{ ELEVEN }

Social-Democratic Societies: Incomplete Equality

In the previous chapter we examined how ownership societies that seemed so prosperous and solid on the eve of World War I collapsed between 1914 and 1945. The collapse was so complete that nominally capitalist countries actually turned into social democracies between 1950 and 1980 through a mixture of policies including nationalizations, public education, health and pension reforms, and progressive taxation of the highest incomes and largest fortunes. Despite undeniable successes, however, these social-democratic societies began to run into trouble in the 1980s. Specifically, they proved unable to cope with the rampant inequality that began to develop more or less everywhere around that time.

In this chapter we will focus on the reasons for this failure. In the first place, attempts to institute new forms of power sharing and social ownership of firms remained confined to a small number of countries (especially Germany and Sweden). This avenue of reform was never explored as fully as it might have been, even though it offered one of the most promising responses to the challenge of transcending private property and capitalism. Second, social democracy did not have a good answer to one pressing question: how to provide equal access to education and knowledge, particularly higher education. Finally, we will look at social-democratic thinking about taxation, especially progressive taxation of wealth. Social democracy did not succeed in building new transnational federal forms of shared sovereignty or social and fiscal justice. Today’s globalized economy is one in which regulation in all its forms has been undermined by free trade and free circulation of capital, instituted by agreements to which social democrats consented or even instigated. In any case they had no alternative to offer. The resulting heightened international competition has gravely endangered the social contract (and consent to taxation) on which the social-democratic states of the twentieth century were built.

On the Diversity of European Social Democracies

In the period 1950–1980, the golden age of social democracy, income equality settled at a level noticeably lower than in previous decades in the United States and United Kingdom, France and Germany, Sweden and Japan, and nearly every European and non-European country for which adequate data are available.1 This reduced inequality was due in part to destruction occasioned by war, which hurt those who owned a great deal much more than those who owned nothing. But a much more important reason for the reduction of inequality was a set of fiscal and social policies that made societies not only more egalitarian but also more prosperous than they had ever been before. To all of these societies we may therefore apply the label “social-democratic.”

Let me be clear from the outset: I am using the terms “social-democratic society” and “social democracy” rather broadly to describe a set of political practices and institutions whose purpose was to socially embed (in Polanyi’s sense) private property and capitalism. In the twentieth century, these practices and institutions were adopted by many noncommunist countries both in Europe and elsewhere, some of which explicitly called themselves social-democratic while others did not. In a narrower sense, only Sweden was ruled more or less continuously by an official social-democratic party (the Swedish Social Democratic Party, or SAP) from the early 1930s to the present (with occasional interludes of so-called bourgeois parties in power after the banking crisis of 1991–1992, about which I will say more later). Sweden is thus the quintessential social democracy, the country that conducted the longest experiment with this type of government. The Swedish case is all the more interesting in that Sweden was, prior to the reforms of 1910–1911, one of the most inegalitarian societies in the world, with voting power concentrated in a tiny stratum of the wealthy.² But from 1950 to 2000 it was the country that claimed the largest share of national income as taxes and had the highest social spending in Europe until France caught up with it in the early 2000s. The notion of social democracy I use in this book is best captured by these indicators, which measure the extent of the fiscal and social state.³

In Germany, the Social Democratic Party (SPD), which by the end of the nineteenth century was the leading social-democratic party in Europe in terms of membership, has only been in power intermittently since the end of World War II. Its influence on the development of the German social state was nevertheless considerable, so much so that the Christian Democratic Union (CDU), which held power continuously from 1949 to 1966, adopted the “social market economy” as its official doctrine. In practice, proponents of the social market economy acknowledge the importance of social insurance and accept some degree of power sharing between shareholders and unions. Note, moreover, that the SPD decided at its Bad Godesberg convention in 1959 to drop all references to nationalizations and Marxism. Thus, there was a certain programmatic convergence of the two leading German parties of the postwar era, both of which were searching for a new developmental model that would enable them to rebuild the country after the Nazi catastrophe—a model that one might characterize as “social-democratic.” Nevertheless, substantial differences remained between the SPD and CDU: they disagreed, for example, about the extent and organization of the social welfare and pension system. But both accepted a broad general framework that included high taxes and social spending compared with the pre–World War I period, to which no political party wished to return (in Germany or any other European country). The political landscape therefore resembled Sweden’s, where the “bourgeois” parties never radically challenged the social state created by the SAP even when they came to hold power after 1991. It also resembled the postwar political landscape of other central and northern European countries with powerful social-democratic parties (such as Austria, Denmark, and Norway).

I also apply the term “social-democratic” (in the broad sense) to various other postwar state models such as those of the United Kingdom, France, and other European countries. These countries have parties that call themselves Labour, Socialist, or Communist and do not explicitly claim the “social-democratic” label. In the United Kingdom, the Labour Party has its own distinctive history, with roots in the trade union movement, Fabian socialism, and British parliamentarism.⁴ Labour followed a distinctive political path: it won a large majority in Parliament in 1945, and Clement Attlee’s government proceeded to establish the National Health Service and lay the foundation of the British social state. Despite subsequent challenges, most notably from the Tories led by Margaret Thatcher in the 1980s, Britain’s fiscal and social state remained large in the first two decades of the twenty-first century (with tax revenues of 40 percent of national income, less than the 45–50 percent that one finds in Germany, France, and Sweden but significantly higher than the mere 30 percent in the United States).

In France, the socialist movement split permanently at the Congress of Tours (1920) into a Communist Party (PCF) that supported the Soviet Union and a Socialist Party (PS) that preferred democratic socialism. The two parties shared power with the Radical Party in the Popular Front government elected in 1936.⁵ They later played a central role in establishing la Sécurité sociale, the French health and pension system, which was adopted after the Liberation in 1945. Like other postwar policies, including the nationalization of many firms and the expansion of the role of unions in collective bargaining, wage setting, and workplace organization, the social security system was partly inspired by the 1944 program of the Conseil National de la Résistance. The Socialists and Communists again governed together in 1981 following the victory of the Union of the Left. In the French context, the label “social-democratic” has often been attacked as a synonym for “centrist,” partly owing to the competition (and verbal one-upmanship) between Socialists and Communists. For instance, nationalizations formed the backbone of the left program in 1981, at a time when the German SPD had long since renounced the practice. In France, “social democracy” was often equated with renunciation of any real ambition to transcend capitalism. Be that as it may, the social and fiscal system that France adopted after World War II puts it in the broad family of European social-democratic societies.⁶

The New Deal in the United States: A Bargain-Basement Social Democracy

One might also characterize as “social-democratic” (very broadly speaking) the social system established in the United States in the 1930s under Franklin D. Roosevelt’s New Deal. This was extended by Lyndon B. Johnson’s War on Poverty in the 1960s. Compared with its European counterparts, however, the social-democratic society that the Democratic Party built in the United States was a bargain-basement version of social democracy, for reasons we will need to understand better. Concretely, European levels of taxation and social spending easily eclipsed those of the United States in the period 1950–1980; no such gap had existed in the nineteenth or early twentieth centuries.7 In contrast to what became the postwar European norm, for example, the United States never established universal health insurance. Medicare and Medicaid, which Congress passed in 1965, are reserved for people over 65 and the poor, respectively, leaving uninsured workers not poor enough to qualify for Medicaid and not rich enough to pay for private coverage. To be sure, there has been much discussion of Medicare for All in recent years, and it is not out of the question that such a reform will pass someday.⁸ Since 1935 the Social Security system has provided pensions and unemployment insurance to Americans. Though less generous than similar programs in Europe, these services have been around longer. As we saw in Chapter 10, income and inheritance taxes were more steeply progressive in the United States than in most European countries in the period 1932–1980. It may seem paradoxical that the United States was more egalitarian than Europe in terms of fiscal progressivity yet less ambitious with respect to its social state; we will look closely at this.

There were also many non-European societies that developed social systems comparable to European social democracies in the period 1950–1980; for example, Latin America, and especially Argentina.⁹ It might also be tempting to see many newly independent countries, such as India between 1950 and 1980, as vaguely belonging to the democratic socialist universe. Bear in mind, however, that India, like most countries in southern Asia and Africa, still had fairly low tax revenues (10–20 percent of national income, sometimes even less than 10), and the trend in the 1980s and 1990s was downward (I will come back to this). It is therefore very difficult to compare such countries to European social democracies. In subsequent chapters, moreover, we will study communist and postcommunist societies and their influence on perceptions of the social-democratic state. More generally, in Part Four, we will take a detailed look at the evolution of voting patterns and “social-democratic” coalitions in Europe, the United States, and other parts of the world, which will help us gain a better understanding of the specificities of these various trajectories and political constructs.

On the Limits of Social-Democratic Societies

At this stage, note simply that in most parts of the world, whether it be social-democratic Europe, the United States, India, or China, inequality has increased since 1980, with a strong rise in the top decile’s share of total income and a significant drop in the share of the bottom 50 percent (Fig. 11.1).¹⁰ Within this broad global landscape, it is true that between 1980 and 2018 inequality increased the least in the social-democratic societies of Europe. In this sense, the European social-democratic model seems to offer greater protection than other models (especially the meager American social state) from the inegalitarian pressures of globalization at work since the 1980s. Nevertheless, it is quite clear that a significant change has occurred compared with earlier periods: 1914–1950 saw a historic drop in inequality, while 1950–1980 was a period of stabilization.¹¹ In a context of increasing fiscal and social competition, which European social-democratic governments themselves did much to create and which has created many problems for African, Asian, and Latin American countries seeking to develop viable social models, it is not out of the question that the inegalitarian trend of the post-1980 period may grow stronger in the future. In addition, most of the countries of the Old Continent have had to contend with growing nationalist and anti-immigrant sentiment since 2000. Clearly, European social democracy cannot afford to rest on its laurels.

FIG. 11.1.  Divergence of top and bottom incomes, 1980–2018

Interpretation: The top decile share increased in all parts of the world. It ranged from 27 to 34 percent in 1980 and from 34 to 56 percent in 2018. The share of the bottom 50 percent decreased: it was between 20 and 27 percent and is now between 12 and 21 percent. The divergence of top and bottom incomes is general, but its amplitude varies with the country: it is greater in India and the United States than in China and Europe (EU). Sources and series: piketty.pse.ens.fr/ideology.

Furthermore, the egalitarian character of the period 1950–1980 should not be exaggerated. For example, if we compare the case of France (which is fairly representative of Western Europe) and the United States, we find that the share of national income going to the bottom 50 percent has always been significantly smaller than the share going to the top 10 percent (Fig. 11.2). At the turn of the twentieth century, the top decile claimed 50–55 percent of total income, and the bottom five deciles had gotten about one-quarter of that (around 13 percent of total income). Since the first group is by definition one-fifth the size of the second, this means that the average income of the top decile was twenty times that of the bottom 50 percent. In the 2010s, this ratio was nearly eight: the average income of the top decile in 2015 was 113,000 euros per adult, compared with 15,000 euros for the bottom 50 percent. Clearly, then, social-democratic society may be less unequal than the ownership society of the Belle Époque or than other social models around the world, but it remains a highly hierarchical society in economic and monetary terms. As for the United States, we find that the ratio is close to twenty: nearly 250,000 euros for the top decile compared with barely 13,000 euros for the bottom half. Later we will see that taxes and transfers only slightly improve this situation for the bottom half of the US population today (and that the gap between the United States and Europe is due to the gap prior to taxes and transfers).

FIG. 11.2.  Bottom and top incomes in France and the United States, 1910–2015

Interpretation: Income inequality in the United States in 2010–2015 exceeded its level in 1900–1910, whereas it was reduced in France (and Europe). In both cases, however, inequality remained high: the top decile, one-fifth the size of the bottom 50 percent, still received a much larger share of total income. The incomes shown are average annual incomes for each group in 2015 euros (at purchasing power parity). Sources and series: piketty.pse.ens.fr/ideology.

Public Property, Social Property, Temporary Property

For all these reasons, it is important to take a fresh look at what social-democratic societies have achieved as well as the limits of those achievements. Social-democratic institutions, including the legal system (especially corporate and labor law), the social insurance system, the educational system, and the tax system, were often put in place under emergency conditions (whether in the immediate aftermath of World War II or during the Depression) and never really conceived as a coherent whole. Countries generally relied on their own experience and took little account of the experiences of others. Sharing and mutual learning were sometimes important, as in the case of setting high top rates on progressive income and inheritance taxes, but played a more limited role when it came to setting social policy or designing the legal system.

Our first priority will be to look at the property regime. To simplify, there are three ways of moving beyond private ownership of firms and shareholder omnipotence. The first is public ownership: either the central government, a regional, state, or town government, or an agency under public control can replace private shareholders and take ownership of the firm. The second is social ownership: the firm’s workers participate in its management and share power with private (and possibly public) shareholders, potentially replacing private shareholders entirely. The third is what I propose to call temporary ownership: the wealthiest private owners must return part of what they own to the community every year to facilitate circulation of wealth and reduce the concentration of private property and economic power. This could take the form of, for example, a progressive tax on wealth, which would be used to finance a universal capital endowment for each young adult. We will look more closely at this option later.12

To sum up: public ownership uses state power to balance the power of private property. Social ownership seeks to share power and control of the means of production at the firm level. Temporary ownership allows private property to circulate and prevents the persistence of excessively large holdings.

History suggests that these three ways of transcending private property are complementary. In other words, the key to transcending capitalism permanently is to rely on a mix of public ownership, social ownership, and temporary ownership. Communist societies of the Soviet type sought to rely almost exclusively on public ownership, indeed, on hypercentralized state ownership of nearly all firms and fixed capital—an experiment that ended in abject failure. Social-democratic societies took a more balanced approach, relying to a degree on all three remedies, but their efforts were insufficiently ambitious and systematic, particularly in regard to social and temporary ownership. Nationalization and state ownership were all too often the primary focus of policy, and ultimately even this option was abandoned after the fall of communism, with nothing worthy of the name to replace it. Hence in the end social democrats almost entirely gave up even thinking about moving beyond private property.

More generally, it is important to note that each of these three ways of transcending private ownership comes in many variants, offering endless scope for political, social, and historical experimentation. My intention here is not to close the debate but rather to open it up and reveal its full complexity. For instance, there are many forms of public ownership, some more democratic and participatory than others. What matters is how the corporate governance of public firms is organized. Are users, citizens, and other stakeholders represented on boards of directors? How are administrators appointed by the state or other public entities, and how is their work monitored? Public ownership can be perfectly justifiable, and it has demonstrated its superiority over private ownership in many sectors, including transportation, health, and education, provided that governance is transparent and responsive to the needs of citizens and users. As for temporary ownership and the universal capital endowment, these may require the institution of some new form of progressive wealth tax, with which we have little experience to date. I will come back to this in greater detail later. Finally, social ownership and power sharing between employees and stockholders can also be organized in many ways, some of which have been practiced in a number of European countries since the 1950s. We will start there.

Sharing Powers, Instituting Social Ownership: An Unfinished History

Germany and Sweden, and more generally the social-democratic societies of Germanic and Nordic Europe (especially Austria, Denmark, and Norway), are the countries that have gone furthest in the direction of co-management (from the German Mitbestimmung, sometimes translated as “codetermination”), which is a specific form of social ownership of firms and institutionalized power sharing between workers and shareholders. To be clear, co-management is not an end in itself. We can go beyond it. But we need first to study this important historical experience in order to gain a better understanding of possible next steps.

The German case is particularly interesting in view of the importance of the German social and industrial model for European social democracy.13 A 1951 law made it mandatory for large firms in the coal and steel industries to reserve half the seats (and voting rights) on their boards of directors for representatives of their employees (generally elected from union slates). In concrete terms, this meant that workers on the board could vote on all of the firm’s strategic choices (including nomination and removal of top executives and certification of financial results) and have the same access to the same documents as the directors chosen by the shareholders. In 1952, another law made it mandatory for large firms in other sectors to set aside one-third of their board seats for worker representatives. These two laws, adopted under Christian Democratic Chancellor Konrad Adenauer (1949–1963), also contained extensive provisions concerning the role of factory committees and union delegates in collective bargaining, especially in regard to wage setting, the organization of work, and occupational training.

The laws were further extended when the Social Democrats came to power in Bonn between 1969 and 1982 (under Willy Brandt and Helmut Schmidt). In 1976 an important law on co-management was passed. In its main outlines this law remains unchanged to this day. It requires all firms with more than 2,000 employees to reserve half their board seats (and voting rights) for worker representatives (one-third for firms with between 500 and 2,000 employees). These seats and voting rights are assigned to the worker representatives as such, regardless of worker participation in the firm’s capital. When workers do own shares in the company (either as individuals or through a pension fund or other collective structure), they may hold additional board seats, potentially commanding a majority. The same is true if a local government, the federal state, or some other public body holds a minority share of the stock.¹⁴

It is important to note that this system, which was given legal force by the laws of 1951–1952 and 1976, is above all the result of the very strong mobilization of German unions since the late nineteenth and early twentieth centuries, combined with Germany’s specific historical trajectory. While these rules are widely accepted in Germany today, including by employers, they were strongly contested in the past by German shareholders and owners, who gave in only after intense social and political struggles waged under historical circumstances in which the balance of power between workers and shareholders was a little less skewed than usual. It was in the aftermath of World War I, in the very unusual (and at times insurrectional) climate of the period 1918–1922, that the German workers’ movement succeeded for the first time in negotiating with employers new rights related to factory committees, union delegates, and wage-setting procedures. These were later incorporated in the 1922 law on collective bargaining and worker representation.

It was also under pressure from the unions and the Social Democrats that the Weimar Constitution of 1919 instituted a much more social and instrumental concept of property than any previous constitution. In particular, the Constitution of 1919 specified that property rights and their limits would henceforth be defined by law, which meant that property was no longer considered a sacred natural right. The text explicitly envisioned the possibility of expropriations and nationalizations if “the good of the community” required it under terms set by law. The law also stipulated that land ownership should be organized in relation to explicit social objectives.¹⁵ The German Fundamental Law of 1949 includes similar language, to the effect that property rights are legitimate only insofar as they contribute to the well-being of the community. The text explicitly mentioned socialization of the means of production in terms that opened the way to measures such as co-management.¹⁶ In many countries, the demand for power sharing in firms, and more generally for redefining ownership and redistributing wealth, have encountered the objection that they are unconstitutional and violate property rights said to be absolute and unlimited; the German Fundamental Law makes this objection moot.

After being suspended by the Nazis from 1933 to 1945, the rights granted to unions by the German law of 1922 were reinstated under the Allied occupation. During reconstruction, from 1945 to 1951, the unions, once again in a relatively powerful position, succeeded in negotiating new rights with employers in the steel and energy sectors, including equal representation in the governing instances of firms. These new rights, obtained through negotiation and struggle, were simply incorporated into the 1951 law. It is worth noting that the 1952 law was seen by German trade union federations (especially the Confederation of German Trade Unions [DGB]) as a disappointment, even a step backward.¹⁷ Worker participation in boards of directors (outside the coal and steel industries) was limited to one-third (in practice, two or three seats), whereas the unions were agitating for universal adoption of the principle of equal representation of shareholders and workers. The law also envisioned separate elections for blue- and white-collar workers, which in the eyes of the unions was tantamount to dividing the firm’s employees and weakening their voice.

Successes and Limitations of German Co-Management

Broadly speaking, it is important to emphasize that one of the main limitations of German co-management is that worker-shareholder parity is in some ways a trap unless workers or the state also own shares in the company. With parity, the directors chosen by the shareholders hold the decisive vote when it comes to choosing the firm’s top executives or deciding on its investment or recruitment strategy. This decisive vote is cast by the chairman of the board, who is always a representative of the shareholders. Another key point to bear in mind is that most German firms are governed not by a single board of directors (as is the case in most other countries) but by a two-headed structure consisting of an oversight committee and a directorate. Worker representatives hold half the seats on the oversight committee, but the shareholders, who have the decisive vote, can name as many members of the directorate as they wish, and this is the operational leadership of the firm. A recurrent demand of German unions, who remain dissatisfied with the system to this day, is for parity in the directorate as well: in other words, worker representatives should be allowed to choose half of the company’s management team and not just the personnel manager or director of human resources (a post often filled by a union representative in large German firms, which already marks a significant departure from standard practice in other countries). These debates show that social ownership and co-management, as currently embodied, should not be regarded as finished solutions. On the contrary, they are still largely unfinished projects, history in progress, whose logic has not been pursued all the way to the end.

In the case of Sweden, the law of 1974, extended in 1980 and 1987, reserves a third of board seats for workers in firms with twenty-five or more employees.18 Since Swedish firms are governed by a single board of directors, this representation, though a minority, sometimes results in more effective operational control than the German parity in oversight committees (which are farther removed from effective management of the firm). The Swedish rules also apply to much smaller firms than the German rules, which are applicable only to firms with 500 or more employees (very restrictive in practice). In Denmark and Norway, workers are entitled to a third of board seats in firms of more than thirty-five and fifty employees respectively.¹⁹ In Austria, the proportion is also one-third, but the rule applies only to firms with more than 300 employees, which considerably limits the scope of its application (almost as much as in Germany).

Regardless of the limitations of German and Nordic co-management as it has been practiced since the end of World War II, all signs are that the new rules have somewhat shifted the balance of power between shareholders and employees and encouraged more harmonious and ultimately more efficient economic development (at least in comparison with firms in which workers enjoy no board representation). In particular, the fact that the unions help to define the firm’s long-term strategy and are given access to all the documents and information they need for that purpose leads to greater employee involvement in the firm and thus to higher productivity. The presence of workers on boards of directors has also helped to limit wage inequality and in particular to control the vertiginous growth of executive pay seen in some other countries. Specifically, in the 1980s and 1990s, executives in German, Swedish, and Danish firms had to make do with far less fabulous raises than their English and US counterparts, yet this did not harm their firms’ productivity or competitiveness—quite the contrary.²⁰

The criticism that a minority presence of workers on boards of directors simply leads to ratification of decisions taken unilaterally by shareholders and therefore reduces union combativeness also appears to be unjustified. To be sure, the co-management system needs to be improved and surpassed. Nevertheless, all countries where it has been introduced have also established collective bargaining systems affording workers representation through factory committees, union delegates, and other entities composed solely of workers and responsible for negotiating directly with management over working conditions and wages (regardless of whether the managers have been approved by boards with worker membership). In Sweden, after the Social Democrats came to power in the 1930s, the unions were quick to take advantage of these entities for capital-labor negotiations. Similar institutions made it possible to develop a true worker “status” with a guaranteed wage income (generally in the form of a monthly wage instead of work paid by the task or the day, as in the nineteenth century) and protection against unjustified dismissal (which also encouraged workers to identify with the long-term interests of their firms) in nearly all developed countries, even where workers were not represented on company boards.²¹ But obtaining board seats offered an additional channel of influence. This is particularly true at times of industrial and union decline and is part of the reason why the German and Nordic social and economic model has been more resilient since the 1980s.²² To sum up, co-management has been one of the most highly developed and durable means of institutionalizing the new balance of power between workers and capital. It came into being in the mid-twentieth century as the culmination of a very long process involving union struggles, worker militancy, and political battles, which dated back to the middle of the nineteenth century.²³

On the Slow Diffusion of German and Nordic Co-Management

To recapitulate: In the Germanic and Nordic countries (notably Germany, Austria, Sweden, Denmark, and Norway) worker representatives fill between a third and a half of the seats on the boards of directors of the largest firms whether or not they own any part of the firm’s capital. In Germany, which led the way on these matters, this system has been in place since the early 1950s. Despite the widely acknowledged success of the German and Nordic social and industrial model, which is noted for producing a high standard of living, high productivity, and moderate inequality, other countries until recently had not followed suit. In the United Kingdom, United States, France, Italy, Spain, Japan, Canada, and Australia, private firms continue to be governed by immutable corporate bylaws: in all these countries, a general assembly of shareholders continues to elect the entire board of directors according to the principle “one share, one vote,” with no representation for employees (except in a few cases that have a merely consultative representation, without voting rights).

Things began to change slightly in 2013, when France passed a law requiring firms with more than 5,000 employees to set aside one board seat out of twelve for a worker representative. This new French rule was nevertheless quite limited compared to the German and Nordic systems (limited in terms of both the number of worker representatives and the scope of firms covered).24 Of course, it is not out of the question that coverage will be increased in the coming decade, not only in France but also in the United Kingdom and United States, where some fairly ambitious and innovative proposals have recently been discussed by Labour and Democratic politicians respectively. If France, Britain, and the United States were to move more decisively in this direction, it is possible that this would lead to a more global diffusion of the model. Nevertheless, as of 2019, if one excepts the meager single board seat introduced in France in 2013, power-sharing and co-management arrangements remain confined to the Germanic and Nordic countries. Co-management is a trademark of Rhenish and Scandinavian capitalism, not of Anglo-American capitalism (or French, Latin, or Japanese capitalism). How can we explain such slow and limited diffusion of the co-management model compared with the rapid and widespread diffusion of large-scale progressive taxation after World War I?

The first explanation is that the decision to give workers voting rights without any corresponding participation in the firm’s capital constituted a fairly radical conceptual challenge to the very idea of private property, which shareholders and owners have always strenuously opposed. It is easy, even for parties with a relatively conservative economic outlook, to defend a certain theoretical diffusion of ownership. For instance, the French Gaullists promoted the idea of “participation” (in the double sense of employee share ownership and potential profit sharing, but without voting rights). Conservatives in Britain and Republicans in the United States have regularly championed the idea of employee stock ownership; this idea was floated in the 1980s, for example, when Thatcher privatized publicly owned firms. But to change the rules linking ownership of capital to the power to decide what use to make of one’s property (a power taken to be absolute in classic definitions of property) and to create voting rights for people who own nothing—these are from a conceptual standpoint highly destabilizing actions, even more so (arguably) than progressive taxation. In Germany and the Nordic countries, such a drastic revision of corporate law and the law of property was possible only in very specific historic circumstances, characterized by unusually strong mobilization of trade unions and social-democratic parties.

The second explanation, which complements the first, is precisely that the political and social forces in other countries did not have the same determination as in Germany for reasons related to the political-ideological trajectory of each country. In France, it is often thought that the enduring socialist preference for nationalizations (which, for example, formed the centerpiece of the Union of the Left program in the 1970s) and the lack of appetite for co-management stem from the supposed statist ideology of French Socialism and its weak ties to the union movement. It is indeed striking that no measure to set aside board seats for worker representatives was proposed between 1981 and 1986, when the Socialists had an absolute majority in the National Assembly. The role of union delegates in negotiating wages and working conditions was expanded, and certain steps were taken to promote decentralization and participation in other sectors (such as increasing the autonomy of local governments), but the link between shareholding and decision-making power within firms was not touched. By contrast, the sweeping program of nationalizations in 1982 sought to complete the nationalizations of the Liberation by incorporating nearly the entire banking sector and major industrial conglomerates in the public sector, which meant appointing directors chosen by the government in place of directors elected by shareholders. In other words, French Socialists believed that the state and its high civil servants were perfectly capable of taking over the boards of directors of all key industries but that worker representatives had no place among them.

Then, in 1986–1988, the Gaullist and liberal parties returned to power in a new context of privatization and deregulation under Thatcher and Reagan, while at the same time the Communist bloc was slowly crumbling. This led to the privatization of most of the companies that had been nationalized between 1945 and 1982. The privatization movement continued, moreover, in the legislatures of 1988–1993, 1997–2002, and 2012–2017, during which the Socialists were in power, yet still no Germano-Nordic-style co-management was ever attempted, apart from the timid and belated law of 2013.²⁵ French Socialists and Communists might also have pushed from co-management in 1945–1946, but they chose instead to concentrate on other battles, including nationalizations and social security, for example.

It is not clear, however, whether the lack of appetite for co-management can be attributed to the weakness of French trade unionism. True, the workers’ movement in France was less powerful and less organized than in Germany or the United Kingdom and less closely tied to French political parties.²⁶ Still, the unions and social mobilizations did play an important role in French political history (especially in 1936, 1945, 1968, 1981, 1995, and 2006). Furthermore, Germano-Nordic co-management did not spread to the United Kingdom either, even though the Labour Party has from its inception been structurally tied to powerful British trade union movement. The more likely explanation for the shared British and French aversion to co-management is that both French Socialists and British Labourites long believed that nationalization and state ownership of large firms was the only way to truly alter the balance of power and move beyond capitalism. This is obvious in the French case (as the Common Program of 1981 indicates), but it is just as obvious for the United Kingdom. The famous Clause IV of the Labour Party’s constitution of 1918 set “common ownership of the means of production” as the party’s central goal (or so it was interpreted). As recently as the 1980s Labour platforms were still promising further nationalizations and indefinite extension of the public sector, until New Labour under Tony Blair finally succeeded in 1995 in eliminating any reference to the property regime from Clause IV.²⁷

Socialists, Labourites, Social Democrats: Intersecting Trajectories

From this point of view, it was the SPD that was the exception. Although the French and British parties waited until the fall of the Soviet Union in 1989–1991 to renounce nationalizations as a central tenet of their programs, the German Social Democrats had already endorsed co-management in the early 1950s and abandoned nationalizations at Bad Godesberg in 1959. In the interwar years things were different: nationalizations were at the heart of the SPD program in the 1920s and 1930s, and, like its French and British counterparts, the party showed little interest in co-management.28 If things changed in 1945–1950, it was because of Germany’s unique political-ideological trajectory. Not only had the very bitter clashes between the SPD and the Communist Party of Germany (KPD) in the interwar years left deep traces,²⁹ but the West-German Social Democrats had every reason in the 1950s to wish to set themselves apart from the Communists in the East and the idea of state ownership. The traumatic experience of hypertrophied state power under the Nazis no doubt also contributed to discrediting nationalizations and state ownership in the eyes of the SPD and the German public, or at the very least to enhancing the appeal of co-management as a solution.³⁰

In any case, it is interesting to note that the abandonment of any reference to nationalizations in the 1990s did not lead either the French Socialists or the British Labour Party to embrace the co-management agenda. In the period 1990–2010 neither party exhibited the slightest desire to transform the property regime. Private capitalism and the “one share, one vote” principle appeared to have become unsurpassable horizons, at least for the time being. Both parties contributed to this state of mind by privatizing some state enterprises and by supporting the free flow of capital and the race to cut taxes.³¹ In the French case, the fact that co-management ultimately resurfaced in the timid law of 2013 owed a great deal to the demands of certain unions (especially the French Democratic Confederation of Labor [CFDT]) and above all to the increasingly obvious success of the German industrial sector. In the late 2000s and early 2010s, when references to Germany and its economic model were ubiquitous, partly for good reasons, it became increasingly difficult for French employers and shareholders to reject co-management out of hand and to insist that the presence of workers on corporate boards would sow chaos.³² The timid advance of 2013—timid by comparison with decades-old German and Nordic practices—tells us a great deal about the political and ideological resistance at work, as well as about the often quite national character of the process of policy experimentation and learning.

In the British case, the need for new approaches in the fight against rising inequality, coupled with the change in the Labour Party’s leadership in 2015 (partly because of dissatisfaction with the Blairite line and the country’s inegalitarian drift), has contributed in recent years to the development of a new political approach. The party is more open to nationalizations (public enterprises are now thought to be desirable in some sectors such as transportation and water supply, reflecting a new pragmatism compared with the preceding era), a new system of labor law, and new forms of corporate governance. The growing popularity of worker representation on boards of directors, an idea that has also been canvased among previously skeptical Democrats in the United States as well as certain British Conservatives, can be explained by the fact that co-management is a social measure that costs the public treasury nothing—a particularly valued quality in these days of growing inequality and rising deficits. For all these reasons, good and bad, it is likely that these issues will continue to be debated in the coming years, although it is impossible at this stage to say when change might occur.

From a European Directive on Co-Management to Proposition “2x + y”

Before we turn to these new prospects, however, it is important to emphasize that the various political-ideological trajectories I have just rehearsed are simply the ones that actually came to pass. Many other paths might have been taken, because the history of property regimes, like the history of inequality regimes in general, contains numerous switch points and should not be seen as linear or deterministic.

One particularly interesting case involves the so-called 2x + y proposal discussed in the United Kingdom in 1977–1978. In 1975 Labour Prime Minister Harold Wilson commissioned a report from a commission chaired by historian Allan Bullock and composed of jurists, trade unionists, and employers. The commission’s conclusions were submitted in 1977. The inquiry was a response to a request from the European Commission, which, under pressure from Germany, was considering a European directive on corporate governance. A draft published by the Brussels authorities in 1972 proposed that all firms with more than 500 employees should have at least one-third of their directors representing workers. Revised drafts were published in 1983 and 1988, but in the end the whole project was abandoned for want of a majority of European countries willing to vote for it.33 I will say more later about how EU rules make it almost impossible to adopt common policies of this type (for reforms of the fiscal and social system as well as the legal system); only a profound democratization of EU institutions can change this. It is nevertheless interesting that a proposal for a European model of power sharing between workers and shareholders did reach a relatively advanced stage in the 1970s and 1980s.

In any case, the Bullock Commission proposed in 1977 that the Labour government adopt the so-called 2x + y system.³⁴ Concretely, in every firm with more than 2,000 employees, shareholders and workers were both to elect a number x of board members, and the government would then top off the board by naming y independent directors, who would cast the decisive votes in case of a stalemate between shareholder and worker representatives. For example, a board of directors might consist of five shareholder representatives, five worker representatives, and two representatives of the government. The numbers x and y could be set by the firm’s bylaws, but the latter could not affect the overall structure or the fact that the board of directors alone had the right to make the most important decisions (such as naming the firm’s executives, approving its financial reports, distributing dividends, and so on). Unsurprisingly, shareholders and the City of London’s financial community outspokenly opposed the proposal, which radically challenged the usual assumptions of private capitalism, potentially going much farther than German or Swedish co-management. By contrast, there was strong support from the unions and the Labour Party, with no compromise in sight.³⁵ In the fall of 1978, James Callaghan, the new Labour prime minister who replaced Wilson in 1976, seriously contemplated calling a snap election at a time when the polls were predicting a Labour victory. In the end, he decided to wait another year. The country was immobilized by numerous strikes during the “Winter of Discontent” (1978–1979) in a period of high inflation. The Tories, led by Margaret Thatcher, won the election in 1979, and the project was definitively buried.

Beyond Co-Management: Rethinking Social Ownership and Power Sharing

In Part Four I will return to the question of how one might move toward a new form of participatory socialism in the twenty-first century, drawing on the lessons of history and, in particular, combing elements of social and temporary ownership.³⁶ At this stage, I simply want to indicate that social ownership—that is, power sharing within the firm—can potentially take forms other than German or Nordic-style co-management. This history is far from over, as any number of recent proposals and debates suggest.

Broadly speaking, one key issue concerns the extent to which it is possible to overcome the automatic majority that shareholders enjoy under the German system of co-management. The Bullock Commission’s 2x + y proposal is one answer to this question, by assigning a major role to the state. This might work with very large firms (where it would be tantamount to making local and national governments minority shareholders), but it might be problematic to apply such a system to hundreds of thousands of small and medium firms.³⁷ One important limitation of the German system is that it applies only to large firms (with more than 500 employees), whereas Nordic co-management applies much more broadly (to firms with more than thirty, thirty-five, or fifty salaries depending on the case). Since the majority of workers work for small firms, it is essential to find solutions applicable to companies of all sizes.³⁸

As a complement to ideas like “2x + y,” one might also want to encourage employee shareholding, which could add seats to those already held by workers without shares, opening the way to worker majorities. Several Democratic senators proposed bills in 2018 to require US firms to set aside 30–40 percent of their board seats for worker representatives.³⁹ Passage of such a law would be revolutionary in the United States, where nothing of the kind has ever existed. There is a certain tradition of employee stock ownership, however, although the influence of the patrimonial middle class has decreased in recent decades as the concentration of wealth has skyrocketed. Fiscal policies less favorable to the highest earners and wealthiest individuals, together with incentives for employee stock ownership, could change this.⁴⁰ Proposals like the one I will discuss later (for a progressive wealth tax coupled with a universal capital endowment) could also create new majorities, alter the balance of power, and equalize participation in the economy. Still, the movement to set aside board seats for workers has not gotten very far in the United States, where it does little good to point to the success of co-management in Germany or Scandinavia or, for that matter, anywhere outside the United States. It might help, however, to call attention to the fact that there is an old (and largely forgotten) Anglo-American tradition of limiting the power of large shareholders: in the early nineteenth century, British and American companies often placed limits on the voting rights of large stockholders.⁴¹

Recent British debates have also suggested new ways of moving beyond existing co-management models. In 2016, for example, a collective of jurists published a “Labor Law Manifesto,” which was partly incorporated into the official platform of the Labour Party. The goal was to revise large parts of labor and corporate law to encourage greater worker participation and improve working conditions and pay while enhancing social and economic efficiency. The manifesto proposed that workers immediately be given a minimum of two board seats (typically 20 percent of the total). The most original proposal was that board members should be elected by a mixed assembly of shareholders and workers.⁴² In other words, workers should be considered members of the firm on the same footing as shareholders—that is, as actors involved in its long-term development. As such, they would enjoy voting rights in a mixed assembly responsible for choosing the firm’s board. Initially, workers would be given 20 percent of the voting rights in this assembly, but this would gradually be increased (possibly to 50 percent or more). These rules would apply, moreover, to all firms, regardless of size, including the smallest; in this respect the manifesto departed from the experience of other countries and offered the potential of involving all workers, not just the employees of large firms.

One virtue of such a system, according to the authors, is that it would oblige would-be directors to address the concerns of both workers and shareholders. Rather than represent solely the interests of one group or the other, directors elected by such a mixed assembly would have to present long-term strategies based on the aspirations and understandings of both. If workers were also shareholders, either individually or through some collective entity such as a pension fund, new dynamics might emerge.⁴³

Cooperatives and Self-Management: Capital, Power, and Voting Rights

Mention should also be made of ongoing reflection on the governance of cooperatives and nonprofit organizations such as associations and foundations, which play a central role in many sectors, including education, health, culture, universities, and media. One of the main limits on the development of cooperatives has been excessive structural rigidity. In the classic cooperative, each member has one vote. This structure is perfectly appropriate for certain types of project, in which each participant does the same amount of work and contributes the same amount of resources. Historically, cooperatives have also demonstrated their ability to manage natural resources in an egalitarian way.44

This structure can lead to complications in many situations, however: for example, when investors in a new venture contribute different amounts to the project. This can be a problem for both large and small ventures. Take a person who wants to open a restaurant or an organic food store and has $50,000 to invest. Suppose the business has three employees: the founder and two other people she recruits to work with her but who contribute no capital. With a strictly egalitarian cooperative structure, each worker would have one vote. The two new hires, who may have joined the business the week before or may be thinking of leaving to start their own businesses the following week, can outvote the founder on all sorts of matters, even though she invested all her savings and may have been dreaming about the business for years. Such a structure might be appropriate in some situations, but to impose it in every case would be neither just nor efficient. Individual aspirations and career paths vary widely, and any power-sharing arrangement must take this diversity into account rather than stifle it. In Chapter 12 I will say more about this important topic in connection with communist and postcommunist societies.

More generally, for projects involving more workers or a more diversified capital structure, there is nothing wrong with giving more votes to individuals who supply more capital, provided that workers are also represented in decision-making bodies (perhaps through representatives chosen according to the rules of the German co-management model or perhaps through a mixed assembly of workers and shareholders) and provided that everything possible is done to reduce inequalities of wealth and to equalize access to economic and social life. One can also set a ceiling on the number of votes that any one stockholder can cast or create several different classes of voting rights.⁴⁵

For example, it was recently proposed to create a class of “nonprofit media companies,” with a ceiling on the voting rights of the largest donors and corresponding extra voting rights for smaller donors (such as journalists, readers, crowdfunders, and so on). For instance, one might decide that only a third of individual contributions above 10 percent of the firm’s total capital should be granted voting rights.⁴⁶ The idea is that it might make sense to give more votes to a journalist or reader who invests $10,000 rather than $100, but it is best to avoid giving all the power to a deep-pocketed investor who invests $10 million to “save” the paper. A firm of this type would be between a traditional joint-stock company, based on the “one share, one vote” principle, and a foundation, association, or other nonprofit to which contributions do not give rise to voting rights (at least not directly).

Initially conceived for the media sector and for a setting in which financial contributions take the form of (nonrecoverable) gifts, a model of this kind might work well for cooperatives in other sectors and might also be applicable in cases where contributions of capital were recoverable. In general, there is no reason to restrict oneself to a choice between a pure cooperative model (one person, one vote) and a pure shareholder model (one share, one vote). The important point is that one needs to experiment with new mixed forms on a large scale. In the past, the idea of worker-managed firms aroused high hopes, for example, in France in the 1970s (where the watchword was autogestion). But many projects did not get much beyond the slogan stage and led nowhere for want of concrete plans.⁴⁷ Any discussion of new enterprise structures must include plans for amending the way nonprofit ventures are taxed. In most countries, tax benefits for giving mostly favor the rich, whose preferences in charity, culture, arts, education, and sometimes politics are de facto subsidized by less well-to-do taxpayers. In Part Four I will say more about how the tax system can be changed to encourage more democratic and participatory outcomes by allowing each citizen to give the same amount to nonprofit ventures of his or her choosing, possibly including gifts to sectors not previously exempt from taxation (such as the media or ventures in sustainable development).

To recapitulate: In the nineteenth century and until World War I, the dominant ideology sacralized private property and owners’ rights. Then, from 1917 to 1991, new thinking about the forms of property was blocked by the bipolar opposition of Soviet Communism and American capitalism. One was either for unlimited state ownership or for full private shareholder ownership. This helps to explain why alternatives such as co-management and self-management were not explored as fully as they might have been. The fall of the Soviet Union inaugurated a new period of unlimited faith in private property from which we have not yet completely emerged but which is beginning to show serious signs of exhaustion. Just because Soviet Communism was a disaster does not mean that we should stop thinking about property and how it might be transcended. The concrete forms of property and power are constantly being reinvented. It is time to take a fresh look at this history, starting with the German and Nordic experiments with co-management, and to ask how these might be generalized and extended to viable, innovative, and participatory forms of self-management.

Social Democracy, Education, and the End of US Primacy

We come now to one of the principal challenges that social-democratic societies must face today, namely, the issue of access to skills and training, especially higher education. Property is important, but education has also played a central role in the history of inequality regimes and the evolution of social and economic inequalities both within and between countries. Two points deserve particular attention. First, throughout much of the twentieth century, the United States has held a significant lead in education over Western Europe and the rest of the world. This US advantage dates back to the early nineteenth century and beyond, and it explains much of the large gap in productivity and standard of living that one observes through most of the twentieth century. In the late twentieth century, the United States lost this lead and witnessed the appearance of a new stratification with respect to education: significant gaps in educational investment separated the lower and middle classes from those with access to the most richly endowed universities. Looking beyond the United States, I will stress the fact that no country has responded in a fully satisfactory way to the challenge of transitioning from the first educational revolution to the second—that is, from the revolution in primary and secondary education to the revolution in tertiary education. This failure is part of the reason why inequality has risen since 1980 and why the social-democratic model (and the electoral coalition that made it possible) seems to have run its course.

FIG. 11.3.  Labor productivity, 1950–2015 (2015 euros)

Interpretation: Labor productivity, measured by GDP per hour worked (in constant 2015 euros at purchasing power parity), rose from 8 euros in Germany and France in 1950 to 55 euros in 2015. Germany and France caught up (or slightly passed) the United States in 1985–1990, whereas the United Kingdom remains 20 percent lower. Sources and series: piketty.pse.ens.fr/ideology.

Let us begin with American primacy. In the early 1950s, labor productivity in Germany and France was barely 50 percent of the US level. In the United Kingdom it was less than 60 percent. Then Germany and France surpassed the United Kingdom in the 1960s and 1970s and ultimately caught up with the United States in the 1980s. German and French productivity subsequently stabilized at roughly the same level as the United States after 1990, while British productivity stagnated at a level 20 percent lower (Figs. 11.3–11.4).

These graphs call for several remarks. First, the productivity measures shown in Figs. 11.3 and 11.4, namely gross domestic product (GDP) divided by total hours worked, are far from completely satisfactory. The very notion of “productivity” is problematic and calls for further discussion. The word might seem to convey an injunction to produce more and more forever and ever, which makes no sense if the result is to make the planet unlivable. Hence instead of reasoning in terms of GDP, it would be far better to use net domestic product—deducting for depreciation and damage to capital, including natural capital—but currently available national accounts do not allow us to do this. Although this does not affect the comparisons we focus on here, its importance for analyzing inequality in the global economy of the twenty-first century remains fundamental.⁴⁸

FIG. 11.4.  Labor productivity in Europe and the United States

Interpretation: Labor productivity, measured by GDP per hour worked (in constant 2015 euros at purchasing power parity), was half of US productivity in Western Europe in 1950. Germany and France caught up (or slightly surpassed) the United States in 1985–1990, while the United Kingdom remained 20 percent lower. Sources and series: piketty.pse.ens.fr/ideology.

Second, it is fairly complex to measure in a reliable and comparable way the number of hours worked in different countries. Since the 1960s there have been of course many surveys that allow us to estimate hours worked per week, vacation time, and other essential data. But these surveys are seldom completely consistent across time and space, and they are far less numerous and comprehensive for years before 1960. In this book I have used data on hours worked compiled by international statistical agencies. These are the best estimates we have, but their accuracy should not be exaggerated. The main fact to bear in mind (which is reasonably well documented) is that the number of hours worked per job was approximately the same in Western Europe and the United States until the early 1970s (1900–2000 hours per year per job); however, a significant gap opened up in the 1980s. By the mid-2010s, the number of hours worked per job per year was 1,400–1,500 in Germany and France; 1,700 in the United Kingdom; and nearly 1,800 in the United States. These differences reflect both the shorter work week and longer vacations in Germany and France.49

Note that working time has decreased over the long run (including in the United Kingdom and to a lesser extent in the United States), which seems logical. As productivity rises, it is natural to work fewer hours to spend more time with family and friends, to discover the world and other people, and to seek entertainment and culture. It may be that this is indeed the goal of technological and economic progress and that the objective of improving the quality of life is better served by the trajectories we see in Germany and France than by those of Britain and the United States. What is the ideal rate of reduction of work time? What is the best way to organize work? These are extremely difficult questions to answer, and I do not intend to do so here. The downward trend in working hours is an eminently political question, which always involves social conflict and ideological change.⁵⁰ Note simply that in the absence of national legislation or collective bargaining for the entire work force, or at least the work force of an entire sector of the economy, it is historically extremely rare to see major reductions of working hours.⁵¹

Note, finally, that the notion of productivity used here, though highly imperfect and unsatisfactory, is more subtle than a simple market-based notion of productivity. In particular, the productivity of the government and nonprofit sectors is taken into account because their “output” is reflected in GDP through production costs; this is equivalent to assuming that the “value” society assigns to teachers, doctors, and so on is equal to the amount of taxes, subsidies, and contributions required to pay for their services. This probably results in underestimation of GDP in countries with an extensive public sector, but the bias is smaller than if the nonmarket sector were simply ignored.

The United States: An Early Leader in Primary and Secondary Education

Returning to the American lead in productivity and its slow reduction after 1950 (Figs. 11.3–11.4), note first that Europe’s low productivity level compared with that of the United States actually dates back to a time well before the middle of the twentieth century. The gap was certainly aggravated by destruction and disruption of Europe’s productive apparatus in two world wars, but the important fact is that it was already fairly large in the late nineteenth and early twentieth centuries. In France and Germany, GDP per capita or per job was 60–70 percent of the US level in 1900–1910. The gap with the United Kingdom was smaller, around 80–90 percent. But the fact is that Britain—which had enjoyed the highest productivity in the world through most of the nineteenth century thanks to the lead established in the first Industrial Revolution (owing largely to British domination of the global textile industry)—had clearly fallen behind the United States by the first decade of the twentieth century, having lost ground at an accelerating rate over the decades prior to World War I.

The evidence suggests that these old, persistent, and growing (at least until the 1950s) productivity gaps were due in large part to America’s historic advance in training its workers. At the beginning of the nineteenth century, the US population was small compared with the populations of Europe, but a larger proportion of Americans went to school. The data we have, mostly taken from census reports, indicates that the primary schooling rate (defined as the percentage of children ages 5 to 11, both male and female, attending primary school) was nearly 50 percent in the 1820s, 70 percent in the 1840s, and more than 80 percent in the 1850s. If we exclude the black population, the primary schooling rate for whites was more than 90 percent by the 1840s. At the same time, the comparable rate was 20–30 percent in the United Kingdom, France, and Germany. In all three countries it was not until the period 1890–1910 that we find the near-universal primary education that the United States had achieved half a century earlier.⁵² America’s educational advance is explained in part by its Protestant religious roots (Sweden and Denmark were not far behind the United States in the first half of the nineteenth century) but also by more specific factors. Germany was slightly ahead of France and the United Kingdom in primary schooling in the mid-nineteenth century but far behind the United States. Another reason for the American lead was a phenomenon we see today among migrants. Individuals in a position to emigrate to the United States in the eighteenth and nineteenth centuries were on average better educated and more inclined to invest in the education of their children than the average European of the time, even controlling for geographic and religious origins.

America’s lead in education, which is very clear at the primary level in the period 1820–1850, coincided with a much more rapid expansion of male suffrage. Alexis de Tocqueville already noticed the connection in 1835: for him, the diffusion of education and landownership were the two fundamental forces responsible for the flourishing of the “democratic spirit” in the United States.⁵³ In fact, we find that the rate of participation of adult white males in US presidential elections rose from 26 percent in 1824 to 55 percent in 1832 to 74 percent in 1844.⁵⁴ Of course, women and African Americans continued to be denied the right to vote (until 1965 for many African Americans). Nevertheless, one had to wait until the end of the nineteenth century or in some cases the beginning of the twentieth to see similar extension of male suffrage in Europe.⁵⁵ Participation in local elections progressed at the same pace, which in turn contributed to greater public support for financing public schools through local taxes.

The key point here is that America’s educational lead would continue through much of the twentieth century. In 1900–1910, when Europeans were just reaching the point of universal primary schooling, the United States was already well on the way to generalized secondary education.⁵⁶ In fact, rates of secondary schooling, defined as the percentage of children ages 12–17 (boys and girls) attending secondary schools, reached 30 percent in 1920, 40–50 percent in the 1930s, and nearly 80 percent in the late 1950s and early 1960s. In other words, by the end of World War II, the United States had come close to universal secondary education. At the same time, the secondary schooling rate was just 20–30 percent in the United Kingdom and France and 40 percent in Germany. In all three countries it is not until the 1980s that one finds secondary schooling rates of 80 percent, which the United States had achieved in the early 1960s. In Japan, by contrast, the catch-up was more rapid: the secondary schooling rate attained 60 percent in the 1950s and climbed above 80 percent in the late 1960s and early 1970s.⁵⁷

Interestingly, voices began to be raised in Europe in the late nineteenth century, especially in the United Kingdom and France, about the lack of investment in education. Many people had begun to see that the world domination of the two colonial powers was fragile. There was of course an obvious moral and civilizational purpose in broadening access to education, but beyond that there was a relatively new idea in the air that skills would play a central role in future economic prosperity. In retrospect, it is clear that the second Industrial Revolution, which took place gradually between 1880 and 1940 with the rise of the chemical and steel industries, automobile manufacturing, household appliances, and so on, was much more demanding in terms of skills than the first. In the first Industrial Revolution, concentrated in coal and textiles, it was enough to mobilize a relatively unskilled work force, which could be overseen by foremen and a small number of entrepreneurs and engineers familiar with the new machines and production processes. Crucially, the whole system relied on the capitalist, colonialist state to organize the flow of raw materials and the global division of labor.⁵⁸ In the second Industrial Revolution it became essential for growing numbers of workers to be able to read and write and participate in production processes that required basic scientific knowledge, the ability to understand technical manuals, and so on. That is how, in the period 1880–1960—first the United States and then Germany and Japan, newcomers to the international scene—gradually took the lead over the United Kingdom and France in the new industrial sectors.

In the late nineteenth and early twentieth centuries, the United Kingdom and France were too confident of their lead and their superior power to take the full measure of the new educational challenge. In France, the trauma of military defeat at the hands of Prussia in 1870–1871 played a decisive role in accelerating the process. In the 1880s the Third Republic passed laws making schooling compulsory and centralizing financing of the primary schools, which had a definite positive effect on primary schooling rates. But it was relatively late in the day, coming as it did after a long period of slow progress in literacy and primary schooling rates, which began in the eighteenth century and gradually accelerated in the nineteenth.⁵⁹

In the United Kingdom, worry about the lack of educational investment began to manifest itself in the middle of the nineteenth century. The country’s political and economic elites remained unconcerned, however, as they were convinced that British prosperity depended above all on the accumulation of industrial and financial capital and on the solidity of proprietarian institutions. Recent work has shown that the results of the British census of 1851 were manipulated to minimize the educational gulf that was opening between the United Kingdom and other countries, especially the United States and Germany. In 1861, an official parliamentary report proudly announced that nearly all children under the age of 11 were in school, but it was contradicted a few years later by a field survey that found that only half of those children were in fact attending classes.⁶⁰

Minds began to change after the North defeated the South in the US Civil War. British and French elites interpreted it as the triumph of educational superiority, just as they would later interpret Prussia’s victory over France in 1871. Nevertheless, budget statistics show that educational investment in the United Kingdom continued to lag until World War I. In 1870, public expenditure on education (at all levels) represented more than 0.7 percent of national income in the United States compared with less than 0.4 percent in France and 0.2 percent in the United Kingdom. In 1910, the comparable figures were 1.4 percent for the United States compared with 1 percent for France and 0.7 percent for the United Kingdom.⁶¹ By comparison, recall that from 1815 to 1914 the United Kingdom spent 2–3 percent of national income year in and year out to serve the interests of its sovereign bondholders, which illustrates the gap between the importance assigned to proprietarian ideology versus that ascribed to education. Recall, too, that public expenditure on education was close to 6 percent of national income in the major European countries in the period 1980–2020.⁶² This shows how much things changed over the course of the twentieth century as well as the potential for divergence between countries and for inequality between social groups within an overall pattern of rising investment in education. The British system in particular remains one of strong social and educational stratification, with stark differences between lavishly endowed private schools and garden-variety public schools and high schools—differences that explain some of Britain’s lag in productivity despite additional school spending since the late 1990s.⁶³

US Lower Classes Left Behind Since 1980

How did the United States, which pioneered universal access to primary and secondary education and which, until the turn of the twentieth century, was significantly more egalitarian than Europe in terms of income and wealth distribution, become the most inegalitarian country in the developed world after 1980—to the point where the very foundations of its previous success are now in danger? We will discover that the country’s educational trajectory—most notably the fact that its entry into the era of higher education was accompanied by a particularly extreme form of educational stratification—played a central role in this change.

Care should be taken not to overstate the importance of a country’s egalitarian roots. The United States has always entertained an ambiguous relationship with equality: more egalitarian than Europe in some respects but much more inegalitarian in others, especially owing to its association with slavery. As noted earlier, moreover, American “social democracy” can trace its ideological origins to a form of social nativism: the Democratic Party was long the segregationist party when it came to blacks and the egalitarian party when it came to whites.64 In Part Four we will take a closer look how electoral coalitions in the United States and Europe evolved in the twentieth and early twenty-first centuries. In particular, we will analyze the extent to which these differences help to explain why the development of the social and fiscal state was more limited in the United States than in Europe and whether similar racial and ethno-religious factors play a comparable role in the European context in the future.

In any case, as recently as the 1950s inequality in the United States was close to or below what one found in a country like France, while its productivity (and therefore standard of living) was twice as high. By contrast, in the 2010s, the United States has become much more inegalitarian while its lead in productivity has totally disappeared (Figs. 11.1–11.4). The fact that European countries like Germany and France have caught up in terms of productivity is not entirely surprising. Once those countries developed a large fiscal capacity in the postwar period and began investing significant resources in education and, more generally, in social spending and public infrastructure, it was only to be expected that they would overcome the educational and economic lag. The rise of inequality in the United States is more puzzling. In particular, while the standard of living of the poorest 50 percent of Americans was higher than that of the equivalent group in Europe in the 1950s, the situation had totally turned around by the 2010s.

Note from the outset that there are many reasons for the collapse of the relative position of America’s lower classes; the evolution of the educational system is only one of them. The entire social system and the way workers are trained and selected must bear a share of the blame. But I want to stress that my use of the word “collapse” is no exaggeration. The bottom 50 percent of the income distribution claimed around 20 percent of national income from 1960 to 1980, but that share has been divided almost in half, falling to just 12 percent in 2010–2015. The top centile’s share has moved in the opposite direction, from barely 11 percent to more than 20 percent (Fig. 11.5). For comparison, note that while inequality has also increased in Europe since 1980 with a significant rise in the top centile’s share and a fall in the share of the bottom half—which has by no means gone unnoticed in a general climate of sluggish growth—the orders of magnitude are not the same. In particular, the share of total income going to the bottom 50 percent in Europe remains significantly larger than the share going to the top centile (Fig. 11.6).

FIG. 11.5.  The fall of the bottom 50 percent share in the United States, 1960–2015

Interpretation: The share of the bottom 50 percent of the income distribution fell from about 20 percent of total income in the United States in the 1970s to 12–13 percent in the 2010s. During the same period, the top centile share rose from 11 percent to 20–21 percent. Sources and series: piketty.pse.ens.fr/ideology.

Note, too, that there is absolutely no reason to think that this divergence between the United States and Europe was inevitable. The two regions are comparable in size, with the US population around 320 million in 2015 and the West European population around 420 million. Levels of development and productivity are similar. Labor mobility is higher in the United States owing to its greater linguistic and cultural homogeneity, which is widely believed to contribute to income convergence. The United States collects taxes at the federal level (both income and estate taxes) and conducts major federal social programs (such as pensions and health insurance), which is not the case in Europe. Clearly, countervailing factors linked to social, fiscal, and educational policies at the national level in Europe played a more important role.65

It is now well known that the explosion of inequality in the United States since 1980 was due to an unprecedented increase in very high incomes, especially the famous “1 percent.” Indeed, for the top centile’s share of total income to exceed the share of the bottom 50 percent, it is necessary and sufficient for the average income of the first group to be fifty times higher than the average income of the second. This is precisely what happened (Fig. 11.7). Until 1980, the average income of the top centile was on the order of twenty-five times the average income of the bottom 50 percent (roughly $400,000 a year for the top centile versus $15,000 for the bottom 50 percent). In 2015, the average income of the top centile is more than eighty times that of the bottom 50 percent: $1.3 million versus $15,000 (all amounts expressed in constant 2015 dollars).

FIG. 11.6.  Low and high incomes in Europe, 1980–2016

Interpretation: The share of the bottom 50 percent by income fell from 26 percent of total income in Western Europe in the early 1980s to 23 percent in the 2010s. Over the same period, the share of the top centile rose from 7 percent to 10 percent. Sources and series: piketty.pse.ens.fr/ideology.

Without a doubt, however, the most striking phenomenon here was not the rise of the one percent but the fall of the bottom 50 percent. Again, this was in no way inevitable: the increase of the top centile share could have come at the expense of those just below them, the people in the ninetieth to ninety-ninth percentile or of the middle 40 percent (fiftieth to ninetieth percentile). But the fact is that it came almost entirely at the expense of the bottom 50 percent. It is particularly depressing to discover that the disposable income of the bottom 50 percent has stagnated almost completely in the United States since the late 1960s. Before taxes and transfers, the average income of the bottom 50 percent averaged about $15,000 per adult per year in the late 1960s, and it is still at roughly the same level in the late 2010s (in 2015 dollars), a half-century later. This is quite remarkable, especially in view of the significant changes in the US economy and society during this time (including a sharp rise in productivity). In a context notable for rampant deregulation of the financial system, this wage stagnation inevitably increased the indebtedness of the poorest households and the fragility of the banking system, which contributed to the financial crisis of 2008.66

FIG. 11.7.  Low and high incomes in the United States, 1960–2015

Interpretation: In 1970, the average income of the poorest 50 percent was $15,200 per year per adult, and that of the richest 1 percent was $403,000, for a ratio of 1 to 26. In 2015, the average income of the poorest 50 percent was $16,200 and that of the richest 1 percent was $1,305,000, for a ratio of 1 to 81. All amounts are in 2015 dollars. Sources and series: piketty.pse.ens.fr/ideology.

If we now take taxes and transfers into account, we find that the situation of the bottom 50 percent improves only slightly (Fig. 11.8).⁶⁷ We look first at the results obtained if we limit ourselves to cash transfers, including food stamps, which are not cash strictly speaking but nevertheless allow more freedom of use than most transfers in kind. We find that average income is not very different after taxes and transfers, which means that the taxes paid by the bottom 50 percent (notably in the form of indirect taxes) are roughly equal to the cash transfers they receive (including food stamps).⁶⁸

FIG. 11.8.  Low incomes and transfers in the United States, 1960–2015

Interpretation: Expressed in constant 2015 dollars, the average annual income before taxes and transfers of the poorest 50 percent stagnated around $15,000 per adult between 1970 and 2015. The same is true after taxes (including indirect taxes) and monetary transfers (including food stamps), with taxes and transfers roughly balancing each other out. It rises to $20,000 in 2010–2015 if one includes transfers in kind in the form of public health spending. Sources and series: piketty.pse.ens.fr/ideology.

If we now include reimbursements from Medicare and Medicaid, we find that the post-tax-and-transfer income of the bottom 50 percent did increase somewhat, from roughly $15,000 in 1970 to $20,000 in 2015 (Fig. 11.8). Over such a long period of time, however, this not only represents a very limited improvement in living standards; it is also hard to interpret. To be sure, this $5,000 of “additional income” for health expenses does represent an improvement in people’s lives in an era of longer life expectancy (less so in the United States than in Europe, however, particularly for the lower classes). But this increase in transfers also reflects the rising cost of health care in the United States, which in practice means higher pay for physicians, higher profits for pharmaceutical companies, and so on—these groups have prospered in recent decades. Concretely, the additional $5,000 a year for the bottom 50 percent corresponds to roughly one week’s pretax income for a caregiver belonging to the top income decile and roughly one day of income for a caregiver belonging to the top centile. This should clarify the difficulties of interpretation that arise when one looks at transfers in kind and not just in cash.⁶⁹

On the Impact of the Legal, Fiscal, and Educational System on Primary Inequalities

In any event, it is clear that no transfer policy (whether in cash or in kind) can deal satisfactorily with such a massive distortion in the distribution of primary incomes (that is, incomes before taxes and transfers). When the share of primary income going to the bottom 50 percent is nearly halved in the space of just forty years and the share going to the top 1 percent is doubled (Fig. 11.5), it is illusory to think that the change can be compensated simply by ex post redistribution. Redistribution is essential, of course, but one also needs to think about policies capable of modifying the primary distribution, which means making deep changes to the legal, fiscal, and educational system to give the poorest people access to better paying jobs and ownership of property.

The various inequality regimes that we find in history are characterized above all by their primary distribution of resources. This is true of trifunctional and slave societies as well as colonial and ownership societies. It is also true of the various types of social-democratic, communist, postcommunist, and neo-proprietarian societies that arose in the twentieth and early twenty-first centuries. For example, if the United States is now more inegalitarian than Europe, it is solely because primary incomes are more unequally distributed there. If we compare levels of inequality before and after taxes and transfers in the United States and France, as measured by the ratio between the average income of the top 10 percent and the bottom 50 percent, we find that taxes and transfers reduce inequality by comparable amounts in both countries (indeed, slightly more in the United States) and that the global inequality gap is explained entirely by the difference observed before taxes and transfers (Fig. 11.9).⁷⁰ In other words, it is at least as essential to look at “predistribution” policies (which affect primary inequality) as at “redistribution” policies (which reduce inequality of disposable income for a given level of primary inequality).⁷¹

Given the complexity of the social systems involved and the limitations of the available data, it is difficult to precisely quantify the degree to which different institutional arrangements explain variations of primary inequality over time and space. It is nevertheless worth trying to describe the principal mechanisms at work. The legal system plays an essential role, especially in the areas of labor and corporate law. The importance of collective bargaining, unions, and, more generally, rules and institutions involved in wage setting has already been discussed. For example, the presence of worker representatives on company boards (under the Germano-Nordic co-management system) tends to limit extravagant executive pay; indeed, it generally results in more compressed and less arbitrary pay scales.⁷² The minimum wage and its evolution also play a central role in explaining variations in wage inequality across time and space. In the 1950s and 1960s, the United States had by far the highest minimum wage in the world. In 1968–1970 the federal minimum wage was more than $10 an hour in today’s dollars. Since 1980, however, the failure to raise the minimum wage regularly gradually eroded its value in real terms: in 2019 it was only $7.20, representing a 30 percent decline in purchasing power over half a century—remarkable for a country at peace and growing economically. This reversal attests to the magnitude of the political-ideological changes that took place in the United States since the 1970s and 1980s. Over the same period, the French minimum wage rose from barely 3 euros an hour in the 1960s to 10 euros in 2019 (Fig. 11.10), advancing at roughly the same rate as the average productivity of labor (Fig. 11.3).

FIG. 11.9.  Primary inequality and redistribution in the United States and France

Interpretation: In France, the ratio of average income before taxes and transfers of the top decile to the bottom 50 percent rose from 6.4 in 1990 to 7.4 in 2015. In the United States, the same ratio rose from 11.5 to 18.7. In both countries, taking account of taxes and monetary transfers (including food stamps and housing allowances) reduces inequality by 20–30 percent. Note: The distribution is annual income per adult. Sources and series: piketty.pse.ens.fr/ideology.

FIG. 11.10.  Minimum wage in the United States and France, 1950–2019

Interpretation: Converted into 2019 purchasing power, the federal minimum wage rose from $4.25 in 1950 to $7.24 per hour in 2019 in the United States, while the national minimum wage (SMIG in 1950, then SMIC after 1970) rose from €2.23 in 1950 to €10.03 per hour in 2019. The two scales represent purchasing power parity ($1.20 to 1€ in 2019). Sources and series: piketty.pse.ens.fr/ideology.

Many works have shown that the drop in the minimum wage in the United States contributed strongly to the declining position of low-wage workers since the 1980s in a general climate of decreased worker bargaining power. The federal minimum wage fell so much relative to the general productivity level, moreover, that several states raised their minimum wage to a much higher level without hurting employment. For example, the minimum wage in California in 2019 is $11 an hour and will gradually rise to $15 by 2023. Similarly, the high federal minimum wage from the 1930s to the 1960s, in a context of high US productivity and skill levels, helped to reduce wage inequality while employment remained high. Recent work has shown that the extension of the minimum wage in the 1960s to sectors in which African American labor was employed more intensively (including agriculture, which had been excluded from the federal minimum wage law in 1938, partly because of hostility from southern Democrats) strongly contributed to reducing wage discrimination and the wage gap between blacks and whites.73

It is interesting to note that several European countries were relatively slow to adopt a national minimum wage. The United Kingdom did so only in 1999 and Germany in 2015. These countries previously relied solely on wage negotiations at the firm and sector level, which could result in high minimum wages but with variations from sector to sector. Changes in the structure of employment, especially the decline of industrial employment and the gradual shift to services coupled with a lower unionization rate, have gradually reduced the role of collective bargaining since the 1980s. This is probably part of the reason for greater reliance on a national minimum wage.⁷⁴ While the minimum wage is an indispensable tool, it is no substitute for wage bargaining and power sharing at the branch and firm level; these could take new forms in the future.

In addition to the legal system (labor and corporate law), the tax system can also have a decisive impact on primary inequalities. This is obviously the case for the inheritance tax. A progressive wealth tax that could be used to finance a universal capital endowment might have a similar effect. Taxing wealth leads to structural reductions of wealth inequality in each new generation, which also helps to equalize investment opportunities and thus the future distribution of labor and capital income. More surprisingly, perhaps, the progressive income tax has also had a very strong impact, not only on after-tax inequality but also on primary inequality (before taxes and transfers).

First, higher tax rates on large incomes helped to limit the concentration of saving and capital accumulation at the top of the distribution, while reduced tax rates in the middle and bottom of the distribution contributed to the diffusion of property. In addition, one of the main consequences of the extremely high marginal rates (70–90 percent) on top incomes between 1930 and 1980, especially in the United States and United Kingdom,⁷⁵ was to put an end to the most extravagant executive pay. By contrast, the sharp reduction of top tax rates in the 1980s strongly contributed to the skyrocketing of executive pay. Indeed, if one looks at the evolution of executive pay in listed companies in all the developed countries since 1980, one finds that variations in tax rates explain much of the variation in executive pay—much more than other factors such as sector of activity, firm size, or performance.⁷⁶ The mechanism at work seems to be linked to the way executive pay is determined and to the bargaining power of executives. How does an executive persuade other relevant actors (including direct subordinates, other employees, shareholders, and members of the firm’s compensation committee) that a pay raise is justified? The answer is never obvious. In the 1950s and 1960s, the top executives of major British and American firms had little interest in fighting for huge raises, and other actors were reluctant to grant them because 80–90 percent of any raise would have gone directly to the government. In the 1980s, however, the nature of the game changed completely. The evidence suggests that executives began to devote considerable effort to persuading others that enormous raises were warranted, which was not always difficult to do, since it is hard to measure how much any individual executive contributes to the firm’s success. What is more, compensation committees were often constituted in a rather incestuous fashion. This also explains why it is so difficult to find any statistically significant correlation between executive pay and firm performance (or productivity).⁷⁷

Since the 1980s, moreover, US production has become more and more concentrated in the largest companies (not just in the information technology sector but across the economy). This increased the bargaining power of executives in the leading firms in each sector and enabled them to compress the bottom and middle portions of the pay scale and increase the profit share of value-added.⁷⁸ This growing concentration reflects the weakness of antitrust policies, the failure to keep up with changing industrial conditions, and above all, the lack of political will on the part of successive administrations to take any action against monopolies. The reasons for this include an ideological context favorable to laissez-faire, heightened international competition, and perhaps a campaign financing system biased in favor of large corporations and their leaders (I will come back to this).

Higher Education and the New Educational and Social Stratification

Last but perhaps not least, in addition to the legal and tax systems, the educational system also plays a crucial role in shaping primary inequalities. In the long run, it is access to skills and diffusion of knowledge that allow inequality to be reduced both within countries and at the international level. Technological progress and transformation of the structure of employment mean that the productive system demands ever higher levels of skill. If the supply of skills does not evolve to meet this demand—for example, if some social groups fail to increase or even decrease their investment in education while others devote an increasing share of their resources to training—wage inequality between the two groups will tend to increase, no matter how good the legal or tax system in place.

The evidence strongly suggests that growing educational investment has played a central role in the particularly sharp increase of income inequality in the United States since the 1980s. In the 1950s and 1960s, the United States was the first country to have achieved nearly universal secondary education. In the 1980s and 1990s, Japan and most countries in Western Europe caught up. All of these countries have now entered the age of mass higher education, in which a growing proportion of each new age cohort attends college or university. In the mid-2010s, the tertiary schooling rate (defined as the percentage of young adults of age 18–21 enrolled in an institution of higher learning) is 50 percent or more in the United States and all the countries of Western Europe and approaching 60–70 percent in Japan and Korea.79 The educational and symbolic order has been turned upside down. In the past, higher education was the privilege of a small fraction of the population: still less than 1 percent at the turn of the twentieth century and less than 10 percent until the 1960s. In the wealthy countries, majorities of the younger generations are now college graduates, and eventually majorities of the entire population will be as well. The process is well under way: given the rate of generational replacement, we find that the proportion of the adult population with a college degree, which is currently 30–40 percent in the United States and in the most advanced European and Asian countries, will rise to 50–60 percent a few decades from now.

This educational upheaval is the source of new kinds of inequality, both between and within countries. The United States lost its educational lead in the 1980s. Many studies have shown how the slowdown in educational investment in the United States contributed to the increase of education-related income inequality since the 1980s and 1990s.⁸⁰ Note, too, that the financing of primary and secondary education, though very largely public (as in most developed countries), is extremely decentralized in the United States. It depends essentially on local property taxes, which can lead to significant inequality depending on the wealth of the community. Compared with European and Asian countries, where the financing of primary and secondary education is generally centralized at the national level, secondary education in the United States is therefore somewhat less universal than elsewhere. Nearly everyone finishes high school, but the variation in the quality and financial resources of different high schools is quite large.

Furthermore, recent research has shown that access to higher education in the United States is largely determined by parental income. More specifically, the probability of attending university in the mid-2010s was 20–30 percent for children of the poorest parents, increasing almost linearly to 90 percent for the children of the richest parents (see Fig. I.8). Similar data for other countries, though quite incomplete (which is itself problematic), suggest that the slope of the curve is less steep. In addition, research comparing the relative income of parents and children shows a particularly steep curve (hence a very low intergenerational mobility rate) in the United States compared with Europe, especially the Nordic countries.⁸¹ Note, too, that the intergenerational correlation between the position of parents in the income hierarchy and that of children has increased sharply in the United States in recent decades.⁸² This significant decrease in social mobility, which contrasts so flagrantly with hypothetical talk about “meritocracy” and equality of opportunity, attests to the extreme stratification of the American educational and social system. It also demonstrates the importance of subjecting political-ideological rhetoric to systematic empirical evaluation, which the available sources do not always permit us to do with sufficient comparative historical perspective.

The fact that access to higher education in the United States is strongly linked to parental income can be explained in many ways. In part it reflects a preexisting stratification: since primary and secondary education is already highly inegalitarian, children from modest backgrounds are less likely to satisfy the admissions requirements of the more highly selective universities. It also reflects the cost of private education, which has attained astronomical heights in the United States in recent decades. More broadly, while all developed countries pay for primary and secondary education almost exclusively with public funds, there is much greater variation in the financing of higher education. Private financing pays 60–70 percent of the cost in the United States and nearly 60 percent in the United Kingdom, Canada, and Australia—compared with an average of 30 percent in France, Italy, and Spain, where tuition is generally lower than in the United States and United Kingdom, and less than 10 percent in Germany, Austria, Sweden, Denmark, and Norway, where higher education is in principle virtually free, just like primary and secondary education (Fig. 11.11).⁸³

FIG. 11.11.  Share of private financing in education: Diversity of European and American models

Interpretation: In the United States, private financing represented 65 percent of total (private and public) financing in higher education and 9 percent of total financing of primary and secondary education. The shares of private financing of higher education vary strongly with country, with an Anglo-American model, a south European model, and a north European model. Private financing is relatively insignificant in primary and secondary education everywhere (2014–2016 figures). Sources and series: piketty.pse.ens.fr/ideology.

In the US case, the importance of private financing of higher education has had two key consequences: first, the best American universities are very rich (which allows them to attract some of the best foreign researchers and students), and second, the system of higher education is extremely stratified. If one considers all resources (public and private) available for higher education, the United States continues to lead the world.84 The problem is that the gap between the resources available to the best universities and those available to less well-endowed public universities and community colleges has grown to abyssal proportions in recent decades. This inequality has been exacerbated by the financial dynamics of global capitalism. The universities with the largest endowments have earned higher yields on their investments than those with smaller endowments, which has widened the gap between them.⁸⁵ If one looks at the available international rankings, as imperfect as they are, it is striking to see that American universities are ultra-dominant among the top twenty in the world but fall well below European and Asian universities if one looks at the top 100 or top 500.⁸⁶ It is likely that the international renown of the wealthiest US universities masks the internal imbalance of the system as a whole. That imbalance would probably be even clearer if US universities were not so attractive to students from the rest of the world. This is a new form of interaction between the global inequality regime and domestic inequality not seen in earlier periods.

Can One Buy a Place in a University?

Furthermore, inequality of access to higher education in the United States is aggravated by the fact that the wealthiest parents can in some cases use financial contributions to win admission to the best universities for children who would not otherwise qualify. Admissions procedures often include not very transparent “legacy preferences” (that is, special advantages for the children of graduates of the institution in question). Unsurprisingly, the American universities where such preferences are allowed claim that the number of students thus favored is ridiculously small—in fact, so tiny that it would be pointless to name them publicly or to explain the algorithms and procedures used to winnow applicants. Indeed, it is likely that the numbers are small and that these opaque practices play a quantitatively less important role than other mechanisms (such as the decentralized public financing of primary and secondary education and the high tuitions and high yields on endowments) in explaining the overall inequality of the system.

The question nevertheless deserves close attention, for several reasons. First, research has shown that the practices may be somewhat less marginal than the universities claim. It turns out that gifts by graduates to their former universities are abnormally concentrated in years during which their children are of an age to apply for admission.87 Furthermore, the lack of transparency is in itself clearly problematic, all the more so in that the new class of inheritors (the beneficiaries of greater US inequality in recent decades) stands out more and more conspicuously in the social landscape; this may stoke resentment of elites.⁸⁸ The lack of transparency shows that the universities are not prepared to defend what they are doing in public; this can only encourage serious doubts about the overall fairness of the system.

It is also striking to discover that American university faculty are increasingly inclined to justify these practices and the secrecy that surrounds them because they are effective in raising funds from the generous billionaires who finance their research and teaching. This ideological evolution is interesting, because it raises a more general question: Exactly how far should the power of money extend, and what institutions and procedures can set limits on that power? We have run into this type of question before: for example, in considering the Swedish practice of awarding voting rights in proportion to wealth in the period 1865–1911.⁸⁹ In the present case, the more apt comparison might be with the imperial Chinese exam system in the Qing era, which allowed wealthy elites to purchase places for their children (in addition to setting aside places for the children of the old warrior class), which undoubtedly weakened the regime and undermined its moral and political legitimacy.⁹⁰

Last but not least, the flagrant lack of transparency in the admissions procedures of America’s leading universities is of concern to all countries because it raises this fundamental challenge: How to define educational justice in the twenty-first century? For example, suppose one wants a quota system with extra points to encourage better representation of disadvantaged social classes, as in India.⁹¹ If every university keeps its admissions algorithm secret, and if that algorithm awards extra points to the children of the rich rather than to the disadvantaged while admissions officials claim that the practice is very rare and must be kept secret, how is democratic deliberation supposed to proceed—especially when the issue is so delicate and complex, affecting the futures of children from lower, middle, and upper classes, and when it is so difficult to construct a standard of justice acceptable to the majority? Yet authorities in the United States have been able to impose much stricter rules and standards on universities in the past.⁹² As always, history shows that nothing is foreordained.

On Inequality of Access to Education in Europe and the United States

As noted, inequality of access to education is quite significant in the United States. It is also significant in Europe. Indeed, throughout the world one finds a wide gap between official rhetoric about equality of opportunity, the “meritocratic” ideal, and so on and the reality of unequal access to education for different social groups. No country is in a position to give lessons on the subject. Indeed, the advent of the era of higher education has posed a structural challenge to the very idea of educational equality everywhere.

In the era of primary and secondary education, there was a fairly obvious rule of thumb for educational equality: the goal was to achieve first universal primary education and then universal secondary education so that every child would receive roughly the same grounding in basic knowledge. With tertiary education, however, things became much more complicated. For one thing, it is not very realistic to think that every child will grow up to receive a PhD, at least not any time soon. Indeed, there are many paths to higher education. In part, this diversity reflects the variety of fields of knowledge and the range of individual aspirations, but it lends itself to hierarchical organization. This in turn influences social and professional hierarchies after graduation. In other words, the advent of mass higher education poses a new kind of political and ideological challenge. One has to live with some degree of permanent educational inequality, especially between those who embark on long courses of study and those who opt for shorter courses. Obviously, this in no way precludes thinking about how to allocate resources more justly or how to devise fairer rules for access to different curricula. But the challenge is more complex than that of achieving strict equality in primary and secondary education.93

In Part Four we will see that this new educational challenge is one of the main factors that led to the breakdown of the postwar social-democratic coalition. In the 1950s and 1960s, the various European social-democratic and socialist parties as well as the Democratic Party in the United States scored their highest percentages of the vote among less educated social groups. In the period 1980–2010, this voting pattern was reversed, and the same parties did best among the better educated. One possible explanation, which we will explore in greater detail later, has to do with changes in the policies backed by these parties, which gradually came to be seen as more favorable to the winners in the socio-educational competition.⁹⁴

At this stage, note simply that even though the educational system is on the whole more egalitarian in Europe than in the United States, European countries too have found it quite difficult to cope with the challenge of educational expansion in recent decades. For instance, it is striking to note that public spending on education, which increased rapidly over the course of the twentieth century from barely 1–2 percent of national income in 1870–1910 to 5–6 percent in the 1980s, subsequently plateaued (Fig. 10.15). In all the countries of Western Europe, whether Germany or France, Sweden or the United Kingdom, we find educational investment stagnating between 1990 and 2015 at about 5.5–6 percent of national income.⁹⁵

This stagnation can of course be explained by the fact that public spending in general stopped growing in this period. In a context marked by the structural and all-but-inevitable increase in spending on health and pensions, some people felt that it was essential to hold the line on educational spending or even decrease it somewhat in relation to national income, relying more instead on private financing and tuition fees. Alternatively, one might have considered (and might in the future still consider) a limited tax increase to pay for additional investment in education, tapping all levels of income and wealth in a fair and equitable manner. In other words, tax competition between countries, combined with the perceived impossibility of devising a fair tax system, may explain both the stagnation of educational investment and the recourse to deficit spending.

In any case, it is important to note how paradoxical this spending freeze was. Just as the developed countries were moving into the era of mass higher education and as the proportion of each age cohort attending college was rising from barely 10–20 percent to more than 50 percent, public spending on education came to a standstill. As a result, some who had believed in the promise of expanding access to higher education—often people of modest or middle-class background—found themselves confronted with dwindling resources and absence of opportunities after graduation. Note, moreover, that even when college is free or nearly free and most of the cost is borne by the government, true equality of access to higher education is nevertheless not guaranteed. Students from privileged backgrounds are often better placed to enter more promising courses of study, thanks both to their family heritage and to prior access to better schools and high schools.

The French case offers a particularly striking example of educational inequality within an ostensibly free and egalitarian public system. In practice, the public resources invested in elitist tracks that prepare students for the so-called grandes écoles (the most prestigious institutions of higher education) are two to three times greater per student than the resources invested in less elitist tracks. This longstanding stratification of the French system became flagrant in the era of mass higher education, especially because promises to equalize investment in less privileged primary, middle, and high schools were never kept; this gave rise to very strong social and political tensions. Beyond the French case, educational justice requires transparency about resource allocation and admissions procedures. This is a fundamental issue, which will become increasingly urgent around the world in years to come. I will have much more to say about it later on.⁹⁶

Educational Equality, the Root of Modern Growth

Note, finally, that the stagnation of educational investment in the rich countries since the 1980s may help to explain not only the rise of inequality but also the slowing of economic growth. In the United States, per capita national income grew at a rate of 2.2 percent per year in the period 1950–1990 but slowed to 1.1 percent in the period 1990–2020. Meanwhile, inequality increased, and the top income tax rate fell from an average of 72 percent in the period 1950–1990 to 35 percent in the period 1990–2020 (Figs. 11.12–11.13). In Europe, we also find that growth was strongest in the period 1950–1990, when inequality was lower and fiscal progressivity greater (Figs. 11.14–11.15). In Europe, the exceptional growth of 1950–1990 can be attributed in part to the need to make up ground lost during the two world wars. This does not apply to the United States, however: growth in the period 1910–1950 was stronger than in 1870–1910 and growth in the period 1950–1990 was even more rapid than in 1910–1950, but the growth rate then fell by half in the period 1990–2020.

This stark historical reality has much to teach us. In particular, it rules out a number of mistaken diagnoses. First, strongly progressive taxes are clearly no obstacle to rapid productivity growth, provided that the top rates apply to sufficiently high levels of income and wealth. If rates on the order of 80–90 percent were applied to everyone even slightly above the mean, for instance, it is quite possible that the effects would be different. But when the top rates apply only to very high levels of income and wealth (typically in the top centile or half centile), the historical evidence suggests that it is quite possible to combine highly progressive taxes, low inequality, and high growth. The strongly progressive tax system that was put in place in the twentieth century helped end the extreme concentration of wealth and income observed in the late nineteenth and early twentieth centuries, and this reduction of inequality opened the way to stronger growth than ever before. At a minimum, this should convince everyone that the very high level of inequality that existed before World War I was in no way necessary for growth, as much of the elite claimed at the time. Everyone should also agree that the conservative Reagan revolution of the 1980s was a failure: growth in the United States fell by half, and the notion that it would have fallen even more in the absence of conservative reforms is not very plausible.⁹⁷

FIG. 11.12.  Growth and inequality in the United States, 1870–2020

Interpretation: In the United States the growth of per capita national income fell from 2.2 percent per year from 1950 to 1990 to 1.1 percent from 1990 to 2020, while the top centile share of national income rose from 12 to 18 percent in the same period. Sources and series: piketty.pse.ens.fr/ideology.

FIG. 11.13.  Growth and progressive taxation in the United States, 1870–2020

Interpretation: In the United States, annual growth of per capita national income fell from 2.2 percent from 1950 to 1900 to 1.1 percent from 1990 to 2020, whereas the top marginal income tax rate fell in the same period from 72 percent to 35 percent. Sources and series: piketty.pse.ens.fr/ideology.

FIG. 11.14.  Growth and inequality in Europe, 1870–2020

Interpretation: In Western Europe, growth of per capita national income fell from 3.3 percent in 1950–1990 to 0.9 percent in 1990–2020, while the top centile share of national income rose over the same period from 8 to 11 percent (average for Germany, United Kingdom, and France). Sources and series: piketty.pse.ens.fr/ideology.

FIG. 11.15.  Growth and progressive tax in Europe, 1870–2020

Interpretation: In Western Europe, annual growth of per capita national income fell from 3.3 percent in 1950–1990 to 0.9 percent in 1990–2020, while the top marginal income tax rate fell over the same period from 68 to 49 percent (average for Germany, the United Kingdom, and France). Sources and series: piketty.pse.ens.fr/ideology.

Last but not least, the historic role played by America’s educational lead in the nineteenth and much of the twentieth centuries shows how crucial egalitarian investment in training and education was. Why was the United States more productive than Europe in the nineteenth and early twentieth centuries, and why did its economy grow faster? Not because property rights were better protected or because taxes were lower; taxes were low everywhere, and property rights were nowhere better protected than in France, the United Kingdom, and elsewhere in Europe. The key point is that the United States in the nineteenth and twentieth centuries had a fifty-year head start on Europe in terms of universal primary and later secondary education. This advance ended toward the end of the twentieth century, and with it ended the productivity gap. At a more general level, the period 1950–1990 saw an exceptionally high level of educational investment in all the rich countries, much higher than in previous periods, which may help to explain the unusually high level of growth. By contrast, the stagnation of educational investment in the period 1990–2020, even as more and more students headed to university, is consistent with slower productivity growth.

To sum up: in the light of the history of the past two centuries, educational equality played a more important role in economic development than the sacralization of inequality, property, and stability. More generally, history demonstrates the recurrent risk of an “inequality trap,” which many societies have faced throughout the ages. Elite discourse tends to overvalue stability, and especially the perpetuation of existing property rights, whereas development often requires a redefinition of property relations and opening up of opportunities to new social groups. The refusal of British and French elites to redistribute wealth and invest in education and the social state continued until World War I. This refusal rested on sophisticated ideological constructs, as is also the case in the United States today.98 History shows that change can come only when social and political struggle converges with profound ideological renewal.

Social Democracy and Just Taxation: A Missed Opportunity

Let us turn now to the question of just taxation, which will lead to the question of transcending the nation-state. We have seen the difficulties that social-democratic societies encountered when they tried to redefine the norms of just property and education after 1980, when the basic agenda of nationalizations ceased to be attractive and the world entered the era of higher education. The same political-ideological limitations hampered new thinking about taxes. Parties of the left—Social Democrats, Socialists, Labour, Democrats—tended to neglect fiscal doctrine and just taxation. The dramatic rise of progressive income and inheritance taxes in the period 1914–1945 generally came about as an emergency response and was never fully integrated into party doctrine, either intellectually or politically. This partly explains the fragility of the fiscal institutions that were put in place and the challenges that were raised against them in the 1980s.

Broadly speaking, the socialist movement grew as a response to the question of the property regime, with the goal of nationalizing privately owned firms. This focus on state ownership of the means of production, which remained strong among French Socialists and British Labourites until the 1980s, tended to foreclose thinking about other issues, such as progressive taxes, co-management, and self-management. In short, faith in state centralization as the only way to transcend capitalism sometimes led to neglect of tax-related issues, including what should be taxed and at what rates as well as issues of power sharing and voting rights within firms.

Among the shortcomings of social-democratic reflection on tax issues, two points warrant special mention. First, parties of the left failed to foster the kind of international cooperation needed to protect and extend progressive taxation; indeed, at times they contributed to the fiscal competition that has proved devastating to the very idea of fiscal justice. Second, thinking about just taxation too often neglected the idea of a progressive wealth tax, despite its importance for any ambitious attempt to transcend private capitalism, particularly if used to finance a universal capital endowment and promote greater circulation of wealth. As we will see in what follows, just taxation requires striking a balance among three legitimate and complementary forms of progressive taxation: taxes on income, inheritance, and wealth.

Social Democracy and the Transcendence of Capitalism and the Nation-State

Twentieth-century social democracy was always internationalist in principle but much less so in political practice. As we saw in Chapter 10, this was the critique that Hannah Arendt leveled at the social democrats of the first half of the twentieth century in 1951. It could equally well be extended to their successors in the second half of the century. After 1950, social-democratic movements focused on building the fiscal and social state within the narrow framework of the nation-state. Although they achieved undeniable success, they did not really try to develop new federal or transnational political forms (such as social, democratic, and egalitarian counterparts to the transnational colonial, Bolshevik, and Nazi regimes analyzed by Arendt). Because social democracy failed to achieve postnational solidarity or fiscality (as the absence of a common European fiscal and social policy attests), it weakened what it had built at the national level, endangering its social and political base.

At the European level, various social-democratic and socialist movements did of course steadfastly support efforts to develop the European Coal and Steel Community in 1952, followed by the European Economic Community (EEC) created by the Treaty of Rome in 1957, and finally the European Union, which succeeded the EEC in 1992. This series of political, economic, and trade agreements, consolidated by treaty after treaty, paved the way to an unprecedented era of peace and prosperity in Europe. Cooperation made this possible, initially by regulating competition in major areas of industrial and agricultural production. The contrast is striking between the 1920s, when French troops occupied the Ruhr to exact payment of a debt-tribute equivalent to 300 percent of German GDP, and the 1950s, when France, Germany, Italy, and the Benelux countries (Belgium, the Netherlands, and Luxembourg) coordinated their production of coal and steel to stabilize prices and ensure the smoothest possible postwar reconstruction. In 1986 the Single European Act established the principle of free circulation of goods, services, capital, and people in Europe (the “four freedoms”).⁹⁹ Then the Maastricht Treaty of 1992 established not only the European Union but also a common currency for those countries that wanted it (the euro came into use by banks in 1999 and entered general circulation in 2002). Since then, member states have increasingly relied on EU institutions to negotiate trade agreements between Europe and the rest of the world in a context of rapidly expanding international economic openness. Scholars have accurately described the construction of Europe in the period 1950–2000 as a “rescue of the nation-state,” a political form that to many people seemed doomed in 1945–1950. In fact, at first the EEC and then the EU allowed Europe’s old nation-states to coordinate their output and trade, initially among themselves and then with the rest of the world while maintaining their role as central political players.¹⁰⁰

Despite its successes, the European construction suffered from many limitations, which today threaten to turn large numbers of people against the entire project as illustrated by the Brexit referendum of 2016. Over the past few decades, the feeling has spread that “Europe” (a word that has come to refer to the bureaucracy in Brussels, ignoring all previous phases of the process) penalizes the lower and middle classes for the benefit of the wealthy and large corporations. This “Euroskepticism” has also fed on hostility to immigration and a sense of lost status (compared with the colonial era in some places or the communist era in others). In any case, European governments have been unable to cope with the combination of rising inequality and lower growth since the 1980s. What are the reasons for this resounding failure? First, Europe has relied almost exclusively on a competitive model pitting region against region and person against person, which has benefited groups perceived to be more mobile. Second, member states have been unable to agree on any kind of common fiscal or social policy. This failure is itself a result of the decision to require unanimity in fiscal matters, a decision perpetuated in treaty after treaty from the 1950s to the present.¹⁰¹

To date, the construction of Europe has been based largely on the hypothesis that free competition and free circulation of goods and capital should suffice to achieve general prosperity and social harmony—on the conviction that the benefits of fiscal competition between states outweigh the costs (the benefits coming from the fact that competition is supposed to prevent states from becoming too bloated or giving in to redistributive fantasies). These hypotheses are not totally indefensible from a theoretical point of view. Indeed, it is not easy to build a political structure with the legitimacy to levy taxes, particularly on a scale as large as Europe. Yet the same hypotheses are also vulnerable to criticism, especially in view of the recent rise of inequality and the dangers it entails as well as the fact that political communities of comparable or larger size, such as the United States and India, have long since adopted common fiscal policies in a democratic framework. The fact that European integration strategy since the 1950s has been based on the construction of a common market can also be explained by the history of the previous decades. In the interwar years, the rise of protectionism and noncooperative mercantilist strategies made the crisis worse. In a way, the ideology of competition is a response to the crises of the past. Yet by proceeding in this way, Europe’s builders forgot another lesson of history: the steady rise of inequality in the years 1814–1914, which demonstrated the need to embed the market in a web of social and fiscal regulations.

It is particularly striking that European social democrats (in particular the German Social Democrats and French Socialists), even though they have regularly held power (sometimes simultaneously) and been in a position to rewrite existing treaties, never formulated a specific proposal to replace the unanimity rule for fiscal policy making. No doubt they were not entirely convinced that the (genuine) complications of a common fiscal policy were worth the trouble. Admittedly, creating a federal structure appropriate to Europe and its old nation-states will be anything but simple. Nevertheless, there are many conceivable ways in which a democratic European federation might have agreed on a common tax policy—a prospect that was already contemplated in 1938–1940 in debates about the Federal Union (Chapter 10). This could quickly become a reality in the years and decades to come (I will come back to this).

However, the fact remains that the unanimity rule and fiscal competition led in the period 1980–2020 to rampant “fiscal dumping” in which countries competed for business by undercutting one another’s tax rates—particularly with respect to corporate tax rates, which gradually fell from 45–50 percent in most countries in the 1980s to just 22 percent on average across the EU in 2018, while overall tax revenues remained stable. Furthermore, there is no guarantee that the long-term decline in corporate tax rates has ended.¹⁰² Rates could still drop toward 0 percent or even become subsidies to attract investment, as is sometimes already the case. Although European states need corporate tax revenues to finance their social benefits, they have been world leaders in reducing corporate taxes, far more than the United States (where corporate taxes, like income and estate taxes, are levied for the most part at the federal level). This attests to the importance of tax competition as well as to the central role of political and electoral institutions for fiscal outcomes.¹⁰³ The fact that the construction of Europe has become synonymous with the defense of “free and undistorted competition” and that the EU is widely perceived as a force hostile or indifferent to the development of the social state also explains why the British Labour Party was divided in the 1972 referendum about whether the United Kingdom should join the EU and again in the 2016 Brexit referendum. Yet between those two dates the party proposed nothing that might have changed the perception of the European Union.¹⁰⁴

Rethinking Globalization and the Liberalization of Capital Flows

Recent research has also shown the central role played by European social democrats and especially the French Socialists in liberalizing capital flows in Europe and the world since the late 1980s.105 Burnt by the difficulties they faced in implementing the nationalizations of 1981, the ill-timed stimulus of 1981–1982, and the exchange controls of 1983, which would have affected the middle class without reducing capital flight by the wealthy, the French Socialists decided in 1984–1985 on a radical change in their economic and political strategy. In the wake of the Single European Act of 1986, they gave in to the demands of the German Christian Democrats for complete liberalization of capital flows, which led to a 1988 European directive that was later incorporated into the Maastricht Treaty of 1992. Its terms were subsequently borrowed by the Organization for Economic Co-operation and Development (OECD) and the International Monetary Fund (IMF) and became a new international standard.¹⁰⁶ According to the accounts of the principal actors in the process, the concessions made by the French Socialists to German demands (which were intended to guarantee full “depoliticization” of monetary and financial questions) were seen as acceptable compromises in exchange for German agreements to a single currency and a shared federal sovereignty over the future European Central Bank (ECB).¹⁰⁷ In fact, the ECB became the only truly federal European institution (neither the German nor the French representative can veto decisions of the majority of the board of directors). As we will see, this allowed it to play a significant role in the aftermath of the 2008 financial crisis.

It is not clear, however, that the principal actors involved fully understood the long-run consequences of completely liberalizing capital flows. The problem was not just short-term flows—the “hot money” that Roosevelt denounced in 1936 and whose destabilizing effects were obvious in the 1930s (especially in the Austrian banking crisis of 1931). These had been regulated, for good reason, between 1945 and 1985 but then liberalized to such an extent that they were partly to blame for the Asian crisis of 1997.¹⁰⁸ More generally, liberalization of capital flows becomes a problem if it is not accompanied by international agreements providing for automatic exchanges of information about who owns cross-border capital assets along with coordinated and balanced policies to regulate and tax profits, income, and wealth. The problem is precisely that when the world moved in the 1980s to free circulation of goods and capital on a global scale under the influence of the United States and Europe, it did so without any fiscal or social objectives in mind, as if globalization could do without fiscal revenues, educational investments, or social and environmental rules. The implicit hypothesis seems to have been that each nation-state would have to deal with these minor problems on its own and that the sole purpose of international treaties was to arrange for free circulation and prevent states from interfering with it. As is often the case with historical turning points of this kind, the most striking thing is how unprepared decision makers were and how much they had to improvise. Note, by the way, that the economic and financial liberalization that began in the 1980s was not entirely due to the conservative revolutions in the United States and United Kingdom: French and German influences also played a central role in these complex developments.¹⁰⁹ The role played by numerous financial lobbies from several European countries (such as Luxembourg) should also be stressed.¹¹⁰

Note, too, that the inability of postwar social democracy to organize the social and fiscal state on a postnational scale was not limited to Europe; we find it in all parts of the world. Attempts to organize regional unions in Latin America, Africa, and the Middle East ran afoul of similar difficulties. We saw earlier how West African leaders, already aware in 1945–1960 of the difficulties their tiny nation-states would face in finding their place and developing viable social models within global capitalism, unsuccessfully sought to develop new types of federations—most notably the Mali Federation consisting of Senegal, Dahomey, Upper Volta, and present-day Mali (see Chapter 7). The ephemeral United Arab Republic (1958–1961), a union of Egypt and Syria (and briefly Yemen), also reflects awareness of the fact that a large community is needed to control the economic forces of capitalism. In this context, the European Union plays a special role owing to the wealth of its members and the potential to inspire emulators by its success.

Furthermore, the magnitude of the European social and fiscal state, which claimed 40–50 percent of national income as taxes in the period 1990–2020,¹¹¹ implies that questions of fiscal justice and consent to taxation should play a crucial role. But consent has been sorely tested, partly because European tax systems are so complex and lack transparency (because they have developed in stages and have never been reformed and rationalized as much as they could have been) and partly because of heightened fiscal competition and lack of coordination between states, which tends to favor those social groups that have already benefited the most from the globalization of trade.

In this connection, bear in mind that the concentration of wealth and income from capital, though less extreme than in the Belle Époque (1880–1914), remained quite high in the late twentieth century and remains high today, higher than the concentration of income from labor (see Figs. 10.6–10.7). This implies that the highest incomes consist in large part of income from wealth, especially dividends and interest on financial capital (Figs. 11.16–11.17). Inequalities of capital and labor income both remain high, but the orders of magnitude are not at all the same. In regard to capital income, the bottom 50 percent account for only 5 percent of all capital income in France in 2015, compared with 66 percent for the top decile (Fig. 11.18). As for labor income, the bottom 50 percent receive 24 percent of the total, or nearly as much as the 27 percent going to the top decile (who are of course one-fifth as numerous). Note, too, that the high concentration of wealth and of the income derived from it is not skewed by the age profile of the wealthy; it can be found in every age cohort, from the youngest to the oldest. In other words, wealth diffuses only very slowly with age.¹¹²

FIG. 11.16.  Composition of income in France, 2015

Interpretation: In France in 2015 (as in most countries for which data are available), low and medium incomes consist mainly of labor income, and high incomes mainly of capital income (especially dividends). Note: The distribution shown here is annual income per adult before taxes but after pensions and unemployment insurance. Sources and series: piketty.pse.ens.fr/ideology.

In view of this very high concentration of wealth (especially financial wealth), it is easy to see why liberalizing capital flows without exchange of information or fiscal coordination can undermine the overall progressivity of the tax system. Beyond the race to the bottom on taxing corporate profits, many European countries allowed dividends and interests to escape progressive taxation in the period 1990–2020. This in turn allowed wealthy people to pay less on their income than a person earning an equivalent amount entirely from labor—a radical change in perspective compared with earlier periods.113

FIG. 11.17.  Composition of property in France, 2015

Interpretation: In France in 2015 (as in all countries for which data are available), small fortunes consist primarily of cash and bank deposits, medium fortunes of real estate, and large fortunes of financial assets (mainly stocks). Note: The distribution shown here is wealth per adult (couples’ wealth is divided in half). Sources and series: piketty.pse.ens.fr/ideology.

FIG. 11.18.  Inequalities with respect to capital and labor in France, 2015

Interpretation: The top decile of capital income accounts for 66 percent of total capital income, compared with 5 percent for the bottom 50 percent and 29 percent for the middle 40 percent. For labor income, these shares are respectively 27, 24, and 49 percent. Note: The distributions shown here are income per adult (couples’ income is divided in half). Sources and series: piketty.pse.ens.fr/ideology.

In fact, if one tries to calculate a comprehensive profile of the tax structure, it turns out that progressivity has decreased significantly since the 1980s. This follows automatically from the fact that the average tax rate has remained stable while rates on the highest income brackets have declined.114 This general factor has been aggravated by various exemptions. In France, the overall tax rate is 45–50 percent on the bottom 50 percent, 50–55 percent on the middle 40 percent, and 45 percent within the wealthiest 1 percent (Fig. 11.19). In other words, taxes are slightly progressive from the bottom to the middle of the distribution but regressive at the top. This is a result of the importance of indirect taxes in France (value-added tax, energy tax, and so on) and of social contributions paid by the lowest earners, with a progressive income tax for the middle and upper-middle classes. For the wealthiest individuals, the progressive tax is not heavy enough to compensate for the lower indirect taxes and social contributions due to numerous exemptions for capital income. The regressivity at the top would be slightly less significant if we measured taxes paid as a function of the taxpayer’s position in the wealth distribution (rather than the income distribution) or if we combined both distributions, which would probably be the best method. Note, finally, that none of these estimates take into account the tax optimization strategies of the rich or the use of tax havens, which also leads to underestimation of the regressivity at the top.¹¹⁵

Of course, the fact that the lower and middle classes pay significant amounts of tax is not a problem in itself. If one wants to pay for a high level of social spending and educational investment, everyone must bear part of the burden. But if citizens are to consent to the taxes they must pay, the tax system must be transparent and just. If the lower and middle classes have the impression that they are paying more than the rich, there is an obvious risk that fiscal consent will be withheld and that the social contract on which social-democratic societies rest will gradually disintegrate. In this sense, the inability of social democracies to transcend the nation-state is the main weakness that is undermining them from within.

FIG. 11.19.  Profile of tax structure in France, 2018

Interpretation: In France in 2018, the overall tax rate was roughly 45 percent for the lowest income groups, 50–55 percent for the middle and upper-middle groups, and 45 percent for the highest income groups. Note: The distribution shown here is annual income for adults aged 25 to 60 working at least part time. Sources and series: piketty.pse.ens.fr/ideology.

The United States, Europe, and the Property Tax: An Unfinished Debate

We have discussed the problem of transcending the nation-state and establishing common taxes and new forms of transnational fiscal cooperation. Now we must delve into the question of just taxation. Broadly speaking, debate about just taxation since the eighteenth century has centered on the idea of progressivity, that is, taxing the poor at a low rate which increases gradually as one moves up the scale of income or wealth. Many progressive tax proposals were debated during the French Revolution.116 Progressive taxes were introduced on a large scale on all continents in the twentieth century.¹¹⁷ This general schema is important, but it does not exhaust the subject because the general idea of progressive taxation hides several different realities.

Broadly speaking, there are three major categories of progressive tax: progressive income tax, progressive inheritance tax, and progressive wealth tax. Each has its justifications and can be seen as complementary to the other two. The progressive income tax falls in principle on all income received in a given year, regardless of its source, including both labor income (wages, pensions, self-employed earnings, and so on) and capital income (dividends, interest, rents, profits, and so on). Each person can thus be taxed as a function of his or her resources at a given point in time and therefore current capacity to contribute to public expenditures. The inheritance tax, which usually includes gifts, is assessed whenever wealth is passed from one generation to the next; it can therefore be used to reduce intergenerational perpetuation of fortunes, thereby reducing the concentration of wealth.¹¹⁸ The wealth tax, which may also go by the name property tax or tax on capital or fortune, is assessed annually on the total value of goods a person owns, which can be seen as a more revealing and stable (and in some respects less manipulable) index of the taxpayer’s capacity to contribute to public expenditure than annual income. It is also the only way to achieve a permanent redistribution of wealth and true circulation of capital.

History suggests that the ideal tax system should seek to strike a balance among these three a priori legitimate forms of progressive taxation, making use of available historical knowledge. This is not an easy goal to achieve, however, because success requires broad social and political engagement with the issues, which (it is true) concern everyone but whose apparent technical complexity can lead even the best intentioned people to rely on others (who, unfortunately, may not be altogether disinterested).

In practice, we find that nearly all developed countries adopted progressive income and inheritance taxes in the late nineteenth or early twentieth centuries, with low rates at the bottom of the wealth and income distributions and rates typically as high as 60–90 percent at the very top.¹¹⁹ In contrast, countries have followed very disparate and hesitant courses with respect to the wealth tax. In a number of countries, exceptional progressive taxes on private wealth have played an important role. Experience with a permanent annual progressive wealth tax is more limited, but the topic has been much debated in both the United States and Europe; there is much to learn from these debates, as well as from occasional attempts to implement such a tax in practice. All signs are that the progressive wealth tax will become a central issue in the twenty-first century owing to the increased concentration of wealth since the 1980s.¹²⁰ Furthermore, as I will explain in detail at the end of this book, a true progressive wealth tax can be used to finance a universal capital endowment and a more egalitarian investment in education. Taken together, these measures could help to counter the inegalitarian and identarian tendencies that we see in globalized capitalism today.

The Progressive Wealth Tax, or Permanent Agrarian Reform

Let us begin by analyzing the case of exceptional taxes on private property. After World War II, a number of exceptional taxes were assessed on real estate and/or professional and financial assets for the purpose of liquidating government debt, most notably in Japan, German, Italy, France, and various other European countries. Assessed just once, these taxes applied rates close to or equal to zero on small to medium fortunes but were as high as 40–50 percent or more on the largest fortunes.121 Despite their shortcomings, including especially the virtual absence of international coordination, these levies on the whole proved to be a great success in the sense that they permitted rapid liquidation of very large debts (in a more just and controlled manner than could have been achieved through a chaotic inflationary process). What is more, the resources derived from these one-time taxes could be used to pay for postwar reconstruction and investments in the future.

In a sense, agrarian reform can also be seen as a type of exceptional tax on private wealth: an agrarian reform policy might involve the seizure of very large tracts of land (perhaps as much as 40–50 percent, often covering entire regions) in order to break it up into small parcels for redistribution to individual farmers. Unsurprisingly, agrarian reform programs frequently give rise to intense social and political struggles. Earlier, I discussed land redistribution during the French Revolution, agrarian reform in Spain, and the seizure of land owned by absentee landlords in Ireland, which was followed by a redefinition of Irish property rights in the late nineteenth and early twentieth centuries.¹²² The large-scale agrarian reforms carried out in Japan and Korea in 1947–1950 are widely considered to have been great successes. They paved the way to a relatively egalitarian distribution of farmland and were combined with social and educational investment strategies that led to subsequent economic takeoff and a consensus development strategy.¹²³ As noted earlier, moreover, the agrarian reforms carried out in India, especially western Bengal in the late 1970s and 1980s (though unfortunately more timid), nevertheless had very positive effects in terms of productivity.¹²⁴ By contrast, agrarian reform in Latin America, especially in Mexico after the revolution of 1910, ran afoul of strong resistance from landowners and very cumbersome and often chaotic political processes.¹²⁵

In general terms, an important limitation of agrarian reform (and, more broadly, of exceptional wealth taxes) is that it offers only a temporary solution to the issue of concentration of wealth and of economic and political power. That is why a permanent and annual progressive wealth tax is necessary. Although the tax rates on the highest concentrations of wealth are of course lower in the case of a permanent tax than an exceptional one, they can still be high enough to shift ownership of large amounts of wealth and prevent it from becoming reconcentrated. If such a tax were used to finance a universal capital endowment for every young adult, it would be tantamount to a permanent and continuous agrarian reform but applied to all private capital and not just farmland.

Of course, it is plausible to argue that land (or natural resources in general) is a special case when it comes to redistribution, since no one made the land or other natural resources, which can be thought of as the common wealth of humankind. Indeed, most countries have special laws pertaining to ownership of underground resources, based on different ideas of communal sharing and appropriation. If a person were to discover in his backyard a new natural resource of exceptional value, essential to preserving life on Earth, and everyone on the planet were about to die unless this new substance were shared immediately, then it is likely that the political and legal system would be amended to allow for such redistribution, whether the fortunate owner of the lucky backyard likes it or not. It would be a mistake, however, to think that such questions arise only in connection with natural resources. Suppose the same lucky individual were to awake from his siesta one day with an idea for a magical medicine that would save the planet; the case for legitimate redistribution of this miracle drug would be just as strong. The question is not so much whether an item of property is a shared natural resource or a private good developed by a single individual, as all wealth is fundamentally social. Indeed, all wealth creation depends on the social division of labor and on the intellectual capital accumulated over the entire course of human history, which no living person can be said to own or claim as his or her personal accomplishment.¹²⁶ The important question to ask is rather this: To what extent does the general interest, and in particular the interest of the most disadvantaged social groups, justify a given level of wealth inequality, regardless of the nature of the wealth in question?¹²⁷ In any case, it would be illusory to think that one could establish a just society by effecting one great agrarian reform, redistributing all land and natural resources in an equitable manner once and for all, and then allowing everyone to exchange and accumulate wealth however they please until the end of time.

In the late nineteenth century, at the height of the Gilded Age, Americans worried about the growing concentration of wealth and the increasing power of large trusts and their shareholders. The autodidact writer Henry George scored a major success with his Progress and Poverty, published in 1879, in which he denounced private ownership of land. In edition after edition over subsequent decades, millions of copies were sold as readers devoured George’s exuberant attacks on the people who had arrogated to themselves the ownership of America’s soil, which had originally been divided up according to the whims of the monarchs of England, France, and Spain and even the Pope. Even as he attacked monarchs, Europeans, and property in general, George denounced the claims of landlords to compensation, going so far as to compare them to the slaveowners who had demanded hefty compensation when the British abolished slavery in 1833–1843.¹²⁸ Yet when it came to proposing a solution for the country’s ills, George in the end showed himself to be fairly conservative. He proposed taking care of everything with a proportional tax on property in land, equal to the total rental value of the land free of any construction, drainage, or other improvement, thus allowing each person to benefit from the fruits of his own labor.¹²⁹ He did not envision any tax on bequests, thus leaving open the possibility of a future reconcentration of wealth in assets other than land. Furthermore, his proposal was impractical because it was virtually impossible to determine the value of unimproved land devoid of the many improvements introduced over the years (unless one was willing to accept a perpetually shrinking tax). This explains why no consideration was ever given to putting George’s proposal into practice. But his book contributed to a revolt against inequality that ultimately led to the adoption of a progressive income tax in 1913 and a progressive estate tax in 1916.

A half century after George published his book, the issue of a property tax returned to the agenda in the United States with the debate over proposals by Louisiana’s Democratic Senator Huey Long. Incensed by the power of stockholders in large corporations, Long tried in the early 1930s to outflank Roosevelt on his left on the issue of progressive taxes, explaining that progressive taxes on income and inheritances were not enough to solve the country’s problems. In 1934 he published a brochure laying out his plan to Share Our Wealth: Every Man a King. The heart of his program was a steeply progressive tax on all private fortunes valued at more than $1 million (around seventy times the average person’s wealth at the time) so as to guarantee each family “a share in the wealth of the United States” at least equal to a third of the national average. To complement this he also proposed higher top income and estate tax rates to pay for higher pensions for elderly people with small savings as well as reduced working hours and an investment plan aimed at restoring full employment.¹³⁰ Born into a poor white family in Louisiana, Long was a colorful character, authoritarian and controversial, who announced his intention to challenge Roosevelt in the 1936 Democratic primary. Partly in response to the pressure, Roosevelt included in the Revenue Act of 1935 a “wealth tax,” which was in fact a surtax on income with a rate of 75 percent on the highest incomes. Long’s popularity was at its height in September 1935 (with more than 8 million members of local “Share Our Wealth” committees and record audiences of 25 million for his radio broadcasts) when he was shot dead by a political opponent in the Louisiana State Capitol in Baton Rouge.

On the Inertia of Wealth Taxes Stemming from the Eighteenth Century

Let us turn now to historical experiments with annual wealth taxes. It is useful to distinguish two groups of countries. In the first group—consisting of the United States, France, and the United Kingdom—the idea of a progressive annual wealth tax long met with stiff resistance from property owners so that the proportional wealth taxes inherited from the eighteenth and nineteenth centuries were never really reformed. By contrast, in the period 1890–1910, the Germanic and Nordic countries—Germany, Austria, Switzerland, Sweden, Norway, and Denmark, the same countries that introduced power sharing between stockholders and employees—introduced a progressive annual wealth tax, usually at the same time as progressive taxes on income and inheritance.

Let’s start with the first group, especially the United States. Although the proposals of Henry George and Huey Long were never enacted, the property tax has played a central role in US fiscal history. It is one of the principal sources of funding for states and municipalities today. Of course, there are many different kinds of property tax. If assessed at a low proportional rate on all property, regardless of its value, it is not much of a threat to people of great wealth, who may well prefer it to an income tax. This is the case with the property tax in the United States as well as the land tax (contribution foncière, today’s taxe foncière) established during the French Revolution, which French property owners viewed as the ideal tax throughout the nineteenth century because its rate was low, it was minimally intrusive, and it encouraged accumulation and concentration of wealth. Along with the inheritance tax, the real estate tax remained the French government’s main source of revenue until World War I.131 The US equivalent was the property tax, which also dates from the late eighteenth century; it was the principal direct tax in the United States in the nineteenth and early twentieth centuries, with the specific feature that it was assessed by states and municipalities and not by the federal government, whose tax revenues remained limited until the creation of the federal income tax in 1913. In France, the real estate tax ceased to be used to finance the central government and became a local tax in 1914, when the income tax was established.

Both the real estate tax (taxe foncière) and the property tax, which still exist today as local taxes yielding substantial revenues (2–2.5 percent of national income in both France and the United States in the 2010s), are assessed not only on housing but also on professional equipment used as productive capital by firms, including office buildings, storage lots, warehouses, and so on.¹³² The main difference between a progressive wealth tax and the real estate tax or property tax is that the latter have always been strictly proportional. In other words, the tax rate is the same whether one owns a single house or a hundred houses.¹³³ The fact that professional assets are taxed at the level of the firm that owns and uses them (or rents them to other users) and not at the level of the shareholder who owns the firm also implies that it is never necessary to list all the properties owned by a given person in a single tax statement (which is comforting for those who own many properties, who might otherwise worry that the tax could quickly become progressive rather than proportional). The fact that the tax is local offers an additional guarantee against any effort to redistribute.¹³⁴ Note, however, that both the French real estate tax and the US property tax are based on the same fiscal philosophy, namely that wealth should be taxed as such, independent of income. No one has ever suggested that a person who owns dozens of apartment buildings or houses or lots or warehouses should be exempt from the property tax or real estate tax because he or she derives no income from the properties (because they are not rented or used). Even if the consensus is rather confused because knowledge of both the tax system and the income and wealth distributions is often highly imperfect, there is in fact a consensus that the owner of a property should either pay the property tax or real estate tax or sell the property to someone who can make better use of it.¹³⁵ In other words, the principle is that wealth should be taxed as such because it is a measure of the taxpayer’s ability to pay that is more durable and less manipulable than income.

The second essential difference between a general progressive wealth tax (ideally including all forms of property) and a real estate or property tax is that the latter leaves many types of assets untouched—especially financial assets, which constitute the lion’s share of the largest fortunes (Fig. 11.17). Of course, it is quite misleading to say that the real estate tax or the property tax falls exclusively on residential property: it also applies to offices, lots, warehouses, and other real estate owned by firms, and shareholders in these firms are therefore also affected. Still, the resulting tax rate on financial assets is much lower than the tax rate on real estate, partly because financial assets invested abroad or in government bonds are totally exempt¹³⁶ and partly because many things that constitute the value of investments in domestic firms escape all or part of the tax (including machinery and equipment as well as intangible assets such as patents).¹³⁷ This hodgepodge is not the result of any preconceived plan. It is the fruit of particular historical processes and specific political-ideological mobilizations (or the absence thereof) around the issue of a wealth tax.

Note, moreover, that the US property tax, as its name suggests, has at times been more ambitious than the French real estate tax. There is considerable variation in the nature of the various property taxes assessed across the United States. Depending on the state or municipality, the property tax may apply not only to “real property” (such as land and buildings, from vacant lots to homes, apartment buildings, office buildings, warehouses, and so on) but also to “personal property” (including cars, boats, furniture, cash, and even financial assets). At the moment, the most common type of property tax applies only to real property, but this has not always been the case.

In this connection, the very lively debates that took place in Boston in the late nineteenth century, recently studied by Noam Maggor, are particularly interesting.¹³⁸ At the time, the property tax levied in the capital of Massachusetts, where much of the country’s high financial and industrial aristocracy resided, fell on both real and personal property, including the financial portfolios of the Boston elite, which were full of investments in other US states and foreign countries. Wealthy Bostonians were up in arms against this tax. They pointed out that they were already paying heavy taxes in the places where their capital was invested, and they demanded that the property tax be limited to real estate, which they saw as a nonintrusive index of their capacity to pay; this was the way things were done in Europe, most notably France.¹³⁹ To support their case, they called upon the help of economists and tax experts from nearby universities, especially Harvard, who praised the wisdom of European tax systems. Thomas Hills, the chief tax assessor of the city of Boston from 1870 to 1900, saw things differently, however. In 1875 he published a white paper showing that real estate accounted for only a tiny fraction of the wealth of the richest Bostonians and that exempting their financial assets from taxation would result in an enormous revenue loss. This would do much harm to the city, which was expanding rapidly at the time, with new waves of Irish and Italian immigrants filling its suburbs, requiring major public investments.¹⁴⁰ The political balance of power at the time was such that the broad wealth tax was maintained. But the debate continued in the 1880s and 1890s, and the wealthy finally carried the day in the early 1900s as various types of personal property were gradually removed from the purview of the property tax. Exemptions were granted to one type of financial asset after another, until in 1915 the Boston property tax was finally limited to real property only.¹⁴¹

These debates are particularly interesting because they illustrate the variety of possible trajectories and switch points. A key element in the controversy was the lack of cooperation between states and municipalities, which refused to share information about who owned what. One way to overcome these contradictions would have been (or might be in the future) to levy a coordinated property tax at the federal level and transform it into a true progressive tax on individual net worth. The choice that the United States made in 1913–1916 was different: the federal government concentrated on federal income and estate taxes, while the annual wealth tax (generally limited to real estate and assessed at a flat rate) was left to states and municipalities.¹⁴²

In the end, both the US property tax and the French real estate tax, neither of which has been comprehensively reformed since the eighteenth century (that is, since the proprietarian-censitarian era), remain today as egregiously regressive taxes, which simply take no account of financial assets and liabilities. Assume for instance that the property tax (or the real estate tax) due for a house worth $300,000 is $3,000—that is, 1 percent of the value of the property. Consider now a person who owns this house but with a mortgage of $270,000 so that his or her net worth is only $30,000. For her, the tax payment will be 10 percent of her net wealth ($3,000 divided by $30,000). Imagine now someone who owns a stock portfolio worth $2.7 million together with this same house (and no mortgage), so that his net worth is $3 million. With the property tax system current applied in the United States, or the land tax system (taxe foncière) applied in France, this person would still pay the same tax ($3,000), although this makes only 0.1 percent of his net worth ($3,000 divided by $3 million). Such a regressive tax system is hard to justify and contributes to undermining fiscal consent and making economic justice seem impossible. It is also striking to discover that surveys on this subject show that most people would prefer a mixed tax system based on both income and net wealth (including both real estate and financial assets, which respondents logically regard as equivalent in terms of fiscal justice).¹⁴³ The only possible (but relatively nihilistic and factually false) justification for not taking financial assets and liabilities into account is that people with financial assets have so many opportunities for tax avoidance that there is no choice but to exempt them entirely from the wealth tax. In fact, financial institutions have long been required to report interest and dividends on financial assets, and there is no reason why they should not be required to report the value of the assets themselves (and not just the income that flows from them). This could be extended to the international level by amending existing treaties concerning capital flows.¹⁴⁴ Remember, too, that the exceptional taxes on private wealth successfully levied in Germany, Japan, and many other countries after World War II obviously applied to financial assets. It would have been totally incongruous to have proceeded otherwise, since the purpose of these taxes was to tap the wealth of the well-to-do.

Collective Learning and Future Prospects for Taxing Wealth

All signs are that this long history is far from over. The existing system is a consequence of sociopolitical processes shaped primarily by the balance of political-ideological power and the mobilization capacities of the various parties in contention, and it will continue to evolve in the same way. The key point is this: the very sharp rise in wealth inequality in the United States in the period 1980–2020, combined with mediocre growth, has created the conditions for a challenge to the conservative ideological turn of the 1980s. Since the mid-2010s, leading Democrats have increasingly called for a return to 70–80 percent top marginal rates on the highest incomes and largest fortunes. The most outspoken of all was Bernie Sanders, who narrowly lost to Hillary Clinton in the 2016 Democratic presidential primary: he proposed a top marginal rate of 77 percent on the largest estates (in excess of $1 billion).

In anticipation of the 2020 presidential election, some Democratic candidates have begun to speak of creating the first US wealth tax, for instance, with a rate of 2 percent on fortunes of $50 million to $1 billion and 3 percent on wealth beyond $1 billion, to quote Elizabeth Warren’s proposal of early 2019.145 The Warren plan includes an exit tax of 40 percent for anyone who decides to give up US citizenship and transfer his or her wealth to another country. The tax would apply to all assets, with no exemptions, and impose dissuasive sanctions on individuals and governments unwilling to share relevant information about assets held abroad.

It is impossible to say at this stage if or when such a proposal might become law and what form it would take if it did. The suggested 3 percent rate on fortunes greater than $1 billion suggests a clear intention to put wealth back into circulation. This rate implies that a static fortune of $100 billion would return to the community after a couple of decades. In other words, the largest fortunes would only temporarily reside in the hands of any given individual. In view of the average rate of increase of large fortunes, however, one would need to consider higher rates on larger wealth holdings: at least 5–10 percent or maybe several dozens percent on multibillionaires so as to facilitate a fast renewal of fortune and power.¹⁴⁶ It might also be preferable to link the rates on the largest fortunes to the much-needed reform of the property tax (with the possibility of reducing the property tax on people with mortgages or seeking to purchase a first home).¹⁴⁷ In any case, these debates are far from over, and their outcome will depend largely on the ability of participants to relate recent developments to past experiences.

In other countries we find a similar need to place current debates in historical perspective. In France as in the United States, there were numerous debates in the late nineteenth and throughout the twentieth centuries about establishing a true progressive wealth tax. There was discussion before World War I, indeed early in 1914, but by the summer of that year the emergency had arrived, and in view of the ideological resistance aroused by the idea of an annual wealth tax, the Senate opted for a general income tax instead. In the 1920s, debate within the Cartel of the Left led nowhere, both because the Radicals did not wish to worry smallholders and because the Socialists were more interested in nationalizations than in tax reform. Indeed, this ideological bias acted as a constant brake on any socialist or social-democratic thinking about a progressive wealth tax: for centrist parties the idea was terrifying, while for parties farther to the left, attached to the idea of state ownership of the means of production, it lacked the power to mobilize the masses. In 1936, at the time of the Popular Front, the Communists agreed to participate in the government; they favored a progressive wealth tax with rates ranging from 5 percent on fortunes of 1 million francs to 25 percent on fortunes larger than 50 million francs (respectively, ten and 500 times the average wealth at the time). But the parliamentary majority depended on the Radicals, who refused to vote for this bill, which they saw as a Trojan horse for socialist revolution. Many other proposals were floated subsequently, especially by the General Confederation of Labour (CGT) in 1947 and by Socialist and Communist deputies in 1972.

Finally, after the Socialists won the presidential and legislative elections in 1981, a “tax on large fortunes” (IGF) was passed by the Socialist-Communist majority, but in 1986 it was repealed by the Gaullist-liberal majority and then subsequently restored by the Socialists as a “solidarity tax on wealth” (ISF) after the 1988 elections.¹⁴⁸ Later, I will come back to the way the government elected in 2017 set about replacing the ISF in 2018 with a tax on real estate (IFI), with complete exemption for financial assets and therefore the bulk of the largest fortunes.¹⁴⁹ At this point, note simply that the very strenuous opposition aroused by this reform suggests that the story is far from over. In any case, bear in mind that the IGF (1982–1986) and ISF (1989–2017) never concerned more than a small minority of taxpayers (less than 1 percent of the population) and that rates were very low (generally 0.2 to 1.5–2 percent), with many exemptions. The result was that the real estate tax (taxe foncière), which in broad outline remained more or less unchanged since the 1790s, continued to be the main French wealth tax.¹⁵⁰

Intersecting Trajectories and the Wealth Tax

In the United Kingdom, Labour governments led by Harold Wilson and later James Callaghan came close to passing a progressive wealth tax in 1974–1976. Urged on by economist Nicholas Kaldor, Labour concluded in the 1950s and 1960s that the tax system based on progressive income and estate taxes needed to be completed by an annual progressive tax on wealth for reasons of both justice and efficiency. In particular, this seemed to be the best way to gather information about the distribution of wealth and its evolution in real time and thus to combat avoidance of the estate tax by way of trusts and similar devices. Labour’s platform in the successful 1974 election campaign included a progressive tax with a rate of 5 percent on the largest fortunes. But the plan ran into trouble, not only because of opposition from the treasury but also because of the consequences of the oil crisis and the ensuing inflation and monetary crisis of 1974–1976 (which led to IMF intervention in 1976), and was ultimately abandoned.151

The United Kingdom thus stands with the United States as the country that has achieved the highest level of fiscal progressivity with respect to income and inheritance yet has never experimented with an annual progressive wealth tax. Recent British experience with the so-called mansion tax bears mention, however. Although the British system of local taxing of houses is particularly regressive, the country does stand out for a strongly progressive system of taxes on real estate transactions. The tax paid on a real estate transaction is zero for transactions up to £125,000, 1 percent for transactions between £125,000 and £250,000, and 4 percent on transactions above £500,000. In 2011, a new 5 percent tax was created for sales of properties with a value greater than £1 million (“mansions”).¹⁵² It is interesting to note that this 5 percent tax, introduced by a Labour government, was at first harshly criticized by Conservatives, who, after coming to power themselves, enacted a 7 percent transaction tax on properties worth more than £2 million. This shows that in a context of rising inequality, especially when wealth is highly concentrated and it is difficult for many people to gain access to the housing market, the need for a more progressive wealth tax can make itself felt across traditional party lines. It also points to the need for a comprehensive reassessment of property and wealth taxes: instead of such high transaction taxes, it would be more just and efficient to have an annual wealth tax with lower rates but based on total asset holdings of all types.

Finally, I should mention the Germanic and Nordic countries, which for the most part did not go as far as the United Kingdom or United States in imposing progressive income and estate taxes but were early to complement those two taxes with annual progressive wealth taxes. Prussia established an annual progressive tax on total wealth (including land, buildings, and professional and financial assets, net of debt) as early as 1893, shortly after it enacted a progressive income tax in 1891. Saxony did the same in 1901, and other German states followed suit, leading to the enactment of a federal wealth tax in 1919–1920.¹⁵³ Sweden enacted a progressive wealth tax in 1911, again coinciding with the progressive income tax reform.¹⁵⁴ In other countries in this group (such as Austria, Switzerland, Norway, and Denmark), similar systems combining progressive taxes on income, wealth, and inheritance were put in place in the same period, generally between 1900 and 1920. Note, however, that these wealth taxes, which generally applied to barely 1–2 percent of the population with rates ranging from 0.1 to 1.5–2 percent (and up to 3–4 percent in Sweden in the 1980s), played a significantly less important role than the income tax.

It is also very important to note that these taxes were repealed in most of these same countries in the 1990s or early 2000s (with the exception of Switzerland and Norway, where they remain in place), partly because of tax competition (in a period marked by liberalized capital flows in Europe after the late 1980s) and an ideological context marked by the conservative revolution in the United States and United Kingdom and the fall of the Soviet Union. In addition to these well-known factors, we should also note the decisive (and instructive) importance of errors in the initial design. Conceived before World War I, at a time when the gold standard was still in effect and inflation was unknown, these Germano-Nordic wealth taxes were mostly based not on the market value of real and financial assets (with an index to prevent unduly abrupt increases or decreases in the amount of tax assessed) but rather on cadastral values—that is, values periodically recorded at intervals of, say, ten years, when all property was inventoried. While such a system is viable in times of zero inflation, it was quickly rendered obsolete by the very high inflation seen in the wake of the two world wars and in the postwar period. Such inflation is already the source of serious problems for a proportional wealth tax (such as the French real estate tax and the US property tax). In the case of a progressive tax, where the problem is to determine who is above each threshold of taxation and who is not, relying on values recorded in the relatively distant past on the basis of comparable local or neighborhood prices is untenable. It was because of this inequity that the German constitutional court suspended the wealth tax in 1997: taxpayers were no longer equal before the law because of inflation. The political coalitions that have held power in Berlin since then have had other priorities than reforming the wealth tax, for reasons we will come back to later.

Finally, note the specific role of the Swedish banking crisis of 1991–1992 in the country’s political-ideological evolution (which had a significant impact on other countries, given the emblematic role of Swedish social democracy). The extreme gravity of the crisis, in which the main Swedish banks nearly went under, raised questions about banking regulation, monetary policy, and the role played by capital flows. This led to a general critique of the alleged excesses of Sweden’s social and fiscal model and, more broadly, to a sense that the country found itself in a very precarious position in a world that had gone over to globalized financial capitalism. For the first time since 1932 the Social Democrats were driven from power and replaced by the Liberals, who in 1991 exempted interest and dividends from taxation and strongly reduced the progressivity of the progressive wealth tax. This tax was finally abolished by the Liberals in 2007, two years after the Social Democrats abolished the estate tax, which may be surprising but reflected the degree to which a country the size of Sweden can be gripped by the fear of fiscal competition as well as the perception that the Swedish egalitarian model is so firmly established that it no longer needs such institutions. There is nevertheless reason to believe that such radical reform of tax policy can have fairly substantial inegalitarian consequences in the long run; this may also help to explain why Swedish Social Democrats appeal more and more to the relatively well-off and less and less to their traditional popular electorate.¹⁵⁵

We will come back to these questions in Part Four, when we examine the evolution of voting patterns and of political conflict in the major parliamentary democracies. At this stage, several lessons can be drawn. Broadly speaking, social democracy, for all its successes, has suffered from a number of intellectual and institutional shortcomings, especially with respect to social ownership, equal access to education, transcendence of the nation-state, and progressive taxation of wealth. On the last point, we have traced a number of trajectories, with multiple switch points. Policies have been highly inconsistent, and there has been too little sharing of experiences across countries. No doubt this is partly because political movements and citizens have not fully engaged with these issues. Recent developments reflect considerable hesitation: on the one hand, rising inequality of wealth clearly calls for the development of new forms of fiscal progressivity; on the other hand, there is a widespread perception that pitiless tax competition justifies less progressivity, even if it contributes to greater inequality.

In reality, refusing to have a rational debate about a progressive wealth tax and pretending that it is wholly impossible to make the largest fortunes contribute to the common good and that the lower and middle classes have no choice but to pay in their place strike me as a very dangerous political choice. All history shows that the search for a distribution of wealth acceptable to the majority of people is a recurrent theme in all periods and all cultures. The thirst for fiscal justice grows stronger as people become better educated and better informed. It would be surprising if things were different in the twenty-first century and these debates were not once again central, especially at a time when the concentration of wealth is increasing. To prepare for this, it is best to begin by delving into past debates—the better to move beyond them. If we are not willing to do this, we risk making people wary of any ambitious effort to achieve fiscal and social solidarity and encouraging instead social division and ethnic and national hostility.

---

1.     1.  See Figs. 10.1–10.2, and the online appendix (piketty.pse.ens.fr/ideology), Figs. S10.1–S10.2.

2.     2.  See Chap. 5.

3.     3.  See Figs. 10.14–10.15.

4.     4.  The Fabian Society, founded in 1884 to promote a gradual reformist transition to democratic socialism without a revolutionary conflagration (hence the choice of namesake, the Roman general Fabius, an adept of the war of attrition in the third century BCE), is still today one of the “socialist societies” affiliated with the Labour Party. The Fabians Beatrice and Sidney Webb founded the London School of Economics in 1895; William Beveridge directed it from 1919 to 1937. On the intellectual history of the Labour Party, see M. Bevir, The Making of British Socialism (Princeton University Press, 2011).

5.     5.  The Radical Party (actually called the Republican, Radical, and Radical-Socialist Party) included the more radical republicans from the first decades of the Third Republic. It championed “social reform with respect for private property” and opposed nationalizations. More conservative than the Socialists and Communists on socioeconomic issues, it lost its central place in the French political system after World War II. Until 1971, the PS was generally known as the SFIO (from the French initials for French Section of the Workers’ International).

6.     6.  For a classic study of social-democratic models and of the French case, see A. Bergounioux and B. Manin, La social-démocratie ou le compromis (Presses universitaires de France, 1979). On the diversity of European social democracy, see H. Kitschelt, The Transformation of European Social Democracy (Cambridge University Press, 1994). See also G. Esping-Andersen, The Three Worlds of Welfare Capitalism (Princeton University Press, 1990).

7.     7.  See Fig. 10.14.

8.     8.  The Affordable Care Act, or “Obamacare” (2010), was intended to make private insurance compulsory with subsidies for those who needed them. Implementation has proved difficult, largely because of opposition from Republican states and because Supreme Court decisions make it difficult for the federal government to impose social programs on the states. More ambitious proposals (including “Medicare for All”) are now being advocated by a growing number of Democratic leaders.

9.     9.  In Argentina and to a lesser degree in Brazil (where inequality was much greater), tax receipts in 1950–1980 reached a level intermediate between the United States and Europe (30–40 percent of national income). By contrast, Mexico and Chile continued to take in much less in taxes (less than 20 percent of national income). See the online appendix and the work of M. Morgan.

10.   10.  Europe, for the purposes of Fig. 11.1, includes both the western and eastern parts of the continent (a total of 540 million people). If we focus on Western Europe, the difference with the United States is even clearer. See Fig. 12.9.

11.   11.  See Figs. 10.1–10.2.

12.   12.  See Chap. 17.

13.   13.  For a recent analysis of German co-determination, see E. McGaughey, “The Codetermination Bargains: The History of German Corporate and Labour Law,” Columbia Journal of European Law, 2017. See also S. Silvia, Holding the Shop Together: German Industrial Relations in the Postwar Era (Cornell University Press, 2013). The German Mitbestimmung system can be translated as “codetermination” or “co-management.” The latter seems more expressive in English.

14.   14.  For example, in 2019, the state of Lower Saxony held 13 percent of Volkswagen’s shares, and the firm’s statutes guarantee it 20 percent of the voting rights.

15.   15.  Article 155: “The distribution and utilization of the land are controlled by the state in such a way as to prevent abuses and achieve the objective of ensuring that every family has a healthy place to dwell, corresponding to its needs.… Land required to satisfy needs resulting from a shortage of housing … or to develop agriculture may be expropriated.”

16.   16.  Article 15: “The soil and land, natural resources and the means of production, may be placed under a regime of collective ownership or other forms of collective management by a law.”

17.   17.  On these debates, see McGaughey, “The Codetermination Bargains.” See also C. Kerr, “The Trade Union Movement and the Redistribution of Power in Postwar Germany,” Quarterly Journal of Economics, 1954, pp. 556–557.

18.   18.  More precisely, the current law provides two seats in firms of 25–1,000 employees and three seats in firms with more than 1,000 employees, which, given the size of boards of directors, usually corresponds to about a third of the seats in both cases.

19.   19.  In both cases, workers are entitled to half the number of directors elected by shareholders or exactly one-third of the total. Firms with thirty to fifty employees are also entitled to a paid director in Norway. See the online appendix.

20.   20.  See E. McGaughey, Do Corporations Increase Inequality? (Transnational Law Institute Think! Paper 32, King’s College London, 2016). Later we will discuss other determinants of executive pay, especially the degree of fiscal progressivity.

21.   21.  On the slow constitution of a “salaried worker” status and veritable “salaried society,” (société salariale), see R. Castel, Les métamorphoses de la question sociale (Folio, 1995), pp. 594–595. It was not until 1969–1977, for example, that monthly pay became the norm in France. See also R. Castel and C. Haroche, Propriété privée, propriété sociale, propriété de soi (Pluriel, 2001).

22.   22.  See McGaughey, Do Corporations Increase Inequality?

23.   23.  Among the many works devoted to this history, see S. Bartolini, The Political Mobilization of the European Left, 1860–1980: The Class Cleavage (Cambridge University Press, 2000). For an analysis of European worker networks and early forms of mutual aid and strike funds dating back to the 1860s, especially in connection with the First International, see N. Delalande, La lutte et l’entraide. L’âge des solidarités ouvrières (Seuil, 2019).

24.   24.  More precisely, the law created one paid directorship when the board of directors consisted of fewer than twelve members and two seats when the board was larger. The 2013 law was to apply to firms with more than 5,000 employees in France or more than 10,000 throughout the world; these thresholds were reduced in 2015 to firms employing more than 1,000 workers in France or 5,000 throughout the world.

25.   25.  François Mitterrand, in a “letter to the French” written in 1988, promised “neither-nor” (neither new nationalizations nor new privatizations). His reelection hinged on this promise of social peace, coupled with his denunciation of police violence against student demonstrators (opposed to the increase in registration fees) and of the suppression of the wealth tax.

26.   26.  The looser ties between unions and parties in France are often attributed to the fact that democracy and universal suffrage preceded social democracy and trade unionism in France (whereas the opposite is to a large extent true in Germany and the United Kingdom); hence, there is a certain wariness on the part of the unions (which were long subject to the ban on professional organizations and guilds enacted in 1791 and were not legalized until 1883) toward parliament and government. See, for example, M. Duverger, Les partis politiques (Armand Colin, 1951), pp. 33–34.

27.   27.  In fact, the Clause IV of 1918 opened the way to various forms of ownership, since it stated the party’s objective as follows: “To secure for the workers by hand or by brain the full fruits of their industry and the most equitable distribution thereof that may be possible upon the basis of the common ownership of the means of production, distribution and exchange, and the best obtainable system of popular administration and control of each industry or service.” The clause adopted in 1995 reads as follows: “The Labour Party is a democratic socialist party. It believes that by the strength of our common endeavour we achieve more than we achieve alone, so as to create for each of us the means to realise our true potential and for all of us a community in which power, wealth and opportunity are in the hands of the many, not the few, where the rights we enjoy reflect the duties we owe, and where we live together, freely, in a spirit of solidarity, tolerance and respect.”

28.   28.  The same was true before World War I, in particular when Bernstein’s “revisionist” faction was outvoted at the Congress of Hanover in 1899. See Chap. 10.

29.   29.  In the decisive 1930–1932 elections, the SPD (Social Democrats) and KPD (Communists) together won more votes than the NSDAP (Nazis): 37 percent of the votes and 221 seats for the SPD and KPD in the November 1932 elections versus 31 percent of the votes and 196 seats for the NSDAP. But the inability of the two left-wing parties to unite allowed the Nazis to take power.

30.   30.  On the intellectual context, see McGaughey, “The Codetermination Bargains.”

31.   31.  The French Socialists in 1997–2002 and the British Labour Party in 1997–2010 also pursued other reforms, including the reduction of the work week (for the French) and educational reform (for the British). But regarding the key issues of the property regime and the international financial regime, both the Socialists and the Labourites adopted a fairly conservative stance.

32.   32.  By contrast, Germany’s economic difficulties, associated with reunification in the 1990s and early 2000s, probably slowed the diffusion of co-management.

33.   33.  The so-called Draft Fifth Company Law Directive also suffered from the fact that the 1972 version favored the German model of dual governance. The 1983 and 1988 versions dropped this but preserved strong representation of workers on corporate boards (from a third to a half of the seats), without success. See the online appendix.

34.   34.  On this proposal and the history of these debates, see E. McGaughey, “Votes at Work in Britain: Shareholder Monopolisation and the ‘Single Channel,’ ” Industrial Law Journal, 2018.

35.   35.  Union and employer representatives had clashed within the Bullock Commission, and it was the jurists and academics who cast the deciding votes in favor of the majority report.

36.   36.  See Chap. 17.

37.   37.  It could be problematic unless the procedures for appointing public board members are spelled out and steps are taken to ensure that the system functions in a satisfactory manner (which is not necessarily impossible but would require concrete historical experimentation).

38.   38.  In 2017, 21 percent of private-sector workers in France worked for firms with fewer than ten employers, 40 percent for firms with 10–250 employees, 26 percent for firms with 251–5,000 employees, and 13 percent for firms with more than 5,000 employees. Self-employment accounted for 12 percent of employment, compared with 21 percent in the private sector (state, municipalities, and hospitals) and 67 percent in the private sector (all types of firms and associations combined). The figures for other European countries are comparable. See the online appendix.

39.   39.  Although limited to large firms, these bills were novel in the American context. The “Reward Work Act” bill (March 2018) would have required listed firms to set aside at least a third of their board seats for worker representatives elected on the basis of “one worker, one vote.” The “Accountable Capitalism Act” bill (August 2018) envisioned 40 percent of “employee directors” for the largest firms (with annual revenues of more than $1 billion), whether listed or not, and would also have required a three-fourths majority of the board to approve political donations (since the Supreme Court had ruled that corporate donations could not be forbidden). Neither of these bills has been adopted so far, but the fact that they are openly discussed in the US Congress is already a major novelty.

40.   40.  See J. Blasi, R. Freeman, and D. Kruse, The Citizen’s Share: Putting Ownership Back into Democracy (Yale University Press, 2013). See also J. Ott, When Wall Street Met Main Street: The Quest for an Investors’ Democracy (Harvard University Press, 2011).

41.   41.  See Chap. 5.

42.   42.  E. McGaughey, “A Twelve-Point Plan for Labour, and a Manifesto for Labour Law,” Industrial Law Journal, 2017. See also K. Ewing, J. Hendy, and C. Jones, eds., A Manifesto for Labour Law (Institute of Employment Rights [IER], 2016); J. Hendy, K. Ewing, and C. Jones, eds., Rolling Out the Manifesto for Labour Law (IER, 2018), pp. 32–33.

43.   43.  See also the propositions of I. Ferreras, Firms as Political Entities: Saving Democracy through Economic Bicameralism (Cambridge University Press, 2017), who envisions firms governed by an assembly of workers and an assembly of shareholders, with neither taking precedence over the other, in the manner of democracies with bicameral legislatures. The advantage would be to encourage actors to reach mutually advantageous compromises; the risk is deadlock.

44.   44.  See D. Cole and E. Ostrom, eds., Property in Land and Other Resources (Lincoln Institute of Land Policy, 2011). See also F. Graber and F. Locher, eds., Posséder la nature. Environnement et propriété dans l’histoire (Editions Amsterdam, 2018).

45.   45.  This was done in the eighteenth and nineteenth centuries in both political assemblies and shareholder assemblies. One could apply the “one share, one vote” principle across the board, or alternatively, one could group shareholders according to their wealth or capital and define several classes of voting rights. See Chap. 5.

46.   46.  J. Cagé, Saving the Media, trans. A. Goldhammer (Harvard University Press, 2016).

47.   47.  On the debates surrounding autogestion, see, for example, P. Rosanvallon, Notre histoire intellectuelle et politique, 1968–2018 (Seuil, 2018), pp. 56–77.

48.   48.  See Chap. 13.

49.   49.  The series used are those of the OECD and the US Bureau of Labor Statistics (BLS). To simplify, an annual duration of 2,000 hours corresponds to 40 hours per week × 50 weeks (two weeks of vacation), while an annual duration of 1,500 hours corresponds roughly to 35 hours per week × 43 weeks (seven weeks of vacation). The average duration in Germany was 1,370 hours per job in 2015 (versus 1,470 in France, 1,680 in the United Kingdom, and 1,790 in the United States), which also reflects the extent of part-time employment. See the online appendix. Available historical research indicates that working durations were significantly shorter in the United States between 1870 and 1914 and then converged with Europe during the interwar period before exceeding European levels after 1970. See M. Huberman and C. Minns, “ ‘The times they are not changing’: Days and Hours of Work in Old and New Worlds, 1870–2000,” Explorations in Economic History, 2007.

50.   50.  For example, the reduction of the legal work week in France to thirty-five hours between 1997 and 2002 coincided with increasingly flexible hours for the lowest-paid workers and a prolonged freeze in disposable income; it turned out to be of greater benefit to management than to workers (with managers receiving additional vacation time). In the United Kingdom and United States, the relatively small decrease in working hours in recent decades coincided with a sharp decline in trade union membership (and failure of government to compensate) and an especially sharp increase in wage inequality. A full analysis of the different national trajectories with respect to work-time reduction and restructuring is beyond the scope of this book.

51.   51.  The reason for this is that it is difficult for an individual worker to negotiate over working hours along with a tendency to aim for a certain standard of living: no worker wants to be the first to sacrifice his disposable income; even if collectively, workers would prefer more leisure time. The decrease observed in the number of hours worked by the self-employed following legislation applicable only to wage workers suggests that the second factor is of some importance. The data available to resolve these questions are imperfect, however.

52.   52.  See the online appendix for the various sources used. The schooling rates cited here are taken from the data in J. Lee and H. Lee, “Human Capital in the Long-Run,” Journal of Development Economics, 2016, which relies on many earlier works.

53.   53.  “But, from the beginning, the originality of American civilization was most clearly apparent in the provisions made for public education.… Municipal magistrates were made responsible for seeing that parents sent their children to school. They were authorized to impose fines on any parent who refused to do so. If resistance continued, society, putting itself in the place of the family, might seize the child and deprive its father of natural rights so egregiously abused.” Alexis de Tocqueville, Democracy in America, trans. A. Goldhammer (Library of America, 2004), pp. 46–47.

54.   54.  For detailed data by state, see S. Engerman and K. Sokoloff, “The Evolution of Suffrage Institutions in the New World,” Journal of Economic History, 2005, p. 906, table 2.

55.   55.  See Fig. 5.3. The contrast is particularly striking with Latin America (esp. Brazil, Mexico, Argentina, and Chile), where the participation rate of adult white males in elections remained below 10–20 percent until 1890–1910. See Engerman and Sokoloff, “The Evolution of Suffrage Institutions in the New World,” pp. 910–911, table 3. On the slow transition from a mercantilist-absolutist ideology to a proprietarian-censitary ideology among Argentine elites in the nineteenth century related to the recomposition of wealth (from silver export to a large agricultural surplus), see J. Adelman, Republic of Capital: Buenos Aires and the Legal Transformation of the Atlantic World (Stanford University Press, 1999). On the absence of an inequality leveling period in Latin America during the twentieth century (similar to Europe or the United States), see J. Williamson, “Latin American Inequality: Colonial Origins, Commodity Booms or a Missed Twentieth Century Leveling,” Journal of Human Development and Capabilities, 2015.

56.   56.  See, in particular, C. Goldin, “America’s Graduation from High School: The Evolution and Spread of Secondary Schooling in the Twentieth Century,” Journal of Economic History, 1998; C. Goldin, “The Human Capital Century and American Leadership: Virtues of the Past,” Journal of Economic History, 2001.

57.   57.  See the online appendix. The available sources are imperfect, but the orders of magnitude and especially the gaps between countries are well established.

58.   58.  See Chap. 9.

59.   59.  After the expulsion of the Protestants in 1685, a first royal edict of 1698 required every parish to have a school to teach the catechism and develop a written religious culture. The use of tax money to pay for compulsory education was approved in 1792–1793 but never applied. In 1883, local governments (communes) were required to pay teachers, with supplementary funding by the state after 1850; the state took full responsibility for paying teachers in 1889 (the same year in which the practice of having priests issue certificates of morality to schoolteachers was ended). See F. Furet and J. Ozouf, Lire et écrire. L’alphabétisation des Français de Calvin à Jules Ferry (Editions de Minuit, 1977). See also A. Prost, Histoire de l’enseignement en France, 1800–1967 (Armand Colin, 1968).

60.   60.  See D. Cannadine, Victorious Century: The United Kingdom, 1800–1906 (Viking, 2017), pp. 257, 347.

61.   61.  See P. Lindert, Growing Public: Social Spending and Economic Growth since the Eighteenth Century (Cambridge University Press, 2005), vol. 2, pp. 154–155.

62.   62.  See Fig. 10.15.

63.   63.  Total educational expenditure in the United Kingdom today is close to that of other European countries (such as Germany, France, and Sweden): around 6 percent of national income. See the online appendix.

64.   64.  See Chap. 6.

65.   65.  In the next chapter we will also see that inequality in Europe is significantly lower than the United States even if one includes Eastern Europe. See Fig. 12.8.

66.   66.  M. Bertrand and A. Morse, “Trickle-Down Consumption,” Review of Economics and Statistics, 2016; M. Kumhof, R. Rancière, and P. Winant, “Inequality, Leverage and Crises,” American Economic Review, 2015. On the history of credit regulation in the United States, see L. Hyman, Debtor Nation: The History of America in Red Ink (Princeton University Press, 2011); L. Hyman, Borrow. The American Way of Debt. How Personal Credit Created the American Middle Class and Almost Bankrupted the Nation (Vintage Books, 2012).

67.   67.  The results summarized here were obtained by combining a variety of available sources: tax records, household surveys, and national accounts. See T. Piketty, E. Saez, and G. Zucman, “Distributional National Accounts: Methods and Evidence from the United States,” Quarterly Journal of Economics, 2018. See the online appendix for the detailed series.

68.   68.  The principal cash transfer to the poor (excluding food stamps) is the Earned Income Tax Credit (EITC), which is a negative tax similar to the prime d’activité in France, the purpose of which is to increase the disposable income of low-paid workers. The extension of the EITC and tax decreases subsequent to the 2008 crisis explain why post-tax-and-transfer income rose slightly above pre-tax-and-transfer income. As in other countries, the pre-tax-and-transfer income considered here includes public pensions (minus the corresponding contributions), without which retiree income would be artificially low. If we look only at the working-age population, we find the same stagnation of the average income of the bottom 50 percent over the past half century. See Piketty, Saez, and Zucman, “Distributional National Accounts,” p. 585, fig. 4. Furthermore, the decrease in the progressivity of the income tax implies that the gap between effective tax rates on the bottom 50 percent and top 1 percent has greatly shrunk compared with the period 1930–1970. See Piketty, Saez, and Zucman, “Distributional National Accounts,” p. 599, fig. 9a, and this work, Fig. 10.13.

69.   69.  One might also include other transfers in kind (such as spending on education and law enforcement), but it then becomes even more difficult to impute and interpret them in a satisfactory way. See the online appendix for detailed results.

70.   70.  As for the United States, the estimates for France were obtained by combing a variety of available sources: tax records, household surveys, and national accounts. See A. Bozio, B. Garbinti, J. Goupille-Lebret, M. Guillot, and T. Piketty, “Inequality and Redistribution in France 1990–2018: Evidence from Post-Tax Distributive National Accounts (DINA),” WID.world, 2018. The results are identical with other indicators of inequality (such as the Gini coefficient) or if we separate out different age groups (by excluding retirees, for example). See the online appendix for detailed series.

71.   71.  The results given here confirm the importance of the notion of “predistribution” (see M. O’Neill and T. Williamson, “The Promise of Predistribution,” Policy Network, 2012; A. Thomas, Republic of Equals: Predistribution and Property-Owning Democracy [Oxford University Press, 2017]). Note, however, that this idea has sometimes been instrumentalized to minimize the importance of redistribution and especially progressive taxes (which was not the intention of its promoters). By contrast, I emphasize the usefulness of progressive taxation (with rates as high as 70–90 percent on astronomical incomes) as one of the most important institutions for influencing “predistribution.” I develop this point further in this and subsequent chapters.

72.   72.  For an analysis of the role of pay scales (and especially minimum and maximum wages) in securing workers and increasing their investment in the firm, especially where the bargaining power of employers is strong, see T. Piketty, Capital in the Twenty-First Century (Harvard University Press, 2014), chap. 9, pp. 307–314. See also H. Farber, D. Herbst, I. Kuziemko, and S. Naidu, Unions and Inequality Over the Twentieth Century: New Evidence from Survey Data (Princeton University, Working Papers 620, 2018).

73.   73.  See E. Derenoncourt and C. Montialoux, Minimum Wages and Racial Inequality (Harvard University, Working Paper, 2018).

74.   74.  In the late 1960s, the United States had the highest real minimum wage in the world. In the late 2010s, the minimum wage is significantly higher in Germany, the United Kingdom, France, Holland, Belgium, Australia, and Canada. See L. Kenworthy, Social-Democratic Capitalism (Oxford University Press, 2019), p. 206, fig. 7.12. The Nordic countries continue to rely on wage bargaining.

75.   75.  See Fig. 10.11.

76.   76.  See T. Piketty, E. Saez, and S. Stantcheva, “Optimal Taxation of Top Labor Incomes: A Tale of Three Elasticities,” American Economic Journal: Economic Policy, 2014.

77.   77.  See Piketty, Saez, and Stantcheva, “Optimal Taxation,” in particular, figs. 3, 5, and A1 and tables 2–5.

78.   78.  See, for example, M. Pursey, “CEO Pay and Factor Shares: Bargaining Effects in US Corporations, 1970–2011” (Paris School of Economics, 2013). See also M. Kehrig and N. Vincent, The Micro-Level Anatomy of the Labor Share Decline (National Bureau of Economic Research, NBER Working Papers 25275, 2018); E. Liu, A. Mian, and A. Sufi, Low Interest Rates, Market Power, and Productivity Growth (National Bureau of Economic Research, NBER Working Papers 25505, 2019). Awareness of the outsized influence of private monopolies in the United States may be on the rise, as indicated, for example, by the increasingly contentious debates around the need for some sort of public control of major information technology firms such as the so-called GAFA (Google, Apple, Facebook, and Amazon).

79.   79.  See the online appendix.

80.   80.  See especially C. Goldin and L. Katz, The Race Between Education and Technology: The Evolution of US Educational Wage Differentials, 1890–2005 (Belknap Press, 2010). See also Piketty, Capital in the Twenty-First Century, chap. 9, pp. 305–307.

81.   81.  In terms of intergenerational mobility, France and Germany seem to fall between the least mobile countries (the United States and United Kingdom) and the most mobile countries (Nordic countries). See the online appendix.

82.   82.  J. Davis and B. Mazumder, The Decline of Intergenerational Mobility after 1980 (Federal Reserve Bank of Chicago, Working Paper WP-2017-5, 2017). See also R. Chetty, D. Grusky, M. Hell, N. Hendren, R. Manduca, and J. Narand, “The Fading American Dream: Trends in Absolute Income Mobility Since 1940,” Science, 2017; F. Pfeffer, “Growing Wealth Gaps in Education,” Demography, 2018.

83.   83.  We find similar diversity in other parts of the world. Private financing is fairly extensive in Japan and Korea as well as Chile and Colombia but limited in China, Indonesia, and Turkey as well as Argentina and Mexico. The share of private financing in primary and secondary education is quite low everywhere (10–20 percent at most). See the online appendix, Fig. S11.11.

84.   84.  The total resources devoted to higher education amount to roughly 3 percent of national income in the United States, compared with 1–1.5 percent in Europe. (Italy spends the least, behind Spain, France, Germany, Sweden, Denmark, and Norway in ascending order.) This figure includes spending on scientific and academic research and on research institutes (spending on universities in the narrow sense is barely 0.5 percent of national income). See the online appendix.

85.   85.  The 850 American universities with capital endowments earned an average real return of 8.2 percent a year between 1980 and 1990 (after correcting for inflation and deducting management fees), with the 498 least well endowed (endowments of less than $100 million) earning 6.2 percent, while the sixty best endowed (with endowments greater than $1 billion) earned 8.8 percent; Harvard, Yale, and Princeton averaged 10.1 percent. The share of the sixty best endowed universities increased from 50 percent of the total endowment in 1980 to more than 70 percent in 2010. These gaps seem to be due to economies of scale: access to the most remunerative investments (unlisted foreign shares, derivatives on raw materials, etc.) is available only to the largest portfolios. See Piketty, Capital in the Twenty-First Century, pp. 447–450, table 12.2.

86.   86.  According to the Academic Ranking of World Universities, 2018 ed., by Shanghai Rankings, sixteen of the top twenty universities are American and four are European, while sixty-nine of the top 200 are American, eighty European, forty Asian-Oceanian, and ten from the rest of the world. Of the top 500, 139 are American, 195 European, 133 Asian-Oceanian, and thirty-three from the rest of the world.

87.   87.  See J. Meer and H. Rosen, “Altruism and the Child Cycle of Alumni Donations,” American Economic Journal: Economic Policy, 2009.

88.   88.  Note an expression in the United States to refer to heirs blessed with all privileges: “trust-fund babies.” In 2018, a boy band (Why Don’t We, or WDW) composed a song entitled “Trust Fund Baby.” The boys in the band, from Minnesota and Virginia, explained that they wanted independent girls who knew how to fix cars and take care of things on their own, unlike “trust fund babies”—heiresses born with silver spoons in their mouths who think of nothing but money.

89.   89.  See Chap. 5.

90.   90.  See Chap. 9.

91.   91.  See Chap. 8.

92.   92.  On relations between Harvard and the state of Massachusetts, see N. Maggor, Brahmin Capitalism: Frontiers of Wealth and Populism in America’s First Gilded Age (Harvard University Press, 2017), pp. 26–28, 96–104.

93.   93.  On the ideological challenge of higher education, see also E. Todd, Où en sommes-nous? Une esquisse de l’histoire humaine (Seuil, 2017).

94.   94.  See Chaps. 14–15.

95.   95.  See the online appendix. The data we have for comparing educational budgets between countries are far from perfect, but the break with previous periods is quite clear.

96.   96.  See especially Chaps. 13, 14, and 17. On inequality of investment in different tracks in France, see Fig. 7.8. See also S. Zuber, L’inégalité de la dépense publique d’éducation en France: 1900–2000 (EHESS, Paper, 2003), and the online appendix.

97.   97.  The idea that Reaganism was a success is based in part on a complex political-ideological construct, which has a lot to do with America’s success in the political and military competition with the Soviet Union (an outcome which itself had little to do with American economic and fiscal policy in the Reagan years), and somewhat less to do with the reduced gap between US and European growth (which would certainly have happened without Reagan, because the postwar catch-up phase had ended).

98.   98.  For other examples of the “inequality trap,” such as the Netherlands in the seventeenth and eighteenth centuries (where the commercial elite largely captured the state, and especially public finances, for its own benefit via accumulation of debt, blocking development), see B. van Bavel, The Invisible Hand? How Market Economies Have Emerged and Declined Since AD 500 (Oxford University Press, 2016).

99.   99.  Note in passing that the “four freedoms” established by the Single European Act of 1986 are rather different from the “four freedoms” evoked by Franklin D. Roosevelt in his famous State of the Union speech in 1941: freedom of speech and expression, freedom of worship, freedom from want, and freedom from fear.

100. 100.  See esp. A. Milward, The European Rescue of the Nation-State (Routledge, 2000).

101. 101.  A single state, such as Luxembourg or Ireland, is enough to block any common fiscal policy. I will come back to this question in greater detail in Chap. 16.

102. 102.  See European Commission, Taxation Trends in the European Union, 2018 ed. (Publications Office of the European Union, 2018), p. 35, graph 17. Some states such as France still have statutory tax rates of 30 percent or more, while Ireland and Luxembourg have rates of 10 percent or less. In a perfectly coordinated international fiscal system, it might be nothing more than a simple withholding against each shareholder’s progressive individual income tax obligation. In practice, given the absence of coordination and exchange of information concerning the ultimate beneficiary and the many opportunities for tax avoidance and evasion, the corporate tax is often the sole tax for which payment is actually guaranteed. See Chap. 17.

103. 103.  The US federal corporate tax rate was 45–50 percent until the 1980s; it fell to 30–35 percent under Reagan. It then remained stable at 35 percent from 1992 to 2017 (with an addition 5–10 percent in state taxes) before falling to 21 percent under Trump in 2018. This may lead to a new race to the bottom with European and other countries.

104. 104.  On British disappointment with respect to social Europe, see A. B. Atkinson, Inequality: What Can Be Done? (Harvard University Press, 2015).

105. 105.  R. Abdelal, Capital Rules: The Construction of Global Finance (Harvard University Press, 2007). This work relies on accounts by officials at the time (especially Jacques Delors and Pascal Lamy). See also N. Jabko, L’Europe par le marché. Histoire d’une stratégie improbable (Sciences Po, 2009).

106. 106.  The insistence of German Christian Democrats on free circulation of capital is often associated with Ordoliberalism. It was incorporated into many bilateral treaties signed by the Federal Republic of Germany in the 1950s and 1960s. See, for example, L. Panitch and S. Gindin, The Making of Global Capitalism: The Political Economy of American Empire (Verso, 2012), pp. 116–117.

107. 107.  The goal was also to reduce the cost of public borrowing on international financial markets. But there was no time for all these different objectives to be spelled out and explicitly debated.

108. 108.  The 1997 crisis led the IMF to reevaluate European rules on short-term capital flows and to rely instead on more flexible principles allowing certain capital controls in the spirit of the Bretton Woods accords of 1944. See Abdelal, Capital Rules, pp. 131–160.

109. 109.  By the same token, the role of German Ordoliberalism should not be overstated. There is also a strong French liberal tradition quite prevalent in the nineteenth and early twentieth centuries, especially in the interwar years, and revived in the 1960s and 1970s by Valéry Giscard d’Estaing, first as secretary of state and then minister of finance more or less continuously from 1959 to 1974 and later as president from 1974 to 1981. In 2001–2004 Giscard chaired the Convention on the Future of Europe, which led to the proposed European Constitutional Treaty (ECT), which de facto sacralized free circulation of capital and the principle of unanimity on tax issues. The ECT was rejected in France in a 2005 referendum but later adopted by parliamentary vote after slight changes in the form of the Lisbon Treaty of 2007. I will say more later about these European treaties and rules. See esp. Chaps. 12 and 16.

110. 110.  See S. Weeks, “Collective Effort, Private Accumulation: Constructing the Luxembourg Investment Fund, 1956–1988” (presentation, Accumulating Capital: Strategies of Profit and Dispossessive Policies conference, Paris, France, Thursday, June 6, 2019).

111. 111.  See Figs. 10.14–10.15.

112. 112.  The concentration is particularly strong in the 20–39 age group, with 62 percent of the wealth held by the 10 percent wealthiest in this group in France in 2015 (because of the importance of inheritances among the few wealthy people in this group) compared with 53 percent in the 40–59 age group, 50 percent in the 60-and-over group, and 55 percent for the population as a whole. In each age group, the poorest 50 percent own almost nothing (barely 5–10 percent of total wealth in all cases). See the online appendix, Fig. S11.18. For detailed results on age profiles and wealth structure by age group, see B. Garbinti, J. Goupille-Lebret, and T. Piketty, “Accounting for Wealth Inequality Dynamics: Methods and Estimates,” WID.world, 2016.

113. 113.  When the progressive income tax was created in the early twentieth century, the main objective was to tax high capital incomes, and most countries had tax laws favorable to income from labor, such as the French cédulaire system. In the 1960s and 1970s, the United States and United Kingdom taxed capital income (“unearned income”) at a higher rate the labor income (“earned income”).

114. 114.  See Figs. 10.11–10.13.

115. 115.  For detailed results, see Bozio et al., “Inequality and Redistribution in France 1990–2018.” See the online appendix for a discussion of methods.

116. 116.  See Table 3.1.

117. 117.  See Figs. 10.11–10.12.

118. 118.  See Chap. 10 for the distinction between the inheritance tax and the estate tax.

119. 119.  See Figs. 10.11–10.12.

120. 120.  See Figs. 10.4–10.5 and 10.8.

121. 121.  See Chap. 10.

122. 122.  See Chaps. 3 and 5.

123. 123.  See, for example, J. You, “Land Reform, Inequality, and Corruption: A Comparative Historical Study of Korea, Taiwan, and the Philippines,” Korean Journal of International Studies, 2014. See also T. Kawagoe, Agricultural Land Reform in Postwar Japan: Experience and Issues (World Bank, Working Paper WPS 2111, 1999). See also E. Reischauer, Histoire du Japon et des Japonais (Seuil, 1997), vol. 2, pp. 22–30. Reischauer, a former US ambassador to Japan who was deliberately condescending to the Japanese and hardly suspect of socialist sympathies, expresses pleasure with the success of agrarian reform and equalization of property at a time when the West was locked in competition with communism.

124. 124.  See Chap. 8 and the references to the work of A. Banerjee. The land redistribution in western Bengal followed the 1977 victory of the Left Front (led by the Communist Party of India), which remained in power until 2011.

125. 125.  In Mexico, where it is estimated that 1 percent of the population owned more than 95 percent of the land on the eve of the 1910 revolution, agrarian reform unfolded over the period 1910–1970. See S. Sanderson, Land Reform in Mexico, 1910–1980 (Elsevier, 1984); P. Dorner, Latin American Land Reforms in Theory and Practice: A Retrospective Analysis (University of Wisconsin Press, 1992).

126. 126.  This “solidaristic” concept of property as social property was proposed in the 1890s by Léon Bourgeois and Émile Durkheim as justification for a progressive income and estate tax. See R. Castel, Les métamorphoses de la question sociale. Une chronique du salariat (Fayard, 1995), pp. 444–449.

127. 127.  I will say more about this (imperfect) definition of justice in Part Four. See Chap. 17.

128. 128.  See H. George, Progress and Poverty (1879), pp. 342–359. On the question of compensation, see Chap. 6.

129. 129.  Concretely, Henry George’s proposal was a tax on the income from land at a rate equal to 100 percent of its rental value (whether actually rented or not) or, equivalently, a tax on capital equal to, say, 4 percent of the land’s value (assuming a rental value of 4 percent).

130. 130.  Long envisioned a progressive wealth tax starting at $1 million (seventy times the average individual wealth at the time) and marginal rates rising gradually to 100 percent, with a maximum wealth of $50 million (3,500 times the average), while noting that the scale could be adjusted if necessary to set the maximum at $10 million (700 times the average). His main goal was to guarantee every American family a fortune of one-third the average ($5,000 for an average of $15,000), and he was careful to make clear that he had nothing against private wealth as long as it remained reasonable and not obscene. The program was riddled with religious references questioning how a small minority had gotten hold of most of the country’s wealth. “God invited us all to come and eat and drink all we wanted. He smiled on our land and we grew crops of plenty to eat and wear. He showed us in the earth the iron and other things to make everything we wanted. He unfolded to us the secrets of science so that our work might be easy. God called: ‘Come to my feast.’ Then what happened? Rockefeller, Morgan, and their crowd stepped up and took enough for 120 million people and left only enough for 5 million of all the other 125 million to eat. And so many millions must go hungry and without these good things God gave us unless we call on them to put some of it back.” See the online appendix.

131. 131.  See Chap. 4.

132. 132.  Revenue from the French land tax in 2018 was about 40 billion euros (2 percent of national income), while the US property tax yielded $500 billion (more than 2.5 percent of national income).

133. 133.  The current rate is about 0.5–1 percent of the value of a property in France and the United States (with variations by state and town). Given the fact that the total value of private property is around five to six years of national income in both countries in the 2010s (see Figs. 10.8 and S10.8 in the online appendix), it is easy to see how receipts can amount to several percent of national income despite exemptions.

134. 134.  Give the competition among local governments to attract wealthy taxpayers, only a tax levied at the national or federal level could be steeply progressive.

135. 135.  In 2007–2011, the French government tried to put in place a so-called tax shield, that is, a ceiling on the total amount of tax any individual should be required to pay in relation to that person’s income (as opposed to wealth). Only the land tax paid on the taxpayer’s principal residence was included in this total.

136. 136.  Foreign assets may of course be subject to a land or property tax in the countries where they are invested.

137. 137.  Machinery and equipment are sometimes included in the property tax or partially taxed via other local business taxes such as the now-defunct taxe professionnelle in France. In practice, such assets are in general taxed at much lower rates than real estate.

138. 138.  See Maggor, Brahmin Capitalism, esp. pp. 76–95 and pp. 178–203.

139. 139.  Though not necessarily everywhere in Europe: to believe Victor Hugo in L’archipel de la Manche, the real estate tax in Guernsey in the nineteenth century fell on the overall wealth of the taxpayer, which the novelist, who was as usual curious about everything, found quite surprising, since he was used to the French system. See V. Hugo, Les travailleurs de la mer (1866; Folio, 1980), p. 67.

140. 140.  Another political battle at the time had to do with extending the city limits to incorporate recently urbanized areas and formerly independent towns. Hills defended this extension, while wealthy residents of central Boston opposed it so as not to have to share the city’s tax revenues with surrounding communities. See Maggor, Brahmin Capitalism. This episode once again illustrates the structural linkage of the tax regime to the political regime and the boundary regime.

141. 141.  On the political and administrative process that led eventually to complete exemption of personal property in 1915, see Maggor, Brahmin Capitalism. See also C. Bullock, “The Taxation of Property and Income in Massachusetts,” Quarterly Journal of Economics, 1916.

142. 142.  On the way property-owning elites were able to mobilize against the extension of property taxation in the nineteenth-century United States, both in the north (where the main issue from the elites’ viewpoint was to avoid the taxation of financial assets) and in the south (where the primary concern of property-owning classes was to avoid what they feared could become an excessive taxation of slave property), see E. Einhorn, American Taxation, American Slavery (University of Chicago Press, 2006).

143. 143.  R. Fisman, K. Gladstone, I. Kuziemko, and S. Naidu, Do Americans Want to Tax Capital? Evidence from Online Surveys (National Bureau of Economic Research, NBER Working Paper 23907, 2017). Specifically, the survey presents pairs of income and wealth and asks what people think would be a fair tax. For a given income (say, $100,000 per year), respondents felt that people with a net worth of $1 million should pay more in taxes than those who owned nothing and less than those with a net worth of $10 million. The same is true if one varies income for a fixed net worth.

144. 144.  I will have more to say on this. See esp. Chaps. 13 and 17.

145. 145.  Warren’s proposal would apply to individuals with wealth more than one hundred times the US average (of roughly $500,000 per couple and $250,000 per adult), or less than 0.1 percent of the population but holding 20 percent of total wealth, which would yield substantial tax revenues, estimated at more than 1 percent of national income. See E. Saez and G. Zucman, How Would a Progressive Wealth Tax Work? (University of California, Berkeley, Paper, 2019); E. Saez and G. Zucman, The Triumph of Injustice: How the Rich Dodge Taxes and How to Make Them Pay (Norton, 2019).

146. 146.  See Table 17.1.

147. 147.  Some researchers recently proposed a high proportional tax (of 7 percent) on all assets in order to force frequent reallocation of property. See E. Posner and E. G. Weyl, Radical Markets: Uprooting Capitalism and Democracy for a Just Society (Princeton University Press, 2018). Given the complete absence of progressivity, however, such a proposal might lead to greater concentration rather than diffusion of wealth. (In any case, the chief goal claimed by the authors is to facilitate rapid reallocation of land and goods.)

148. 148.  For a detailed analysis of Socialist and Communist programs and debates on progressive wealth taxes from the interwar years to the 1980s, see T. Piketty, Top Incomes in France in the Twentieth Century (Harvard University Press, 2018), pp. 367–380. Regarding Joseph Caillaux’s proposal in 1914 and the 1947 and 1972 bills, see J. Grosclaude, L’impôt sur la fortune (Berger-Levrault, 1976), pp. 145–217.

149. 149.  See esp. Chap. 14.

150. 150.  On the eve of the 2018 reform, the ISF yielded revenues of about 5 billion euros (less than 0.3 percent of national income), compared with 40 billion euros from the taxe foncière (more than 2 percent of national income).

151. 151.  See H. Glennerster, A Wealth Tax Abandoned: The Role of UK Treasury 1974–6 (London School of Economics, CASE Paper 147, 2011).

152. 152.  More precisely, the rate is 0 percent when the property is worth less than £125,000, 1 percent when it is worth £125,000–250,000, 3 percent between £250,000 and £500,000, 4 percent between £500,000 and £1 million, 5 percent between £1 million and £2 million (a new tax introduced in 2011), and 7 percent above £2 million (introduced in 2012). This progressive system is fairly surprising if one considers that such transaction fees are proportional in most countries (including France) and that the local “council tax,” which replaced the “poll tax” in 1993 (and cost Margaret Thatcher her post) was in reality almost as regressive as the latter (the rate of the council tax rose far less than proportionately to the rental value of the main residence). On this, see Atkinson, Inequality, pp. 197–199, fig. 7.3.

153. 153.  On the evolution of the German tax system since 1870, see F. Dell, L’Allemagne inégale. Inégalités de revenus et de patrimoine en Allemagne, dynamique d’accumulation du capital et taxation de Bismarck à Schröder 1870–2005 (EHESS, 2008).

154. 154.  The Swedish system was unique in imposing a joint tax on income and wealth from 1911 to 1947 before evolving into two separate systems in 1948. For details, see G. Du Rietz and M. Henrekson, “Swedish Wealth Taxation (1911–2007),” in Swedish Taxation: Developments Since 1862, ed. M. Henrekson and M. Stenkula (Palgrave, 2015), pp. 267–302.

155. 155.  See Chap. 16.