CEO, whether the venture manager works to obtain buy-in, the importance of venture culture, the venture's position on the parent's business agenda, whether the venture is protected from politics, and the parent's track record of meeting commitments to the venture. Analysis The mean scores to the survey items, for the high and low performers, respectively, in each of the three stages of venture development are presented in Table 2.2. Simple means tests allowed us to evaluate hypotheses 1, 2, and 7, but the remaining hypotheses were operationalized by several multi-stage variables that should be tested concurrently. For this, we employed two 26 tests, one on the mean scores themselves and the other on composite factor scores. In the first test, we used Hotelling's T-squared in each of the three stages of venture maturity. Results of this procedure for Hypotheses 3, 4, 5, 6, and 8 are presented in Table 2.3. For these multi-item hypotheses, we also used principal components analysis to derive factor scores for the combined variables associated with each hypothesis (the item regarding venture autonomy was reverse coded for consistency with the other variables in H4). Eigenvalues and alpha values for the factors are included in the Appendix (Table A2.3). Once factor scores were derived, we ran logistic regression analyses for each of the three stages of venture development. The results of the logistic analyses are presented in Table 2.4. We also used Hotelling's T-squared in our evaluation of the dynamic propositions of our model (Pl, P2). Table 2.5 contains results of our within-firm comparisons of responses to the multi-stage survey items. First, the difference scores for each venture were calculated between the early and middle stages and between the middle and established stages. Only middle and established firms could be evaluated on their evolution from early-to-middle stage and only for established firms could these results be compared with the middle-to-established stage transition phase. The means of the difference scores, for all ventures in each transition phase, are presented in the columns labeled "Diff." in Table 2.5. Two separate tests were then performed for each variable. First, the mean difference was evaluated against a null hypothesis that the mean is equal to zero. If we could not reject the null hypotheses, then we could conclude that the item varies with increasing venture maturity. Second, we compared the relative change between stages across the groups of high and low performing ventures. In this case, the null hypothesis was that the change among the high-performers is the same as the change among the low-performers. F-statistics and p-values are provided for each test. 27 In addition to testing the changes of individual variables as the ventures mature, we were also interested in the aggregate changes of the variables that jointly formed our multivariate hypotheses. Tests were performed for the joint changes of the variables which represent Hypotheses 3, 4, 5, 6, and 8 (i.e., the vectors of mean difference scores are compared). Finally, we combined all economic variables and all relational variables, respectively, to evaluate the net change of the economic and relational dimensions in the parent-venture relationship. RESULTS Although firms have numerous reasons for engaging in corporate venturing, our results indicate that a venture's anticipated rate of return significantly distinguishes the high performers in our sample from the low performers, as predicted by HI (Table 2.2). Our hypothesis regarding the use of financial targets (H2), however, is not supported by the data. The differences in mean scores indicate a greater use of financial targets by low performers in all three stages, although the results are not statistically significant. We observe, however, that the use of financial targets increases among all firms as the ventures mature. Our expectation that venture success would be associated with significant, financial resource commitments by the corporate parent (H3) is not supported by the data. In fact, an inspection of the mean responses among high and low performers (Table 2.1) indicates that, contrary to our predictions, the low-performers reported sunk funds and non-diversion of funds to a greater extent than did the high-performers. The logistic regression also fails to support H3 -none of the coefficients for the three stages of venture maturity is significant for the factor variable on investment (Table 2.4). One other noteworthy finding is that high-performers relied less on parent resources as they matured, while the opposite is evident for the low-performers, 28 Table 2.2 Mean Scores of Survey Items by Venture Stage and Performance Category Early Stage Middle Stage Established Stage High Low Diffa High Low Diff High Low Diff Economic Dimension Venture offered best rate of return (Hl)b 5.56 4.89 0.67** Venture must meet financial targets (H2) 4.77 4.86 -0.09 5.24 5.65 -0.41 6.14 6.28 -0.13 Level of investment in venture (H3) 2.77 2.48 0.29 Venture restricts alternate venturing (H3) 3.86 3.43 0.43 3.50 2.91 0.59* 2.80 2.42 0.38 Investments in the V are specialized (H3) 5.52 5.40 0.12 5.39 5.03 0.36 5.50 4.09 1.41** Funds in the venture are sunk (H3) 5.54 6.24 -0.70 4.92 5.62 -0.69 4.10 4.93 -0.83 Funds are not diverted from venture (H3) 6.04 6.26 -0.22 6.15 6.06 0.09 6.06 6.36 -0.30 Importance to V of Parent's resources (H3) 6.33 5.76 0.57* 5.75 6.00 -0.25 5.07 6.00 -0.93 Venture budgeting is more flexible (H4) 4.33 4.43 -0.10 3.83 3.94 -0.11 3.15 3.40 -0.25 Venture compensation is different (H4) 3.35 3.49 -0.14 3.14 3.35 -0.21 3.45 3.71 -0.26 Venture has larger training budget (H4) 2.82 3.56 -0.74* 2.79 3.46 -0.67* 2.68 3.06 -0.39 Venture must use parent's systems (H4) 4.60 3.74 0.86* 4.57 4.00 0.57* 4.56 3.53 1.03* Relational Dimension CEO actively supports venture (H5) 6.38 6.39 -0.01 6.26 5.95 0.32 6.27 5.53 0.74* CEO runs interference for venture (H5) 4.39 5.07 -0.68 4.07 4.69 -0.61 3.41 3.81 -0.40 V manager works to obtain buy-in (H6) 5.75 4.85 0.90** 5.74 4.94 0.80* 5.82 4.80 1.02* V near top of P's business agenda (H6) 4.83 4.38 0.45 4.90 4.18 0.73* 4.73 3.71 1.01* P would withdraw if V in trouble (H7) 3.39 4.31 -0.92** P has record of meeting commitments (H7) 5.67 5.50 0.17 5.67 5.46 0.21 5.76 5.27 0.49* Decision making power of V EEs (H8) 4.21 4.89 -0.67* 4.10 4.97 -0.88* 4.28 5.00 -0.72* Decision making power of V manager (H8) 4.38 4.73 -0.35 4.05 4.94 -0.89* 4.24 4.87 -0.63 Importance of venture culture (H8) 6.18 5.93 0.25 6.07 5.71 0.37 6.03 5.87 0.16 Venture is protected from politics (H8) 4.16 4.29 -0.13 4.05 4.34 -0.29 3.85 5.06 -1.21* Identification with V as distinct entity (H8) 4.90 6.17 -1.27** 4.66 5.57 -0.92* 4.96 5.80 -0.84 V EE sense of autonomy from Parent (H8) 4.49 5.07 -0.58* 4.50 5.24 -0.74* 4.41 5.60 -1.19* (a) Superscripts indicate p-values for 1-tailed t-test: t < 0.10; * < 0.05; ** < 0.01 (b) All responses are on a seven-point scale anchored by strongly disagree (1) and strongly agree (7) except the investment question which is scaled as follows: (1) <1%; (2): l%-5%; (3): 6%-10%; (4): >10% of parent's capital budget. Items not reported for the middle and established stages were not framed in a multi-stage format. Table 2.3 Composite Mean Score Comparisons between High and Low Performing Ventures Early Middle Established # F d Probe # Var F Prob # Var F Prob Varc Economic H3: Large, specialized investments 5 1.69 0.15 5 1.22 0.32 5 1.08 0.40 H4: Uniform financial practices 4 1.92 0.12 4 1.02 0.40 4 0.91 0.47 Relational H5: CEO protection and support 2 1.70 0.19 2 1.96 0.15 2 1.99 0.15 H6: Venture preeminence 2 3.57 0.03 2 3.39 0.04 2 3.30 0.05 H8: Low venture autonomy 6 3.99 0.00 6 1.75 0.13 6 1.20 0.34 (c) # Var indicates the number of dynamic survey items used to evaluate the hypothesis (e.g., H3 evaluates restriction of alternate venturing, specialization of investments, sunk funds, diversion of funds, and importance of parent resources; a total of 5 dynamic variables. The level of investment by the parent is not multi-stage and is not included in the calculation of Hotelling's T2). (d) The F-test statistic is derived from Hotelling's T2 by the formula F=((N-p-l)/(N-2)*p))*T2. In this equation, N = n\ + n2 = the total number of observations; p indicates the total number of variables being evaluated (e.g., p=5 for H3) (Stevens, 1996) (e) P-values are based on a null hypothesis that the mean vectors of the high and low performers are equal. (HQ: Hi=Lo) 29 Table 2.4 Logistic Regression of Factor Scores Early1 Middle Established H3: Large, specialized investments -0.16 0.26 -0.35 H4: Uniform financial practices -0.47 -0.34 -0.96 H5: CEO protection and support -0.49f -0.68t -1.40 H6: Venture preeminence 0.92** 1.14* 1.02 H8: Low venture autonomy -0.13 -0.30 0.78 Constant 0.33 0.44 1.88* Observations 61 48 27 Chi-squared 13.81* 11.63* 5.74 Pseudo 0.17 0.18 0.22 Log Likelihood -34.71 -26.40 -10.07 (f) Superscripts indicate p-values: t < 0.10; * <0.05; ** < 0.01 although this finding is not statistically significant when tested for within-venture mean differences. Our prediction that successful ventures would be distinguished by uniform financial practices between parent and venture (H4) is partially supported by the data. Use of the parent's systems is positively and significantly associated with high performance in all three stages. Differences between parent and venture in their budgeting, compensation, and training practices are associated with low performance, as predicted, yet the differences are statistically significant only in the case of training budgets in the early and middle stages. Tests of the combined mean vectors of the H4 variables indicated a low level of statistical significance in the early stage and virtually no significant differences between high and low performers in the middle and established stages, although the sign of the differences does correspond to our hypothesis. The logistic regression coefficients are not statistically significant for the H4 factor variables. The items regarding protection and support of the venture at the CEO level (H5) yield mixed results. Active CEO support is significant in the established stage of successful ventures. However, having the CEO run interference for the venture is more pronounced for the less 30 successful ventures. The combined test of the two variables yields results of weak statistical significance, yet because the two variables are quite different in nature (CEO support predominates among high performers while having the CEO run interference is more characteristic of the low performers), the meaning of this result is not clear. The factor variables in the logistic analysis produce negative coefficients of weak statistical significance, consistent with the influence of the CEO interference item among the sub-sample of low performers. Venture preeminence in the eyes of the parent (H6) does, however, distinguish between firms that met milestones on schedule and those that did not. Ventures that were near the top of the parent's business agenda and whose managers worked to obtain buy-in of the parent are among the success stories of our sample. The univariate results are confirmed by Hotelling's T-squared, yielding significance at the 0.05 level in all three stages. As well, the logistic analysis indicates significant, positive coefficients in the early and middle stages and a non-significant positive coefficient in the established stage. Our expectation that successful firms would believe that the parent would not withdraw support if the venture experienced adverse conditions (H7) is supported by a single-stage survey item. As well, the direction of the difference between high and low performers on a multi-stage assessment of the parent's track record of meeting commitments is consistent with our prediction, although the difference is statistically significant only in the established stage. Finally, our hypothesis regarding the level of autonomy of the venture (H8) is partially supported by the data. Indications of high venture autonomy are characteristic of the low performers, although the degree of statistical significance varies across the individual items. The strongest differences are evident in the items pertaining to relative decision making authority of venture managers and employees, the level of identification of employees relative to parent or venture, and the perceived sense of autonomy of the venture from the parent. In these instances, 31 the data strongly and significantly support the notion that close ties are associated with venture success. The combined tests of mean differences indicate a strong difference between high and low performers in the early stage, and moderate to weak differences in the middle and established stages, respectively. The combined factor variables run in the logistic analysis are not significant predictors of venture performance. Our proposition that the economic ties between parent and venture would diminish with venture maturity (Pl) is strongly supported (Table 2.5). In the early-to-middle stage transition phase, the combined test of H3 and H4 variables (Hotelling's T 2) indicate decreasing mean responses with significance at the 0.01 level. The combined test of all multi-stage economic variables is also negative (as predicted) and significant at the 0.01 level. These results are also found across the middle-to-late transition phase. The sole exception to the pattern of decreasing economic connection between parent and venture is found in the H2 variable - use of financial targets - which is strongly and significantly negative (significant at the 0.001 level). It appears that while economic dependence decreases with increasing venture maturity, accountability increases. The level of economic change across the phases of venture maturity is not significantly different between the high and low performers. The predicted stability of relational ties (P2) is also generally supported although, as is the case for P l , there is one important exception. The individual and combined significance tests for the relational, multi-stage variables for H6, H7, and H8 do not allow us to reject the null hypothesis of no change from the early-to-middle stage. The same is true of the middle-to-late stage transition. We infer from this that the relational ties do not differ across stages. The one relational aspect that does evolve across the stages of venture maturity is H5 - CEO interference and support. In this case, the mean scores of both variables decrease with increasing venture 32 T a b l e 2.5 D y n a m i c E v o l u t i o n o f P a r e n t - V e n t u r e R e l a t i o n s h i p Early to Middle Stage Middle to Established Stage HQ: Diff=0 H 0 : Hi=Lo H 0 : Diff=0 HQ: Hi=Lo Diffg Fh Prob F Prob Diff F Prob F Prob Economic H2 Venture must meet financial targets 0.92 60.60 0.00 0.00 0.95 1.00 36.03 0.00 0.73 0.40 H3 Venture restricts alternate venturing -0.39 9.77 0.00 0.03 0.85 -0.56 7.58 0.01 2.54 0.12 H3 Investments in the V are specialized -0.11 0.76 0.39 0.01 0.92 -0.12 0.45 0.51 7.62 0.01 H3 Funds in the venture are sunk -0.58 17.14 0.00 0.00 0.96 -0.82 17.95 0.00 0.88 0.35 H3 Funds are not diverted from venture -0.16 3.74 0.06 1.35 0.25 0.00 0.00 1.00 0.76 0.39 H3 Importance to V of P's resources -0.47 10.71 0.00 1.98 0.17 -0.56 12.35 0.00 0.76 0.39 Test of combined H3 Diff scores 7.90 0.00 0.74 0.60 7.63 0.00 0.54 0.75 H4 Venture budgeting is more flexible -0.43 13.00 0.00 0.15 0.70 -0.42 11.63 0.00 1.13 0.24 H4 Venture compensation is different 0.11 2.33 0.13 0.05 0.83 0.11 0.42 0.52 1.64 0.21 H4 Venture has larger training budget -0.13 1.95 0.17 0.50 0.48 -0.13 1.52 0.22 0.01 0.91 H4 Venture must use parent's systems 0.11 0.65 0.42 0.80 0.37 0.17 1.00 0.32 0.53 0.47 Test of combined H4 Diff scores 4.50 0.00 0.50 0.74 3.63 0.01 1.23 0.31 Test of combined Economic hypotheses 5.70 0.00 0.86 0.56 5.20 0.00 0.40 0.92 Relational H5 CEO actively supports venture -0.19 6.52 0.01 4.68 0.03 -0.09 0.66 0.42 1.58 0.22 H5 CEO runs interference for venture -0.30 5.55 0.02 0.31 0.58 -0.38 8.52 0.01 2.37 0.13 Test of combined H5 Diff scores 5.16 0.01 2.10 0.13 3.77 0.03 3.10 0.06 H6 V manager works to obtain buy-in 0.12 1.33 0.25 0.26 0.61 0.02 0.03 0.87 0.71 0.41 H6 near top of P's business agenda 0.00 0.00 1.00 0.40 0.53 0.04 0.08 0.78 1.19 0.28 Test of combined H6 Diff scores 0.66 0.52 0.36 0.70 0.04 0.96 1.47 0.24 H7 CP record of meeting commitments -0.10 0.74 0.39 0.26 0.61 -0.10 0.30 0.59 0.13 0.72 H8 Decision making power of V EEs -0.08 0.64 0.43 1.00 0.32 0.02 0.02 0.88 0.05 0.82 H8 Decision making power of V mgr -0.05 0.31 0.58 0.24 0.63 0.06 0.19 0.67 0.00 0.96 H8 Importance of venture culture -0.01 0.04 0.84 0.05 0.82 -0.10 1.09 0.30 0.06 0.81 H8 Venture is protected from politics 0.00 0.00 1.00 0.04 0.85 0.02 0.01 0.93 0.02 0.89 H8 Identify with CV as distinct entity -0.35 4.19 0.05 3.76 0.06 0.26 2.56 0.12 5.15 0.03 H8 CV sense of autonomy from CP 0.02 0.02 0.88 1.06 0.31 0.00 0.00 1.00 1.19 0.28 Test of combined H8 Diff scores 1.23 0.31 1.30 0.28 0.58 0.74 1.16 0.36 Test of combined Relational hypotheses 1.48 0.18 1.16 0.34 1.47 0.22 1.13 0.39 (g) The difference score is obtained by subtracting the earlier period from the later period for a given venture. The number reported in the Diff column of this table is the average of the within venture difference scores for all ventures. Mean difference scores for the high and low performers, respectively, are not reported. (h) The F-test statistic is derived from Hotelling's T^ by the formula F=((N-p-l)/(N-2)*p))*T2. In this equation, N = n) + n2 = the total number of observations; p indicates the total number of variables being evaluated (e.g., p=5 for H3) (Stevens, 1996) maturity. The relational difference scores across the transition phases do not differ significantly when we compare the high and low performing ventures. D I S C U S S I O N The hypotheses developed in this paper address the effects of the economic and relational dimensions of parent-venture strategic fit on venture performance. The eight hypotheses fall into 33 3 general categories: HI, H2, H5, and H6 pertain to characteristics of a specific venture; H3 and H7 address the level of parent-venture commitment; and H4 and H8 speak directly to the level of connection or autonomy that exists between parent and venture. The hypotheses and the results of our empirical investigation are summarized in Table 2.6. Table 2.6 Summary of Hypotheses and Results ECONOMIC DIMENSION RELATIONAL DIMENSION Hypotheses Results Hypotheses Results Characteristics HI: Success assoc. with Supported H5: Success assoc. with Mixed results of the Venture ROR selection criteria CEO protection and support H2: Success assoc. with Not supported H6: Success assoc. with Supported economic performance preeminence in eyes of criteria parent Commitment H3: Success assoc. with large, sunk investments Not supported H7: Success assoc. with commitment and trust Supported Connection H4: Success assoc. with uniform financial practices Partial support H8: Success assoc. with low autonomy Partial support Dynamic Pl: Diminishing economic ties as venture matures Supported P2: Consistently, strong relational ties as V matures Supported Of the four hypotheses related to venture-specific characteristics, two are supported by the data, one gives mixed results, and one - use of economic performance criteria - is not supported. The hypotheses on commitment and connection are fully or partially supported in the relational dimension, yet they receive only partial support or no support when evaluated in economic terms. In fact, sunk funds are significantly associated with low performance, contrary to our expectations. From these findings, we conclude that the degree of fit between a corporate parent and venture does affect the success of a venture, and that success is associated with high levels of awareness, commitment, and connection. Further, the relational dimension of the parent-venture 34 interface appears to have a greater association with venture success than does the economic dimension. Our second research question, regarding the dynamic nature of the parent-venture relationship (Pl and P2) is also addressed in Table 2.4. Support is found for our model in that ventures generally lessened their economic connections with their parents as they mature (or vice-versa) while the relational bonds remain more or less intact. The exceptions to these general trends are an increasing emphasis on financial targets and decreasing CEO involvement as ventures mature. Both of these findings make intuitive sense. Greater financial independence (the defining characteristic of the stages in our model) is accompanied by greater financial accountability. And, as a venture gains in both independence and accountability, there is less need for the CEO to provide "air cover." These two issues aside, the basic model of enduring relational ties and diminishing economic ties is supported most notably in the areas of commitment and connection. As well, the increasing accountability is consistent with our expectation that close connection is preferable to high venture autonomy. Our finding that the degree of change in the parent-venture relationship does not differ between high and low performers may be due to a number of factors. The simplest of these is that there is no difference. However, it may be that our relatively crude measurement instrument, coupled with a fairly small sample size, is not sensitive enough to allow us to detect differences that may be quite subtle. Larger-scale studies of a truly longitudinal design may address this issue with greater power and precision. Another area that may prove to be of value in future research is the possible interaction of the relational and economic aspects of the CP-CV interface. While beyond the scope of this paper, it is an issue that may yield interesting insights into the dynamic processes of corporate venture development. 35 Two aspects of this research that differ from other studies are our use of a multi-stage questionnaire format and our criteria for venture success. With respect to the first issue, the dynamic trends within the parent-venture managerial processes support our use of the multi-stage format. Although our survey is strictly a cross-sectional method of data collection, containing all of the weaknesses of retrospective, self-reported data, the format was well received and allowed respondents to indicate which practices changed over time and whether they had improved or degraded. Future development of this data collection method may yield additional insights into the evolutionary nature of management and decision processes. The issue of how to define "success" is still unresolved (see Miller, Wilson, and Adams, 1988). Any method or measure of venture performance has both strengths and weaknesses and an "ideal" criterion is yet to be defined. The prevalent use of milestones by corporations, and the near unanimous support for this method of performance evaluation in the academic and practitioner press, convinced us that this was a useful way to categorize corporate ventures. CONCLUSION We have taken a dynamic, multi-dimensional approach in our investigation of corporate venturing. We have identified distinct relational and economic dimensions of the parent-venture relationship. Contrary to conventional wisdom, our data indicate that a close fit between a corporate parent and its venture is positively associated with venture performance. Further, our multi-stage survey instrument allows us to confirm that the relationship between parent and venture evolves as the venture matures, and that the nature of the changes are consistent with consistently strong relational ties, low venture autonomy, and decreasing economic connection. While there is still a pressing need for longitudinal studies of corporate ventures, this paper hints at the type of findings that may result as we move from static, cross-sectional research designs to those that can capture the dynamic processes underlying the growth of 36 corporate ventures. Like any organization, corporate ventures have the potential to grow and flourish or contract and wither away. As we improve our ability to identify and measure dynamic elements within organizations, we will improve our ability to understand the dimensions that underlie the processes of venture growth. Given the prevalence of corporate venturing as a growth mechanism, it is important to continue to explore these dimensions, their components, and the way they interact during venture growth and development. 37 REFERENCES Arrow, K. J. (1982). Innovation in large and small firms. New York: Price Institute for Entrepreneurial Studies. Baden-Fuller, C. (1995). Strategic innovation, corporate entrepreneurship and matching outside-in to inside-out approaches to strategy research. British Journal of Management, 6(Special Issue), S3-S16. Barney, J. B., & Hansen, M . H. (1994). Trustworthiness as a source of competitive advantage. Strategic Management Journal, 15(Special Issue). 175-190. Biggadike, R. (1979). The risky business of corporate diversification. Harvard Business Review. May-June. 103-111. Birkinshaw, J. (1997). Entrepreneurship in multinational corporations: the characteristics of subsidiary initiatives. Strategic Management Journal. 18(3). 207-229. Block, Z. (1983). Can corporate venturing succeed? The Journal of Business Strategy. 3(2). 21-33. Block, Z. (1989). Damage control for new corporate ventures. Journal of Business Strategy, 10(2), 22-28. Block, Z., & MacMillan, I. C. (1993). Corporate venturing: Creating new businessses within the firm. Boston: Harvard Business School Press. Block, Z., & Ornati, O. A. (1987). Compensating corporate venture managers. Journal of Business Venturing, 2(11. 41-51. Burgelman, R. A. (1983). A process model of internal corporate venturing in a diversified major firm. Administrative Science Quarterly. 28(1), 223-244. Caruana, A., Morris, M . H., & Vella, A. J. (1998). The effect of centralization and formalization on entrepreneurship in export firms. Journal of Small Business Management. 36(1), 16-29. Chung, L. H. , & Gibbons, P. T. (1997). Corporate entrepreneurship: The roles of ideology and social capital. Group & Organization Management, 22(1). 10-30. Cooper, A. C. (1979). Strategic management: New ventures and small business. In D. E. Schendel, & C. W. Hofer (Eds.), Strategic Management: A New View of Business Policy and Planning (pp. 316-327). Boston: Little, Brown, and Company. Covin, J. G., & Slevin, D. P. (1989). Strategic management of small firms in hostile and benign environments. Strategic Management Journal. 10(1). 75-87. 38 Covin, J. G., & Slevin, D. P. (1994). Corporate entrepreneurship in high and low technology industries: a comparison of strategic variables, strategy patterns and performance in global markets. Journal of Euromarketing, 3(3), 99-127. Day, D. L. (1994). Raising radicals: different processes for championing innovative corporate ventures. Organization Science, 5(2), 148-172. Dess, G. G., & Robinson, R. B. (1984). Measuring organizational performance in the absence of objective measures: the case of privately held firms and conglomerate business units. Strategic Management Journal, 5(2), 265-73. Dougherty, D. (1995). Managing your core incompetencies for corporate venturing. Entrepreneurship: Theory & Practice, 19(3), 113-135. Fiorito, S. S., & LaForge, R. W. (1986). A marketing strategy analysis for small retailers. American Journal of Small Business, 10(4), 7-17. Frost, P. J., & Egri, C. P. (1990). Influence of Political Action on Innovation: Part I. Leadership & Organization Development Journal, 11(1), 17-25. Garud, R., & Van de Ven, A. H. (1992). An empirical evaluation of the internal corporate venturing process. Strategic Management Journal, 13(Special Issue), 93-109. Ghemawat, P. (1991). Commitment: the dynamics of strategy. New York: Free Press. Ginsberg, A., & Hay, M. (1994). Confronting the challenges of corporate entrepreneurship: guidelines for venture managers. European Management Journal, 12(2), 382-389. Glaser, G.B. (1978). Theoretical Sensitivity: Advances in the Methodology of Grounded Theory. Mill Valley, CA: Sociology Press. Glaser, G. B., & Strauss, A. L. (1967). The discovery of grounded theory: strategies for qualitative research. Chicago: Aldine. Guth, W. D., & Ginsberg, A. (1990). Guest editors' introduction: corporate entrepreneurship. Strategic Management Journal, 11 (Special Issue), 5-15. Hornsby, J. S., Naffziger, D. W., Kuratko, D. F., & Montagno, R. V. (1993). An interactive model of the corporate entrepreneurship process. Entrepreneurship: Theory & Practice, 17(2), 29-37. Jennings, D. F., & Lumpkin, J. R. (1989). Functioning modeling corporate entrepreneurship: an empirical integrative analysis. Journal of Management 15(3), 485-502. Kanter, R. M. (1985). Supporting innovation and venture development in established companies. Journal of Business Venturing, 1(1), 47-60. Knight, R. M. (1989). Technological innovation in Canada: a comparison of independent entrepreneurs and corporate innovators. Journal of Business Venturing, 4(4), 281-288. 39 Lengnick-Hall, C. A. (1992). Strategic configurations and designs for corporate entrepreneurship: exploring the relationship between cohesiveness and performance. Journal of Engineering & Technology Management, 9(2), 127-154. MacMillan, I. C , Block, Z., & Narasimha, P. N. S. (1986). Corporate venturing: alternatives, obstacles encountered, and experience effects. Journal of Business Venturing. 1 (2). 177-191. McGrath, R. G. (1995). Advantage from adversity: learning from disappointment in internal corporate ventures. Journal of Business Venturing, 10(2), 121-142. McGrath, R. G., Venkataraman, S., & MacMillan, I. C. (1994). The advantage chain: antecedents to rents from internal corporate ventures. Journal of Business Venturing, 9(5), 351-369. Miller, A., & Camp, B. (1985). Exploring determinants of success in corporate ventures. Journal of Business Venturing. 1(1), 87-105. Miller, A., Gartner, W. B., & Wilson, R. (1989). Entry order, market share, and competitive advantage: a study of their relationships in new corporate ventures. Journal of Business Venturing, 4(3). 197-209. Miller, A., Spann, M . S., & Lerner, L. (1991). Competitive advantages in new corporate ventures: the impact of resource sharing and reporting level. Journal of Business Venturing. 6(5). 335-350. Miller, A., Wilson, B., & Adams, M . (1988). Financial performance patterns of new corporate ventures: an alternative to traditional measures. Journal of Business Venturing. 3(4). 287-300. Miller, D., & Toulouse, J. M . (1986). Strategy, structure, CEO personality and performance in small firms. American Journal of Small Business, 10(3), 47-62. Naman, J. L. , & Slevin, D. P. (1993). Entrepreneurship and the concept of fit: a model and empirical tests. Strategic Management Journal, 14(2), 137-153. Pearce, J. A. I., Kramer, T. R., & Robbins, D. K. (1997). Effects of managers' entrepreneurial behavior on subordinates. Journal of Business Venturing, 12(2), 147-160. Penrose, E. (1959). The Theory of the Growth of the Firm . London: Basil Blackwell. Sandberg, W. R., & Hofer, C. W. (1987). Improving new venture performance: The role of strategy, industry structure, and the entrepreneur. Journal of Business Venturing, 2(1), 5-28. Sapienza, H. J., Smith, K. G., & Gannon, M . J. (1988). Using subjective evaluations of organizational performance in small business research. American Journal of Small Business, 12(3). 45-53. 40 Schrader, R. C , & Simon, M . (1997). Corporate versus independent new ventures: Resource, strategy, and performance differences. Journal of Business Venturing, 12(1), 47-66. Sorrentino, M . , & Williams, M . L. (1995). Relatedness and corporate venturing: does it really matter? Journal of Business Venturing. 10(1). 59-73. Stevens, J. (1996). Applied Multivariate Statistics for the Social Sciences. Mahwah, NJ: Lawrence Earlbaum Associates. Sykes, H. B. (1986). Lessons from a new ventures program. Harvard Business Review. 64(3). 69-74. Sykes, H. B., & Block, Z. (1989). Corporate venturing obstacles: sources and solutions. Journal of Business Venturing, 4(3). 159-167. Thornhill, S., Amit, R., & Belcourt, M . (1998). Determinants of corporate venture performance: An empirical study of firm and industry effects. Proceedings of the Administrative Sciences Association of Canada Annual Conference. 19(21). 80-90. Tsai, W. M.-H. , MacMillan, I. C , & Low, M . B. (1991). Effects of strategy and environment on corporate venture success in industrial markets. Journal of Business Venturing. 6(1), 9-28. Tushman, M . , & Nadler, D. (1986). Organizing for innovation. California Management Review. 3(1), 73-84. Venkataraman, S., MacMillan, I. C. & McGrath, R. G. (1992). Progress in research on corporate venturing. In D. Sexton and J. Kasarda (Eds.) State of the Art in Entrepreneurship Research (pp. 487-519). Boston: PWS-Kent Publishing Co. Venkatraman, N. (1990). Performance implications of strategic coalignment: A methodological perspective. Journal of Management Studies. 27(1). 19-41. Venkatraman, N., & Ramanujam, V. (1987). Measurement of business economic performance: an examination of method convergence. Journal of Management. 13(1). 109-22. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal. 5(2). 171-180. Zahra, S. A. (1991). Predictors and financial outcomes of corporate entrepreneurship: an exploratory study. Journal of Business Venturing. 6(4). 259-285. Zahra, S. A. (1993). Environment, corporate entrepreneurship, and financial performance: a taxonomic approach. Journal of Business Venturing. 8(4). 319-340. Zahra, S. A. (1995). Corporate entrepreneurship and financial performance: the case of management leveraged buyouts. Journal of Business Venturing. 10(3). 225-247. 41 Zahra, S. A. (1996a). Governance, ownership, and corporate entrepreneurship: the moderating impact of industry technological opportunities. Academy of Management Journal. 39(6). 1713-1735. Zahra, S. A. (1996b). Technology strategy and new venture performance: A study of corporate-sponsored and independent biotechnology ventures. Journal of Business Venturing. 11(4), 289-321. Zahra, S. A. & Covin, J. G. (1995). Contextual influences on the corporate entrepreneurship-performance relationship: a longitudinal analysis. Journal of Business Venturing. 10(1). 43-58. 42 APPENDIX TABLE A2.1 Summary of Corporate Venturing Research Author(s) Research Question(s) Finding(s) Baden-Fuller (1995) Proposition: competitive advantage flow from the capacity to manage internal change, a capacity which is closely connected to corporate entrepreneurship Strategic innovations are not necessarily profitable. Corporate entrepreneurship is not the only way to stimulate innovation in established firms. Biggadike (1979) What are the performance implications of corporate venturing? Of the ventures studied, 18% achieved profitability in two years, 38% in four years. Median performance was 7% ROI in years seven and eight. Companies should start fewer ventures with more resources rather than many with less. The time it takes to reach profitability can be reduced by spending more earlier to obtain market share. Birkinshaw (1997) An exploratory study on initiatives in MNC subsidiaries Two distinct entrepreneurial processes: (1) internal -initiatives subject to corporate selection mechanisms, and (2) external - initiatives subject to local environmental selection mechanisms (e.g., customer acceptance) Block (1983) How can corporate ventures succeed? There are five pivotal decisions which can impact venture success: format, management and compensation, venture plan approval, positioning, and financing triggers. Block (1989) How can firms reduce the costs of corporate venture failure? Two principal causes of large losses: (1) incorrect assumption and (2) pressures within the parent which inhibit altering or aborting venture strategies. Block and Ornati. (1987) What compensation practices are in use among corporate ventures and do they impact performance? "... the incentives and compensation used are not correlated with the rate of success ... (but) the compensation systems used weren't much of an incentive" (p. 44) Burgelman (1983) What are the processes by which a large diversified firm transforms new technology into new businesses through internal corporate venturing? "Because corporate entrepreneurship ... seems to differ from traditional individual entrepreneurship, as well as from traditional organizational economic activity, it may be necessary to devise different arrangements between the corporate resource providers and their entrepreneurial agents" (p. 243) Caruana, Morris and Vella(1998) How do centralization and formalization impact entrepreneurial behavior in export firms? "The results suggest that entrepreneurial behavior is negatively affected by increases in centralization and size but positively influenced by increased formalization" (p. 24) Chung and Gibbons (1997) What is the role of culture in corporate entrepreneurship? 1. As superstructure, culture provides an ideology for organizational members 2. As sociostructure, culture provides social capital which in turn enables the emergence of competitive advantage. Covin and Slevin (1994) What is the impact of industry technological sophistication and the strategy-related characteristics of entrepreneurial firms? Industry technological sophistication moderates the strength of the relationships between many strategic and firm performance. The modes of competition differ for entrepreneurial firms in high and low technology industries. Day (1994) How does the championing process explain innovativeness in corporate venturing? Ventures requiring modest resources may survive with low level champions and have the potential to be radically innovative. Top level champions are required for high-profile, expensive, innovative ventures. 43 TABLE A2.1 (continued) Author(s) Research Question(s) Finding(s) Dougherty (1995) How can managers assure an effective connection between ventures and the firm's core competencies? Connecting core competencies with new products is problematic because of the existence and persistence of rigid core incompetencies with which the competencies have little interplay. Incompetencies must be managed so that competencies can be accessed. Garud, and Van de Ven(1992) What guides the development of a venture under conditions of uncertainty and ambiguity? Three different periods: agenda setting, expansion, and contraction. Negative outcomes lead to greater involvement by corporate sponsors. Hornsby, Naffziger, Kuratko and Montagno (1993) Introduction of a model that describes the intrapreneurship process from initial decision to implementation "Intrapreneurship is multidimensional and relies on the successful interaction of several activities rather than events occurring in isolation" (p. 35) Jennings and Lumpkin (1989) What differences exist between conservative and entrepreneurial firms? In entrepreneurial organizations: decision making is more participative, performance objectives are jointly determined, and managers are not penalized for the failure of risky projects. Knight (1989) Examination and comparison of innovative practices in large and small firms CV obstacles includes: lack of entrepreneurial talent, lack of CV fit with corporate strategy, and lack of commitment to the venture. IVs encountered problems with market assessment and operational and financial issues. Lengnick-Hall (1992) Can configurations for strategy and design of CVs be defined and tested and does the level of cohesion between strategy and CV design affect market performance and profitability? Firms that are highly cohesive experience fewer obstacles and problems. Cohesiveness may be more advantageous for lean, value driven firms than for rigid firms requiring more resource diversity and organizational slack. MacMillan,, Block, and Narasimha (1986) What obstacle face firms wishing to launch CVs? Firms should attempt to acquire experience by starting with joint ventures or small ventures. McGrath (1995) What are the management opportunities inherent in CV disappointments? Redirection is an important alternative to shutting down unsuccessful CVs. Linkages between a CV and other SBUs can facilitate the benefit of knowledge transfer if a CV fails. McGrath, Venkataraman, and MacMillan (1994) Development of a framework which combines extant CV theory with the RBV to depict CV potential for yielding future rents. Ventures evolve through a series of developmental stages. CV progress should be assessed in terms of the development of new competencies and their subsequent exploitation. Miller and Camp (1985) What differences exist between the strategies of successful and unsuccessful corporate ventures? Successful strategies differ between young ventures and matures SBUs (e.g., low cost strategy is better for mature SBUs while differentiation is more suitable for adolescent CVs) Miller, Gartner, and Wilson (1989) What is the effect of entry order on market share and competitive advantage among corporate ventures? "Overall, pioneers typically gain significantly greater market share and some types of competitive advantage compared to followers (specifically, product and technological advantages) " (p. 203) 44 TABLE A2.1 (continued) Author(s) Research Question(s) Finding(s) Miller, Spann, and Lerner (1991) What is the impact of resource sharing, reporting level, and the relationship between sharing and reporting level? Resource sharing benefits quality advantages but hurts cost advantages. No main effects for reporting level. Reporting level moderated quality and cost advantages of resource sharing (highest performance with lowest reporting levels). Miller, Wilson and Adams (1988) What is an appropriate measure for evaluating corporate venture success? Velocity (V = product of regression coefficient of ROI over time (beta) and coefficient of determination (r-sq.) fit the observed data in 78% of the cases. Pearce, Kramer and Robbins (1997) What managerial behaviors indicate an entrepreneurial orientation and how are such behaviors evaluated by subordinates? "Overall, this study developed and validated a scale of entrepreneurial behaviors and found that corporate entrepreneurship was well received by subordinates even when such behaviors were counter to the preexisting culture" (p. 158) Schrader and Simon (1997) 1) Do I Vs and CVs emphasize different resources? 2) Do IVs and CVs pursue different strategies? 3) Do IVs and CVs differ in performance? 4) Do IVs and CVs differ in the relationship between resources, strategies, and performance? 1) IVs placed importance on external capital and brand development; CVs developed proprietary knowledge and marketing. 2) IVs concentrated on customer service & specialty products & had greater strategic breadth. 3) No performance differences 4) Broad strategies improved IV performance and degraded CV performance. Sorrentino and Williams (1995) Does relatedness determine CV success? "The degree to which a venture is related to its parent firm does not explain performance results or the entry strategy decisions at the venture level" (p.70) [Performance is evaluated by market share) Sykes (1986) What factors determine corporate venture success?" "The data provide strong evidence that venture managers' prior experience in the venture's target market area and their general managerial experience are the factors most important to venture financial success" (290) Thornhill, Amit and Belcourt (1998) What are the influences of venture strategies, venture capabilities, and environmental hostility on corporate venture performance? Corporate venture profitability is positively associated with venture age and negatively associated with environmental hostility. Differentiation appears to be an appropriate strategy in benign environments, while aggressive marketing and innovation are called for in hostile competitive conditions. Tsai, MacMillan and Low (1991) What is the relative importance of environment and strategy for CV performance? Both environment and strategy are important to CV success. Munificent markets are good; hostile markets are not. Quality strategy improves market share at the expense of ROI. Low prices and high promotion improve both market share and ROI. Zahra (1991) (1) What are the antecedents of corporate entrepreneurship? (2) What is the association between CV and company performance? (1) CV activities increased as environments were perceived as increasingly dynamic, hostile, and heterogeneous. (2) There was a positive relationship between CV and company performance. Zahra (1993) What is the relationship between external environment, corporate entrepreneurship and financial performance? CV activities differ in different environments. CV is associated with company financial performance. Zahra (1995) What is the impact of LBOs on corporate entrepreneurship? Changes in corporate entrepreneurship activities after LBOs were positively associated with changes in company performance. 45 T A B L E A2.1 (continued) Au thor (s ) Research Question(s) Finding(s) Zahra (1996a) What is the relationship between corporate ownership, industry, and corporate entrepreneurship? Executive stock ownership and long-term institutional ownership are positively associated CV. Conversely, short-term institutional ownership is negatively associated with it, as is a high ratio of outside directors on a company's board. Finally, an industry's technological opportunities moderate the associations observed between corporate governance and ownership variables and corporate entrepreneurship. Zahra (1996b) (1) Do CVs and IVs vary in their technology strategies? (2) Do the dimensions of technology strategy influence the performance of CVs and IVs differently? Both research questions can be answered in the affirmative: "The results suggest that the two venture types follow different paths to achieve success. The high performing IV focuses its R&D spending on pioneering a few new products. Conversely, the CV benefits from investing more in R&D to develop many products and by protecting these products with patents ... the CV uses both internal and external R&D sources" (p. 310) Zahra and Covin (1995) What is the CV - performance relationship in different industry contexts? CV is positively associated with firm performance and the strength of the relationship increases over time. Hostile environments reward CV activities more than do benign environments. 46 Table A2.2 Sample Items from Corporate Venturing Survey Many of the following questions ask for a response in three time periods. Please indicate, for your venture, approximately when each stage as defined below occurred. If your venture has not reached the Established Stage, for example, respond only for the stages appropriate to your venture. Early Stage - First financial investment in the venture: 19 Middle Stage - The venture has begun to generate revenue: 19 Established Stage - The venture has become profitable: 19 28. Venture employees are empowered with greater decision making authority than are their counterparts in the parent company: Strongly 7 O 7 O 7 O Agree 6 O 6 O 6 O 5 O 5 O 5 O 4 O 4 O 4 O 3 O 3 O 3 O Strongly 2 O 2 O 2 O Disagree 1 O 1 O 1 O Early Middle Estab. Stage Stage Stage 54. The budget for training venture personnel is proportionally greater than the training budget of the parent company: Strongly 7 O 7 O 7 O Agree 6 O 6 O 6 O 5 O 5 O 5 O 4 O 4 O 4 O 3 O 3 O 3 O Strongly 2 O 2 O 2 O Disagree 1 O 1 O 1 O Early Middle Estab. Stage Stage Stage 47 Table A2.3 Alpha Scores and Eigenvalues for Composite Variables Alpha Eigenvalue Eigenvalue Eigenvalue Coefficient (Early) (Middle) (Established) Economic H3: Large, specialized investments 0.74 1.41 1.44 1.16 H4: Uniform financial practices 0.85 1.60 1.59 1.86 Relational H5: CEO protection and support 0.80 1.27 1.09 1.09 H6: Venture preeminence 0.85 1.27 1.30 1.43 H8: Low venture autonomy 0.91 2.69 2.88 2.94 48 CHAPTER 3: W H Y DO YOUNG FIRMS FAIL? MANAGERIAL CAPABILITIES, ORGANIZATIONAL ASSETS, AND THE L IAB IL ITY OF NEWNESS ABSTRACT This paper models the growth and decline of young firms as a function of their initial asset stocks, initial capabilities, rate of capability development, rate of asset depletion, and failure threshold. Data from 246 Canadian corporate bankruptcies confirm that young firms fail due to insufficient organizational capital at start-up and inadequacies in managerial knowledge, financial management skills, and marketing abilities. Older firms, on the other hand, are more prone to failure due to environmental change. INTRODUCTION There are two reasons why the study of business failures is important. First, significantly more companies fail than survive. In Canada, only one new firm in five survives for ten years or more. And, of the 80% of new ventures that cease operations in the first decade, half fail within the first two years. Each year, many firms with lengthy records of success cease to exist. The magnitude of the turnover of firms is significant: in a given year 15-20% of firms in the Canadian economy are less than two years old. Also worth noting is the rise in the rate of business bankruptcies: in 1980, 10 firms per 1000 in Canada entered bankruptcy proceedings; in 1995, the rate was 14 firms per 1000. Yet most business research considers only the minority of firms that stays in business (in many cases, this focus is due to data availability). There is a paucity of research that considers what happens when firms die and why they cease to exist. The present study begins to fill this gap by developing a theoretical model of young firm survival and growth and testing the resulting hypothesis on a set of firms (of varying ages) that went bankrupt. 49 Second, the scant research into business failures undertaken so far has been within the distinct fields of strategy and organizational ecology, and their theoretical and methodological approaches have exhibited little convergence. Recent exceptions include Gimeno, Folta, Cooper, & Woo's (1997) threshold model of firm survival and Pennings, Lee, & van Witteloostujin's (1998) examination of the role of human and social capital in firm dissolution. The present paper seeks to bridge these academic disciplines and to further our understanding of how firm-level processes relate to population-level dynamics. We focus on the "liability of newness," that is, on the well-established inverse relationship between firm age and mortality risk. We are particularly interested in the phenomenon known as the "liability of adolescence" where firms exhibit low mortality rates in their early years followed by an increasing rate, after which the risk declines as the firm ages (Bruderl & Schussler, 1990; Fichman & Levinthal, 1991). Why firms fail early in life has long been a question of interest among organizational ecology scholars. Management strategists, on the other hand, have tended to concentrate their organizational-level studies on the general questions of survival and failure at any age. What is missing is a firm-level examination of why young firms in particular are so prone to failure. Drawing on the resource-based view of the firm, we propose that firm survival during the formative years is a function of five factors: (1) initial asset stocks, (2) initial capabilities, (3) rate of capability development, (4) rate of asset depletion, and (5) failure threshold. Later in a firm's life, we concur with Aldrich & Auster (1986) that inertia in the face of external change may be a significant contributor to firm failure. Given that so many firms fail early in their life, we believe there is value in trying to understand this phenomenon and in identifying how new firms can position themselves to beat the odds against their survival. In the following section, we review two branches of literature. The first focuses on the liabilities of newness and adolescence. The second examines the general phenomenon of firm failure, its definition, and its causes. We then propose a simple model of firm survival from which we generate hypotheses regarding the liabilities of newness versus the liabilities of age and inertia. These hypotheses are then tested using data on 246 Canadian corporate bankruptcies. The data generally support our hypotheses that young firms fail due to insufficient organizational capital at start-up and inadequate managerial capabilities. Older firms, on the other hand, fail due to environmental change. L I T E R A T U R E R E V I E W Th e Liability of Newness In his seminal paper, Stinchcombe (1965) articulated the theoretical foundations of the liability of newness. He identified four aspects of new organizations that make them more prone to early failure than are older, more established organizations: (a) new organizations must get by with general knowledge until members learn new, specific roles and functions, (b) during the role identification and formation process, there may be conflict, worry, and inefficiency, (c) relations with outside individuals and organizations must be forged, and an initial lack of trust may prove a liability, and (d) new organizations lack stable ties with the customers they wish to serve. Jovanovic (1982) proposed a theory of "noisy" selection to account for empirically observed deviations from the proportional growth law. He argued that a firm's level of efficiency accounts for its survival or demise. Since the slowly growing small firms are precisely those that are selected out of the population, the large firms that remain will exhibit higher rates of growth; it is through their efficiency-driven high growth rates that they became large survivors in the first place. Carroll (1983) and Freeman, Carroll, & Hannan (1983) both provided empirical support for the liability of newness. Carroll (1983) examined 52 data sets of organizational death event histories, including 23 in retail and 20 in manufacturing. From this analysis, he concluded that the data supported Stinchcombe's (1965) liability of newness hypothesis and that Makeham's Law was an appropriate way to model organizational mortality. Freeman et al. (1983) extended the scope of the empirical work with their investigation of US labor unions, US local newspapers, and US semi-conductor producers. They confirmed the liability of newness effect and were able to disentangle the effects of aging from those of initial size. They also determined that the strength of the age-dependent relationship depends on whether organizations exit the population through dissolution or merger. More recent research has refined the age dependency model to account for the existence of organizational adolescence. Bruderl & Schussler (1990) examined more than 150,000 German business foundings and determined that firms exhibit a period of adolescence in their mortality patterns. While Schussler (1988) described this phenomenon as liability of newness with delayed onset, Fichman & Levinthal (1991) offered the alternative proposition that the liability of newness is not a monotonically decreasing function of firm age, but that there is an initial "honeymoon" period during which initial assets buffer the new organization. They argue that variation in the levels of initial assets affects the way in which time affects mortality rates. The time dependence occurs because the longer an organization survives (due to initial capital endowments), the more it will be able to develop relationship-specific capital and adapt to the environment. 52 The honeymoon period, noted in the German and US data, is also present in the Canadian economy. A plot of hazard rates for young Canadian firms (shown in Figure 3.1), extracted from the Longitudinal Employment Analysis Program (LEAP) 2, exhibits the now-familiar pattern of organizational adolescence. (Data values for Figure 3.1 may be found in the Appendix, Table A3.1.) Figure 3.1 H a z a r d rates for Canadian firms 0.30 Goods Svces | 0.05 0 F i r m A g e 10 Source: Baldwin, Dupuy & Gelatly (1998) Figure 3.1 illustrates hazard rates by goods industries and service industries. The hazard rate peaks and then declines for each case. For manufacturing firms, the peak occurs in year one. In the service industries, the hazard rate increases to a maximum value in year two and declines thereafter. After roughly the sixth year, the steady decline in hazard rates gives way to a fairly constant mortality hazard in the 9% to 13% range.3 2 For a detailed description of the LEAP database, see Brander, Hendricks, Amit, & Whistler (1998). 3 The estimates of hazard rates for years 8-11 should be viewed with caution due to the increasing influence of cohort effects in these years. The year 11 data represents only those firms born in 1983; the year 10 data is comprised of the 1983 and 1984 cohorts, and so forth. Thus, lingering business cycle effects from the cohort birth 53 In another interpretation of the observed mortality patterns of young organizations, Levinthal (1991) developed a random walk model comprised of the initial stock of organizational capital and the process governing the change in organizational capital over time. He found that the model generates not only the commonly observed negative duration dependence relationship, but also a honeymoon period prior to a liability of adolescence. Levinthal argues that, according to the random walk model, there is no direct link between an organization's age and its reduced mortality risk. Older firms experience reduced mortality rates not because they are old but because their prior successes buffer them from current pressures. Thus, organizational changes such as increased reliability and competence are not necessary to explain the declining mortality risk that accompanies age in most empirical studies. Levinthal concluded that age may be a proxy for the broader construct of organizational capital. Contrasts between Organizational Ecology and the R B V Definition of Failure While both organizational ecology and the resource-based view (RBV) of the firm have examined firm failure, their perspectives differ in the level of analysis used: the ecologists observe entire populations, while RBV researchers tend to examine individual firms. As a result of this disparity, each school defines failure differently. Our review of 42 empirical papers on firm failure revealed that population-level studies use firm discontinuance as the operational indicator of firm failure (see Appendix, Table A3.2). Firms are thus either members of a population or they are not; the nature of the exit is seldom addressed at the population level of analysis. year contribute to the erratic pattern of hazard rates evident in the right hand side of Fig. 1 where the data is relatively scarce and representative of only a few birth year cohorts. 54 Firm-level studies, on the other hand, vary in the way they operationalize firm failure. Discontinuance is the most popular definition, but bankruptcy is also frequently used. Another popular definition is that used by Dun & Bradstreet: "those businesses that cease operations following assignment or bankruptcy; ceased with loss to creditors after such actions as execution, foreclosure, or attachment, voluntarily withdrew leaving unpaid obligations; were involved in court actions such as receivership, reorganization or arrangement, or voluntarily compromised with creditors (D & B Annual, 15)" (Fredland & Morris, 1976: 7). Berryman (1983) highlighted four common definitions of failure: (1) significant and continually low returns on capital (Altman, 1971), (2) the Dun & Bradstreet criteria listed above, (3) bankruptcy or insolvency, and (4) liquidation to avoid losses (Ulmer & Nielson, 1947). Cochrane (1981) depicted failure as a series of nested conditions. The most general definition is discontinuance. Then, in increasing order of specificity and decreasing order of subset size are failures as opportunity costs, termination with losses or to avoid losses, and, finally, bankruptcy. Performance thresholds are also important to consider in the context of business failures. Gimeno et al. (1997) established that a significant factor in the continuance-discontinuance decision for many entrepreneurs is their own acceptable threshold of performance. Firms that appear to be underperformers may persist if their thresholds are sufficiently low while other, relatively superior performers may exit if their thresholds are sufficiently high. The mechanisms by which firms leave the roster of active enterprises include voluntary retirement by proprietors, merger and acquisition, declaring bankruptcy, and orderly cessation of operations for various reasons. Our empirical analysis, described below, investigates only bankrupt firms. This restricted set of business exits excludes other types of firm exits (i.e., discontinuances) that are typically captured in organizational ecology studies. However, it does capture failure in the extreme, and ensures that we do not confound orderly discontinuance or 55 exit due to high performance thresholds with true failure. Firms that are insolvent to the point of legal proceedings have clearly failed to meet the market's threshold of meeting financial obligations. Causes The level-of-analysis difference between the ecological and resource-based perspectives has also led researchers in each school to identify different causes of failure. Population ecologists have been constrained to studying characteristics that can be identified for entire populations of organizations. Firm age and size are thus among the most commonly studied phenomena, and the results have generally been consistent across populations (e.g., Bates & Nucci, 1989; Carroll, 1983). Simply put, firm mortality risks increase if firms are young and small. Freeman et al. (1983) disentangled these effects, and concluded that age effects were not confused or confounded with historical processes. Amburgey, Kelly, & Barnett (1993) studied the effects of organizational change on mortality using a population of Finnish newspapers. The authors concluded that older organizations may be more affected by change than are younger organizations. However, the resource bases of more established organizations better position them to withstand the stresses of change. Other findings within the population-level perspective include a positive association between owners' education levels and firm survival (Bates, 1990) and a concave relationship between survival and owner age (Bates, 1990; Preisendorfer & Voss, 1990). At the firm level of analysis, the research designs have been more varied, as have the findings. Methodologies employed include archival studies, large-scale surveys, and in-depth case studies. In our review, failure at the level of the firm has been attributed to: (1) managerial deficiencies (Fredland & Morris, 1976; Gaskill, Van Auken, & Manning, 1993; Larson & Clute, 1979; Litvak & Maule, 1980; McKinlay, 1979); (2) undercapitalization (Cooper, Gimeno-56 Gascon, & Woo, 1994; Fichman & Levinthal, 1991; Hall, 1992; Larson & Clute, 1979); (3) inadequate business/financial planning (Gaskill et al., 1993; Lussier, 1995); (4) environmental conditions (Gaskill et al., 1993; Hall, 1992); and (5) the pursuit of growth (Boardman, Bartley, & Ratliff, 1981; Gaskill et al., 1993). Clearly, there are many factors that inhibit firm survival. Firms that are young and small, that lack financial capital and human capital, that exist in hostile competitive environments, and that pursue growth too zealously are at risk of joining the large majority of firms that fail to exist for even a single decade. In summary, convincing empirical evidence for the liability of adolescence abounds. However, while macro-level factors such as firm size and industry characteristics have been identified as contributors to the phenomenon of young firm failure, only a limited number of such firm-level effects have been identified (Pennings et al., 1998). We now turn to the resource-based view of the firm to try to understand why young firms are so prone to failure and how the factors enumerated above differentially affect firms of different ages. We begin with a simple model of young firm survival and growth and then develop a number of hypotheses for empirical analysis. THEORETICAL DEVELOPMENT Organizational Mortality and the Resource-Based View The resource-based view of the firm depicts firms as heterogeneous bundles of idiosyncratic, hard-to-imitate resources and capabilities (e.g., Barney, 1991; Conner, 1991; Rumelt, 1984, 1991; Wernerfelt, 1984). Amit & Schoemaker (1993) posit that a firm's ability to capture economic rents is a function of how such capabilities can be used to deploy and utilize a firm's resources. Resources in their view are "stocks of available factors that are owned or 57 controlled by the firm" (1993: 35). Capabilities, on the other hand, are "information-based, tangible or intangible processes that are firm-specific and are developed over time through complex interactions among the firm's resources" (1993: 35). Teece, Pisano, & Shuen (1997) add a dynamic element to this view by arguing that "the competitive advantage of firms lies with its managerial and organizational processes, shaped by its (specific) asset position, and the path available to it. By managerial and organizational processes, we refer to the way things are done in the firm, or what might be referred to as its routines, or patterns of current practice and learning" (1997: 518). Levinthal (1991), in his development of the random-walk model, observed that firms fail when they can no longer meet their financial obligations. Consequently, failure can be averted in the presence of either strong current performance or substantial asset stocks. In this context, assets are not only financial, but also include market position, distribution systems, manufacturing infrastructure, and technological capabilities. Levinthal refers to this conglomeration of financial and non-financial assets as "organizational capital." In essence, this construct corresponds to Amit & Schoemaker's (1993) "firm resources." A Model of Young Firm Survival and Growth. The model of firm survival and growth advanced in this paper incorporates the concept of organizational capital (Lenvinthal, 1991) and the role of initial endowments (Fichman & Levinthal, 1991) along with the concepts of resources and capabilities (Amit & Schoemaker, 1993), managerial and organizational processes (Teece et al., 1997) and organizational thresholds of performance (Gimeno et al., 1997). A firm has an initial asset stock A 0 when it enters an industry. This asset stock consists of not only financial and physical capital, but also the human capital and other intangible elements of organizational processes and procedures. Over time, net increases in asset stocks indicate firm 58 growth, while decreases represent firm decline. The change in assets from one period to the next is a function of the ability of the entrepreneur to learn routines, environmental contingencies, supplier relations, customer preferences, and the myriad other elements that enable economic rents to be extracted from an industry. Specifically, we make the following assumptions: (1) Assets in period t [A(t)J decrease as a constant function (D, where 0<D<1) of asset level in period t-1. This is comparable to capturing fixed costs to scale. Such an assumption is clearly simplistic, but it reflects the unrelenting costs of staying in business. (2) Assets in period t increase as a function of organizational capabilities [C(t)] and assets in period t-1. (3) Capabilities change over time. Thus, the level of organizational capabilities in a given period is a function of the rate of capability development (learning). (4) Organizations cease to exist when the asset level falls below a given threshold of survival. For firms with high thresholds, discontinuance may be voluntary and early. For others, creditors may force closure through bankruptcy proceedings. A simple version of this model may be expressed as: A t = A , , * [ l - D + C(t)] [3.1] For a given initial asset level, firms in this model with depletion rates that are significantly larger than the rates of replacement [through capabilities, C(t)] will be discontinued very quickly. Such firms succumb to the liability of newness because initial asset stocks are inadequate to sustain them while learning takes place. On the other hand, firms with higher rates of learning are able to use capability development to compensate for the depletion of assets, and are able to turn their firms into growing enterprises after an initial period of decline. In exceptional cases, when the capability term is initially greater than the depletion term, growth will be immediate and dramatic. Slater & Narver emphasized the importance of such learning: 59 "the ability to learn faster than competitors may be the only source of sustainable competitive advantage" (1995: 63). This model is consistent with Fichman and Levinthal's (1991) proposition that initial asset levels will affect the duration of organizational adolescence and the subsequent nature of the time dependent hazard rate. For given levels of depletion and capability acquisition, higher initial endowments give organizations more time for the learning curve to offset the depletion rate. While we have set the depletion rate as a constant function of prior asset levels, the idea of constant learning, and thus the potential for constantly increasing growth, is more problematic. A more reasonable representation would capture the "learning curve" where there is a period of rapid initial learning, followed by a plateau at which additional gains become rare and difficult (Levitt & March, 1988). The capability function is better represented as C(t) = G/[ l+L*Exp(- t ) ] [3.2] where G (with 0<G<1, although G>1 is possible) indicates growth to a maximum asymptote and L (with 0< L < oo) is the coefficient of learning. While the log-linear depiction of learning curve effects is relatively common, our emphasis on time and learning coefficients is not. A more conventional approach represents learning with a cost-based function of production such as C n = C,n- a [3.3] where C is direct labor cost of the n t h unit and a is a constant (Levitt & March, 1988). However, we feel that the cost-based production function is too narrow and restrictive to accommodate the broad construct of organizational assets featured here. A production-based learning function also neglects the current composition of the Canadian economy, which is dominated by service firms. Equation 3.2 enables us to capture non-production-specific learning such as that involved in 60 building relationships with customers and suppliers, while still retaining the log-linear form of the traditional learning curve. The denominator of equation 3.2 asymptotically approaches one, at which point the capability factor approaches the constant value, G, and the overall growth rate of the firm equals the difference between the maximum capability level (C m a x = G) and the rate of depletion, D. The initial capability level is jointly determined by G and the learning parameter L, which also represents the rate at which C approaches its maximum value G. The L term is thus the steepness of the learning curve, while G represents the height of the plateau. A small value for L in this formulation indicates rapid learning. Firms in stable environments have the potential to grow indefinitely at the rate at which firm capabilities exceed depletion rates (C m a x - D). However, competitive conditions are rarely, if ever, stable. Changes in the environment have the effect of reducing capabilities and thus slowing or reversing growth. Changes within a firm have the same effect. Thus, the loss of a major customer or key employee would each have the net effect of reducing C(t) and, in essence, returning the firm to the position of having to cope with newness. Large asset stocks buffer the firm against such shocks while low levels of organizational assets increase the likelihood of failure (Amburgey et al., 1993). Figure 3.2 illustrates the growth trajectories of three hypothetical firms. Assume that firms will discontinue if assets drop below 30% of the initial endowment level.4 (The model parameters, asset values, and period-to-period growth rates may be found in the Appendix, Table A3.3.) 4 Note that while Levinthal (1991) sets the dissolution threshold as organizational capital reaching zero, our model approaches zero assets only asymptotically. Given that even bankrupt firms still retain organizational skills, routines, and some assets until the point of dissolution, a non-zero threshold of discontinuance seems appropriate. Our choice of a 30% level was arbitrarily chosen for the purpose of the illustrative example depicted in Figure 2. 61 Figure 3.2 Illustrative growth profiles 1.10 ., : 0.20 0.10 -J , , , , , 1 0 2 4 6 8 10 12 Years All firms in this example have a fixed depletion rate of 60% of prior period assets. Firm one (represented by squares in Figure 3.2) has a maximum capability level of 64%, which means that it can aspire to stable growth of 4% if it survives long enough. As can be seen, however, the learning rate of this firm (operationalized by L=10 in equation [3.2]), is relatively slow and the firm fails when assets fall below 30% of initial stocks in period 4. Firm 2 (triangles) has the same learning factor, but greater potential capabilities (as indicated by the maximum G level of 70%); it is able to arrest the net depletion without falling below the failure threshold of 30%. Firm 3 (crosses) has the same maximum capabilities as Firm 1 (and thus approaches 4% growth asymptotically), but it has a quicker learning factor (L = 5) and thus is able to achieve optimal growth without venturing near the failure threshold. If any one firm had started with a larger initial endowment than the others, survival chances would be enhanced. It is also worth noting that after 12 periods, Firm 3 has a greater stock of assets than Firm 2. Given Firm 2's higher maximum capability level (and consequently higher maximum growth rate), it would catch Firm 3 in period 15 and continue to outgrow it beyond that point. Of course, this scenario does not include environmental or other shocks that 62 would disrupt the smooth growth trajectories and force the firms to cope with new pressures and new learning. Hang gliding provides a useful metaphor for the process outlined above. The glider must first achieve a certain altitude, usually accomplished by driving to the top of a cliff. Altitude is analogous to the level of organizational capital required to initiate a new enterprise. Once the glider is launched, altitude is sacrificed in exchange for velocity, which in turn imparts lift. The new venture sacrifices capital in exchange for production runs, advertising, clients, and supplier relationships, which in turn lead to revenues. The trick for gliders is to achieve enough lift to offset the inexorable downward force of gravity without getting too close to the ground, i.e., running out of altitude. An entrepreneur's challenge is to generate sufficient revenue streams to offset the costs of doing business, before asset stocks are depleted. In summary, organizations begin life with asset stocks that are subject to depletion and replacement. Firms that fail early in their existence do so because (i) initial assets were low, (ii) initial depletion rates were high, (iii) initial capabilities were low, and/or (iv) the rate of capability acquisition (learning) was low. By extension, firms that fail after an initial period of existence may have been unable to adjust to changed internal or external conditions. From these general propositions, we next develop a series of testable hypotheses. We develop a set of hypotheses concerning the liability of newness and another set concerning the liability of age. The liability-of-newness hypotheses encompass initial endowments of human capital, financial capital, general management skills, financial management skills, and marketing skills. The liability-of-aging hypotheses capture environmental change, managerial inertia, and the pursuit of growth. Each are described in detail below. Hypotheses Newness Firms with experienced management should have relatively high initial levels of both organizational capital and capabilities. An alternative position holds that experienced managers may have preconceptions about the way things are done and are thus closed to learning (Hall, 1994). Bruderl, Preisendorfer, & Ziegler (1992) found a concave relationship between work experience and the survival chances of new businesses. However, given substantial empirical evidence of a positive association between survival odds for an organizational and prior managerial experience (Carter, Williams, & Reynolds, 1997; Cooper et al., 1994; Pennings et al., 1998), we anticipate an inverse relationship between experience and early mortality risk of firms. HI: Low levels of prior managerial experience are more prevalent among firms that fail early than among firms that fail late. Our model of firm survival and growth includes the broad construct of organizational capital. Yet much of the asset depletion in the early stages of a firm's life is due to outlays of cash. The costs of doing business are as onerous as they are varied. Lack of financial capital when firms are nascent may thus have fatal consequences (Cooper et al., 1994; Gaskill et al., 1993; Venkataraman, Van de Ven, Buckeye, & Hudson, 1990). The following hypothesis reflects the expectation that firms with low initial capital or unbalanced capital structures will number among the early exits from a given population. H2: Young firms are more vulnerable to capital structure inadequacies than are older firms. Our model predicts that early failure may be due to low levels of managerial capabilities. The literature makes it clear that the fitness of a firm is damaged by managerial deficiencies in a number of areas (Gaskill et al., 1993; Larson & Clute, 1979; McKinlay, 1979). The following 64 hypotheses reflect this multi-dimensional view, and suggest that inadequacy in general management, financial management, or marketing may each contribute to early failure. H3: Young firms are more vulnerable to failure due to low levels of general management skills than are older firms. H4: Young firms are more vulnerable to failure due to low levels offinancial management skills than are older firms. H5: Young firms are more vulnerable to failure due to low levels of marketing skills than are older firms. Age Firms that survive through the early years face very different issues than do young enterprises. As noted by Aldrich & Auster "the major problem facing smaller and younger organizations is survival, whereas larger and older organizations face the problem of strategic transformation" (1986: 193). The established routines of older organizations, which in many cases were critical to their initial survival, can become liabilities in the face of changing competitive conditions. Amburgey et al. (1993) noted that, while older organizations may be severely affected by change, they also are often the best suited to withstand shocks by virtue of their accumulated asset stocks. However, holding levels of asset stocks constant, we anticipate that the firms in the older cohort are more susceptible to failure due to rigidities of inertia than are firms in the younger cohort. H6: Old firms are more vulnerable to changes in industry conditions than are younger firms. H7: Old firms are more vulnerable to managerial inertia than are younger firms. An additional cause of failure noted in the literature is the pursuit of growth (Boardman et al., 1981). If one envisions growth as the voluntary pursuit of newness, one would expect older 65 firms, with their entrenched routines and procedures, to be more vulnerable than are younger firms to which growth is simply the only way to survive. H8: Old firms are more vulnerable to growth hazards than are younger firms. M E T H O D O L O G Y Sample All corporate bankruptcies in Canada are processed through the office of the Superintendent of Bankruptcies, which assigns a trustee to each case. Trustees are required to become familiar with the operations of each bankruptcy. The data used in this study were drawn from a survey of bankruptcy trustees who responded to a survey while they were handling active bankruptcy files (see Appendix, Table A3.4). The survey was developed with the Canadian Insolvency Practitioners Association and compiled by Statistics Canada (Baldwin et al., 1997). There were a total of 1910 bankruptcies in Canada in the six-month period from March to August 1996. Surveys were sent to trustees for 1085 of these cases and 550 valid responses were obtained (51%). A total of 426 of these surveys contained data on the age of the business at the time of bankruptcy. The age distribution of the bankrupt enterprises is shown graphically in Figure 3.3 (for data values, see Appendix, Table A3.5). It can be seen in Figure 3.3 that young firms are heavily represented in the sample, providing confirmation that young firms are the highest risk category for failure in general and bankruptcy in particular. Twenty-nine percent of the firms in the sample were one or two years old at the time of bankruptcy, 42% were in the three to nine year old range, and the remaining 29%o were ten years old or more. The mean age of the firms in our sample was 7.7 years (median 10.0). Net losses incurred had a mean value of $426,000 (median $139,300). Eighty percent of the bankrupt firms were in service industries, which is representative of the general composition 66 Figure 3.3 Age distribution of bankrupt firms S u o S s 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 A g e a t B a n k r u p t c y of the Canadian economy (Johnson, Baldwin, & Hinchley, 1997). As well, the composition of the sample is generally consistent with the industry distribution of incorporated bankruptcies in 1992, (see Appendix, Table A3.6), although there appears to be slight over-representation in wholesale and retail and under-representation in the construction industry. The survey asked respondents to use a 5-point Likert scale to indicate to what extent a given factor (e.g., under-capitalization) contributed to the bankruptcy. Responses ranged from "Not at all" (1) to "A great deal" (5). Means, standard deviations, and inter-item correlation coefficients are presented in Table 3.1. Three items on the survey pertained to managerial experience: the number of years a manager worked for the bankrupt firm, the number of years a manager was in the industry, and the number of years a manager had been a manager. Since we were interested in the contribution of prior managerial experience to firm survival, we eliminated from the data set those managers who had not been with the firm at inception. This was accomplished by comparing the number of years that an individual had been with the firm with the age of the firm. This reduced the number 67 3 |3 O O 0 0 O cn — O o o cn —' © o o — o cn oo ov m C N — — o o o o o © q V 0 f N O — d d o o in 1 - cn f N —* d d d o q 0 0 •>* •t ro VO —• d d d d o O •n VO Ov r-cn —' d d d d d O o in OV ( N f N 0 0 cn cn vo —• .d d d d d d o o f N O o o o 0 0 o q f N O f N O — d 1 d • d d d d d • o q cn in cn o OV o o o o cn o o — o d d d d d d i d • in cn in 0 0 0 0 t-; o r-; 0 0 in f N in vq vd VD r- OV O •* 0 0 r-OV cn VD VD cn VD 0 0 cn in f N f N f N f N f N f N t N CN cn f N in cn f N cn f N cn f N f N cn f N cn ^ -f N OO cn f N fN o m "* fN O —' ci ci ci ci in r- o fNI ~- fN fN ci ci ci ci o o © o N N - -ci ci ci ci •* ov ov t-~ fN O O O ci ci ci ci o o o o o o Ov f N OV 1.00 -0.06 0.11 O q f N cn VO o — d d d O q OV VD 0 0 VD VD o f N q —1 d d d d o o f N VD o 0 0 o — d d d d d O q VO cn cn t N in cn o •n o — d d d d d d o o o V 0 cn o f N 0 0 ov o o — d d d d d d d cn fN f N cn OV f N cn f N Ov f N OO o VO o d d d d d d d • d • ov cn in cn cn cn f N f N f N f N VD o d d d d d d d d cn VO OV cn f N o f N cn o •<*• o d d d d d d d • d • cn f N •n vo r-f N o cn o d d d d d d d d cn vo in f N o cn f N cn f N O d d d d d d d • d t--o o o o cn f N t— o oo o d d d d d d d i d 0 0 fN vo in r-- OV in o o fN d d d d d d d d 0 0 o o o o (N •n o VO o d d d d d d d d o vo o t N o vo cn cn o cn o d d d d d d d d OV >n OV o 0 0 0 0 VD cn •n o d d d d d d d i d • vo f N t N ov cn CN o CN m q t N d d d d d d d d o • o cn •n o o vo o o d d d d i d • d d d o VO o vo VD o o OV o - o d d d d • d d d d VO o f N 0 0 ro cn cn cn o q •n cn o o ov in 0 0 in o O OV vo OV — — d f N — —1 f N CN f N o f N f N 0 0 f N f N cn f N cn f N cn 0 0 r-u y cd ' — N 3 O 4= — x C ^ H •I u -s — fN cn in vd a. c u o Q U o o CQ 13 M u oo 9- -a -s J3 on -J <C _ . u-o 1—I O r~ oo ov - tN c i * g in vo oo vo o A lo Id 1 V O A lo Id 1 V D , Id Icn CN of observations with which we could test Hypothesis 1 to 108 (68 young failures with incumbent managers and 40 older failures with incumbent managers). Analysis To test the hypotheses, we split the sample. "Young" firms are defined as those that were less than three years old when they failed (122 firms) and "old" firms are those that were older than nine years of age at the time of bankruptcy (124 firms). This grouping reflects the high-density, peak segment of the hazard curve in the case of the young failures and the stable, flatter portion of the curve (see Figure 3.1). We then employed three separate tests. We began with a series of means tests, comparing characteristics of the young failures with those of the older bankrupts. For each hypothesis, item-by-item difference scores were calculated, along with p-values for the Student's T distribution. Second, since each hypothesis was operationalized by at least two items, combined means tests were performed using Hotelling's T 2 as the criterion of significance. Finally, we ran a series of maximum likelihood logit regressions, in which each run included only those variables for a specific hypothesis5. Results of our analyses are presented in Table 3.2. Note that the first five hypotheses pertain to the liabilities of newness, while hypotheses six through eight deal with the liabilities of age and inertia. In addition to examining the differences between young and old bankruptcies on a hypothesis-by-hypothesis basis, we also wished to see how well the various factors distinguished among the young and old sub-samples when evaluated together. To this end, we combined the variables in a series of logit regression models. The first of these included all variables that were hypothesized to be elements of the liability of newness (excluding managerial experience). The 5 Although the firms in our sample all failed within the same six-month period in 1996, they were born in different years and thus experienced different market conditions over their respective life histories. There may thus be structural determinants that are not accounted for in our models. Pooling the young and old firms for comparison and analysis may violate the assumptions of maximum likelihood estimation, particularly those pertaining to the independence of error terms. Results should be interpreted with this caveat in mind. 69 u In CO 3 1 Sb . i . o ^ l—i U X i , S3 CD '5 oo « o o I J U o CD > o ft P -o w 00 > < w (50 t/> 3 M l , ° > > O N H r o T f C N O N O T f 00 r- T f O o o r o T f o o (-• O N T f C N r H 00 O N O 00 m N O Tf' || o d N O || O d II O d N O II o d C N || p d N O || o d C N I  T f d C N V C N v • C N V C N V <N V <N V C N V X a. X ft X & ft X n. X ft X ft 8 I & CL) S S IP 73 s •2 « « o 111 ft 1 5 CD ^ - 3 S cy s So oo s -I JJ -a 2 CD bO • T3 CD Is O Si S T3 o 43 -H* ft CD -a s I s I 3 2 is ts ts ti ca x> ts D o «s <u Q ft co « o o X> H o o ft & ts '9- S 00 .2 .S o M 3 H i - r t > i-H ^ 5* '2 § C o (U cJ5 £ .S so JJ CO CO S =* a cr e JJ I s I 5» 00 IS JJ £ 2 ^ to «3 o '5 * J +-> +-» O «J •§ 2 ft JJ H , 3 O .-73 O CO ft tH s •S *S s a •I •2 >, C oo o> o P s u ^ e TT o Tf' CN v X ft * * * # * o c n oo CN CN O N T f f -00 CN o o NO m # N O •/I m T f # o 00 m r _ r ^ r _ H o 00 CN T f o N O CN CN o o T f O CN o N O N O O o CN O O T f o O o CN o d d i d d • d • d d d • d d d • d d • d d d d d d d d >n CN NO o o T f <N <—1 CN in CN oo CN oo CN CN r n r n CN N O CN d CN _ II CN II •/-> II O N II o I  I  »n II CN CO CN 0 S T f T-H r ? oo O N " o G " oo T ? o r ^ m O CN o m p CN o m o CN T f NO d CN d CN d CN d CN d CN d <N d d CN V CN V r n V r n V r n V r n V r n V CN V OH ft bv, ft ft HH ft P H ft P H ft P H ft rC ft t ~ T f ON m NO m O N m o o N O f - m oo NO cn OO l-« ^ H o T f >o o 00 T f O N CN o <N o m r- O N r ^ m O N oo T f O o o m o - H *—H O N p p p CN o o r H o i—i d d d d d d d d d d d d d d d d d d d d d O O O O O O O O O O O O O O O O O O O O O V V A A A A A A A A A A A V V V V V V V V >< >- >- >- > >- >- > > > > > > > >- > NO N O N O >o m T f m m T f CN o m o m m m m o T f O '—1 - H T—H ^ H T-H *—< »—' r - l '—i o i—i r H i—i i—f d d d d d d d d d d d d d d d d d d d d in CN T f T f o ^ H m 00 m CN O N oo CN N O T f m o o oo m T f m oo t-; O N 00 r n in '—i oo '—' p p cn oo CN CN CN CN CN CN CN d CN CN CN CN CN m N O m T f T f <r> NO T f ON T f ON m oo T f CN CN m oo oo - H *—' - H *—' T—H r— O O '—' o I—1 i—| r-l -—i d d d d d d d d d d d d d d d d d d d d d o ^ H C D CN _i oo _ NO m T f m N O CN r n O N CO T f oo NO OO <N ON r n T f ON r H o r H CN oo CN OO °) oo CN N O in CN r n CN CN CN CN <N CN CN d d CN CN 2 -a " 2. 5" E II o .2 > to -a 3 CO r-S ^ < N Z O O ft rZ 5 C r --a ^ § .22 O 00 H 1) <u — CO Xl - CO e .5 .S <U CD 00 00 c a co S U U U 5 I N J S -t C5 co .: CH. C H o o *t ^ o o CO CO —I J p CD C3 co J S —1 2 o 5 oo co C XI 3 O O HH O o U H * •e c ft T3 <~ J 3 c/l C H ft eb cj •a CO to _3 CO o d V o d V o V o 00 o CO -3 — •S 3 J 3 CO ft H H J 2 next model included those factors that we identified as contributors to the liabilities of age and inertia. The third model combined the variables from Models 1 and 2. Finally, the fourth model incorporated all elements of Model 3 as well as the variables for "previous industry experience" and "previous managerial experience." These variables were not included in the earlier models because of their effect on the degrees of freedom. Since the firms that are retained differ in a meaningful way from those that are excluded (i.e., they had managers with continuous tenure during the life of the firm), results of this final model should be interpreted with caution5. Results of the joint logit estimation model are presented in Table 3.3. Table 3.3 Combined Logit Analysis Model 1 Model 2 Model 3 Model 4 Coef. P>|z| Coef. P>|z| Coef. P>|z| Coef. P>|z| Previous managerial experience 0.00 0 978 Previous industry experience -0.10 0 235 Capital structure 0.09 0 432 0.10 0.566 -0.26 0 364 Under-capitalized -0.23 0 020 -0.21 0.162 -0.20 0 491 Breadth of knowledge -0.59 0 005 -0.58 0.05 -0.66 0 240 Depth of knowledge 0.27 0 174 0.04 0.888 -0.01 0 990 Managerial control 0.06 0 622 -0.28 0.193 0.52 0 189 Bookkeeping -0.16 0 159 -0.44 0.015 -0.67 0 099 Working capital management 0.21 0 133 0.25 0.257 0.46 0 269 Prod cost control -0.04 0 751 -0.05 0.767 -0.32 0 366 Pricing strategy 0.02 0 900 0.08 0.666 0.55 0 184 Product quality 0.47 0 020 0.24 0.406 -0.28 0 701 Niche -0.45 0 000 -0.58 0.002 -1.31 0 003 Technological change -0.07 0.744 -0.19 0.532 -0.34 0 679 Market conditions 0.48 0.009 0.70 0.006 1.88 0 004 Labor legislation -0.27 0.227 -0.33 0.242 -1.77 0 056 Adaptability -0.04 0.821 0.24 0.296 -0.66 0 200 Initiative 0.15 0.465 0.06 0.824 0.43 0 348 Flexibility 0.29 0.151 0.81 0.004 1.04 0 055 Market share 0.42 0.049 0.46 0.103 1.25 0 032 Growth -0.04 0.782 -0.09 0.664 -0.26 0 474 Constant 1.26 0 005 -2.05 0.002 -0.17 0.854 -0.36 0 869 Observations 218 160 156 80 Chi-square 39.09 0 22.82 0.004 62.55 0.000 50.24 0 000 Pseudo R-sq 0.129 0.103 0.291 0.463 Log Likelihood -132 -99.2 -76.4 -29.10 6 Models 1, 2, and 3 were also run on the subset of firms that claim uninterrupted managerial tenure, as does Model 4. The results are comparable for the full and reduced data sets. Coefficients did not differ in sign, although there were some differences in level of significance. 71 RESULTS The results presented in Table 3.2 provide varying degrees of support for six of our eight hypotheses. Hypothesis 2 regarding capital structure was supported on the strength of the significant difference between the coefficients on "under-capitalization" for the two groups. There was little difference between the young and old failures on the issue of capital structure balance, but the younger bankrupts were clearly more prone to problems associated with under-capitalization. The results indicate that the young failures were significantly more deficient in their breadth and depth of general management knowledge, as indicated by the univariate means tests. However, in the logit model, only breadth of knowledge was a significant predictor of early failure. The combined results generally confirm Hypothesis 3's expectation that deficiencies in general management would be an element of the liability of newness. Young failures were below par at bookkeeping, relative to the older firms in our sample. This specific finding was significant at the p < 0.01 level in both the means comparison test and the logit regression. The mean differences for working capital management and product cost control were of the expected signs, but neither was significant at even the 10% level. The combined means test and the logit analysis provided marginal support (p < 0.10) for Hypothesis 4. Of the three variables that we employed to test Hypothesis 5 (Marketing), only "failure to establish a niche" was significantly higher for the young failures. Surprisingly, poor product quality was a significant predictor of failure late in life - an opposite finding to that predicted by Hypothesis 5. The Chi-square results for the logit analysis, like Hotelling's T 2 , was significant at only the p < 0.10 level. And, given the mixed results, we can only say that Hypothesis 5 was partially supported by the data. 72 The univariate means tests for Hypothesis 6 were all of the expected sign, and both change in market conditions and change in technology were significant at the p < 0.05 level or better. The logit analysis indicated significance for only a single measure (change in market conditions) and the combination of the three yielded a Chi-Square and Hotelling's T 2 significance of only p < 0.10. There is tentative support for our expectation that environmental change would be of greater impact to firms later in their life than earlier. Contrary to our expectations, the data indicate that the managers of the young failures had greater prior managerial and industry experience, at start-up, than did those of the firms that stayed in business for at least a decade. This result has marginal significance when evaluated by Hotelling's T 2 and by the Chi-squared value for the logit analysis of Hypothesis 1. A nai've test of the hypothesis, in which previous industry and managerial experience means are compared without controlling for involvement at start-up yields the expected result that older failures were managed by more experienced individuals than were firms that failed young. However, our hypothesis was based on the presumption that managerial experience would contribute to the organizational assets of a new firm and provide a buffer against the liabilities of newness. This was not the case among the firms in our sample that had continuous management from start to finish. It is worth noting, though, that we were unable to contrast the net experience of the early failures with the managerial experience resident in those firms that went on to eventual success. Our seventh hypothesis, on the liability of inertia, was not supported by our sample. Each of the three individual variables was only marginally significant, and the combined means test and the logit model were non-significant. Finally, our prediction that older firms would be more vulnerable to growth hazards (Hypothesis 8) received only marginal support. The individual and combined differences on the two variables only were weakly significant, although of the expected sign (i.e., predictive of failure as older firms). 73 Not surprisingly, the results from the composite models run as logit regressions (Table 3.3) were generally consistent with the findings from our individual hypothesis tests. In the first model, which includes variables that were hypothesized to capture the liability-of-newness effects, the constant term was significantly positive. In the second model, which included liability-of-age variables, the intercept was significantly negative. Thus, absent the independent variables of Model 1, firms would have a tendency to fail at a late age, while in Model 2 the default condition was early failure. Each of these models also had a Chi-square significance value of p < 0.01 or less. The most striking predictors of early failure were under-capitalization, deficient breadth of managerial knowledge, and failure to establish a niche. On the age side of the failure model, changes in market conditions, lack of managerial flexibility, and over-pursuit of market share were most strongly associated with the failure of firms that were ten or more years of age. The most surprising result, as noted above, was the finding young failures were staffed by more experienced managers, at start-up, than were the firms that endured through the early years of adolescence but failed later in life. However, in the full logit model (Model 4), neither of the coefficients for managerial experience was a significant predictor of early failure. Managers with prior experience, particularly experience in the same industry, may be in a position to acquire capital through reputational effects, or may be more selective in taking on projects that are undercapitalized. There is also the issue of substitutability of human and financial capital. Chandler & Hanks (1998) found that financial capital and founder human capital are at least partially substitutable. Accordingly, we wished to explore whether there was an interaction effect between managerial experience and degree of capitalization. We therefore created principal components factors for each of the eight sub-groups of variables associated with our hypotheses. We first ran a logit model with these eight variables, and then re-ran the analysis including an 74 interaction term for managerial experience and capitalization. The results indicated that the interaction term was non-significant, although the model was significant at the p < 0.02 level (see Appendix, Table A3.7). Thus, the possible interaction of experience and capitalization was not supported as a determinant of early failure within our sample. One other area of interest pertains to possible presence of industry-specific effects. Mitchell (1994) observed that manufacturing firms in financial distress are more likely to be sold than shut down, as compared with service firms. This may be due to the greater transferability of physical assets in manufacturing, relative to the personnel-based assets of a service provider. An examination of the relative age distribution within industries in our sample (see Appendix, Table A3.8) reveals two industries that stand out - retail and accommodation, food, and beverages (AFB). The former has a large number of failures in the older firm category (25 young, 40 old), while the latter is strongly skewed within the young failures sub-sample (28 young, 5 old). In order to evaluate the effect of these specific industries on the overall results, we recalculated the Hotelling's T 2 statistics and associated p-values for each of our eight hypotheses, first holding out the AFB firms, and then holding out the retail firms from the sample (see Appendix, Table A3.9). The sample reductions had little effect on H3 (general management) and H5 (marketing management), both of which were strongly significant before and after the industry hold-outs. Likewise, HI (managerial experience) and H7 (inertia) were not supported with or without the sample modification. However, both H4 (financial management) and H8 (pursuit of growth) were less significant when the retail sub-sample was withheld from the analysis. The significance level of H6 (environmental change) degraded under both conditions of sample reduction, and H2 (capitalization) was adversely affected by the omission of the AFB firms. While some of the changes in significance levels may be attributable simply to the reduction in sample size, the 75 nature of the changes also speak to differences in bankruptcy causes between industry sectors. This issue is discussed in greater detail below. When taken together, the findings enumerated above support the general propositions that young firms fail due to low initial endowments of capital and managerial ability, while older firms fail due to environmental change. DISCUSSION In the model presented here, new firms initially shrink due to resource depletion. Some then acquire sufficient capabilities to reverse the trend, while others fail if they do not develop capabilities sufficient to offset the rate of depletion. This model recognizes the role of initial endowments of organizational assets in determining the survival chances of new firms. It also incorporates the concept of failure thresholds. Some firms may choose early exit in the face of initially poor returns, while others with lower performance thresholds may persist even with sustained poor performance. Our model thus offers a descriptive account of the performance of young firms that conforms to observed population-level patterns of exit. Specifically, hazard rates increase during the first few years and decrease thereafter. This pattern is consistent with survival as a function of initial endowments of assets, resource depletion, and capability acquisition. We hypothesized that firms that failed soon after start-up differ from those that failed after ten or more years of operation. We predicted that young failures would be characterized by low initial endowments of human and financial capital and by low levels of managerial capabilities. By extension, we anticipated that older failures would be more likely to have succumbed to environmental change and/or due to inertia and inflexibility. We tested these hypotheses using a unique data set of Canadian bankruptcies. Contrasts of the nature of the 76 young firm failures with those of the older failures provided support for all but two of our hypotheses. The glaring incongruity among otherwise consistent empirical support lay in the area of human capital endowments. While we anticipated that firms with inexperienced management would be more prone to early failure, we found instead that the managers of the young failures had more industry and managerial experience at start-up than did the management of firms that endured for a decade or more. Because the data were cross-sectional, we were not able to evaluate the learning-curve effects captured by the conceptual model. It may be that the managers of the young failures, while they had more experience, were deficient in certain critical capabilities and/or lacking in the ability to acquire new skills. There is also the possibility that adverse selection is at work. If the experienced managers who are available to new firms are available because they performed poorly in their prior employment, we may be capturing the effects of a pool of "bad" managers. The highly skilled managers would not be among the group of candidates from which the start-ups firms drew their managerial talent. Also, while managerial experience is often used as a proxy for human capital, it is at best a blunt instrument for measuring a complex construct. It could be argued that the variables for the dimensions of managerial capability (e.g., breadth of knowledge, bookkeeping ability) are, in fact, better proxies for human capital, in which case the argument that human capital aids survival probability would be supported by the data. The result may also be an artifact of the data. The incidence of continuous managerial tenure among the young failures is greater, and the measures of central tendency less skewed, than are found among the subset of older failures with incumbent management dating to the start of business. Absent this result, the data generally supported our contention that under-capitalization and shortcomings in marketing, financial management, and general management characterize 77 firms that failed early. Our analysis also confirmed our expectation that changing industry conditions inhibit more established firms. We found only marginal support for our hypothesis that older firms are more prone to failure in the pursuit of growth, and there was little support for our hypothesis regarding the effects of managerial inertia. Our exploration of industry-specific effects revealed a high incidence of older failures in retail and a high incidence of younger failures in accommodation, foods, and beverages. These trends are consistent with anecdotal evidence and with implications of our model. The case of the retail industry is one in which there have recently been many changes in the competitive environment - a condition hypothesized to adversely affect older firms. The emergence of discount chains such as Chapters and Home Depot have changed the landscape and sharply increased the competitive pressure on stand-alone booksellers and hardware stores. The emergence of Internet retailers, such as Amazon.com, is further exacerbating the challenge of staying in business that confronts traditional retail outlets. In the food, accommodation, and beverage industry, one can think of restaurants as representative of the failure mode described by our model. This industry is characterized by early failures, as one would expect from our model of asset depletion and organizational learning. When a new restaurant opens for business, it has a fairly short window in which to establish a reputation and build a clientele - a window proportional in length to the level of initial assets. A restaurant that has a high level of quality and service (capabilities) may do well from the outset. Absent a strong starting position, only rapid acquisition of these capabilities (organizational learning) will give the business a chance at survival. Taken together, these results enhance our understanding of firm failure in general and the liability of newness in particular. It has been more than a decade since Aldrich & Auster (1986) suggested that research would benefit from a dual consideration of population- and organization-78 levels of analysis. We began with the observed liability of adolescence phenomenon at the population level and then developed a firm-level model that features organizational capital endowments and managerial capabilities. Our empirical analysis generally supported the model. These findings reinforce the importance of initial capitalization when launching a new venture. Also, given the importance of general, financial, and marketing management skills, the results speak to the importance of seeking outside advisors to bolster deficiencies in the knowledge base of managers and entrepreneurs. The finding that growth ambitions may have played a role in the demise of the mature firms in our sample may serve as a warning that growth should not be an objective to be pursued without carefully weighing the consequences and the potential impact on organizational survival. This research represents only an initial step along the road to uniting the organizational-and population-level views of firm survival and performance. Two future lines of study that are suggested by this inquiry include (1) broadening the scope of exits under study to include modes of discontinuance other than bankruptcy and (2) contrasting young and old failures with a comparable population of young and old survivors. In each case, efforts should me made to reconcile observed population level dynamics with characteristics and behaviors that reflect the way firms do business. 79 R E F E R E N C E S Aldrich, H. E. , & Auster, E. R. (1986). Even dwarfs started small: Liabilities of age and size and their strategic implications. Research in Organizational Behavior, 8(11, 165-198. Altman, E. I. (1971). Corporate bankruptcy in America. Lexington, MA: D.C. Heath and Co. Amburgey, T. L. , Kelly, D., & Barnett, W. P. (1993). Resetting the clock: The dynamics of organizational change and failure. Administrative Science Quarterly, 38(1), 51-73. Amit, R., & Schoemaker, P. J. H. (1993). Strategic assets and organizational rent. Strategic Management Journal, 14(1), 33-46. Baldwin, J., Dupuy, R., & Gelatly, G. (1998). The determinants of survival for new Canadian firms: A characteristics-based approach. Working Paper. Ottawa: Minister of Industry. Baldwin, J., Gray, T., Johnson, J., Proctor, J., Rafiquzzaman, M . , & Sabourin, D. (1997). Failing concerns: business bankruptcy in Canada. Ottawa: Minister of Industry. Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99-120. Bates, T. (1990). Entrepreneur human capital inputs and small business longevity. Review of Economics & Statistics, 72(4), 551-559. Bates, T., & Nucci, A. (1989). An analysis of small business size and rate of discontinuance. Journal of Small Business Management, 27(4), 1-7. Berryman, J. E. (1983). Small business failure and bankruptcy: A survey of the literarture. European Small Business Journal, 1 (4), 47-59. Boardman, C. M . , Bartley, J. W., & Ratliff, R. B. (1981). Small business growth characteristics. American Journal of Small Business, 5(Winter), 33-42. Brander, J., Hendricks, K., Amit, R., & Whistler, D. (1998). The engine of growth hypothesis: On the relationship between firm size and employment growth. Working Paper: The University of British Columbia. Bruderl, J., Preisendorfer, P., & Ziegler, R. (1992). Survival chances of newly founded business organizations. American Sociological Review, 57(1), 227-242. Bruderl, J., & Schussler, R. (1990). Organizational mortality: The liability of newness and adolesence. Administrative Science Quarterly, 35(3), 530-547. Carroll, G. R. (1983). A stochastic model of organizational mortality: Review and reanalysis. Social Science Research, 12(4). 303-329. 80 Carroll, G. R., & Delacroix, J. (1982). Organizational mortality in the newspaper industries of Argentina and Ireland: An ecological approach. Administrative Science Quarterly, 27(2), 169-198. Carroll, G. R., & Huo, Y. P. (1986). Organizational task and institutional environments in evolutionary perspective: Findings from the local newspaper industry. American Journal of Sociology. 91(4). 838-873. Carter, N. M . , Williams, M . , & Reynolds, P. D. (1997). Discontinuance among new firms in retail: The influence of initial resources, strategy, and gender. Journal of Business Venturing. 12(1). 125-145. Chandler, G. N., & Hanks, S. H. (1998). An examination of the substitutability of founders human and financial capital in emerging business ventures. Journal of Business Venturing. 13. 353-369. Cochrane, A. B. (1981). Small business mortality rates: A review of the literature. Journal of Small Business Management. 19(4), 50-59. Conner, K. R. (1991). A historical comparison of resource-based theory and five schools of thought within industrial organization economics: do we have a new theory of the firm? Journal of Management, 17 (1), 121-154. Cooper, A. C , Gimeno-Gascon, F. J., & Woo, C. Y. (1994). Initial human and financial capital as predictors of new venture performance. Journal of Business Venturing. 9(5). 371-395. Daily, C. M . (1995). The relationship between board composition and leadership structure and bankruptcy reorganization outcomes. Journal of Management. 21(6), 1041-1056. Dun and Bradstreet. (1981). Business failure record. Dunn, T., Roberts, M . J., & Samuelson, L. (1988). Patterns of firm entry and exit in US manufacturing industries. Rand Journal of Economics. 19(4), 495-515. Fichman, M . , & Levinthal, D. A. (1991). Honeymoons and the Liability of Adolescence: A New Perspective on Duration Dependence in Social and Organizational Relationships. Academy of Management Review. 16(2), 442-468. Fredland, J. E. , & Morris, C. E. (1976). A cross-sectional analysis of small business failure. American Journal of Small Business. 1(1). 7-18. Freeman, J., Carrol, G. R., & Hannan, M . T. (1983). The liability of newness: Age dependence in organization death rates. American Sociological Review. 48(4). 692-710. Gaskill, L. R., Van Auken, H. E. , & Manning, R. A. (1993). A factor analytic study of the perceived causes of small business failure. Journal of Small Business Management, 31(4). 18-31. 81 Gimeno, J., Folta, T. B., Cooper, A. C , & Woo, C. Y. (1997). Survival of the fittest? Entrepreneurial human capital and the persistence of underperforming firms. Administrative Science Quarterly, 42(4), 750-783. Hall, G. (1994). Factors distinguishing survivors from failures amongst small firms in the U K construction sector. Journal of Management Studies, 31(5), 737-760. Hall, G. (1992). Reasons for insolvency amongst small firms - a review and fresh evidence. Small Business Economics, 4(3), 237-250. Hambrick, D. C , & D'Aveni, R. A. (1988). Large corporate failures as downward spirals. Administrative Science Quarterly, 33(1), 1-23. Johnson, J., Baldwin, J., & Hinchley, C. (1997). Successful entrants: creating the capacity for survival and growth. Ottawa: Minister of Industry. Jovanovic, B. (1982). Selection and the evolution of industry. Econometrica, 50(4), 649-670. Kalleberg, A. L. , & Liecht, K. T. (1991). Gender and organizational performance: Determinants of small business survival and success. Academy of Management Journal, 34(1), 136-161. Keasey, K., & Watson, R. (1987). Non-financial symptoms and prediction of small company failure: A test of Argenti's hypothesis. Journal of Business Finance and Accounting, 14(3), 335-354. Larson, C. M . , & Clute, R. C. (1979). The failure syndrome. American Journal of Small Business. 4(2), 35-43. Levinthal, D. A. (1991). Random walks and organizational mortality. Administrative Science Quarterly, 36(3), 397-420. Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology, 14, 319-340. Litvak, I. A., & Maule, C. J. (1980). Entrepreneurial success or failure - ten years later. Business Quarterly. 44, 68-78. Lussier, R. N. (1995). A nonfinancial business success versus failure prediction model for young firms. Journal of Small Business Management, 33(1), 8-20. McKinlay, R. A. (1979). Some reasons for business failures in Canada. Cost and Management, 53(3), 17-21. Mitchell, W. (1991). Dual clocks: Entry order influences on incumbent and newcomer market share and survival when specialized assets retain their value. Strategic Management Journal. 12(1), 85-100. Mitchell, W. (1994). The dynamics of evolving markets: The effects of business sales and age on dissolutions and divestitures. Administrative Science Quarterly, 39(4), 575-596. 82 Mitchell, W., Shaver, J. M . , & Yeung, B. (1994). Foreign entrant survival and foreign market share: Canadian companies' experience in United States medical sector markets. Strategic Management Journal 15. 555-567. Moulton, W. N., & Thomas, H. (1993). Bankruptcy as a deliberate strategy: Theoretical consideration and empirical evidence. Strategic Management Journal, 14(1), 125-135. O'Neill, H. M . , & Duker, J. (1986). Survival and failure in small business. Journal of Small Business Management. 24(1). 30-37. Pennings, J. M . , Lee, K., & Witteloostujin, A. 1. (1998). Human capital, social capital, and firm dissolution. Academy of Management Journal. 41(4), 425-440. Phillips, B. D., & Kirchhoff, B. A. (1989). Formation, growth, and survival: Small firm dynamics in the US economy. Small Business Economics, 1(1), 65-74. Preisendorfer, P., & Voss, T. (1990). Organizational mortality of small firms: the effects of entepreneurial age and human capital. Organization Studies. 11(1). 107-129. Rumelt, R. P. (1984). Towards a strategic theory of the firm. In R. Lamb (Ed.) Competitive Strategic Management, (pp. 556-570). Englewood Cliffs, NJ: Prentice-Hall. Rumelt, R. P. (1991). How much does industry matter? Strategic Management Journal. 12(3). 167-185. Schussler, R. (1988). The early death of organizations - problems of the liability of newness. Unpublished Manuscript. University of Munich. Singh, J. V. , House, R. J., & Tucker, D. J. (1986). Organizational change and organizational mortality. Administrative Science Quarterly. 31 (4). 587-611. Singh, J. V., Tucker, D. J., & House, R. J. (1986). Organizational legitimacy and the liability of newness. Administrative Science Quarterly. 31 (2). 171-193. Slater, S. F., & Narver, J. C. (1995). Market orientation and the learning organization. Journal of Marketing. 59(3). 63-74. Statistics Canada. (1985). Canada Year Book 1985. Ottawa: Minister of Supply and Service Canada. Statistics Canada. (1994). Canada Year Book 1994. Ottawa: Minister Industry, Science and Technology. Stearns, T. M . , Carter, N. M . , Reynolds, P. D., & Williams, M . (1995). New firm survival: Industry, strategy and location. Journal of Business Venturing. 10(1). 23-42. Stinchcombe, A. L. (1965). Organizations and social structure. In J.G. March (ed.), Handbook of Organizations, (pp. 142-193). Chicago: Rand-McNally. 83 Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509-533. Ulmer, K. J., & Nielsen, A. (1947). Business turnover and causes of failure. Survey of Current Business Research, 4(4), 10-16. Venkataraman, S., Van de Ven, A. H., Buckeye, J., & Hudson, R. (1990). Starting up in a turbulent environment: a process model of failure among firms with high customer dependence. Journal of Business Venturing, 5(5), 277-295. Watson, J., & Everett, J. E. (1993). Defining small business failure. International Small Business Journal, 11(3), 35-48. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal, 5(2), 171-180. 84 APPENDIX Table A3.1 Data for Figure 3.1 Duration Goods Services 0 0.00 0.00 1 0.24 0.22 2 0.20 0.23 3 0.16 0.18 4 0.16 0.16 5 0.14 0.15 6 0.11 0.14 7 0.12 0.10 8 0.10 0.11 9 0.12 0.13 10 0.09 0.10 11 0.10 0.11 Source: Baldwin, J., Dupuy, R. & Gelatly, G. (1998). 85 Table A3.2 Empirical Studies of Organizational Failure Author (s ) | Definition 1 Kev Findinq(s) POPULATION-LEVEL STUDIES Amburgey, Kelly &Barnett(1993) Voluntary cessation of operations. "Our analyses largely support the structural inertia model ... The risk of failure is strongly age-dependent and ... age acts as a powerful moderator of the relationship between organizational age and failure." (p. 69) "If older organizations are more disrupted by changes but the lower base rate produced by the main effect of age (or other factors) results in a lower net risk of failure, then .... The organizations most likely to survive fundamental change are old organizations, not young organizations." (70) Bates (1990) Discontinuance Young firms were (1) most likely to fail, (2) smaller, and (3) less profitable. Highly educated owners are more likely to survive. Education and initial financial capital appear to be interrelated and jointly predict survival. Owner age has an inverted U-shaped relationship to failure. Bates & Nucci (1989) Discontinuance Firm size and age both affect the likelihood of survival. When age is controlled for, a strong direct relationship exists between survival and size. Bruderl & Schussler(1990) Discontinuance "... there is a very strong liability of smallness in our data ... The more employees a firm has, the longer it will survive." (p. 540) There is evidence of a period of adolescence, the duration of which varies according to type of business. Carroll (1983) Discontinuance "The empirical estimates obtained for the three models showed that Makeham's Law is the best-fitting model for organizational mortality ... The empirical findings support the liability-of-newness arguments advanced by Stinchcombe and they provide no evidence for the obsolescence argument. In comparing the estimates for different types of organizations, the capital-intensive manufacturing organizations generally showed lower death rates than other types of organizations." (pp. 326-326) Carroll & Delacroix (1982) Discontinuance "... the (mortality) rate is highest when newspapers are young and gradually declines with age until the asymptote is approached at approximately age fifteen." (p. 178) "... the major variation in organizational death rates occurs in the early stages of organizational life ... (indicating) that environmental selection actions with greatest force on new organizations." (p. 190) Carroll & Huo (1986) Discontinuance "...our analysis suggests that institutional environmental variables, especially political turmoil, more strongly affect the founding and death rates of organizations in the newspaper industry, whereas task environmental variables more strongly affect the performance of ongoing organizations." (p. 867) Dunne, Roberts & Samuelson(1988) Exit from an industry "... industry-specific factors play an important role in determining entry and exit patterns." (p. 496) "... small, young firms (have) the highest failure rates. Both the probability of survival and the size of surviving firms vary significantly across the three entrant types (new firms, diversifying firms-new production, diversifying firms-new product mix)." (p. 496) 86 Author(s) Definition Kev Findinq(s) Freeman, Carrol & Hannan(1983) Dissolution or merger 1. "... there is indeed a liability of newness - death rates at early ages are much higher than those at later years" (p. 706) 2. "... the strength of age dependence differs for two different kinds of organizational death, dissolution and absorption by merger" (p. 706) 3. "... the effects of aging can be separated from those of initial size within a plausible parametric model" (p. 706) 4. "The plots of the integrated hazards over historical time ... do not reveal any sharp discontinuous jumps or dips in the rate ... Therefore, it seems unlikely that we have mistaken historical processes for aging processes" (p. 706) Levinthal (1991) Discontinuance "What the random walk model suggests is that it is critical to consider the history dependence of the selection process. While the current literature has focused on organizational age as the critical attribute of prior history, the analysis ere suggests that this variable may, in part, be a proxy for a broader construct, which I have termed organizational capital." (p. 418) Phillips & Kirchhoff(1989) Discontinuance "... new small firms show an average survival rate of 39.8 percent after six years ... survival rates among new firms vary depending upon whether or not the new firm grows. The greater the growth as measured by employment, the greater the firm survival rate." (p. 68) "... although larger initial size assures a greater survival rate for zero or low growth firms, initial size does not substantially affect survival rates for firms with significant growth." (p. 69) Preisendorfer & Voss(1990) Discontinuance "(I)t emerges that survival times of manufacturing firms are longer than those of trading firms." (p. 116) "(W)e can conclude that, even after controlling for other factors affecting the survival chances of firms, there is a convex founder-age-mortality profile and, vice versa, a concave founder-age survival profile." (p. 121) Stearns, Carter, Reynolds & Williams (1995) Discontinuance "No significant relationships were found for the distribution industry. In manufacturing, quality proponents are more likely to survive. In retail, we find price competitors and quality proponents to be prone to discontinuance, whereas a technology value strategy promotes survival. However, technology value is likely to lead to discontinuance in the service industry. The niche purveyor strategy promotes survival in the service industry." (p. 36) MULTI-LEVEL STUDIES Fichman & Levinthal (1991) Discontinuance "We think that our proposed baseline model is more appropriate than an alternative approach that was suggested by Schussler (1988). What we have characterized as a liability of adolescence, Schussler (1988) has treated as liability of newness with a delayed onset." (p. 461) "Aldrich and Auster (1986) suggested that the size of the organization can have an effect on an organization's probability of survival. We interpret this size effect as a reflection of the organization's endowment." (p. 461) Gimeno, Folta, Cooper & Woo (1997) Discontinuance "Our theoretical model and empirical results suggest that survival is enhanced by economic performance but not uniquely determined by it. Rather, organizations have different required thresholds of performance, and survival (or exit) is determined by whether performance falls above (or below) the threshold. In small and new ventures, the threshold of performance is fundamentally influenced by the human capital characteristics of the entrepreneur, including the values of this capital in alternative uses, psychic income, and switching costs." (p. 774) 87 Author(s) Definition Kev Findinq(s) Pennings, Lee & van Witteloostuijn (1998) Discontinuance "This study produced major evidence for the contention that a firm s human and social capital have important implications for performance. Applying human and social capital theory to firms, we found that the specificity and nonappropriability of such capital diminished dissolutions of professional service firms." (p. 437) Singh, House & Tucker (1986) Ceasing to exist as a distinct legal entity "An important implication of this finding is that a complete treatment of organizational change should consider both adaptation and selection views simultaneously without assuming that either one alone explains organizational change. As this study has shown, selection and adaptation are complementary rather than contradictory views." (pp. 608-609) Singh, Tucker & House (1986) Ceasing to exist as a distinct legal entity "Opposed to an invariant law-like status for the liability of newness, our study suggests an external contingency view. The liability of newness does exist in organizational populations, but it is not constant or uniform across all organizations. It is variable and is contingent on factors such as external legitimacy." (p. 191) FIRM-LEVEL STUDIES Baldwin et al. (1997) Bankruptcy The main reasons for failure are inexperienced management and/or poor financial planning. Sixty-one percent of the failures occurred within the first five years of operation. Boardman, Bartley & Ratliff (1981) Insufficient capital to meet obligations "The failed firms showed an ability to maintain or increase sales, yet experienced increasingly lower profit margins on sales due to proportionately larger increases in costs of goods sold." (p. 39) "The failed companies may have pursued sales growth at all costs." (p. 40) Bruderl, Preisendorfer & Ziegler(1992) Discontinuance 37% of businesses failed within the first five years. No differences between specialists-generalists or between conventional-innovative organizational structures. "Increased schooling and work experience are associated with lower failure rates ... there is a concave relationship between work experience and survival chances. Among specific human capital variables, industry-specific experience is the most important effect." (237) Carter, Williams & Reynolds. (1997) Discontinuance "...women-owned retail businesses have higher odds of discontinuing than those owned by their male counterparts and that the lack of human and financial resources significantly increase the odds of businesses discontinuing." (pp. 139-140) "... whereas men use human and financial resources to enhance the chances of their firms' survival, women find strategic choice more beneficial." (p. 141) Cooper, Gimeno-Gascon, & Woo (1994) Discontinuance "Resources that were statistically significant for marginal survival included education, minority status, parents who had started ventures, experience in similar business, capital and use of professional advisors. Contributing to growth were education, gender and minority status, experience in similar business, partners, capital, and industry sector ... most resource variables appeared not to demonstrate differential impacts on whether a venture marginally grew or survived" (p. 386) Daily (1995) Chapter 11 bankruptcy reorganization "Successfully reorganized firms were characterized by boards which were dominated by outside directors (65 percent), while firms which were liquidated were dominated by inside directors (51 percent insiders)." (p. 1052) "No support was found for a systematic relationship between board leadership structure and successful reorganization." (p. 1052) 88 Author(s) Definition Kev Findinq(s) Fredland & Morris (1976) Dun& Bradstreet Failure is negatively associated with firm age and managerial experience. A categorical analysis of size indicates that "there is, if anything, a greater likelihood of failure as size increases. Apparently the advantages of bigness in isolating firms from failure do not exist until firms become quite large." ( p - 1 2 ) • • u u "Apparently, firms are more likely to enter business in growing, high income areas. Since new firms are more likely to fail than older ones, failure rates are higher in economically expanding areas, in spite of strong demand." (p. 12) Gaskill, Van Auken & Manning (1993) Lack of profitability Factor analysis of survey data reduced to four factors: (1) managerial and planning functions, (2) working capital management, (3) competitive environment, and (4) growth and overexpansion. Hall (1992) Involuntary insolvency "It would appear that the owners of young firms were more likely to suffer from inadequate funding, poor products and inefficient marketing. As their companies aged, however, they were more likely to be buffeted by strategic and environmental shocks for which they did not have the managerial skills to respond." (p. 244) Hall (1994) Deregistration from UK VAT registry. "(T)he negative relationship between the age of owners when they started, or assumed control, of their business and the probability of its survival, may be taken as evidence that there is more to the quality of human capital than might be measured by education levels or years in which it would have been possible to accumulate experience."(p. 750) Hambrick & D'Aveni (1988) Bankruptcy There were no significant effects for domain initiative (inertia) or industry carrying capacity. "... the survivors experienced the same environmental downturns ... However the survivors had far greater slack for withstanding such periods." (p. 17) Kalleberg & Leicht (1991) Discontinuance "Older companies were less likely to go out of business" (p. 152) "(W)e found that the determinants of survival and success operated in much the same way for men and women, suggesting that the processes underlying small business performance are similar irrespective of an entrepreneur's gender." (p. 157) Keasey & Watson (1987) Discontinuance "The results, whilst being of a tentative nature, indicate that marginally better predictions concerning small company failure may be obtained from non-financial data as compared to those which can be achieved from using traditional financial ratios." (p. 351) Larson & Clute (1979) Discontinuance "The authors found that those who have failed in business or who are about to fail have many symptoms in common. These are grouped into three categories: (1) Personal characteristics, (2) Managerial deficiencies, and (3) Financial shortcomings." (p. 37) Litvak & Maule (1980) Discontinuance "In brief, in at least 50 percent of the 18 (failures), the owner-managers were not sufficiently well-prepared managerially, psychologically and financial to cope with protracted difficulties." (p. 71) Lussier(1995) Dun& Bradstreet "Successful firms made greater use of professional advice and developed more specific business plans; failures had more education and less difficulty staffing." (p. 18) McKinlay (1979) Bankruptcy "... there is no doubt that management shortcomings are the overriding cause of business failures ... By this is meant inadequacies in administrative functions including such items as the planning and control of financial operations, and to a lesser degree the planning and control of physical operations." (p. 17) 89 Author (s ) Mitchell (1991) Definition Discontinuance Kev Findinq(s) "Including all entrants and controlling for dual entry order influences, the study finds that industry newcomers and industry incumbents have been subject to different timing clocks when entering new subfields of the diagnostic imaging industry ... (N)ewcomers are subject to the influence of their entry order relative to all entrants, while industry incumbents mainly are affected by other incumbents." (p. 97) Mitchell (1994) Discontinuance "The dissolution rate declined with greater initial sales, related experience, foreign experience, and market age at entry. The dissolution rate rose with entry density and had a U-shaped relationship with density." (p. 586) Mitchell, Shaver & Yeung(1994) Discontinuance "In this paper, we identified a negative influence of low and also of high foreign share at the time of entry on the length of survival of businesses that entered a U.S. market when foreign businesses held only a moderate share." (p. 565) Moulton & Thomas (1993) Chapter 11 bankruptcy "Large firms had a better chance of reorganizing than small firms tor both positive and negative reasons. With their large and varied assets, they are better able to survive substantial losses and decreases in size than small firms ... Also, their very large size makes liquidation or acquisition less likely." (p. 130) O'Neill & Duker (1986) Discontinuance "(S)uccessful firms had a lower percentage of inferior products, a lower level of debt, and a lower level of capital intensity ... The only group of ^ outside advisers that appeared to affect a firm's success was accountants." (30) Venkataraman, Van de Ven, Buckeye & Hudson(1990) Discontinuance "... a small firm is very vulnerable to a failed transaction; irrespective oi the source of the failure. The degree of vulnerability experienced in new small firms, even if they are well managed, is great because, by their very nature, they lack the resources and the to build redundancy - to create a broad, well-balanced product line, or to diversify into other business activities." (p. 293) Watson & Everett (1993) Multiple definitions "... no one definition is clearly superior on all the criteria identified as being important in choosing a measure of failure." (p. 46) "... results indicate that reported small business failure rates are highly sensitive to the definition of failure used."(p. 47) 90 Table A3.3 Illustrative Growth Profiles Firm 1 Firm 2 Firm 3 a(.64), b(10) a(.70), b(10) a(.64), b(5) Years Threshold Net Growth Net Growth Net Growth Assets Assets Assets 0 0.30 1.00 1.00 1.00 1 0.30 0.54 -46.3 0.55 -45.0 0.70 -29.6 2 0.30 0.36 -32.8 0.38 -30.3 0.60 -14.5 3 0.30 0.30 -17.3 0.33 -13.3 0.58 -4.3 4 0.30 0.28 -5.9 0.33 -0.8 0.58 0.7 5 0.30 0.28 0.0 0.35 5.6 0.60 2.7 6 0.30 0.29 2.5 0.38 8.3 0.62 3.5 7 0.30 0.30 3.4 0.41 9.4 0.64 3.8 8 0.30 0.31 3.8 0.45 9.8 0.67 3.9 9 0.30 0.32 3.9 0.50 9.9 0.69 4.0 10 0.30 0.33 4.0 0.55 10.0 0.72 4.0 11 0.30 0.35 4.0 0.60 10.0 0.75 4.0 12 0.30 0.36 4.0 0.66 10.0 0.78 4.0 13 0.30 0.38 4.0 0.73 10.0 0.81 4.0 14 0.30 0.39 4.0 0.80 10.0 0.84 4.0 15 0.30 0.41 4.0 0.88 10.0 0.88 4.0 16 0.30 0.42 4.0 0.97 10.0 0.91 4.0 17 0.30 0.44 4.0 1.07 10.0 0.95 4.0 18 0.30 0.46 4.0 1.17 10.0 0.99 4.0 19 0.30 0.47 4.0 1.29 10.0 1.02 4.0 20 0.30 0.49 4.0 1.42 10.0 1.07 4.0 91 Table A3.4 Details of Bankruptcy Survey Items Survey items and inter-item alpha scores HI: Managerial Experience Previous managerial experience = number of years as a manager - number of years with the firm Previous industry experience = number of years in the industry - number of years with the firm Inter-item coefficient alpha = 0.69 H2: Capitalization To what extent did the following contribute to insolvency: (Not at all (1) to (5) A great deal) a) Unbalanced capital structure (e.g., excessive reliance on short term debt) b) Under-capitalization Inter-item coefficient alpha = 0.62 H3: General Management To what extent was bankruptcy cause by deficiencies in: (Not at all (1) to (5) A great deal) a) Breadth of knowledge (across financing, marketing, operations, etc.) b) Depth of knowledge (within financing, marketing, operations, etc.) c) Control Inter-item coefficient alpha = 0.79 H4: Financial Management To what extent did the following contribute to insolvency: (Not at all (1) to (5) A great deal) a) Poor bookkeeping b) Inability to manage working capital c) Inadequate information on production costs Inter-item coefficient alpha = 0.69 H5: Marketing Management To what extent was bankruptcy caused by: (Not at all (1) to (5) A great deal) a) Poor pricing strategy (over- or under-pricing) b) Inferior or poor quality of product c) Failure to establish a market niche Inter-item coefficient alpha = 0.64 H6: Impact of Industry Conditions To what extent did the following factors contribute to bankruptcy: (Not at all (1) to (5) A great deal) a) Fundamental change in technology within the industry b) Fundamental change in market conditions within the industry (such as product obsolescence) c) Labour or industrial relations legislation Inter-item coefficient alpha = 0.70 H7: Managerial Inertia To what extent was bankruptcy cause by deficiencies in: (Not at all (1) to (5) A great deal) a) Adaptability b) Initiative c) Flexibility (in decision-making) Inter-item coefficient alpha = 0.87 H8: Pursuit of Growth How much emphasis was placed on: (Too little (1) to (5) Too much) a) Market share b) Expansion/growth Inter-item coefficient alpha = 0.67 92 Table A3.5 Data for Figure 3.3 Age Count Percent Cum % 16+ 58 13.6% 13.6% 15 10 2.3% 16.0% 14 15 3.5% 19.5% 13 12 2.8% 22.3% 12 6 1.4% 23.7% 11 9 2.1% 25.8% 10 14 3.3% 29.1% 9 10 2.3% 31.5% 8 16 3.8% 35.2% 7 19 4.5% 39.7% 6 20 4.7% 44.4% 5 29 6.8% 51.2% 4 34 8.0% 59.2% 3 52 12.2% 71.4% 2 61 14.3% 85.7% 1 61 14.3% 100.0% 93 T a b l e A3.6 Industry Composition of 1996 Sample and Canadian Economy from 1980 and 1992 1980 1980% 1992 1992% 1996 1996% Population1 Population2 Sample Primary 843 12.8% 2033 14.2% 32 13.0% Retail 2120 32.1% 3192 22.3% 65 26.4% Wholesale3 984 6.9% 30 12.2% Construction 1301 19.7% 1992 13.9% 17 6.9% F.I.R.E. 254 3.9% 623 4.4% 8 3.3% Accom., Food & Bev.4 1874 13.1% 33 13.4% Business Services 232 3.5% 976 6.8% 20 8.1% Other Services 1113 16.9% 1361 9.5% 26 10.6% Other 732 11.1% 1282 9.0% 15 6.1% Total 6595 100.0% 14317 100.0% 246 100.0% 1. Source: Statistics Canada (1985). 2. Source: Statistics Canada (1994). 3. The 1980 data includes Wholesale and Retail in a combined category (Trade). 4. There is no comparable category in 1980 for Food, Accommodation, and Beverages. 94 Table A3.7 Logit Regression of Composite Factor Variables Composite Factors Coef. Std Err P>|z| Coef. Std Err P>|z| Manager Experience -0.28 0.22 0.201 -0.27 0.23 0 240 Capitalization -0.39 0.30 0.196 -0.40 0.31 0 194 General Management -0.30 0.29 0.303 -0.27 0.29 0 357 Financial Management -0.01 0.31 0.987 0.00 0.32 0 998 Marketing Management -0.57 0.30 0.059 -0.51 0.30 0 092 Environmental Change 0.31 0.23 0.185 0.30 0.23 0 205 Inertia 0.42 0.27 0.126 0.40 0.28 0 154 Pursuit of Growth 0.48 0.23 0.037 0.51 0.23 0 028 Experience * Capitalization -0.22 0.19 0 267 Constant -0.23 0.27 0.398 -0.25 0.27 0 352 Observations 80 80 Chi-square 19.2 0.014 20.5 0 015 Pseudo R-sq 0.177 0.189 Log Likelihood -44.6 -44.0 95 Table A3.8 Age Distribution of Sample by Industry Industry All Firms % Young % Old % Primary 32 13.0 20 8.1 12 4.9 Retail 65 26.4 25 10.2 40 16.3 Wholesale 30 12.2 8 3.3 22 8.9 Construction 17 6.9 7 2.8 10 4.1 F.I.R.E. 8 3.3 4 1.6 4 1.6 Accom., Food & Beverage 33 13.4 28 11.4 5 2.0 Business Services 20 8.1 10 4.1 10 4.1 Other Services 26 10.6 13 5.3 13 5.3 Other 15 6.1 7 2.8 8 3.3 Total 246 100.0 122 49.6 124 50.4 96 Table A3.9 Combined Means Tests with Industries withheld from Sample All Firms . Omit A.F.B. Omit Retail (n = 246) (n = 33) (n = 65) HI: Prior Experience 0.121 0.124 0.053 H2: Capitalization 0.042 0.184 0.076 H3: General Management 0.015 0.016 0.022 H4: Financial Management 0.089 0.054 0.261 H5: Marketing Management 0.000 0.001 0.008 H6: Environment 0.087 0.269 0.265 H7: Inertia 0.475 0.775 0.505 H8: Growth 0.112 0.115 0.242 Note: The values in each column represent the p-value associated with the Hotelling's T 2 of combined mean differences. The scores for All Firms may also be found in Table 3.2. 97 CHAPTER 4: DETERMINANTS OF YOUNG FIRM GROWTH: EVIDENCE FROM CANADA ABSTRACT This paper utilizes detailed survey data from a proportionally stratified, representative sample of 3,000 Canadian firms to evaluate industry- and firm-level determinants of young firm growth. The competitive environment is found to be a poor predictor of the growth of young firms. In general, growth of the seven to ten year old firms in our study did not follow the growth trends of the industries in which they operated. Among firm strategies, innovation was the strongest predictor of revenue growth. Also of note was the finding that different types of managerial experience were significant in different sectors. For service firms, general management experience was positively associated with growth, while for goods-producing firms industry experience was a more important factor. INTRODUCTION The growth prospects for new firms are daunting. Nearly 470,000 new businesses were launched in Canada between 1983 and 1986; only 95,000 were still in operation in 1993 - a mortality rate of 80% (Johnson, Baldwin & Hinchley, 1997). Of the 20% that did survive, less than one third (27%) grew by more than five employees in the first decade of operation, only one in ten grew by 15 employees or more, and only two in 10,000 grew to have more than 100 employees. Yet, it is today's young firms that, through their growth and development, enable the economy to grow and prosper. This paper addresses the important and somewhat exceptional phenomenon of young firm growth. Specifically, we seek to understand why some young firms grow more than others. 98 We draw on a unique, representative sample of young Canadian firms to evaluate the contribution of industry and firm-specific effects to firm growth, as well as how these determinants interact with the strategic choices that firms put into action. The general conceptual framework of this paper is depicted graphically in Figure 4.1. The solid lines leading to firm growth represent direct effects that have been discussed in prior research, but which have not before been evaluated on a comprehensive database of small firms. These direct relationships form the basis for a series of hypotheses that we test on our full sample of firms and within the sub-populations of goods producers and service providers. We then build upon our findings by exploring interrelationships among the dominant main effects of the model. COMPETITVE ENVIRONMENT • Dynamism • Hostility • Concentration • Market Stage Figure 4.1 Multi-level Structural Model of Young Firm Growth A FIRM STRATEGY • Product Differentiation • Service Differentiation • Innovation GROWTH FIRM RESOURCES • Human Capital • Financial Capital 99 At least three features distinguish this paper from previous inquiries into the growth of young firms. First, we consider a wide range of growth determinants, spanning industry- and firm-level effects and allowing us to evaluate the unique and relative contribution of each. Second, we use a longitudinal, representative sample of firms upon which we test our models. Data from a stratified sample of 2,962 Canadian firms, proportionally weighted to represent the full population of firms from which the sample is drawn, provides us with a rare opportunity to assess the determinants of young, firm growth. This data combines secondary source data from business tax records and national employment statistics with detailed survey data to provide a comprehensive source of information. Third, owing to the stratification of the data sample, we have equal representation of firms from the goods-producing and service-providing sectors of the economy. In Canada, the vast majority of new firms are in service industries. More than eighty percent of new business starts are in the service sectors, and this trend shows no sign of slowing down. We are able to evaluate growth determinants on the economy as a whole and within the sub-populations of the goods and services sectors. The paper is organized in five sections. In section two, following this Introduction, dominant themes within strategic management are briefly reviewed. While a great deal of strategy research has focused on large firms, a not inconsiderable body of work has recently emerged on the topic of small firm performance. This literature encompasses several fields including strategy, entrepreneurship, organizational behavior, and economics. The third section, Method, describes a large database that contains information about the industry characteristics, firm characteristics, and firm strategies of 2,962 Canadian businesses. The measures used in our study, dependent and independent, are introduced in this section. As well, we describe the dual methodologies that are employed in our analysis: weighted ordinary least squares regression and structural equation modeling. These tools are used to evaluate the 100 explanatory power of firm and industry characteristics with respect to the growth of these young firms, within the theoretical framework outlined in Section Two. Section Four presents the results of the analysis in two stages. First, the hypotheses are tested on the full data set and within the goods and services subsets. These allow us to evaluate the direct effects of the competitive environment, firm strategy, and firm resources on firm growth (the solid lines in Fig. 4.1). Subsequently, we combine these determinants in an exploratory analysis of their combined effects. A discussion of these results follows in the final section. The concluding section also addresses limitations of the paper and discusses directions for future research. THEORETICAL FOUNDATIONS The purpose of this paper is to shed light on the poorly understood phenomenon of young firm growth. To do so, we have included a broad range of determinants in our analysis. Unlike previous efforts to assert the primacy of firm or industry effects (McGahan & Porter, 1997; Rumelt, 1991; Schmalensee, 1985), we are interested in obtaining a more holistic view of the growth process. We begin by reviewing research on the phenomenon of growth itself, followed by the competitive environment, strategy, innovation, and firm resources. In each of these sub-sections, hypotheses are developed from extant research for subsequent testing with our unique, representative sample of young Canadian firms. Firm Growth It is becoming clear that venture initiation is considerably less problematic than is firm growth. Recent empirical findings strongly suggest that there is an abundance of firm starts, coupled with an equal or greater number of exits. What appears to be the principal challenge for new entrants, and one which is of considerable interest to policy-makers, is how to become 101 larger, create more employment, and increase contribution to GDP. Brander et al. (1998) found that net employment growth in Canada is due to the entry of a large number of very small firms each year. Pre-existing firms tend to have net shrinkage over time; the overall increase in employment is due to new entrants, not to the growth of survivors. A similar finding was reported by Mata (1994) based on Portuguese data. He observed that, "Although not all entrants are small, the general idea one gets from the data is that the most important challenge to market leaders may come from the post-entry penetration rather than from entry itself (1994: 27). A number of other studies have focused on the size-growth relationship in an attempt to understand the dynamics of entry, exit, and the evolution of the economy (Dunne, Roberts & Samuelson, 1989; Evans, 1987; Hall, 1987), although much of this prior work has focused only on the manufacturing sector. Both the Brander et al. (1998) paper and the present study encompass all sectors of the Canadian economy. The issue of firm growth has been investigated from a number of theoretical perspectives (Brock & Evans, 1986; Storey, 1994). They include the stochastic growth model of Simon and Bonini (1958), Lucas' (1978) human capital model, and the efficiency/learning model put forth by Jovanovic (1982). Organization theory has a tradition of research into the issue of firm survival (e.g., Carroll, 1983; Hannan & Freeman, 1984; Stinchcombe, 1965), although the level of analysis has been constrained largely to that of the population rather than the organization. Researchers have also studied the life-cycle processes within which firm expansion and contraction take place (Arbaugh & Camp, 1999). Davidsson and Wiklund (1999) distinguished between process and factor models of firm growth and proposed that theoretical perspectives can be categorized on the basis of whether they concern firm resources, motivation, strategic adaptation, or configuration. They suggest that different units of analysis are "optimal" for different theoretical investigations. Specifically, they 102 identify the firm as the appropriate unit for a resource-based analysis. Individuals are recommended for analysis in motivational studies and the governance-structure unit is suggested as appropriate for study under the configuration or strategic adaptation perspectives. Research in the motivation arena includes that of Davidsson (1991) and Cliff (1998), both of whom study the growth intentions and growth willingness of individual lead entrepreneurs. The current study spans the resource-based, strategic adaptation, and configuration approaches in an attempt to provide a comprehensive picture of young firm growth in Canada. Given the nature of the young firms under study (i.e., little, if any, separation between ownership and management), the governance structure is synonymous with the firm as the unit of analysis, in accordance with the recommendation of Davidsson and Wiklund (1999). Our integrative approach is similar to that employed by Cooper (1993), who considered four sets of variables: entrepreneurs' characteristics, founding processes, environmental conditions, and initial firm attributes. Sandberg (1986) also encompassed a wide range of determinants in his investigation of industry structure, entrepreneur attributes, and new firm strategy. These studies differ from the current work in the selection of measures, the size and scope of their samples, the duration of the period under study, and the statistical methodologies employed. Growth Measurement An important issue in the study of firm growth is that of measurement. Firms can be described in terms of their level of employment, assets, revenues, or a variety of financial ratios such as return on investment. Growth can thus be computed as the relative change in any of several state variables such as those listed above. For policy makers, employment growth is often the focal measure. For investors, ROI may be of interest. Among academics, however, there is a growing consensus that sales growth is the preferred measure of firm growth, and one that is the most common performance indicator among entrepreneurs themselves (Barkham et al., 1996; 103 Hoy, McDougall & Desouza, 1992; Davidsson & Wiklund, 1999). Utilizing sales as the underlying metric of growth is also consistent with a desire to measure growth as an output measure; revenues are the result of how firm assets (including employees) have been utilized to capture economic rents. A focus on employment or asset change would thus capture input measures, rather than the consequences of their implementation. In a review of empirical entrepreneurship literature that appeared between 1987 and 1993, Murphy, Trailer, and Hill (1996) found that growth, efficiency, and profit were the most commonly considered dimensions of new firm performance. And, of the growth indices, change in sales was the dominant measure, appearing in 66% of the cases. High performers can be clearly identified by evaluating sales growth relative to the average for a firm's industry or sector (Hoy, McDougall & Desouza, 1992; Merz, Weber & Laetz, 1994). The present study follows this convention by utilizing sales growth (net of industry growth) as the dependent variable. Another issue that pertains to the measurement and analysis of growth is that of the longitudinal nature of the phenomenon itself. By its very definition, growth is the change in a variable (in this case, revenue) over time. However, research into firm growth, including the present study, must often rely on predictor variables that are measured at a single point in time. Thus, there is the quandary of trying to understand a longitudinal process with cross-sectional determinants. Some research (e.g., Cooper, 1993) has relied on initial conditions to predict subsequent growth. The present study relies on data collected at the end of the growth period. While such a situation is clearly sub-optimal, it is neither without worth nor without precedent. For example, it is relatively common practice to use cross sectional determinants in attempts to understand firm profitability. In such cases, profits are measured at a point in time concurrent with the observation of the independent measures. However, designs of this sort seldom acknowledge that profits result from myriad operations, decisions, and actions that occurred prior 104 to the point of measurement. A firm's profits are the result of the individual and organizational activities that preceded the earning of revenues and the disbursement of expenses. As such, profit is a snapshot indicator of complex, longitudinal processes which, like firm growth, provides a relative measure of the ongoing performance of a given firm. The present study observes a number of industry, strategy, and firm-specific characteristics in an effort to try to understand firm growth. An assumption of this approach is that the factors identified by respondents as important to firm success are those which were important during the period of time in which the growth took place. Two steps have been taken to mitigate the imperfect temporal match between the independent and dependent measures of this study. First, growth was calculated as a rate over the life of each firm in the sample, rather than as an absolute measure of growth from birth year until 1995. This controls for the both the initial and final states of firm revenue as well as the total length of time that each firm was in operation. Second, the growth rates used are net of industry effects, thus accounting for cyclical differences in various industries during the period under study. It is worth noting that industry effects account for very little variance among the growth of Canada's young surviving firms. The correlation between firm growth rates and industry growth is less than 0.01. This is consistent with evidence from the US manufacturing sector in which no relationship was found between industry growth and small firm growth (Acs & Audretsch, 1990). One possible reason for this finding is that asymmetrical technological changes may have conferred different types and levels of benefits to firms of different sizes (Shepherd, 1982). Table 4.1 contains summary statistics for Canadian industry growth and small firm growth rates. 105 Table 4.1 Growth Rates by Industry Industry Industry Std. Industry Young Std. Young Growth Dev. Rank Firm Err. Firm (GDP) Growth Rank Accommodation, food & beverages 6.78 0.72 9 -0.57 3.39 25 Agriculture, fishing & trapping 1.61 1.43 22 9.61 5.5 13 Business services 12.81 1.37 2 -1.08 1.53 26 Chemicals & refined petroleum 4.79 1.37 15 13.61 2.05 10 Clothing 0.97 0.62 23 18.6 5.61 6 Communication & other utilities 6.65 0.51 10 1.32 4.77 24 Construction 4.46 0.4 17 3.15 2.07 21 Electrical and electronic products 3.1 1.21 19 20.86 3.91 4 Finance, insurance & real estate 9.6 0.66 4 -3.22 1.62 28 Food & beverages 5.59 0.53 13 15.61 3.41 8 Furniture & fixtures 2.48 1.73 20 15.34 3.38 9 Logging & forestry 7.71 0.81 7 2.54 1.96 22 Machinery 5.26 1.37 14 8.46 1.58 16 Non-metallic mineral products 0.42 1.55 24 32.9 9.45 1 Other manufacturing 17.95 25.62 1 2.09 2.14 23 Other services 8.75 0.84 5 -2.11 4.09 27 Paper & allied products -0.55 1.31 26 19.28 2.95 5 Petroleum, natural gas & mining -2.83 1.8 28 29.48 4.45 2 Primary & fabricated metal products 1.93 1.18 21 9.54 2.01 14 Printing & publishing 5.87 1.12 12 10.36 2.91 12 Retail trade 4.79 1.08 16 8.05 4.19 19 Rubber & plastics 8.52 1.52 6 9.45 2.76 15 Services incidental to mining -1.38 1.17 27 8.27 3.88 18 Textiles & leather 0.28 0.61 25 27.77 5.58 3 Transportation & storage 3.48 0.76 18 11.61 4.44 11 Transportation equipment 6.51 2.13 11 17.61 2.73 7 Wholesale trade 6.84 1.25 8 8.3 2.94 17 Wood 10.15 1.58 3 7.27 3.15 20 Competitive Environment Industrial organization economics and corporate strategy literature each emphasize the importance of industry competitive conditions on firm performance, although differences exist as to the appropriate level of analysis, the specific variables to include, and the relative contribution of firm vs. industry effects. There has been convergence, however, on a number of aspects of environmental analysis, including the relevance of market maturity, the number of competitors in the market, and the degree of competitive hostility and dynamism in an industry or market 106 segment. An exception to the view that industry matters was offered by Acs and Audretsch (1990). They found that average small firm growth and general industry growth were largely unrelated, suggesting that they may be two distinctly different phenomena. As noted above (Table 4.1), young Canadian firms do not follow the general growth trends of their respective industries. The present analysis evaluates characteristics of the competitive environment, rather than simply controlling for the industry in which a given firm competes. The dimensions of environmental dynamism and hostility were introduced by Miller and Friesen (1983). Dynamism is characterized by "the rate of change and innovation in an industry as well as the uncertainty or unpredictability of the actions of competitors and customers (Lawrence and Lorsch, 1967; Thompson, 1967; Burns and Stalker, 1961)" (1983: 222). Environmental hostility captures "the degree of threat to the firm posed by multifacetedness, vigour, and intensity of the competition and the downswings and upswings of the firm's principal industry (cf. Khandwalla, 1973; Miller and Friesen, 1978)" (1983: 222). Covin and Slevin (1989) found environmental hostility to be a significant predictor of poor performance among small firms. We anticipate that young firms, which should be possessed of flexibility and responsiveness relative to larger competitors, should be well-suited to dynamic environments, while their low initial resource endowments may make hostile environments less conducive to growth. HI: Firm growth is positively related to the level of dynamism of the competitive environment. H2: Firm growth is negatively related to the level of hostility of the competitive environment. The presence of competitors in a market is a critical factor in understanding the nature of competition (Davig, 1986). As an industry moves from a monopolistic scenario involving no 107 competitors along the spectrum to a perfectly competitive environment consisting of many identical firms, assumptions about the power of a firm and the ability to extract rents differ substantially. Competitors are one of the five forces in Porter's (1980) well-known framework for industry analysis. An increase in the number of competitors generally implies that buyer power is enhanced, rivalry is intensified, and barriers to entry may be low, all of which suggest that economic rents will be more difficult to capture and sustain. We thus anticipate a negative relationship between firm growth and the number of competitors sharing the same market. H3: Firm growth is negatively related to the number offirms with which a firm competes. Chaganti (1987) observed that prescriptions for firm growth generally counsel firms to steer clear of low growth industries, and to concentrate instead on introductory or expanding markets. He offered the alternative suggestion that small firms may be successful in low-growth environments, although different strategies will be necessary from those that would be successfully employed in a growing market. Although specific industry membership has low explanatory power with respect to firm growth (Table 4.1), we expect to see higher rates of growth among firms in introductory and growth markets than among firms competing in mature or declining markets. H4: Firm growth is negatively related to the maturity of the industry in which it competes. It is also important to consider the nature of an industry's principal output. The vast majority of young Canadian firms are in service-based industries (84% of our sample frame). This was not always the case. Between 1951 and 1981 the share of employment represented by service industries increased from 48% to 68%. And, since the turn of the century, services have grown from contributing 30% of GDP to 70% (McCharles, 1990). Numerous explanations have been 108 offered to account for this phenomenon, including lower entry barriers to service industries, high demand for services, and the relative competitive advantage of nations (Caves et al., 1980). McCharles (1990) observed that the strongest growth has been in those services that support, in one way or another, the production of goods and services, or "producers' services." The two other major groups in the service sectors are consumers' and government services. The data used in the present study includes both producers' and consumers' services, but not government and education. It seems reasonable to anticipate that, due to differences in the nature of inputs and outputs, there will be differences between the determinants of growth for goods and services firms. In the analysis that follows, we test all models and hypotheses on the full sample of young Canadian firms and we also evaluate each set of variables within the goods and services sub-populations to allow cross comparison for similarities and differences. Competitive Strategy Differentiation Strategies It is a fundamental assumption of strategic management research that the strategies that firms employ in the market determine, to some extent, their eventual success or failure. Speaking directly to the issue of strategy in a small-firm context, Cooper, Willard and Woo (1986) noted that small firms have typically been encouraged to specialize their outputs and concentrate on factors such as quality and service which are not consistent with the economies of scale of larger competitors. Young firms, in addition to their cost disadvantages due to scale, are seldom as far along the learning curve as are their more established competitors. Therefore, young firms should seek ways to differentiate themselves in order to garner economic rents (Carter et al., 1994). However, there are many ways in which a firm may differentiate itself from its competitors. Carter et al. (1994) identified six dimensions of competitive strategies for new ventures. Other than price, which poses high risks as discussed above, two themes that emerge from the remaining factors are an emphasis on either (1) meeting customer needs through quality or high market sensitivity or (2) providing a greater selection of products or services. Gertz and Baptista (1995) reported that growing companies often pursue either "customer franchise management" or a "new products/services development strategy." The former indicates that firms "focus selectively on better-chosen customers, know everything they can know about those customers and their needs, and serve those needs with intense dedication" (1995: 4). Alternately, the latter strategy places emphasis on "rapidly developing large numbers of new products that offer superior value to customers" (1995: 4). These classification schemes correspond to the results of principal components analysis of the strategies of the young firms in our sample (described below). Whether a firm chooses to emphasize the core product/service offering, or instead to differentiate itself on the basis of supporting characteristics such as customer service, is an essential strategic choice. While each strategy may be successful for a given firm in a given environment, there may be systematic differences based upon whether a firm is essentially a goods producer or a service provider. Recall that the Canadian economy has shifted to predominantly service-based firms during the past half century. Among new entrants, services firms outnumber goods producers four to one. At issue, is how these firms can carve out a market niche. For service firms, quality and customer service may present greater opportunities for building a loyal customer base than would depth or breadth of product range. These techniques, however, may be eminently suited to the aims of goods producers who can build a complementary line of products as they endeavor to obtain market share. Yet, while service firms n o may be constrained to service-centric differentiation strategies, goods producers may be able to successfully build clientele through the implementation of both product- and service-centric strategies. This reasoning leads to the following hypotheses: H5a: Product-centric differentiation strategies are positively related to firm growth in goods-producing industries. H5b: Product-centric differentiation strategies are negatively related to firm growth in service-based industries. H6a: Service-centric differentiation strategies are positively related to firm growth in goods-producing industries. H6b: Service-centric differentiation strategies are positively related to firm growth in service-based industries. Innovation Innovation is arguably the mostly highly regarded and frequently suggested prescription for firms with growth aspirations. Wolfe (1994) reported that in a five-year span, there were 6,244 published journal articles and 1,336 dissertations on the general topic of innovation. A recent search of the ABI Inform database of business abstracts yielded 12,512 hits on the subject word "innovation." Though the range of subjects addressed in these thousands of papers is far beyond the scope of the current paper, it is instructive to note the degree of interest that has been dedicated to the topic in recent years. Empirical support for a positive relationship between innovation and firm performance is extensive, although there are indications that the relationship between innovation and performance may be moderated by the industry in which a firm competes (Acs & Audretsch, 1987; Banbury & Mitchell, 1995; Malerba & Orsenigo, 1996). Innovative output is also markedly higher among small enterprises. According to the US Small Business Administration, i l l new small firms produce 24 times more innovation per research dollar than do those on the Fortune 500 (Keats & Bracker, 1988). In their study of US SMEs, Lefebvre and Lefebvre (1993) reported that firms with a strong competitive position exhibited the strongest efforts to be innovative. Thus, the relationship between growth and innovation may be a virtuous cycle of positive reinforcement. Innovation has also been cited as a means by which small firms can offset competitive advantages that size and experience have conferred upon their larger rivals (Acs & Audretsch, 1990; Brock, 1981; Scherer, 1980). We expect to find a positive relationship between the innovative activity of young Canadian firms and their rates of growth. H7: Firm strategies featuring innovation are positively related to firm growth. F i r m Resources Three characteristics of firms that have been extensively evaluated as main effects of research, or control variables in related empirical studies, are the age of a firm, the initial size of a firm, and the experience of a firm's manager(s). Firm age is generally held to increase the probability of a firm's continued survival (Jovanovic, 1982; Stinchcombe, 1965). The liability of newness argument posits that, with increasing age, firms gain resources and knowledge that enhance their prospects for continued survival and prosperity. However, with increasing age (and size), the rate of growth may decrease (Evans, 1987; Mata, 1994). Our study focuses on firms that are similar in age, ranging from seven to ten years. There is, however, somewhat greater heterogeneity of initial size and managerial experience in our sample. Initial size has been associated with firm growth, although conflicting results have been reported by Cooper and his colleagues. In a study of two-year old start-ups, ventures which were smaller at birth (measured by number of employees) exhibited greater rates of sales and employment growth (Cooper, Woo & Dunkelberg 1989). However, in a study of new ventures, 112 tracked over a three-year period, larger stocks of initial resources were found to be positively associated with firm performance as measured by employment growth (Cooper, Gimeno-Gascon, & Woo, 1994). Given that growth represents change from an initial resource base (Cooper et al., 1989), it is reasonable to anticipate that firms that are smaller at start-up will exhibit greater relative growth. H8: Firm growth is negatively related to initial size. Mixed results have also been reported with respect to the issue of managerial experience. Two recent studies have established a positive relationship between managerial experience and new venture performance (Cooper et al., 1994; Gimeno, Folta, Cooper & Woo, 1997). As well, both Aislabie (1992) and Acs and Audretsch (1990) reported positive relationships between levels of human capital (measured as experience) and growth of small firms. Contrasting results were reported by Acar (1993), who did not find firm age or owner experience to be related to size or sales performance. One component of the liability of newness is associated with insufficient knowledge of business and industry conditions. Thus, we anticipate that managerial experience will be an asset to young firms, first in fostering their survival and then in facilitating growth. H9: Firm growth is positively related to managerial experience. Configuration Strategy research has long recognized the relevance of configuration or fit - the match of organizational attributes, such as structure and strategy, to environmental characteristics7. The identification of optimal configurations, and a theoretical framework for replicating configurations as environments change, would be of considerable value to managers and academics alike. 7 See Venkatraman and Camillus (1984) or Ward, Bickford and Leong (1996) for a review. 113 Miller and Friesen (1983) tackled the related, but different issue of how strategy-making processes within organizations relate to characteristics of the environment. They found that innovation was beneficial in dynamic environments, while a more analytical approach was appropriate in the face of environmental hostility. The results were ambiguous with regard to the linkage between hostility and innovation. Miller (1987) investigated the relationships among the strategies of large firms and a series of structural and environmental variables. Data from Canadian, Australian, and US firms confirmed his model, which predicts the structural and environmental correlates of a strategy based on the number and degree of uncertainty of the available contingencies. Powell (1992) studied two US industries, wood upholstered furniture and women's dresses, and found support for his thesis that some organizational alignments constitute a source of competitive advantage. Venkatraman and Prescott (1990), in their examination of PIMS data, concurred that finding an appropriate match between environment and strategy has systematic performance implications. Venkatraman (1990) focused on the strategic business units of US corporations as the unit of analysis. He found support for the existence of a strategic coalignment construct and for the presence of a positive and significant effect of coalignment on SBU performance. Naman and Slevin (1993) collected data from 82 manufacturing firms that encompassed environmental turbulence, entrepreneurial style, organization structure, and mission strategy. They found that financial performance was positively related to the measurement of fit. In sum, prior studies of this important issue have spanned a wide range of issues, and explored them empirically on a diverse collection of data sets, but a comprehensive picture of small firm coalignment does not number among them. This paper will proceed by first analyzing, separately, environmental, strategic, and firm resource determinants of growth. The hypotheses enumerated above will be tested and the results discussed. Subsequently, more complex models featuring interactions among the main effects 114 will be derived and evaluated, thus providing a more comprehensive picture of the interrelationships that exist within the population of Canada's emerging enterprises. The first step, then, will be an attempt to confirm the predictions derived from prior research, while step two will attempt to build new knowledge inductively from interrelationships within the data. METHOD Sample Data for this paper was drawn from the Survey of Operating and Financing Practices (SOFP), a survey instrument which we developed in conjunction with the Micro-Economic Analysis Division of Statistics Canada (Johnson et al., 1997). The Longitudinal Employment Analysis Program (LEAP) (Statistics Canada, 1988) provided the frame for our research sample. From 1983-1986, inclusive, there were 469,114 new business starts in Canada. Of these, 95,302 were still operating in 1993, representing an average survival rate of 20.3% across the four birth-year cohorts. Firm-level information was available in the LEAP file for 39,675 of these survivors. These are the firms that form the survey frame. A representative sample of 3,991 firms was drawn from the longitudinal data base. Of this initial target sample, 301 were out of scope (i.e., no longer part of the population), reducing the functional sample size to 3,690. Usable responses were obtained from 2,962 firms, representing an overall response rate of 80.5%. Missing values were imputed to generate a complete data matrix (Johnson et al., 1997). The sample was stratified on the basis of firm size, firm growth, financial structure, and by industry group. Firm size was measured by Average Labor Units (ALUs) in 1993 as a proxy for number of employees8. Firm growth was measured 8 ALUs are calculated by dividing a firm's annual payroll by the average wage earnings per employee in the same industry-province group. Industry-province groups are defined at the SIC 3-digit level. 115 by the change in ALUs from birth year to 19939. Financial structure was measured by debt/asset ratio in 1993. Finally, four industry groups were constructed on the basis of whether a firm was primarily a goods-producer or a service-provider and whether the firm's industry was one with a relatively high or low knowledge base. Industry knowledge base was classified on the basis of indices specific to the goods and services sub-populations. The goods-producer index was composed of a multi-factor productivity score (Statistics Canada, 1996), the proportion of workers with post-secondary education, the percentage of sales devoted to R&D in the industry, the percentage of firms in the industry using high technology (Lee and Has, 1996), and an innovation index (Robson, Townsend & Pavitt, 1988). The service index was based on three factors: GDP per hour worked in 1992, proportion of workers with post-secondary education, and industry average wage rates. Principal components analysis was used to assign scores to industries at the two-digit SIC level. The industries were then classified as high or low knowledge on the basis of whether they were above or below the median factor score. A list of the industries in each classification is presented in Table 4.2. The data was collected using a stratified sampling design under which approximately equal numbers of firms were sampled from within each strata. Consequently, it is necessary to apply proportional sampling weights when performing statistical analysis. For example, low-knowledge service firms represent 45% of the sample frame but comprise only 27% of the actual survey sample. On the other hand, high-knowledge goods producers, which make up less than 4% of the population, were over-sampled to represent 23% of the sampled firms. The strata proportions and cell counts for the frame and the sample are presented in Table 4.3. 9 This stratification variable thus reflects the change in employees - an input measure. We will be testing a number of independent variables for their influence on firm growth as measured by revenue - an output measure. In any case, the strata merely determine the proportion of the sample that is represented by firms of a given characteristic, and weighting procedures, described below, account for the sampling design. 116 Table 4.2 SOFP Industry Classifications HIGH-KNOWLEDGE INDUSTRIES LOW-KNOWLEDGE INDUSTRIES GOODS Chemical and chemical products Agriculture Crude petroleum and natural gas Beverages Electrical and electronic products Clothing Fabricated metal products Fishing & trapping Machinery Food Mining Furniture & fixtures Non-metallic mineral products Leather Paper & allied products Logging & forestry Plastics Other manufacturing Primary metals Printing & publishing Refined petroleum & coal products Quarry & sand pits Rubber Services incidental to mining Transportation equipment Textiles Tobacco Wood SERVICES Business services Accommodation, food & beverages Communication Amusement & recreational services Construction Other services Finance, insurance & real estate Retail trade Other utilities Storage Pipeline transportation Transportation Wholesale trade Source: Johnson, Baldwin & Hinchley (1997) Two methods of analysis are employed in this paper: weighted ordinary least squares (OLS) regression and structural equation modeling. Each are described briefly below. Weighted OLS Regression Analysis Due to the stratified sample design under which the data was collected, it is important to incorporate sampling weights and strata parameters in the regression analysis (Sarndal, Swennsson & Wretman, 1992). When one employs a simple random sampling (SRS) procedure, any observation has an equal chance of being included in the sample. Under stratified random sampling, each stratum represents a sub-population within which any observation has an equal chance of being selected. However, across the entire data set, the probability of inclusion varies as a function of the proportion of observations within a stratum that are sampled. Consequently, Table 4.3 Population and Respondent Counts by Strata Population Percent Respondent Percent of Percent of Count Count Sample Pop. Cell1 Total 39675 100 2962 100 7.47 INDUSTRY Goods, high knowledge 1386 3.49 679 22.92 48.99 Goods, low knowledge 4853 12.23 733 24.75 15.10 Services, high knowledge 15544 39.18 754 25.46 4.85 Services, low knowledge 17892 45.10 796 26.87 4.45 FIRM SIZE (ALUs) 0-9 31569 79.57 1121 37.85 3.55 10-24 5618 14.16 1056 35.65 18.80 25 + 2488 6.27 785 26.50 31.55 FIRM GROWTH (ALUs) Declined 8164 20.58 723 24.41 8.86 0-4 20837 52.52 769 25.96 3.69 5-14 7214 18.18 755 25.49 10.47 15 + 3460 8.72 715 24.14 20.66 DEBT/ASSET RATIO Low(0- 19 th percentile) 7899 19.91 737 24.88 9.33 Medium-low (20-59m percentile) 15974 40.26 790 26.67 4.95 Medium-high (60-79m percentile) 7902 19.92 791 26.70 10.01 High (80- 100th percentile) 7900 19.91 644 21.74 8.15 1. Calculated as proportion of strata cell in sample (e.g., Goods, high knowledge = 48.99% = 679/1386) Source: Johnson, Baldwin & Hinchley (1997) ignoring sampling design parameters can have the effect of introducing significant bias into point estimates and reported standard errors. The analyses reported below utilize the weighting and stratification parameters in the OLS regression procedure. In essence, this simulates the results that would be obtained on the representative sample frame of more than 30,000 firms from which the sample was drawn. The Appendix (Table A4.1) contains data on the extent of misspecification of variance estimates that would occur if the assumption of simple random sampling were utilized on this data set. The "misspecification effect" is the ratio of the complex-design-based variance estimate to the variance that would be estimated under the assumption of SRS data collection (Scott & Holt, 1982; Skinner, 1989). The average magnitude of the 118 misspecification effect across the 19 independent variables used in this study is 7.2 for the full data set, 3.4 for the subset of firms in goods-producing industries, and 5.4 for service firms. The misspecification effect is reduced as smaller subsets of the data are parsed out for examination. The misspecification effect for goods firms is less that than for service firms because a greater proportion of goods producers in the sample frame was captured within that stratum. Structural Equation Modeling In addition to the weighted OLS regression analysis described above, structural equation models are used to evaluate latent effects within the data and interactions among the variables. Again, the sampling design is employed in the analysis. This is accomplished by creating weighted variance-covariance matrices from the raw data. The weighted matrices then serve as the computational base for the structural equation models. Structural equation modeling has been used with increasing frequency in organizational behavior and marketing (Bagozzi & Yi, 1988; Gatignon & Xeureb, 1997), but it has appeared only occasionally in the entrepreneurship and strategy literature. Cool, Dierickx, and Jemison (1989) utilized the technique to investigate the relationships among risk-return outcomes, market share, firm conduct attributes, and inter-firm rivalry. Drawing on data from the Indiana commercial banking industry, they were able to identify indirect and direct components of total effects and thus disentangle some of the ambiguity surrounding risk-return outcomes. In their work, they employed Wold's (1980, 1982) method of partial least squares (PLS). The PLS analytical technique was also featured by Davidsson (1991) in his evaluation of the effects of growth motivation on small firm growth. While the focus of Davidsson's research is similar to that of the current paper (small firm growth), his predictor variables (ability, need, and opportunity) and dependent measure (growth over a three-year period) differ substantially from this research. His findings indicated that ability, need, and opportunity explain 25% of the 119 variation in observed growth among young Swedish firms. A structural equation framework incorporating personality traits, ability, and motivation of entrepreneurs was also proposed by Herron and Robinson (1993), although it has yet to be tested with empirical data. Keely and Roure (1990) utilized independent variables similar to those in the present study in their investigation of how industry structure, business strategy, and manager characteristics affected the performance of 36 new, technology-based ventures. Unlike the current study, however, they drew on initial decisions of the start-up team (as documented in business plans) to predict performance (ROI) measured several years later. They concluded that industry structure and firm strategy had greater predictive power than did the measures of managers' characteristics (e.g., prior experience). Venkatraman (1990) incorporated aspects of marketing, manufacturing, and administration as latent dimensions of a construct labeled "strategic coalignment." He found support for the existence of the coalignment construct and a positive relationship between coalignment and performance. This study employs the SEPATH module of STATISTICA statistical software (Steiger, 1995). The analysis protocol performs 5 iterations using the Generalized Least Squares estimation procedure and then switches to Wishart Maximum Likelihood Estimation until convergence is achieved or deemed infeasible. The criterion for convergence is met when the residual cosine does not exceed 0.0001 for successive iterations. All figures depicting structural models conform to SEPATH conventions (see Appendix, Table A4.2). Measures Growth The dependent variable in our models is growth in firm revenue, net of industry growth. This measure was obtained by calculating compound annual growth rates for each firm from 120 birth-year-plus-one until 1993. We used birth-year-plus-one as the base for our growth calculations because we were unable to determine the number of months a firm had been in operation in its initial year. Thus, we use the subsequent year as the first full year of operation. Revenue data were obtained from the business tax records (T2) for each firm. Industry growth rates were calculated based on GDP at factor cost for the industry in the same period for which a given firm was in existence (Statistics Canada CANSIM Data Base). For example, in the transportation equipment industry, the 1983-1993 growth rate was 10.6%, while the same industry grew only 4.8% in the 1985-93 period. Thus, firms that were born in 1983 in the transportation equipment industry have their growth rates adjusted downward by 10.6% while a 4.8% net calculation is applied to the cohort of firms born in that industry in 1985. This method of growth rate adjustment was selected due to the potential lack of comparability that could result if firms from many different industries were pooled together. An alternative tool for industry control would be to introduce separate dummy variables for each industry into the regression models. However, such an approach would make already complex models even more cumbersome, especially in the structural equation analysis. By adjusting firm growth rates on the basis of GDP growth for their given industry-cohort group, we retain the relative growth ordering of firms within a given industry-cohort, while adjusting for the relative growth rates between industry-cohort groups. Firms that grew in excess of industry average will remain in the data as high growth firms, while those that were sub-par performers will likewise remain at the lower end of the distribution. We controlled for outliers in the population by calculating standardized net growth rates for the firms in our sample. Standardized rates with an absolute value greater than 3.0 were classified as outliers. This criterion caused us to omit 62 firms from the sample. Four had growth rates less than - 68% and 54 had rates exceeding 87%. Means, standard deviations, and inter-item 121 correlation coefficients for the remaining 2,899 firms in the sample are presented in Table 4.4 (see p. 123). For exploratory purposes, we also divided the sample into thirds based upon net revenue growth. The weighted mean scores and standard errors for the high, medium, and low growth segments are presented in Table 4.5. Table 4.5 Weighted Means by High, Medium, and Low Growth Categories All Firms High Growth Medium Growth Low Growth Mean Std. Err. Mean Std. Err. Mean Std. Err. Mean Std. 1 NETNCPSG 4.61 1.10 28.37 1.86 3.87 0.40 -10.75 0.43 2 FIRM_EXP 3.70 0.04 3.60 0.08 3.74 0.04 3.72 0.09 3 INDEXP 3.81 0.04 3.74 0.06 3.87 0.02 3.81 0.09 4 M G R E X P 3.73 0.04 3.77 0.05 3.79 0.05 3.66 0.09 5 LNASBRTH 4.41 0.08 4.39 0.15 4.25 0.16 4.56 0.10 6 PRE) CHNG 2.70 0.08 2.75 0.15 2.67 0.15 2.69 0.12 7 TEC CHNG 3.35 0.08 3.26 0.17 3.42 0.13 3.34 0.12 8EQP CHNG 3.83 0.08 3.82 0.14 3.59 0.15 4.04 0.12 9 PRE) SUBS 3.73 0.08 3.70 0.12 3.74 0.14 3.75 0.14 10 SUP_SUBS 3.64 0.08 3.63 0.13 3.57 0.15 3.70 0.14 11 NEWCOMP 3.79 0.08 3.78 0.14 3.68 0.14 3.88 0.13 12 INNOVATE 0.22 0.03 0.30 0.05 0.25 0.05 0.15 0.04 13 FLEXIBLE 4.22 0.06 4.18 0.15 4.33 0.07 4.16 0.11 14 QUALITY 4.48 0.05 4.50 0.11 4.66 0.06 4.32 0.09 15 CST_SVCE 4.42 0.08 4.45 0.13 4.37 0.18 4.45 0.10 16 CUSTOM 2.46 0.12 2.45 0.20 2.89 0.19 2.13 0.20 17 RANGE 2.65 0.11 2.62 0.22 3.01 0.18 2.36 0.18 18 NEW PROD 2.24 0.11 2.42 0.21 2.63 0.18 1.80 0.18 19 MKT_STAGE 0.32 0.03 0.37 0.05 0.36 0.06 0.25 0.05 20 # COMP 3.39 0.08 3.62 0.11 3.30 0.10 3.32 0.15 21. DYNAMIC 0.00 0.07 -0.02 0.10 -0.09 0.11 0.08 0.12 22. HOSTILE 0.00 0.06 -0.01 0.11 -0.03 0.11 0.03 0.10 23. PRDSTRT 0.00 0.05 0.03 0.11 0.24 0.09 -0.22 0.08 24. SRV STRT 0.00 0.05 -0.00 0.12 0.05 0.08 -0.04 0.08 25.NET EXP 0.00 0.06 -0.08 0.10 0.09 0.05 -0.03 0.12 Growth categories based on sample split at into thirds where lowest third net growth < -2.94%, medium net growth -2.93% < M< 11.90%) and high net growth > 11.92% 122 u a © CU S. S-© u T3 S — H « tzi « S a s V J3 O CN o o o o O — NO O (N O o o — </3 o o o o — o oo O cn —' d o T t r~ O 00 NO o o o — d o o T t T t — o o — d d — o — NO r-; oo T t m rn o o o ON o NO CN ON o m o d d d o o o in T t o in o CN in o d d d d o o p CN o CN CN ON m CN CN O o r-ON o d d d d d o o o NO T t T t T t T t CN CN in CN in NO CN in r~-o o r-o —• d d d d d d o o © C— m m CN r-m m CN in CN CN T t CN ON CN O T t CN O —• d d d d d d d • o o p CN O NO T t o in o o 00 T t m in CN T t O d • d d d d d d d o o p NO o o NO NO o in T t o m ON o T t o o o m CN in o in in o T t o —• d d d d d d d d d O o © <N m <N o o CN O o ON ON o t -T t O in r-o NO o oo NO O o oo NO o —' d d 1 d d d d d d d d o o © T t o T t ON m CN o NO CN O 00 o m m o oo o o o 00 o CN CN O m o o o o —• d d d d d d d i d d d d o o © oo m CN T t 00 <N in m NO m o (N o r-(N o in o o 00 T t o 00 o o o r-o o o CN rn o —• d d d d • d • d • d • d d d d d O o © m CN CN T t •n CN T t OO o CN o o o in NO o T t NO ON 00 o CN T t ON o CN m —• d d d d d d • d d d d d d d • o © p ON m <n m cN m r-m o o CN (N 00 o ON NO o CN T t O NO O T t T t o oo ON © CN CN © m NO o —• d d d d d d d • d • d d d d d i d • o o p o o 00 o o T t o ON <N O m o in o NO o o in o r-T t r-o NO NO o m o in o CN r-o -• d • d d • d d • d • d d • d • d • d i d i d i d i d i o o p T t T t o CN o m in o 00 oo o m T t NO o in o o m o in NO o NO T t O CN in o in T t o m T t O CN O m CN o in T t o *•" d • d • d d i d d d d d • d d d d d d d r-CN NO CN r-o CN o T t O 00 m o CO in m o T t o ON o o 00 m O NO O NO T t o CN in o T t in o NO o 00 m o ON in o d d d • d p d d d • d d • d d d d • d • d d T t T f m NO O NO NO o o m o NO O CN o ON m o •n CN © NO o p ON CN O ON m o m CN o CN CN O m o T t CN o o NO o d d 1 d d d d d d d d • d • d • d d d • d • d ON m o CN 00 CN O in in o o ON CN O T t o o o ON m o NO o CN O o T f T t o o o d d i d d • d • d • d i d • d d d d • d d d d d T t o 00 o 00 o 00 o 00 o 00 o 00 o 00 o m o NO o in o 00 o <N - m o oo o d d d d d d d d d d d d d d d d d m T f o t - ; >n rn m 00 m t - ; T t NO ON CN (N CN CN 00 T t CN T t NO T t in NO T f CN CN m ON m T t CN rn rn rn rn rn d T t T t T t CN CN CN d rn IP X w £5 Z E — C N & fi « S. -'I 111 m *3* u-> o o a 0- H a OH S o w Z CO >< LU o > c<o < a o Tt in o H C/3 D U NO Q S u Z w o < H H 1 s ON o d v a. c C3 'c 00 o vo cG CD CN O m " II § K J3 ° N <u 00 H CN O 3 m CN Determinants of Growth The 19 independent variables in our data set were obtained from the survey, with two exceptions: initial size of the firm at birth and industry membership. Size of the firm in its first full year of operation was measured as the natural log of the firm's assets as determined from the firm's T2 record. The broad sector in which the firm competes (goods vs. services) was determined from the LEAP file and was a determinant of the stratification used for data collection. For the analysis of environmental effects, six items were drawn from a survey question that asked respondents to agree or disagree (5-point Likert-type scale anchored by "Agree" and "Disagree") with a series of statements describing the characteristics of the industry in which they operated. They included the rate of product obsolescence, the rate of change of production technology, the rate of obsolescence of production equipment, the availability of substitute products, the availability of alternate suppliers for competitors, and the threat of arrival of new competitors. Principal components analysis of these six items yielded two factors with eigenvalues greater than one (see Appendix, Table A4.3). The first three items load heavily on a factor that we have labeled environmental Dynamism, while the latter three represent the degree of environmental Hostility. These dimensions of the competitive environment are consistent with the definitions for dynamism and hostility described in the previous section (Miller & Friesen, 1983). The grouping of items identified by the principal components analysis is utilized in the environmental effects structural model, in which Dynamism and Hostility are included as latent variables. We also utilized the derived factor scores for weighted regression analysis. Two other items were tested as environmental determinants of young firm growth. One assessed the number of competitors a firm faced in its primary market (0; 1-4; 5-19; 20-99; 100+). The other, Product Market Stage, classified firms according to whether their primary 124 product market is in the early (introductory or growth) or late (mature or declining) stages of its life cycle. Respondents were asked to rate six items according to the importance of each item to their firm's competitive strategy (5-point Likert-type scale anchored by "High Importance" and "Low Importance"). The items were: flexibility in responding to customers' needs, quality, customer service, product customization, offering a wide range of related products, and frequently introducing new/improved products. Principal components analysis was again used to identify latent dimension of strategy (see Appendix, Table A4.4). Firms with a service-centric strategy placed greater emphasis on the first three strategic elements, while those with a product-centric strategy tended to load more heavily on the latter three items. These are consistent with the dimensions identified and discussed in the preceding section (Carter et al., 1994; Gertz & Baptista, 1995). One additional measure of strategy was a dichotomous measure of innovation. The innovation measure was coded according to whether a firm had introduced any innovations in the 1992-1994 period (data was collected in 1995). Innovations were defined in the survey as new or improved products/services or processes, excluding esthetic changes that did not change technical construction or product/service performance. The final set of predictor variables pertains to firm resources - human and real capital. Real capital is operationalized as the log of assets at start-up. Three other variables tap different dimensions of managerial experience: experience with the firm, experience in the industry, and total experience as a manager. In each case, the response is a categorical variable (0-2 years; 3-5; 6-9; 10+). These three measures are combined in the structural models within the latent variable labeled Managerial Experience (see Appendix, Table A4.5). R E S U L T S The analyses proceeded in two stages, and are reported accordingly. The first step involved testing the hypotheses enumerated above. Following a review of these findings, we engaged in a second, exploratory stage designed to probe the interactions among the main effects that were tested in stage one. The results are discussed briefly as they are presented. A more comprehensive discussion appears in the following section. Hypothesis Tests The Competitive Environment The first four hypotheses pertain to the effects of the competitive environment on the growth of young firms. Table 4.6 contains the results of the weighted OLS regression analysis. The structural model of only the environmental effects is presented in Figure 4.2, which conforms to SEPATH conventions. The results of the structural model analysis are shown in Table 4.7. Note that the paths among variables listed in Table 4.7 correspond to the path Table 4.6 Weighted OLS Regression of Environmental Effects All Firms All Firms All Firms All Firms Goods Services P R D C H N G T E C C H N G E Q P C H N G P R D S U B S SUP_SUBS NEW_COMP Intercept M K T _ S T A G E #COMP 4.288' 4.636* 4.289* 4.631* 6.302 1.598' 1.510* 1.661* 1.581* 1.501 0.011 1.043 -1.091 -1.118 -1.065 0.171 -0.190 -0.053 1.270 3.912* 1.449 1.536 -1.376 -0.843 -1.240 0.350 1.213 -0.341 0.046 1.77* -0.523 -0.485 2.431* -0.114 -0.086 -0.318 -0.017 -0.052 -2.631* R-Squared F (d.f.) -2.176 -0.245 0.046 1.043 1.831 0.024 0.030 0.026 0.032 0.089 4.23* 2.34* 1.88* 1.50 0.98 (2,2886) (5,2883) (5,2 883) (8,2880) (8,1351) (8, 1532) tp<0.10 *p<0.05 ** p<0.01 *** p<0.001 126 Figure 4.2 Schematic Diagram of Environmental Effects Model c G > H Product ) * Obsolescence 10 82 53 17 12 13 14 54 85 Technological Change Rapid Depreciation Availability of Substitutes Supplier Avail to Competitors Threat of New 1 r Entrants J 8 Market Stage 16 Table 4.7 Structural Equation Model of Environmental Effects All Firms (n=2899) Goods (n=1359) Services (n=1540) Parameter Estimate Parameter Estimate Parameter Estimate (DYNAMIC)- 1->[PRD CHNG] 0 777*** 0.575*** 0.842*** (DYNAMIC)-2->[TEC CHNG] 0.753*** 0.898*** 0.715*** (DYNAMIC)-3->[EQP CHNG] 0.468*** 0.427*** 0.468*** (HOSTILE)-4->[PRD_SUBS] 0.900*** 0.431*** 1.016*** (HOSTILE)-5->[SUP SUBS] 0.735*** 0.822*** 0.685*** (HOSTILE)-6->[NEW_COMP] 0717*** 0.910*** 0.659*** [Bl]-7->[NETNCPSG] 1.657*** 1.950*** 1.604*** [STAGEDUM]-8->[NETNCPSG] 4.315*** 6.432*** 3.668*** (DYNAMIC)-18->[NETNCPSG] 0.047 -0.063 0.412 (HOSTILE)- 19->[NETNCPSG] -0.880* -2.668*** -0.856 ML Chi-Square 699.3*** 469.0*** 480.4*** tp<0.10 *p<0.05 **p<0.01 p<0.001 127 identifiers in Figure 4.2. For simplicity of presentation, the variance pathways identified in the schematic diagram are not included in the table of results. More detailed output from the structural equation analyses, including variance estimates may be found in the Appendix (Tables A4.6 - A4.8). These conventions are also observed for the strategic and firm resource effects, as well as for the combined models presented below. The OLS regression provides support for Hypothesis 4's prediction that higher growth is associated with firms in introductory or early stage markets, rather than mature or declining markets. This is further supported by the strongly significant coefficient in the structural equation model (Path 8).10 Hypothesis 3, which predicts decreasing firm growth as the number of competitors increases, was not supported by the data. The results indicate that there is a marginally significant positive relationship between the increasing presence of competitors and the growth rates of the young firms in our sample. One plausible explanation for such a phenomenon would be that more competitors are present in early stage markets - a common occurrence prior to the consolidation of an industry and the shakeout of weak firms. However, inspection of Table 4.4 reveals a very low correlation (r = 0.002) between the number of competitors and the variable for market stage. A large number of competitors may also indicate an attractive industry, irrespective of the market's product life cycle stage. Such a situation would be consistent with the observed positive relationship between firm growth and competitor density. Neither of the hypotheses regarding environmental dynamism (HI) or hostility (H2) were supported by the analysis, although the OLS regression of the individual variables provided some 1 0 The structural equation technique is based on the covariance structure, it does not capture the stratification effects in a way that is comparable to the weighted OLS analysis. Since the omission of sample design effects can cause variances to be underestimated by a factor of seven on average, the significance levels of the structural model coefficients should be interpreted with caution. 128 interesting insights into differences between the goods-producing and service-providing firms in our sample. The structural analysis first served as a confirmatory factor analysis for the dimensions of environmental dynamism and hostility (Paths 1-6). The latent variable, Dynamism, was not significant as a determinant of growth (Path 18). The latent variable, Hostility, was significant at the p < 0.05 level (Path 19), however given the underestimation of standard errors due to omission of stratification, this is at best a tentative finding. In order to verify this, factor variables were created for Hostility and Dynamism and models were tested in a weighted OLS regression that incorporated the stratification design. These results (Table 4.8) indicated that while Hostility was consistently negative, it lacked statistical significance. T a b l e 4.8 W e i g h t e d O L S R e g r e s s i o n o f D e r i v e d P r i n c i p a l C o m p o n e n t s F a c t o r s All Firms All Firms All Firms All Firms Goods Services M K T S T A G E 4.312* 3.098 3.132 2.723 #COMP 1.653* 1.222 0.775 1.046 DYNAMIC -0.669 -0.491 0.731 -0.257 HOSTILE -0.034 -0.635 -1.946 -0.577 INNOVATE 8.384* 7.825* 11.51** 5.773 PRD STRAT 0.568 0.552 -1.305 1.179 SRV_STRAT 0.288 0.051 -2.006 1.574 LNASBRTH -1.391** -1.217* -2.389** -1.461* EXPERIENCE -0.965 -1.062 -1.457 -0.968 Intercept -2.369 2.731** 10.74*** 3.094 14.76** 4.048 R-Squared 0.025 0.040 0.015 0.070 0.142 0.071 F 2.290* 3.310* 3.860* 3.150*** 3 go,*** 3.160**** (d.f.) (4, 2884) (3, 2885) (2, 2884) (9, 2877) (9, 1348) (9, 1529) tp<0.10 *p<0.05 **p<0.01 ***p<0.001 An additional test was employed to evaluate the characteristics of the high and low growth firms in our sample, absent the medium growth firms in the "belly" of the distribution. A dummy variable was created for firms in the top and bottom thirds of the sample, by net growth (Table 4.5). Weighted logistic regression was then performed on the specified high and low 129 growers (Table 4.9). The logistic regressions essentially confirmed the weak relationships between market stage and number of competitors, and reinforced the non-significance of the dynamism and hostility factors. Table 4.9 Weighted Logistic Regression of Derived Principal Components Factors All Firms All Firms All Firms All Firms Goods Services M K T _ S T A G E 0.59' 0.47 -0.27 0.53 #COMP 0.21' 0.18 0.42* 0.13 DYNAMIC -0.15 -0.16 0.09 -0.15 HOSTILE -0.02 -0.09 -0.14 -0.08 INNOVATE 0.81* 0.72' 1.60*** -0.44 PRD STRAT 0.21 0.22 0.12 0.31' SRV_STRAT 0.04 0.03 -0.02 0.29' LNASBRTH -0.08 -0.08 -0.12 -0.15 EXPERIENCE -0.05 -0.04 -0.40** -0.03 Intercept -1.29** -0.56** -0.04 -0.95 -0.35 -0.64 F 1.74 2.66* 0.51 1.39 2 35*** 1.53 (d.f.) (4, 1919) (3, 1920) (2, 1921) (9, 1914) (9, 877) (9, 1037) tp<0.10 *p<0.05 **p<0.01 *** pO.001 Logit Coding Based on Top and Bottom Thirds of Sample Net Growth. High and Low Growth categories based on sample split at into thirds where lowest third net growth < -2.94% and high net growth > 11.92% Although the main hypotheses were not supported, the items on the competitive environment serve to highlight some intriguing differences between the goods and service sectors. Among goods producers, an environment characterized by rapid product changes and a high threat of new entrants was detrimental to growth, while these characteristics signaled positive growth prospects for service firms. In a similar vein, growth was positively associated with the availability of substitutes among goods producers, while service providers exhibited lower growth in such an environment. These differences can also be seen in the factor variables. While dynamism had low significance in the full sample, it carried a positive sign when run among goods-producers and a negative sign among service firms. Taken together, these results indicate that an industry-level analysis that does not account for the specific competitive 130 characteristics facing a firm may miss important elements of the growth process. However, it is also important to recognize the poor explanatory power of environmental effects with respect to the growth of young firms in Canada. This serves to reinforce Acs & Audretsch's (1990) conclusion that general industry trends may be less influential on small firms than is the case for larger enterprises. Competitive Strategy Our hypothesis that innovative firms would exhibit higher growth rates was supported by the data. However, while innovation was strongly positive and significant for goods firms and the aggregate sample, it was not significant at the p < 0.05 level when run within the sub-population of service firms. The strength of the coefficient for innovation is worthy of note. In both the weighted OLS regression and the structural model (Path 27), firms that had introduced an innovation in the 3 years prior to data collection had compound net growth rates eight times greater than their non-innovative counterparts. It is also noteworthy that only one in five firms were innovators (22%). From Table 4.5 it can be seen that among the high growth third of firms, Table 4.10 Weighted OLS Regression of Strategic Effects All Firms All Firms All Firms All Firms Goods Services INNOVATE FLEXIBLE QUALITY CST_SVCE C U S T O M R A N G E 8.646** 8.525** 8.401* 8.348* 11.328** 0.156 0.104 1.111 1.255 1.130 0.585 -0.688 -0.668 -3.103* 6.350 0.609 1.084 0.482 -0.899 -0.147 1.392 -8.079 0.054 1.37 NEW_PROD -0.667 -0.662 0.567 -0.259 -0.305 -0.226 1.260 1.217 -0.496 Intercept R-Squared F (d.f.) 2.673** -0.540 2.228 -0.102 12.874* 0.039 0.043 0.050 0.053 0.121 6.62* 2.01f 2.84* 1.79* 1.87* (1,2887) (4,2884) (4,2884) (7,2881) (7,1351) (7, 1532) tp<0.10 *p<0.05 **p<0.01 ***p<0.001 131 Figure 4.3 Schematic Diagram of Strategic Effects Model 29 30 -51 ^ V 5 l 2 J?13. 37 32 33 34 Sj4_ -515 S l 6 _ Flexibility to ) r Customers Quality Customer Service Offering Wide Product Range SERVICE S T R A T E G Y ^ Innovation 27 GROWTH Si Product ) * Customization Introducing New Products PRODUCT ""N STRATEGY ) Table 4.11 Structural Equation Model of Firm Strategy Effects All Firms (n=2899) Goods (n=1359) Services (n= 1540) Parameter Estimate Parameter Estimate Parameter Estimate (SRVSTRAT)-21 ->[FLEXIBLE] 0.675*** 1.268*** 0 4 7 1 * * * (SRVSTRAT)-22->[QUALITY] 0.648*** 0 7 3 4 * * * 0.689*** (SRVSTRAT)-23->[CST_SVCE] 0.896*** ] 411*** 0.702*** (PRDSTRAT)-24->[CUSTOM] 1.256*** 1.601*** 1.185*** (PRDSTRAT)-25 -> [RANGE] \ 4 3 9 * * * 1.600*** 1 3 9 7 * * * (PRDSTRAT)-26->[NEW PROD] 1 5 3 7 * * * 1.679*** 1.507*** [INNOVATE]-27->[NETNCPSG] 8.274*** 11.68*** 5.996*** (SRVSTRAT)-38->[NETNCPSG] 0.091 -3.262*** 1.960*** (PRDSTRAT)-39->[NETNCPSG] 0.892* 0.272 1.168* M L Chi-Square 688.4*** 690.4*** 320.4*** tpO.10 *p<0.05 **p<0.01 pO.001 132 30% had introduced recent innovations, as compared with 25% of the medium-growth firms and only 15% of the low-growers. The two dimensions of strategy identified in this research, product-centric and service-centric, were not significant determinants of young firm growth. Thus, hypotheses 5 and 6 were not supported. As was the case with the environmental effects, however, an examination of the individual survey items illuminates differences in competitive practice between goods producers and service providers. For example, in both sectors an emphasis on quality and flexibility were positively (though not significantly) associated with growth. Yet in goods industries, an emphasis on customer service and frequent new product introductions had negative growth coefficients, quite the opposite of what can be observed among the service industries. Conversely, the coefficient for product customization was positive for goods-producers and negative for service providers. Firm Resources The final two hypotheses concern the financial and human capital endowments of young firms. As predicted by Hypothesis 8, initial asset levels were inversely related to firm growth T a b l e 4.12 W e i g h t e d O L S R e g r e s s i o n o f F i r m R e s o u r c e s All Firms All Firms All Firms Goods Services LNASBRTH -1.345* -1.436** -1.582* FIRM_EXP -2.067 -2.157 -6.027** IND_EXP -3.129 -3.330* 3.574* M G R E X P 3.744*** 3.763*** 0.455 Intercept 10.54*** 10.20* 17.56** 25.50** R-Squared 0.012 0.020 0.034 0.035 F 6.27* 7.48*** 6.95*** 2.97* (d.f.) (1,2885) (3,2885) (4,2885) (4,1353) -1.846*** -1.810 -4.388* 4.422*** 18.36** 0.048 6.28*** (4, 1534) tp<0.10 * p<0.05 **p<0.01 ***p<0.001 133 Figure 4.4 Schematic Diagram of Firm Resources Model 49 50 51 0 ) fe Years with the firm §22. § 2 3 . Years in the industry ) fe Years as a 1 * manager 57 Table 4.13 Structural Equation Model of Firm Resource Effects All Firms (n=2899) Goods (n=1359) Services (n=l540) Parameter Estimate Parameter Estimate Parameter Estimate (NETEXP)-41->[FIRM_EXP] 0.495*** (NETEXP)-42->[IND_EXP] 0.568*** (NETEXP)-43->[MGR_EXP] 0.491*** [LNASBRTH]-47->[NETNCPSG] -1.403 * * * (NETEXP)-58->[NETNCPSG] -1.282*** M L Chi-Square 69.75*** 0.639*** 0.611*** 0.282*** -1.552*** -1.689** 27 95*** 0.472*** 0.552*** 0.541*** -1.804*** -1.035* 47 07*** tpO.10 *p<0.05 **p<0.01 p<0.001 rates. Although this may be merely an artifact of the calculation of growth from an initial base level, it is nonetheless instructive to confirm the finding, which further reinforces the merit of including initial size as a control variable in studies of firm growth. The findings regarding prior experience are mixed. The three survey items, general managerial experience, industry experience, and firm experience were designed to capture dimensions of human capital that progressively decreased in generality and increased in specificity. Firm-level experience implies more specific knowledge than industry-level 134 experience, which is in turn more specific than managerial experience of any kind. The most general of these, experience as a manager, was positively and significantly associated with the growth of service firms, but its relationship with the growth of goods-producers was negligible. Industry experience was positively associated with growth in the goods industries and negatively associated with growth in the service sector. Finally, firm-specific experience negatively associated with the growth of both types of firms, although this finding was significant only among goods-producers. The net result of the latent variable for prior experience in the structural model (Path 58) was negative. In sum, Hypothesis 9 was not supported, yet the nature of the findings paints an intriguing picture. From the data, one may infer that the most valuable experience for managers in the goods-producing sector is that which is derived from working in a particular industry. However, for service firms with growth aspirations, general management skills may be of greatest value. Managerial skills may be more mobile across industry boundaries within the general category of service firms, whereas knowledge-specificity may limit the mobility of managers in goods-producing industries. A manager of a financial services firm may be able to readily adapt to a new role in telecommunications or retail, while a manager in an electronics firm may have greater difficulty re-establishing him or herself in mining or textiles, for example. Summary of Hypothesis Tests The variables for environmental, strategic, and firm resources effects were combined in a series of nested models for weighted OLS regression analysis (Table 4.14). From these results, and those presented above, a number of general observations can be made prior to the commencement of the exploratory analysis. For convenience, the results of the hypothesis tests have also been summarized in Table 4.15. 135 oo C N — o r-i >2 wo co ro T I -ro wo in K „ X ° •* tN — o T T wo (N — • _• ^ CN vq r - oo h i d d d d d ffl m - -CN ro * m « o m ^ . vo — o • Ov in co ( N *"vo * It «^ — Ov m _^ \ ! in ov — M= • d d <N ov CQ ro' CQ tN o o ro co vo •>t ^ co oo cN d T i —« C N d +- +-r- o 00 o> <n ro ro © — O O O CQ CQ CQ tN vo VO b -- o\ ^ " oo vo — wo vo ov <3v tN tN v o - ~ - ^ t N 0 0 CO co —: — ' T i i d d I— * T l - tN tN in „ m « » J „ O Ov o v ° ° — O O f N O O — oo • co — in • • CQ vd — CQ in vo T | -T f co wo ro ro _ -7 d d • d 1 d I CQ •*- — 00 ro ro vo vo S tN 00 tN Co r?, rv1 JC! — ^ ^ - d -N q q d — in 00 vo TL 0 0 i n wo °° o — t N wo £ ! „ — r-~ — Ov — • • vo ro ro CQ ro d T 1 CQ D a w o < H 1/3 ffl 5 S I o o O D W (- W 0. c/3 Z Ov ro T f VD VD O d d vo - § VD OV 2 o r- ^ d d 9 1^ o 0 fN vo in d 9 9 t-~ o * OV t N O r - T I - C2 d d ^ 1 1 »—-ov T l - OV VD ro * * to ro —• 00 I to in vo C N Jo — °°. 00 CN i ^ , ov O M 6 m C * +• I — -rr r - * i~- S o < N — r-— 00 T i — o  — ro Ov tN —* vo Ov" co Ov O tN w „ _ . n tN O (N 00 ?2 ^ £ 000 S o — — CN OV CO 00 00 ov wo • • ro • • o d d T 9 - V 9 T T CN ro ro r-~ in — TJ-— tN — OV d d ro OV 00 T r 00 ov "3" —' CN t— ro OV °0 "* in tN (N CN _; O O CN VD CN 00 VO tN r~ ro T T T t Ov ro ro in o o • — — * I t N ro Ov Ov I «"> Tt- vo Ov Tn —.. vq vq q o — T T wo 0 n C vo 00 Ov r - J vo r - ov vo ' n. •* o< — -ro — tN O TT rv, o r- T I - J 00 00 o wo 00 5> CN ^ <*>. ro °. - -1 ro — O ro C-O —• * _ ro ro — C f£ in Tt- tN vo — CN ro CN i i 1 T r — VO CN O tN O ro w-i * _ TT 00 co T ) -00 r~ r-Ov q ^ — wi t N ro O •— O O to ov * 4 -CN r - — * O VO 00 <—> <1 ^ 00 — tN IN • S ov ov rr °. °°. m. {2 o d d 9 9 -o VD tN T l - OV T f 00 00 — tN P- T l -. . . ^  0 - 9 — tN O 1 — — O T t d *" +" T h f N OO 00 * ro o v o o J J c M r s r 2 ? 5 — — T f wo • ro • • • ro T f - d T d 9 9 9 K - I - — ^ 0 S S 0 CN ro § £ ^ d d 9 9 9 vo t N . O ro t N — O ro w Ov _ —., O ov VD wo .j. S! ^ ov VO — ov 00 0 — b 0 * S Jr. 00 tN CN • O Ov 1/0 ^ 0 - ^ S 3 - — vo r o o o 0 V a. o V a. wo q o V a. o V Table 4.15 Summary of Hypotheses and Results All Firms Goods Services HI: Firm growth is positively related to the level of dynamism of the competitive environment. Not Supported Not Supported Not Supported H2: Firm growth is negatively related to the level of hostility of the competitive environment. Not Supported Not Supported Pro1 Subs [+] NewjOomp [-] Not Supported H3: Firm growth is negatively related to the number of firms with which a firm competes. Not Supported Not Supported Not Supported H4: Firm growth is negatively related to the maturity of the industry in which it competes. Marginal Support Not Supported Not Supported H5a: Product-centric differentiation strategies are positively related to firm growth in goods-producing industries. Not Supported Not Supported Not Supported H5b: Product-centric differentiation strategies are negatively related to firm growth in service-based industries. Not Supported Not Supported Not Supported New_Prod [+] H6a: Service-centric differentiation strategies are positively related to firm growth in goods-producing industries. Not Supported Not Supported Cst_Svce [-] Not Supported H6b: Service-centric differentiation strategies are positively related to firm growth in service-based industries. Not Supported Not Supported Not Supported H7: Firm strategies featuring innovation are positively related to firm growth. Supported Supported Not Supported H8: Firm growth is negatively related to initial size. Supported Supported Supported H9: Firm growth is positively related to managerial experience. Not Supported Partial Support Firm Exp [-] Ind_Exp [+] Partial Support Mgr_Exp [+] Ind Exp [-] 137 Among the environmental effects only one determinant, market stage, had predictive power with respect to firm growth. Neither hostility nor dynamism was significantly associated with growth, and the number of competitors had a weak positive relationship, contrary to our expectations. The weighted OLS regression model of all environmental effects had an R 2 of only 0.032 for all firms (Table 4.6) and the F-statistic for the environmental effects was not significant. The strategic effects model (Table 4.10) had greater explanatory power than did the model of environmental effects, due largely to the strongly positive and significant influence of innovation as a predictor of firm growth. However, even innovation became non-significant when run within the sub-population of firms in service industries. This result underscores the differences that seem to exist between the growth determinants of firms in goods-producing industries and those in the service sectors. While one should not conclude that growth is unrelated to firm strategy, our findings indicate that the differentiation schemes of product- and service-centricity were poor predictors of the growth rates of young survivors. The firm resource models were generally significant, as indicated by the F-statistics (Table 4.12), although they also had poor explanatory power as measured by their R 2 results. Nevertheless, a significant relationship between initial size and growth was established, and sector-specific results emerged with respect to prior experience. In the firm resources models, the explanatory power was greater within the sub-population of service firms, unlike the environmental and strategic effects models in which the explanatory power was considerably greater for firms in goods-producing industries. Exploratory Analysis From the results presented above, we elected to develop a structural model in which firm resources have direct effects on growth as well as indirect effects via their influence on firm 138 strategy. Dynamism and hostility are not included in this model due to their general lack of explanatory power, although market stage and number of competitors are retained in the analysis for control purposes. As such, only their direct effects on growth are specified in the model. A schematic diagram of the combined firm resources and firm strategy model is presented in Figure 4.5. F i g u r e 4.5 S i m p l i f i e d S c h e m a t i c D i a g r a m o f F i r m R e s o u r c e s a n d S t r a t e g i c E f f e c t s M o d e l Initial trials of the structural model failed to achieve convergence for all specified paths, although a number of relationships proved to be robust across a wide range of initial conditions and search algorithms. Most notably, the direct effects of number of competitors, market stage, innovation, and initial assets on growth were remarkably consistent, both within runs of the 139 model depicted in Figure 4.5 and in comparison to the simpler models presented above. However, it is the indirect relationships in Figure 4.5 that are of the most interest, and it is those which bring the most new information to the analysis. Specifically, we were interested in how initial assets and net managerial experience influenced the three strategies of innovation, product-focus, and service-focus. In order to achieve robust findings for these intermediate paths, it was necessary to specify fixed effects for the pathways from the latent variables for product- and service-strategy to net growth (Paths 38 and 39). The values were fixed at the levels determined from the structural analysis for strategic effects alone (see Figure 4.3 and Table 4.11). The results of this analysis are presented in Table 4.16. As an additional step, we also ran a series of weighted OLS regressions in which innovation, product strategy, and service strategy were each specified as dependent variables, with initial assets and prior experience as predictors. The results, for the full sample and within the goods and services subsets, are presented in Table 4.17. Service-centric strategies were negatively related to initial assets for the full sample, and within both the goods and services sub-populations. This relationship was observable in both the structural equation results (Table 4.16) and the weighted OLS regressions (Table 4.17). Among goods-producing firms there was also a negative relationship between prior managerial experience and service differentiation indicated in the structural equation findings, although this result was not significant in the OLS runs. Product-centric strategic choice was negatively associated with both prior experience and initial asset levels among firms in goods-producing industries, although these relationships did not hold for service firms. From these results, one may infer that firms that start smaller, and survive, may be more likely to specify distinct product or service-based niche strategies, relative to firms that entered their industries with larger initial endowments of assets. 140 T a b l e 4.16 S t r u c t u r a l E q u a t i o n M o d e l o f F i r m S t r a t e g y a n d R e s o u r c e s All Firms Goods Services Parameter Parameter Parameter Estimate Estimate Estimate [Bl]-7->[NETNCPSG] 1.225*** 1.428*** 1.102** [STAGEDUM]-8->[NETNCPSG] 3.064*** 2.131* 2.804** (SRVSTRAT)-21 ->[FLEXIBLE] 0.483*** 0.248*** 0.050 (SRVSTRAT)-22->[QUALITY] 0.562*** 0.112*** 0.654 (SRVSTRAT)-23->[CST SVCE] 0.369*** 0.158*** 0.253 (PRDSTRAT)-24->[CUSTOM] 0.179*** 0 777*** -0.109 (PRDSTRAT)-25->[RANGE] 0.126** 0.785*** -0.005 (PRDSTRAT)-26->[NEW PROD] 0.228*** 0.665*** -0.270 [INNOVATE]-27->[NETNCPSG] 7.907*** 11 70*** 6.313*** (NETEXP)-41->[FIRM EXP] 0.493*** 0.664*** 0.468*** (NETEXP)-42->[lND EXP] 0.572*** 0.576*** 0.559*** (NETEXP)-43->[MGR EXP] 0.487*** 0.281*** 0.534*** [LNASBRTH]-47->[NETNCPSG] -0.878*** -4.965*** -1.479 [LNASBRTH]-61 ->(SRVSTRAT) -0 177*** -1.146*** -0.127 [LNASBRTH]-62->[INNOVATE] 0.002 0.024** -0.008 [LNASBRTH]-63->(PRDSTRAT) -0.377*** -0.326*** 0.165 (NETEXP)-64->(SRVSTRAT) 0.013 -1.201*** 0.114 (NETEXP)-65->[INNOVATE] 0.004 -0.039** 0.015 (NETEXP)-66->(PRDSTRAT) -0.686*** -0.575*** 0.460 (SRVSTRAT)-(O.l }->[NETNCPSG] 0.100 -3.300 2.000 (PRDSTRAT)-{0.9}->[NETNCPSG] 0.900 0.300 1.200 (NETEXP)-58->[NETNCPSG] -0.794* -4.478*** -2.031*** ML Chi-Square 4919.2*** 4090.5*** 2427.0*** tp<0.10 * p<0.05 **p<0.01 ***p<0.001 The strength of the relationships between net experience and innovation and between initial assets and innovation were quite small for goods-producers and negligible in the case of service firms. These results failed to achieve statistical significance in the weighted OLS analyses. The coefficients for the goods industries, though small, were unsurprising in sign -innovation was negatively related to prior experience and positively associated with initial asset levels. Thus, at least in the case of goods-producers, firms that are initially small may exhibit more net growth, but better-endowed firms may be better positioned to pursue growth through innovation. Prior experience appears to be detrimental to the pursuit of innovation or clear product-focused or service-focused differentiation strategies. 141 Table 4.17 Weighted OLS Regression of Firm Resources on Firm Strategies Innovate All Firms Goods Services Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Initial Assets 0.002 0.014 0.024 0.024 -0.008 0.015 Prior Experience 0.008 0.016 -0.038 0.026 0.016 0.017 Constant 0.215** 0.072 0.197 0.098* 0.239** 0.078 F-Stat 0.12 2.68* 0.57 Product Strategy All Firms Goods Services Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Initial Assets -0.025 0.039 -0.118" * 0.052 -0.004 0.044 Prior Experience -0.041 0.042 -0.174** 0.052 -0.018 0.050 Constant 0.111 0.195 0.498* 0.237 0.034 0.217 F-Stat 0.66 6.11** 0.06 Service Strategy All Firms Goods Services Coeff. Std. Err. Coeff. Std. Err. Coeff. Std. Err. Initial Assets -0.091** 0.032 -0.186" 0.078 -0.054' 0.031 Prior Experience 0.007 0.047 -0.037 0.064 0.014 0.056 Constant 0.401** 0.140 0.557* 0.275 0.303* 0.145 F-Stat 4.04* 2.92f 1.54 tp<0.10 *p<0.05 ** p<0.01 *** pO.001 DISCUSSION The research question introduced at the beginning of this paper was: "Why do some young firms grow more than others?" In our efforts to understand the phenomenon of young firm growth, we have focused on three classes of determinants - industry/environmental, competitive strategy, and firm resources. We have drawn on a large, representative sample of Canada's seven to ten year old enterprises, and employed multiple methodologies in our empirical analysis of the data. One important feature of our stratified data set is the comparable representation of firms from goods and services industries, thus enabling us to evaluate the determinants of growth within these broad sectors, rather than simply controlling for goods or services industry membership through the use of dummy variables. We utilized the data set to evaluate a series of 142 hypotheses pertaining to the growth of Canada's young firm population. The hypotheses, and the results of our analyses, are summarized in Table 4.15. Our examination of industry/environment effects brought to light a number of interesting findings. First, we were able to confirm, through principal components analysis, the constructs of environmental dynamism and hostility (Miller & Friesen, 1983). However, while the constructs themselves were evident in our data, they were found to be poor predictors of firm growth. Our hypothesis that growth would be diminished in the presence of an increasing number of competitors was also not supported by the data. There was marginal support for our expectation that growth would be positively associated with early stages of an industry's life cycle. In sum, the industry and environmental determinants used in our models explained very little of the variance in young firm growth rates. Recall that the growth rates of young firms, sorted by industry membership, do not follow the general pattern of industry growth as measured across the entire Canadian economy (Table 4.1). It may be that industry characteristics, such as entry barriers and levels of competitive rivalry, are better predictors of entry and exit than they are predictors of growth. Since growth over time is a characteristic unique only to survivors, there may be factors beyond industry membership that play an important role in determining the growth trajectories of young, surviving businesses. The cohort of firms that survives for at least one decade represents only 20% of the start-ups that enter the economy in a given year. The inclusion of other predictor variables may improve the explanatory power of our models, and our understanding of firm growth. For example, measures of industry concentration (e.g., Herfindahl index) are currently available for only a limited number of manufacturing industries (Statistics Canada, 1995). The development of concentration measures for the full range of industries under study (especially the oft-neglected service industries) may enable us to 143 deepen the scope of our inquiry. Other measures for future study could include an industry's knowledge base of competition or the average level of investment required to initiate a new venture. Among the strategic effects examined in this research, none was as effective a predictor of young firm growth as was innovation. Firms that had introduced innovations in the three years prior to data collection clearly outperformed the majority of firms that were non-innovators. This finding may relate to the observed lack of concurrence between industry-wide and young firm growth trends. If, as suggested by Shepherd (1982), there are discontinuities to scale in the application of emerging technologies, then the young firms that take advantage of innovative technologies or business models may be those best suited to carve out a new niche in an otherwise complacent industry. And, since new firms generally lack the resources of incumbents, the ability to do something differently or to introduce a new product or service may be essential to both survival and growth. We were also able to confirm the constructs of product-centric and service-centric competitive orientations, although neither of these measures proved to be significant predictors of young firm growth. The signs of the regression coefficients, while not statistically significant, are worthy of mention. Both tactics had negative coefficients within the sub-population of goods industries and positive coefficients within service industries. This could indicate that the nature of a specific product offering matters more in goods-producing industries, while the positioning and support of a firm's principal offering may bear a stronger relationship to the growth prospects of service firms. An additional observation on product-centric strategy can be made from an inspection of the mean scores of the sample as split into high, medium, and low growth subsets (Table 4.5). There appears to be an inverted U-shaped relationship between product-144 focus and growth, indicating that such product-centric strategies may be appropriate for moderate growth ambitions, if not for high growth. Absent the strong findings for innovation, the strategic effects tested in this study were quite poor at explaining young firm growth. The strategic effects models could undoubtedly benefit from the inclusion of additional predictor variables. For example, information about the marketing tactics and distribution channels used by young firms may shed more light on their relative performance over time. It would be premature to conclude that growth is unrelated to strategy. An innovation-based strategic orientation is clearly important. However, it does not appear that the dimensions of product- and service-based differentiation are effective predictors of the growth of young Canadian firms. The issue of survivor bias is germane to the discussion of initial size and growth. The firms that were initially the smallest in our sample were those that exhibited the greatest net revenue growth. However, within the population of initial entrants, the inefficient small firms would likely have exited from the population before reaching the minimum seven-year age of the young firms in our data set. Thus, we may be capturing a biased subset of initially small, highly efficient survivors. This issue is further complicated by the fact that the small start-ups have small denominators in the growth equation. So, for a given absolute increase in annual revenue, smaller firms will exhibit high rates of growth relative to their initial starting position and relative to firms that were larger at the outset. While the meaning of this result is somewhat cloudy, its significance should serve as a reminder of the importance of controlling for initial size in firm growth research. The other firm resource determinant in our models was prior managerial experience. Although the composite construct of net experience was not statistically significant, specific types of prior experience were important within the goods and services industry subsets. For 145 goods-producers, growth was positively associated with prior industry experience and negatively associated with firm-level experience. The former finding is intuitively sensible, while the latter is somewhat puzzling. A tentative explanation may be that the rapidly growing firms had recently brought new management on board to help cope with the challenges of expansion. In this case the presumed causality between growth and firm-specific experience would be reversed, viz., experience or the lack thereof would be a function of growth. Among service firms, a very different picture emerged from the data. For this subset, growth was negatively associated with prior industry experience and positively associated with general experience as a manager. These findings paint a picture in which the skills for successfully managing service firms are general in nature and transferable across industry boundaries. The finding that different types of experience appear to be differentially related to growth in different industry sectors underscores the importance of distinguishing between goods firms and service firms in our analysis. In Canada, young service firms outnumber young goods-producers four to one. Yet, national data (e.g., Industry Canada monthly summaries) is available in much greater detail for manufacturing firms than is the case for firms in the service sectors. This may stem from the post-WWII era during which manufacturing played a more dominant role in the Canadian economy. The goods producers in our sample have higher average growth rates (Avg 10.72, S.E. 2.20) than do the service firms (Avg. 3.48, S.E. 1.24), but the level of economic activity (business starts, level of employment) is much greater in the service sectors. Extensions of this research will attempt to draw out additional distinctions between these two very different sectors of the Canadian economy. The exploratory phase of this paper featured a structural equation model designed to capture the combined effects of firm resources and competitive strategy. Overall, the results of 146 this analysis were disappointing. The fit of the model was poor, and the relationships between resources and strategy that we had hoped to identify proved to be largely insignificant. Neither prior experience nor initial asset levels were associated with whether a firm was an innovator. Nor was prior experience related to competitive strategy. The one tentative finding to emerge from our analysis was the negative relationship between initial size and service-centric strategy. This may be indicative of a higher degree of focused differentiation among the firms that start small and manage to survive. Recall that these small survivors were also the firms that exhibited substantial net growth. The lack of a relationship between firm resources and innovation may be welcome news to nascent entrepreneurs. While innovation may be critical to survival and growth, it appears that innovative activity is not a function of initial firm size. Clearly, much remains to be learned about the determinants of young firm growth in Canada. The overall explanatory power of our models was quite low, due to a number of potential causes. One concern, noted previously, is the temporal mismatch between the cross sectional survey data and the longitudinal measure of performance. This characteristic of the data limits the scope of the study and restricts the degree to which inference can be drawn from the findings. As well, young firms are a heterogeneous population, and there may be limits to the amount of variance that can be explained across a wide range of business models and competitive conditions. Future research into this topic would benefit from the inclusion of several measures that are absent from the present study. In addition to specific determinants noted above, subsequent work could incorporate measures of growth intentions or motivation, and financial issues such as the extent to which capital availability constrains growth ambitions. Controlling for the survivorship bias would also be of benefit, although it is not clear at this time how such controls might be implemented. 147 In conclusion, we have determined that the surviving firms that start small are the firms that grow the most. We have found that, while industry and environmental conditions are poor predictors of firm growth, there are differences between what matters in the goods and services sectors. Specifically, general knowledge is associated with growth in service industries, while industry-specific knowledge appears to be more critical to goods-producers. And, we have been able to confirm that innovative activity is much more prevalent among the young high-growth firms in Canada than it is among the low growers. 148 REFERENCES Acar, A . C. (1993). The impact of key internal factors on firm performance: An empirical study of small Turkish firms. Journal of Small Business Management 31(4), 86-92. Acs, Z. J. & Audretsch, D. B. (1987). Innovation, market structure, and firm size. Review of Economics & Statistics. 69(4). 197-210. Acs, Z. J. & Audretsch, D. B. (1990). The determinants of small-firm growth in US manufacturing. Applied Economics. 22(2). 143-153. Aislabie, C. (1992). Sudden change in a model of small firm growth. Small Business Economics, 4(4), 307-314. Arbaugh, J. B. & Camp, S. M . (1999). Managing growth transitions: Theoretical perspectives and research directions. Florida: Presented at the State of the Art in Entrepreneurship Conference. Bagozzi, R. P. & Youjae Y i . (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16(1), 74-94. Banbury, C. M . & Mitchell, W. (1995). The effect of introducing important incremental innovations on market share and business survival. Strategic Management Journal, 16(Special Issue), 161-182. Barkham, R., Gudgin, G., Hart, M . & Hanvey, E. (1996). The Determinants of Small Firm Growth. Gateshead, Tyne and Wear, U K : Athanaeum Press. Brander, J., Hendricks, K., Amit, R. & Whistler, D. (1998). The engine of growth hypothesis: On the relationship between firm size and employment growth. Working Paper: The University of British Columbia. Brock, G. W. (1981). The Telecommunications Industry. Cambridge, M A : Harvard University Press. Brock, W. A. & Evans, D. S. (1986). The Economics of Small Businesses. New York: Holmes & Meier. Burns, T. & Stalker, G. M . (1961). The Management of Innovation. London: Tavistock. Carroll, G. R. (1983). A stochastic model of organizational mortality: Review and reanalysis. Social Science Research, 12(4). 303-329. Carter, N . M . , Stearns, T. M . , Reynolds, P. D. & Miller, B. (1994). New venture strategies: Theory development with an empirical base. Strategic Management Journal, 15(1), 21-41. Caves, R. E., Porter, M . E., Spence, A. M . & Scott, J. T. (1980). Competition in the Open Economy. Cambridge, M A : Harvard University Press. 149 Chaganti, R. (1987). Small business strategies in different industry growth environments. Journal of Small Business Management, 25(3). 61-68. Cliff, J. E. (1998). Does one size fit all? Exploring the relationship between attitudes towards growth, gender, and business size. Journal of Business Venturing. 13(6). 521-542. Cool, K., Dierickx, I. & Jemison, D. (1989). Business strategy, market structure and risk-return relationships: A structural approach. Strategic Management Journal. 10(6). 507-522. Cooper, A. C. (1993). Challenges in predicting new firm performance. Journal of Business Venturing. 8(3). 241-253. Cooper, A. C , Gimeno-Gascon, F. J. & Woo, C. Y. (1994). Initial human and financial capital as predictors of new venture performance. Journal of Business Venturing. 9(5). 371-395. Cooper, A. C , Willard, G. E. & Woo, C. Y. (1986). Strategies of high-performing new and small firms: a reexamination of the niche concept. Journal of Business Venturing. 1(3). 247-260. Cooper, A. C , Woo, C. Y. & Dunkelberg, W. C. (1989). Entrepreneurship and the initial size of firms. Journal of Business Venturing. 4(5). 317-332. Covin, J. G. & Slevin, D. P. (1989). Strategic management of small firms in hostile and benign environments. Strategic Management Journal. 10(1). 75-87. Davidsson, P. & Wiklund, J. (1999). Conceptual and empirical challenges in the study of firm growth. Florida: Presented at the State of the Art in Entrepreneurship Conference. Davidsson, P. (1991). Continued entrepreneurship: Ability, need, and opportunity as determinants of small firm growth. Journal of Business Venturing, 6(6). 405-429. Davig, W. (1986). Business strategies in smaller manufacturing firms. Journal of Small Business Management. (January), 38-46. Dunne, T., Roberts, M. J. & Samuelson, L. (1989). The growth and failure of US manufacturing plants. Quarterly Journal of Economics. 104(4), 671-698. Evans, D. S. (1987). The relationship between firm growth, size, and age: Estimates for 100 manufacturing industries. Journal of Industrial Economics. 35(4). 567-581. Gatignon, H. & Xuereb, J.-M. (1997). Strategic orientation of the firm new product performance. Journal of Marketing Research. 34(1). 77-90. Gertz, D. L. & Baptista, J. P. A. (1995). Grow to be Great. New York: The Free Press. Gimeno, J., Folta, T. B., Cooper, A. C. & Woo, C. Y. (1997). Survival of the fittest? Entrepreneurial human capital and the persistence of underperforming firms. Administrative Science Quarterly. 42(4). 750-783. 150 Hall, B. H. (1987). The relationship between firm size and firm growth in the US manufacturing sector. Journal of Industrial Economics. 35(June), 538-605. Hannan, M. T. & Freeman, J. (1984). Structural inertia and organizational change. American Sociological Review. 49(1), 149-164. Herron, L. & Robinson, R. B. J. (1993). A structural model of the effects of entrepreneurial characteristics on venture performance. Journal of Business Venturing. 8(3). 281-294. Hoy, F., McDougall, P. P. & Desouza, D. E. (1992). Strategies and environments of high growth firms. In D.L. Sexton & J.D. Kasarda (Eds.), The State of the Art in Entrepreneurship (pp. 341-357). Boston: PWS-Kent Publishing. Johnson, J., Baldwin, J. & Hinchley, C. (1997). Successful Entrants: Creating the Capacity for Survival and Growth (Catalogue No. 61-524-XPE). Ottawa: Minister of Industry. Jovanovic, B. (1982). Selection and the evolution of industry. Econometrica, 50(3), 649-670. Keats, B. W. & Bracker, J. S. (1988). Toward a theory of small firm performance: A conceptual model. American Journal of Small Business, 12(Summer), 41-58. Keeley, R. H. & Roure, J. B. (1990). Management, strategy, and industry structure as influences on the success of new firms: A structural model. Management Science, 36(10), 1256-1267. Khandwalla, P. N. (1973). Effect of competition on the structure of top management control. Academy of Management Journal. 16(2), 285-310. Lawrence, P. & Lorsch, J. (1967). Organization and Environment. Boston: Harvard University. Lee, F. C. & Has, H. (1996). A quantitative assessment of high-knowledge versus low-knowledge industries. In Peter Howlett (ed.), The Implications of Knowledge-Based Growth for Micro-Economic Policies. Calgary: University of Calgary Press. Lefebvre, L A. & Lefebvre, E. 1993. Competitive positioning and innovative efforts in SMEs. Small Business Economics, 5(4), 297-305. Lucas, R. (1978). On the size distribution of business firms. The Bell Journal of Economics, 9(Autumn), 508-523. Malerba, F. & Orsenigo, L. (1996). Schumpeterian patterns of innovation are technology-specific. Research Policy. 25(3), 451-478. Mata, J. (1994). Firm growth during infancy. Small Business Economics, 6(1), 27-39. McCharles, D.C. 1990. Knowledge Production in Manufacturing: Its Impact on Business Services. Vancouver: The Fraser Institute. McGahan, A. M. & Porter, M. E. (1997). How much does industry matter, really? Strategic Management Journal, 18(Special Summer Issue), 15-30. 151 Merz, G. R., Weber, P. B. & Laetz, V. B. (1994). Linking small business management with entrepreneurial growth. Journal of Small Business Management. 32(4). 48-60. Miller, D. (1987). The structural and environmental correlates of business strategy. Strategic Management Journal. 8(1). 55-76. Miller, D. & Friesen, P. H. (1978). Archetypes of strategy formulation. Management Science. 24(9), 921-933. Miller, D. & Friesen, P. H. (1982). Innovation in conservative and entrepreneurial firms: Two models of strategic momentum. Strategic Management Journal. 3(1). 1-25. Miller, D. & Friesen, P. H. (1983). Strategy making and environment: The third link. Strategic Management Journal, 4(3). 221-235. Murphy, G. B., Trailer, J. W. & Hill, R. C. (1996). Measuring performance in entrepreneurship research. Journal of Business Research. 36(1). 15-23. Naman, J. L. & Slevin, D. P. (1993). Entrepreneurship and the concept of fit: A model and empirical tests. Strategic Management Journal. 14(2). 137-153. Porter, M. E. (1980). Competitive strategy. New York: Free Press. Porter, M. E. (1991). Towards a dynamic theory of strategy. Strategic Management Journal, 12(1), 95-117. Powell, T. C. (1992). Organizational alignment as competitive advantage. Strategic Management Journal. 13(2). 119-134. Robson, M., Townsend, J. & Pavitt, K. (1988). Sectoral patterns of production and use of innovations in the UK. Research Policy. 7(1). 1-14. Rumelt, R. P. (1984). Towards a strategic theory of the firm. In R.Lamb (Ed.) Competitive Strategic Management (pp. 556-570). Englewood Cliffs, NJ: Prentice-Hall. Sandberg, W. R. (1986). New Venture Performance. Lexington, MA: D.C. Heath and Company. Sarndal, C. E., Swennsson, B. & Wretman, J. (1992). Model Assisted Survey Sampling. New York: Springer. Scherer, F. M. (1980). Industrial Market Structure and Economic Performance. Chicago: Rand-McNally. Schmalensee, R. (1985). Do markets differ much? American Economic Review, 75(3). 341-351. Scott, A. J. & Holt, D. (1982). The effect of two stage sampling on ordinary least squares methods. Journal of the American Statistical Association. 77. 848-854. Shepherd, W. G. (1982). Causes of increased competition in the US economy. Review of Economics and Statistics. 64(November), 613-626. 152 Simon, H. A. & Bonini, C. P. (1958). The size distribution of business firms. American Economic Review, 48(September), 607-617. Skinner, C. J. (1989). Introduction to Part A. In C.J. Skinner, D. Holt & T.M.F. Smith (Eds.) Analysis of Complex Surveys (pp. 25-38). New York: John Wiley and Sons. Statistics Canada. (1988). Developing a Longitudinal Database on Businesses in the Canadian Economy: An Approach to the Study of Employment (Catalogue No. 18-501). Ottawa: Minister of Supply and Services. Statistics Canada. (1995). Industrial Organization and Concentration in Manufacturing Industries (Catalogue No. 31C0024). Ottawa: Minister of Industry. Statistics Canada. (1996). Aggregate Productivity Measures (Catalogue No. 15-204E). Ottawa: Minister of Industry. Steiger, J. H. (1995). Structural equation modeling. In Statistica for Windows (Vol. Ill) (pp. 3539-3688). Tulsa, OK: Statsoft. Stinchcombe, A. L. (1965). Organizations and social structure. In J.G. March (ed.). Handbook of organizations (pp. 142-193). Chicago: Rand-McNally. Storey, D. J. (1994). New firm growth and bank financing. Small Business Economics, 6(2), 139-150. Thompson, J. D. (1967). Organizations in action. New York: McGraw-Hill. Venkatraman, N. (1990). Performance implications of strategic coalignment: A methodological perspective. Journal of Management Studies, 27(1), 19-41. Venkatraman, N. & Camillus, J. C. (1984). Exploring the concept of "fit" in strategic management. Academy of Management Review, 9(3), 513-525. Venkatraman, N. & Prescott, J. E. (1990). Environmental-strategy coalignment: An empirical test of its performance implications. Strategic Management Journal, 11(1), 1-23. Ward, P. T., Bickford, D. J. & Leong, G. K. (1996). Configurations of manufacturing strategy, business strategy, environment and structure. Journal of Management, 22(4), 597-626. Wold, H. (1980). Soft modeling: Intermediate between traditional model building and data analysis. Mathematical Statistics. 333-346. Wold, H. (1982). Systems under indirect observation using PLS. In C. Fornell (Ed.), A Second Generation of Multivariate Analysis (Vol. 1). (pp. 325-347). New York: Praeger. Wolfe, R.A.. 1994. Organizational innovation: Review, critique and suggested research directions. Journal of Management Studies. 31(3). 405-431. 153 a o 03 u s o <u S < o e N P - p - i n p i c N o o — oo m oo — — ov P I — vo ^ j U i n p ^ c n o o r n i n t ^ p g - ^ T r o o o p i o r ^ o v O T f O J ) ^ OO N -i) 3 Ifi «i O " 1 N -O n - ; - ; in in p» vo — Ov rs vq « r; « h • • in vo ov in Ov cn (N cn CN 0 0 in o in vd a\ * *  o OV o  * in H — OV cn — in Ov — vq ov TT d d CN J 3 •SP * g ^ TT S u m C O <N D O N c o ' i IS o <u g- t3 -n •a & § 2 W vo # rr P->n oo cn cn d d •n vo ov in p-o ov m m d d vo i n ov Ov vo m p . cn © v-i d 0 0 o d — ov — p- p- m © — — Xj in ov vo m vo ov vo 3 m - SS d i '— CN O T f OV C N O T t O v r ^ c N - ^ - t N r ^ ' S - o o o v r ^ c n v o o o o v o v o ov — v o v o o v v o o v o o v c n c N o o o o o v v o m i n T t r ^ m v q o q i n r n — o q v q o q — < ov oq — r-; ov in oo p ) T J -c n i o d c N - ^ r n r N r n T t c N c n - ^ — ; CN —; cn p i — - p i ] OO <U — ' o Ov ov o * JH c o t cn CN o o cn cn vo • ^ ^ ^ r n oq CN © T 1 i — pi d ov m T f vo vq q d d r o B '.SP S , I P e o o IS o Ct-i ° <D CN O T f U -t— P- cn d ^ S in « o CN CN ! P - P I 9 © 9 d © OV * oo vo Co in CN OV vd TT oo cn CN * * * * 0 0 c n o m o v * „ 2 ; „ * cn v o m T f O v O ^ c n o o o * _ d d H v d ^ ^ « i ^ ^ 2 i i i T c n O C N O m w CN fN o o _• — * * * * # # VO +-VO cn T f CN in p- p- 0 0 p- in in CN CN d i CN d cn oo c n O d in CN CN vo cn i n T J -t - ~ c n c n c n v o i n c N o o o v ( N — o v c N c n o o — i n r - -© © " • t f - a - v o o o t ^ — p - o v v o p - c n p - o o — © — W v o i n — — vovo-* CN i n — P j N » ^ ^ Oj H i n o q o v C - ; v d i n i n ' r - ; v d v d i n v d v d T j - r ~ ; T3 OB ' o o v !S r s - >^  p~ u 0 0 vo Ov £S in cn in • — p-cn — TT — d d ro £ 1-5 .SP <3 , , c P 0 0 OV o ov i n U I N d o cn o m CN CN cn — CN © © d o © v o v o - ^ ^ ^ O ^ ^ S g p ^ P ^ c n ^ o q ^ d d ? ? 9 - 7 7 *- it CN — m — OV CN p-* VO * oo P- <~i VO OV m ov o CN cn r f PJ T f — — CN OV OV v o ° : — ' | o o o v o N ' ; r ' ! l ? i C 1 N v d © 9 9 © 9 9 ^ ^ VO Ov OV P- Ov <=>. p i o — -5 < •S c 5 ov o m TO CO c . ( U '3 o u OJ) 00 \q <- s > a -o 00 1 3 t3 > c o o c c <u •is T3 c D 03 T3 C C S C 3 J3 S <D o u -o — o S q aa d '3 V % * 3 * CJ * 4= in o o « d « v c » 3 — o V c c s - 1 U <D _ - .2 o on 8 o .1 & CO • — > E .£ u o o (D ID > ~o * 3 c O 00 H c<3 § s C O 2^ cn <D * •C I •*-» 1 00 fN .s ^ *c •—1 C3 — .SP •^  I tS 3 S.S s !> ov <u Pi s -a u c -S re <^ vo o S .2 p- ca CD — C 3 "3 > > SO 'S c ° . = <D -o o. c o ^ s SS Ui 0 S , o °°-O 0 0 <u T f in Table A4.2 Rules for SEPATH Structural Model Diagrams (1) Manifest variables are always represented in boxes (squares or rectangles) while latent variables are always in ovals or circles. (2) Directed relationships are always represented explicitly with arrows between two variables. (3) Undirected relationships need not be represented explicitly. (See rule 9 below regarding implicit representation of undirected relationships.) (4) Undirected relationships, when represented explicitly are shown by a wire from a variable to itself, or from one variable to another. (5) Endogenous variables may never have wires connected to them. (6) Free parameter numbers for a wire or arrow are always represented with integers placed on or slightly above the middle of the wire or arrow line. (7) Fixed values for a wire or arrow are always represented with a floating-point number containing a decimal point. The number is generally placed on or slightly above the middle of the wire or arrow line. (8) Different statistical populations are represented by a line of demarcation and the words Group 1 (for the first population or group), Group 2, etc., in each diagram section. (9) All exogenous variables must have their variances represented either explicitly or implicitly. If variances and covariances are not represented explicitly, then the following rules hold: a) For latent variables, variances not explicitly represented in the diagram are assumed to be 1.0, and covariances not explicitly represented are assumed to be 0. b) For manifest variables, variances and covariances not explicitly represented are assumed to be free parameters each having a different parameter number. These numbers are not equal to any number appearing explicitly in the diagram. Source: Steiger (1995: 3559-3560) 155 Table A4.3 Principal Components of the Competitive Environment Dynamism Hostility PRD CHNG 0.059 0.804 TEC CHNG 0.068 0.791 EQP CHNG 0.435 0.457 PRD SUBS 0.796 0.010 SUP SUBS 0.727 0.049 NEW_COMP 0.697 0.233 Eigenvalue 2.219 1.162 % Variance Explained 36.99 19.36 Cronbach's Alpha 0.66 Table A4.4 Principal Components of Competitive Strategy Product Strategy Service Strategy FLEXIBLE 0.178 0.696 QUALITY 0.171 0.758 CST^SVCE 0.143 0.795 CUSTOM 0.757 0.190 RANGE 0.837 0.129 NEW_PROD 0.850 0.141 Eigenvalue 2.676 1.164 % Variance Explained 44.59 19.4 Cronbach's Alpha 0.75 Table A4.5 Principal Components of Managerial Experience Manager Experience FIRM_EXP 0.798 IND_EXP 0.887 MGREXP 0.826 Eigenvalue 2.104 % Variance Explained 70.15 Cronbach's Alpha 0.78 156 o T t 1 0 C O o • f e ON i n o O s ON 00 CN 03 TJ 0) "5 s Ctf i -cd ft. CL> > —1 > _ CJ P H - J o o o o o © © © © © © © © © © © © © T t — © o o o o © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © T t —< o © © © © © © © © © © © ©' © © © © © © © © ON CN CN ON VO NT) CN cn VO VO o o VO 00 VO m CN i n i n CO IT) ON t-- c n 00 m i n i n © 00 CN o O o © © © ON © © © © © © © © o m m ©' © © © © © © © © © © © © ©' ©' © © ©' CN m o o NO i n ON T t OO c n o o cn CN —* VO ,—p T t OO CN T t —HH NO T—I 00 i n © v o ON v o t"- NO T t T— T— 00 l > T t © v q v q v q v q 00 © c n i n »—j CN i n CN o T t © © © © © T-H CO © T-H --1 © ' T-H © cn © VO u - , T t r n o o o ° O 0 0 ^ I T t © © © © © © © © © © o © © © © © © o v o © © © © © © © © © © © © © © © © © © © CN © © © © © © © © o © © © © © © © © © © ON © © © © ' © © o © © © © ' © ' © © © © ' © ' © ' o © © ' © oo i n © T t VO cn oo 00 ON v o i n i n i n i n ON v o cn m T t NO T t T t i n v o CN T t i n © T t VO oo © v o © oo oo © © © © © © T t O © © © --H © © i n NO NO © © ' © © © © ' © T-H © © © © © © © © © © i n 00 r - - CN © © CN o o o o T t T t ON cn cn 00 r - ON CN cn CN i n cn © © CN 00 © © CN i n o o T t T t 00 ON ON T t © VO © i n © r - ; CN © © © © © o i-l v d T^ © — —'< © —! T-H © cn o o _ NO NO © © VO ON d CN ^ T t © CCJ TJ i -C O do W 8 <o cd .5 L- 4-1 crj cn P H W © © © © © © © © © © © © © © © © © T t T t © © © © © © © © © © © © © © © © © © cn © © © © © © © © © © © © © © © © © © ON © © © ' © © © © © © © © © © © © © © © ' © © © © © ON © cn © 00 ON ON VO ON CN T t VO VO ON m T t cn cn cn c n c n VO f-H i n i n c n m T t T t T t © T t cn © © © © © © CN © © © © © © © © i n T t T t d © © © © © © © © © © © © ' © ' © © o o © © c n o o © i n m 00 T t CN T^ NO VO t~-ON i n VO © c n T—1 m T—* m VO T-H cn 00 T t VO —~ ^—• T t T t ON vq cn ON ON c n 00 © CN i n CN CN © © © © © © © i-l T t © © —I © T-H —1 © cn © ° ° ON g O ON Q o m CU •s +-» < I J2 O o c o a4 W "3 - H 3 o CO o Tt in CO |co O s in e n ON ON OO CN CO > +3 C co W -S « s e a .5 cd 03 2 & CTJ c o e8 £ CO tl) o o © © © © © © © © © © © © © © ON © o © © © © © © © © © © © © © © © © o © © © © © © © © © © © © © © © © © d © © © © © © © © © © d d d d © © © VO © c n CN CN in T t ON OO © V0 v o ON ON o o m •*t T f in in in 00 in r-H © © v q T t ON o © © © © © © © © © r-H ' ' © d in T t © © © © © © • 1 © © © © d d d d d ON CN m NO m CN © T t ON T t c n © © o o T t — OO © OO ON © ON t-- ON ON v o .• NO NO d < ^ T t V  c n in ON © T t 00 in v q CN T T ON ON ' — 1 _ ^ CN . d © © • 1 in • © © CN • • © CN c n © o © © © © © © © © © © © © © © © c n © o © © © © © © © © © © © © © © © c n © o © © © © © © © © © © © © © © © v o © d © © © © © © © © © © © © d d d d d ON T f © NO m © o o T t ON OO © ON o o ON CN ON m T t T f T t o o T t OO o o r - v o © ON VO o © © © © © © © © © © © © T t in in d © © © © © © © © © © © © © © tti o o T t r-H r-H © ON 00 v o VO © 00 . 1 00 VO m r— © © v o T t CN VO CN c n CN T t v q v q v q , ; T t ON t - - r~- o o CN ^ H d —' —< r-H —< r-H r ^ © © —< —< © © <N ON NO CN CN CN T t ON CN © ON ON © VO © CO > CO —1 a -a u c o co w B <o ecS .5 1 - +H crj cn P H t i l © © © © © © © © © © © © © © © CN © © © © © © © © © © © © © © © o CN © © © © © © © © © © © © © © © © OO © © © © © © © © © © © © © © © © © © © © oo •tf 00 vo VO ON r-- 00 CN in in ON vo c n CN CN m c n m ON c n CN T f © © © © © © © © © © © © © © © © T f T t c n © © © © © © © © © © © © © © OO © © in oo v o v o ON r~ - r t ON in c n r n v q v q oo CN T t in © © © r - . r-^ T f CN r-~ VO ON v q oo CN 00 © T t in ON m 00 CN CN T f 19.9 ON © oo r ^ © © CN r—' © c n © CN T t T t ON OO OO o o c n © vo © oo m ^ f—1 f—1 H f—1 on co co co £ £ £ ai pi ai . CO CO CO P H o T * o ON i n c n ON ON OO CN 4) > o o o o o o o o © ,—1 © o o o o o o o © © m © o o o o o o o © © © © © © © © © © © © © © © ca © NO OO i n 1—1 m CN © CN r— ON T—< 00 OO o o © © © CN o © © © o T t © © © © © © © © ' © CN CN —> i n i n i n © ^ r © o o © © i n c n © ,—i NO 00 c n i n c n © CN CN © © © CN ON ON ON i n m © r - ON © c n <=> TJ- © 2 Z P-c - J CS S= O ca £ © © © © © © © © T t © © © © © © © © © © © © © © © © © © © © © © © © © ' © © © © © © © © ' © CN ON o o © ON NO c n 00 o o i n CN 1—H ^ H ON 1-H »-H r—l p ~ I - 1 o o © © © m © © © © NO* i n © ' © © © © ' © © © © g . s m — ca cu ON m NO NO © © C N i n m NO Tf _ 1.689 i n CN © Tt CN 1.689 ON CN © m © CN 1.689 © © © © CN •vT CN © 2 5 ca c ca c/3 W © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © © o © ' © © © © © © © © © CN c n CN ON ON © c n r-H CN r— © © NO CN NO © © © CN © © © © NO m © © © © © © © © 00 © <o B 2 ca PH C i n o o • - ON NO ON Tt i n TI-LL) © © © ' © 00 m NO r - NO ON c n © CN CN © © © CN I NO © ON i n COMPREHENSIVE REFERENCES Acar, A. C. (1993). The impact of key internal factors on firm performance: An empirical study of small Turkish firms. Journal of Small Business Management. 31(4). 86-92. Acs, Z. J. & Audretsch, D. B. (1987). Innovation, market structure, and firm size. Review of Economics & Statistics. 69(4). 197-210. Acs, Z. J. & Audretsch, D. B. (1990). The determinants of small-firm growth in US manufacturing. Applied Economics, 22(2), 143-153. Aislabie, C. (1992). Sudden change in a model of small firm growth. Small Business Economics, 4(4), 307-314. Aldrich, H. E. & Auster, E. R. (1986). Even dwarfs started small: Liabilities of age and size and their strategic implications. Research in Organizational Behavior, 8(1), 165-198. Altman, E. I. (1971). Corporate Bankruptcy in America. Lexington, MA: D.C. Heath and Co. Amburgey, T. L., Kelly, D. & Barnett, W. P. (1993). Resetting the clock: The dynamics of organizational change and failure. Administrative Science Quarterly, 38(1), 51-73. Amit, R. & Schoemaker, P. J. H. (1993). Strategic assets and organizational rent. Strategic Management Journal, 14(1). 33-46. Arbaugh, J. B. & Camp, S. M. (1999). Managing growth transitions: Theoretical perspectives and research directions. Florida: Presented at the State of the Art in Entrepreneurship Conference. Arrow, K. J. (1982). Innovation in large and small firms. New York: Price Institute for Entrepreneurial Studies. Baden-Fuller, C. (1995). Strategic innovation, corporate entrepreneurship and matching outside-in to inside-out approaches to strategy research. British Journal of Management. 6(Special Issue), S3-S16. Bagozzi, R. P. & Youjae Yi. (1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science. 16(1). 74-94. Baldwin, J., Dupuy, R. & Gelatly, G. (1998). The Determinants of Survival for New Canadian Firms: A Characteristics-Based Approach. Working Paper. Ottawa: Minister of Industry. Baldwin, J., Gray, T., Johnson, J., Proctor, J., Rafiquzzaman, M. & Sabourin, D. (1997). Failing Concerns: Business Bankruptcy in Canada (Catalogue No. 61-525-XPE). Ottawa: Minister of Industry. Banbury, C. M. & Mitchell, W. (1995). The effect of introducing important incremental innovations on market share and business survival. Strategic Management Journal, 16(Special Issue), 161-182. 160 Barkham, R., Gudgin, G., Hart, M. & Hanvey, E. (1996). The Determinants of Small Firm Growth. Gateshead, Tyne and Wear, UK: Athanaeum Press. Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management. 17(1). 99-120. Barney, J. B. & Hansen, M. H. (1994). Trustworthiness as a source of competitive advantage. Strategic Management Journal. 15(Special Issue). 175-190. Bates, T. (1990). Entrepreneur human capital inputs and small business longevity. Review of Economics & Statistics. 72(4). 551-559. Bates, T. & Nucci, A. (1989). An analysis of small business size and rate of discontinuance. Journal of Small Business Management. 27(4), 1-7. Berryman, J. E. (1983). Small business failure and bankruptcy: A survey of the literature. European Small Business Journal, 1 (4), 47-59. Biggadike, R. (1979). The risky business of corporate diversification. Harvard Business Review, May-June. 103-111. Birkinshaw, J. (1997). Entrepreneurship in multinational corporations: The characteristics of subsidiary initiatives. Strategic Management Journal. 18(3). 207-229. Block, Z. (1983). Can corporate venturing succeed? The Journal of Business Strategy, 3(2). 21-33. Block, Z. (1989). Damage control for new corporate ventures! Journal of Business Strategy, 10(2), 22-28. Block, Z. & MacMillan, I. C. (1993). Corporate Venturing: Creating New Businesses Within The Firm. Boston: Harvard Business School Press. Block, Z. & Ornati, O. A. (1987). Compensating corporate venture managers. Journal of Business Venturing. 2(1). 41-51. Boardman, C. M., Bartley, J. W. & Ratliff, R. B. (1981). Small business growth characteristics. American Journal of Small Business. 5(Winter), 33-42. Brander, J., Hendricks, K., Amit, R. & Whistler, D. (1998). The engine of growth hypothesis: On the relationship between firm size and employment growth. Working Paper: The University of British Columbia. Brock, G. W. (1981). The Telecommunications Industry. Cambridge, MA: Harvard University Press. Brock, W. A. & Evans, D. S. (1986). The Economics of Small Businesses. New York: Holmes & Meier. 161 Bruderl, J., Preisendorfer, P. & Ziegler, R. (1992). Survival chances of newly founded business organizations. American Sociological Review, 57(1), 227-242. Bruderl, J. & Schussler, R. (1990). Organizational mortality: The liability of newness and adolesence. Administrative Science Quarterly, 35(3), 530-547. Burgelman, R. A. (1983). A process model of internal corporate venturing in a diversified major firm. Administrative Science Quarterly, 28(1). 223-244. Burns, T. & Stalker, G. M. (1961). The Management of Innovation. London: Tavistock. Carroll, G. R. (1983). A stochastic model of organizational mortality: Review and reanalysis. Social Science Research. 12(4). 303-329. Carroll, G. R. & Delacroix, J. (1982). Organizational mortality in the newspaper industries of Argentina and Ireland: An ecological approach. Administrative Science Quarterly. 27(2). 169-198. Carroll, G. R. & Huo, Y. P. (1986). Organizational task and institutional environments in evolutionary perspective: Findings from the local newspaper industry. American Journal of Sociology. 91(4). 838-873. Carter, N. M., Stearns, T. M., Reynolds, P. D. & Miller, B. (1994). New venture strategies: Theory development with an empirical base. Strategic Management Journal. 15(1). 21-41. Carter, N. M., Williams, M. & Reynolds, P. D. (1997). Discontinuance among new firms in retail: The influence of initial resources, strategy, and gender. Journal of Business Venturing. 12(1). 125-145. Caruana, A., Morris, M. H. & Vella, A. J. (1998). The effect of centralization and formalization on entrepreneurship in export firms. Journal of Small Business Management. 36(1), 16-29. Caves, R. E., Porter, M. E., Spence, A. M. & Scott, J. T. (1980). Competition in the Open Economy. Cambridge, MA: Harvard University Press. Chaganti, R. (1987). Small business strategies in different industry growth environments. Journal of Small Business Management, 25(3). 61-68. Chandler, G. N. & Hanks, S. H. (1998). An examination of the substitutability of founders human and financial capital in emerging business ventures. Journal of Business Venturing. 13. 353-369. Chung, L. H. & Gibbons, P. T. (1997). Corporate entrepreneurship: The roles of ideology and social capital. Group & Organization Management, 22(1), 10-30. Cliff, J. E. (1998). Does one size fit all? Exploring the relationship between attitudes towards growth, gender, and business size. Journal of Business Venturing. 13(6), 521-542. 162 Cochrane, A. B. (1981). Small business mortality rates: A review of the literature. Journal of Small Business Management, 19(4). 50-59. Conner, K. R. (1991). A historical comparison of resource-based theory and five schools of thought within industrial organization economics: Do we have a new theory of the firm? Journal of Management, 17 (1), 121-154. Cool, K., Dierickx, I. & Jemison, D. (1989). Business strategy, market structure and risk-return relationships: A structural approach. Strategic Management Journal, 10(6), 507-522. Cooper, A. C. (1979). Strategic management: New ventures and small business. In D. E. Schendel & C. W. Hofer (Eds.), Strategic Management: A New View of Business Policy and Planning (pp. 316-327). Boston: Little, Brown, and Company. Cooper, A. C. (1993). Challenges in predicting new firm performance. Journal of Business Venturing. 8(3), 241-253. Cooper, A. C , Gimeno-Gascon, F. J. & Woo, C. Y. (1994). Initial human and financial capital as predictors of new venture performance. Journal of Business Venturing. 9(5), 371-395. Cooper, A. C , Willard, G. E. & Woo, C. Y. (1986). Strategies of high-performing new and small firms: A reexamination of the niche concept. Journal of Business Venturing, 1(3). 247-260. Cooper, A. C , Woo, C. Y. & Dunkelberg, W. C. (1989). Entrepreneurship and the initial size of firms. Journal of Business Venturing. 4(5). 317-332. Covin, J. G. & Slevin, D. P. (1989). Strategic management of small firms in hostile and benign environments. Strategic Management Journal. 10(1). 75-87. Covin, J. G. & Slevin, D. P. (1994). Corporate entrepreneurship in high and low technology industries: A comparison of strategic variables, strategy patterns and performance in global markets. Journal of Euromarketing. 3(3). 99-127. Daily, C. M. (1995). The relationship between board composition and leadership structure and bankruptcy reorganization outcomes. Journal of Management. 21(6). 1041-1056. Davidsson, P. (1991). Continued entrepreneurship: Ability, need, and opportunity as determinants of small firm growth. Journal of Business Venturing. 6(6). 405-429. Davidsson, P. & Wiklund, J. (1999). Conceptual and empirical challenges in the study of firm growth. Florida: Presented at the State of the Art in Entrepreneurship Conference. Davig, W. (1986). Business strategies in smaller manufacturing firms. Journal of Small Business Management. 24(January). 38-46. Day, D. L. (1994). Raising radicals: Different processes for championing innovative corporate ventures. Organization Science. 5(2). 148-172. 163 Dess, G. G. & Robinson, R. B. (1984). Measuring organizational performance in the absence of objective measures: The case of privately held firms and conglomerate business units. Strategic Management Journal, 5(2), 265-73. Dougherty, D. (1995). Managing your core incompetencies for corporate venturing. Entrepreneurship: Theory & Practice, 19(3). 113-135. Dun and Bradstreet. (1981). Business failure record. Dunne, T., Roberts, M. J. & Samuelson, L, (1989). The growth and failure of US manufacturing plants. Quarterly Journal of Economics. 104(4), 671-698. Dunne, T., Roberts, M. J. & Samuelson, L. (1988). Patterns of firm entry and exit in US manufacturing industries. Rand Journal of Economics, 19(4). 495-515. Evans, D. S. (1987). The relationship between firm growth, size, and age: Estimates for 100 manufacturing industries. Journal of Industrial Economics. 35(4). 567-581. Fichman, M. & Levinthal, D. A. (1991). Honeymoons and the liability of adolescence: A new perspective on duration dependence in social and organizational relationships. Academy of Management Review, 16(2), 442-468. Fiorito, S. S. & LaForge, R. W. (1986). A marketing strategy analysis for small retailers. American Journal of Small Business, 10(4), 7-17. Fredland, J. E. & Morris, C. E. (1976). A cross-sectional analysis of small business failure. American Journal of Small Business. 1(1). 7-18. Freeman, J., Carrol, G. R. & Hannan, M. T. (1983). The liability of newness: Age dependence in organization death rates. American Sociological Review. 48(4), 692-710. Frost, P. J. & Egri, C. P. (1990). Influence of political action on innovation: Part I. Leadership & Organization Development Journal, 11(1), 17-25. Garud, R. & Van de Ven, A. H. (1992). An empirical evaluation of the internal corporate venturing process. Strategic Management Journal. 13(Special Issue), 93-109. Gaskill, L. R., Van Auken, H. E. & Manning, R. A. (1993). A factor analytic study of the perceived causes of small business failure. Journal of Small Business Management. 31(4). 18-31. Gatignon, H. & Xuereb, J.-M. (1997). Strategic orientation of the firm new product performance. Journal of Marketing Research. 34(1), 77-90. Gertz, D. L. & Baptista, J. P. A. (1995). Grow to be Great. New York: The Free Press. Ghemawat, P. (1991). Commitment: the dynamics of strategy. New York: Free Press. 164 Gimeno, J., Folta, T. B., Cooper, A. C. & Woo, C. Y. (1997). Survival of the fittest? Entrepreneurial human capital and the persistence of underperforming firms. Administrative Science Quarterly. 42(4). 750-783. Ginsberg, A. & Hay, M. (1994). Confronting the challenges of corporate entrepreneurship: Guidelines for venture managers. European Management Journal. 12(2). 382-389. Glaser, B. G. (1978). Theoretical Sensitivity: Advances in the Methodology of Grounded Theory. Mill Valley, CA: Sociology Press. Glaser, G. B. & Strauss, A. L. (1967). The Discovery of Grounded Theory: Strategies for Qualitative Research. Chicago: Aldine. Guth, W. D. & Ginsberg, A. (1990). Guest editors' introduction: corporate entrepreneurship. Strategic Management Journal, 11 (Special Issue), 5-15. Hall, B. H. (1987). The relationship between firm size and firm growth in the US manufacturing sector. Journal of Industrial Economics. 35(June). 538-605. Hall, G. (1994). Factors distinguishing survivors from failures amongst small firms in the UK construction sector. Journal of Management Studies. 31(5). 737-760. Hall, G. (1992). Reasons for insolvency amongst small firms - A review and fresh evidence. Small Business Economics. 4(3). 237-250. Hambrick, D. C. & D'Aveni, R. A. (1988). Large corporate failures as downward spirals. Administrative Science Quarterly. 33(1). 1-23. Hannan, M. T. & Freeman, J. (1984). Structural inertia and organizational change. American Sociological Review. 49(1). 149-164. Herron, L. & Robinson, R. B. J. (1993). A structural model of the effects of entrepreneurial characteristics on venture performance. Journal of Business Venturing. 8(3), 281-294. Hornsby, J. S., Naffziger, D. W., Kuratko, D. F. & Montagno, R. V. (1993). An interactive model of the corporate entrepreneurship process. Entrepreneurship: Theory & Practice. 17(2), 29-37. Hoy, F., McDougall, P. P. & Desouza, D. E. (1992). Strategies and environments of high growth firms. In D.L. Sexton & J.D. Kasarda (Eds.). The State of the Art in Entrepreneurship (pp. 341-357). Boston: PWS-Kent Publishing. Jennings, D. F. & Lumpkin, J. R. (1989). Functioning modeling corporate entrepreneurship: An empirical integrative analysis. Journal of Management. 15(3). 485-502. Johnson, J., Baldwin, J. & Hinchley, C. (1997). Successful Entrants: Creating the Capacity for Survival and Growth (Catalogue No. 61-524-XPE). Ottawa: Minister of Industry. Jovanovic, B. (1982). Selection and the evolution of industry. Econometrica. 50(4). 649-670. 165 Kalleberg, A. L. & Liecht, K. T. (1991). Gender and organizational performance: Determinants of small business survival and success. Academy of Management Journal, 34(1), 136-161. Kanter, R. M. (1985). Supporting innovation and venture development in established companies. Journal of Business Venturing, 1(1), 47-60. Keasey, K. & Watson, R. (1987). Non-financial symptoms and prediction of small company failure: A test of Argenti's hypothesis. Journal of Business Finance and Accounting. 14(3), 335-354. Keats, B. W. & Bracker, J. S. (1988). Toward a theory of small firm performance: A conceptual model. American Journal of Small Business. 12(Summer), 41-58. Keeley, R. H. & Roure, J. B. (1990). Management, strategy, and industry structure as influences on the success of new firms: A structural model. Management Science, 36(10), 1256-1267. Khandwalla, P. N. (1973). Effect of competition on the structure of top management control. Academy of Management Journal, 16(2), 285-310. Knight, R. M. (1989). Technological innovation in Canada: A comparison of independent entrepreneurs and corporate innovators. Journal of Business Venturing, 4(4), 281-288. Larson, C. M. & Clute, R. C. (1979). The failure syndrome. American Journal of Small Business. 4(2), 35-43. Lawrence, P. & Lorsch, J. (1967). Organization and Environment. Boston: Harvard University. Lee, F. C. & Has, H. (1996). A quantitative assessment of high-knowledge versus low-knowledge industries. In Peter Howlett (ed.), The Implications of Knowledge-Based Growth for Micro-Economic Policies. Calgary: University of Calgary Press. Lefebvre, L. A. & Lefebvre, E. (1993). Competitive positioning and innovative efforts in SMEs. Small Business Economics, 5 (4), 297-305. Lengnick-Hall, C. A. (1992). Strategic configurations and designs for corporate entrepreneurship: Exploring the relationship between cohesiveness and performance. Journal of Engineering & Technology Management. 9(2), 127-154. Levinthal, D. A. (1991). Random walks and organizational mortality. Administrative Science Quarterly. 36(3). 397-420. Levitt, B. & March, J. G. (1988). Organizational learning. Annual Review of Sociology. 14, 319-340. Litvak, I. A. & Maule, C. J. (1980). Entrepreneurial success or failure - ten years later. Business Quarterly, 44. 68-78. 166 Lucas, R. (1978). On the size distribution of business firms. The Bell Journal of Economics, 9(Autumn), 508-523. Lussier, R. N. (1995). A nonfinancial business success versus failure prediction model for young firms. Journal of Small Business Management 33(1), 8-20. MacMillan, I. C , Block, Z. & Narasimha, P. N. S. (1986). Corporate venturing: Alternatives, obstacles encountered, and experience effects. Journal of Business Venturing, 1(2), 177-191. Malerba, F. & Orsenigo, L. (1996). Schumpeterian patterns of innovation are technology-specific. Research Policy. 25(3). 451-478. Mata, J. (1994). Firm growth during infancy. Small Business Economics. 6(1), 27-39. McCharles, D. C. (1990). Knowledge Production in Manufacturing: Its Impact on Business Services. Vancouver, BC: The Fraser Institute. McGahan, A. M. & Porter, M. E. (1997). How much does industry matter, really? Strategic Management Journal. 18(Special Summer Issue), 15-30. McGrath, R. G. (1995). Advantage from adversity: Learning from disappointment in internal corporate ventures. Journal of Business Venturing, 10(2), 121-142. McGrath, R. G., Venkataraman, S. & MacMillan, I. C. (1994). The advantage chain: Antecedents to rents from internal corporate ventures. Journal of Business Venturing, 9(5), 351-369. McKinlay, R. A. (1979). Some reasons for business failures in Canada. Cost and Management, 53(3), 17-21. Merz, G. R., Weber, P. B. & Laetz, V. B. (1994). Linking small business management with entrepreneurial growth. Journal of Small Business Management, 32(4), 48-60. Miller, A. & Camp, B. (1985). Exploring determinants of success in corporate ventures. Journal of Business Venturing, 1(1), 87-105. Miller, A., Gartner, W. B. & Wilson, R. (1989). Entry order, market share, and competitive advantage: A study of their relationships in new corporate ventures. Journal of Business Venturing. 4(3). 197-209. Miller, A., Spann, M. S. & Lerner, L. (1991). Competitive advantages in new corporate ventures: The impact of resource sharing and reporting level. Journal of Business Venturing, 6(5), 335-350. Miller, A., Wilson, B. & Adams, M. (1988). Financial performance patterns of new corporate ventures: an alternative to traditional measures. Journal of Business Venturing. 3(4), 287-300. Miller, D. (1987). The structural and environmental correlates of business strategy. Strategic Management Journal, 8(1), 55-76. 167 Miller, D. & Friesen, P. H. (1978). Archetypes of strategy formulation. Management Science. 24(9), 921-933. Miller, D. & Friesen, P. H. (1982). Innovation in conservative and entrepreneurial firms: Two models of strategic momentum. Strategic Management Journal. 3(1). 1-25. Miller, D. & Friesen, P. H. (1983). Strategy making and environment: The third link. Strategic Management Journal, 4(3), 221-235. Miller, D. & Toulouse, J. M. (1986). Strategy, structure, CEO personality and performance in small firms. American Journal of Small Business, 10(3), 47-62. Mitchell, W. (1991). Dual clocks: Entry order influences on incumbent and newcomer market share and survival when specialized assets retain their value. Strategic Management Journal. 12(1), 85-100. Mitchell, W. (1994). The dynamics of evolving markets: The effects of business sales and age on dissolutions and divestitures. Administrative Science Quarterly, 39(4), 575-596. Mitchell, W., Shaver, J. M. & Yeung, B. (1994). Foreign entrant survival and foreign market share: Canadian companies' experience in United States medical sector markets. Strategic Management Journal, 15, 555-567. Moulton, W. N. & Thomas, H. (1993). Bankruptcy as a deliberate strategy: Theoretical consideration and empirical evidence. Strategic Management Journal, 14(1), 125-135. Murphy, G. B., Trailer, J. W. & Hill, R. C. (1996). Measuring performance in entrepreneurship research. Journal of Business Research, 36(1), 15-23. Naman, J. L. & Slevin, D. P. (1993). Entrepreneurship and the concept of fit: A model and empirical tests. Strategic Management Journal, 14(2), 137-153. O'Neill, H. M. & Duker, J. (1986). Survival and failure in small business. Journal of Small Business Management, 24(1), 30-37. Pearce, J. A. I., Kramer, T. R. & Robbins, D. K. (1997). Effects of managers' entrepreneurial behavior on subordinates. Journal of Business Venturing, 12(2), 147-160. Pennings, J. M., Lee, K. & Witteloostujin, A. 1. (1998). Human capital, social capital, and firm dissolution. Academy of Management Journal, 41(4), 425-440. Penrose, E. (1959). The Theory of the Growth of the Firm . London: Basil Blackwell. Phillips, B. D. & Kirchhoff, B. A. (1989). Formation, growth, and survival: Small firm dynamics in the US economy. Small Business Economics. 1(1). 65-74. Porter, M. E. (1980). Competitive Strategy. New York: Free Press. Porter, M. E. (1991). Towards a dynamic theory of strategy. Strategic Management Journal, 12(1), 95-117. 168 Powell, T. C. (1992). Organizational alignment as competitive advantage. Strategic Management Journal. 13(2). 119-134. Preisendorfer, P. & Voss, T. (1990). Organizational mortality of small firms: The effects of entepreneurial age and human capital. Organization Studies. 11(1). 107-129. Robson, M., Townsend, J. & Pavitt, K. (1988). Sectoral patterns of production and use of innovations in the UK. Research Policy, 7(1). 1-14. Rumelt, R. P. (1991). How much does industry matter? Strategic Management Journal. 12(3), 167-185. Rumelt, R. P. (1984). Towards a strategic theory of the firm. In R.Lamb (Ed.) Competitive Strategic Management (pp. 556-570). Englewood Cliffs, NJ: Prentice-Hall. Sandberg, W. R. (1986). New Venture Performance. Lexington, MA: D.C. Heath and Company. Sandberg, W. R. & Hofer, C. W. (1987). Improving new venture performance: The role of strategy, industry structure, and the entrepreneur. Journal of Business Venturing, 2(1), 5-28. Sapienza, H. J., Smith, K. G. & Gannon, M. J. (1988). Using subjective evaluations of organizational performance in small business research. American Journal of Small Business. 12(3), 45-53. Sarndal, C. E., Swennsson, B. & Wretman, J. (1992). Model Assisted Survey Sampling. New York: Springer. Scherer, F. M. (1980). Industrial Market Structure and Economic Performance. Chicago: Rand-McNally. Schmalensee, R. (1985). Do markets differ much? American Economic Review, 75(3), 341-351. Schrader, R. C. & Simon, M. (1997). Corporate versus independent new ventures: Resource, strategy, and performance differences. Journal of Business Venturing, 12(1), 47-66. Schussler, R. (1988). The early death of organizations - Problems of the liability of newness. Unpublished Manuscript. University of Munich. Scott, A. J. & Holt, D. (1982). The effect of two stage sampling on ordinary least squares methods. Journal of the American Statistical Association, 77, 848-854. Shepherd, W. G. (1982). Causes of increased competition in the US economy. Review of Economics and Statistics. 64(November). 613-626. Simon, H. A. & Bonini, C. P. (1958). The size distribution of business firms. American Economic Review. 48(September), 607-617. Singh, J. V., House, R. J. & Tucker, D. J. (1986). Organizational change and organizational mortality. Administrative Science Quarterly, 31(4). 587-611. 169 Singh, J. V., Tucker, D. J. & House, R. J. (1986). Organizational legitimacy and the liability of newness. Administrative Science Quarterly, 31 (2), 171-193. Skinner, C. J. (1989). Introduction to Part A. In C.J. Skinner, D. Holt & T.M.F. Smith (Eds.) Analysis of Complex Surveys (pp. 25-38). New York: John Wiley and Sons. Slater, S. F. & Narver, J. C. (1995). Market orientation and the learning organization. Journal of Marketing, 59(3), 63-74. Sorrentino, M. & Williams, M. L. (1995). Relatedness and corporate venturing: Does it really matter? Journal of Business Venturing, 10(1), 59-73. Statistics Canada. (1985). Canada Year Book 1985. Ottawa: Minister of Supply and Services Canada. Statistics Canada. (1988). Developing a Longitudinal Database on Businesses in the Canadian Economy: An Approach to the Study of Employment (Catalogue No. 18-501). Ottawa: Minister of Supply and Services. Statistics Canada. (1994). Canada Year Book 1994. Ottawa: Minister of Industry, Science and Technology. Statistics Canada. (1995). Industrial Organization and Concentration in Manufacturing Industries (Catalogue No. 31C0024). Ottawa: Minister of Industry. Statistics Canada. (1996). Aggregate Productivity Measures (Catalogue No. 15-204E). Ottawa: Minister of Industry. Stearns, T. M., Carter, N. M., Reynolds, P. D. & Williams, M. (1995). New firm survival: Industry, strategy and location. Journal of Business Venturing, 10(1), 23-42. Steiger, J. H. (1995). Structural equation modeling. In Statistica for Windows (Vol. Ill) (pp. 3539-3688). Tulsa, OK: Statsoft. Stevens, J. (1996). Applied Multivariate Statistics for the Social Sciences. Mahwah, NJ: Lawrence Earlbaum Associates. Stinchcombe, A. L. (1965). Organizations and social structure. In J.G. March (ed.), Handbook of Organizations (pp. 142-193). Chicago: Rand-McNally. Storey, D. J. (1994). New firm growth and bank financing. Small Business Economics, 6(2), 139-150. Sykes, H. B. (1986). Lessons from a new ventures program. Harvard Business Review, 64(3), 69-74. Sykes, H. B. & Block, Z. (1989). Corporate venturing obstacles: Sources and solutions. Journal of Business Venturing, 4(3), 159-167. 170 Teece, D. J., Pisano, G. & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal. 18(7). 509-533. Thompson, J. D. (1967). Organizations in Action. New York: McGraw-Hill. Thornhill, S., Amit, R. & Belcourt, M . (1998). Determinants of corporate venture performance: An empirical study of firm and industry effects. Proceedings of the Administrative Sciences Association of Canada Annual Conference. 19(21). 80-90. Tsai, W. M.-H. , MacMillan, I. C. & Low, M . B. (1991). Effects of strategy and environment on corporate venture success in industrial markets. Journal of Business Venturing. 6(1). 9-28. Tushman, M . & Nadler, D. (1986). Organizing for innovation. California Management Review. 3(1), 73-84. Ulmer, K. J. & Nielsen, A. (1947). Business turnover and causes of failure. Survey of Current Business Research. 4(4). 10-16. Venkataraman, S., MacMillan, I. C. & McGrath, R. G. (1992). Progress in research on corporate venturing. In D. Sexton and J. Kasarda (Eds.) State of the Art in Entrepreneurship Research (pp. 487-519). Boston: PWS-Kent Publishing Co. Venkataraman, S., Van de Ven, A. H., Buckeye, J. & Hudson, R. (1990). Starting up in a turbulent environment: A process model of failure among firms with high customer dependence. Journal of Business Venturing, 5(5), 277-295. Venkatraman, N. (1990). Performance implications of strategic coalignment: A methodological perspective. Journal of Management Studies. 27(1), 19-41. Venkatraman, N. & Camillus, J. C. (1984). Exploring the concept of "fit" in strategic management. Academy of Management Review, 9(3), 513-525. Venkatraman, N. & Prescott, J. E. (1990). Environmental-strategy coalignment: An empirical test of its performance implications. Strategic Management Journal, 11(1), 1-23. Venkatraman, N. & Ramanujam, V. (1987). Measurement of business economic performance: An examination of method convergence. Journal of Management, 13(1), 109-22. Watson, J. & Everett, J. E. (1993). Defining small business failure. International Small Business Journal. 11(3). 35-48. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal. 5(2), 171-180. Wold, H. (1980). Soft modeling: Intermediate between traditional model building and data analysis. Mathematical Statistics. 333-346. Wold, H. (1982). Systems under indirect observation using PLS. In C. Fornell (Ed.), A Second Generation of Multivariate Analysis (Vol. 1), (pp. 325-347). New York: Praeger. 171 Wolfe, R. A. (1994). Organizational innovation: Review, critique and suggested research directions. Journal of Management Studies, 31, 405-431. Zahra, S. A. (1991). Predictors and financial outcomes of corporate entrepreneurship: An exploratory study. Journal of Business Venturing, 6(4), 259-285. Zahra, S. A. (1993). Environment, corporate entrepreneurship, and financial performance: A taxonomic approach. Journal of Business Venturing, 8(4), 319-340. Zahra, S. A. (1995). Corporate entrepreneurship and financial performance: The case of management leveraged buyouts. Journal of Business Venturing, 10(3), 225-247. Zahra, S. A. (1996a). Governance, ownership, and corporate entrepreneurship: The moderating impact of industry technological opportunities. Academy of Management Journal, 39(6), 1713-1735. Zahra, S. A. (1996b). Technology strategy and new venture performance: A study of corporate-sponsored and independent biotechnology ventures. Journal of Business Venturing, 11(4), 289-321. Zahra, S. A. & Covin, J. G. (1995). Contextual influences on the corporate entrepreneurship-performance relationship: A longitudinal analysis. Journal of Business Venturing, 10(1), 43-58. 172 

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.831.1-0099468/manifest

Comment

Related Items