Sunday, September 28, 2008

A Pragmatic Pluralist Response to Daniels & Trebilcock, Rethinking the Welfare State: The Prospects for Government by Voucher.


This paper seeks to develop several observations and arguments made in my review of Daniels’ and Trebilcock’s book, which was published in Canadian Public Policy in 2007. The factors adduced that could limit the use of tied demand-side subsidies in reforming the welfare state include: the degree of ‘voucherness” already in existence; the degree of individualism characteristic of a given social policy sub-field; nature and strength of external preferences; market structure of potential suppliers; potential exposure to international trade rules, the likelihood of design challenges and design failures in voucher experiments, and, at the highest level of generality, the relationship between the means and ends of social policy (e.g. the analytical separability, logical connection, and dynamic interactions which exist between means or instruments and goals of social policy).

Just as the strength of any critique based upon any analysis of these factors can be expected to vary between various social policy areas, so too can the strength of the voucher argument be said to vary between policy sub-fields. The resultant analysis eschews rationalist neo-liberalism in the sense of either strong value neutrality or uniform individualism, which are both prevalent in Daniels and Trebilcock’s work, as well as the work of the Fraser Institute. It does however embrace the liberal value of pragmatic pluralism, and therefore likely implies the endorsement of voucher experiments in at least a couple of areas.


The creation of public markets (i.e. inviting providers to compete for public monies to supply a public good) has been described as “the most significant change in the face of the state since the creation of merit-based bureaucracies to manage the delivery of public goods a hundred years ago (Stein, 116).” It has developed in association with the broader trend known as New Public Management, according to which “the central failure of government today is one of means, not ends” (Osbourne & Gaebler, xxi), and governments are deemed to be better at “steering” rather than “rowing”. In Rethinking the Welfare State: Prospects for Reform By Voucher, Ron Daniels and Michael Trebilcock ask what the state would look like if it relied on vouchers to deliver public goods and services across the broad spectrum of activities that comprise the modern welfare state.They argue that in 8 different social policy fields (food stamps, housing, legal aid, health care, early childhood education, K-12 education, post-secondary education and labour market training) “providing that sponsoring governments are steadfastly attentive to demonstrating fidelity to public goals and values through their design, implementation and monitoring of voucher programs, ... [v]ouchers can, we believe, achieve a more satisfactory optimization of efficiency and equity goals than many competing instruments” (Daniels & Trebilcock, 221). In a slightly more exuberant formulation, that authors approvingly quote Julian Le Grand in saying that “vouchers convert citizen beneficiaries of the welfare state from pawns, the least powerful piece on the chess board, to queens, the most powerful. Thus, vouchers demonstrate considerable promise in terms of their ability to realize public goals. Indeed, given this potential, the adoption of voucher programs may be able to mute, or even obviate, the great efficiency-equity trade-off of which Arthur Okun wrote so eloquently in his celebrated essay by that title” (Daniels & Trebilcock , 2).

This paper argues that the considerable merit of Daniels’ and Trebilcock’s premises must be accepted if we are to make any headway in making an incisive and fair critique. They are right to remind us what vouchers are, the considerable extent to which voucherness is already a characteristic of the welfare state, and that there are considerable gains to be made in terms of both efficiency and choice from expanding and carefully designing voucher programs. What we need to ask is (1) whether the ends of public policy in these various fields are suffciently clear and discrete so as to permit the “optimization of efficiency and equity goals”; (2) to what extent it is realistic or feasible to consistently achieve consumer sovereignty and tame the power of suppliers through intelligent design; and (3) beyond the question of the poverty of ends and the pragmatic difficulties of turning pawns into queens, we also need to raise the discussion to the level of political theory: does the ‘obviation of the great efficiency-equity trade-off’ represent a defensible ideal for perfection of the welfare state?

In this regard, there is a noteworthy ambiguity surrounding the relationship of the voucher ideal to the liberal ideal of pluralism. On the one hand, as Joseph Heath has put it, “[e]fficiency has replaced religion as the primary source of value in our culture. This is because striving for efficiency is one of the few ways we have to achieve social order in the context of a pluralistic society” (Heath, 9). On the other hand, there is another tradition within liberalism, in the thought of Isaiah Berlin and Bernard Williams, and currently exemplified by Joseph Raz and John Gray, which is skeptical both of utilitarian and Rawlsian tendencies to develop a common metric for ‘solving’ political problems (Gray 1995: 8-16). These authors argue for a deeper pluralism, and a richer concept of politics, which I argue is antagonistic to such rationalist reconstructions of the welfare state as represented in Rethinking the Welfare State. Neverthless, a truly context-sensitive, deeply pluralistic perspective must allow for the possibility that vouchers do represent the best policy instrument in certain areas, and I argue that for the most part these are the same areas indicated by the practical considerations of avoiding design failure.

I. The Structure of the Argument

Daniels and Trebilcock begin by pointing out that vouchers (defined as “tied demand-side subsidies”) are simply a policy instrument. As such, they are compatible with any legitimate goal of the welfare state regulation of public morality; building social solidarity; insuring individual risk; promoting economic stability, and providing an equitable distribution of resources. The authors provocatively assert that “a decision to invoke the voucher instrument does not in itself reveal much about the underlying character of the policies being pursued by the sponsoring government” (p.2), and that opponents who think otherwise routinely make an “illogical connection between means and ends” (p.12). Their primary claim is that choice by individual consumers of public services on the demand-side, coupled with competition on the supply-side, with explicit, targeted subsidies following consumers rather than suppliers, ensures that decisions regarding the consumption and production of public services will be made efficiently, and that citizen beneficiaries will be empowered. Their secondary claim is that, where their primary claim does not appear to be borne out by the evidence, “this is often because of design failures” (p.13).

The authors admit “that is not to say that reliance on vouchers is bereft of any normative content whatsoever” since enabling citizen choice in the consumption of publicly funded goods and services increases “the scope for individual autonomy” (p.2). Nevertheless ,

“as against alternative instruments, the case for the voucher is relatively straightforward: once governments have decided to intervene in a given policy conferring explicit, targetted subsidies on individual citizens, decisions regarding the consumption and production of public goods and services will be made more efficiently. This results from the fact that vouchers place resources directly into the hands of citizens, where they, rather than a governmental agent, will determine the precise goods and services that will be consumed from among a number of competing suppliers, thereby increasing the likelihood that citizen preferences in the consumption of publicly supported goods and services will be more effectively vindicated. Specifically, the voucher concept introduces the prospect of market exit and hence failure on the supply side (e.g. bankruptcy) for institutions that are unable to attract and retain sufficient voucher-assisted consumers to ensure their survival and thus face a discipline that is markedly absent from many current government programs that depend exclusively or primarily on political voice to ensure accountability (p.2).
“From a Public choice perspective, one of the significant political attractions of government supply is that it allows governments to obscure differential entitlements that are predicated on politically contentious cross-subsidies. In contrast, because vouchers place government subsidies in the hands of citizens, there is a need for government to be more explicit in defining the precise nature of the benefit conferred. ...[T]he conferral of vouchers provides citizens with greater information respecting the distribution of government benefits, and, in particular, it facilitates critical evaluation of whether the distribution of benefits accords with legitimate normative principles or is simply the outcome of unprincipled political rent-seeking” (pp.39-40).

Thus, in order to counter what they see as largely irrational emotional resistance to the concept, the authors stress that vouchers are essentially value-neutral policy instruments consistent with a wide range of policy objectives, and that they serve to reduce waste and principal-agent problems inherent in public bureaucracy, both by empowering and informing citizens and by disciplining providers. They also point out that most of the policy areas being considered display degrees of “voucherness” already.

The remainder of this paper (1) looks at how each of these principal variables—the degree of voucherness already extent in particular policy fields, the degree of individualism in preference and choice on the part of citizen/consumers, the value and character of market accountability as opposed to political accountability, and the character and magnitude of design challenges—vary markedly between social policy sub-fields; and (2) questions the value of minimizing the equity/efficiency trade-off as a single over-arching value in the choice of policy instruments. The upshot of both analyses is to recommend the use of vouchers in a few limited contexts—in particular, postsecondary student aid and labour market training.

II. Voucherness

Rethinking the Welfare State makes the interesting point that many of our existing policies already display degrees of “voucherness”. Student aid programs give conditional grants and awards that are in effect “tied demand-side subsidies”. They also secure private loans to students by backing loan applications up to a certain amount, and by guaranteeing to banks that students will not be able escape debt, (e.g. by declaring bankruptcy). Food stamp programs in the United States are a clear example of tied demand-side subsidy. Even Canada’s medicare system is in effect a tied demand-side subsidy paid by a single government payer directly to physician and hospital-suppliers on behalf of patients, making it a limited form of “voucher” which nonetheless restricts consumer choice and supplier competition.

One purpose of identifying existing degrees of voucherness is to reduce the prejudice towards the concept that one finds in public sector unions and the ideological left. But there are a couple of other considerations that also relate to the cost and feasibility of vouchers. First, the pre-existence of voucher-like arrangements make it far easier to design and sell what is essentially an extension or refinement of the concept. Second, there is the possibility that voucher-like arrangements have sprung up spontaneously in those contexts where they are most suitable—as, for example, where the need for political/ administrative agency is low, there are low costs to market entry for potential suppliers, and the self-interest of citizen/consumers does not carry serious externalities for other citizens. There is some possibility of a divergence between Hayekian arguments (for grown or spontaneous market ordering) as opposed to the more rationalist/economistic designs in the Milton Friedman vein.

III. Individualism on the Demand Side and Market Accountability on the Supply Side

Worries about the civic consequences of privatization, such as those expressed by Martha Minow (“testing accountability requirements in the light of public values is not a task to be contracted out to private enterprise”—Minow:2003 p.1270) are met by the point that “agency problems that afflict public provision also afflict public accountability mechanisms” (p.20). Furthermore...because of the public sector’s reliance on certain processs values, “agencies are often free to pursue a substantive agenda that may or may not advance the public interest” (p.20). Daniels and Trebilcock look at the debate over the relative merits of public and private accountability from the perspective of principal-agency dilemmas, and are impressed by the potential of vouchers to ‘reconcile the interests of principals and agents’. Citizens as consumers have limited influence or oversight of services to them, and there is little incentive for public agencies to be highly responsive to them: the interests of providers and government agents are not strongly aligned with those of consumers as principals (p.23).

One can challenge the characterization of accountability as essentially a principal –agent problem in this way. I suggest that it has greater applicability in those circumstances where demands are not strongly influenced by negative preferences and the public goods and services are consumed individually rather than collectively. So, for K-12 education there may be strong external preferences by parents for schools that exclude or harm the interests of children other than their own; K-12 may also be a ‘social’ public good in the sense that is intended to be consumed collectively and interdependently. Post-secondary education, which involves adult students choosing from a menu of courses, exhibits a more essential individualism. Essential individualism is in turn more likely to be suited to market mechanisms of accountability, where a public interest (i.e. one that is not merely the sum of individual private interests) is less likely to be in conflict with the object of straightforward consumer satisfaction.

IV. Market Structure and the Characteristics of Suppliers

Daniels and Trebilcock mention that among the central design issues are “determining the qualifications of suppliers who can compete for voucher-assisted customers; ...... permissibility of extra-billing; permissibility of cream-skimming; information failures; [and] concerns about the inadequacy of supply-side market responses to vouchering , and the possibility that vouchers, under certain conditions,will simply entrench existing monopolies, or create new ones.” Theirs is an honest accounting of these design challenges, but one must ask: how often do the conditions for a truly competitive supply market truly exist? Even where markets are large and lucrative, how easily can corporations with deep pockets for lobbying, advertising, and litigating be subordinated to sovereign consumers? The authors admit that “the success of any given mode of delivery depends not only on the characteristics of the instrument itself, but also on the market conditions in which it will be operative” (p.36). It is relevant to ask just how successful the state has been in taming corporations in other regulatory contexts; or why, if the state is so prone to inefficient and rent-seeking, we have any reason to expect that the regulation required to make a voucher system work will work any better. Suffice to say that the challenges posed by regulatory capture and market structure can be expected to vary between social policy sub-sectors. One area where large transnational corporations and pharmaceutical companies can be expected to influence their regulatory conditions and not simply compete for vouchers is the health care sector.

V. Likelihood of Exposure to International Trade Rules

Daniels and Trebilcock state that “[w]e see no reason why for-profit entities should be prevented from participating in [hospitals] or indeed any other segment of the health care sector” [p.119, emphasis added]. This is a sweeping statement to make in the absence of any discussion of how a voucher-based welfare state might be more exposed to international trade rules (an ironic omission in view of Trebilcock’s recognized expertise in international trade law). My best guess is that the WTO General Agreement on Trade in Services (GATS) Article I:3 exemption for services “supplied in the exercise of governmental authority” would be much less likely to extend to Canadian public services under voucher-based policy regimes because of the way that that GATS provision is defined in Article I:3 (c) in terms of the conditions of supply, i.e. “supplied neither on a commercial basis, nor in competition with one or more service suppliers”.

Would Canadian public health and education measures still be exempted from NAFTA national treatment provisions under Annex I of NAFTA after they have been “revolutionized” in the way that Daniels and Trebilcock advocate? A single payer voucher scheme serving clearly defined public ends might meet the Annex II definition of a “social service established or maintained for a public purpose”, although the U.S. Office of the Trade Representative has already made statements which indicate that it would likely have an opposing view. In addition, the NAFTA Chapter Eleven expropriation clause enables U.S. or Mexican companies to bring compensation claims for any “indirect expropriation” of their investments in violation of Canada’s national treatment obligations.

These concerns reinforce the larger point that foreign corporate for-profit suppliers of public services would not just become efficient “instruments” of competitive delivery. They would also become powerful rights-bearing political actors , who would use legal and political channels to advance their own ends, making attempts by government to reverse competitive and private provision or extend public provision more expensive and difficult to do.

VI. The Poverty of Ends: The Case for Pragmatic Pluralism in the Delivery of Welfare State Programs

Simple, discrete policy objectives are more amenable to schemes for efficient modes of delivery, such as those afforded by well-designed voucher schemes. More complex policy objectives are better-suited to mechanisms of political accountability. Student postsecondary aid may be an example of the former, notwithstanding its many different functions and stakeholder interests: how to fund and empower student consumers, while achieving a balance between demand-side and supply-side funding. Healthcare is again an apposite example of the latter.

With respect to post-secondary funding, there are numerous efficiency and quality arguments against having exclusive reliance upon supply-side funding. Public universities have difficulty competing with elite private institutions for the best faculty and the highest levels of research funding; lower tuition fees are linked with higher drop-out rates, perhaps explaining why Quebec has not realized great improvements in access despite having very low tuition fees. Equity concerns are raised by the degree of taxpayer subsidization of students from affluent families when postsecondary tuition is free or very low. And, as no less an authority than Adam Smith once observed, "Beyond some point the higher level of endowment to any university, the lower its efficiency" (p.186).

But wait a minute--if supply-side funding out of general consolidated revenue is both inequitable and inefficient, doesn't that imply greater reliance on demand-side funding? And doesn't that mean--gulp--higher tuition fees? Yup. But that need not be a barrier to access, provided there is a generous income-contingent grant -and -loan program, or tied demand-side subsidy that is targetted to tuition fees. I suggest that at the federal level there be a $2 billion national tuition voucher scheme (or $1.65 billion, in view of the recent boost to the Canada Student Aid program), giving a tuition voucher up to $5,000 annually to the 400,000 neediest students , as based on family income and/or existing debt load, that would both significantly reduce average student loan debt and simultaneously increase university revenues. It would differ from both existing policy and the NDP platform in that it would allow universities greater latitude to compete for students on the basis of quality (paid for by higher fees, up to a voucher-related limit) as well as on the basis of price.

Daniels and Trebilcock state that “a universal voucher regime in health care provision, if properly designed, can provide effective, if bounded, choices to consumers, effective competititon by health care and health plan providers for consumer patronage, and effectively address concerns over equitable access to health care services” (p.127). Healthcare is not nearly so amenable to voucher solutions, because it has no such clearly defined objective, and probably cannot have. Bob Evans has noted that we do not want to buy health care, we want to buy health, but do not know what economists call the "production function" connecting them (Evans 1984). “But healthcare adds in another wrinkle – need”. (Deber 2008). Achieving a better efficiency-equity trade-off in this context might not be nearly as meaningful an achievement as in a simpler context where the goals are more clearly defined.

VII Conclusions

It is desirable, ceteris paribus, to achieve goals in a way that are both equitable and efficient. But it is easy to imagine scenarios where equitable and efficient modes of delivery frustrate political and/or technical processes of adjustment and negotiation. Where market actors are competing to supply consumer wants, ends that are complex, changing and/or ill-defined –that relate to larger political interests and which need to arbitrate deep philosophical divides—are unlikely to be well-served, regardless of either efficiency or equity. If there is a weakness in the argument of Daniels and Trebilcock, it may be that in such contexts, the efficient achievement of equity (or the equitable achievement of efficiency) may still be open to the riposte, “So what?” Added to this is the more mundane consideration, culled from the public administration literature, that governments which have trouble managing internal principal-agent problems will probably have difficulty managing external contracts as well; and the notable paradox that while vouchers elevate and arguably embody the principle of demand-side autonomy and choice, they are also susceptible to a high degree of paternalism---the conditions for demand-subsidies can be tied as tightly as the government wishes, as the case of food stamps indicates.

It would appear from this analysis that the appropriate place for voucher schemes are post-secondary student aid and labour market training, while vouchers for food aid and perhaps legal aid merit some consideration. Defenders of the existing welfare state will see it as the top of a slippery slope, which will inevitably weaken the public sphere and eventually infect the sacred citadels of health and K-12 education. Advocates of rationalist neoliberal reform will simply see the objections raised to vouchers in this paper as indicative of design challenges which can be met. Certainly, much more research will have to be done in each of these sub-fields as well as a deeper meditation on the relationship between ideas and interests in public policy, before making more than provisional pronouncements about the future of the welfare state in Canada. Nevertheless, I suggest that my analysis represents a vision for the improvement and perfection of the welfare state that is both pragmatic and pluralist, attentive to the need for efficiency without impoverishing the discussion of ends or vitiating the possibilities for richly political civic engagement, and which is hopefully deserving of the label “postliberal”.