Introduction: Symposium on Robust Political Economy

ABSTRACT

Mark Pennington’s Robust Political Economy is a systematic exposition of a framework for analyzing institutional performance. The Robust Political Economy framework evaluates institutions according to their ability to solve knowledge and incentive problems. On grounds of robustness, Pennington combines insights from Austrian market-process theory and public-choice theory to defend classical liberalism from several compelling critiques. These include theories of market failure in economics; communitarian, deliberative-democratic, and liberal-egalitarian theories of justice; and concerns with social capital, domestic and international poverty, and ecology.

Keywords:

• epistemic challenge
• Hayek
• incentive problem
• knowledge problem
• Mark Pennington
• Robust Political Economy

Mark Pennington’s Robust Political Economy (Edward Elgar, 2011) is a wide-ranging application of comparative institutional analysis to contemporary public policy. It is, in addition, an attempt to revive classical liberalism. In the five years since its publication, it has inspired and provoked scholars across the disciplines of social science. Pennington has contributed, in particular, to debates about the epistemic capacities and limits of political institutions. This makes Critical Review an ideal forum to explore and criticize this influential work, so I am grateful to the editor for encouraging me to organize this symposium on the book for publication in the journal.

In this introduction to the symposium, I attempt to put Pennington’s work into intellectual context and to summarize his contribution. The papers by Victoria Bateman, Michael Bennett, Andrew Gamble, and Daniel Layman, and the reply by Pennington speak for themselves.

Circumstances of Time and Place

The fall of the Iron Curtain put an end, for the moment, to the debate over “communism.” In its stead emerged several more nuanced critiques of “capitalism.” Economists revived various theories of market failure, bolstered by new empirical discoveries in behavioral economics suggesting that markets produce predictably inefficient outcomes that can be readily improved through judicious regulation. In political philosophy, egalitarians and deliberative democrats found the underlying values, outcomes, and decision processes within market societies desperately wanting. In public policy, market institutions were criticized for their incapacity to relieve poverty and ecological problems, both domestic and international, and for undermining civic virtues. These criticisms pointed not toward the abolition of market institutions altogether but toward political efforts to complement, correct, and perfect them. To borrow a phrase from Albert Métin and revived by John Quiggin (Hazledine and Quiggin 2006), they exemplified a “socialism without doctrines.”

At the same time, the world witnessed the apparent success of social-democratic welfare states in many advanced societies—and then a prolonged recession that emerged out of catastrophic failings in financial markets. This left classical-liberal defenses of limited government and market institutions looking, at best, irrelevant to current policy debates, and at worst simply wrong (Alves and Meadowcroft 2013).

In response, Pennington’s book offers a systematic defense of classical-liberal institutions that engages concerns in political philosophy while drawing on insights from empirical social science. His case for classical liberalism is based not on abstract moral grounds but on the alleged impracticability of socialism without doctrines.

The Idea of “Robustness”

The notion of robustness as applied to political economy has its origins in a theoretical parallel that David M. Levy (2002) drew between the public-choice economist James M. Buchanan’s case for limited government and “robust” approaches to statistical analysis. Levy explained that robust statistical approaches sacrifice accurate point estimation in order to minimize measurement error. In parallel, he suggested that robust institutions would sacrifice some of the capacity of well-intentioned public officials to do good in order to minimize the possibility that they would do harm in cases where they are imperfectly motivated, as public-choice theory holds is usually the case. While robust statistics are designed to cope with measurement error, robust institutions are designed to cope with political actions that are not conducive to the public good.

Peter J. Boettke, Peter T. Leeson, and J. Robert Subrick then expanded this conception of robustness into the notion of Robust Political Economy (RPE), an approach to the comparative analysis of institutions with a specific emphasis on their ability to cope with not only incentive but also knowledge problems (Boettke and Leeson 2004; Leeson and Subrick 2006). This dual focus was partly historical, partly analytic in nature. The historical component can be traced back to the famous “socialist calculation debate” between neoclassical market-socialist economists and the Austrian economists Ludwig von Mises and F. A. Hayek (see Hayek 1935). The neoclassical economists argued that socialist planners could, in principle, make use of the price mechanism to allocate resources efficiently, so long as the prices were given and all actors were perfectly informed (Taylor 1929). This left motivational problems as the most important factors that might keep socialism from performing at least as efficiently as capitalist institutions. Thus, if a socialist regime could impose an adequate incentive structure or inspire economic actors to shrug off their self-interested ways, commercial society could be effectively replaced.

However, Mises (1920) and Hayek (1935 and 1940) bracketed the question of motives. Granting for the sake of argument that incentive problems could be solved, they contended that benevolent social planners would lack the knowledge needed to coordinate production for a whole society. Eventually, Hayek (1937 and 1945) refined this argument into the assertion that the knowledge essential for social cooperation is dispersed across a society and is inaccessible to any single individual or agency. Each individual, he maintained, has access to only a small portion of this knowledge—mostly knowledge connected to the specific circumstances of one’s time and place. Such knowledge is often tacit and therefore impossible to articulate through formal communication.Moreover, he contended, the natural and social worlds are dynamic and ever-changing, such that (1) economic actors face uncertainty even with respect to “given” knowledge and (2) new knowledge has to be discovered (Hayek [1968] 1978).

Following Hayek’s lead, several scholars have used the notion of robustness to support institutions of private property and voluntary exchange within a framework of the rule of law. They contend that these institutions are uniquely capable of dealing with the epistemic barriers to human cooperation, as they allow individuals to engage in socially desirable activity without explicitly coordinating with everyone else in the economy. Individuals can specialize in ways that make use of their specific knowledge and engage in mutually beneficial trade. When these institutions are combined with a price system, individuals can cooperate much more extensively, using publicly available estimates of the scarcity of particular resources, as embodied in prices, to engage in systematic rational calculation and planning. In this way, market institutions allow individuals to make use of knowledge that others possess but to which they do not have personal access.

This account relies on the assumption that prices are not “given” parameters, as the neoclassical economists had assumed in the socialist calculation debate, but are discovered through trial-and-error entrepreneurial learning, with feedback provided by the realization of profit and loss. In this view, entrepreneurs use their personal knowledge to discover otherwise scarce resources and put them to social use (Hayek 1945). However, there are also continuous sources of disequilibrium, including changes in relative scarcities of resources and new goods, technologies, and production processes introduced by other economic actors. Perfect competition and competitive equilibrium are merely explanatory ideal types that we should not expect to observe in practice. Competition is never perfect, but has a tendency to push prices in the direction of equilibrium (Kirzner [1963] 2013). This tendency overcomes the “knowledge problem” first explored in the socialist calculation debate.

When incentive questions are unbracketed, they provide the other half of the RPE framework. This is where public-choice research enters the picture. Public choice is the application of neoclassical economic analysis to non-market situations (Ostrom and Ostrom 1971, 205). Buchanan and Gordon Tullock ([1962] 1999) showed that, given the assumption that political and economic agents have symmetrical motivations, political processes will allow some people to impose costly externalities on others. For example, in a regime governed by simple majority voting and few constitutional limits on economic legislation, 51 per cent of the public can vote to impose a tax on the remaining 49 per cent, with the proceeds to be distributed to the majority alone. The same behavior that generates collective-action problems between self-interested agents in market situations can produce similarly perverse and harmful outcomes through the political process. In the worst-case scenario, these collective outcomes could cause even formally democratic regimes to fall into desperate conflict over the community’s resources.

Should the assumption of motivational symmetry be made? Some have argued that it should not, because in the public sphere, cultural norms militate against the self-interest that is encouraged or allowed in the private sphere (Friedman 1997, 441–42). Indeed, public-choice scholars have had to acknowledge that the empirical evidence does not confirm symmetrically self-interested behavior among political actors (Mueller 2011; Munger 2011). However, RPE does not depend on the prediction that self-interest will be just as pervasive in the public sphere as it is in a capitalist economy. It is enough that self-interest is a potential problem that we need to guard against. Few would deny that it is. Thus, the question is how best to guard against it.

RPE scholars have followed public-choice scholars in exploring constitutional mechanisms as a way to prevent self-interest from overrunning politics. When they are well designed, constitutional rules constrain individual political actors so that their aggregate impact is to produce socially beneficial outcomes, just as private actors are constrained in a capitalist economy. Thus, competitive elections that lead to regular exchanges of power between parties can help share collective resources more widely. Super-majoritarian provisions and other constitutional limits on majority rule can prevent popular coalitions from exploiting minorities (Buchanan 2001 and 2008; Meadowcroft 2014). The separation of powers, typically between executive, legislative, and judicial branches, can check arbitrary conduct. Most significantly, federal institutions, as well as other forms of polycentrism or overlapping governance, allow individuals to exercise an effective right of exit from one local regime to another (Buchanan and Congleton 2003; Weingast 1995). This allows for local variation and experimentation in policy making, and crucially forces local polities to compete for inhabitants, reducing their capacity to exploit residents.

This neat outline may obscure the fact that RPE is more than the sum of its parts: it can explore how incentive and knowledge problems interact with each other. For example, Pennington (2015) has suggested that hypothetical constitutional framers, just like real-world policymakers, face a significant knowledge problem of their own that challenges Buchanan’s constitutional political economy and its reliance on the rational design of restraints on self-interest.

Pennington’s Contribution to RPE

Pennington’s contribution is unique in the depth and range of its applications of the RPE framework. Robust Political Economy targets not only key economic debates about markets but also criticisms of classical liberalism from moral philosophers, sociologists, and ecologists. It applies the stress tests of robustness to these approaches to see if they can match the performance of classical liberal institutions under symmetrical conditions. The result is not only a defense of capitalism in the economic sphere, but also of polycentricism more broadly understood. Pennington describes the role of communal, civic, and other sub-national institutions in overcoming epistemic challenges to successful public policy.

In the introductory chapter, Pennington describes the three key features of his argument. First is the premise that institutions should be robust to realistic problems of human social life: limited knowledge and self-interest. Second is the method of comparative institutional analysis by which different policy regimes can be evaluated in light of these realistic conditions. Pennington’s comparative approach allows him to accept many of the premises of his interlocutors, for example, that market institutions produce manifest imperfections. In response, he does not try to defend the perfection of market outcomes but asks why we should expect alternative institutions to perform better. Third is a reiteration of the distinctively epistemic case for liberal market institutions. This is different from market institutions’ more widely recognized role of aligning private interests with socially beneficial outcomes. A common theme throughout Robust Political Economy is that many proposals appear superior to markets only when the epistemic dimension is ignored.

Market Failure versus Government Failure

Having laid out the principles of his argument, Pennington dedicates the second chapter to criticizing a popular economic case for substantial market regulation: the theory of market failure, both in old and new forms. Old and new theories of market failure share this common ground: the neoclassical economic practice of taking as an analytic baseline a static equilibrium characterized by perfect competition and complete, given information. The model of perfect competition then explains the allocative efficiency of markets. In a market where there are sufficient potential buyers and sellers, and where everyone has perfect information, competitive prices will reflect their marginal products.

These conditions are rarely, if ever, instantiated even for most private goods, where information asymmetries and other forms of ignorance are virtually ubiquitous. The neoclassical model is rendered even more irrelevant when faced with externalities, where the social costs of producing goods are not appropriately reflected in the price; and public goods, where the benefits of production cannot be excluded from free riders. In these cases, it is thought, only state allocation can produce efficient outcomes, because the state can account for the full social costs of production and compel everyone to contribute to the costs fairly. Old market-failure theories tend to take rationality and knowledge as given and construct scenarios, such as scale monopolies or public-goods provision, under which economic actors would fail to coordinate optimally. New market-failure theories observe that real-world actors deviate from economic rationality and thus fail to coordinate in cases where a perfectly rational and informed Homo economicus would do so.

Pennington notes an important link between market-failure theory and the neoclassical case for market socialism. Because available resources and their possible productive ends are already given data, in the neoclassical account, the calculative task of the market economy is relatively mundane, at least in the ideal scenario. It is possible, in principle, to produce efficient allocations under a variety of institutional frameworks. While entrepreneurs carry out the task of allocating resources under a private-property regime, this role could equally well be carried out by managers of state-owned or worker-owned firms. In addition, prices not only represent information but are also an incentive for self-interested actors to undertake the most efficient allocation. But there is a wide range of economic tasks that private-property markets simply cannot do, while there is nothing, in principle, that socialist alternatives could not do so long as incentives are taken into account in some way that simulates private markets. While the popularity of market socialism as a substantive alternative to private-property markets has waned, the theory that underlies it has continued to inform less systematic critiques of markets.

Pennington detects an irony here. Although both old and new market-failure arguments support state regulation, they are frequently based on opposed premises. Old market-failure theories assume that essentially, people are narrowly rational, self-interested utility maximizers. Therefore, they fail to reach Pareto-efficient outcomes when facing collective-action problems. New market-failure theories, by contrast, assume that people are predictably ignorant, with the result that many of their purely private economic decisions will be suboptimal. A more symmetrical approach would recognize that the apparent irrationality of economic agents found through empirical observation may represent the types of extra-economic motivations that would allow them to overcome collective action problems that purely self-interested actors would find intractable.

Critically, both new and old approaches assume that economic actors are helpless—unable to develop their own solutions to manifest social problems. By contrast, policy makers are modelled as exogenous observers of suboptimal decisions within markets, who will faithfully correct them if granted sufficient power to do so. Pennington’s argument is that we should not posit the ignorance and self-interest of economic actors as barriers to coordination without also acknowledging the parallel constraints that policy makers face when attempting to correct market outcomes. To do so is to compare a worst-case market scenario against a best-case political scenario. He maintains that there should be some independent reason to suggest that political actors will behave more beneficently, or with greater knowledge, in a particular domain of activity before commending a policy alternative to the market.

In order to evaluate institutional performance in cases of widespread ignorance, Pennington sets out an alternative way of conceptualizing an economy: as exchange activity between individuals, or catallaxy (Mises 1963, 233). On this account, knowledge of available resources and their possible uses does not exist prior to trade, and thus the conditions of perfect competition never obtain. The primary economic problem is that the knowledge necessary for beneficial cooperative production is dispersed across a whole community (Hayek 1945; Lavoie 1986). The puzzle is how individuals in some institutional circumstances, despite their manifest ignorance, nevertheless manage to coordinate and cooperate to produce socially desirable outcomes in the absence of the omniscient intelligence that would be necessary to produce a perfectly efficient allocation of resources.

Pennington contrasts his anti-perfectionist case with what Boettke (1997) identifies as the defense of capitalism that emerged out of the “Chicago School,” which assumes that any existing institutions and outcomes are necessarily efficient (because if they were not efficient, adjustments would already have been made to make them efficient). By contrast, the market-failure model assumes that any real-world situation not obviously characterized by a competitive equilibrium is a sign of inefficiency that government intervention can correct. Both models assume stasis: there is no spontaneous adjustment to problems of knowledge and coordination over time. Adjustment is either instantaneous or it is impossible.

To avoid these two extremes, Pennington rejects the assumptions of perfect competition, complete information, and perfectly rational actors. Institutions must work in an environment of bounded rationality, pervasive ignorance, and dispersed knowledge. Market institutions can never produce equilibrium outcomes, but through a process of trial and error and adjustment to new information, they may approach equilibrium in some circumstances. The advantage of this perspective is that it can explain the reality of economic coordination that is commonly observed in real-world economies without recourse to the Panglossian assumption that institutions are always already optimally coordinating. It is this anti-perfectionist description of markets against which alternative institutions should actually be compared.

Situated Selves, Social Capital, and Social Justice

In chapter three, Pennington turns to communitarian and deliberative-democratic critiques of market institutions. For this range of scholars, which includes Habermas, MacIntyre, and Iris Marion Young, the case for market institutions rests on an empirically unrealistic and morally unattractive conception of people as autonomous, atomistic individuals. In reality, these thinkers contend, people are radically situated within their social world. Attempts to impose rationalist liberal institutions premised on individual autonomy do not neutrally respect people as separate persons, but privilege selfish behavior and ultimately alienate people from their surrounding community. Furthermore, unconstrained economic liberalism allows the incursion of inappropriate values, attitudes, and power relations into communal spaces. Instead these should be sites of moral development and reflection, where people are respected as persons rather than mere means.

What is the non-market alternative? Scholars in the communitarian and deliberative traditions often affirm a dialogical, as opposed to an individualist, understanding of the self, as well as a range of democratic and collective political institutions that emphasize the importance of voice in political decision-making. Democratic and collective institutions allow people to justify their beliefs and actions in a public forum, rather than act on private interest alone. They let people widen their conception of the self, allowing them to coproduce and identify with the shared values of their community. At its most radical and utopian, this approach reflects a Marxian conception of freedom as individuals acting in a spontaneous, authentic social unity. In its less radical form, this approach advises that market decisions be constrained by overriding democratic institutions that are uniquely responsive to a community’s interests and values (Knight and Johnson 2007).

Paralleling his rejection of agents as rational utility maximizers in chapter two, Pennington accepts that atomistic individualism is indeed a flawed conception of human beings. A society premised on untrammelled selfishness would indeed be empirically and normatively bleak. The bounded rational agents described by Pennington share much in common with the radically situated agents that communitarians and democrats affirm. Like the democratic citizens that communitarians valorize, Pennington’s agents participate in social interaction not only to achieve given ends, but also to discover new ends that are worth pursuing with others. The question is not metaphysically or psychologically whether human beings are more or less social, but practically how to enable them to cooperate peacefully under realistic conditions.

Pennington points out that it is not self-interest that presents the primary barrier to establishing and pursuing a shared common good. Instead, it is the epistemic challenge that is inherent to cooperation in any complex social order. He argues that it is impossible for a deliberative institution to give voice to all the relevant considerations for establishing a common good. He draws on empirical research on voter ignorance to show that even minimal facts relevant to important policy issues are never understood by any but a minority of those participating in a political process. As a result, when asked to make decisions that have impacts beyond their personal experience and expertise, as they are required to do by centralized political institutions, people systematically err.

While ignorance is the more significant challenge facing public decisions, incentive problems compound it. In the absence of public access to essential knowledge, deliberative settings easily succumb to organized groups with articulate representatives who can best persuade others that their special interests are aligned with the public good. While the demand for democratic voice supports equality in the best-case scenario, it can easily diminish and silence relevant perspectives in reality.

Pennington contrasts these voice-based institutions with exit-based institutions, including those of the market. In theory, decisions based on exercising purchasing power through the market undermine equal participation. In practice, such decisions compare favorably because at least they allow everyone to participate in generating social outcomes through a shared process. By contrast, democratic institutions at any significant scale necessarily involve alienating all but a fraction of the population from actual political influence. Markets have the additional advantage of being able to communicate the tacit knowledge of participating agents. Democratic theorists tend to assume that markets supply personal incentives—motives, but not reasons—for action. Pennington argues that the price mechanism has the more fundamental role of supplying information that is critical for reasoning about public decisions. The estimates of the scarcity and availability of particular resources embedded in market prices represent the best available information about the social costs and benefits of particular policies. Leaving incentives aside, this information is critical for making the best use of society’s available resources. Moreover, Pennington points out that communities seldom agree what the best public policy might be on the basis of open debate and discussion alone. Instead, exit-based institutions allow people to experiment locally with different sets of policies and communal norms, generating new knowledge about how they work in practice in a way that is absent when decisions are taken collectively.

In chapter four, Pennington turns to a parallel critique of classical liberalism that has emerged out of theories of social capital. In this case, the notion that markets are artificial and have to be imposed is used not to condemn markets intrinsically but to argue that they are precarious and unstable. The basic argument is that peace, security, and a basic justice system alone are inadequate for commercial societies to operate. Markets only function in circumstances of widespread trust and social cohesion. Because social capital is a public good, individuals can reap the benefits of widespread trust without contributing to it, leading to its slow erosion. The state must constantly intervene in civil society in order to shore up the public environment that makes commerce possible.

In response, Pennington draws on historical evidence to show that markets have frequently emerged through endogenous evolutionary social processes rather than formal imposition by a modern state. This suggests that there are circumstances where markets can operate without their continued imposition by a powerful state. Developing previous work with John Meadowcroft (Pennington and Meadowcroft 2007), Pennington describes the role of the “bridging social capital,” as opposed to closer social ties, that market activity is capable of sustaining. Market activity in a civil society generates contact outside of one’s immediate family and social circle. People are confronted with situations where they are rewarded for recognizing and satisfying the interests of those from a diverse range of backgrounds (Meadowcroft 2005, 153). This interaction produces loose ties that permit people from markedly different backgrounds to trust each other sufficiently to engage in commercial relations despite social distance. In essence, a functional market process produces a continual stream of beneficial interactions between relative strangers. In turn, it draws in and rewards individuals who offer beneficial interactions with others, while excluding people whose practices harm or disadvantage would-be cooperators. In this way, trust between acquaintances and strangers is cultivated and social assurance problems are ameliorated (Al-Ubaydli et al. 2013; Berggren and Jordahl 2006). This creates an environment where people can more readily contribute to social insurance schemes and public good provision (Brosig-Koch et al. 2011; Ockenfels and Weimann 1999).

In chapter five, Pennington discusses the issue of social justice. He considers in particular the arguments of Rawls, Dworkin, and Young. From these liberal-egalitarian perspectives, markets may occasionally produce efficient outcomes. However, that is insufficient to establish those outcomes as just. In Rawls’s terminology, market processes reward individual assets that are arbitrary from a moral point of view, such as talents, class position, and inherited resources. For these scholars, the only just solution is to empower state institutions to allocate resources in a way that can be justified, at least in principle, to all members of a political community.

In order to deduce a fair way to distribute resources, scholars in this tradition have made use of hypothetical rational-choice mechanisms. In Rawls’s case, this is the famous original position, where rational, reasonable representatives of a community, deprived of knowledge of their personal circumstances and individual beliefs about the good behind a veil of ignorance, agree to the principles on which their political institutions should be based. Similarly, Dworkin uses a hypothetical desert-island auction that requires equally endowed members of a new isolated community to bid on all social resources. The result of these thought experiments is a set of principles that reserve a range of civil liberties that facilitate these rational agents’ autonomous pursuit of their personal conceptions of the good. This is combined with a stringent requirement for the state to control the distribution of all social resources in a way that respects the equality of all persons.

Underlying these arguments is a model of idealized agents who are rational, reasonable, and far-sighted. The assumption is that human beings should be modeled as rational choosers whose imperfections and biases can be abstracted away along with their personal histories and cultural peculiarities. The residual is universal agreement on what justice requires. As in chapter three, Pennington offers an alternative notion of radically situated, boundedly rational agents. As individuals with their own unique motivations, histories, and cultural backgrounds, they cannot be expected to agree to a single universal conception of justice. In this respect, Pennington notes an important overlap between his and Young’s critique of constructivist rationalism in theorizing about social justice. Young is concerned that a supposedly universal scheme of social justice will inadequately weight the claims of minority communities who have radically different ways of valuing aspects of the social world. It is for this reason that she supports deliberative democratic processes that give a specific voice to the historically oppressed and marginalized. Pennington agrees that centralized systems of distribution do indeed risk ignoring the legitimate claims of minority groups. However, he suggests that Young does not go far enough. One important step is giving sufficient weight to minority groups who might value certain sorts of resources and locations spiritually, or as constitutive of their identities. The question is, then, why not take the further step of giving weight to the ultimate minority, the individual with unique ways of valuing the world and its resources—which cannot necessarily be articulated or weighed in any centralized process, no matter how democratic?

In more practical terms, Pennington argues that theories of social justice are too epistemically ambitious. To allocate resources according to a systematic rule, we must have knowledge of all social resources (both actual and potential under different rules of cooperation) in order to judge which institutional framework will optimally serve the disadvantaged. In Rawls’s theory, this knowledge includes “basic facts of social theory,” including scientific knowledge of psychology and economics. Yet knowledge in the human sciences is contested because its subject matter is essentially complex. Meanwhile, more mundane facts on the ground, regarding available resources and the most efficient production techniques, are subject to change and uncertainty. Once again, the problem of incentives aggravates this epistemic challenge. Granting states a principled claim on all productive resources in society, Pennington argues, would discourage individuals from maintaining and developing useful assets.

What sort of institutions can cope with these challenges? Pennington proposes a scheme of minimal government that will allow people to discover better ways of engaging in fair social cooperation through an evolutionary process of “ethical learning.” This process is less reliant on policy makers possessing all the necessary knowledge of a whole political community, and allows for decisions to respond to facts on the ground and to changing circumstances. Such institutions are also able to cope with the fact that reasonable people inevitably disagree about the value of equality compared to other social goals. Rather than demanding that society conform to a single notion of social justice, freedom of exit and entry into different communities would, according to Pennington, allow people to experiment with different understandings and implementations of social justice, with the more successful gaining new members, the less attractive losing popularity.

Poverty, Economic Development, and Pollution

In part two, Pennington establishes the public-policy implications of his theoretical argument. First he applies his systematic comparative analysis to the issue of welfare provision.

Proponents of welfare-state institutions tend to assume that classical liberal institutions will ignore the plight of the poor and sustain rigid class distinctions. Welfare-state institutions are, therefore, assumed essential to extend the benefits of social cooperation to all. Pennington argues that this assumption is not borne out by historical experience. Nascent welfare states did not fill an institutional vacuum in industrializing societies but were imposed on a complex overlapping scheme of mutual-aid societies and private charities. Because the latter schemes were not imposed from above on poor communities, they were more aligned with their needs and priorities as they (and their benefactors) saw them. While poverty relief may have been relatively meager by today’s standards, so too were general standards of living. Pennington suspects that, had private welfare provision been allowed to continue to develop within a framework conducive to economic growth, it could have matched or improved on the outcomes of the modern welfare state. This is because the case for welfare states assumes that market outcomes are static, unable to respond endogenously to new social problems and challenges as they arise, while the democratic process is dynamic and responsive to the needs of the whole community. Both assumptions are questionable.

Pennington then considers the question of international development. In a parallel to claims that poor individuals face poverty traps that can be escaped only through state intervention, many theories of international development argue that the burdened positions of developing countries within the global economic order can be fixed only by the deliberate intervention and support of policy makers in the Western world. Pennington notes, however, that economic research has established the importance of institutions of the rule of law, freedom of contract, and secure property rights for sustained economic growth (Acemoglu and Johnson 2005; Shleifer 2009). Addressing this literature, Pennington adds a specific epistemic understanding of the case for property rights. In the neoclassical literature, these rights are primarily established to resolve incentive problems (Acemoglu, Johnson, and Robinson 2001). For Pennington, another key advantage is that they reduce uncertainty for entrepreneurs in the developing world and the likelihood of systemic error by policymakers who may be tempted to engage in big, expensive interventions that crowd out local, endogenous sources of development. An appropriate acknowledgement of this epistemic challenge of institutional reform would encourage policy makers in the West to take special care when advising or offering support to developing countries. The very incomplete information that remote officials necessarily have means that their interventions could easily cause as much harm as good.

Finally, Pennington turns to the question of environmental protection. The basic rationale for state involvement is that environmental challenges are classic public goods or common-pool resource problems that market institutions are ultimately incapable of addressing.

Pennington delineates two mainstream approaches to environmental policy. The first are “command and control” measures associated with direct state regulation and prohibition of harmful activities. The second, “price-based” measures, are more often associated with neoliberal attempts to use market mechanisms to ration the right to pollute or use up a resource, so as to ensure this is not done casually or wastefully. Drawing on his critique of market-failure theory, Pennington points out that both policy approaches miss the epistemic challenge inherent to environmental protection. They both assume that the policy makers designing the interventions already know what the correct outcome is, and are merely establishing an incentive and enforcement structure to achieve it. In fact, the preferred outcome is very often not “given” because no agent could possess knowledge of all the costs and benefits of particular production processes, uses of resources, and their impact on the environment.

As in later work, Pennington describes how property rights may help overcome this epistemic challenge (Pennington 2013). Rather than pricing rights to pollute or banning particular activities across a whole community, property rights grant individuals and groups ultimate control over specific areas and resources. The prospect of continuous possession means that property owners benefit directly from preserving and cultivating these resources but are also free to experiment in order to discover better ways of doing so. Property owners may choose to ban certain activities on the territory they control or set a price for use or access, but they will do so from a much more informed standpoint than a policy maker with only a remote interest in the particular resource. To illustrate the point, Pennington points to the better performance, compared to public regulation, of private property in conserving the quality of rivers and forests.

Pennington also addresses more radical proposals for protecting the environment, and in particular deliberative-democratic and participatory forms of governance. He suggests plausible cases where mechanisms of democratic participation will stymie innovation in ecological living. Traveler communities and deep ecologists who experiment in alternative, low carbon-footprint lifestyles can be understood to be entrepreneurs of a kind. They act independently to discover and disseminate attractive ways of living that are attuned to environmental values. Their attempts to challenge existing social norms are most plausibly blocked not by overwhelming consumerist practices, which are essentially voluntary, but by political processes, whose outcomes are compulsory. Pennington draws on his research on housing regulation to point out that land-use planning regulations typically have a conservative bent (Pennington 1997 and 2000). They are often used to restrict new kinds of developments aiming at ecological sustainability, rendering such lifestyle experimentation inaccessible to all but the marginal and the wealthy.

Pennington does not claim that any particular configuration of property rights will solve all problems, especially in this domain, where externalities are always present. Indeed, no theorist or policymaker could possibly possess sufficient knowledge to design institutions that ultimately guarantee environmental protection. Thus, Pennington introduces the notion of proprietary governance. In essence, having endorsed institutions that permit an evolutionary process of trial and error amongst participants, he suggests that such a process should also be commended for discovering the institutional frameworks themselves.

RPE as an Open-Ended and Progressive Research Program

Throughout his work, Pennington uses the concept of robustness to defend classical liberal institutions. However, it is critical to distinguish RPE as a framework of analysis from Pennington’s substantive argument that classical liberal institutions are more robust than others. It is quite possible within the RPE framework to argue for institutions that are not classical liberal insofar as that argument considers the performance of the institutions in worst-case scenarios. RPE is not pre-committed to a particular set of institutional answers to the problems of knowledge and self-interest. As Bateman suggests in this volume, epistemic challenges are not a trump card held by classical liberals. Instead, their consideration opens up another perspective on how we should consider institutional performance.

Conversely, a particular policy—classical liberal or otherwise—might be robust when holding a particular institutional or cultural environment constant, while being apt to fail elsewhere. These considerations make Pennington hesitant to propose any particular instantiation of classical liberal institutions in contexts where they have not emerged endogenously. Thus, he opposes the imposition of marketization by an external agency with limited knowledge of the specific circumstances of the actors expected to participate in them. This applies especially to international development. Instead, Pennington suggests that institutions that have endured over long periods are more likely to possess some characteristics of robustness, and that commendable meta-institutional arrangements are capable of evolving in response to stress, change, and uncertainty.

Somewhat paradoxically for a framework premised on examining worst-case scenarios, this approach allows scholars to escape the narrow confines of an institutional analysis based on the pessimistic assumption that the main barrier to social cooperation is the rivalrous, selfish nature of human beings. It shows that poor social outcomes are not necessarily the result of irreconcilable conflicts between competing interests that can be resolved only through sheer compulsion. They can instead be the result of ignorance, which can be ameliorated by better rules, norms, and institutions.