Capital in the Twenty-First Century

by Thomas Piketty

  This is perfectly legal and not inherently problematic.14 Nevertheless, it does present a challenge to the tax system. If some people are taxed on the basis of declared incomes that are only 1 percent of their economic incomes, or even 10 percent, then nothing is accomplished by taxing that income at a rate of 50 percent or even 98 percent. The problem is that this is how the tax system works in practice in the developed countries. Effective tax rates (expressed as a percentage of economic income) are extremely low at the top of the wealth hierarchy, which is problematic, since it accentuates the explosive dynamic of wealth inequality, especially when larger fortunes are able to garner larger returns. In fact, the tax system ought to attenuate this dynamic, not accentuate it.

  There are several ways to deal with this problem. One would be to tax all of a person’s income, including the part that accumulates in trusts, holding companies, and partnerships. A simpler solution is to compute the tax due on the basis of wealth rather than income. One could then assume a flat yield (of, say, 5 percent a year) to estimate the income on the capital and include that amount in the income subject to a progressive income tax. Some countries, such as the Netherlands, have tried this but have run into a number of difficulties having to do with the range of assets covered and the choice of a return on capital.15 Another solution is to apply a progressive tax directly to an individual’s total wealth. The important advantage of this approach is that one can vary the tax rate with the size of the fortune, since we know that in practice larger fortunes earn larger returns.

  In view of the finding that fortunes at the top of the wealth hierarchy are earning very high returns, this contributive argument is the most important justification of a progressive tax on capital. According to this reasoning, capital is a better indicator of the contributive capacity of very wealthy individuals than is income, which is often difficult to measure. A tax on capital is thus needed in addition to the income tax for those individuals whose taxable income is clearly too low in light of their wealth.16

  Nevertheless, another classic argument in favor of a capital tax should not be neglected. It relies on a logic of incentives. The basic idea is that a tax on capital is an incentive to seek the best possible return on one’s capital stock. Concretely, a tax of 1 or 2 percent on wealth is relatively light for an entrepreneur who manages to earn 10 percent a year on her capital. By contrast, it is quite heavy for a person who is content to park her wealth in investments returning at most 2 or 3 percent a year. According to this logic, the purpose of the tax on capital is thus to force people who use their wealth inefficiently to sell assets in order to pay their taxes, thus ensuring that those assets wind up in the hands of more dynamic investors.

  There is some validity to this argument, but it should not be overstated.17 In practice, the return on capital does not depend solely on the talent and effort supplied by the capitalist. For one thing, the average return varies systematically with the size of the initial fortune. For another, individual returns are largely unpredictable and chaotic and are affected by all sorts of economic shocks. For example, there are many reasons why a firm might be losing money at any given point in time. A tax system based solely on the capital stock (and not on realized profits) would put disproportionate pressure on companies in the red, because their taxes would be as high when they were losing money as when they were earning high profits, and this could plunge them into bankruptcy.18 The ideal tax system is therefore a compromise between the incentive logic (which favors a tax on the capital stock) and an insurance logic (which favors a tax on the revenue stream stemming from capital).19 The unpredictability of the return on capital explains, moreover, why it is more efficient to tax heirs not once and for all, at the moment of inheritance (by way of the estate tax), but throughout their lives, via taxes based on both capital income and the value of the capital stock.20 In other words, all three types of tax—on inheritance, income, and capital—play useful and complementary roles (even if income is perfectly observable for all taxpayers, no matter how wealthy).21

  A Blueprint for a European Wealth Tax

  Taking all these factors into account, what is the ideal schedule for a tax on capital, and how much would such a tax bring in? To be clear, I am speaking here of a permanent annual tax on capital at a rate that must therefore be fairly moderate. A tax collected only once a generation, such as an inheritance tax, can be assessed at a very high rate: a third, a half, or even two-thirds, as was the case for the largest estates in Britain and the United States from 1930 to 1980.22 The same is true of exceptional one-time taxes on capital levied in unusual circumstances, such as the tax levied on capital in France in 1945 at rates as high as 25 percent, indeed 100 percent for additions to capital during the Occupation (1940–1945). Clearly, such taxes cannot be applied for very long: if the government takes a quarter of the nation’s wealth every year, there will be nothing left to tax after a few years. That is why the rates of an annual tax on capital must be much lower, on the order of a few percent. To some this may seem surprising, but it is actually quite a substantial tax, since it is levied every year on the total stock of capital. For example, the property tax rate is frequently just 0.5–1 percent of the value of real estate, or a tenth to a quarter of the rental value of the property (assuming an average rental return of 4 percent a year).23

  The next point is important, and I want to insist on it: given the very high level of private wealth in Europe today, a progressive annual tax on wealth at modest rates could bring in significant revenue. Take, for example, a wealth tax of 0 percent on fortunes below 1 million euros, 1 percent between 1 and 5 million euros, and 2 percent above 5 million euros. If applied to all member states of the European Union, such a tax would affect about 2.5 percent of the population and bring in revenues equivalent to 2 percent of Europe’s GDP.24 The high return should come as no surprise: it is due simply to the fact that private wealth in Europe today is worth more than five years of GDP, and much of that wealth is concentrated in the upper centiles of the distribution.25 Although a tax on capital would not by itself bring in enough to finance the social state, the additional revenues it would generate are nevertheless significant.

  In principle, each member state of the European Union could generate similar revenues by applying such a tax on its own. But without automatic sharing of bank information both inside and outside EU territory (starting with Switzerland among nonmember states) the risks of evasion would be very high. This partly explains why countries that have adopted a wealth tax (such as France, which employs a tax schedule similar to the one I am proposing) generally allow numerous exemptions, especially for “business assets” and, in practice, for nearly all large stakes in listed and unlisted companies. To do this is to drain much of the content from the progressive tax on capital, and that is why existing taxes have generated revenues so much smaller than the ones described above.26 An extreme example of the difficulties European countries face when they try to impose a capital tax on their own can be seen in Italy. In 2012, the Italian government, faced with one of the largest public debts in Europe and also with an exceptionally high level of private wealth (also one of the highest in Europe, along with Spain),27 decided to introduce a new tax on wealth. But for fear that financial assets would flee the country in search of refuge in Swiss, Austrian, and French banks, the rate was set at 0.8 percent on real estate and only 0.1 percent on bank deposits and other financial assets (except stocks, which were totally exempt), with no progressivity. Not only is it hard to think of an economic principle that would explain why some assets should be taxed at one-eighth the rate of others; the system also had the unfortunate consequence of imposing a regressive tax on wealth, since the largest fortunes consist mainly of financial assets and especially stocks. This design probably did little to earn social acceptance for the new tax, which became a major issue in the 2013 Italian elections; the candidate who had proposed the tax—with the compliments of European and international authorities—was roundly defeated
at the polls. The crux of the problem is this: without automatic sharing of bank information among European countries, which would allow the tax authorities to obtain reliable information about the net assets of all taxpayers, no matter where those assets are located, it is very difficult for a country acting on its own to impose a progressive tax on capital. This is especially unfortunate, because such a tax is a tool particularly well suited to Europe’s current economic predicament.

  Suppose that bank information is automatically shared and the tax authorities have accurate assessments of who owns what, which may happen some day. What would then be the ideal tax schedule? As usual, there is no mathematical formula for answering this question, which is a matter for democratic deliberation. It would make sense to tax net wealth below 200,000 euros at 0.1 percent and net wealth between 200,000 and 1 million euros at 0.5 percent. This would replace the property tax, which in most countries is tantamount to a wealth tax on the propertied middle class. The new system would be both more just and more efficient, because it targets all assets (not only real estate) and relies on transparent data and market values net of mortgage debt.28 To a large extent a tax of this sort could be readily implemented by individual countries acting alone.

  Note that there is no reason why the tax rate on fortunes above 5 million euros should be limited to 2 percent. Since the real returns on the largest fortunes in Europe and around the world are 6 to 7 percent or more, it would not be excessive to tax fortunes above 100 million or 1 billion euros at rates well above 2 percent. The simplest and fairest procedure would be to set rates on the basis of observed returns in each wealth bracket over several prior years. In that way, the degree of progressivity can be adjusted to match the evolution of returns to capital and the desired level of wealth concentration. To avoid divergence of the wealth distribution (that is, a steadily increasing share belonging to the top centiles and thousandths), which on its face seems to be a minimal desirable objective, it would probably be necessary to levy rates of about 5 percent on the largest fortunes. If a more ambitious goal is preferred—say, to reduce wealth inequality to more moderate levels than exist today (and which history shows are not necessary for growth)—one might envision rates of 10 percent or higher on billionaires. This is not the place to resolve the issue. What is certain is that it makes little sense to take the yield on public debt as a reference, as is often done in political debate.29 The largest fortunes are clearly not invested in government bonds.

  Is a European wealth tax realistic? There is no technical reason why not. It is the tool best suited to meet the economic challenges of the twenty-first century, especially in Europe, where private wealth is thriving to a degree not seen since the Belle Époque. But if the countries of the Old Continent are to cooperate more closely, European political institutions will have to change. The only strong European institution at the moment is the ECB, which is important but notoriously insufficient. I come back to this in the next chapter, when I turn to the question of the public debt crisis. Before that, it will be useful to look at the proposed tax on capital in a broader historical perspective.

  Capital Taxation in Historical Perspective

  In all civilizations, the fact that the owners of capital claim a substantial share of national income without working and that the rate of return on capital is generally 4–5 percent a year has provoked vehement, often indignant, reactions as well as a variety of political responses. One of the most common of the latter has been the prohibition of usury, which we find in one form or another in most religious traditions, including those of Christianity and Islam. The Greek philosophers were of two minds about interest, which, since time never ceases to flow, can in principle increase wealth without limit. It was the danger of limitless wealth that Aristotle singled out when he observed that the word “interest” in Greek (tocos) means “child.” In his view, money ought not to “give birth” to more money.30 In a world of low or even near-zero growth, where both population and output remained more or less the same generation after generation, “limitlessness” seemed particularly dangerous.

  Unfortunately, the attempts to prohibit interest were often illogical. The effect of outlawing loans at interest was generally to restrict certain types of investment and certain categories of commercial or financial activity that the political or religious authorities deemed less legitimate or worthy than others. They did not, however, question the legitimacy of returns to capital in general. In the agrarian societies of Europe, the Christian authorities never questioned the legitimacy of land rents, from which they themselves benefited, as did the social groups on which they depended to maintain the social order. The prohibition of usury in the society of that time is best thought of as a form of social control: some types of capital were more difficult to control than others and therefore more worrisome. The general principle according to which capital can provide income for its owner, who need not work to justify it, went unquestioned. The idea was rather to be wary of infinite accumulation. Income from capital was supposed to be used in healthy ways, to pay for good works, for example, and certainly not to launch into commercial or financial adventures that might lead to estrangement from the true faith. Landed capital was in this respect very reassuring, since it could do nothing but reproduce itself year after year and century after century.31 Consequently, the whole social and spiritual order also seemed immutable. Land rent, before it became the sworn enemy of democracy, was long seen as the wellspring of social harmony, at least by those to whom it accrued.

  The solution to the problem of capital suggested by Karl Marx and many other socialist writers in the nineteenth century and put into practice in the Soviet Union and elsewhere in the twentieth century was far more radical and, if nothing else, more logically consistent. By abolishing private ownership of the means of production, including land and buildings as well as industrial, financial, and business capital (other than a few individual plots of land and small cooperatives), the Soviet experiment simultaneously eliminated all private returns on capital. The prohibition of usury thus became general: the rate of exploitation, which for Marx represented the share of output appropriated by the capitalist, thus fell to zero, and with it the rate of private return. With zero return on capital, man (or the worker) finally threw off his chains along with the yoke of accumulated wealth. The present reasserted its rights over the past. The inequality r > g was nothing but a bad memory, especially since communism vaunted its affection for growth and technological progress. Unfortunately for the people caught up in these totalitarian experiments, the problem was that private property and the market economy do not serve solely to ensure the domination of capital over those who have nothing to sell but their labor power. They also play a useful role in coordinating the actions of millions of individuals, and it is not so easy to do without them. The human disasters caused by Soviet-style centralized planning illustrate this quite clearly.

  A tax on capital would be a less violent and more efficient response to the eternal problem of private capital and its return. A progressive levy on individual wealth would reassert control over capitalism in the name of the general interest while relying on the forces of private property and competition. Each type of capital would be taxed in the same way, with no discrimination a priori, in keeping with the principle that investors are generally in a better position than the government to decide what to invest in.32 If necessary, the tax can be quite steeply progressive on very large fortunes, but this is a matter for democratic debate under a government of laws. A capital tax is the most appropriate response to the inequality r > g as well as to the inequality of returns to capital as a function of the size of the initial stake.33

  In this form, the tax on capital is a new idea, designed explicitly for the globalized patrimonial capitalism of the twenty-first century. To be sure, capital in the form of land has been taxed since time immemorial. But property is generally taxed at a low flat rate. The main purpose of the property tax is to guarantee property rig
hts by requiring registration of titles; it is certainly not to redistribute wealth. The English, American, and French revolutions all conformed to this logic: the tax systems they put in place were in no way intended to reduce inequalities of wealth. During the French Revolution the idea of progressive taxation was the subject of lively debate, but in the end the principle of progressivity was rejected. What is more, the boldest tax proposals of that time seem quite moderate today in the sense that the proposed tax rates were quite low.34

  The progressive tax revolution had to await the twentieth century and the period between the two world wars. It occurred in the midst of chaos and came primarily in the form of progressive taxes on income and inheritances. To be sure, some countries (most notably Germany and Sweden) established an annual progressive tax on capital as early as the late nineteenth century or early twentieth. But the United States, Britain, and France (until the 1980s) did not move in this direction.35 Furthermore, in the countries that did tax capital, the rates were relatively low, no doubt because these taxes were designed in a context very different from that which exists today. These taxes also suffered from a fundamental technical flaw: they were based not on the market value of the assets subject to taxation, to be revised annually, but on infrequently revised assessments of their value by the tax authorities. These assessed valuations eventually lost all connection with market values, which quickly rendered the taxes useless. The same flaw undermined the property tax in France and many other countries subsequent to the inflationary shock of the period 1914–1945.36 Such a design flaw can be fatal to a progressive tax on capital: the threshold for each tax bracket depends on more or less arbitrary factors such as the date of the last property assessment in a given town or neighborhood. Challenges to such arbitrary taxation became increasingly common after 1960, in a period of rapidly rising real estate and stock prices. Often the courts became involved (to rule on violations of the principle of equal taxation). Germany and Sweden abolished their annual taxes on capital in 1990–2010. This had more to do with the archaic design of these taxes (which went back to the nineteenth century) than with any response to tax competition.37