Piketty’s Capital in the 21st Century, 2014

You are here:
< Back

In 2014 Thomas Piketty published Capital in the 21st Century, the state of the art of global data compilations of trends in the distribution of income and of capital accumulation. Much of Piketty’s underlying academic work (and that of colleagues) is open access, with links available on Piketty’s web-page: a model of best research practice. Perhaps the most remarkable feature is that such fundamental material is only now being pulled together, along with all the inevitable gaps filled, and assumptions made. The general response (see The Economist and the Financial Times and Piketty’s response to the FT) has been favourable. Indeed many trends are so marked, and consistent across nations, that they are unlikely to be overturned, although the explanations, projections and presumptions for the future, suggesting serious problems ahead, along with the solutions offered, will stir a debate, as intended.

The political and policy question arises from rate of return expected on investments, compared to rate of growth of the economy being invested in. Do past policies tell us anything about an optimum distribution of wealth that gets the fastest rate of growth and spread of wealth generation for the greatest number, given the future challenge of achieving 9-10 billion living well by 2050?

Last, note that the data below illustrates trends in monetary wealth, from which the social consequences can be inferred; but it doesn’t tell us much about how to decouple global growth from unsustainable resource use.

We start with very long term trends in total capital of a nation (i.e. private and public), expressed as a percentage of annual national income generated, for the example of France from 1700 to today (see first graph below).

Above: Evolution of French national capital as a % of annual national income in 1700–2010, corrected for inflation. The key features, typical of Western European countries in general are, first, that the relative total capital value of agricultural land declining throughout three centuries and, second, the devastating impact on France of two world wars and the 1930s depression. In 1910 France’s total capital was worth almost seven years of national income. By 1950 it had fallen to less that 300% and the value of French foreign assets was virtually nil. Source: Fig 3.2 in Piketty 2014 data taken from Piketty’s data depository.

Three marked trends are clear. First, the relative contribution of agricultural land to national capital declined dramatically through the period, a consequence of additional capital value created by the industrial revolution. Second, is the astonishing destruction of assets 1914-1945. Although not shown in this graph, most capital was privately owned, and the losses particularly affected the wealthy rentier class living off capital income. This was partly as a result of physical destruction, but also through government borrowing through bonds that were then cancelled or inflated away, along with the loss of colonial capital. Piketty shows the same broad trends for Germany and Britain. In the US, while the decline in the relative contribution of agricultural land is similar, total national capital as a percentage of annual national income was always less: ranging from around 300% of annual national income in 1700, peaking at 525% in 1930 and was 450% in 2010.

The second graph shows Piketty’s analysis of the annual per capita (per person) growth rate (not absolute size) in the output of the global economy over the last two millennia, and a projection for the future growth rate through to 2100.

Above: Growth rate of per capita GDP from Antiquity, and projected to 2010. The global growth in annual income, at least as measured by money and economists, was extraordinarily low until 300 years ago. It then started to grow, reaching 2% from 1950. Many economists assume this will continue until 2050 as the ripples of the explosion of innovation and growth in wealthy nations in the twentieth century is integrated into a now globalised economy. Assuming no new innovations, and no natural or human induced disasters, the growth rate (note, not absolute wealth) then declines. Shading indicates future projections, to emphasise that the decline is based on assumptions, not on trends to date. Source: Fig 2.4 in Piketty 2014 data taken from Piketty’s data depository.

This assumes that global growth rate will reach a historic peak of 2.5% per annum or greater to 2050, and then sharply falls. This is as a result of assumptions concerning the growing middle class in developing nations, described earlier.

Whether this actually happens depends on many factors. Insights come from a more detailed look at the US distribution of income and GDP growth rate over the last century already graphed in the main text. This is done over through the following graphs, first concentrating on the US, and then comparing and contrasting with other economies. The graph below includes the same data as the graph in the main text. It adds the total share of income of the top 10% including capital gains (showing the impact of stock market and property value fluctuations) and also the contribution from the wages – top executive pay. We concentrate on the period through to the 1970s, and then subsequent events.

Above: Share of total national income the wealthiest 10% of the US population. This is broken down between total income (magenta), income excluding capital gains (such as stock market fluctuations) shown in blue (the same graph as in the main text), and income from wages (executive pay) to the top 10%. The original data gathered by Kuznets in the 1950s led him to highlight the decline in the relative share of total wealth by the wealthy, coupled with growing national prosperity. This was advanced as an irreversible trend of advanced capitalism, the American Way, and an unanswerable response to Communism. The ‘shock’ finding of Piketty and colleagues was to show just how dramatically the trend had reversed, since the 1980s, with the result that total wealth is now as unequal in the US as peak inequality in the 1920s. This change came about largely through increased income from wages, itself the result of the progressive tax rates on higher pay rate. Source: Fig 8.7 in Piketty 2014 data taken from Piketty’s data depository.

First, though, the ‘real’ per capita GDP may need some explaining. It is shown as the logarithm of GDP, so that the same rate of growth of GDP show up as the same slope in the line. It is the per capita GDP – Gross Domestic Product – i.e. the value of output of the US economy every year, per person. It is ‘real’ GDP, meaning that it adjusted for inflation, and expressed in 2009$s (bearing in mind the limitations, see John Kay in the main text) using data from MeasuringWorth. This indicates that the average GDP generated was around $6,241 in 1910. As seen in the graph, by 1933 it was still around $6,192. But it then grew dramatically, to $14,398 by 1950, £28,325 by 1980 and $47,724 in 2010 (note that the precision is spurious, it is given to help locate the data in MeasuringWorth)

Simon Kuznets was the first to explore the bare bones of such information for the US, England and Germany in 1955. He concluded that the historic trajectory of income equality for these countries over the previous century was first one of increasing inequality, a plateau, and then a move towards equality, particularly since the 1920s accompanied by increases in real average income per capita. This, he suggested, was the consequence of a switch from an agricultural to an industrial society, first resulting in increasing inequality, but subsequently the “growing political power of the urban lower-income groups” meant that they could claim “more adequate shares of the growing income”. The other factor reducing inequality that Kuznets singled out was the “very importance” of the “recent phase” of progressive income taxes and benefits. The wider significance of Kuznets is that this ‘Kuznets Curve’ was promoted by Cold War politicians as the underlying economic proof that  ‘American Way’ of capitalism, promised widespread growing wealth for everyone, contradicting the basic message of Communism.

The graph below shows the changes in progressive taxation (higher tax levels for higher income bands) as measured by changes to the top rate of tax, as collated by Piketty. Bear in mind that the top tax rate can be set so high that it affects few people, and that setting a rate does not mean that it is collected! Note also that it may neither be necessary, not the intent, of a tax to raise revenue, but rather to change behaviour, in this to change expenditure patterns towards investments that generate future wealth and employment, and away from the consumption of goods and services that do not. On the other hand, many people clearly believe that extreme unearned difference in income are immoral.

Above: Top Marginal Income Rates for the US, UK, France and Germany 1910-2010. Broadly top rates increased from 1910 to the 1940s, plateaued, and then started to fall broadly to levels last seen in the 1920s. However there was marked variation between countries. In the US the rate rose, fell back to 25% or less 1925-30 and then rose. Source: Fig 14.1 in Piketty 2014 data taken from Piketty’s data depository.

Still concentrating on the period to the 1950s and the US, it is evident that progressive marginal rates of taxation were introduced and then grew during the 1914-18 war, fell, and then rose again during the 1930s peaking during the 1939-45 war. It is also clear from the previous graph that while the growth rate of the economy was high from the mid-1930s, it was only during the war that the share of national income taken by the top 10% fell sharply. The economic stimulus of the war, not discussed by Kuznet (but who does note the destruction of national wealth by the 1914-18 war), complicates the analysis. War Bonds and other government borrowing of the US and allies would also have been a major stimulus to US GDP.

Turning now to US history after the 1950s, and the data gathered by Piketty and colleagues, the consequences of the massive shift in US taxation policy in the 1970s by President Reagan (Margaret Thatcher in the UK then made similar changes) are clear in the graph. The top marginal rate of tax, having already been reduced in the 1960s, was reduced twice more in the 1980s, to bring the top rate down to the lowest level since the late 1920s. The justification was that high marginal tax rates on executive pay were impeding entrepreneurial activity of the rich. As a result wages then started to increase, making up a notable part in increasing total income going to the top 10%, along with highly variable amounts from capital gains.

The total take continued to increase until by 2010 they had returned to levels last seen in the 1930s. If the intent was to increase the rate of economic growth, it failed as the rate of GDP growth did not increase. Inspection of the slope suggests that GDP growth may even have slowed. On the other hand, GDP growth did not stall, as it had between 1910 and the 1930s. The next question would be, what is the current contribution of the real economy, and of the financial sector, to total US GDP, and the extent to which the contribution of the financial sector could be realised.

Data from the World Top Incomes Database, the ‘bottom’ 90%, shows average income rising from under $5000 in 1933 (in 2012 $s) to $36,000 in 1973, it then fell, to $31,000 in 2012. Taking all of the data into account, it would seem that 90% of the US population would likely conclude they would benefit from restoration of progressive taxation, and that — to put it at its weakest — there was no historical evidence that the rate of growth of GDP would fall. They might also be thinking about the nature of an equivalent economic stimulation to the real economy that was such an important part of the American Dream.

Now turning to the broader context, of the relative position of wealth distribution in different countries. The graph below shows the proportion of annual income taken by the top 1% in ten countries over the last hundred years.

Above: Proportional of total national income taken by wealthiest 1% of ten countries 1910-2010. Excludes capital gains where data distinguishes this, such as the US. Although there is a lot of data overlaid, the lines are transparent and provide a number of insights. Source: Fig 8.8 and 9.2-4 in Piketty 2014 data taken from Piketty’s data depository.

In the US, the top 1% took, at its greatest, took 20% of total income in the 1920s and again in the last decade, vs. ca 45% for top 10% (excluding capital gains). The US, UK and Canada show sharp rises from the 1980s after being centre trend in the 1950s-70s, but other Anglophone countries such as Australia and New Zealand similar to non-Anglophone countries. Denmark and Sweden show their journey through the century from a high share of annual income to low, as relatively egalitarian countries, while the post war recovery in Germany is extremely marked. Japan starts high, falls dramatically in the 1940s and then is similar to other non-Anglophone countries.

An additional perspective comes from comparing performance of companies in countries with different management cultures, and of historical performance with current. This is done in the next graph by comparing share value with book value of companies in different countries.

Above: Company share value as a percentage of audited book value in wealthy nations between 1970 and 2010. Source: Fig 5.6 in Piketty 2014 data taken from Piketty’s data depository.

When a company is created with share capital, initially the book value, as it appears in the company accounts, will exactly match the share value. Over time these will drift apart for various reasons. There are obviously speculative movements on the stock market unrelated to intrinsic value of the company, and also informed reasons why the future value may be higher or lower than actually shown on the books. But another important reason is that the profits of the company do not only go to shareholders. They may be retained for investment, including investing in human resources, including performance related pay for workers and executives, or more indirectly, including employee health schemes.

So, in Germany, stock value is consistently low against book value, yet German businesses have a global reputation for premium products at premium prices. But companies must integrate the opinions and interests of workers, represented on the board, but also, for example, members of civil society such as environmental NGOs. In essence, Germany ‘Rhenish Capitalism’ was a form of legally binding CSR well before the glossy brochure version became widespread. Wages are negotiated regionally between employers and employees, making apprenticeships easy (because companies cannot free-ride the training schemes of others) but also requiring investment. But this also means that profits are more widely shared. It is notable that the top 1% in Germany score lower than the US or UK. Similarly, for the other country with a notably low stock to book value, Japan there was a well established tradition that companies look after employees for life, and a non-confrontational industrial relations between managers and workers. But Japanese companies are facing major challenges, and German business is not without problems. And there are other circumstances that are different, such as the importance of customer-owned banks and the KfW for business lending in Germany, so any conclusions have to be rather cautious.

The other obvious trait is that prior to 1980s the share value of US and UK companies were consistently less than book value, then became consistently greater than book value, but have now fallen back towards parity. This is consistent with trends in company governance, where the power of employees to bargain for a greater share of profits, noted by Kuznets, has been diminished, and speculation on stock prices have become detached from company book value.

Income is only half of the story. The other half arises from changes in the distribution and growth of capital, shown in the next graph.


Above: Changes in of private and public capital in wealthy countries 1970–2010 as a percentage of annual national income. Private capital grew from between 250-350% of annual national income in 1970 to between 400-700% in 2010. Public capital from ca. 50% to 0. Source: Fig 5.5 in Piketty 2014 data taken from Piketty’s data depository.

The graph above shows the dramatic growth in private capital in rich countries over the last forty years. There is an underlying trend and fluctuations about that trend. The fluctuations include a massive bubble in the Japanese economy in the late 1980s, a real estate bubble in Italy in the 1990s, another, in Spain, that took nominal private capital up to 800% of annual national income until the crash of 2007/08, and the US and UK the internet bubble of 2000-01, the US real estate and stock market boom to 2007, and the subsequent recession.

The underlying trend was that private wealth was worth 2–3.5 years of national income in 1970 and 4–7 years in 2010. Behind this are differences between the growth rate, per capita, of national income, which for all these countries was under 2%, and of annual private savings, which ranged between 7.3 and 15%. Public capital started this period at between +10 and +90% of annual income, and ended at between +90 and -90%. So in these wealthy countries, the bulk of wealth is private wealth.

We can now paint the bigger picture. There was the destruction of capital in two wars and the intervening Great Depression. There was redistribution of wealth by taxation. There was also the massive rebuilding in western Europe and Asia of infrastructure, and re-investment and redirection towards peacetime industrial capacity in these countries and in North America, ignored during the wars and Depression. That lasted from the 1950s through to the 1970s. Labour was relatively scarce and wages relatively high. That rebuilding came (as it did in the war-time economy expenditure) through direct investment in manufacturing, services and the building of infrastructure. That contrasts with the current crisis, where quantitative easing (printing money) and government bonds go to banks and other investors in the hope that they will in turn lend to the ‘real’ economy. Occupational pensions also contributed to the increasing proportion of wealth held by the middle classes, with the result that the third quarter of the twentieth century is now seen as a golden era of growth in Western Europe, North America and Japan.

But what we now know that there is nothing inevitable about the Kuznets Curve, of prosperity growing evenly across the board. Possibly growth would have slowed in any case. The subsequent Information Revolution has been a mixed blessing for low and medium wage earners. But the major change was the winding back of progressive taxation, with a minor contribution from the privatisation of nationalised industries in Europe, producing benefits which went principally to the better-off. The positive feedback of interest on investments has widened the gap still further, although the various stock market busts and property bubbles have added a capital-gains saw tooth edge (red line on the graph) to the underlying growing wealth (blue line) of the richest top 10% of those living in the US during the first decade of the twenty first century.

So were these wealthy countries the entire world, this would speak directly to Piketty’s concern of the growing concentration of capital, from which governments would now be seeking to borrow and pay interest for measures to stimulate the economy, This is a different strategy from that of the 1930s-1970s, where this was funded from a combination of progressive taxation and (during the 1940s) War Bonds and other government borrowing.

However one last point is that capital resources of the global economy now include significant capital resources beyond these countries, and the distribution of capital is projected to change in coming decades. The final graph summarises trends in value and distribution between regions of total global capital since 1870 and a projection forward to 2100.

Above: Growth of private capital in wealthy countries 1970–2010 as a percentage of annual national income. Private capital grew from between 250-350% of annual national income in 1970 to between 400-700% in 2010. Source: Fig 5.3 in Piketty 2014 data taken from Piketty’s data depository.

The first stand out feature of this graph is to underline the consequences for Europe of the two wars and the Great Depression, previously illustrated for France in the first graph. The wiping out of European capital value between 1910 and 1920, and then a further fall between 1940 and 1950 is astonishing. To this day Europe has scarcely recovered a third of its pre 1914 global capital share. There can be no starker illustration of the folly of Europe’s 20th century history. The American continent came through the twentieth century unscathed in terms of global capital share. Africa’s share of global capital has been minimal: the projections that Piketty has speculate that Africa’s share will start to increase significantly in the second half of the 21st century. Asia, including the sovereign wealth funds of the oil-rich states, the accumulated capital of savers in China’s growing economy, as well as that of India, has been increasing since the 1960s, and the accumulation of wealth has become obvious in the last two decades.

Piketty also provides some indication of the share of that wealth in both India and China (Piketty’s Fig. 9.9). For India, the share of the top 1% by income of total annual income between 1920 and today peaked in the British Raj in the late 1930s at 18%, a level comparable to that in the US today. The income of the top 1% then declined, reaching a minimum of around 4% in the early 1980s. It then started to grow again and in 2010 the top 1% share of annual income was estimated to be around 12%. For China data is available from the late 1980s. Since that time the income share of the top 1% has grown from some 4% to around 11% in 2010. The data for South Africa, Indonesia and Argentina are also consistent with a declining share of the top 1% in total income through to the 1980s, since which time it has been growing.

So, in other parts of the world, as well as the wealthy countries of the twentieth century, the income saved disproportionately by the already well off can be expected to be contribution to capital accumulating at rates far greater than the projected global growth rate, creating levels of inequality last seen in the early twentieth century. To avoid a repeat of the conflict and destruction of value that then followed, Piketty calls for globally progressive taxation of wealth, in effect to learn the lessons behind the success of the American Way, but not to assume that the progression to increasing wealth, increasingly evenly distributed, is an automatic consequence.

Source: Trends to Bend, modus vivendi, 2014, MMG