Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Implementing welfare-to-work schemes in British Columbia Spence, Robin Kirsten 1994

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

Item Metadata


831-ubc_1994-0623.pdf [ 1.68MB ]
JSON: 831-1.0087466.json
JSON-LD: 831-1.0087466-ld.json
RDF/XML (Pretty): 831-1.0087466-rdf.xml
RDF/JSON: 831-1.0087466-rdf.json
Turtle: 831-1.0087466-turtle.txt
N-Triples: 831-1.0087466-rdf-ntriples.txt
Original Record: 831-1.0087466-source.json
Full Text

Full Text

IMPLEMENTING WELFARE-TO-WORK SCHEMES IN BRITISH COLUMBIAbyROBIN KIRSTEN SPENCEB.A., The University of North Carolina at Chapel Hill, 1993A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF ARTSinTHE FACULTY OF GRADUATE STUDIESDepartment of Political ScienceWe accept this thesis as conformingTHE UNIVERSITY OF BRITISH COLUMBIAOctober 1994© Robin Kirsten Spence, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of 9p’ hco ScJ-c .4)The University of British ColumbiaVancouver, CanadaDate Ock.DE-6 (2188)11AbstractThe successful implementation of the Canada/British ColumbiaAgreement to Enhance the Employability of Social Assistance Recipients (the“SAR”, or Four Corner” Agreement) can be explained by a revised version ofMazmanian and Sabatier’s 1983 theory of implementation. This framework isalso able to account for some of the limitations that the initiative faced. Theanalysis of the SAR Agreement is placed in the context of the on-going dilemmaof work and welfare and in the evolving ideological climate in the B.C. welfaresystem from an ideology of redistribution, to one of liberal developmentalism,emphasizing opportunity before work.The case-study provides a history of the implementation of the SARAgreement in B.C. at both the policy-formulation and field levels ofgovernment, and gives an overview of the agreement and its results. Thisinformation is gained through reports, government documents and interviewswith officials involved in the SAR Agreement. Application of the revisedtheory of implementation to the agreement illuminates the ingredients criticalto the success of the SAR initiative in British Columbia. Among the mostimportant determinants of success were the intensive cooperation betweenfederal and provincial agencies, the amount of discretion given to local officialswhen combined with the expertise and resources of those officials, the correctcausal theory underlying most project designs, and the flexibility of theagreement respond to past successes and failures. The agreement was limited bythe lack of general guidelines to provide officials with a sense of direction, by thepossibility of conflicting goals of outside agencies, and by problems with theinvalid causal theory underlying a few programs.111Table of ContentsAbstract iiTable of ContentsPreface ivIntroduction 1Chapter OneIn Search of a Comprehensive Theory of Implementation 4Laying the Foundation 7Building a Comprehensive Theory 14Policy Breeding Theory: Insights From Workfare Research 23The Preferred Framework and Suggestions for Modification 25Chapter TwoThe Context of the Work/Welfare Problem 30The Work/Welfare Debate: An Overview 31Public Context: A Battle of Social Constructions and Ideologies 33The Policy-maker’s Context: Political Constraints and Conflicting Goals 38The Social Assistance Recipient’s Context: The Pushes and Pulls of 42Welfare RulesWelfare-to-Work Experiments: Modest Successes and Inevitable 46Trade-OffsChapter ThreeA Modest Success Story: The Canada/B.C. SAR Agreement 48The Roots of Reform 49The Agreement 52British Columbia Implements the Agreement 56Implementation of the B.C. Programs at the Local Level 61The Impact of the SAR Agreement in B.C. 63Chapter FourExplaining the Fate of British Columbia’s SAR Agreement 70Applying the Revised Theory of Implementation to the Case-Study 72Explaining Individual Programs’ Successes and Failures 93Conclusion 97Bibliography 102ivPrefaceThe roots of the Canada/British Columbia Agreement to Enhance theEmployability of Social Assistance Recipients are well documented in scholarlyliterature and government documents. Similarly, the outcomes of this initativeare preserved in numbers, reports, and in the evolving SAR Agreement itself. Itis the space between the design and outcome that is relatively uncharted. Forthis information I am heavily indebted to government officials and members ofanti-poverty organizations, whose anonymity this thesis will respect.During the months of July to September I conducted either in-person ortelephone interviews with fifteen individuals involved in the SAR Agreement.They came from all “four-corners” of the Agreement -- Health and WelfareCanada, Employment and Immigration Canada, British Columbia’s Ministry ofSocial Services, and British Columbia’s new Ministry of Skills, Labour andTraining -- and represented both policy-formulators and field-level workers. Inaddition, I heard the voice of the agreement’s critics through the opinions of aVancouver anti-poverty organization, and I gained a valuable nationalperspective from Professor Michael Prince. I am grateful to these individuals, allof whom were extremely generous with their time, and with their knowledge.I would like to thank, particularly, my supervisor, Professor GeorgeHoberg, for his financial support, his critical eye, his time, and his constantinterest.IIntroductionIn the mid-1980s the social assistance system in British Columbia, as in all ofCanada, reached a critical point. It was no longer serving those for whom it wasdesigned -- those unable to support themselves who turn to welfare as a last resort --but had become a stopping place for those who were considered by the governmentto be “employabl&. Policy makers throughout Canada focused on two theories toexplain this phenomenon. The first arose from the recognition that fallingunemp1oymnt did not reduce the number of employables on welfare. Politiciansconcluded that the economy’s labour needs were changing rapidly and that theskills of workers on social assistance were not able to meet these needs. Additionalmeasures to help this group to overcome their barriers to employment wererequired. The second theory was that this welfare dependence was the result of aparticular dilemma of work and welfare that had been created by the interaction ofthe rules of the welfare system and the reality of work conditions. Many workers inlow-paying jobs with few benefits believed (correctly in many cases) that they wouldbe better off on welfare, with its income and in-kind benefits, than working. Thispeculiar relationship between work and welfare created disincentives for welfarerecipients to seek to become self-sufficient. Policy-makers knew that to address theproblem of welfare dependence of employables they had to find ways to train socialassistance recipients to meet the needs of the changing labour market and they hadto provide incentives to encourage them to engage in this training.The result of these imperatives was the Agreement to Enhance the Employabilityof Social Assistance Recipients, (the “SAR”, or “Four corner” Agreement), agreed to byCanadian and provincial officials responsible for social services and labour in 1985and signed by British Columbia in May of 1986. The SAR Agreement generated anextensive network of federal and provincial training programs in British Columbia2tailored to serve social assistance recipients. Like other welfare-to-work schemesthroughout North America, the SAR strategy for British Columbia mixed authority,incentive and capacity tools to create an environment in which welfare recipientswould be motivated to become self-sufficient. However, unlike many previouswelfare-to-work experiments in British Columbia and in other regions, the initiativewas successful in meeting many of its objectives.The success of the complex initiative, in which any of its many actors could havethwarted the implementation process, is intriguing. Examination of its particularfeatures yields insights into the important features of policy and implementationdesign that influenced implementing officials and the target group to behave inconformity with the initiative’s objectives.In order to achieve the overall goal of identifying conditions for success ofwelfare-to-work initiatives, several preliminary inquiries are necessary. First, acredible theory of implementation must be found or developed which is appropriatefor application both to the Canadian policy environment, and to the field of socialpolicy. The first chapter of this thesis reviews a body of literature on policyimplementation and extracts the important contributions of each author to create amore comprehensive theory of effective implementation.Next, the intricacies of the policy problem must be understood. Chapter twoprovides a theoretical and practical background to the B.C. case-study by discussingthe three contexts-- that of the public, of the policy-makers and of the socialassistance recipient -- in which the welfare-to-work problem exists. The chapterexamines the transformation in the ideology underlying the welfare system inCanada from redistribution to liberal developmentalism. Finally, it provides asummary of the findings of recent experiments in mixing welfare with work andconfirms the existence of important goal trade-offs in this policy area.3The case-study of the implementation of the SAR Agreement in BritishColumbia makes up chapters three and four. Chapter three provides an account ofhistory of the agreement in both Canada and British Columbia, describesimplementation of the initiative at upper levels of government as well as at the field-level, and finally, considers different evaluations of the impact of the agreement,noting areas of weakness and failure in a package of programs that in general was asuccess.Chapter four draws the three previous chapters together. In this chapter, therevised theory of implementation devised in the first chapter is applied to BritishColumbia’s attempt to address its work/welfare dilemma through the SARAgreement. The eight conditions of the revised theory are effective in explaining theoutcome of the initiative and the successes and failures of its various components.Overall, the application of a theory of implementation suitable to Canadian socialpolicy to the B.C. case-study helps to open up the secrets of effective welfare-towork initiatives. These lessons might provide valuable guidance for future welfareto-work reforms, as well as for problem-solving in other areas of social policy.4Chapter OneIn Search of a Comprehensive Theory of ImplementationThere is a great distance separating a paper agreement between politiciansand the day-to-day reality of Canada’s poorest citizens. Between the years of 1986and 1988 ten provinces and the Northwest Territories signed an agreement with thefederal government to enhance the employability of social assistance recipients byproviding them with more opportunities for training. Since that date, many of the“employment enhancement” initiatives have led to increases in both employabilityand earnings for a large number of social assistance recipients, and have shifted theemphasis in social services from simply providing transfers to the poor, to providingthem with opportunities for self-sufficiency. The ministers’ agreement has, in fact,changed the landscape of social services for the poor and over time has led tomodest gains for this population. The successful transformation of a legislativeobjective into a social and economic reality for the poor marks the effectiveimplementation of a government’s decision. The Federal/Provincial/TerritorialAgreements to Enhance the Employability of Social Assistance Recipients (The SARAgreement) was a significant achievement in a policy arena in which many visionsof social reform have been shattered by the force of the social and economicobstacles that they must confront.Policy, whether in the form of a bill, or in this case a signed agreementbetween governments, is born at the top levels of government. However, its lifedepends on the successive stages that follow its birth until it reaches and influencesits target population. These stages constitute the implementation of a policy, aprocess with tremendous potential for variation and unpredictability. A welldesigned policy with enthusiastic support may nonetheless fail to accomplish itsobjectives; the clue to this failure is to be found somewhere in the stages of5implementation. In order to design and enact effective policy, the various obstaclesto effective implementation must be understood. The process of implementationshould be structured, alongside other elements of policy design, in the policyformulation stage. The ability to connect intentions with results through applicationof a sound theory of implementation is certainly one key to competent governance.The potential power of such a theory has recently drawn many scholars to itspursuit.Amid their shared claim that implementation is a field of study in its infancy,the various students of implementation have amassed a considerable body ofknowledge about the fate of government policies once they leave the legislatures.From case-studies to more comprehensive theoretical frameworks, academics’ questfor an understanding of policy has extended beyond the traditional subject of theforces at work in the policy-formulation and adoption process to the more cloudyarea of the implementation stage. Answers to several key questions are sought inthe literature, among them the questions of the true source of policy, whether theformulators or forces within the implementation process generate policy, andwhether formulation and implementation can be regarded as separate processes.The paramount project, however, of the various researchers is practical; they areconcerned with identifying the conditions which must be met in order for a policy tobe implemented successfully.A chronological account of several major pieces of research captures thebreadth and diversity of the implementation debate. At the same time, theconsiderable overlap in the conclusions of the authors suggests that the key elementsof successful policy implementation are not as elusive as an observer ofgovernment’s disability in several arenas might have once thought. In fact, despitemany researchers’ assertions of the distinctiveness of their variables or approach,and the innovative vocabulary that they employ, the findings reveal that6implementation can be studied systematically and that an empirically testableimplementation hypothesis is within reach. Implementation, initially defined byPressman and Wildavsky as “a process of interaction between the setting of goalsand actions geared to achieving them” (Pressman and Wildavsky 1973: xv),maintains its definition, in all of the research examined in this chapter, as a dynamicprocess in which policy results are rarely the mirrors of policy designs. During thelast two decades the pressing debate between scholars of implementation is overwhich circumstances or factors are most significant in their impact on the fate of apolicy.Eight studies that have contributed new elements to the debate will beexamined in this chapter. First, Pressman and Wildavsky’s Implementation presentsa hypothesis on the complexity of joint action and its relation to implementationprocess. Second, Eugene Bardach’s The Implementation Game examines theinfluence that self-interested individuals exert on the process. Third, an article byRichard Elmore “Backward Mapping: Implementation Research and PolicyDecisions” presents a bottom-up analysis of implementation. This perspective onimplementation is explored further in a fourth work, Michael Lipsky’s Street-LevelBureaucracy. A fifth work, Mazmanian and Sabatier’s Implementation and PublicPolicy, advances a comprehensive, mostly top-down framework to explain policysuccesses and failures. Sixth, Goggin et al.’s Implementation Theory and Practice,tries to improve on Mazmanian and Sabatier’s work by synthesizing top-down andbottom-up approaches and by making implementation analysis more “scientific”.Seventh, Robert Stoker’s Reluctant Partners retreats from the comprehensiveperspective to deal with one element in the implementation process, that is, thecooperation between government and implementing agencies. Finally, in FromWelfare to Work Judith M. Gueron and Edward Pauly outline a framework forunderstanding the impacts of welfare-to-work programs specifically, shedding light7on particular factors which influence the implementation of social policy. Theseeight works, taken together, testify to the progress made in this “new” field and thedegree to which a substantial body of knowledge on implementation now exists.Even while the perspectives differ, at the end of two decades of research the findingsare converging.From these works a framework can be developed which combines now“conventional” implementation wisdom with the clever insights of particular authorsto create a framework which is appropriate to both Canada and to the realm of socialpolicy. This framework will provide the basis for examining the success of the SARinitiatives in British Columbia in later chapters.Laying the FoundationThe work that is often cited as having been most responsible for thedevelopment of the field of implementation research is Jeffrey Pressman and AaronWildavsky’s Implementation. While examining the fate of a single policy -- theEconomic Development Agency’s employment initiative in Oakland California-- theauthors developed a theory of implementation that has implications beyond theircase study.In their detailed five-year account of the government’s initiative the authorsquestion how a program on which all participants initially could agree could fail tobe implemented successfully. Their focus is on the numerous and ever-changingplayers involved in implementation, on their varying goals and perspectives, on thepassage of time, and on the consequent evolution of the policy throughoutimplementation. The conclusion that they reach is that the fate of a policy isdetermined by a number of complex forces, none of which are governed by a singleparticipant in the implementation process.8The theory of implementation that they develop is “The Complexity of JointAction”. It is easily arrived at by their description and analysis of the Oaklandproject. Generally, it says that the more parties involved in implementation, the lesschance that implementation will run smoothly. More specifically, they note that asparticipants change, so do the perspectives, goals and measures of success that theybring to the implementation process. As participants change, so do theunderstandings between them. They explain that “participants may agree with thesubstantive ends of a proposal and still oppose (or merely fail to facilitate) the meansfor effectuating it” (Pressman and Wildavsky 1973: 99), due to different preferencesregarding the priority of various goals and the methods used to promote them. Themore actors there are, the more clearances are necessary for any decision to be made,the more potential veto points exist. Decisions also take time. The more clearancepoints there are, the more potential there is for delay and for the priorities of actorsto change, hindering the implementation process. According to Pressman andWildavsky, the Oakland project was jeopardized by the vast number of agencieswith conflicting priorities that were involved in its implementation.Another significant contribution that the authors make is their recognitionthat the “separation of policy design from implementation is fatal” (ibid: xvii). Theylament the fact that the formulators of the policy did not give the implementorssufficient guidance. “Congress may have written the language, but theadministrators were doing the translating” (ibid: 75). If bills are too vague anddetails of policies are not worked out in advance then it is the various participantswho must contend with one another to establish a bill’s exact meaning. These kindsof struggles hinder implementation.Further concerns of the researchers include resources and theoreticalproblems. Implementing agencies need sufficient funding and personnel to dealwith unforeseen circumstances as well as the initial program requirements. There9must also be a sound theory underlying the policy. Pressman and Wildavskyattribute part of the failure of the Oakland project to a failure on the part offormulators to choose the means that most directly leads towards the desired ends.Implementation is a “top-down” study. The authors begin with a bill andfollow it through the chain of actors involved in “translating” and enacting it. Theirfocus is on how a policy is structured to control the implementation process. The“advice” that policy formulators are left with concerns elements of policy design:improve the clarity of the legislation, provide an adequate budget, minimize thenumber of agencies involved, promote unity at the local level and ensureconsistency in the leadership (a “fixer” -- an individual or agency that stays with theproject throughout implementation-- is key to program success). The authorsrecognize the power that local “street-level” officials exercise, since federal officialsultimately are dependent on them if federal objectives are to be achieved, and theytherefore seek “federal modesty” (ibid: 146). However, rather than taking the“bottom-up” approach of suggesting that more local discretion and informalauthority might help to deal with problems, Pressman and Wildavsky put theirlimited hope for improved implementation in improvements at the centre -- in thepotential of a policy to create more favourable implementation conditions.Four years after Implementation was published, Eugene Bardach respondedwith The Implementation Game. He commends Pressman and Wildavsky for theirinsight into the anatomy of delay and conflicting interests in policy implementation,and criticizes them only for not going far enough. Whereas Pressman andWildavsky hold a somewhat benign view of the complexity of joint action --implying that delays are a natural result of participants all working in their ownways to address a problem— Bardach has a view of implementation participantsthat is distinctly suspicious and negative. In his view, actors or agencies aremotivated solely by their own interests. He characterizes the implementation10process as “highly defensive. A great deal of energy goes into maneuvering to avoidresponsibility, scrutiny, and blame” (Bardach 1977: 37). His theory ofimplementation suggests that a policy will fail unless the motives of individuals caneither be forced to conform to the objectives of a policy, or if they are in conflict, canbe suppressed.Bardach provides a two—part definition of the “implementation processT’: “1. aprocess of assembling the elements required to produce a particular programmaticoutcome, and 2. the playing out of a number of loosely interrelated games wherebythese elements are withheld from or delivered to the program assembly process onparticular terms” (ibid: 57-8). It is the second part of the definition that characterizesBardach’s distinctive perspective -- participants sell-consciously manipulate theimplementation process to further their own interests. Such unconstrainedbehaviour, described by Bardach as “games”, impedes policy implementation in fourways.First, four games have the effect of diverting resources away from thoserecipients designated by a policy. These include overcharging for sub-standardservices, (“Easy Money”); performing only to meet the requirements of thosecontrolling the budget, (“The Budget Game”); under-performance due to jobsecurity, (“Easy Life”) and allocating funds based on political grounds, rather thanout of concern for efficiency (“Pork Barrel”).A second set of three games serve to deflect the policy goals outlined in theoriginal policy by adding on additional goals to the original ones outlined in thepolicy, (“Piling On”); by fashioning unclear elements of mandates to suit one’s needs(“Up for Grabs”); and by attempting to recapture in implementation what was lost inpolicy formulation, (“Keeping the Peace”).Third, three games create the dilemmas of administration: the ability ofimplementors to use symbolic and inadequate gestures to quell demands11(“Tokenism”); the ability of those opposed to a policy to resist together therebymaking enforcement impossible (“Massive Resistance”); and the more impersonalfactors of incompetence, variability and lack of coordination that weakenimplementors’ capacities (“Social Entropy”).Finally, participants engage in five games which have the effect of dissipatingthe energies of those involved in implementation. Participants may hold theirpositions until others give into their demands (“Tenacity”); they may compete forjurisdiction (“Territory”); they might shirk responsibility for the elements ofprograms assigned to them (“Not Our Problem”); they attempt to maintain theiroption for backing out should the costs seem too great for them (“Odd Man Out”); orthey may act more out of concern for their own prestige than for the program itself(“Reputation”).All fifteen games are outlined by Bardach in detail in order to illustrate thenegative force that participants (or “human nature”, it seems at many points),unchecked, can exert on government initiatives. Bardach’s solution is thatgovernment policies should be designed so as to constrain the behaviour of thevarious actors. Through use of authority and through foresight -- “risk assessment”-- on the part of policy formulators, the effect of the games of individual participantscan be avoided. The possible range of behaviour would be determined beforeimplementation and the proper incentives and rules integrated into theimplementation process during the policy design phase in order to keep behaviourconsistent with the goals of policy-makers. This is the ultimate in top-downprescription. Bardach would have a implementation arena in which all discretion istaken away from the local levels and centralized. He champions the notion of apowerful fixer -- an individual or organization in whose self-interest it is thatimplementation be successful.12The “gam& metaphor used by Bardach is clever, and indeed it does describethe behaviour of many actors some of the time. However, it is shallow in itsportrayal of the motivations of participants in the implementation process.Nowhere does he allow for true commitment to an issue inspired by anything otherthan self-interest. He ignores the notion of public good and service which pervadesthe policy arena. By failing to acknowledge that some participants will help tofacilitate an implementation process even when it imposes costs on them he suggeststhat the only likely outcome for a controversial policy is failure.Interestingly, almost as an afterthought, Bardach expands his vision. In asection in which he deals with “non-purposive delays”, he acknowledges that factorsapart from individual self-interest can interfere with program implementation. Hebows to Pressman and Wildavsky in acknowledging the time-consuming nature ofcollective-decision making, regardless of the disposition of each participant.Similarly, only in the last pages of the book does he mention the need for a goodtheory underlying a chosen policy. Bardach offers a useful typology of the kinds ofstrategic behaviour engaged in by participants; his theory elaborates on the actualmechanics of the “complexity of joint action” that Pressman and Wildavsky describe.While his description of the selfish behaviour of some implementing officials iscredible, he does little to convince the reader that this behaviour is pervasive. Histheory of implementation has one overwhelming and narrow focus in its concern forthe self-interested actions of implementors. By ignoring some of the unselfishmotives of those engaged in public service he portrays a very stilted policy world. Ifthe reader does not accept this view, the theory then serves only to explain a fractionof policy outcomes. In passing over consideration of further personal motivationsand other impersonal factors affecting the policy process Bardach fails to provide acompelling theory.13While Bardach advocates constraining behaviour through centralizing formalauthority in policy implementation, Richard Elmore seeks the opposite response tothe problem of government’s inability to achieve the results it seeks from its policies.In his article, “Backward Mapping: Implementation Research and Policy Decisions”,Elmore explores the “bottom-up’ approach to implementation analysis in which thefocus is not the policy but the societal problem itself. The starting point is “the pointat which administrative actions intersect private choices” (Elmore 1979: 604). Ratherthan examining what he terms the “command and control” concerns of top-downanalysis, he advocates an approach that emphasizes local expertise and discretion,bargaining among actors, and incentive structures. His is a perspective that putsfaith the capacity of “street-level” officials to solve a problem, as they are closer to itssource. Rather than seeing the complexity of joint action as a impediment tosuccessful implementation as Pressman and Wildavsky do, or sharing Bardach’sdistrust of the varying behaviours and motivations of participants, Elmore seesdecentralization as a resource and suggests that by extending local discretion andinformal authority, innovations in problem-solving can be nurtured.Like Elmore, in Street-Level Bureaucracy Michael Lipsky argues that theactions of street-level bureaucrats are critical to the success or failure of a policy,since it is they who have the most contact with the target group whose behaviour apolicy is designed to influence. Beyond their role in affecting the implementation ofa policy, Lipsky suggests that they actually create the policy that is received by theclient. He calls street-level bureaucrats “policy-makers” for two reasons. “Theyexercise wide discretion in decisions about citizens with whom they interact. Then,when taken in concert, their individual actions add up to agency behaviour” (Lipsky1980: 13). The chief resources which street-level bureaucrats have at their disposalare discretion and “relative autonomy from organizational authority”. Theseresources can be used either to further or to hinder policy objectives. Lipsky14contends that in the current institutional environment in which these bureaucratswork, effective implementation of a policy is likely to be hindered rather thanhelped.Lipsky characterizes the street-level bureaucracy in terms of five conditions ofwork in addition to discretion, autonomy and a high degree of interaction withclients. These include chronically inadequate resources relative to performancedemands; demand for service that exceeds the supply; goal expectations which areambiguous or conflicting; performance goals which are difficult to measure, andclients who are non-voluntary (ibid: 27-28). In an institutional environmentcharacterized by poor conditions of work, street-level bureaucrats need to find waysto adapt to these conditions, and to maintain their sense of the value of their ownwork. “Patterns of practice”, as Lipsky calls these adaptations, tend to routinize thework of the bureaucrat and to lower their expectations of both their target group andof their own role in an organization in terms of the objectives they set out to achieve.These patterns of practice lead to impacts of a policy being implemented by officialsthat are far beneath the potential of the policy.Lipsky’s theory holds that street-level bureaucrats will continue to be policy-makers so long as discretion and relative autonomy is allowed. So long asinadequate working conditions prevail in their bureaucracies, street levelbureaucrats will develop patterns of practice which will put limits on effectiveimplementation. As such, more effective implementation can be expected wheneither discretion or autonomy is reduced or working conditions in street-levelbureaucracies improved.Building A Comprehensive TheoryIt is Daniel Mazmanian and Paul Sabatier who manage to move beyond thecase-study approach of Pressman and Wildavsky and the narrow theoretical focus of15Bardach to provide a comprehensive theory emphasizing several variables affectingimplementation in order to apply them to a variety of case-studies. InImplementation and Public Policy they argue that “the basic components of policyimplementation are the same across most policy areas and ... the effectiveness oftheir implementation results from the presence of more, or less, of the relevantfactors” (Mazmanian and Sabatier 1983: 3). Mazmanian and Sabatier’s book attemptsto name and define these factors.Theirs is a mostly top-down approach that seeks solutions in the creation of apolicy environment that will facilitate implementation. However, their theory canbe interpreted to incorporate some of Elmore and Lipsky’s bottom-up concerns inrelation to “critical linkages” affecting target group compliance. They define the keydependent variable, implementation itself, as a series if five stages. The first stage isthe policy output or decision. This is followed by the compliance of the targetgroups. The next stage is the actual impact of the decisions, followed by theperceived impacts of the decisions. Finally, the initial decision is evaluated andrevised. Unlike the bottom-up theorists and Bardach, Mazmanian and Sabatiermaintain the distinction between policy formulation and implementation; this ispossible through their use of the feedback loop to account for modifications (ibid: 9).Mazmanian and Sabatier echo Bardach in their assertion that “the energizingforce of the implementation process is the rational pursuit by individuals of theirdesires for power, security, and well-being” (ibid: 20). However, they differ fromBardach in recognizing that this “pursuit” occurs within “broad historical, culturaland economic conditions and organizational imperatives” (ibid). In other words, thecontext that shapes individuals’ interests is the independent variable withinMazmanian and Sabatier’s framework, whereas individuals’ interests wereportrayed as independent variables within Bardach’s theory. That individuals’16interests are shaped by context is a possibility overlooked by Bardach who only sawcontext as dependent on interests.Out of this understanding, Mazmanian and Sabatier develop an elaborateframework, and a corresponding simpler set of “six conditions of effectiveimplementation”. In their framework they divide the independent variables intothree categories. The first concerns the “tractability of the problem”. They posit thateffective implementation is directly related to the tractability of the problem.Effective implementation becomes more likely as 1) the technical difficultyassociated with solving a problem decreases; 2) the diversity of the proscribedbehaviour decreases; 3) the size of the target population decreases; and 4) the extentof the behavioural change required decreases.In a second category of variables, Mazmanian and Sabatier posit a series ofrelationships between effective implementation and the “ability of a policy decisionto structure implementation”. In this case they argue that the chance of successfulimplementation increases if 1) the statute contains clear and consistent objectives;2a)the causal theory on which the policy is based is valid and b) the implementingofficials have adequate jurisdiction over important linkages in the implementationprocess; 3) the level of funding is adequate; 4) “hierarchical integration” is ensuredby a) minimizing the number of veto points, and b) providing incentives for thosewith a veto to agree to implement the policy; 5) the decision rules are unambiguousand facilitate the achievement of objectives; 6) Officials are committed to the goalsoutlined in the policy; and 7) Opportunities for public participation are biased infavour of the policy’s supporters.A final group of non-statutory variables influence implementation. Effectiveimplementation can be attributed in part to 1) favourable local socio-economicconditions that cause the issue to have high salience; 2) positive public opinion andmedia attention; 3) favourable attitudes and and strong financial resources of17constituency groups; 4) support for the program from those who control theresources, (Mazmanian and Sabatier call this actor or group “sovereigns”); and 5)skilled leadership committed to the policy’s objectives.Mazmanian and Sabatier reduce their framework to six conditions that can beused to determine the chances of success of any given policy. The first condition isthat the statute contain “clear and consistent” objectives. Second, the policy must bebased on a sound causal theory and it must grant adequate jurisdiction to relevantagencies to pursue the objectives. (This is, in fact, two separate conditions whichhereafter will be treated separately). Third, the implementation process must bestructured in such a way as to “maximize the probability that implementing officialswill perform as desired”. For Mazmanian and Sabatier, this includes “assignment tosympathetic agencies with adequate hierarchical integration, supportive decisionrules, sufficient financial resources and adequate access to supporters”. (Thisvariable is very similar to Pressman and Wildavsky’s concern with minimizing thenumber of clearance points in decision making). A fourth condition is strongmanagerial and political skill and commitment in the leadership. Fifth, activesupport by key groups and legislatures, and neutral courts facilitatesimplementation. Finally, the priority of the policy’s objectives must be maintainedover time and must not be undermined by changes in socio-economic conditions“which weaken the statutes causal theory or political support” (Mazmanian andSabatier 1983: 41-42).By recognizing the various stages in the implementation process, Mazmanianand Sabatier are better able than previous researchers to identify the variety of forcesthat act on a policy at different times. For example, they note the equal importancestrong initial financial support as well as the necessity of sustained interest fromconstituents over the long-term. Their framework is convincing in the breadth ofvariables that it touches. As they apply their framework to several case-studies its18flexibility becomes apparent. Each policy area is influenced to a different degree bysome or all of their conditions. However, their general hypotheses about therelationships between the variables and implementation appear to be valid.The main weakness of Mazmanian and Sabatier’s theory is that it does notgive sufficient attention to local factors in implementation. Although they can becommended for their discussion of tractability issues insofar as they discuss thedifficulty of changing behaviour, they do not address adequately the issue ofincentives for the target population. In the case of social policy, the importance ofthe structuring of incentives in the policy and the perception of these incentives bytarget groups must not be overlooked, nor can it be lumped together under the“theoretical validity” or “adequate jurisdiction” variables, neither of whichadequately addresses the interaction of policy with the target population. In view ofthis concern, an additional variable should be added to those related to tractability.This variable would help to predict the response of target population to incentivestructures outlined in policies.In addition, although the authors deal with the commitment and skills ofimplementing officials, their bias is toward agency officials closer to the policy ratherthan “street-level bureaucrats” closer to the target population. They neglect issuessuch as the competence, skill, concern and discretion of individuals at the local levelin facilitating program implementation. The first three of these variables are notlikely to be generated through the careful structuring of a statute, and yet they areinstrumental to the faithful implementation of a policy. Mazmanian and Sabatier’stheory could benefit from the addition of a condition relating to the capabilities oflower-level implementors.Since Mazmanian and Sabatier’s work was published, several furtherattempts to develop a theory of implementation have occured and are worth noting.Malcolm Goggin et al.’s Implementation Theory and Practice, and Robert Stoker’s19Reluctant Partners each claim to bring a new perspective to the inquiry. Goggin etal. claim to advance a more scientific approach than Mazmanian and Sabatier, togive operational definitions of variables with specifications for their measurementand to synthesize the top-down and bottom-up approaches. In his work, Stokeradvocates a “regime paradigm” for implementation analysis that looks at differencesamong participants as a resource, rather than an obstacle to effectiveimplementation.Goggin et al.’s work needs only a few comments. Although they present acomprehensive theory of implementation with a wide range of variables, they haveborrowed heavily from Mazmanian and Sabatier-- their variables often differ onlyin name-- and thus their theory seems more a rewriting of previous work than “thethird generation” of implementation literature that they claim that it represents.Their primary independent variable categories are “federal level inducements andconstraints”, and “state and local inducements and constraints”. Also influencingimplementation is organizational structure and state capacity (both political andeconomic).Their “communications model for intergovernmental implementation” is avehicle for explaining the flow of decision-making as the sending of messages, andserves the same purpose as Mazmanian and Sabatier’s feedback ioop, which dividesimplementation into five stages. However, it is more successful than the feedbackloop in acknowledging that messages flow up through the implementation process,as well as down. In fact, the strength of Goggin et al.’s work is that they includeamong their factors several of the bottom-up concerns that Mazmanian and Sabatierignored. In particular, they argue that where administrative flexibility is needed inorder for implementors to achieve policy objectives, the effectiveness ofimplementation might be directly proportional to the degree of decentralization. In20this case, complexity of joint action is a resource, generating “assistance structures”,rather than a liability.Robert Stoker responds more to Bardach than to Mazmanian and Sabatier.He agrees with Bardach that exchanges between actors cause policy to evolve duringthe implementation stage, and that a top-down approach of command and controlcan be used to constrain behaviour that deflects from the policy objectives.However, Stoker has a dramatically different view of the actions of actors. Ratherthan seeing them as self-interested, a view that suggests that individuals operate in a“moral vacuum”, he recognizes that there is more at stake in policy implementationthan the satisfaction of the participants in implementation. “In matters of policy, thewelfare of others -- those who are the ostensible beneficiaries of the policy -- is, orought to be, central” (Stoker 1991: 42). As such, conflict of interest is not somethingthat is “politically illegitimate” that should be suppressed. Rather, Stoker viewsconflict of interest as arising out of a variety of legitimate and sincere approaches toan issue; different interests should be regarded an opportunity (ibid: 29). The properresponse to conflict of interest is to seek cooperation rather than to seek to suppressone view in favour of another. Out of this assertion comes Stoker’s view thatimplementation must not be judged based on how well it satisfies participants’interests. Rather, a good policy continually evolves based on the interactionbetween the public program and citizens.Stoker identifies two parties involved in implementation in which he has aparticular interest. It is the federal program sponsor who must convince anintermediary or “reluctant partner”, (for instance, a state or province, or an agencythat might have different objectives), to implement a program. Each party has theoption either to cooperate or to defect. If both decide to defect, the result is policystagnation. If the intermediary defects and the federal program sponsor cooperates,the program is co-opted by the intermediary. If the federal sponsor defects and the21intermediary cooperates the result is the compliance of the intermediary to federalimperatives. The ideal situation, mutual adaptation, results from the cooperation ofboth parties.According to Stoker, formulators should strive to create a strategic andinstitutional context in which mutual adaptation can be achieved. In order to wishto cooperate, each party must prefer cooperation to defection, and there must be nobenefit to a party in policy stagnation. Stoker asserts that implementation regimescan be designed so as to induce cooperation. An implementation regime is “anarrangement among implementation participants that identifies the values to beserved during the implementation process and provides an organizationalframework to promote these values” (ibid: 55). These regimes resolve a tension thathe sees as implicit in the modem political system -- a tension between centralizationand diffuse authority. Stoker identifies eight regime types, each of which embodiesdifferent values and thus generates a different type of policy.The point of the regime framework is to demonstrate that cooperation cantake many forms and that the central paradox of governance in a liberal, federalpolity can be resolved. This paradox is that in order to achieve their objectives,governments must empower other agencies; however, once empowered thesegroups have the power to undermine the central government’s initiatives. Instructuring the the strategic and institutional context in which the policy isimplemented, policy-makers can protect against this scenario by increasing thebenefits of cooperation for the reluctant partner, and creating a set of rules or normsthat governs the implementation process.An insightful point that he makes, (neglected by previous researchdiscussed), is that cooperation is most likely within the context of an ongoingrelationship. When two parties have a history of cooperation, or when there is anexpectation of future positive interaction, parties are less likely to defect than in the22absence of these circumstances. In the absence of such institutional context, severalstrategic tactics can induce cooperation. These include linking other issues ofconcern to a reluctant partner to the pressing issue; boosting the rewards ofcooperation; dividing large, controversial programs into several smaller exchanges,making each more palatable; or using the commitment of one reluctant partner toinfluence the involvement of another. It is also important for the government tochoose its implementation partners carefully, preferably involving itself with‘participants whose interests reflect and reinforce the existing distribution of power”(ibid: 184).In addition, a favourable institutional context can be created throughattention to process, rules, standards of conduct, oversight of projects andinformation-sharing; through these mechanisms the behaviour of implementationpartners can be controlled to a certain degree. “The regime may alter the costs oftransactions, availability of information of level off uncertainty in the decisionprocess; it may promote cooperation by making the relationship betweenparticipants more regular and predictable” (ibid: 55). The regime is thus a system ofnorms and values that governs the behaviour through incentives, (strategic context),and rules (institutional context).In his work, Stoker does not attempt to develop a comprehensive theory ofimplementation. However, by seeking to determine how cooperation between thefederal government and implementing agencies can be induced he develops andadds to Mazmanian and Sabatier’s fourth variable concerning the ways in which thegovernment can “maximize the probability that implementing officials will performas desired”. Stoker’s “cooperation” focus is an important addition to the study ofimplementation, for it acknowledges the incentive structures acting upon agencies,as agencies, rather than as a collection of Bardach’s self-interested individuals. Itadds a new calculus: that of delayed reciprocity in policy decisions. Stoker’s23understanding of established relationships between agencies helps to explain whyagencies may cooperate even when the immediate costs of this decision are great.Similarly, Stoker’s emphasis on standards of behaviour within the institutionalcontext is important. Like partisanship, which was briefly discussed by Goggin etal., standards allow for predictable parameters of behaviour to emerge within whichpolicy formulators can expect cooperation.Stoker makes a strong case for the mingling of the realms of policyformulation and implementation, arguing that within the regime perspective,implementation is “part of the governing process and so is an arena in which policymust evolve to meet the legitimate concerns of participants” (ibid: 185). As such, hedoes not believe that improvements in policy design can altogether remove obstaclesto implementation. Rather, some flexibility must be allowed within a broadimplementation regime.Policy Breeding Theory: Insights From Workfare ResearchFinally, From Welfare To Work, an in-depth survey of the impacts of variousU.S. “workfare” programs yields a framework to predict implementation outcomesthat is more particular to social policy. In their summary of the findings of theManpower Demonstration Research Corporation, Judith Gueron and Edward Paulyhypothesize that two sets of variables influence the likelihood of effectiveimplementation of welfare-to-work programs.First, a set of variables define “the external context in which the programintervention is implemented which determine[s] the underlying pattern of welfaredynamics” (Gueron and Pauly 1991: 60). These include the labour market conditionsand area characteristics; characteristics of the welfare population; characteristics ofthe welfare program (that is, the way in which eligibility is determined and benefitsallocated); and the other employment and training and support services that exist in24the community which could benefit the control group. These factors determine howsusceptible welfare recipients will be to any program introduced to them, whatincentives influence their behaviour, as well as how appropriate such programs are,given prevailing social and economic conditions in the region.The second set of factors affecting implementation are those which define“the nature and strength of welfare-to-work interventions”. These include thestrength of a program’s funding; the design and content of the program; theexistence of other community employment and training services to which clients canbe referred, thus increasing the size of the case-load that can be served; the existenceof services, such as day care, in the community; the degree of mandatoriness of theprogram; case management and monitoring capacity; and the target population forthe program.The variables which Gueron and Pauly deem as important are those whichdeal with causal theory, that is, with assumptions about the source of the problem,and about how the target population, (both administrators and recipients), willrespond to incentives within the policy. Extensive and accurate knowledge of themotivations and incentive structures of the lower-level players is therefore essentialto the design of a successful policy. This factor is a key part of the bottom-upapproach to implementation research. On the other hand, their concern withfunding and the features of the program model reflects a top-down approach.Finally, they recognize that forces beyond the control of both the policy formulatorsand the target groups, such as labour market conditions and characteristics of thetarget population, can exert a powerful influence of implementation. A usefultheory of implementation must respond to the particular insights gained through anextensive evaluation of a specific policy area. Although its conditions are too tied toa specific policy arena to be used as a comprehensive theory, Gueron and Pauly’swork will help to complete Mazmanian and Sabatier’s framework.25The Preferred Framework and Suggestions for ModificationOf the eight works, Mazmanian and Sabatier’s study is most useful in itscomprehensiveness, its simplicity and its breadth of coverage of all policy areas. Itsthree-part framework encompasses the concern for clarity in the statute, theoreticalvalidity, and simplicity in organization introduced by Pressman and Wildavsky. Inaddition, it recognizes the importance of individual motivation, as Bardach andStoker do. And, like Elmore and Lipsky, Mazmanian and Sabatier’s theory leavesroom for consideration of street-level players who influence the implementationprocess, although they do not give the attention to this perspective that Goggin et Finally, they include an entire category of variables which are absent from theanalyses of the other authors. The issue of the tractability of the problem is barelytouched upon by the other researchers with the exception of Gueron and Pauly.This is perhaps because it is beyond the control of those involved in the policyprocess, and therefore is not a problem that can be “solved” by the authors.Nevertheless, its relative absence as a set of predictive factors in the other worksonly yields inadequate portraits of the forces at work in the policy environment.One factor not addressed by any of the authors is the importance of “learning’in policy-making. Although Mazmanian and Sabatier provide a “feedback loop” intheir model, implying that policy-makers will change a policy when appropriate, inreality, such changes are extremely difficult to make formally, given the intenselegislative activity which this would necessitate. A policy which is truly able toadapt is one in which the mechanisms for both learning which changes should bemade, and then making the changes are included in the original policy design. Inother words, a policy that has both a method of evaluation included in its designand the flexibility to allow periodic changes to its content is more likely to beadaptable and to respond to changing conditions or to past mistakes. Such a policyis more likely to meet with success than a policy without such attributes.26The application of Mazmanian and Sabatier’s framework to Canada’s policyenvironment requires that a few modifications be made to their conditions ofeffective implementation. In his paper “Public Policy Implementation: AnAssessment of the Applicability to Canada of the Mazmanian-Sabatier Model”Stephen Phillips argues that the “comparative imperviousness of the Canadian stateapparatus may offset to some degree the significance of such factors as the paucity offixers in the implementation process and the lack of clearly stated statutoryobjectives and duties” (Phillips 1992: 13). Similarly, since outsiders have less accessto participation in the process of policy implementation in Canada than in theUnited States, this variable can be eliminated from a Canadian framework. Finally,the relative autonomy of the provinces compared the states makes the federal task ofachieving cooperation more difficult. This increases the relative importance ofMazmanian and Sabatier’s variable concerned with “maximizing the probability thatimplementing officials will perform as desired” in cases requiring federal-provincialagreement.A new framework, more appropriate to the Canadian environment and tosocial policy, adds two new categories of factors to Mazmanian and Sabatier’sframework, eliminates one of their categories, splits their second category into twoseparate categories and changes slightly the content of other categories. A Canadianframework omits the importance of the access of supporters to the policy process, ofthe courts and of key legislators, and places political variables such as the support ofconstituency groups into Mazmanian and Sabatier’s final category, “the relativepriority of statutory objectives”. In this category variables such as media supportand conflicting policies would determine the place that the initiative would have onthe government agenda and therefore the likelihood that its implementation wouldproceed with considerable support. To this category is also added Gueron’s27“external” factors, including the characteristics of the target population and of theeconomic conditions prevailing in the region in which policy is implemented.Three new perspectives complete a revised version of Mazmanian andSabatier’s framework. First, out of Robert Stoker’s work, a history of cooperation orexpectations of future cooperation, as well as favourable institutional contextinfluence the probability that implementing officials will perform as desired. Thesevariable are added to Mazmanian and Sabatier’s fourth category concerned withfostering cooperation. Second, out of the work of the bottom-uppers, the variable“incentives for the target group to participate” is included in order for the theory topredict whether the legislation will be able to induce the target group to engage inthe desired behaviour. This is separate from Mazmanian and Sabatier’s categoryconcerning “sufficient jurisdiction” which deals with organizational issues asopposed to how the needs of individuals are met by a policy. Also out of the workof Elmore and Lipsky, Mazmanian and Sabatier’s category concerning the skill of theleaders of implementation is expanded to include the skills and attitudes of street-level workers. Finally, the framework includes a condition that the process of“learning” be facilitated by policy design. The revised framework reflects morecompletely the many forces involved in the implementation process than doesMazmanian and Sabatier’s 1986 theory. The revisions and additions are in boldtype:1. The enabling legislation or other legal directive mandates policy objectives whichare clear and consistent or at least provides substantive criteria for resolving goalconflicts.2. The enabling legislation incorporates a sound theory identifying the principalfactors and causal linkages affecting policy objectives.3. The enabling legislation gives implementing officials sufficient jurisdiction overtarget groups and other points of leverage to attain, at least potentially, the desiredgoals.284. The enabling legislation structures the implementation process so as to maximizethe probability that implementing officials and target groups will perform asdesired. This involves assignment to sympathetic agencies with adequatehierarchical integration, with a history of cooperation, or expectations of futurecooperation, supportive decision rules, sufficient financial resources and adequateinfrastructure.5. The leaders and street-level workers of the implementing agency possesssubstantial managerial and political skill and are committed to statutory goals.6. The relative priority of statutory objectives is not undermined over time by theemergence of conflicting public priorities, changes in constituency groups’ supportor by changes in the relative socio-economic conditions which weaken the statute’scausal theory or political support. External conditions are conducive to theachievement of the policy’s objectives.7. The enabling legislation and the institutions and norms created by it providethe target group with sufficient incentives to engage in the desired behaviour, andremove any disincentives to target group compliance created by factorsboth”external” and political that define the experience of the target group.8. The enabling legislation ensured that “learning” could take place and that theimplementation process was flexible enough to incorporate past lessons.Two decades after the Pressman and Wildavsky’s study, implementationliterature retains its earnest and yet detached character. There has emerged a greatersense of the process that follows policy formulation, as a complex whole involvingmany actors and socio-economic forces. Even the main division in the literature,between the top-down and the bottom-up approaches to implementation analysis, isdisappearing as researchers forge a synthesis between the two by adopting elementsfrom each type of framework in order to present a complete picture of the forcesaffecting public attempts to solve problems. The danger that these researchers faceis in including too many variables; indeed, what is needed is not a descriptive studyof implementation but a prescriptive theory -- one that can account for a broad rangeof policy consequences and can serve both policy-formulators and “street-level”citizens to design and manage solutions that will be effective. Through the dialogueof a number of scholars such a theory begins to take shape.2930Chapter TwoThe Context of the Work/Welfare ProblemImplementation is the translation of a policy idea into results. Althoughtheories and strategies abound, very few paper solutions manage to achieve theirobjectives in practice. Nowhere is this more evident than in the realm of socialwelfare policy, where for decades the ‘war on poverty” has been foughtcontinuously with innovative attacks that have resulted in little more thansuperficial changes to the well-being of welfare recipients or to the financial burdenof governments. These shallow forays into the deepening crisis of poverty have, iflittle else, yielded lessons about the complexity of the problem and have broughtinto focus the tangled web of influences acting upon low-income North Americansand upon the policy-makers trying to change the circumstances of the poor.The implementation of a welfare initiative must be understood in terms of thecontext in which it occurs. As background to chapter three’s case-study of aparticular initiative to enhance the employability of welfare recipients in BritishColumbia, this chapter will outline the various contexts in which such a welfare-to-work initiative is implemented. Three contexts are particularly important. Theyeach relate to the perspectives of one of the three key players in the social policyrealm. First there is the context of public debate, relating to the balance of pervasivebeliefs in society. Second is the context of the policy-makers, related to their need toachieve multiple, sometimes conflicting, objectives. A third context is that in whichsocial assistance recipients operate, specifically related to their personalcircumstances and to the incentive structures acting upon them through existingpolicies. Finally, the welfare-to-work issue must be seen as just one problemarising from the greater structural context of society that perpetuates conditions ofpoverty even for those working full-time in low-wage employment. This context is31too large, too entrenched and too expensive to change for the SAR Agreement tohave any possibility of redefining it.The interaction of the public, policy-maker’s and social assistance recipient’scontext have led to numerous attempts to resolve the work/welfare dilemma. Atthe end of this chapter a brief overview of the past history of policy successes andfailures in this field illuminates the issues with which policy-makers must inevitablywrestle, and foreshadows the examination of the B.C. initiatives in chapters 3 and 4.The Work/Wefare Debate: An OverviewThe social policy debate has always concerned the interaction of two groupswithin the population: those who earn an income and pay taxes which aretransferred to others, and those who do not earn their income but receive transfers.The key difference between these two sets of players is the fact that for one groupwork is the norm, while for the other group it is not. Traditionally, such a split inroles was acceptable. The poor were considered to be “deserving”, that is,comprised of widows, or handicapped individuals, who, through no choice of theirown were incapable of supporting themselves. Over time, however, thecharacteristics of those on welfare have changed, and with them, the legitimacy ofthe notion of transfers without a reciprocal obligation from the recipient has beenundermined.Between 1980 and 1988 the percentage of the Canadian welfare case-loadmade up of “employable” individuals -- those under 65 years of age having neitherphysical impediments, nor sole responsibility for young children -- grew from 35 to46 percent, even as unemployment fell throughout the 1980s (CLMPC 1990: 136).Increasingly, transfers were going from those who worked to those capable ofworking but who did not. As the case-load sizes grew so did the financial burdenon taxpayers. Politicians began searching for ways to reduce the numbers of32‘employables” on welfare; they soon found themselves embroiled in one of the mostintense debates in the domestic politics of Canada and the United States-- thework/welfare dilemma. The controversy surrounds the new element that has beenadded to the transfer bargain, that of reciprocal obligation. Employable socialassistance recipients are no longer regarded in some jurisdictions as being Ientit1edHto benefits, without some effort on their part to move towards self-sufficiency.Welfare-to-work programs are the vehicle that government has designed to facilitatethe fulfillment of reciprocal obligations, placing these programs in the eye of thework/welfare storm.Although the “welfare-to-work” program used to conjure up visions of“workfare”, (the tying of unpaid work to the receipt of welfare benefits), the phrasehas now come to be associated with any kind of work-related activity that makes asocial assistance recipient more job-ready. As such, job-search clubs, compulsoryeducation known as “learnfare”, training, work experience and community work arejust a few examples of frequent attachments of work to welfare. Welfare-to-workprograms are variously voluntary or coercive depending on the features of policydesign. According to Patricia Evans, there are four elements of welfare-to-workprograms that determine the degree to which they impose an obligation on welfarerecipients. They can be summarized by the following four questions: Who is to betargeted? What must they do? What is the extent of the monitoring? Are theresanctions for non-compliance and will they be applied? (Evans 1993: 57). Thesequestions are central to the debate surrounding policy design, and are to beanswered in different ways, depending on which perspective one takes, whetherthat of the public, the policy-makers, or the social assistance recipient.33Public Context: A Battle of Social Constructions and IdeologiesThe facts of poverty are not hard to grapple with. That poverty has ill effectson the health and educational achievement of children is accepted. That poverty iscommon for certain groups -- for example, single parents throughout NorthAmerica, and blacks and Hispanics in the United States-- is a matter of statistics.That social assistance benefit levels rarely lift recipients above the local poverty lineis well-documented. Society agrees, in general, about the devastating effects ofpoverty on the individual and of welfare dependency on government’s budget.However, in spite of this shared understanding of the effects of poverty there ismuch less consensus about its causes and therefore a tremendous debate over theappropriate policy initiatives to attack both dependence and poverty.In the absence of a well-defined causal theory for poverty, support for variousapproaches to poverty-reduction and dependency-reduction are founded instead onthe public’s perception of the source of the problem. The public constructs theirview of poverty issues based on their perceptions of the key players involved. Intheir work, Anne Schneider and Helen Ingram argue that target populations areeither positively or negatively constructed. This social construction, as well as thestrength of the group in terms of resources, determines the distribution of rewardsand punishments that flows from governments to target groups (Schneider andIngram 1993). The social construction also affects the types of policy tools that areperceived to be useful in influencing the behaviour of these socially constructedgroups. Schneider and Ingram identify five different types of tools -- “authority”,“incentive” and “capacity” tools which progressively replace control withopportunity in order to control behaviour; “symbolic and hortatory” tools whichappeal to the target group’s system of beliefs, and “learning” tools which allow forexperimentation to determine the best policy approach to a problem (Schneider andIngram 1990). Each of these tools has a place in welfare-to-work strategies.34The social construction of the poor in a particular policy environment has asignificant effect on the choice of policy tools used to influence their behaviour. Forinstance, in the United States the poor are negatively constructed as inner-city blackunwed mothers with no political resources. As a result they receivedisproportionate punishment through tools of the “authority” type.Part of the social construction of these groups is that they respondmainly to punishment... .Persons in powerful groups are constructed asintelligent and able to make good choices. Powerless groups are notusually constructed this way but are viewed as needing direction(Schneider and Ingram: 1993: 340).In Canada, by contrast, the poor population is not trapped in a stereotype tothe same extent that American welfare recipients are. Canada’s poor population isperceived as spanning all races, ages, regions and family types and both sexes. As aresult, welfare recipients are constructed less negatively in Canada than in theUnited States, although they have traditionally been viewed as weak, politically.The result of a weak and relatively positive construction have for yearsbeen”Symbolic policies [that] permit elected officials to show great concern butrelieve the need to allocate resources” (ibid: 338). As such, before the mid-1980s,policies emphasized financial and service incentives. However, they were often sopoorly funded that they are of little value.As the percentage of employables in the welfare case-load soared throughoutthe 1980s, the political imperative to reverse this trend grew and consequently thepoor, by their sheer numbers, had a greater potential to cause political upheaval.The political strength of the poor, as a target group for policy, increased. Symbolicpolicies targeted at the poor would no longer pacify the public in their demands foreither anti-dependence or anti-poverty measures. The government began to searchfor tools to direct at a stronger, slightly positively constructed target group in order35to change their behaviour. By mid-1980s, effective incentive and capacity tools weresought to address the work/welfare dilemma.Policy tools are implemented within the context of broad public ideologieswhich interact with the social constructions described by Schneider and Ingram.Ideology both conditions, and is fashioned by, prevailing perceptions about groupsin society. Several different ideologies prevail today in Canada and the UnitedStates. In “The Family Support Act of 1988: Federalism, Developmental Policy, andWelfare Reform”, Mark Rom takes as his starting point the redistributivedevelopmental split in governments’orientation towards welfare:Developmental policies attempt to enhance the economiccompetitiveness of a political jurisdiction, while redistributive policiesfocus government spending on some specific, typically needy, group.Redistributive policy takes from some and gives to others, while indevelopmental policy, those who pay the costs reap the benefits (Rom1989: 61).When the need for reciprocal obligation between government and welfare recipientswas introduced in Canada, the country’s long-time focus on redistributive policywas overshadowed by an emerging developmental emphasis.Two notions of developmentalism now serve as the poles between whichwelfare-to-work initiatives are pulled in the work/welfare debates. Developmentalpolicy is any government initiative which “emancipates recipients from the welfarerolls and brings them into the productive workfare so that they pay taxes, ratherthan receive tax dollars” (ibid: 62). However, developmentalism has a verydifferent look depending on whether it is in the hands of conservatives or liberals.Developmental strategies are not merely administrative frameworks. They arecomplete philosophies about human nature and motivation.Conservative developmentalists believe that individuals will not work unlessthey are forced to, and that the best strategy for reducing welfare dependence is toremove any benefits that the government provides, thus forcing individuals to36provide for themselves, or starve (ibid: 63). With such incentives, conservativedevelopmentalists claim, individuals will become self-sufficient, will have anincreased sense of dignity which arises from conforming to the Protestant WorkEthic. In the long term, conservative developmentalists argue, reduced welfaredependency will lead to reduced poverty since individuals will learn to hold evenlow-wage, low-skill jobs whose wages exceed welfare benefits in most states (Morrisand Williamson 1987: 950). The flaw in this argument is that conservativedevelopmentalists do not take into account the value of non-cash benefits, such assubsidized housing and medicaid, that social assistance recipients lose when theyleave welfare for low-wage employment, which is often without benefits. Such atransition to work may in fact make recipients poorer.Beyond the financial arguments, conservatives’ concerns are ideological.Mickey Kaus, in his controversial 1986 article “The Work Ethic State” claims that:The long run savings, of course, would be huge if the welfare culturewere absorbed into the working, taxpaying culture. In the short run,however, welfare savings would be less... .The point isn’t to savemoney. The point is to enforce the work ethic (33).Through the enforcement of the work ethic they envision that other social problems,such as high illegitimacy rates among welfare mothers, will fade away.On the other hand, liberal developmentalists promote anti-poverty measuresahead of those emphasizing reduced dependence on welfare. They believe thatindividuals want to work and that they merely face barriers, imposed by theconditions of their poverty, to gaining and maintaining employment. Liberaldevelopmentalists “believe that providing incentives to welfare recipients is the bestpolicy , at least in the long run, to get people off welfare, save the governmentmoney, and promote economic growth” (Rom 1989: 63). They are advocates ofgenerous education and training programs and of the provision of incentives towork within the welfare benefit structure. They believe that it is society’s37responsibility to build human capacity, and that reducing benefit levels withoutproviding meaningful opportunities for individuals adequately to supportthemselves can only reduce human capacity and drive welfare recipients deeper intopoverty, both economic and personal.The two disparate ideologies have one area of agreement. In both strategieswork and training opportunities can play a significant role in achieving thestrategies’ objectives. For both liberal and conservative developmentalists, work isthe preferred state of affairs for individuals. Conservatives champion thereinforcement that work brings to the values of individualism and autonomy,whereas liberals prefer work for its power to integrate individuals into thecommunity, to reduce stigmatization and to provide dignity. However, the contentof these welfare-to-work initiatives, and degree to which work should be tied towelfare are tricky matters over which the two groups clash.Conservatives and liberals disagree about which welfare-to-work activitiesare valuable. While liberals support initiatives to increase the capacity ofindividuals, through education, training experiences and job-preparedness classes,they adamantly oppose workfare of the “make-work” variety, in which the jobprovides the recipient with no useable skills and no sense of accomplishment.Conservatives, on the other hand, believe that it is the act of working itself,regardless of the quality of the work, that is important. Similarly, conservative andliberal developmentalists have very different views of the value of coercion. Eachgroup favours certain policy tools based on their social construction of the poor.Conservatives favour “authority” tools to influence the poor to leave welfare. Suchstrategies include limiting eligibility for benefits, tying benefits to work, reducingbenefit levels to force social assistance recipients to work and enforcing a penalty fornon-compliance. Liberals tend to favour “incentive” and “capacity” tools. Incentivetools increase the financial incentives for social assistance recipients to work by38allowing them to keep more of their earnings, for example. Capacity tools aredeveloped out of a recognition that welfare recipients face many severe obstacles toemployment; they attempt to help them to overcome these obstacles through suchprograms as training and education, as well as day care and transitional supports.These programs serve to “encourage” welfare recipients to work through extendinghelp without touching benefit levels.In Canada, compulsion to do work in exchange for welfare benefits is illegal.The Canada Assistance Plan (CAP), the nation’s principal welfare legislation, sharesthe cost of social assistance with the provinces provided that eligibility is based onneed alone, and that “no person shall be denied assistance because he refuses or hasrefused to take part in a work activity project” (Canada Assistance Plan, article 15-3a).However, while unpaid work for welfare benefits is illegal, “CAP does not object to arequirement for work-seeking activities as a part of a process of determining need: arefusal to seek work may be taken as evidence that there is in fact no need”(ibid:121). Canada’s policy is thus a mixture of encouragement and with a small amountof coercion. The federal government’s strict prohibition against funding “workfare”initiatives has kept policy-makers focused on creating incentive and capacity-building tools, responding to, and reinforcing the liberal developmentalist ideologyin the country.The Policy-makers’ Context: Political Constraints and Conflicting GoalsPolicy-makers are influenced by the ideology which is pervasive in theirparty, agency, or constituency. However, they are influenced even more strongly bya context which is particular to their roles as policy-makers, as those who define theproblem to be solved and attempt to solve it by defining and pursuing goals.Deborah Stone argues that problem-definition, that which helps to create theparticular context in which the policy-maker works, is an intensely political process:39In the polis...problem definition is never simply a matter of defininggoals and measuring our distance from them. It is rather the strategicrepresentation of situations. Problem definition is a matter ofrepresentation because there is no objective description of a situation;there can only be portrayals of people’s experiences andinterpretations. Problem definition is strategic because groups,individuals, and government agencies deliberately and consciouslydesign portrayals so as to promote their favoured course of action(Stone 1988: 106).Problem definition creates the context in which a policy maker functions. Inthe work/welfare debate, problem definition often flows from ideologicalorientation, but this is not always the case. Although liberals tend to define thework/welfare problem as one of poverty and structural barriers to employment, aspolicy-makers seeking re-election or promotion they must also be concerned withlessening government’s burden through a reduction of dependence. In addition,through the legitimating power of work, they seek to improve popular support forwelfare by ensuring the worthiness of the clientele (Evans 1992: 58). Conservativessee the problem more as one of excessive dependence on welfare, fiscal strain, and apoor work ethic in the target group. However, at the same time most moderateconservatives recognize the need for the government to finance welfare-to-workactivities and transitional supports, involving an increase in spending, in order toachieve their goal of dependence reduction. Both conservatives and liberals sharethe goals of increasing overall employment, and of increasing the dignity of welfarerecipients by bringing them into the mainstream (ibid).The main split in goals, between dependence-reduction and povertyreduction, results from different definitions of the work/welfare problem. This splitis particularly troubling for policy-makers. Since most initiatives require broadsupport in order to be implemented, a welfare-to-work program must satisfy thegoals of as many of the major players is possible. The goals of dependencereduction and poverty-reduction are difficult to satisfy simultaneously and often40appear to stand in conflict with one another. For example, to reduce poverty in theshort-term means to increase the benefits of training for social assistance recipients.Increased benefits reduce the incentives of social assistance recipients to leavewelfare, since the rewards of work become less appealing. Similarly, generousbenefits might attract the working poor to leave work for welfare. Anti-povertymeasures thus could have the effect of increasing dependence (Weaver 1994: 19). Onthe other hand, measures to reduce dependence are those which decrease therewards of welfare relative to work. These measures generally take the form ofreduced benefits, deepening the poverty of the welfare recipient.Dependence-reduction and poverty-reduction can be reconciledsimultaneously in the long-term. Conservative developmentalists argue that ifindividuals are forced off of the welfare rolls they will become self-sufficient; asindividuals become more disciplined and hard working they will lift themselves outof poverty. Liberal developmentalists, on a similar note of optimism, claim that ifthe government reduces the poverty of individuals it will remove many of thebarriers to employment that these individuals face. As these people’s lifecircumstances improve they will be better able to enter the workforce and tomaintain employment. They will experience an increase in earnings which willdecrease their dependence on welfare.The main problem with pursuing either of these courses in view of thepositive long-term results is the short-term consequences. The liberaldevelopmentalist course would be enormously expensive. The conservativedevelopmentalist strategy would deprive Canadians of any safety net and wouldleave the fate of the poor to the whim of charity. Neither option is politicallyfeasible. As a result, policy-makers must attempt, as best they can, to reconcile twocompeting objectives, aware all the while that in the short-term one objective cangenerally only be achieved by sacrificing the other.41In addition, policy-makers can not afford to wait for long-term successes.Policy-makers depend on others recognizing their achievements in order to securetheir chance for re-election, for retaining their jobs, or for promotion. In general theydo not have the job-security to wait for long-term results. They are undertremendous pressure to demonstrate early that their policies are effective. Thenegative consequence of this reality in the welfare-to-work field is that it can lead topressures to “cream”, that is, to spend resources on those who would be most likelyto succeed on their own, in order to generate favourable statistics and thus keepprogram support high. A program in which creaming is taking place is unlikely toreach the target population that most needs the help it provides. That policy-makersmust pursue short-term achievements limits their ability to develop effectivewelfare-to-work programs.Finally, and most important to understanding their particular context, is therecognition that policy-makers are limited in terms of how they can define thework/welfare problem. They must attempt to keep the problem distinct from otherproblems which are beyond their jurisdiction, or their means, to influence. To bemore specific, one promising way in which to reduce both poverty and dependencesimultaneously is to define the work/welfare problem as “the inadequate conditionsand rewards of low-wage work”. In 1986, a welfare recipient with dependentswould had to have to earned an income of $8.OO/hr, far above the provincialminimum wage, in order to be better off working than on welfare (Evans 1988: 132).This creates a “perverse incentive” for individuals to choose welfare over work(Weaver 1994: 19).There is an invisible wall that welfare reform scales but is wary ofsurmounting. This wall, well-recognized by those policy-makers familiar with theway incentives within the welfare system influence behaviour, is the current state oflow-paid work in Canada. The conditions of work include a low minimum wage,42few benefits provided to workers in the labour force and discrimination faced bymany workers for reasons of gender, race, or physical disability. What is necessaryin order to remove the perverse incentives of welfare is a complete reconstruction ofthe wall -- entailing a higher minimum wage, subsidized child care, health care andhousing for the working poor and legislation that ensures pay equity for all workers-- a far too expensive proposition for either Canadian or American governments toconsider. So long as this is true, welfare reform will always have boundaries.The social assistance recipients’ context: the pushes and pulls of wefare rulesWelfare is more than a cash-transfer. It is a complete network with thepotential to govern an individual’s behaviour through the incentives that it provides.It is the “safety net” of last resort, meaning that it provides income to those who haveexhausted all of the accepted means of supporting themselves. The primary featureof the system is therefore income. It is assumed that it is to individuals’ benefit tohave the greatest income possible and that individuals will respond to incentives inthe welfare system which allow them to be financially better-off. The typical NorthAmerican welfare system is made up of several components that influence theamount of income that individuals receive, as well as what the source of this incomeis, whether transfers or work. Governments manipulate these components in theirattempts to influence the behaviour of social assistance recipients, through changingtheir incentives.First, government decides who will be eligible for welfare based on the levelof their assets. For example, in British Columbia, single individuals are not beeligible for welfare if their assets exceed $2500, and families are ineligible if theirassets exceed $5000 plus $500 per dependent (G.A.I.N. Regulations 1993).Governments also determine whether or not welfare recipients will be expected towork. “Employables” are separated from “unemployable” persons. Unemployable43individuals in British Columbia are those over 65 years of age, individuals with amental or physical infirmity that renders them incapable of accepting employment,and women who bear sole responsibility for young children. Those who are deemed“employable” generally receive lower welfare payments than “unemployables”andare subject to work-seeking requirements in British Columbia, meaning that if theyfail to demonstrate that they are making “reasonable efforts to secure employment”their benefit level or eligibility might be affected (G.A.I.N. Act Section 18.4). This is apolicy tool in the “restrictive” vein.Manipulating this last element involves a particular moral dilemma. First, indetermining eligibility, policy-makers are passing judgement on the worthiness ofeach individual to receive support. Handler and Hasenfeld write “The fundamentaldistinction between the deserving poor and the undeserving poor is whether acategory is morally excused from work” (1991: 229). While the distinction isrelatively simple when applied to persons with severe physical or mentaldisabilities, it becomes controversial when applied to single mothers, for example,who until recently have been regarded as worthy of society’s support due to theirrole of raising children. As more women have entered the labour force it hasbecome less clear why the state should continue to support some mothers and notothers. As single mothers increasingly are regarded as undeserving they will becategorized as “employables” and thus find their benefits reduced in many cases,and the pressure on them to work, increased.If a work requirement requires work experience, training or placement, anintervening force acting upon welfare recipients is the availability of such positions.The adequacy of the position, in terms of providing financial and in-kind benefits,influences the participants’ willingness to comply with the requirement, as does thedegree of monitoring and the likelihood and magnitude of the sanction for failing toparticipate.44Another common reform to encourage welfare recipients to work is to alterthe rewards of work. This “incentive” strategy has three parts. First, an EarningsExemption designates a fixed amount of earnings from work that a welfare recipientis allowed to keep. For example, previous to March 1992 British Columbia allowedsingle social assistance recipients to keep $50 of their earnings per month, whereas afamily could keep $100. In 1992 these exemptions were doubled (Bailey 1994: 2).The more income from work that welfare recipients are allowed to keep without areduction in their benefit levels, the more likely they will be to work. Second, thelower the tax-back rate, that is, the amount that the government takes of the incomethat social assistance recipients earn in excess of the flat rate, the more likely theindividual will be to work. Tn 1986 British Columbia reduced its tax-back rate from100% to 75%, meaning that instead of losing one dollar in benefits for every dollarearned through working, welfare recipients now can keep 25 cents out of everydollar earned above the flat-rate exemption. This is called the “enhanced” earningexemption. A third component in the financial strategy is the waiting period beforethese exemptions are applied. The shorter the waiting period, the more likely thesocial assistance recipient will be to work. fri 1986 British Columbia reduced itswaiting period from eight months to three, encouraging welfare recipients to worksooner. These three elements are continually manipulated by government in orderto tempt welfare recipients to enter the workforce or to increase their hours of work.Beyond cash transfers, the social assistance system also provides a variety ofin-kind transfers, which vary from one jurisdiction to another but often includesubsidized day care, medical and dental coverage, transportation and clothingallowances, and moving allowances. These transfers have no cash value but increasethe well-being of welfare recipients. In-kind transfers provide a major disincentiveto individuals considering leaving welfare to take a job since most low-paying,temporary or part-time jobs do not provide such benefits, thus transferring a45considerable and usually unmanageable expense to the income assistance recipient.This is especially the case for single parents whose day care expenses would likelyconsume more than the increased income that they would receive from working. Inmany jurisdictions, including British Columbia, policy-makers have tried to erasethis disincentive by providing these services during the first year of transition intowork. However, this reform merely delays inevitable costs for welfare recipientsand thus does not change the work-or-welfare calculus for most recipients. A true‘capacity” strategy with a real likelihood of success would ensure that essentialservices would be guaranteed in low-paying jobs.A final force influencing individuals’ decision to participate in a work-relatedactivity is the capacity of the individuals themselves. For example, if someindividuals lacks basic literacy skills, work skills, social skills, or even the confidenceto meet with an employer or trainer, the incentives in the system will be meaninglesswith respect to whether welfare recipients are capable of executing the activitiesrequired of them in a job.This last factor that has been receiving increasing attention in both Canadaand the United States. As such, “workfare” has taken on its broader meaning as anyactivity that prepares a welfare recipient for work. Canadian provinces andAmerican states have been experimenting with a variety of “capacity”-buildingapproaches -- beyond their conventional authority and incentive strategies. Theseinclude training, education, and work experiences, designed to provide the welfarerecipient with both the skills and the confidence to become self-sufficient, withoutthreatening welfare benefits. These strategies are becoming commonplace in mostwelfare-to-work designs.46Welfare-to-Work Experiments: Modest Successes and Inevitable Trade-offsEvery jurisdiction with a welfare system has as part of its design a strategy tohelp recipients to leave the welfare rolls. A major component of these strategies arework-related activities, whether mandatory or voluntary, costly or inexpensive,targeted or broad-coverage. Relatively little is known about the design andeffectiveness of Canadian programs. However, in the United States, work andtraining programs have been the subject of extensive study. Since the Americanexperience can serve as a model for Canadian experiments, the American findingsconcerning the successes and failures of work and training programs provide auseful background for a study of the fate of Canadian implementation of suchprograms.The most extensive American studies to date were conducted by theManpower Demonstration Research Corporation (MDRC) whose findings from its“multi-state field research project of State Work/Welfare Initiatives”, aresummarized by its president, Judith Gueron, along with its Senior ResearchAssociate, Edward Pauly, in the book From Welfare to Work. The authors aregenerally positive yet modest in their assessment of the accomplishments ofwork/welfare initiatives in the 1980s, noting that there is a great variety in thedesign of programs, in the local context in which they are implemented and in theirrange of impacts. However, they are able to draw several overarching lessons fromthe experiences of the 1980s.Gueron and Pauly reach a number of conclusions based on their evaluation ofa largenumber of American welfare-to-work programs. They conclude thatwelfare/work programs, if adequately funded, are capable of reaching a largenumber of welfare recipients, and that participation rates are generally high inbroad-coverage programs. Many programs result in both employment andearnings increases for social assistance recipients, the latter through increased hours,47rather than increased wages. Voluntary, selective, higher-cost programs dooccasionally allow recipients to secure better jobs with higher pay. Those who aremoderately disadvantaged experience the largest increase in earnings. However, thelargest savings to government result from the increased work effort of the mostseverely disadvantaged recipients. There was no significant earnings or welfaresavings impact produced by those who are most job-ready. Government’s savingsgenerally exceed their expenditures over the long-term, with public investmentusually returned within five years, in terms of increased tax revenue and a reductionin welfare transfers. Finally, the researchers confirmed that the multiple goals ofwelfare-to-work programs-- providing earnings gains for participants, creatingwelfare savings for government, and reducing long-term dependence of welfare --can often not be achieved simultaneously within a single program and thereforeconstitute a trade-off.A continuous stream of innovative programs is being developed in bothCanada and the United States in an attempt to arrive at the best mix of authority,incentive and capacity tools to influence social assistance recipients to leave thewelfare rolls for work. Despite the considerable theoretical debate pervadingCanadian literature on the place of work in welfare reform, there is actually verylittle empirical research of the type done in the United States to determine thesuccess of the developmental strategy as it continues to evolve in its Canadiancontext. In the next chapter, the case of British Columbia’s welfare-to-workprograms will be examined, both to explore the transformation of the developmentalideal of the attachment of work to welfare in a nation in which the notion ofentitlement still lingers, and to test several hypotheses about implementation in theCanadian policy environment.48Chapter ThreeA Modest Success Story: The Canada/B.C. SAR AgreementIn both Canada and the United States the developmental strategy for welfarepolicy, both liberal and conservative, is being realized as welfare-to-work initiativesbecome central to federal and regional agendas. At the same time, the philosophicaldifference between the two countries is reflected in the tone of the programs. For themost part, welfare-to-work programs are voluntary in Canada, where workprograms are sold as “opportunities” which should have no bearing on entitlement.In the United States, states frequently implement mandatory programs in which therecipients’ “responsibility” to participate is tied to their entitlement to socialassistance benefits. Nevertheless, three values ever-present in modem welfaredebates, are paramount in two countries’ growing reliance on welfare-to-workinitiatives. They are outlined by Golding and Middleton in their description of theimagery of welfare:Three key ideas formed the tripod on which public understanding ofpoverty and welfare rested. These were.. .the efficiency of the labourmarket and the economy; the morality of the work ethic and selfsufficiency; and the pathology of individual inadequacy as the cause ofpoverty (Golding and Middleton 1982:478).Although the majority of Canadians cite “circumstances” as the primary cause ofpoverty, as opposed to the majority of Americans who cite “lack of effort” on the partof the poor, the tripod of values outlined by Golding and Middleton have a clearplace in the Canadian welfare state’s emerging liberal developmentalist ideology(Lightman 1991a: 13).This chapter will address the ways in which these values have been expressedand modified in Canada’s recent, and most ambitious welfare-to-work initiative, theFederal/ProvincialiTerritorial Agreement to Enhance the Employability of Social Assistance49Recipients (The SAR Agreement). The context from which this initiative emerged,the design and implementation of the programs that flowed from the agreement,and their impact, all reflect the growing emphasis on developmental values over thestrictly entitlement-based approach. In addition to providing an overview of theinitiative at a national level, this chapter will focus on the implementation of theagreement in British Columbia -- the first jurisdiction to sign the accord-- todetermine the factors relevant to the success of the program, and to the significantcomponents of the design and policy climate which limit the effectiveness of theinitiativeThe Roots of ReformThe need for measures to improve the employability of welfare recipients inCanada arose out of a convergence of economic factors in the early 1980s. Severeeconomic recession in 1982, rapidly changing labour market needs, soaringunemployment rates, and growing welfare case-loads warned the government of aneed for dramatic reorientation of its labour market policy. As the recession eased, aperiod of “jobless” recovery failed to bring many of those who had lost their jobsback into the marketplace. Even as unemployment levels dropped, the percentageof “employable” individuals on welfare grew, demonstrating that measures beyondthe cycling of the economy were necessary to reduce the dependence of“employable” individuals on social assistance.The strategy adopted by the Conservative government to deal with thechanging labour iriarket was an example of what Mike Prince and James Rice havetermed “supply side social policy”. Rather than involving itself in creating a demandfor jobs, the government’s gamble in 1985 was that by increasing the quality ofworkers themselves the labour market would then be able to absorb these workersand meet the growing need for a skilled labour force in Canada. As opposed to50continuing the patchwork of job creation and training programs which had existedin the past, described by the Neilson Task Force as having “often been of the nature and [having] been insufficiently coordinated with the private sector andthe provinces”(1985: 19), the government sought a comprehensive strategy in which the variety ofneeds of workers, business and the regions could be both recognized and met.Prince and Rice explain that the Mulroney government’s strategy was areflection of its four major priorities of national reconciliation, economic renewal,social justice and constructive internationalism. The Canadian Jobs Strategy,announced in June of 1985, reflected Mulroney’s concern for national reconciliationin its emphasis on federal-provincial cooperation and respect for national diversityof needs. The second theme, economic renewal, could be seen in the government’semphasis on job creation in the private sector and its support for competition andentrepreneurship. Third, the social justice priority was reflected in the government’srecognition that certain groups of the unemployed faced larger than average barriersto employment that had to be addressed through participation targets. Finally, thegovernment was aware of the need for job creation and security in the face ofinternational trade liberalization. The Canadian Jobs Strategy was seen as effectiveplanning in the face of increased potential economic instability. (Prince and Rice1988?: 249-50).The Canadian Jobs Strategy (CJS) was heralded as a flexible, coordinated,regionally responsive strategy of programs that were simple to understand and fairto all groups. It consists of six programs, each targeting a different labour marketneed. Three had the potential to affect social assistance recipients. First, JobEntry/Re-entry is targeted at women and youth and is designed to help those whohad been out of the workforce for a considerable period of time (3 years for women),or who had not worked before. It consists of classroom pre-employment51preparation as well as life-skills training and work experience. Second, JobDevelopment, the most heavily funded of the CJS programs, frequently provides asubsidy or financial contribution to an employer or training host to provide acombination of training and work experience either on-site or off-site for the long-term unemployed. Third, Innovations is a program designed to finance creativepilot projects, designed to alleviate labour-market problems. In the first two of theCJS programs, the government set participation targets for women, the disabled,aboriginal peoples and visible minorities-- groups who are traditionallydisadvantaged in the economy.One group of employment disadvantaged citizens who were not includedformally in the original CJS targets were social assistance recipients. Although“employable” social assistance recipients made up 30 percent of Canada’sunemployed population (Health and Welfare Canada 1985), their particular trainingneeds were not addressed in the CJS strategy. To ignore social assistance recipientsas a distinct disadvantaged population is to fail to address the particular social andeconomic barriers which stand in the way of a considerable segment of society’sreintegration into the mainstream. Without government intervention to removethese barriers “an ever-increasing number of individuals and families will becometrapped in a costly, long-term cycle of poverty and welfare dependency” (Prince1989: 5).In order to ensure that social assistance recipients would not be denied thebenefits of CJS, the federal and provincial ministers of social services met in April1985 in Ottawa to discuss a strategy for improving the self-sufficiency of socialassistance recipients. Each jurisdiction came to the table with its own history oftraining and employment services for social assistance recipients, which rangedfrom well-developed programs in Saskatchewan to the very few bare-bones andpoorly funded programs of British Columbia (SP Research Associates 1988: 23). At52this meeting the ministers agreed to develop proposals to increase the coordinationof the training efforts of various levels and departments of government and todevelop creative funding mechanisms to increase the effectiveness of employmentenhancement efforts nation-wide. In September of 1985 they met again andapproved the Minister’s Agreement to Enhance the Employability of Social AssistanceRecipients. In subsequent months each province and the Northwest Territoriesnegotiated its own Letter of Understanding with the Federal government in order tomake the agreement appropriate to its particular regional context.The AgreementThe SAR Agreement is essentially a framework whose content is fleshed out inthe provincial Letters of Understanding. Nevertheless, the original agreementbetween National Health and Welfare, Employment and Immigration Canada andthe provincial departments of social services and of labour, sets out as its generalgoal “to maximize the use of the Canadian Jobs Strategy, corresponding provincialtraining and employment programs, social services programs and the CanadaAssistance Plan to help income assistance recipients to obtain and maintainemployment” (Canada/B.C. SAR Agreement 1986). The agreement outlines sevenpoints to the governments’ new strategy to assist social assistance recipients tobecome self-sufficient:1. Negotiate for each province and territory participation targets for socialassistance recipients in the Canadian Jobs Strategy; these targets to become effectivefrom April 1, 19862. Make immediate changes to the Guidelines of the Canada Assistance Plan toencourage greater participation of social assistance recipients in training programsand employment; the impact of this initiative to be reviewed in three years.3. Use a portion of the Innovations Program of the Canadian Jobs Strategy tosupport innovative initiatives for enhancing the employability of social assistance53recipients, with federal and provincial officials to prepare a specific set of proposalsby the end of 1985, for review by the National Innovations Advisory Committee.4. Initiate federal/provincial/territorial employment and training pilot projectsunder the Canada Assistance Plan. These projects will be designed to determine theeffectiveness of experimental approaches to promoting self-sufficiency among socialassistance recipients. Criteria for such projects will be developed by December 31,1985.5. In addition to item 1 above, expand opportunities for social assistance recipientsto participate in training and employment by permitting provinces, who wish to doso, to refer social assistance recipients to Canadian Jobs Strategy programs or similarprovincial programs, and also to transfer to those programs those funds that theprovince wQuld have spent on social assistance payments to those programs. Thefederal government will then transfer an equal amount of money, which would havebeen paid under the Canada Assistance Plan as the federal share of the socialassistance payment, to the programs involved.The design, eligibility criteria, administrative requirements and evaluationmethodology relating to these new arrangements will be developed by November30, 1985, and the impact of this initiative will be reviewed in three years.6. Cooperate in gathering the information needed to assess the benefits and costs ofthe program to erthance the employability of social assistance recipient; particularlythose described in items 2,3,4 and 5 of this agreement.7. Maintain the four-cornered” collaboration among National Health and Welfare,Employment and Immigration Canada and provincial department with socialservices and labour market responsibilities in the implementation of the presentagreement.(Ministers’ SAR Agreement, September 1985)Two features of the SAR Agreement were particularly innovative. The firstwas the “four cornered” collaboration between the various ministries at two differentlevels of government. Departments which formerly had neither the informationabout, nor the interest in, the activities of other department came, through theagreement, to cooperate to establish a more effective array of programs whichavoided unnecessary duplication or gaps in programming. Because of this renewed54federal-provincial cooperation, the agreement is often simply referred to as the “FourCornered Agreement”.The second innovation was the particular funding mechanism for theemployability initiatives. Due to the climate of fiscal restraint as Canada came out ofthe recession, there were few new funds to allocate to welfare-to-work programs. Assuch, governments were forced to make use of existing resources, but wereencouraged to use them in more creative ways. At the same time, the governmentsagreed that neither the federal nor the provincial governments were allowed toreduce their efforts as a result of the increased efforts of the other party. Thefunding solution developed in SAR Agreement has become known as theEmployability or Diversionary Fund. It allowed money which is normally spent onincome assistance, through the Canada Assistance Plan and the provincial welfarelegislation, to be spent instead on allowances and wages for training.The Canada Assistance Plan (CAP) is the mechanism by which the federaland provincial governments share the cost of welfare. Welfare is administeredprovincially. The federal government matches dollar-for-dollar the funds that theprovince spends on income assistance and social services. (Recently three provinces,including B.C., have been denied full cost matching under a “cap” on CAP funds).The Canada Assistance Plan regulations had prohibited National Health andWelfare from sharing with the provinces the cost of wages for welfare recipients orsubsidies for training. The 1986 SAR Agreement, as a temporary agreement that didnot require a permanent change to CAP regulations, led to the extension of costsharing to welfareto-work programs.At the same time, the SAR Agreement attempted to protect social assistancerecipients who chose to train from being worse-off financially as a result. It led tothe extension of earnings exemptions under CAP to cover training allowances andincome earned in these programs. It also led to the retention of non-cash benefits for55those who train during the first year that the trainee leaves social assistanceThrough the Ministers’ creative initiative, CAP became a vehicle more conducive tothe promotion of employability for welfare recipients.Overall, $600 million dollars of income assistance funds over three yearsbecame available to finance new federal and provincial training initiatives for socialassistance recipients in Canada. In addition to these initiatives, targets set for theparticipation of social assistance recipients in CJS programs ensured that theywould not be shut out of mainstream training programs. In 1986, social assistancerecipients, agroup that had been virtually invisible in training agendas, suddenlywas presented with what seemed in contrast to be a deluge of opportunities.Like the Canadian Job Strategy, the national SAR Agreement, from which theprovincial Letters of Understanding emerged, embodies the principles of the welfarestate outlined by Golding and Middleton. First, it emphasizes the efficiency of thelabour market in its assumption that once individuals are properly trained, thelabour market will absorb these workers. Second, In Minister of Employment andImmigration Flora MacDonald’s address at the September meeting, she reinforcedGolding and Middleton’s principle of the morality of the work effort by saying that“not only are these problems [of high welfare case-loads] placing a heavy burden onprovincial and federal treasuries, they are also depriving individuals concerned ofthe opportunity to achieve self-fulfillment through satisfying and productive work”(Ottawa, Sept. 17, 1985). Finally, the nature of most of the programs initiated by theSAR Agreement is to compensate for “individual inadequacy” — to make up fordeficits in their education and work-skills, rather than to combat structural factorsthat might reduce their incentive to leave social assistance, such as low minimumwage or a shortage of jobs.In addition, the agreement embodies many principles reflecting the particularneeds of its target group. These include the recognition of the primary influence of56economic factors, such as earnings exemptions and in-kind benefits, on socialassistance recipients’ decisions to participate in welfare-to-work activities. Also, theagreement recognizes the importance of individual choice by mandating thatprograms be voluntary in accordance with CAP guidelines, and that programs beflexible, reflecting the diverse needs of social assistance recipients, businesses andregions. In this way, the Canadian welfare-to-work initiatives which were born inSeptember of 1985 recognize the importance of circumstances in the phenomenon ofpoverty and seek client-sensitive ways to alleviate both dependency and poverty.British Columbia Implements the AgreementAfter the ministers consented to the agreement, the next crucial stage in theimplementation of the principles was for the federal government to enter intonegotiations with the provinces and territories to produce a more detailed Letter ofUnderstanding which would provide the authority for employment enhancementinitiatives in each province. There were several issues which had not raised anyeyebrows in the initial agreement, but which at the negotiation stage were thesubject of considerable debate and hindered smooth and rapid implementation ofthe agreement. These issues were identified and explored by SP Associates in its1988 study of the first years of the Employment Enhancement Agreement. First,many of the features of the Agreement, especially that of fund diversion, were newto the provinces and so took a good deal of “selling” to be accepted. Second, thetraining arena was one of considerable tension in several jurisdictions, in which thesocial service and labour departments each wanted primary control over theinitiative, and in which provincial authorities were unwilling to see their IncomeAssistance funds diverted to the federal Canadian Jobs Strategy programs.A third, extremely controversial issue concerned the distribution of divertedmoney between federal and provincial programs. What was at issue was which57government’s programs should have precedence in the province. The resolution tothis debate in many provinces, including British Columbia came in the form of “cost-matching”, in which each level of government diverts the same amount of incomeassistance funds to its own programs. The danger of “cost-matching” is that thecooperation championed by the original agreement is weakened since each level ofgovernment appears be running its own programs with its own money, rather thanseeking to share the costs of particular programs (SF Associates 1988).While British Columbia was the second province to begin discussions todevelop the province’s Letter of Understanding, it was the first province to sign. Itmoved rapidly ahead to implement its programs while other regions were stillnegotiating the terms of their agreement.British Columbia’s experience with employment and training for welfarerecipients was limited prior to the signing of its Letter of Understanding on May 29,1986. The province has a long history of tying strict and punitive requirements forparticipation in work activities to welfare eligibility (Callahan et al 1990). Althoughnever mandating that work be performed in exchange for welfare benefits, BritishColumbia has an increasingly stringent eligibility policy in which work-seekingactivities and availability for and acceptance of a job are conditions upon whicheligibility is maintained (GAIN Regulations, Section 3.0-9). It has continued toexpand eligibility for welfare-to-work activities over time by adding new categoriesof individuals who are considered to be “employable”-- a title which allows theprovince to reduce welfare benefits and to compel social assistance recipients to seekwork. GAIN regulations allow this work-seeking requirement to be met throughparticipation in employment-preparation programs (GAIN Regulations Section 3.2-3).Prior to 1986, B.C.’s welfare-to work policy was characterized by seasonalwork projects in the 1960s in which “Picking berries in the Fraser Valley or fruit inthe Okanagan, fire-fighting and tree planting were typical jobs that employables58receiving assistance had to accept in some areas and some seasons or be cut offwelfare” (Callahan et al. 1990: 16). Throughout the 1970s and early 1980s workactivity projects continued to be introduced as the categories of those who were“employable” grew. These programs, although introduced with strong words,notably Premier Vander Zaim’s statement that he would provide social assistancerecipients with “a shovel” and not a hand-out, had little impact on increasing theself-sufficiency of welfare recipients. Programs were only minimally funded and notpart of any comprehensive strategy to help the recipients to overcome the barriersthat they faced. In 1984-85, for example, only $637,00 was spent on “Work ActivityProjects”, and $315, 000 on Joint Training Assistance with the Ministry of Labour(Ministry of Human Resources 1984-5: 27).The agreement increased dramatically the resources allocated to trainingsocial assistance recipients in British Columbia. In the first year of the agreement,training received 15 times as much as it had in previous budgets. The SARAgreement for British Columbia initiated a great surge in programming for thisgrowing target group. In each of the three years of the initial agreement, $30 millionwas diverted from CAP and GAIN funds to support the federal and provincialgovernments’ initiatives. The federal government funded and ran its own CJSprograms in the province, alongside the province’s initiatives, funded andadministered by the provincial government.In the first year the British Columbia government chose to spend $13 millionof its diverted funds on a program called Forsar and $2 million of its funds on theJobTrac Employment Subsidy Program. Forsar was a forestry enhancement projectadministered by the Ministry of Forests and Lands in which that ministry hired acrew from among the social assistance recipients referred to them by the localministry of Social Services and Housing, on behalf of a contracted forest company.These social assistance recipients trained in the forest industry while the government59paid a full wage subsidy to the employer out of the diverted funds. In the JobtracEmployment Subsidy Program, private sector employers received a partial wagesubsidy for hiring social assistance recipients and for providing them with jobtraining.For its part, the Federal government provided $15 million each of the threeyears of the agreement for Job Development and Job Entry/Re-entry projects in theprovince. The package of programs included a Severely Disadvantaged Option ofJob Development (later transferred to Job-Entry) in which social assistance recipientsexperiencing “formidable” barriers to employment were paid a wage while receivingboth in on-the-job experience and basic skills upgrading. Also under JobDevelopment and Job Entry were Direct Purchase Option programs, in which thegovernment purchased seats in a training classroom rather than providing on-sitework experience. R.I.S.E. projects were added in later years, which focused onprogramming to meet the needs of a specific target group in a specific community.In addition, the federal government set participation target levels for socialassistance recipients in mainstream CJS programming of 30% for Job Developmentand 35% for the Re-entry Option of the Job Entry program.In the 1987/88 year, the province renamed and expanded its programs.Under the broad umbrella heading of “Jobtrac”, the Ministry renamed Forsar“Forestry JobTrac” and added two new programs. “Community Jobtrac” provided afull subsidy to non-profit agencies for hiring and training social assistance recipientsin community, tourism, sport, culture and heritage projects. Environment JobTracprovides funding, In the form of a full wage subsidy, to two non-profitorganizations, the B.C. Conservation Foundation, and the Outdoor RecreationFoundation of B.C. “to hire and train social assistance recipients to deliver parkenhancement and fish and wildlife projects” (SP Research Associates 1988: 72). TheJobTrac Employment Subsidy Program remained the same. The province allocated60over $40 million to their programs for 1987-88 year, even though only $15 million ofthese funds would be cost-matched by the federal government. For administrativepurposes, only portions Forestry JobTrac and the Employment Subsidy Programwere declared under the agreement. That year marked the beginning of theprovince’s commitment, far above and beyond the mandate of the SAR Agreement,to training welfare recipients.Over time, the names of the provincial programs have changed, (the umbrellaprogram called Jobtrac as later changed to “Employment Plus”), programs have beenterminated, reinstated, and remodeled. A new program, Job Action, an intensivecourse on job-search skills, soon became part of the welfare-to-work mix, as didRegionally Initiated Special Employment Programs (R.I.S.E.), which allowed regionsto identify a specific target group and to design projects to meet their particularneeds. Over the past decade, SAR Programs have changed names, but the content isessentially the same. Below, the original programs are listed opposite their variousincarnations.Federal ProgramsJob Development Job OpportunitiesJob Entry/Re-entry Project-Based TrainingDirect Purchase of TrainingProvincial ProgramsJobtrac: Employment Subsidy Program----Employment Opportunity ProgramForsar Forest Enhancement Program-Forest Worker DevelopmentProgramCommunity Jobtrac Community Tourism EmploymentProgramEnvironment Jobtrac Environment Youth ProgramEnvironment Youth CorpsRegionally Initiated Special Employment Program (R.I.S.E.)Job Action61Over time, British Columbia has shown an increasing commitment to fundemployment enhancement initiatives. This commitment was not matched by anincrease in funds on the federal side until the Minister’s Agreement was actuallyrenegotiated and resigned, in 1991, after having been extended at 1986 levels duringthe years of renegotiation. In the 1991 agreement the fund diversion for the Federaland British Columbia governments was $28 million apiece. The programs coveredby the 1991 agreement are more numerous and varied than those covered in theearlier agreement. It is important to note, however, that while funds were increasingunder the SAR Agreement, the federal commitment to mainstream programs fromwhich social assistance recipients benefit was eroding.Implementation of the B.C. programs at the local levelThe objectives of the SAR Agreement are achieved through using programfunds to deliver training projects in which individual social assistance recipients canparticipate. In both the federal and provincial training systems the funds that flowfrom the SAR Agreement are allocated to local offices, whether CJS operation centersor Canada Employment Centers (CECs) on the federal side, or Ministry of SocialServices (MSS) offices on the provincial side. These funds are directed at certainbroad programs, such as at Job Development or Wage Subsidy Programs. However,the content of these programs, that is, the individual projects, is the responsibility ofthe officials in the local offices. Their responsibilities range from approvingapplications from contractors or employers for projects, to helping in the design ofprojects that are reponsive to the needs of target groups, to monitoring the progressof the trainees in the projects, to deciding whether to continue the relationship withthe contractor or employer.Local officials have considerable autonomy from the higher levels ofgovernment in choosing their projects. Since projects are generally small, serving62only a few clients at a time, their cost usually falls below the financial level at whichlocal officials would have to seek approval for projects from higher level officials. Inthe federal training system, local officials are constrained by the fact that they musthave the blessing of the local Member of Parliament. This approval is generallygiven. If it is withheld, local officials retain the option to over-rule this objection.However, the situation in which an MP rejects a project application that is supportedby local officials does not often arise. There is no comparable role for MLAs in theprovincial training system.A unique feature of project administration that arose from the SARAgreement was the evolution of joint planning between the federal and provincialministries involved in training social assistance recipients. This happened at twolevels. First, upper levels of government formed a management committee on whichofficials from the “four corners” involved in the SAR Agreement sat. As federal andprovincial officials plan their yearly mix of programming, they consult with eachother with the goal of avoiding gaps and duplication in programming in local areasin British Columbia. Second, at the local level, officials from the CEC and MSSoffices are encouraged to communicate often, to share information about projects, torefer clients to the other government’s programs, and to coordinate their efforts inorder to best serve the social assistance population in their area.Some of the most important stages of implementation occur once the fundsare allocated and the welfare-to-work programs designed. The interaction betweenlocal officials and clients is critical. Local officials determine which clients will beallowed to participate in welfare-to-work programs. This means that they mustassess the needs of clients and their capabilities in order to determine whether theywould benefit from a particular project. At the same time the provincial field-levelworkers determine the level of benefits, in-kind transfers and transitional supportsthat the individual will receive. The tasks of local officials involve tremendous63discretion. The effectiveness of the referrals and of the income assistance decisionsmade is a function of the knowledge, sensitivity, intuition and experience of theindividual caseworker.In order for implementation to be effective, the client must have the incentiveto participate in a program. In British Columbia the incentives that have beendesigned by the government to encourage clients to participate in either federal orprovincial programs include the enhanced earnings exemption which, in conjunctionwith the October 1985 changes to the CAP guidelines, allows social assistancerecipients to retain 25 percent of their earnings above a flat rate exemption, (either$100 or $200 depending on family size), while they train. Another incentive is theprovision of the non-cash benefits, such as medical and day care subsidies, for thosein training, and the compensation to income assistance recipients of the increasedcost of training through clothing and transportation allowances. Although theseincentives impose a large cost on government, officials administer these incentives inthe belief that the savings to the government generated from the increased workeffort of social assistance recipients will offset this cost.The Impact of the SAR Agreement in British ColumbiaThe 1986 Canada-British Columbia Agreement To Enhance The Employability ofSocial Assistance Recipients was successful in meeting many of its objectives. First, itincreased the priority and resources given to the issue of training social assistancerecipients and helping them to overcome barriers to employment. In BritishColumbia the increase in the province’s commitment to this issue was dramatic.Even without federal matching funds, the province increased its commitment toemployment enhancement well-beyond the $15 million per year to which it hadoriginally agreed, spending in excess of $40 million on programs in 1987/88.64The goal of maintaining “Four Cornered” collaboration was also attained inthe province. SP Research Associates describes a high degree of cooperation and a“very good working relationship” between the four departments involved in theagreement (1988: 75). In fact, the collaboration on the social assistance recipientsissue initiated further cooperation in other areas. “Officials from the four corners areusing the committee to address a number of difficult issues troubling the twosystems, including some unrelated to the Letter of Understanding” (ibid).As a package the SAR Initiative represents true progress in government’sattempts to address the needs of social assistance recipients and they must be judgedas a success. However, the impacts are modest and not without limitations. Someindividual programs within the SAR package have been judged to impose morecosts than benefits on society. Moreover, although the majority of SAR projectsapproved at the local level had some positive impacts, within every program typeare individual projects that fail, and within successful projects, some individualsocial assistance recipients whose experience with training is not positive. Inconsidering the factors influencing the success of the implementation of SAR, it isequally important to understand why some programs, projects and social assistancerecipients also fail to achieve their objectives.For government, the ultimate measurement of the effectiveness of theemployment enhancement initiatives is the direct impact on the earnings andwelfare dependency of social assistance recipients, and on the savings to thegovernments providing training. Routes to Independence, a 1992 Ministry of SocialServices study of the provincial initiatives in British Columbia, found a mixture ofsuccesses and failures, but overall had an optimistic tone about the potential of welldesigned welfare-to-work initiatives. Evaluation of Employability Initiatives ForSocial Assistance Recipients in CJS. a 1994 Human Resources Development report onthe fate of social assistance recipients in federal programs, was even more positive.65On the provincial side, the Employment Subsidy Program, (now theEmployment Opportunity Program) produced the most markedly positive results.The program moved participants off of income assistance more quickly than non-participants. For two years following the completion of the program participantshad a 10 to 15 percent lower rate of dependency on welfare than non-participants.The greatest impact in terms of dependence-reduction occurred for those socialassistance recipients who had longer, rather than shorter, welfare histories. Theinitiative increased earnings through increasing the hours of labour forceparticipation of social assistance recipients. However, the average wage of programparticipants was no higher than that of non-participants, indicating that there waslikely no “training effect” of participation in the program which would have allowedparticipants to secure higher-paying employment after completing the program.The benefits of the Employment Subsidy Program far exceeded its costs. Forthe provincial government, the dramatic decrease in welfare dependence resultingfrom the wage subsidy program resulted in substantial savings. Even thoughparticipation in the program was characterized by 15 to 20 percent greaterdependence on the federal government’s Unemployment Insurance program for thetwelve months of Unemployment Insurance eligibility compared to non-participants, over the long term (65 months), there was still a net savings of $890 perparticipant to the governments (Ministry of Social Services 1992: 18). Similarly,although the displacement of current workers by subsidized social assistancerecipients did occur, the Ministry concluded that there was a higher level ofemployment overall, since these displaced workers “would have found other jobsmore easily than would the subsidized employees who replaced them” (ibid: 16).The Wage Subsidy Program was thus a very successful program element of the SARAgreement.66The three on-the-job programs that involved training in public projects --theCommunity Tourism Employment Training Program (CTETP), the ForestEnhancement Program (FEP) and the Environment Youth Corps (EYC)-- werejudged by the Ministry to have been unsuccessful in meeting many of the crucialobjectives of the SAR initiatives. Although they helped to reduce welfaredependence in the short-term, the programs had “little long-term impact”. Rather,they increased Unemployment Insurance significantly, by 25 to 30 percentage pointsover non-participants. The projects, like the Employment Opportunity Program, didnot have a training effect resulting in higher wages. Even when the value of thework that the social assistance recipients did was added to the welfare savings,program costs and increased Unemployment Insurance dependence exceeded themonetary benefits of the program to social assistance recipients and to government.A final provincial program, Job Action, was evaluated by the Ministry ashaving been highly successful in reducing dependence and producing governmentsavings. This intensive job-search course helped to get participants off of welfaremore quickly than non-participants. The ministry calculated that Job Action savedthem an average of $367 per participant, which, taken over a large case-load resultedin considerable savings to governments. Program benefits far exceeded costs.The federal government evaluated its initiatives to enhance to employabilityof social assistance recipients nationally. Although impacts in B.C. are citedfrequently, the federal results are more general than the provincial results, andtherefore cannot necessarily always to taken to represent British Columbia. Thefederal government found that “participation of social assistance recipients in theCanadian Jobs Strategy provided substantial gains for participants and for theCanadian economy” (Human Resources Development Canada 1994: i). Assessing allof their CJS programs for social assistance recipients as a group, they report that thefederal programs resulted in increases in employability, annualized earnings and,67for those who participated in mainstream CJS programming under the target rates,increases in wage rates. Social assistance recipients spent 24 percent more timeemployed following participation in CJS programs than non-participants.Although the federal report does not present its quantitative results in ascomprehensive a fashion as the provincial government report does, it does separateprogram impacts to some degree. Job Entry and Job Development were bothconsidered to be very successful, in view of the numerous barriers to employmentfaced by the target population. Both programs are reported to have “significant andcomparab1e’ employability impacts and positive earning impacts. Participation inJob Development had a significant effect of reducing future social assistancedependence. The Direct Purchase Option had a “smaller and less significantemployability impact” but a positive impact, nonetheless, in terms of earnings.The federal report concludes that although, from a government accountingpoint of view, welfare-to-work programs are not cost-effective over the short term,such a method of analysis fails to take into account the benefits to the recipient oftraining. When all social benefits are weighed against costs, they calculate that thewelfare-to-work programs for social assistance recipients that federal governmenthas been running have a benefit-cost ratio of 2.0, and are quite successful inproducing a positive impact on society.Both the federal and provincial reports used comparison group data in orderto determine what the employment and earning outcomes would have been forsocial assistance recipients had they not participated in welfare-to-work programs.The provincial report, with its more complete numerical portrayal of the impacts ofprograms, appears to be more credible in its assessment of program impacts. That itignores qualitative measures of success, such as increased self-esteem forparticipants or a revitalization of a community, suggests that its evaluation errs onthe side of portraying program impacts as less positive than was actually the case.68The federal report, by contrast, takes into account the criteria absent in the provincialreport. Together, the reports provide a good discussion of a variety of impacts ofwelfare-to-work initiatives, from governments’ point of view. Many of theirfindings are consistent with research done in other jurisdictions and compiled byGueron and Pauly. This consistency increases the credibility of the federal andprovincial results for B.C.By qualitative measures, the Employment Enhancement Agreement was asuccess. Employment and training programs were found to both increase the self-confidence of income assistance recipients (NAPO 1989), to reduce the bias ofemployers against this target group, and to enhance communities through thecontributions of this new portion of the labour force (SP Research Associates 1993).These impacts all have positive effects on the possibility of new initiatives reachingtheir objectives, since communities and employers will be more supportive ofinitiatives, and social assistance recipients more convinced of their own abilities andof the opportunities that employment enhancement programs can provide them.Providing a counterweight to this optimism, anti-poverty groups warn thattraining programs pacify the public, thus serving the needs of the government, andmaintain a supply of low-wage workers, serving the needs of business. While theywould not dispute the governments’ numbers or their contention that theseprograms improve the skills of social assistance recipients, critics point out thatsuccess should only be measured by the end result: are social assistance recipientsbetter-off following training than they were before? This means, specifically, dothey have a job that pays wages on which they can live with some degree of dignity?The problem with the SAR Agreement, according to those who measure success inthis way, is that it fails to address the problems of few jobs and fewer well-payingjobs in British Columbia. It does not provide the training that would allow mostsocial assistance recipients to qualify for the better-paying jobs. Moreover, they69contend that training is too narrow to allow trainees to pursue several differentavenues in their search for a job (Interviews 1994).The assessment of government and its critics together provides a balancedpicture of the impact of the SAR Agreement in British Columbia. The Agreementwas successful in bringing training opportunities to social assistance recipients.Many of the provincial programs and all of the federal program designs wereeffective in improving their employability and increasing their earnings. Withinthese successful programs some individual projects still failed. Some of theunsuccessful provincial programs or the unsuccessful projects within the successfulprogram types were too limited in nature, too narrow in the training they providedor too unresponsive to the needs of clients and to the local economy to providesignificant benefits to either social assistance recipients or to society in general.However, the majority of SAR programming was implemented effectively and metthe objectives of the original agreement. That there were few decent jobs at the endof training has less to do with the implementation of SAR and more to do with thelarger issues, beyond the purview of the agreement, of the structure of work inCanadian society.70Chapter FourExplaining The Fate of British Columbia’s SAR AgreementIn choosing the SAR Agreements as their vehicle for improving theemployment rates of social assistance recipients, for reducing welfare dependencyand for alleviating poverty among this target group, the Ministers engaged in agamble that a number of actors and circumstances would come together to producethe desired outcomes. The existence of a federal-provincial agreement to fundinitiatives to increase the employability of social assistance recipients does not initself bring about this goal. Rather, successful implementation of welfare-to-workprograms is a process with many stages. In the broadest terms there are six stages,each of which has to be implemented effectively in order for the entire process tomeet with success, defined in terms of increased earnings and employability for thesocial assistance recipient, and reduced expenditures for the state. Each of thestages attempts to answer a question crucial to implementation with a governmentresponse.In the first stage the question, “what resources are provided?”, is answered atthe upper levels of government through the capacity of the governments thatformulated the policy or initiative to provide the direction, guidelines and resourcesto enable the next level of implementors, in this case the local level officials withinBritish Columbia, to carry out their jobs effectively. The second question, “how arethe resources used?,” is responded to at the local level, where the local MSS and CECoffices generate and support projects which are designed to meet the needs of socialassistance recipients. “Who gets to use these resources?” is a third concern, which isaddressed by the field-level workers who decide who will have access to trainingprograms and support services. The soundness of rationales behind who is includedand who is excluded from which programs will greatly affect the programs’ ability71to achieve the agreement’s objectives. Once referrals are made the new issuebecomes “will these groups use these resources, and why?” At this stage the socialassistance recipient makes the individual decision about whether to participate in, orcomplete, a given program. This decision has much to do with the client’sperception of the program and of the effectiveness of the incentive structure of theprogram in addressing the client’s needs. A fifth question is, “when they doparticipate, will the resources help these groups to achieve the important goals forwhich the resources were allocated to begin with?” This question is answered by theprogram design, which must provide the necessary elements for a client to becomeself-sufficient. After these five questions have been answered there remains onefinal one: “are all of the previous decisions meaningful in light of the externalconditions in the implementation environment?” Even if all of the earlier stagesfacilitate the achievement of the agreement’s objectives, successful implementationof the agreement can nevertheless be undermined by factors external to theimplementation process. The environment in which all the decisions are made canmake the decisions irrelevant. Thus, a final stage of implementation involves thetrained social assistance recipient having access to employment opportunities in thecurrent social and economic environment.In addition to the six stages at which conditions conducive to implementationmust prevail, the causal theory underlying the intervention, that is, the premises thatform the basis of the SAR Agreement must be sound. A clear understanding ofbehaviour, and of the factors which can be manipulated so as to alter behaviour, isthe basis for any policy intervention. Thus, if the original behavioural assumptionsare invalid, effective implementation of the six stages will fail to resolve the policyproblem.An analysis according to the theory of the “complexity of joint action”,advanced by Pressman and Wildavsky, would suggest that the federal-provincial72initiative was destined to fail, given the number of contingencies which had to bemet in order for a social assistance recipient ultimately to become independent ofwelfare. Eugene Bardachs theory, similarly, would predict failure, given the manyopportunities for implementation games to be played by implementing officials atthe numerous points at which discretion is exercised. However, in spite of thecomplexity and relatively decentralized nature of the implementation processcreated by the SAR Agreement, it was, and continues to be, a notable success. Foran explanation of the success of the initiative, it is helpful to turn instead to therevised version of Mazmanian and Sabatier’s framework, outlined in Chapter one,which also takes into account the contributions of Elmore, Lipsky, Goggin et al.,Stoker and Gueron and Pauly, all of whom recognize that complexity and discretioncan in some cases facilitate effective implementation. The revised frameworkprovides a key to understanding why the particular programs generated by theCanada/British Columbia SAR Agreement were generally successful in helpingsocial assistance recipients to become more self-sufficient than they would have beenin the absence of such programs. The same framework also suggests reasons for thefailure of particular programs, and for the limitations that even the successfulprograms face.Applying the Revised Theory of Implementation to the Case-studyWhile Mazmanian and Sabatier’s theory acknowledges the importance ofbottom-up factors to implementation, their original theory was biased in terms of theimportance of top-down features of policy design and centralized control. Therevised framework includes several distinctly bottom-up factors; in addition, in thisanalysis, each set of Mazmanian and Sabatier’s original variables will be consideredat the field, as well as at the policy-formulation, level.73The conditions of the revised framework are effective in explaining thegenerally successful implementation of the agreement, the particular mixture ofsuccesses and failures in the programs generated by it, and the limitations which theprograms consistently faced. The importance of these conditions is verified bygovernment and anti-poverty group reports on the implementation and outcomes ofthe SAR Agreement and by interviews with officials at the policy formulation andfield levels in both the federal and provincial government.1. The enabling legislation structures the implementation process so as tomaximize the probability that implementing officials and target groups willperform as desired. This involves assignment to sympathetic agencies withadequate hierarchical integration, with a history of cooperation, or expectations offuture cooperation, supportive decision rules, sufficient financial resources andadequate infrastructure.The most common explanation for the success of the welfare-to-workprograms developed through the SAR initiatives in British Columbia is that theinitiative achieved an unprecedented level of cooperation between the variousparties involved in training and income assistance in British Columbia. Before 1986,the interaction between Employment and Immigration Canada and the provincialministries of Social Services and of Labour, (later Advanced Education and Trainingand now Skills, Labour and Training), had been haphazard. Each department hadits own training agenda, and conducted its planning at the upper levels ofgovernment with little coordination with other departments. At the field level,coordination between local Canada Employment Centres and the field offices of theMinistry of Social Services was fairly common. However, this cooperation wasbased on the initiative taken by individual workers to seek out contacts with theother local office. Where cooperation was inadequate each office planìned its projectsseparately, leaving gaps and duplication in programming. As a result, trainingresources were often not directed to where they were most needed.74The 1986 agreement set up a process in which federal-provincial cooperationwas formally encouraged. This cooperation took place first at the upper levels ofgovernment in the negotiation process. Muironey’s support for the revitalization offederal-provincial relations imbued the negotiations with what participantsremember as a “strong atmosphere of cooperation,” in which participants sawthemselves as partners, rather than as antagonists.A management committee composed of representatives from both levels ofgovernment has remained active ever since the introduction of the SAR Agreement,engaging in joint planning and assessment of the effectiveness of programs inmeeting the agreement’s objectives.One of the chief responsibilities of the committee in the first years of theagreement was to hold meetings in the various field level offices of bothEmployment and Immigration Canada and B.C.’s Social Services. These visits by theexecutive of the management of the SAR Agreement had two effects. First, theydemonstrated to field workers that a united front existed at the upper levels ofgovernment and thus encouraged them to cooperate with their local sister office ofthe other level of government. Second, they ensured that all field-level workerswould receive identical messages about the purpose of the SAR Agreement, unitingthe field offices and the upper levels of governments. The management committeewas reported to have been excellent at “selling” the objectives of the agreement.Through the work of the management committee responsible forimplementing the SAR Agreement, hierarchical integration, in the sense of sharedvalues and orientation towards the initiative, was improved at both levels ofgovernment. This may be the key as to why Pressman and Wildavsky’s “Complexityof Joint Action” was avoided. Following Mazmanian and Sabatier, this integrationincreases the possibility of effective implementation. The shared orientationtowards the agreement decreased the likelihood that destructive implementation75games, of the types described by Bardach, would be played. At the same time,following Stoker, the emphasis, in both the original document and the actions of themanagement, on mutual adaptation, and on creating a ‘shadow of the future” inwhich cooperation is the key to both parties achieving their objectives, encouragedthe development of an environment in which joint planning could be pursued inorder to address the needs of clients most effectively. In addition, the decision rulescreated by the agreement mandated that the programs of one level of government bereviewed by the other level. These features of the implementation all emphasizedcooperation in an environment in which previous to 1986 contact between agencieshad been sporadic, at best.Funding arrangements also helped to ensure that the various actors inthe implementation process would “perform as desired”. First, the generous fundinglevels for programs ensured that officials would be able to deliver what theypromised, both at the executive and field levels of implementation. The agreementensured that there were real opportunities for training for many social assistancerecipients. Second, through the mechanism of “matched funding” federal andprovincial governments had strong incentives to spend all of the money in theirbudgets on their own programs. If a government fell short of spending its $15million per year it was required to compensate the other government for half of theamount that the other government surpassed it in spending. For example, if BritishColumbia only spent $10 million on its training programs in a given year, and thefederal government spent its full $15 million on its CJS programs for socialassistance recipients, the provincial government would owe the federal government$2.5 million. Rather than “subsidize” the other government’s programs, eachgovernment had the incentive to spend all of its own resources on its own initiatives,thus maximizing the opportunity provided by the SAR Agreement.76In addition, the fact that the four departments involved in the SARAgreement were agencies sympathetic to the goals of the agreement-- all committedto improving the life circumstances of social assistance recipients-- increased thelikeithood that they would perform in ways which advanced the agreements’objectives. For example, all three of the departments involved in arranging contractswith job training hosts, that is, Employment and Immigration Canada, the B.C.Ministry of Social Services and B.C.’s Advanced Education and Training (formerlyLabour), actively sought contractors who could provide social assistance recipientswith the work and training experience that they would need to progress towardsself-sufficiency. This commitment helped ensure that Job Entry and JobDevelopment on the federal side, and the Wage Subsidy Program on the provincialside, would be successes.The compatibility of agency goals could not be as assured for organizationsother than the original “four corners” participants. In fact, consistent with therevised framework, program failures and limitations on the effectiveness ofsuccessful programs can be explained by the multiple and conflicting goals ofdifferent implementation partners. For example, the early Forestry programs forsocial assistance recipients were administered through the Ministry of Forests, ratherthan through social services. This ministry did not have as its primary objectivemeeting the needs of social assistance recipients and preparing them for futureemployment. As such, the programs that they administered were often notappropriate for their clients. They were make-work programs which did notprovide trainees with a marketable skill, or with the increased self-respect thatcomes from doing valuable work. The different orientation of the ministryresponsible for this program made it unlikely that their officials’ behaviour would bein-line with that of the ministries from the “four corners”.77Beyond accounting for program failures, the notion of incompatible goals alsoexplains limitations on successful programs. Although in general the province’sWage Subsidy Program helped to reduce dependence and increase employabilityand income for social assistance recipients, many individual projects were notsuccessful because the goals of the employer contracted to provide the training wereat odds with those of the SAR Agreement. Field-workers reported a fair amount ofabuse of the program on the part of employers, in which they took advantage of theopportunity for cheap labour without providing real training for the socialassistance recipients. Similarly, employers sometimes displaced current workers inorder to receive a subsidy for the trainees (Interviews 1994). This behaviour isdifficult to constrain, given the lack of shared orientation and goals between theministry of social services and some businesses. When conditions do not exist inwhich the probability is maximized that implementing officials “will perform asdesired” the objectives of the SAR Agreement are often undermined and it impactsare less positive than they could be.Finally, the absence of a key variable hindered implementation in severalareas of B.C. in the early years of the agreement. The lack of adequate start-up timeand the immaturity of the infrastructure necessary to implement the variousprograms initiated by the agreement was the cause of several early failures of federalprograms to achieve positive results for clients. Reports of projects starting withoutthe proper forms having been delivered to the field offices or without adequate staffbriefing limited officials’ ability to perform as desired. Over time, the necessaryinfrastructure was developed to handle the increased case-load and newprogramming, and implementing officials’ behaviour became more in-line withwhat was desirable.Successful implementation of the SAR Agreement was helped by thosefactors which took away from the “complexity of joint action”. These included78hierarchical integration fashioned by a strong communication network between theupper and lower levels of an agency, fostering a shared orientation of majorimplementing actors towards helping social assistance recipients, cooperation on ahorizontal level between different agencies, decision rules supportive of theinitiative, adequate resources, and assignment to sympathetic agencies. Wherethese conditions did not exist the complexity of joint action again became a problemthat hindered the effectiveness of the implementation, and limited the ability of thetwo governments to deliver the most programming possible using maximum funds.2. The enabling legislation gives implementing officials sufficient jurisdictionover target groups and other points of leverage to attain, at least potentially, thedesired goals.The success of the agreement can be attributed to its ability to influenceimplementing officials to act in particular ways. At the same time, however, it alsoresulted from the agreement’s ability to allow officials to act in a variety of ways,unconstrained by the reins of executive control. This seeming paradox of controlversus discretion was actually a ingenious mix which was one of the strengths of theagreement. Although there was strong hierarchical integration in terms of sharedvalues and orientation, the decision-making as to how the clients would be servedby the agreement was left almost entirely to the field-level workers.At the field level in the provincial system, and in the federal systemincreasingly over time, local officials at the Canada Employment Centres, (and CJSoperation centres), and Ministry of Social Services offices were responsible fordeveloping and monitoring the projects to which social assistance recipients werereferred. Each area office had its own budget, the autonomy to approve projectswith little involvement from higher levels of government and the discretion to referwhichever clients it deemed to be most suitable for these projects.. This freedomgave street-level officials control over several of the key linkages in theimplementation process. By being able to respond to the labour market needs of79their particular community, to the training needs of a specific target population, tothe income assistance needs of individuals, and to past successes and failures in thisarena, street-level officials were able to design and approve projects that had agreater chance of success than programs imposed upon a local area from higherlevels of government. In this way, true to Lipsky’s thesis, street-level bureaucratsactually created the policy with which the clients came into contact. Due to theamount of discretion given local officials, the policy that clients encountered couldbe carefully designed with them in mind, and thus had the greatest possibility ofhelping them to achieve self-sufficiency.Discretion, mixed with inter-agency cooperation at a local level, had apositive effect on the implementation of the SAR Agreement. Field workers soughttheir guidance from the cooperative planning structures that they created at the locallevel more than from higher levels of government. Local discretion lent thecooperation at the local agency level a greater sense of importance, since field-workers truly had the capacity to influence their community, but could only do soeffectively through making the most of the entire range of resources available tothem in the community, including the knowledge, energy and resources of the othergovernment office.The problem with discretion, and with Mazmanian and Sabatier’s “sufficientjurisdiction” condition is that it is meaningless to the implementation process if theofficials who have jurisdiction over the important linkages -- the discretion to designand monitor programs, and to refer clients -- lack the expertise and commitment todo so. Thus, effective implementation depends on the satisfaction of this conditionin conjunction with the satisfaction of the next condition. In the absence of skilledand committed staff discretion can be abused. In such a case “lack of discretion”could become an important condition for implementation success.803. The leaders and street-level workers of the implementing agency possesssubstantial managerial and political skill and are committed to statutory goals.In British Columbia, the wide discretion of street-level bureaucrats is apositive condition of implementation due to the fact that implementing officials atboth the federal and provincial level and at the executive and at the field levelpossessed both the political and managerial know-how and the enthusiasm andcommitment to facilitate the effective implementation of the SAR Agreement.Officials at all levels are quick to point to the contagious enthusiasm of theoriginal management committee in creating and maintaining the momentum toimplement effectively a complex and often unwieldy set of initiatives. Members ofthe SAR Management Committee themselves note that the strong support of thepolitical executive of their ministries or departments was a factor in the success ofthe first stages of implementation. While the upper level of management of theagreement was credited by most as being clever at handling political problems, itseffect was mostly to remove obstacles which might have stood in the way ofeffective implementation of the agreement at lower levels of government. Forexample, in the early years the management committee responded to local workers’reports of a lack of sufficient supports to social assistance recipients during trainingby ensuring that child care subsidies, transportation allowances and clothingallowances would be available to those trainees for whom these costs posed asignificant disincentive to their participation in employment enhancement activities.The fact that capable management guided the process both inspired street-levelworkers and facilitated their jobs.It is at the “street level” that this condition of effective implementation becamemost meaningful. The wide discretion given to field workers allowed them todetermine how the governments’ policies would be experienced by clients. Streetlevel workers recruited and accepted contracts for training from training hosts,81helped to determine the content of the projects, decided which clients would bereferred to projects, monitored the training and then decided which trainingopportunities would continue to receive funds based on their perceptions of theprojects’ success. Although constrained by broad guidelines CEC officials and MSSofficials were able in general to create their own training systems. For example,among provincial “street level” workers, there were no consistent criteria as to whowould be referred for training from among those assessed as “employable”.According to one provincial evaluation, of those exercising discretion 47 percent offinancial assistance workers referred clients who already possessed good job skills,whereas 44 percent referred clients precisely because these skills were poor.Similarly, officials responsible for selecting contractors or job training hosts used avariety of criteria in making their decision, beyond the formal criteria of the province(DPA Group 1989). The success of the projects is thus based on the official’sexperience and intuition as to what works for whom. Over time, those involved intraining social assistance recipients in British Columbia have developed a moremature understanding of the variety of impacts of welfare-to-work programs andhave used their discretion more wisely.In addition to having gained the skills which enable them to choose correcttraining hosts and make appropriate referrals, field level workers havedemonstrated a tremendous commitment to the goals of the SAR Agreement. Oneprovincial office manager described the reaction of officials to the new resourcesgenerated by the agreement. “We were always dealing with problems; for once wewere dealing with opportunities” (Interview 1994). Most field level workers alreadyhad a real commitment to helping social assistance recipients become self-sufficient.The agreement finally removed some of the barriers that their clients andconsequently, they as field workers, faced. A limitation to effective implementationarose in some areas when caseworkers became overwhelmed by the sheer increase82in the volume of their case-loads. However, most field-workers were extremelysatisfied with the increased funds and training opportunities overall. A keyingredient in the success of the SAR Initiative was the degree of enthusiasm withwhich it was received from officials at all levels of government.4. The enabling legislation incorporates a sound theory identifying the principalfactors and causal linkages affecting policy objectives.The SAR Agreement as it was implemented in British Columbia reflected theliberal developmental approach to work for welfare recipients elaborated in chapter2. Rather than assuming that welfare recipients will only seek self-sufficiency whenforced by utter destitution or administrative requirements, the agreement and theefforts of those involved in the agreement reflect a belief that welfare recipientswould work if given the opportunity and if opportunities offered to them improvetheir circumstances instead of driving them further into poverty. This fundamentaltenet is evident in the program designs emerging from the SAR Agreement. Manyof the programs were successful because they helped participants to overcomemultiple barriers to self-sufficiency. The causal theory underlying the initiativewas that by addressing the particular needs of the clients governments can providethem with the resources to become self-sufficient.First, British Columbi&s federal and provincial programs are formallyvoluntary. Although some social assistance recipients felt pressure to participate inprograms or lose their welfare eligibility (SP Research Associates 1993), demand fortraining exceeded supply in the province. Most field-workers referred to trainingand upgrading programs those clients who were motivated to participate, as thesefield-workers felt that these clients had higher success rates than those with lessmotivation (Interviews 1994).Second, the projects that met with the greatest success in both the federal andprovincial training streams in British Columbia were those that provided real job83experience in a field in which there was a demand for the learned skill, or providedtraining with an employer with whom there was a possibility of on-goingemployment after the training is complete. Federal Job Development and provincialEmployment Subsidy programs included these elements.For people who were less job-ready, a more integrated approach was neededto address the multiple barriers which most social assistance recipients face. Onetype of upgrading was often not enough to prepare the more severely employmentdisadvantaged people for employment. The most successful programs incorporatedjob experience, classroom training, life-skills training and counselling. These werethe features of the federal Job Entry programs, including the Severely EmploymentDisadvantaged (SED) Option programs, which had a large social assistance base.These programs had a more incremental view of what constitutes a success and tooka client gradually towards employability. Officials matched program designs toclient’s needs rather than the other way around; this more client-centred approach iscredited with the success of these CJS options.This “valid causal theory” condition of implementation is effective inexplaining some of the failures in programming as well. Federal Direct PurchaseOption (DPO) programs under Job Development provided only classroom training,with no assistance in finding the client a job. By failing to provide integratedtraining to address multiple aspects of job-readiness they failed to prepare workersfor the real on-the-job challenges they would face. As a result, they were lesssuccessful in providing trainees with the confidence and skills they would need toacquire and maintain a job.Similarly, the provincial public employment programs were less able thanthe Employment Subsidy Program to take people off of the welfare rolls by helpingthem to enter permanent employment. These programs often did not encompass acommitment to training individuals to perform real jobs. Thus, when the training84was over, the trainees had not developed marketable skills. For example, traineesmight spend their training period clearing a trail under a public employmentprogram. However, at the end of this training they would not have learned skillsenough to qualify them for a job with another employer. The training was toonarrow. Their original training hosts would not have the same capacity as theEmployment Subsidy private employers to continue to employ the social assistancerecipients after training, given that many of the organizations are non-profit anddepend on the provincial wage subsidy for its staff funds. Once training was over,so was the job.More recently, the various public employment training programs have beenmodified, in an attempt to make training more relevant to real work situations. Forexample, the Forest Worker Development Program now hires social assistancerecipients for positions for which they would otherwise have to hire non-socialassistance employees. These trainees perform meaningful work and are trained in aspecific field for which there is labour market demand. This increases the morale ofthe individual worker, as well as the attitude of the contractor towards the work thatis done by the trainee, since it is now of value to the contractor’s enterprise. Otherchanges to this program include a mandate that trainees be selected from a local areafor projects in that area, increasing the willingness of the employer to continue toemploy the trainee once the government funding is terminated. A beneficial sideeffect of this program is increased direct benefits to community resulting inenhanced community support for training initiatives for social assistance recipients(Interviews 1994)..Experience with welfare-to-work programs has shown that the moreprograms provide meaningful work experience in fields of labour market demandwith the possibility of future employment, the more these program address themultiple barriers of social assistance recipients, the longer the duration of the85program and the more integrated an initiative is in terms of providing an array ofprograms and supports, the more likely it is that a welfare-to-work initiative willsucceed. Through emphasizing many of these elements in its programs, B.C.’sliberal developmental approach is vindicated to some extent by the successes ofinitiatives that attempt to provide opportunities, not merely work.On the other hand, the critics of the SAR Agreement contend that it isanother facet of the causal theory that limits the true potential of government to helpsocial assistance recipients. “Supply-side social policy” does little to affect the largerconditions in society that perpetuate poverty. One of the theories underlying theSAR Agreement is that social assistance recipients lack the necessary skills to get ajob. Anti-poverty advocates argue that this viewpoint stigmatizes social assistancerecipients by portraying them as deficient in some way, and deflects from the trueproblem, that is, the lack of decent jobs in society. In fact, the most commonly heardcomplaint from both social assistance recipients and field workers is that the traineesneed a job after training (NAPO 1989). This job must not make them worse off thanthey were on welfare. Although it did result in an overall increase in employment,the SAR Agreement did not lead to an increase in the wage paid to individuals inlow-paying work in the vast majority of cases. This desired “training” effect wasvirtually absent perhaps because of the short-term nature of most training. Toprepare social assistance recipients for more skilled and highly-paid work wouldrequire training programs that are longer and more intensive.The two causal theories underlying the SAR Agreement are implicit in itsdesign. One theory, that by addressing the particular needs of income assistancerecipients governments can help them to become self sufficient, seems valid. Theother, “supply-side social policy” seems weaker: there simply are not jobs for alltrained social assistance recipients.865. The enabling legislation and the institutions and norms created by it providethe target group with sufficient incentives to engage in the desired behaviour, andremove any disincentives to target group compliance created by factorsboth”extemal” and political that define the experience of the target group.Consistent with the liberal developmentalist approach, the framers of the SARAgreement did not believe that social assistance recipients needed to be coerced intoparticipating in employment enhancement activities. Rather, they saw theimportance of removing significant disincentives to leaving welfare for work. Thegreatest disincentive experienced was child care. A federal report indicated that 40per cent of mothers interviewed maintained that without support for child care theywould not be able to participate in training and educational upgrading(Employment and Immigration 1993: 46). Other significant barriers faced were theincreased costs of transportation to the training location, as well as the loss ofmedical and dental benefits that they would have received while on welfare. Overtime as these disincentives were revealed to the SAR Management Committee byfield workers, the committee increased the supports available to trainees and toindividuals during the year of transition from welfare to work. The objective was toensure that social assistance recipients would not be made worse-off by engaging intraining. The Enhanced Earnings Exemption introduced by British Columbiaallowed trainees to keep more of what they earned, and the 1991 Agreement wasupdated to explicitly provide that participants would receive “at a minimum, theequivalent of cash and non-cash benefits that they would have been eligible for onsocial assistance as well as contributions towards the increased costs associated withparticipation in a particular program” (SAR Agreement 1991, Section 5.01).Federal research found that a large number of those who took part in trainingprograms would have participated even if they had been made worse off financiallyby it (Employment and Immigration 1989). In the SED option programs, 60 per centof participants claimed that they would have engaged in training and upgrading in87the absence of a wage or allowance; 35 percent of Job Entry trainees claimed thesame thing (Employment and Immigration 1993: 43; 1989: 106-7). This evidenceindicates that the primary incentive to participate was not immediate financial gain.Federal and provincial officials were quick to point out that the greatest incentivewas the real opportunity that these programs were perceived to provide. Officialsbelieved that part of the reason for the success of the various SAR initiatives was thepositive light in which upper management portrayed the program to “street-level’Tbureaucrats and in which they, in turn, portrayed them to their clients. Theprograms offered clients a chance to better themselves through increased self-confidence, employability and annual earnings, and at the same time removed someof the major disincentives to participating in training.It is important to recognize, however, that the group asked about theincentives to participate in the programs were those who chose to engage in training.There is no way to determine how many potential trainees still do not have adequateincentives to participate. In fact, those with the most barriers may not be able tobenefit from training since to move off of welfare would result in too manyhardships. For example, for social assistance recipients with many dependents thewage earned through the Employment Subsidy Program would be significantly lessthan the income assistance levels provided for a large family, since wages do notvary according to family type. Incentives, thus, are sufficient to attract enoughtrainees to more than fill welfare-to-work programs, but they may not be sufficientto attract all of those for whom the SAR Agreement was originally intended.6. The relative priority of statutory objectives is not undermined over time by theemergence of conflicting public priorities, changes in constituency groups’support or by changes in the relative socio-economic conditions which weakenthe statute’s causal theory or political support. External conditions are conduciveto the achievement of the policy’s objectives.88From the point of view of those not on income assistance there were strongincentives to support welfare-to-work training in British Columbia. Rapidlyincreasing welfare case-loads placed an increasingly great financial strain on taxpayers throughout the 1980s. The notion of giving welfare recipients “a shovel” andnot “a hand-out” gained increasing salience for the population. As such, the SARinitiative was well met by British Columbia’s politicians and quickly became apriority in the province, which has not been undermined by other priorities. In 1986,distinct from the SAR management committee, B.C. set up its own EmploymentInitiatives Projects to explore new possibilities in provincial programming. Rapidgrowth in provincial spending, from a few million to over $40 million by 1988, eventhough it was not matched by the federal government, demonstrates the provincialgovernment’s commitment to the principles of the SAR Agreement.On the federal side, the commitment to assisting welfare recipients to gainindependence was also strong, although funds did not flow as freely as those of theprovincial government towards this target group. Nevertheless, the SAR budgetwas maintained throughout the first five years of the Agreement and then increasedto $28 million from $15 million in 1991 even as mainstream CJS training fundsdeclined over these years. Political support for the SAR Agreement was maintainedduring this period. At the local level, the approval of Members of Parliament wassought for federal CJS projects. In general this approval was easily won, since job-creation schemes in a Member’s riding are popular politically. Only more recently isthe political support wavering as Reform Party members challenge what they regardas “liberal” programs.Finally, at the local level employment enhancement activities have lastingpopularity. Training projects not only bring benefits to the job-training host and tothe social assistance recipient, they often also provide direct benefits to thecommunity through the fruits of the trainees’ labour. For example, the revamped89Forest Worker Development Program has produced a more efficient corps ofworkers who often finish their contracted work before the training period iscomplete. The team manager then consults with the local community to decide on aproject upon which the trainees can then embark. In the remaining days of theirtraining period trainees thus produced tangible benefits, such as cleared trails or theconstruction of a shelter, for their communities.The troubling socio-economic trend that prompted the original SARAgreement --a trend towards the welfare case-load being composed of“employables” -- did not disappear over the past decade. The programs developedto combat this trend have met with a good deal of success. These two facts havesecured the relative priority of welfare-to-work initiatives on the social policyagenda as well as their popularity in the communities into which they areintroduced.One possibility that should not be ignored is that the existence of externalconditions conducive to the achievement of objectives of the SAR Agreementinfluenced its success. In spite of the agreement’s ability to confirm many of the keyhypotheses about effective policy implementation, the existence of severalfavourable external conditions casts some doubt as to whether the satisfaction of theother seven conditions of implementation in any jurisdiction would ensure thesuccess of welfare-to-work initiatives. In fact, there are two conditions external tothe policy process which might have accounted for the success of the B.C. SARAgreement.First, in order for the goal of self-sufficiency to be achieved by welfarerecipients they must find permanent employment. In the absence of a relativelystrong economy even a well-trained individual might not be able to findemployment. In British Columbia, with one of the lowest unemployment rates andthe strongest economies of the Canadian provinces, jobs have been relatively90plentiful. As such, the success rates of the programs in British Columbia may be aresult of a strong economy rather than a particularly strong package of welfare-to-work programs. A study by Employment and Immigration Canada does suggestthat this hypothesis is partly true. A comparison group of non-CJS social assistancerecipients saw their employment prospects increased during the period of theexperiment, but to a lesser extent than CJS participants on social assistance(Employment and Immigration Canada 1994).The statistics from the federal study suggest another external factor thatmight have influenced the success of SAR in B.C. The target group referred to CJSprograms in the province were generally more job-ready than the national average.Both the CJS participant group and the comparison group had an average of an extrayear of education in comparison to the national average, a characteristic whichmight have led them to require less upgrading than other social assistance recipientsin Canada before they were ready for employment and thus might have facilitatedtheir transition from welfare to work. This characteristic of the welfare populationmight have resulted from the tremendous migration of workers into the province inthe last decade. These workers might already have been fairly job-ready but turnedto social assistance to ease their transition during the first year of their move. (Thisspeculation is supported by the statistic that in-migrants are twice as likely to receivesocial assistance in the first year after moving as they are in the second.)In general, then, the favourable external conditions in which the B.C.initiatives were implemented likely played some role in their success. However, thefederal CJS analysis using comparison groups in the province suggests that withinthe province social assistance recipients who train still have an advantage over thosewho do not, regardless of the similar external circumstances affecting both groups(Employment and Immigration Canada 1994).917. The enabling legislation ensured that “learning” could take place and that theimplementation process was flexible enough to incorporate past lessons.In addition to mandating increased activity to enhance the employability ofsocial assistance recipients the Canada/British Columbia SAR Agreement also calledfor a program of evaluations to be set up to assess the impact of these programs. Ina system in which so many unrelated activities are taking place at the local level aformal plan for evaluation and information-sharing was necessary to assess wheremoney was being well spent, and where it was improperly targeted. Health andWelfare Canada provided support for a large portion of the evaluations;Employment and Immigration Canada conducted their own assessments of CJS.The capacity of the evaluations to generate useful information about theeffectiveness of employment enhancement initiatives was complemented by theflexibility of the initiatives to adapt to the lessons learned. Because the managementcommittee met frequently and produced a new mix of programming every year, andbecause projects were approved and renewed frequently at a local level, the federaland provincial training programs in B.C. became better over time as a result of pastexperience. The evaluation component of the initial agreement was a feature new tomost Canadian welfare-to-work schemes.8. The enabling legislation or other legal directive mandates policy objectiveswhich are clear and consistent or at least provides substantive criteria forresolving goal conflicts.The need for clear guidelines and standards might appear at first to stand inconflict with the second condition of the revised framework which suggests thatlocal discretion is one of the main reasons for the success of SAR. In fact, as waspointed out at the end of the discussion of the second condition, discretion isvaluable if the staff has the expertise to make the appropriate decisions to advancethe objectives of the agreement. In the absence of this expertise, discretion canactually work to undermine effective implemention, a danger against which clear92guidelines serve as a buffer. Guidelines can provide a general direction withinwhich street-level bureaucrats can use their discretion.The SAR Agreement was hurt by a lack of clear guidelines to provide field-level officials with a direction during the early years of implementation. The 1986agreement was a rather vague document with few guidelines, open to manyinterpretations at the field level. The province maintained this ambiguity in layingout its own programs. The federal training system was far more regulated.Although the enormous discretion permitted by a lack of guidelines could be seen asan advantage in terms of allowing local offices the flexibility to design programs thatwere appropriate to their particular areas and target populations, federal officialsclaim that they did not feel constrained by their guidelines. Rather, they providedthem with greater sense of the purpose of the SAR Agreement and allowed them touse the discretion that they did have more wisely (Interviews 1994). Whereguidelines did not inform street-level officials of how best to make use of the SARAgreement in their area there was an increased likelihood of waste, of misdirectionof funds and poor targeting (SP Research Associates 1988: 88-90).A significant disadvantage of a lack of clear guidelines and standards wasthat it resulted in inconsistency from one training area to the next. Whereas a socialassistance recipient might have access to a variety of programs and supports in oneoffice, if she moves to another town she might find that such services are notavailable or that she is not deemed eligible for those programs by the local criteria.This problem was cited more by provincial officials than by federal officials.Federal officials at the local level felt that their useful guidelines were a result of thematurity of the federal training apparatus. They felt that the guidelines allowed foraccountability, and guided field workers in their decision-making, while stillallowing them sufficient discretion to respond to local needs.93In the absence of guidelines at the provincial level, projects were approvedperiodically which lacked a genuine training component, and inappropriate clientswere sometimes referred to projects. Field-level workers might be left to their ownresources without a clear sense of direction in their decision-making. In addition, inthe absence of standards, it is difficult to maintain accountability for the decisionsthat are made. In general, the potency of the SAR Agreement was watered down bythe absence of guidelines and standards in the training system.Explaining Individual Programs’ Successes and FailuresThe success of the SAR Agreement overall in promoting self-sufficiency forsocial assistance recipients must not overshadow the important lessons about thedeterminants of successes and failures of specific programs within the large packageof programming that resulted from the agreement. The most successful programs,that is, Job Entry, Job Development and the Severely Employment Disadvantagedprograms (SED) on the federal side and the Wage Subsidy programs, R.I.S.E.programs and Job Action on the provincial side demonstrate the importance ofcertain of the implementation variables. On the other hand, the less successfulprograms, that is, the three provincial public employment programs and the FederalDirect Purchase Option failed to meet several conditions critical to effectiveimplementation.Job Entry, SED programs and R.I.S.E. programs were successful largelybecause of the validity of the assumptions which underlay their designs. Theirincremental and integrated approach to the upgrading of the individual through avariety of supports and services addressed effectively the multiple barriers faced bysocial assistance recipients. They were often small programs, designed for aparticular target group, responsive more to individual than to local needs. A varietyof supports gave social assistance recipients the incentives to enter one of these94programs. The programs were closely supervised by a field-level worker, and thecontractors were carefully assessed before a project was accepted or renewed. Theconsiderable discretion given to the local staff to approve, design and overseeprojects led to creative approaches to addressing the needs of clients. Finally,realistic expectations about what constituted a success led to programs providing avariety of opportunities that were not strictly related to jobs. In some cases,counselling and life-skills instruction was added to more formal instruction andtraining and led to a more integrated approach to improving the situation of theindividual social assistance recipient.Job Development and the Employment Subsidy programs were similarlyaffected by a valid causal theory and skilled and committed field-level workers. On-the-job placements which included meaningful training in a field in which futureemployment was likely provided workers with marketable skills and increasedconfidence, increasing the likelihood that they could find another job. Field-levelstaff from CEC and MSS offices monitored the projects and formed valuable linkswith employers. The work that trainees did was valuable to the employers and ingeneral employers were pleased with their productivity (Ministry of Social Services1992). Since the job experiences were ‘real”, the satisfaction level for all partiesinvolved was higher than it would have been in make-work projects. JobDevelopment and Employment Subsidy contracts were developed at the local levelso that they could be most responsive to the needs of the target population as well asthe labour market.The danger to the wage subsidy programs is that the differing motivations ofimplementing actors might undermine the objectives. Projects which wereunsuccessful were often ones in which the employer was not committed toproviding training that would lead the social assistance recipient to futureemployment. Employers might be more interested in acquiring cheap labour than in95helping the social assistance recipient. In cases in which the orientations of actorsare different it is a real challenge to avoid these kind of implementation games. It isup to the field-worker to assess the potential of the employer to behave as desired,and to try to foster a cooperative relationship in which behaviour will fall withinpredictable limits.Although Job Action was a short program with limited resources, theattention it gave to assisting social assistance recipients to find work, rather thanleaving them to their own resources, was valuable. Participants moved off ofwelfare into work faster than non-participants. This could be attributable to theincreased confidence, contacts and support gained through this program, all ofwhich were part of a program design flowing from a valid causal theory whichstates that given opportunities social assistance recipients will choose to pursue self-sufficiency.The programs that were less successful were so in general because of theincorrect causal theory underlying the program and the fact that they were assignedto agencies whose behaviour could not be controlled through the intensecooperation of the four partners involved in the SAR Agreement. The threeprovincial public employment programs provided work without providingmeaningful training in fields in which future employment could be expected. Thosetrained in the tourism or environmental fields, or in the early years of the forestryprograms, for example, often performed tasks which were not in great demand inthe labour market. The agencies which trained them did not have the funds toemploy them after training to continue to use the narrow training that they receivedfrom the program. In addition, they were often given make-work” which neitherbenefited the employer particularly nor did it give the trainee the sense ofaccomplishment that arises from performing meaningful work.96The private employment programs were overseen by ministries other thanSocial Services or Advanced Education and Training. The priorities of thoseministries overseeing the public programs, including the Ministries of Forests, ofEnvironment, and of Tourism, are different from those ministries whose primaryconcern is helping individuals to become self-sufficient. As such there was lessunderstanding of the needs of social assistance recipients and less commitment toaddressing their particular needs than would be the case in a program administeredby the Ministry of Social Services. In addition, these ministries were not partners inthe SAR Agreement, thus they did not have the same mandate for constantconsultation and cooperation. There was little way of ensuring that officials andtarget group would behave in the desired manner.Finally, the relatively low success rate of Direct Purchase Options can beexplained by the same factors that explained the success of Job Entry options.Unlike Job Entry, DPO did not provide the integrated supports necessary for thesocial assistance recipient to become job-ready. It merely provided classroomtraining and then left social assistance recipients to fend for themselves. DPOprograms were not as closely monitored by field workers as Job Entry Options, andthey were more structured, in a strictly classroom environment, so programs weregenerally less able to respond to the individual needs of social assistance recipientsthan were other training options.The variations of project successes within the larger program types can beexplained by the presence or absence of any number of the eight conditions foreffective implementation. For example, one region in B.C. might have moresuccessful programs than another region with the same type of program because ofbetter coordination between the local offices of both governments, because of theexpertise of a few staff members or because the target population is more job-ready.With a seemingly infinitely divisible initiative such as the SAR Agreement, the97conditions which contribute to the success of the whole package at a provincial levelalso contribute to successes or failures at the interface between caseworker andclient.ConclusionMazmanian and Sabatier’s framework, with a few revisions and additions, isvery effective at explaining the fate of the implementation of the SAR Agreement. Itis most powerful when its conditions are expanded and interpreted to apply to field-level bureaucrats as well as to the higher-level officials who formulate policy. Infact, by integrating the lessons of Elmore and Lipsky-- the bottom-uppers—Mazmanian and Sabatier’s theory takes on a meaning that they did not intend.Policy is no longer made at the upper reaches of government but rather is shaped bythe decisions of lower-level bureaucrats. Through the addition of a new variable,that of “learning” in policy, the theory no longer allows for the separation of policyformulation and implementation that Mazmanian and Sabatier had intended.Rather than compromising the integrity of Mazmanian and Sabatier’s model,the contributions of other authors to it makesthe model more widely applicable. Forexample, their model, concerned as it was with the “command and control” issues ofpolicy formulation, does not deal with the realities of the implementation of socialpolicies. It does not acknowledge the separate motivations acting upon two distincttarget groups -- street-level bureaucrats, and the ultimate recipient of the policy’sprovisions who in this case-study are social assistance recipients. Social policy mustcreate an environment in which street-level bureaucrats will want to carry-out themandate of higher level officials. At the same time it must be designed to takeadvantage of the incentive structures acting on the ultimate target population inorder to ensure that their behaviour can be made consistent with policy objectives.98This entails understanding the social and economic context in which the targetgroup lives from day-to-day.The revised framework can easily be applied to the Canadian policyenvironment when the United States-specific conditions are made more general.Mazmanian and Sabatier’s concerns about the role of fixers and constituency groupscan be captured within broader categories about the skill and commitment ofimplementing officials, and about the priority of statutory objectives, respectively.The new categories are broad enough to permit such additions, but discreet enoughfor the framework to be prescriptive, rather than merely descriptive.Despite the complexity of the problem that it addressed, and the fact thatmany similar programs had failed before it in Canada and the United States, theSAR Agreement was responsible for measurable improvements in the employabilityand earnings of social assistance recipients, and for benefits to society as a whole. Byapplying the revised framework to the SAR case-study, a set of recommendationscan be generated for the creation of successful welfare-to-work programs in Canada.Successful welfare-to-work programs result from intensive cooperationbetween the various agencies and levels of government involved in training socialassistance recipients. This cooperation maximizes the resources available, helps toavoid duplications and gaps in programming, and keeps the communication ofinformation on individual projects and welfare recipients strong. Local discretionallowed by extensive decentralization also helps to generate successful programs,since officials very familiar with the needs of the target population, the trends in thelocal labour market, and the activity of other local agencies are able to designprograms that are responsive to all of these variables. At the same time asproviding them with discretion, a successful initiative provides officials withsufficient guidelines and standards to give them a clear sense of purpose, and ofprofessional accountability. Standards also breed programs that treat clients with99consistency throughout the region. At the same time, the managerial and politicalskills and commitment of field-level workers influence the quality of the projectsdesigned to serve social assistance recipients and the appropriateness of the referralsmade by these workers to particular training projects. Successful programs are thosethat embody a valid causal theory in the design, by providing opportunities forclients that include work experience in fields in which there is a labour marketdemand and a chance of future employment, that are responsive to client needs,consist of an integrated array of training and upgrading activities, and that aresufficiently long in duration to upgrade the trainees skills. Successful programsprovide a range of supports and transitional benefits that are sufficient to at leastreduce the disincentives, if not actually provide incentives, to the target group toparticipate. Well-funded programs have more opportunities to achieve success. Aprogram is more likely to meet with success if it includes in its design on-goingevaluations, and if it is flexible enough to adapt to past lessons and thus improveover time. Finally, if the local economy is strong, and the target population more jobready, a program is more likely to achieve its objectives of moving social assistancerecipients into the workforce.Beyond these conditions for success lie further conditions, even morechallenging to meet, and a definition of success which is beyond that provided in theSAR Agreement. True progress for social assistance recipients would mean theelimination of their poverty. However, the conditions that would satisfy this goalare not those that can be manipulated within the implementation process for the theSAR Agreement. They include raising the minimum wage significantly,guaranteeing pay equity, reducing discrimination in the workforce, and providingfree, quality child care. The realization of these goals would entail an enormousrestructuring of the way Canadian society now functions, and a tremendous expensefor the government. Given the virtual revolution that it would take to satisfy the100conditions as a package, it is wise to pursue these various goals individually, inconjunction with continuing the pursuit of the objectives of the SAR Agreement, aninitiative that has carved out modest, but significant gains for the population that itwas designed to serve.Welfare-to-work initiatives are here to stay in British Columbia and inCanada. Given both governments’ perceptions of the success of many of theirprograms, and the continuing phenomenon of tremendous welfare dependence foremployable welfare recipients, training programs are a promising route out of bleakemployment situation of the country. Programs will continue to evolve asgovernments examine the lessons of past successes and failures. Strategies willchange as politicians confront new social and economic realities.The British Columbia government has recently moved all of its trainingprograms for social assistance programs out from under the wing of the Ministry ofSocial Services to the Ministry of Skills, Training and Labour. This ministry isattempting to redesign the province’s training system to provide a training andsupport system that is ongoing, that supports the social assistance recipient aftertraining and into employment and that continues to emphasize “capacity” tools in itsapproach to the work/welfare dilemma. This will entail an even greatergovernment investment, and will renew British Columbia’s commitment to thistarget group.The federal training system, similarly, will undergo restructuring. Theconcept of “mutual responsibility” between governments and social assistancerecipients is becoming increasingly salient for Canadians.This restructuring will only serve to increase the importance of welfare-to-workinitiatives, which embody this notion of responsibility. The precise content ofreforms is as yet uncertain, but the direction is assured. In response to the success ofsome of the most innovative programs in social policy-- those crafted at the local101level and serving local needs-- decentralization will continue to characterize thetraining system. In addition, training will become even more central to the socialassistance agenda and may even be tied to welfare benefits in the future, heraldingthe ultimate eclipse of the classic Canadian redistributive approach to welfare by theemerging liberal developmentalist ideology.102BibliographyArmitage, Andrew. (1991) “Work and Welfare” in Alternatives to SocialAssistance in Indian Communities, Frank Cassidy and Shirley Seward,eds. Halifax: Institute for Research on Pulbic Policy.Bailey, Nicholas. (1994) The Effect of the B.C. Enhanced Earnings Exemption on theProbability of Leaving Social Assistance. Prepared for B.C. Ministry ofSocial Services.Bardach, Eugene. (1977) The Implementation Game: What Happens After a BillBecomes A Law. Cambridge: MIT Press.British Columbia, Ministry of Human Resources. (1986) Annual Report 1985-86.British Columbia, Ministry of Social Services and Housing (1986-90) AnnualReports.British Columbia, Ministry of Social Services. (1989) Final Report of the ProcessEvaluation of the Employment Plus Program. Prepared by The DPAGroup, Vancouver.British Columbia, Ministry of Social Services. (1992) Routes to Independence: TheEffectiveness of Employment and Training Programs for IncomeAssistance Recipients in British Columbia. Victoria: Research,Evaluation and Statistics Branch, Corporate Services Division.British Columbia Ministry of Social Services. (1993) A System Under Siege.Callahan, Marilyn et al. (1990) “Workfare in British Columbia: SocialDevelopment Alternatives” in Canadian Review of Social Policy, No. 26.Canada Assistance Plan (Consolidated 1990)Canada/British Columbia Agreement to Enhance the Employability of Social AssistanceRecipients, 1986. (Revised 1991).Canada/British Columbia SAR Agreement Annex ifi. 1993/94 and 1994/95.Canadian Labour Market and Productivity Centre (1990) Report of the CLMPC TaskForces on the Labour Force Development Strategy. Ottawa.Eliwood, David T. (1988) Poor Support: Poverty in the American Family.New York: Basic Books.103Elmore, Richard. (1979) “Backward Mapping: Implementation Research andPolicy Designs” in Political Science Quarterly Volume 94 Number 4.Employability Resources Network. (1990) Selected Bibliography onEmployability. Ottawa.Employment and Immigration Canada. (1987) The Canadian Jobs Strategy: AReview of the First Eighteen Months. Ottawa.Employment and Immigration Canada (1989) Evaluation of the JobDevelopment Program, Final Report. Prepared by Goss, Gilroy &Associates Ltd. Ottawa.Employment and Immigration Canada. (1989) Evaluation of the Job EntryProgram. Final Report. Prepared by Abt Associates of Canada.Employment and Immigration Canada (1992) Employment Services andPrograms.Employment and Immigration Canada (1993) Evaluation of the SeverelyEmployment Disadvantaged (SED) Option of the Job Entry Option.Ottawa: Program Evaluation Branch, Strategic Policy and Planning.Evans, Patricia M and Eilene McIntyre. (1987) “Welfare, Work Incentives and theSingle Mother: An Interprovincial Comparison”, in The Canadian WelfareState. ed. Jacqueline Ismael, Calgary: University of Calgary Press.Evans, Patricia M. (1992) “Targeting Single Mothers for Employment:Comparisons from the United States, Britain and Canada” in Social ServicesReview, September.Evans, Patricia M. (1993) “From Workfare to the Social Contract: Implication forCanada of Recent U.S. Welfare Reforms” in Canadian Public Policy No. 19.Federal-Provincial-Territorial Conference of Ministers of Social Services (1986)Agreement Regarding Enhancing the Employment Opportunities forSocial Assistance Recipients. Ottawa, Sept 17-18.Goggin, Malcolm et al. (1990) Implementation Theory and Practice: Toward aThird Generation. Glenview: Scott, Foresman/Little, Brown HigherEducation.Golding, Peter and Sue Middleton (1982) Images of Welfare: Press and PublicAttitudes to Poverty. Oxford: Martin Robertson, 1982.Guaranteed Available Income For Need Act (1989) and Regulations (1993)104Gueron, Judith M. and Edward Pauly. (1991) From Welfare to Work. NewYork: Russell Sage Foundation.Handler, Joel and Yeheskel Hasenfeld. (1991) The Moral Construction of Poverty.Newbury Park: Sage Publicatons.Health and Welfare Canada. (1985) Statement by The Minister of NationalHealth and Welfare on Enhancing Employment Opportunities for SocialAssistance Recipients. Ottawa, Sept. 19.Human Resources Development Canada. (1994) Evaluation of EmployabilityInitiatives for Social Assistance Recipients (SAR5’) in CJS. Ottawa: ProgramEvaluation Branch.Kaus, Mickey (1986) “The Work Ethic State”, in The New Republic, July 7.Lightman, Ernie. (1991) “Work Incentives Across Canada” in Journal of CanadianStudies, Vol. 26, No.1.Lightman, Ernie (1991a) “Support for Social Welfare in Canada and the UnitedStates” in Canadian Review of Social Work, Vol. 28.Lipsky, Michael. (1980) Street-level Bureaucracy. New York: Russell SageFoundation.MacDonald, Hon. Flora. (1985) Addess at Federal-Provincial-TerritorialConference of Ministers of Social Services, September 17-18.Mazmanian, Daniel and Paul Sabatier. (1983) Implementation and PublicPolicy. Chicago: Scott Foresman and Co.Morris, Michael and John B. Williamson. (1989) “Workfare: ThePoverty/Dependence Trade-Off” in Policy Studies Review Annual, Vol. 9.National Anti-Poverty Organization. (1989) You Call It A Molehill. I’ll Call It AMountain: Job Training For People on Social Assistance. Ottawa.Ontario: Ministry of Community and Social Services. (1988) Transitions,Summary. Toronto: Queen’s Printer for Ontario.Ontario. Social Assistance Review Committee. (1987) A Review of Trainingand Educational Programs for Social Assistance Recipients Entering theLabour Force. Prepared by Burt Perrin Associates. Toronto.105Phillips, Stephen. (1992) “Public Policy Implementation: An Assessment of theApplicability to Canada of the Mazmanian-Sabatier Model”.Pressman, Jeffrey and Aaron Wildavsky. (1973) Implementation. Berkeley:University of California Press.Prince, Michael. (1989) Background Paper on Programs for Increasing theEmployability of Social Assistance Recipients. Ottawa: Canadian LabourMarket and Productivity Centre.Prince, Michael and James Rice (1990) “The Canadian Jobs Strategy: SupplySide Social Policy” in How Ottawa Spends. Katherine A. Graham, ed.Ottawa: Carleton University Press.Rom, Mark. (1989) “The Family Support Act of 1988: Federalism, DevelopmentalPolicy, and Welfare Reform” in Publius: The Journal of Federalism, Summer1989.Schneider, Anne and Helen Ingram. (1990) “Behavioural Assumptions ofPolicy Tools. In Journal ofPolitics Vol. 52, No. 2.Schneider, Anne and Helen Ingram (1993) “Social Construction of TargetPopulations: Implications for Politics and Policy” in American PoliticalScience Review, Vol. 87, No. 2.SP Research Associates. (1988) A Review of the Employability EnhancementInitiatives for Social Assistance Recipients agreed to byFederal/Provincial/Territorial Ministers in September. 1985. Regina.SP Research Associates. (1989) National Program Meeting on theEmployability Enhancement Initiative in Ottawa, Report. June 12-15,1989.SP Research Associates. (1993) “What Works For Whom”: EvaluatingEmployment and Training Programs for Social Assistance Recipients inCanada, A Review of the Literature. Prepared for Health and WelfareCanada.Stoker, Robert. (1991) Reluctant Partners: Implementing Federal Policy.Pittsburgh: University of Pittsburgh Press.Stone, Deborah A. (1988) Policy Paradox and Political Reason. Harper CollinsPublishers.Task Force on Program Review. (1985) Jqh Creation, Training and EmploymentServices. Ottawa: Ministry of Supply and Services.Weaver, Kent. (1994) “Old Traps, New Traps” in The Brookings Review,Summer.Yalnizyan, Armine and David A. Wolfe. (1989) Target on Training: MeetingWorkers’ Needs in a Changing Economy. Toronto: Social PlanningCouncil of Metropolitan Toronto.106


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items