Will Wilkinson’s new essay in the New York Times has sparked a great deal of discussion about the relationship between free markets and welfare state. Will’s view (and mine) comes down to a rejection of the usual right-left ideological axis, on one end of which lies anti-regulation and anti-welfare sentiments, and at the other, pro-welfare and pro-regulation sentiments. What if instead, as Wilkinson puts it, “the regulatory state and the size of government are intertwined, but they contain logically and practically separable strands”?

As Ed Dolan has deftly showed, there’s plenty of reason to think they do. A small government as a percentage of GDP is simply not correlated with economic freedom, nor with broader measures of the quality of governance. And at least from my liberaltarian point of view, what matters most for liberty is whether you are free to act and transact as you wish — “capitalist acts between consenting adults” — acts that are jeopardized by rules prohibiting the free movement of labor, or the diffusion of new technologies, or the exchange of benign substances, far more than they are by social insurance programs.

Indeed, there are even good reasons to think the welfare state and economic freedom are natural complements, both conceptually and in terms of stable political-economic outcomes. To see why requires unpacking exactly why the welfare state exists in the first place, and through that historical reconstruction, deriving, in some sense, what it is for.

“Three Normative Models of the Welfare State”

This is precisely what the philosopher Joseph Heath tries to do in his article “Three Normative Models of the Welfare State.” Heath’s writings on normative economics and business ethics are required reading for anyone interested in the connection between markets, ethics and the role of the state. This article is no exception.

In “Three Models,” Heath’s goal is to derive the normative underpinnings of the welfare state through its historical reconstruction. The idea is that normative theories are relatively moot if they are unable to describe, ex post, the emergence of the very sorts of institutions they purport to recommend. One by one, Heath moves through the leading normative models and rejects them as failing to make sense of actually existing social programs, from social security on down.

The leading model, that the welfare state exists to redistribute from the haves to have-nots, is wrong, according to Heath, despite being embraced by both its egalitarian supporters and its conservative critics. The core problem with the redistributive view is that it ascribes fundamentally different normative logics to the state and the market. The market is justified because it enhances productivity and efficiency, creating win-win outcomes for market participants. This is great, but inequalities inevitably arise which, for reasons of distributive justice, must be corrected after the fact by win-lose redistributions. In short, the market is for efficiency, the welfare state for fairness — two great tastes that, to liberal egalitarians, taste great together.

The Public Goods Model of Social Insurance

This view is problematic because it fails to explain much, if not most, of what the state actually does. Social Security retirement benefits, for example, represent a transfer from people who die young to those who live a long, long life — a structure that may even be mildly regressive in its first-order effects. Every year, thousands of millionaires collect unemployment insurance between gigs, having paid an outsize share of payroll taxes. And then there are infrastructure, policing, national defense, and all the other things the state provides out of general revenue, not due to fairness, but because they are public goods.

Heath argues that a reconstructed history the welfare state shows that it should be thought of in terms of public-goods provision, as well. For example, he notes that

the single most significant economic risk faced by most women in our society is divorce – especially in the years immediately following the birth of a child. Rising divorce and abandonment rates have been the single most important factor driving the so-called “feminization of poverty” in our society over the past three decades. Thus any economically rational woman should want to buy “divorce insurance.” Yet there are no private markets for such insurance, for fairly obvious reasons. Such insurance would be subject to extreme moral hazard and adverse selection problems. Thus one of the functions that the welfare system has increasingly taken on has been to provide a (very low) baseline level of divorce insurance.

If Heath is right, the normative logic of the social safety net and the market are not at all in tension. States provide public goods for the very same win-win efficiency reasons that they enforce property rights and a robust legal system (public goods in their own right). We all benefit from things like national defense, but high transaction costs prevent us from organizing to provide them privately. We also all stand to benefit from the ability to insure against economic risks like the loss of a job — millionaires have bills to pay, too — even when information asymmetries make private-sector unemployment insurance a challenge. Too much insurance has costs, as well, like inducing overly risky behavior. Yet such “moral hazards” are a generic feature of all insurance, and an inevitable part of making trade-offs.

As Ronald Coase pointed out years ago, given non-zero transaction costs there will always be areas where hierarchy is efficient, even if public choice constraints prevent democracies from achieving first-best outcomes. That is doubly true for many insurance markets, which are particularly prone to failure or even non-existence due to forces like adverse selection. The growth of the welfare states across the developed world strongly suggests that various forms of social insurance are another such area. The upshot is that the welfare state is better thought of as a kind of democratic mutual insurance scheme rather than, as it is often portrayed, a nationalized form of private charity or a tool to soak the rich. We fail to realize this because the ex ante win-win nature of social insurance (and, indeed, insurance more generally) is obscured by the ex post win-lose transfer to the recipient. The Swedes’ concept of “the people’s insurance,” which developed alongside the construction of their own formidable welfare state, shows that this notion is not even foreign to the places that strong egalitarians look up to.

The Republicans’ Missed Opportunity

Of course, nothing I’ve said denies the possibility of egalitarian concerns, or genuinely nefarious ulterior motives, seeping into the deeper efficiency origins of social insurance. But it does raise a problem for the quest to abolish the safety net as we know it. If the welfare state were merely a matter of redistributing from the “makers” to the “takers,” its abolition would come with no long-run cost — in a zero-sum world, the transitional losers are canceled out by the winners. But if, instead, the welfare state is for facilitating higher forms of win-win cooperation — what economists call Pareto improvements — then destroying it could leave society as a whole permanently worse off. This is not to say that every social insurance program is properly funded, or perfectly designed to make some people better off and no one else worse off. But, as a normative model, that is what we should strive for. Arguing about the empirical extent of the market failure in social insurance is not the point of this post. That empirical debate only comes after conceding the normative ground that efficiency matters in the first place.

In a more rational world — the sort of world Wilkinson calls for in his Times piece — the Republican position on entitlement spending would not be so existential. Rather, they would concede its normative consistency with the market, and move to debate its ideal scope and design, prioritizing certain conservative principles and pushing back on egalitarian excesses. On that point, I could see conservatives embracing the power of social insurance to reinforce the market’s dynamism along three key dimensions:

  • As a substitute for protectionist and freedom-reducing regulation;
  • As a tool for risk transfer, to encourage innovation and entrepreneurship;
  • And as a compensatory mechanism, bringing the losers of free trade and creative destruction into the “Pareto improvement” space.

Each of these pro-market features of social insurance is worth a blog post of its own. My only goal here is to illustrate the problem with the current anti-entitlement orientation of the Republican party. It means Republicans squander the opportunity to be more effective and creative reformers — to be on the side of the continuous normative logic of the market and the social-insurance state, while working to improve their efficient operation.

It also means that they play into the hands of the progressive left, which revels in the myth that social insurance is an egalitarian game of Robin Hood with conservatives taking the side of King John. The result is what I call America’s reluctant welfare state: A hot mess of programs that just barely satisfies the country’s bottom-up demand for social insurance, but which creates new inefficiencies and injustices through its self-sabotaging and incoherent design.