Capital in the Twenty-First Century

by Thomas Piketty

  Beyond these definitional quibbles, a more pertinent criterion for characterizing different types of tax is the degree to which each type is proportional or progressive. A tax is called “proportional” when its rate is the same for everyone (the term “flat tax” is also used). A tax is progressive when its rate is higher for some than for others, whether it be those who earn more, those who own more, or those who consume more. A tax can also be regressive, when its rate decreases for richer individuals, either because they are partially exempt (either legally, as a result of fiscal optimization, or illegally, through evasion) or because the law imposes a regressive rate, like the famous “poll tax” that cost Margaret Thatcher her post as prime minister in 1990.3

  In the modern fiscal state, total tax payments are often close to proportional to individual income, especially in countries where the total is large. This is not surprising: it is impossible to tax half of national income to finance an ambitious program of social entitlements without asking everyone to make a substantial contribution. The logic of universal rights that governed the development of the modern fiscal and social state fits rather well, moreover, with the idea of a proportional or slightly progressive tax.

  It would be wrong, however, to conclude that progressive taxation plays only a limited role in modern redistribution. First, even if taxation overall is fairly close to proportional for the majority of the population, the fact that the highest incomes and largest fortunes are taxed at significantly higher (or lower) rates can have a strong influence on the structure of inequality. In particular, the evidence suggests that progressive taxation of very high incomes and very large estates partly explains why the concentration of wealth never regained its astronomic Belle Époque levels after the shocks of 1914–1945. Conversely, the spectacular decrease in the progressivity of the income tax in the United States and Britain since 1980, even though both countries had been among the leaders in progressive taxation after World War II, probably explains much of the increase in the very highest earned incomes. At the same time, the recent rise of tax competition in a world of free-flowing capital has led many governments to exempt capital income from the progressive income tax. This is particularly true in Europe, whose relatively small states have thus far proved incapable of achieving a coordinated tax policy. The result is an endless race to the bottom, leading, for example, to cuts in corporate tax rates and to the exemption of interest, dividends, and other financial revenues from the taxes to which labor incomes are subject.

  One consequence of this is that in most countries taxes have (or will soon) become regressive at the top of the income hierarchy. For example, a detailed study of French taxes in 2010, which looked at all forms of taxation, found that the overall rate of taxation (47 percent of national income on average) broke down as follows. The bottom 50 percent of the income distribution pay a rate of 40–45 percent; the next 40 percent pay 45–50 percent; but the top 5 percent and even more the top 1 percent pay lower rates, with the top 0.1 percent paying only 35 percent. The high tax rates on the poor reflect the importance of consumption taxes and social contributions (which together account for three-quarters of French tax revenues). The slight progressivity observed in the middle class is due to the growing importance of the income tax. Conversely, the clear regressivity in the top centiles reflects the importance at this level of capital income, which is largely exempt from progressive taxation. The effect of this outweighs the effect of taxes on the capital stock (which are the most progressive of all).4 All signs are that taxes elsewhere in Europe (and probably also in the United States) follow a similar bell curve, which is probably even more pronounced than this imperfect estimate indicates.5

  If taxation at the top of the social hierarchy were to become more regressive in the future, the impact on the dynamics of wealth inequality would likely be significant, leading to a very high concentration of capital. Clearly, such a fiscal secession of the wealthiest citizens could potentially do great damage to fiscal consent in general. Consensus support for the fiscal and social state, which is already fragile in a period of low growth, would be further reduced, especially among the middle class, who would naturally find it difficult to accept that they should pay more than the upper class. Individualism and selfishness would flourish: since the system as a whole would be unjust, why continue to pay for others? If the modern social state is to continue to exist, it is therefore essential that the underlying tax system retain a minimum of progressivity, or at any rate that it not become overtly regressive at the top.

  Furthermore, looking at the progressivity of the tax system by examining how heavily top incomes are taxed obviously fails to weigh inherited wealth, whose importance has been increasing.6 In practice, estates are much less heavily taxed than income.7 This exacerbates what I have called “Rastignac’s dilemma.” If individuals were classified by centile of total resources accrued over a lifetime (including both earned income and capitalized inheritance), which is a more satisfactory criterion for progressive taxation, the bell curve would be even more markedly regressive at the top of the hierarchy than it is when only labor incomes are considered.8

  One final point bears emphasizing: to the extent that globalization weighs particularly heavily on the least skilled workers in the wealthy countries, a more progressive tax system might in principle be justified, adding yet another layer of complexity to the overall picture. To be sure, if one wants to maintain total taxes at about 50 percent of national income, it is inevitable that everyone must pay a substantial amount. But instead of a slightly progressive tax system (leaving aside the very top of the hierarchy), one can easily imagine a more steeply progressive one.9 This would not solve all the problems, but it would be enough to improve the situation of the least skilled significantly.10 If the tax system is not made more progressive, it should come as no surprise that those who derive the least benefit from free trade may well turn against it. The progressive tax is indispensable for making sure that everyone benefits from globalization, and the increasingly glaring absence of progressive taxation may ultimately undermine support for a globalized economy.

  For all of these reasons, a progressive tax is a crucial component of the social state: it played a central role in its development and in the transformation of the structure of inequality in the twentieth century, and it remains important for ensuring the viability of the social state in the future. But progressive taxation is today under serious threat, both intellectually (because its various functions have never been fully debated) and politically (because tax competition is allowing entire categories of income to gain exemption from the common rules).

  The Progressive Tax in the Twentieth Century: An Ephemeral Product of Chaos

  To gaze backward for a moment: how did we get to this point? First, it is important to realize that progressive taxation was as much a product of two world wars as it was of democracy. It was adopted in a chaotic climate that called for improvisation, which is part of the reason why its various purposes were not sufficiently thought through and why it is being challenged today.

  To be sure, a number of countries adopted a progressive income tax before the outbreak of World War I. In France, the law creating a “general tax on income” was passed on July 15, 1914, in direct response to the anticipated financial needs of the impending conflict (after being buried in the Senate for several years); the law would not have passed had a declaration of war not been imminent.11 Aside from this exception, most countries adopted a progressive income tax after due deliberation in the normal course of parliamentary proceedings. Such a tax was adopted in Britain, for example, in 1909 and in the United States in 1913. Several countries in northern Europe, a number of German states, and Japan adopted a progressive income tax even earlier: Denmark in 1870, Japan in 1887, Prussia in 1891, and Sweden in1903. Even though not all the developed countries had adopted a progressive tax by 1910, an international consensus was emerging around the principle of progressivity and its application to overall in
come (that is, to the sum of income from labor, including both wage and nonwage labor, and capital income of all kinds, including rent, interest, dividends, profits, and in some cases capital gains).12 To many people, such a system appeared to be both a more just and a more efficient way of apportioning taxes. Overall income measured each person’s ability to contribute, and progressive taxation offered a way of limiting the inequalities produced by industrial capitalism while maintaining respect for private property and the forces of competition. Many books and reports published at the time helped popularize the idea and win over some political leaders and liberal economists, although many would remain hostile to the very principle of progressivity, especially in France.13

  Is the progressive income tax therefore the natural offspring of democracy and universal suffrage? Things are actually more complicated. Indeed, tax rates, even on the most astronomical incomes, remained extremely low prior to World War I. This was true everywhere, without exception. The magnitude of the political shock due to the war is quite clear in Figure 14.1, which shows the evolution of the top rate (that is, the tax rate on the highest income bracket) in the United States, Britain, Germany, and France from 1900 to 2013. The top rate stagnated at insignificant levels until 1914 and then skyrocketed after the war. These curves are typical of those seen in other wealthy countries.14

  FIGURE 14.1. Top income tax rates, 1900–2013

  The top marginal tax rate of the income tax (applying to the highest incomes) in the United States dropped from 70 percent in 1980 to 28 percent in 1988.

  Sources and series: see piketty.pse.ens.fr/capital21c.

  In France, the 1914 income tax law provided for a top rate of just 2 percent, which applied to only a tiny minority of taxpayers. It was only after the war, in a radically different political and financial context, that the top rate was raised to “modern” levels: 50 percent in 1920, then 60 percent in 1924, and even 72 percent in 1925. Particularly striking is the fact that the crucial law of June 25, 1920, which raised the top rate to 50 percent and can actually be seen as a second coming of the income tax, was adopted by the so-called blue-sky Chamber (one of the most right-wing Chambers of Deputies in the history of the French Republic) with its “National Bloc” majority, made up largely of the very delegations who had most vehemently opposed the creation of an income tax with a top rate of 2 percent before the war. This complete reversal of the right-wing position on progressive taxation was of course due to the disastrous financial situation created by the war. During the conflict the government had run up considerable debts, and despite the ritual speeches in which politician after politician declared that “Germany will pay,” everyone knew that new fiscal resources would have to be found. Postwar shortages and the recourse to the printing press had driven inflation to previously unknown heights, so that the purchasing power of workers remained below 1914 levels, and several waves of strikes in May and June of 1919 threatened the country with paralysis. In such circumstances, political proclivities hardly mattered: new sources of revenue were essential, and no one believed that those with the highest incomes ought to be spared. The Bolshevik Revolution of 1917 was fresh in everyone’s mind. It was in this chaotic and explosive situation that the modern progressive income tax was born.15

  The German case is particularly interesting, because Germany had had a progressive income tax for more than twenty years before the war. Throughout that period of peace, tax rates were never raised significantly. In Prussia, the top rate remained stable at 3 percent from 1891 to 1914 and then rose to 4 percent from 1915 to 1918, before ultimately shooting up to 40 percent in 1919–1920, in a radically changed political climate. In the United States, which was intellectually and politically more prepared than any other country to accept a steeply progressive income tax and would lead the movement in the interwar period, it was again not until 1918–1919 that the top rate was abruptly increased, first to 67 and then to 77 percent. In Britain, the top rate was set at 8 percent in 1909, a fairly high level for the time, but again it was not until after the war that it was suddenly raised to more than 40 percent.

  Of course it is impossible to say what would have happened had it not been for the shock of 1914–1918. A movement had clearly been launched. Nevertheless, it seems certain that had that shock not occurred, the move toward a more progressive tax system would at the very least have been much slower, and top rates might never have risen as high as they did. The rates in force before 1914, which were always below 10 percent (and generally below 5), including the top rates, were not very different from tax rates in the eighteenth and nineteenth centuries. Even though the progressive tax on total income was a creation of the late nineteenth and early twentieth centuries, there were much earlier forms of income tax, generally with different rules for different types of income, and usually with flat or nearly flat rates (for example, a flat rate after allowing for a certain fixed deduction). In most cases the rates were 5–10 percent (at most). For example, this was true of the categorical or schedular tax, which applied separate rates to each category (or schedule) of income (land rents, interest, profits, wages, etc.). Britain adopted such a categorical tax in 1842, and it remained the British version of the income tax until the creation in 1909 of a “supertax” (a progressive tax on total income).16

  In Ancien Régime France, there were also various forms of direct taxation of incomes, such as the taille, the dixième, and the vingtième, with typical rates of 5 or 10 percent (as the names indicate) applied to some but not all sources of income, with numerous exemptions. In 1707, Vauban proposed a “dixième royal,” which was intended to be a 10 percent tax on all incomes (including rents paid to aristocratic and ecclesiastical landlords), but it was never fully implemented. Various improvements to the tax system were nevertheless attempted over the course of the eighteenth century.17 Revolutionary lawmakers, hostile to the inquisitorial methods of the fallen monarchy and probably keen as well to protect the emerging industrial bourgeoisie from bearing too heavy a tax burden, chose to institute an “indicial” tax system: taxes were calculated on the basis of indices that were supposed to reflect the taxpayer’s ability to pay rather than actual income, which did not have to be declared. For instance, the “door and window tax” was based on the number of doors and windows in the taxpayer’s primary residence, which was taken to be an index of wealth. Taxpayers liked this system because the authorities could determine how much tax they owed without having to enter their homes, much less examine their account books. The most important tax under the new system created in 1792, the property tax, was based on the rental value of all real estate owned by the taxpayer.18 The income tax was based on estimates of average rental value, which were revised once a decade when the tax authorities inventoried all property in France; taxpayers were not required to declare their actual income. Since inflation was slow, this made little difference. In practice, this real estate tax amounted to a flat tax on rents and was not very different from the British categorical tax. (The effective rate varied from time to time and département to département but never exceeded 10 percent.)

  To round out the system, the nascent Third Republic decided in 1872 to impose a tax on income from financial assets. This was a flat tax on interest, dividends, and other financial revenues, which were rapidly proliferating in France at the time but almost totally exempt from taxation, even though similar revenues were taxed in Britain. Once again, however, the tax rate was set quite low (3 percent from 1872 to 1890 and then 4 percent from 1890 to 1914), at any rate in comparison with the rates assessed after 1920. Until World War I, it seems to have been the case in all the developed countries that a tax on income was not considered “reasonable” unless the rate was under 10 percent, no matter how high the taxable income.

  The Progressive Tax in the Third Republic

  Interestingly, this was also true of the progressive inheritance or estate tax, which, along with the progressive income tax, was the second important fiscal innovation of the early
twentieth century. Estate tax rates also remained quite low until 1914 (see Figure 14.2). Once again, the case of France under the Third Republic is emblematic: here was a country that was supposed to nurse a veritable passion for the ideal of equality, in which universal male suffrage was reestablished in 1871, and which nevertheless stubbornly refused for nearly half a century to fully embrace the principle of progressive taxation. Attitudes did not really change until World War I made change inevitable. To be sure, the estate tax instituted by the French Revolution, which remained strictly proportional from 1791 to 1901, was made progressive by the law of February 25, 1901. In reality, however, not much changed: the highest rate was set at 5 percent from 1902 to 1910 and then at 6.5 percent from 1911 to 1914 and applied to only a few dozen fortunes every year. In the eyes of wealthy taxpayers, such rates seemed exorbitant. Many felt that it was a “sacred duty” to ensure that “a son would succeed his father,” thereby perpetuating the family property, and that such straightforward perpetuation should not incur a tax of any kind.19 In reality, however, the low inheritance tax did not prevent estates from being passed on largely intact from one generation to the next. The effective average rate on the top centile of inheritances was no more than 3 percent after the reform of 1901 (compared to 1 percent under the proportional regime in force in the nineteenth century). In hindsight, it is clear that the reform had scarcely any impact on the process of accumulation and hyperconcentration of wealth that was under way at the time, regardless of what contemporaries may have believed.