It is a commonplace that “neoliberal” thought and practice has an anti-redistributive bent. But dig a little deeper, and neoliberalism—or, at least, the mainstream economic thinking that has underwritten much policy in the neoliberal era—is shockingly pro-redistributive. The issue, instead, is how and when redistribution takes place. In this thinking, there is a clear preference for redistributing income exclusively through the tax system, while avoiding supposedly less efficient tools for reducing inequality, such as the minimum wage, collective bargaining, or housing regulation. In other words, there is a preference for redistribution rather than what has been termed predistribution.
This preference for tax redistribution has had palpable political and economic consequences. Recent research shows that, since the 1970s, working-class voters moved away from the Democratic Party as it abandoned its traditional support for predistribution, through pro-union policies and public employment, and embraced redistributive policies instead. Other research concludes that greater predistribution, not redistribution, explains why Europe is less unequal than the United States.
One of the most refined arguments for using redistribution rather than predistribution to address income inequality is the “double distortion” argument, which says that predistribution is always more economically costly than redistribution. This argument, however, fails because it assumes that labor supply responses to either approach will be identical or that the costs of predistribution are “additive.” Neither assumption need always be true. I explore these issues at length in my new book, Ending Income Inequality: A Critical Approach to the Law and Economics of Redistribution, where I contend that the scope for predistribution is much larger than neoliberal thought and modern economic theory allow. Looking beyond the book, predistribution should be only the starting point for rethinking capitalism’s institutions.
Predistribution versus Redistribution
To see why modern economic theory is congenial to redistribution, consider the views of its progenitors. Jeremy Bentham, following his utilitarian principle of maximizing the greatest good for the greatest number, suggested that taxing the wealthy and transferring to the poor could increase aggregate happiness. Later, the intuition underlying this conclusion was formalized as “diminishing marginal utility.” Happiness rises with each increment of wealth, but the gain diminishes as wealth grows. A person earning $15,000 values an additional $1,000 far more than a billionaire, so a transfer of $1,000 from the ultra-wealthy to the poor raises total well-being.
But why stop at merely taxing billionaires? Extending this idea in the 1890s, the English economist F. Y. Edgeworth explained, “the richer should be taxed for the benefit of the poorer up to the point at which complete equality of fortunes is attained.” In other words, the distribution of income that would maximize happiness in society is the one where everyone has exactly equal incomes.
The idea of engineering social happiness through income transfers was later expressed in modern economic theory’s two “fundamental theorems of welfare economics,” the two doctrines that provide the foundation for economists’ policy prescriptions. The first theorem says that a perfectly competitive market economy maximizes well-being in society, in the “Pareto efficient” sense—no one can be made better off without making someone else worse off. Yet there are many such Pareto efficient outcomes, depending on how society’s productive assets are initially distributed. The second theorem says that with redistributive “lump sum” transfers—taxes that do not influence behavior—any desired Pareto efficient outcome can be reached.
These two theorems of welfare economics fundamentally shape how the economics profession understands the world. They encourage a kind of “division of labor” in economic policy. Some economists even talk about a government department of “efficiency,” which promotes competitive markets and a big economic pie, and a government department of “distribution,” sealed off from the first, which divides the economic pie with entirely different tools. The perspective also encourages a skepticism about direct intervention in market activities, especially those with distributive objectives, as pithily expressed by political scientist Jacob Hacker: “just let the market rip and clean up afterward.”
Yet despite being pro-redistribution in principle, mainstream economics in practice has often been associated with hostility to it. One reason is selective use. Free trade, for instance, was promoted as a universal wealth generator, but when globalization produced winners and losers, powerful interests ignored the redistributive half of the theory. The two theorems’ division of labor makes such cherry picking easy.
Another reason is the costliness of redistribution. Edgeworth himself immediately cautioned that perfect income equality would undermine the motivation to work. An equalizing of the earnings distribution has what economists call a “substitution” effect: additional time working is made less remunerative, so people engage less in paid work and increase their leisure time. So, while redistributive taxes may increase society’s total well-being, they might also reduce economic output. How one balances that trade-off leaves a lot of room for extraneous ideological importations. Indeed, when James Mirrlees first formalized the utility benefits of redistribution and the economic costs of taxation in his 1971 article, “An Exploration in the Theory of Optimum Income Taxation,” he was notoriously inconclusive about exactly how much redistribution through taxation should take place. Conservatives, nevertheless, seized on the paper’s inclination for a flat tax rate: because progressivity increases the marginal cost of taxation, it is thought to exacerbate the substitution effect.
The Double-Distortion Argument
Despite its predisposition for redistribution over predistribution, the two fundamental theorems of welfare economics never unequivocally mandated the exclusive use of taxes and transfers to pursue distributive goals. It was law and economics scholars Louis Kaplow and Steven Shavell who put forward an explicit reason to confine redistribution to taxes. Their 1994 article, “Why the Legal System Is Less Efficient than the Income Tax in Redistributing Income,” advanced the “double distortion” argument (a phrase coined by their critics).
Kaplow and Shavell’s argument is simple but abstract. It begins with the idea that taxation has economically harmful labor supply effects, as mentioned above. As they point out, any policy that redistributes income, whether through taxes or other “legal rules” (such as a minimum wage or collective bargaining), will have similar substitution effects. In addition, however, Kaplow and Shavell claim that redistribution using legal rules will always have added negative economic consequences. In a nutshell, while redistributive policies of any kind affect labor supply, legal rules also affect—“distort” is Kaplow and Shavell’s word of choice—the behaviors regulated by the legal rules themselves, whatever those may be.
Put simply, redistribution through taxation has one economic cost, while distribution through legal rules has two. And since one cost is better than two, redistribution is always cheaper—less inefficient—than predistribution. Moreover, because the tax code can always be crafted to achieve the same amount of redistribution as any predistributive legal rule, taxation can always redistribute just as effectively as legal rules, but without any additional economic costs. Indeed, owing to the relative cheapness of tax redistribution, replacing a redistributive legal rule with an efficient form of taxation can result in a “win-win.” The replacement’s economic savings can make the poor better off than the redistributive legal rule can. It’s a hard argument to ignore, if one cares about the welfare of the economically disadvantaged.
Kaplow and Shavell illustrate their argument with an example from tort law. Consider two different legal rules, strict liability and a wealth-dependent damages rule. Think of someone who injures another person in a car accident. Under strict liability, injurers pay for exactly the amount of harm the accident causes, to person or possessions, neither more nor less. By contrast, under a wealth-dependent rule, damages are greater than harm caused when a rich person injures a poor person and lower when a poor person injures a rich person.
This latter rule redistributes income but distorts “safety” behavior: the rich overinvest in precautions, while the poor underinvest, creating a net economic loss. For example, a wealthy person might spend more than otherwise on a vehicle with extra safety or warning features. By contrast, keeping strict liability and raising taxes slightly on the rich achieves the same amount redistribution, yet without any of the distortions to precautionary costs. These savings redound to the government in a budget surplus that it can use to reduce everyone’s taxes or increase transfers to the poor.
I am not the first to challenge Kaplow and Shavell’s double-distortion thesis. Yet many responses remain on the same abstract plane, reworking their tort example. This has limited the debate, leaving it detached from legal areas where distribution matters most, such as the minimum wage, collective bargaining, antitrust, intellectual property, and housing regulation. The goal of my book is to engage directly with these more tangible areas of predistributive law. When one does so, I argue, we discover unique, even idiosyncratic, reasons why legal rules can redistribute more cheaply than the tax system.
I argue that, overall, there are two reasons that Kaplow and Shavell’s prescription can fail. First, the distributive, and hence labor supply, effects of taxes and legal rules may not be equivalent. When deciding how much paid labor to supply, a person may respond very differently, for example, to a certain increase of $1,000 in taxes than to a 10 percent increased chance of being assessed $10,000 in damages—even though in terms of expected value (0.10 x $10,000 = $1000) they are equal. The forms of income may matter, too. Stricter enforcement of antitrust law can reduce the capital income received from owning interests in monopolistic firms. But we would not expect identical labor-responses to changes in labor income and capital income, because capital income has, at best, a complicated relationship to labor supply. Wealth can be inherited, rather than earned, and even one’s savings account can change in unanticipated ways that have nothing to do with one’s labor supply.
The second, more important fault in the theory is that there is no reason to assume that the double distortion will be additive rather than subtractive. That is, Kaplow and Shavell simply assume that the two distortions will move in the same direction. But it is completely possible for distortions to counteract each other. Consider, for instance, the minimum wage. The minimum wage has well-known distributive effects, which we can assume will reduce labor supply, just as Kaplow and Shavell assert. And, as every critic of the minimum wage will also tell you, the minimum wage lowers the demand for low-wage labor.
But there is a third effect as well. As more realistic models of the labor market recognize, labor contracts are not perfectly enforceable. This means that the wage can play an important role in regulating worker behavior. For example, higher wages can increase worker effort on the job, reduce turnover, or lower the employer’s supervision costs (because a worker wants to hang on more to a better-paying job, the employer’s threat of termination, and therefore supervision, is less necessary). As a result, a minimum wage may well increase employers’ wage costs, but these costs may be counterbalanced or even outweighed by savings from greater productivity, lower turnover, or reduced supervision.
In the aggregate, these effects can make the minimum wage more distributively cost effective than the income tax. Say we raise the minimum wage by one dollar and use this to measure its labor supply cost, which, along with Kaplow and Shavell, we will assume is the same as a distribution-equivalent change to the income tax. Let’s also say that the minimum wage reduces supervision costs by $0.75. To keep matters simple, suppose that the wage increase and reduced monitoring costs result in a slight decrease in the demand for low-wage labor, producing a net labor demand cost of 25 cents.
By Kaplow and Shavell’s own claim, taxes don’t have any of the behavioral effects in the areas regulated by the legal rules. Thus, a distribution-equivalent change to the income tax will have a $1.00 labor supply cost, but it won’t have the labor demand cost of $0.25, nor the supervision savings of $0.75. That makes the minimum wage a better, more efficient distributive deal. The minimum wage cost comes to a total of 50 cents: one dollar in distributive (labor supply) cost, 75 cents savings in supervision costs, and 25 cents labor demand cost (– $1.00 + $0.75 – $0.25 = – $0.50). In contrast, a distribution-equivalent income tax change will cost one dollar (exclusively the labor-supply cost). Without the offsetting savings in supervision costs, Kaplow and Shavell would be right: the double distortion of the distributive (labor supply) and labor demand effects would be $1.25, higher than the tax alternative. But Kaplow and Shavell simply ignore the possibility of any offsetting “distortions.”
It is important to note that this justification for the minimum wage differs from the most common defense offered by economists. Many economists still think of the minimum wage as distorting a presumed perfectly competitive labor market. The only justification for the minimum wage would be as a corrective to employer market power, or monopsony. To the extent that the minimum wage is such a corrective, it has no economic cost at all, but only enlarges the economic pie, so to speak. But then the justification for the minimum is not a redistributive one, and Kaplow and Shavell have no objection to it. By contrast, Ending Income Inequality engages with a less appreciated debate where the explicit justification for the minimum wage is distributive. In this case, the crucial question is not whether the minimum wage makes the economic pie bigger, but whether it distributes more effectively—or, less inefficiently—than the tax and transfer system.
Beyond Distribution
My book challenges Kaplow and Shavell’s double-distortion thesis, which argues redistribution is always superior to predistribution. Because they are economists, with a formal theory and model to justify their policy preferences, I respond on the same ground. For the sake of argument, I grant all their basic assumptions about individual economic behavior. This prevents the objection that I have simply swapped in different premises. Even on their terms, their conclusions do not necessarily follow. I intend that strategy to provide support for those who want to advance a predistribution agenda, done so in a language that still dominates policy discourse.
This critical method is useful for those purposes but has clear limits: the foundations of modern economics are not beyond critique. On the contrary, writing this book has deepened my skepticism. Economic theory’s framework of utility, interpersonal welfare comparisons, and social welfare maximization—abstracted from history and specific social relations of production—produces an overly technical, even technocratic, vision of the economy. With today’s breakdown of both institutions and the very idea of policy discourse, it is clear we must look beyond economics’ worldview and beyond distribution alone.
Toward the end of his life, Karl Marx wrote, in what has come to be called the Critique of the Gotha Programme:
Any distribution whatever of the means of consumption is only a consequence of the distribution of the conditions of production themselves. The latter distribution, however, is a feature of the mode of production itself. The capitalist mode of production, for example, rests on the fact that the material conditions of production are in the hands of nonworkers in the form of property in capital and land, while the masses are only owners of the personal condition of production, of labor power. If the elements of production are so distributed, then the present-day distribution of the means of consumption results automatically. If the material conditions of production are the co-operative property of the workers themselves, then there likewise results a distribution of the means of consumption different from the present one. Vulgar socialism (and from it in turn a section of the democrats) has taken over from the bourgeois economists the consideration and treatment of distribution as independent of the mode of production and hence the presentation of socialism as turning principally on distribution. After the real relation has long been made clear, why retrogress again?
Marx may overstate the case, but his core insight here is sound: distribution cannot be separated from the mode of production that generates it. Distributional “fixes” that leave production relations intact both sustain capitalism and contain the seeds of their own undoing. If production relations shape income distribution, then even successful efforts to reduce inequality only treat a symptom, not the cause.
Those concerned with distribution should therefore turn their attention to transforming this mode of production itself. We have no shortage of initial ideas for public ownership of society’s productive assets, but much more discussion is needed, and such proposals should move to the center of the political agenda. Their marginal status reflects an economics discipline— and an encompassing liberal discourse, found in both left- and right-wing versions—that treats production as purely technical, devoid of any normative or social status, while restricting policy choice to distribution. This outlook underlies the economists’ division between “efficiency” and “distribution” and explains the policy bias toward redistribution over predistribution. Arguments for predistribution begin to challenge that divide, but staying within the realm of distribution, they tend to presuppose the same technical view of the economy. A more comprehensive challenge would put public ownership of production at the heart of the political agenda.
** ** **
This post is co-published with our comrades at Jacobin.
