The Piketty Observation against the Institutional Background: How Natural is this Natural Tendency and What Can We Do about it?

1 Introduction

The Piketty observation, needs no introduction: r > g; the rate of return on capital (g) is greater than the growth rate in the economy (r). Thomas Piketty’s book, Capital in the Twenty-First Century (Piketty, 2014), provides a great deal of historical support for the observation, and derives many valuable insights from it. Two of these insights are particularly important: entrepreneurs eventually become rentiers, and except during unusual circumstances, inequality between capital owners and wage earners tends to rise over time. If entrepreneurs eventually become rentiers, there is little merit in the tremendous inequalities in our economic system. Even if a fortune begins with ownership, eventually the holder of that fortune is rewarded more just for having wealth than for what one does to become wealthy initially. If inequality tends to rise over time because of the high returns to capital, the organization of the economy threatens both equality and merit.

Piketty’s book is extremely valuable as a guide to good policymaking. Toward that effort, this paper views Piketty’s observation against the institutional setting that has prevailed over the period of his study and makes two observations to help understand how to use it as policy guide. First, whether Piketty’s two insights follow from his observation depends not simply on whether r is greater than g, but on whether the difference between the two is greater than the consumption of the capital-owning group. If capital owners spend more than the difference between r and g, inequality between capital owners and workers will not increase over time. The relative size of capitalists’ consumption and capital income is not obvious, and therefore, more evidence is needed to confirm the connection between Piketty’s observation and his insights. Second, the statement r has been greater than g is more accurate than simply r is greater than g. Whether r continues to exceed g depends crucially on the political and institutional environment in question. There is no natural set of property institutions, and those that have prevailed over the two-hundred-year period from which Piketty draws his observations are extremely favorable to capital owners. Because the prevailing set of property institutions is within society’s power to control, whether r has exceeded g and by how much is not the result of a natural tendency but of (witting or unwitting) institutional design. If so, whether r continues to exceed g is ultimately up to the people to decide.

These observations do not overturn the importance of Piketty’s observations, but awareness of the flexibility of potential property institutions raises many ethical questions and makes many tools available to address inequality. The observation that r has tended to exceed g, and awareness of the dangers of allowing it to do so should be used as a guide to public policy making. This paper will argue that, rather than viewing r > g as natural tendency of capitalism, we should use the relationship as a policy guideline, targeting policies over time to keep r – g low or perhaps negative.

2 A more complex relationship

The observation that r > g is not enough to show that the entrepreneur eventually becomes a rentier or that inequality (between capital owners and wage-earners) tends to rise over time. For one thing, r must be defined in after tax terms. If the return on capital exceeds the growth rate in the economy, but we tax the difference away, there is clearly no opportunity for average property owners to gradually capture a larger and larger share of total wealth simply by reinvesting their returns. Piketty is aware of this issue. He discusses taxation at various points in the book, but pointing out the effect of taxes on capital accumulation in his most basic equation, shows how much power government has to control the relationship between r and g.

If r exceeds g, then property owners can gradually capture a larger-and-larger share of total wealth. However, only if the difference between r and g is greater than consumption out of capital income, do owners actually capture a larger-and-larger share of total wealth. Again, Piketty is aware of this issue. He discusses the savings rate at various points in the book, but we can emphasize this point by considering consumption in the simple formula relating r and g.

Use C to stand for consumption out of capital income:

inequality between capital and labor income tends to rise over time, all else equal.

capital owners spend the difference between their income and the economic growth rate, and so that, over time, their relative wealth neither falls nor rises in relation to total wealth.

despite the potential for frugal rentiers to capture a larger share of ownership, typical rentiers eventually spend down their fortunes and wealth becomes more equal over time.

My guess is that Piketty didn’t bring this complication into his simple equation because he thought it was obvious that some fortunes are so large that no one could spend it all on consumption. It might be obvious for the very largest fortunes, but it is not obvious for the average capital owner. A multi-million dollar fortune split between ex-wives, children, and grandchildren, could leave all heirs in the middle class. It is possible that capital-owning families tend to have children more rapidly than their capital grows, and possibly even in low-fertility families, wealth might eventually tend to reach a spendthrift heir who squanders it. Therefore, more research is needed to establish the historical relationships between r, g, and C.

There are important normative questions whether policymakers should target r – g or r – C – g. The variable to target remains r. If society can increase g without overstretching our workers or our resource base, there’s no reason not to, but presumably we’re doing this to the best of our ability already, and it can’t be used as policy target for this issue. Little reason exists to maintain equality by encouraging greater consumption by capital owners because investment is something we want capital owners to do. We need to reduce r without hurting incentives so much that they have a significant negative effect on investment and growth. This will take experimentation.

Can we use policy to reduce r to the point where r – C – g = 0, or r – g = 0, or even to the point at which r – g < 0 without unacceptable negative effects on investment and growth? Trial and error is probably needed, and it is up to the people in a democratic society to determine what negative effects are unacceptable. If we reduce r to the point where r – C = g, we have decided to allow rentier dynasties to continue to grow (in relative terms) only in families that are above average in their frugality. If we reduce r to the point where r = g or less, we have decided that frugality is not enough for a family to capture a raising share of economic; they must be above average in their entrepreneurship. In this setting, it would only be a few families – those lucky or talented enough to make above-average returns for generations at a time – who could continue to capture a larger share of income across generations. A nation concerned with merit and equal opportunity is likely to want to reduce r to this level as long as it does not come with an unacceptable cost in reductions of investment and growth. I will return to this issue below.

3 There is no natural institutional setting

Piketty makes it clear that he has not established r > g as an immutable law but only as a historical fact that has prevailed in a large number of countries for a long time. He does not list it among his “laws of capitalism.” Thus, he’s well aware that r > g is merely an observation. Although he argues that there is good reason to expect the inequality to continue or even to widen in the future, he is aware that the observation is something that has happened; not necessarily something that will always happen. Nevertheless he tends to treat r > g as if it were an observation of capitalism in its natural setting.

Daron Acemoglu and James A. Robinson have criticized Piketty for attempting to identify any laws or natural tendencies of capitalism (Acemoglu & Robinson, 2014). They argue that the institutional structure of capitalism is too important and too variable to identify any natural laws of capitalism. Institutional factors affect r and g and the relationship between them; and institutional factors can alter inequality even while the relationship between r and g remains constant. Although Acemoglu and Robinson are right that institutions are important and variable, their criticism of Piketty goes too far in two senses. First, the relationship between r and g is undeniably important in determining the level of inequality between capital owners and workers, even if at some times and places it is not the most important factor, and therefore, it should be a focus of policy. Second, their criticism attributes to Piketty what is in fact a long-time convention of economics. The first of these responses needs no elaboration, but the second does.

Ever since Adam Smith wrote about the “system of natural liberty,” (Smith, 2005, p. 560), economists have tended to view one specific institutional setting, some idea of laissez-faire capitalism, as natural. This convention allows economists to view the economy as a system to be studied, diagnosed, and treated while they view government policy as an outside force that administers treatment – whether it works or not. Although Piketty doesn’t discuss the issue of whether the institutional setting in which r > g is any more natural than any other, at worst, he seems to be doing no more than following a long established convention in the field.

The convention of viewing laissez-faire as natural capitalism has its uses. For example, suppose the government has never had any specific policy toward cigarette smoking, and then decides, in the interest of pubic health, it should do something to reduce smoking. It then seems appropriate to view the cigarette market as relatively laissez faire with natural tendencies and to look at possible tax and regulation policies and specifically targeted treatments meant to alter those natural tendencies.

But for such a fundamental, system-wide issue such as the relationship between the rate of return on capital and the growth rate of the economy as a whole, there is no natural structure of capitalism. The market system is a complex set of institutions that have grown out of a long history and are dependent on the laws of society. Even within the realm of “laissez-faire,” there can be no trading system for the government to leave alone unless resources are privatized first, but the privatization of resources is dependent on a system of laws, which could be structured in many – if not an infinite number – of different ways. For example, an economy in which anyone who wishes can dump sewage into the river we all drink from is no more laissez faire or natural than one in which anyone who wishes to dump sewage into the river must first buy the right from everyone who drinks from it.

In each possible property rights regime the relationship between r and g is likely to differ. The laws determining how it is structured should be subject to democratic debate in which it is up to the people to decide what it means to own a resource or something we make out of a resource. If the r exceeds g by so much that entrepreneurs eventually become rentiers and the divide between rentiers and the rest of us continually widens, it is not because the natural economic system has this tendency; it is because policymakers have made laws creating an institutional background on which trade has this tendency. We might not have understood what we were doing when we made it, but we can learn, and we can unmake it. That is, we should make the relationship between r and g a policy guideline. We should be aware of this relationship and when it is one of the sources of unacceptable inequality, we need to use taxation and other policies to change the institutional setting to bring equality in line.

4 What can we do about it?

When we take the view that there is no natural makeup of capitalism and the structure of the property rights regime is subject to democratic debate, the amount of tools at our disposal to maintain economic equality increases enormously. For example, in most places land rights, mineral rights, water rights, broadcast-spectrum rights and a host of other rights were privatized years ago for little or no charge with little or no democratic debate. Any individual can partially privatize the atmosphere as her personal sinkhole free of charge simply by purchasing an automobile with an internal combustion engine and accompanying tailpipe. If we consider different rules for how all these resources are privatized, we consider the issue of predistribution: the distributional effects of rules made before rather than after trade (O’Neill 2012; O’Neill & Williamson, 2012). Different predistributional policies will have different effects on r, g, C, and inequality in general.

One predistributional policy we should consider is the principle of pay-for-what-you-take-out-of-the-commons. Imagine a system in which every private holder makes a regular payment for the resources they own. The revenue goes into a fund that in turn finances an Unconditional Basic Income for all. You pay everyone else for the resources you own; and you receive a payment from everyone else for the resources they own. If you own an average amount of resources, the two payments would be exactly equal. If you own more than average, you pay more than you receive, as you should if you are claiming a more-than-average share of resources in a society of equal citizens. If you own less than average, you will receive more than you pay in compensation for having access to fewer resources than other citizens. For options along these lines see (Widerquist, 2006, 2012, Forthcoming).

This kind of policy would have an obvious direct effect promoting greater equality, and it will also have an indirect effect by putting workers in a better bargaining position in the market. The total effect on the rate of return on capital and on inequality is difficult to estimate in advance, but some good attempts exist (Flomenhoft, 2012), and the overall effect is likely to be substantial. This predistributional strategy should not be the only strategy to address economic inequality, but it could be a prominent tool in the policymaker’s kit. This resource-based strategy has an advantage against the “race to the bottom” issue that Piketty is concerned with, because the resources in question tend to be immobile. They cannot easily escape across national boundaries and avoid taxation.

I am in one way more optimistic and in one way less optimistic than Piketty on the issue of international flows of capital. Piketty is quite worried that capital fight is a major problem with taxation. His book reads as if the governments of the world would like to promote greater equality but are trapped by the threat of capital flight into this race to the bottom. More optimistically than Piketty, I do not believe that the race to the bottom is a significant reason taxes on the wealthy are so low, because only some kinds of taxes are vulnerable to it, and governments have many other options for promoting equality.

Less optimistically, I am concerned that governments lack the will: the main reason taxes on capital income are so low is that our governments disproportionately represent the wealthy. Whether it is through campaign donations, lobbying, similarity of interests, proximity to power, or many other reasons, the wealthy have enormous influence over the political process, far beyond their numbers. A short article like this one is not the place to discuss all the reforms to address this enormous problem, but we need to recognize that noneconomic reforms toward greater political equality and a deeper, more thorough democracy are important reforms in the effort to maintain economic equality.