We live in an era of popular disenchantment with political institutions and hostility to government’s role in the economy. Neoliberal critics of political institutions argue that politicians have insulated themselves from popular control, acting on behalf of special interests, including interests in their own power, privilege, and financial gain. Guided by such private interests, government economic intervention generates inefficiency. From this analysis, they draw what is now a familiar political conclusion: that we need political reforms which limit government’s capacity for economic intervention.

The truth in neoliberalism is that improving economic performance requires political reform. But neoliberals wrongly conclude that improved performance — even measured by their own standard of efficiency — would be best achieved by reducing the economic role of government. To make the political economy work better, we need instead to ensure that government conduct is subject to vigilant oversight by citizens. We should, in short, reform our economy by improving our democracy, ensuring that citizens can effectively hold governments accountable for their economic activities.

That is a large thesis, and here I can only sketch a case for it. My purpose is not to provide operational answers, but to formulate an agenda for debate. And the central point on the agenda is, in a phrase: Put democracy first.1

 

State and Market

To decide which political institutions are best suited to improving economic performance, we need first to ask what role government should play in a modern market economy. Debate on this question has been raging at least since Adam Smith’s Wealth of Nations, and the arguments now seem to run around in circles, with every claim about the deficiencies of markets countered by a corresponding point about the horrors of regulation. As one reviews the successive rounds of controversy, they appear almost like a boxing match, with the state and the market alternately on the ropes. Here is a brief sketch of this history.

Round 1: When does the invisible hand of the market generate efficient allocations of resources? That was Adam Smith’s question, and modern neoclassical economics has worked out a precise answer to it. Markets are sure to generate efficiency when (but only when) several stringent conditions hold:

  • Complete markets: There must be markets for everything, now and for the future.
  • Complete information: Everyone must know everything (and therefore know the same as everyone else).
  • Complete independence: There can be no public goods, externalities, or increasing returns.

A few words of explanation: “Public goods” are goods whose consumption by one person does not prevent consumption by others. A lighthouse, for example, can be seen by one ship without preventing other ships from seeing it. “Externalities” are effects of one person’s actions on the welfare of others. They can be “negative” — for example, health damage due to passive smoking — or “positive” — for example, the fact that when I cooperate with a better educated co- worker, I work better. “Increasing returns” occur when the cost of producing an additional unit declines as the volume of production increases. Increasing returns give rise to “natural monopolies.” What unites these assumptions is that, in a world of perfectly competitive markets, no one can impose costs on anyone else: thus “complete independence.”

Under these special assumptions, Adam Smith’s invisible hand works. Market allocations are optimal — the separate actions of individuals aggregate to the common benefit. But state intervention of any kind transfers income from one person to another. And income transfers reduce incentives and misinform about opportunities. So in this special world, the state cannot improve on market efficiency, only reduce it. To be sure, people might decide that efficiency is not the only thing that matters — that they want an economy which is fair, and that they are prepared to sacrifice some efficiency to achieve fairness. But for now I propose to keep the focus entirely on efficiency.

Round 2: The case for market efficiency depends on these strong conditions; so when any of the conditions fail, as they inevitably do, markets no longer allocate efficiently. This was the basis of the social- democratic doctrine of state intervention. Enshrined in the 1959 Bad Godesberg Programme of the German Social Democratic Party — “markets whenever possible, the state when necessary” — the idea was twofold: let markets do what they do well (allocate private goods in those cases when private and social benefits are the same) but have the state provide public goods, correct externalities, and regulate monopolies due to increasing returns. This second round goes to the state.

Round 3: This social- democratic argument for intervention has an apparently fatal logical weakness: It starts from the premise that markets fail, and concludes that states should act. But just because markets fail to achieve efficiency does not mean that states will do any better. Of course, we can make the social- democratic argument sound by assuming an omniscient and benevolent state aiming to promote the good of citizens. But government conduct in the economy appears to be driven by the same self- interests that motivate all economic action; so its conduct can and does damage the economy. Here lies the fundamental dilemma of economic liberalism, identified by Richard Posner: “The economist recognizes that government can do some things better that the free market can do but he has no reason to believe that democratic processes will keep government from exceeding the limits of optimal intervention.” Indeed, analyses of the downfall of Keynesianism presented in the middle 1970s, whether from the left, the center, or the right, were almost identical. The state got its hands on everything, and thus offered an attractive target for the pursuit of private advantage. Because the state was permeated by special interests, private logic prevailed, and the internal cohesion of state interventions disintegrated. Round 3 ends with the state back on the ropes.

Round 4: But the theory of markets as efficient resource allocators is equally battered. Earlier I mentioned the importance of assumptions of complete markets and complete information in making the case for market efficiency. These assumptions are as implausible as the idea of an omnicompetent state, and — as the last 20 years of economics have shown — their failure has devastating implications. Markets are incomplete because we know that we will be making transactions in the future, and cannot perform them all now. Information is imperfect because we do not know everything about the others with whom we do business; more particularly, information is “endogenous” because we learn about others — their preferences, capacities, and beliefs — by observing their conduct in the market, and changing our prior beliefs in light of those observations. The inefficiencies originating from incomplete markets and endogenous information are profound. In a recent summary, economist Joseph Stiglitz put it bluntly: “Once information imperfections (and the fact that markets are incomplete) are brought into the analysis, as surely they must be, there is no presumption that markets are efficient. Absent that presumption, `the standard neoclassical model,’ the formal articulation of Adam Smith’s invisible hand, the contention that market economies will ensure economic efficiency provides little guidance for the choice of economic systems.” When some markets are missing, as they inevitably are, and when information is endogenous, then markets do not work their special magic: Supply may not equal demand (even in equilibrium); prices do not uniquely summarize opportunity costs and can even misinform; externalities result from most individual actions; information is often asymmetric; market power is ubiquitous; and “rents” abound. These are no longer “imperfections,” correctable departures from a perfect market. There is no unique, ideal market out there to be blemished, but lots of possible institutional arrangements, each with different consequences for social welfare.

Moreover, when markets are incomplete and information imperfect, some forms of state economic action are inevitable. For example, a modern economy can only function if the state insures investors (limited liability), firms (bankruptcy), and depositors (two- tier banking system). But once the state provides insurance, it must pay claims, even if those claims are the product of negligence resulting from the insurance itself. The result is that ideal resource allocations are unattainable.

Of course, even the most ardent neoliberals think that governments should provide law and order, safeguard property rights, enforce contracts, and defend the state from external threats. And once we acknowledge the importance of these minimal functions — once we agree that they provide benefits to individuals or inputs to private production — then the entire framework of analysis is transformed: Governments can do something to improve market allocations. But the economics of incomplete markets and imperfect information opens room for a much greater role for the state. Textbook complacency about markets is untenable, since markets simply do not allocate efficiently. Even if governments have only the same information as the private economy, some interventions by governments would unambiguously increase social welfare.

Thus the state has a positive role to play. Still, Round 4 ends at best in a draw. Acting as it could and should, the state is able to improve on the market. But the consequences of the neoliberal counter- punch linger. How can we get the state to act when it should, and refrain when it should not?

 

Back to Political Economy

Because of incomplete markets and imperfect information, there is no such thing as “the” market, only differently organized economic systems. Indeed, the very language of “the market” subject to interventions by “the state” is misleading, and the political project of “freeing the market” makes no sense. The problem we face is not to liberate “the market” from “the state,” or regulate “the market” through “the state,” but to design specific institutional mechanisms that induce individual agents, operating under conditions of incomplete markets and imperfect information, to behave in a collectively beneficial manner. This organizational problem is the fundamental concern — intellectual and practical — of contemporary political economy.

To clarify the general form of the problem, let’s start with a simple case. Suppose your car has been making funny noises. You go to a mechanic, explain the problem, leave the car, and wait for the result. One day later, the car is ready, the mechanic tells you that shock absorbers needed changing and that it took five hours to fix. You pay, drive out of the garage, and the noise is gone. You can reward the mechanic by going back to him if you are satisfied, or punish him by going somewhere else if you are not. Should you be satisfied? The noise is gone, but that doesn’t answer the question. The mechanic knows much more than you: whether he wanted to do the best job possible or as little as he could get away with; whether the car required a major repair or a minor adjustment; whether he did the work in an hour or needed five. To be sure, you are the “principal,” and he is the “agent.” You hired him to act in your best interest. What affects his behavior, however, is not the formal contractual relation, but that he has interests of his own, and therefore may well not act in your best interest. Of course, you can provide incentives — rewards and punishments — for him to act for your interests. But when you decide to reward or punish, you have imperfect information about his preferences, capacities, and activities. And that sets your problem: Faced with such limits, what can you do to induce him to work for you as well as he can?

The problem generalizes from auto mechanics to economic and political organizations. When some markets are missing and people have access to different information, relations between classes of actors are those of principals and agents. We can think of the economy itself as a network of diverse relations between different kinds of principals and agents; between managers and employees, owners and managers, investors and entrepreneurs, but also between citizens and politicians, politicians and bureaucrats. The agents know their own motivation, have privileged understanding of their capacities, and may have a chance to observe some things principals cannot. Principals, acting on limited information, try to get agents — assumed to be acting in their own interests — to act in the principal’s interest as well. And it matters to economic performance — efficiency and growth — whether they succeed: whether employees have incentives to maximize effort (or shirk); whether managers have incentives to maximize profits (or their own income); whether entrepreneurs have incentives to take only good risks (or to throw good money after bad); whether politicians have incentives to promote public welfare (or their own income); and whether bureaucrats have incentives to implement goals set by politicians (or enhance the power of their bureaus).

Whether they have these desirable incentives depends on the institutions that organize all these relations. Economic institutions, such as firms and equity markets, organize relations between employers and employees, owners and managers, or investors and entrepreneurs; political institutions, including parties and elections, organize relations between citizens and governments or politicians and bureaucrats; and political-economic institutions, including regulatory agencies, organize relations between governments and private economic agents. If the economy is to operate efficiently, all these relations must be structured appropriately.

But the role of the state is unique. The state sets the incentive structures confronting private agents in their relations, by mandating or prohibiting some actions by law, changing relative prices via the fiscal system, facilitating transactions, and coercing participation. Suppose, for example, that I buy car theft insurance. I drive to my destination and have a choice to park a few blocks away from where I am going, in a place where the car is unlikely to be stolen, or park just in front, in a place where the car is more likely to be stolen. Because I am insured, I take the risk and park in the more dangerous place. Now the state comes in. It taxes me and uses the tax revenue to place a policeman in the dangerous place. As a result, car theft is less likely, the insurance company loses less money, and my premium goes down, more than compensating for the increased tax. The state is deeply implicated in my “market” relation with the insurer. Although our relation is strictly “private,” it is shaped by the state.

The story about the insurance case generalizes. The state permeates the entire economy; it is a constitutive factor of private relations. Problems about the design of economic institutions cannot be avoided by throwing the state out of the economy, and thus they must be confronted more directly.

 

Effective State Action

My central thesis, then, is that the efficiency of an economic system depends on the design of relations between the state and private economic agents as well as between citizens and the state. Private agents must benefit by behaving in the public interest and must suffer when they do not, and so must governments.

The conditions for effective state intervention are that governments must be able to control economic actions of private actors, and citizens must be able to control governments. These conditions must be fulfilled simultaneously. A government with the power to intervene in the economy is likely to act in its own interest unless it is accountable to citizens. In turn, if citizens fully control a government that can do nothing to affect their welfare, democracy is impotent.

To understand how political accountability affects the quality of state economic intervention, consider the following situation (the description will be very abstract and oversimplified, but it brings out the essential elements of the economic effects of accountability).2 Suppose there is a firm that is a natural monopoly, and which can have either high costs or low costs of production. The government observes these costs, but the public does not. A firm with high costs can invest to lower the costs. This investment is socially beneficial.

In a good intervention, the government subsidizes investment if the firm has high costs but does not pay for investment otherwise. Bad interventions fall into two types: Either the government fails to subsidize investment by a firm with high costs or the government subsidizes a firm with low costs and splits the benefits with the firm. Hence different “interventions” — actions of government directed at private agents — result in different behaviors of the firm, with consequences for public welfare.

But governments in democracies need to be elected, and different public reactions to government behavior affect the quality of intervention. Assume that governments want to be reelected. To induce governments to intervene well, citizens must make it known that — despite the limits on their information — governments will be reelected if they intervene well and will be thrown out if they intervene badly.

To improve on markets, then, governments must be able to tell when their interventions would increase social rates of return, and must have instruments of effective intervention. And citizens must be able to discern good from bad governments and be able to sanction them appropriately, so that those incumbents who act well win reelection and those who do not lose. Elected politicians must want and be able to control bureaucracies, which are not subject to direct popular sanctions. These conditions are stringent and can never be fully met. But the key point is that the details of institutional design matter to their satisfaction. As studies of government accountability indicate, the ability of citizens to control politicians depends on specific institutional arrangements as well as on the structure of the party system. As analyses of administrative oversight show, minute institutional arrangements affect the ability of legislators to control bureaucrats. In all these relations, properly designed institutions can improve on market allocation.

In short, neoliberals are right that we should not assume that government is an agent of the general welfare, and trust it to act as such. But that insight does not settle the issue of state and market. The question is whether mechanisms of democracy — institutions that keep government accountable to citizens — can provide incentives for governments to act as they should.

 

Can Citizens Control Governments?

The crucial question, therefore, is whether and under what institutional conditions citizens can control governments. This broad question comprises two others:

  • Are citizens able to retain governments that act in their best interest and throw out governments that do not?

     

  • Is the threat of being thrown out sufficient to induce governments to behave well?

Let’s say that governments are accountable if citizens can tell whether governments are acting for the good of citizens and sanction them appropriately: Incumbents who act in the best interest of citizens win reelection, and those who do not lose them. Though important, accountability is insufficient for making governments act in the best interest of citizens. Even if citizens are able to hold governments accountable for their conduct — by punishing them if they fail to do what citizens want — governments may fail to do everything possible to further citizens’ interests. Incumbents may accept the possibility of defeat and dedicate themselves to pursuing their own advantage rather than the electorate’s. So let’s say that a government is “responsive” if it acts in the best interest of citizens — specifically, if it chooses policies that an assembly of citizens, acting with the information available to the government, would have chosen by a majority vote under the same constitutional constraints. Hence, governments may be accountable but not responsive.3 And, in the end, responsiveness is what matters.

Responsiveness is problematic because politicians have interests of their own as well as some private information not observed by citizens. Even if politicians are interested only in reelection, they can collude with special interests and use the resources raised in this way to make themselves visible to voters or to persuade voters that they are acting responsively. The result will be less than optimal for voters. And politicians may have other goals than simply being reelected; they can use office to pursue some private benefits — so- called “rents” — which are costly to citizens. These rents may involve simple theft, a cushy job once out of office, private perks while in office, or shirking on the job. Such politicians still care about reelection, but at least in part because they hope to continue to extract benefits.

The problem for citizens is thus to induce governments to act for the general welfare, rather than to pursue self- interest, perhaps by colluding with special interests. And there are two discernable ways in which citizens might achieve this: prospective and retrospective devices.

The idea of the prospective mechanism is that elections define a mandate for governments to pursue. Parties or candidates propose policies during elections and explain how these policies will affect citizens’ welfare; citizens decide, by majority rule, which of these proposals they want implemented. Thus, the winning platform becomes the “mandate.” But while the notion of responsiveness as the implementation of mandates is often endorsed by theorists of democracy, Bernard Manin has pointed out that no actual democratic institutions work this way. In no existing democracy do citizens instruct representatives how to vote on particular issues. No national level democratic constitution allows for recall. Binding referenda based on citizens’ initiative are rare at the national level. And, except for the US House of Representatives, electoral terms tend to be long, at least four years. Once citizens elect representatives, they cannot force them to adhere to instructions. Voters can punish politicians who betray mandates only at the time of the next election, after the effects of such a betrayal have been experienced. And because deviations from mandates produce outcomes which affect these retrospective judgments (in addition to the mere passage of time), citizens cannot enforce the adherence to mandates per se.

Why are there no imperative mandates? Historically, the main argument was that legislatures were thought of as deliberative bodies. People want representatives to discuss issues and learn one from another. Moreover, when people are uncertain about their judgments, they may want representatives to consult experts.

Another historical argument was self- restraint. People may be afraid of their own passions; moreover, they know that they do not know whether their views about the impact of policies are correct. Hence, they have reasons to give government latitude to govern.

Finally, institutions must allow for changing conditions. No electoral platform can specify ex ante what the government should do in every possible situation. Governments must have some flexibility in coping with changing circumstances. If, when adopting a constitution, citizens expect that conditions will continue to change and governments are likely to be responsive, they will not want to bind governments with specific instructions.

These are reasons why democratic institutions contain no mechanisms enforcing prospective representation. We choose policies that represent our interests, or candidates who represent us as persons, but democratic institutions do not ensure that our choices will be respected.

The question thus remains whether we can enforce responsiveness through retrospective mechanisms. Elections force governments to account for their past actions. Anticipating the retrospective judgments voters will make during the next election, governments choose policies which in their best judgment lead to outcomes that citizens will evaluate positively. Accountability works, then, through government anticipation of citizens’ retrospective judgments and through retrospective voting by citizens. Citizens set some standard of performance to evaluate governments. They decide to vote for the incumbent if, but only if, their income increased by at least four percent during the term, for example, or if the inflation rate is less than six percent. The government, wanting to be reelected and knowing the citizens’ decision rule, does what it can to satisfy these criteria.

The problem facing citizens, then, is to set up a trade- off for politicians. The aim is to present politicians with the following choice: extract rents and lose office, or do not extract rents and stay in office. The difficulty, however, is that citizens may not observe what officials are doing, or know what governments know. Moreover, citizens have limited knowledge of the causal relations between policies and outcomes. Citizens care only about results, but they want to know if they are as well off as possible under conditions which they do not observe. Hence, they must make inferences about the types of governments they face from what they do observe. And such inferences are easier under some conditions than others.

To illustrate the point, suppose citizens are not certain at the time of the election whether government will be responsive. Suppose further that nature throws dice to determine whether conditions are “good” or “bad.” Governments observe these conditions but citizens do not. For example, governments know if the state coffers are full or empty, whether the negotiating posture of international financial institutions is accommodating or belligerent, while citizens observe neither. Governments decide whether to follow a policy which is better for citizens when conditions are good, or a policy which is better when conditions are bad. Suppose further that rents for the government are higher when they pursue B rather than A, no matter what the conditions are, so that governments always have an incentive to choose B. Finally, assume that reelection gives a bonus, V, to the incumbent. (I will use numerical examples, in which V = 3.)

Now, suppose that the structure of pay- offs is as follows (the first number in each pair is government rents, but remember that citizens observe only their welfare, which is the second number):

Under these conditions, citizens are ex post certain that the government was responsive if it generated an outcome of at least 3. And they can enforce responsiveness by a simple retrospective rule: Vote “Yes” (retain incumbents) if welfare is at least 3, vote “No” otherwise. When conditions are bad, governments will want to do what is better for citizens anyway. When conditions are good, governments know that if they generate more than 3 for citizens (in fact 5), they will be reelected and will get a bonus of 3; while if they concentrate on extracting rents, they will get 2 and not be reelected. Since outcomes for citizens are uniquely mapped on government actions, retrospective voting will enforce responsiveness.

Yet even if citizens have good theories, if they know the effects of policies as well as the government does, they may still be unable to judge if the government was responsive. Suppose that the observational structure is the following:

Now citizens do not know if they got 3 because conditions were good but the government unresponsive, or because conditions were bad and the government responsive. While the mechanism of accountability is necessarily retrospective, responsiveness cannot always be enforced by retrospective voting based on the observed outcomes alone. To induce responsiveness, citizens must know not only how well off they were, but also whether they were made as well off as possible. And looking at the outcomes alone may be insufficient to arrive at this judgment.

Under such conditions citizens must base their voting decisions on something in addition to the past record of the incumbent: messages of the incumbents, messages by challengers, information about campaign contributions, promises of the incumbents or the challengers, perhaps even identities or symbols. They must use whatever information they can get.

When retrospective judgments are not sufficient, citizens will end up punishing some governments that in fact are acting in good faith. They may also reward some governments that are acting in bad faith. Hence, complete accountability is not possible. But if voters adopt an appropriate voting rule, they can reduce the risk of rent extraction.

Yet even if voters adopt an appropriate voting rule, governments are accountable only when several institutional conditions are present:

1. Voters must be able to assign responsibility for government performance. Their ability to do so is limited when the government is a coalition. It is also limited when the presidency and the congress are controlled by different parties. It would take an elaborate theory of government to figure out who is responsible for what under such conditions.

2. Voters must be able to vote parties responsible for bad performance out of office. This may appear to be a universal feature of democracy, but under some electoral systems it becomes next to impossible — witness the continued tenure of the Italian Christian Democrats or Japan’s Liberal Democrats.

3. There must be an opposition which monitors the performance of the government and informs the citizens. Indeed, any reasonable understanding of control over politicians must focus on the role of the opposition. Citizens have two agents, not one: the incumbent government which chooses policies and the opposition which wants to become the government. The opposition is an agent of citizens since it wants to win office, and in order to win office it must also anticipate the retrospective judgments which voters will make about the incumbents at election time. Anticipating these judgments, the opposition has incentives to monitor the government and inform citizens (truly or not) about bad performance on the part of the incumbents. It can win elections if it persuades voters that the incumbent government was not responsive. Even if, to begin with, citizens care only about outcomes rather than about the policies that generated them, the opposition can induce voters to care about policies if it succeeds in persuading them that different policies would have led to better outcomes. And if opposition parties inform citizens about the misdeeds of the government or just about the sources of their money, they lower the cost of information to voters.

Yet the existence of an opposition which wants and is able to monitor government performance should not be taken for granted. If all parties are hostage to the same interests, they will collude in hiding it. Incumbents may prefer to collude against all challengers rather than compete with one another. The opposition can be so divided that it spends most efforts on internal fights rather than focusing attention on the incumbents. The opposition may see no chance of winning and do something else rather than monitor the government. And it may or not may not have resources to do so.

4. Politicians must have incentives to want to be reelected. This condition becomes problematic when there are limitations on re-eligibility and, particularly, when political parties are not ongoing, bureaucratic organizations that offer careers to their militants.

5. Mechanisms of accountability must be not only “vertical” — of elected politicians to voters — but also “horizontal” — of different branches of the government to each other. The legislature must play an active role in deliberating and formulating policies. In turn, the executive must be able to control the bureaucracy.

6. There must be institutions, independent of other organs of the government, that provide citizens with information needed to improve their posterior evaluation of governmental actions, not only outcomes. Such institutions may include a body that would insure the transparency of campaign contributions, such as an independent auditing branch of the state, an Auditor General, a statistical office independent of government control (in the vein of the Chilean contraloria), and an office of the ombudsman that would serve as a vehicle of popular control over the bureaucracy.

7. Voters must have some institu- tional instruments to reward and punish governments for outcomes they generate in different realms. Yet elections are inherently a blunt instrument of con- trol. Voters have only one decision to make with regard to the entire package of government policies.

The sum of these conditions appears formidable: Democracy is hard. But there are conditions in which governments can be controlled by citizens. What’s needed, in a nutshell, is a clear party system with stable parties, a vigorous opposition, an effective system of checks and balances, a decent level of information that focuses on general economic performance, and non-electoral mechanisms for control over specific policy realms or particular organs of the government.

 

Other Projects

The question of whether or not a neoliberal state is superior to an interventionist one cannot be resolved in general: The welfare consequences of state intervention depend on the specific institutional design as well as on the inherited economic conditions. But the neoliberal state is at best a benchmark against which to measure the quality of state intervention. Given that market allocations are not efficient, disabling the state is not a reasonable goal.

To conclude, I want to clarify the relation between efficiency and other criteria as goals of state intervention in the economy, and the relation between this vision of political reform and some alternative projects.

I have couched my discussion in terms of efficiency because this is the only common yardstick to compare the neoliberal prescription with alternatives. But reasons other than efficiency may lead governments to intervene in the economy.

Capitalist economies use two mechanisms to allocate resources and distribute incomes: markets and the state. Markets are mechanisms in which scarce resources are allocated by their owners. Individuals cast votes for allocations with the resources they own, and these resources are always distributed unequally. The state is also a system which allocates resources, including those it does not own, with rights distributed differently from markets.

The allocation of resources that would result from an unfettered operation of markets is likely to differ from that resulting from a democratic process. This is not only because markets may be inefficient, but also because the people, as the ultimate authority in a democracy, may decide as citizens to pursue different goals than they would choose as market agents. For instance, considerations of distributive justice might cause preferences expressed through the political process to differ from those actualized via markets. The distribution of income generated by markets depends on initial endowments of inherited resources and natural abilities that people bring to the market, and this distribution may be collectively deemed unjust. Economic security may be another reason for this divergence. Markets do not and cannot insure against all risks since some forms of insurance involve too much moral hazard; that is, they encourage people to take excessive risks. There is no private market for employment insurance because the moral hazard entailed in such insurance requires coercive monitoring. Yet people may want to have more security than markets can provide.

While justice and economic security are obvious candidates for collective goals that would not be satisfied by markets, other such goals — for example, beauty — are non- economic. The people may want to preserve opera as an art form even if economic agents are not willing to pay what it takes. Opera may be an ecological good, a cultural stock people do not want to deplete out of respect for the potential tastes of future generations. Justice, security, or beauty are goals the people may decide to pursue even if markets were efficient in the standard sense, and even if their pursuit were economically costly.

A word of clarification on these other goals: I am not claiming that all the verdicts of popular will, as expressed through democratic process, should be implemented. Government intervention must be limited by individual rights, including economic rights. Protection of some aspects of property rights is normatively justifiable and economically necessary. Constitutional principles are necessary to limit the realm of state intervention. The state must be not only effective and accountable but also limited. But “limited” does not mean that it has no economic role.

Finally, a few words on the location of this version of political reform among other political projects. One alternative is the neo-liberal project: to disable state intervention altogether, restricting its economic functions to a minimum and constraining economic policy by rules or delegating it to organs that are not responsive to the political process. In this view the state must be institutionally constrained so that it is unable to yield to political temptations. I rejected this prescription with the argument that an economy cannot function efficiently under market discipline: We don’t get the best allocation of resources because some forms of insurance are required if economic agents are to be willing to take necessary risks. And even if insurance inevitably creates bad incentives, the purpose of political reform, I have argued, is to find institutional mechanisms that would allow the state to distinguish and appropriately sanction private actions that are or are not collectively desirable, as well as to find institutions that allow citizens to punish governments that abuse their power to intervene. If such institutions are well-designed, then even if they yield allocations which are only second best, these can be superior to the unfettered rule of markets.

Another alternative project is to enhance participation. Democracy is a system which grants citizenship rights to most individuals, but it does not automatically generate the social and economic conditions necessary for an effective exercise of these rights. Hence, to the extent that social and economic inequalities limit access to the political system, even well-designed accountability mechanisms may end up merely perpetuating class relations. Widespread participation, and also a better distribution of assets to facilitate the exercise of citizenship, is thus necessary for accountability mechanisms to work. Even so, participation is not enough. Unless participants can effectively monitor the performance of the bodies they supervise, and unless they have instruments with which to reward and punish government, participation will remain merely symbolic. The question of whether governments can be controlled is prior to the question of who exercises this control.

Hence, in my view, the quality of state intervention in the economy depends on the quality of democratic institutions. Political reform should be guided by the goal of designing institutions that permit the society to pursue, under constitutional limitations, its collective goals, by enabling the government to intervene in the economy and by subjecting governments to popular control.

Originally published in the April/ May 1996 issue of Boston Review