In a series of recent studies by Thomas Piketty and Gabriel Zucman, new evidence on historical trends in aggregate national and private wealth–income ratios in France, Germany, the UK and the US are presented. Their series shows that the importance of wealth has fluctuated grossly over time in Europe, whereas it has been both lower and more stable in the US (Piketty and Zucman 2014, Piketty 2014). Accounting for wealth accumulation trends, the authors find that new savings stand for the major part, whereas asset price gains are relatively unimportant over the long run.
These findings doubtless offer many important insights, but they still leave some key questions unresolved. First, do these patterns carry over to smaller and less-populated countries that also industrialised at a later stage than the countries studied by Piketty and Zucman? A priori, this is an open question. According to most trade models, for example, a technology shock that raises the demand for capital (e.g. industrialisation) should increase values of wealth in a large economy, whereas it generates capital imports in small economies where factor prices are fixed. A second issue concerns the role of economic and political institutions. While there is a large literature showing how universal suffrage, redistributive taxation or various welfare state arrangements can have real long-term effects on the economy, much less is known about whether they also shape, and are shaped by, the evolution of aggregate wealth in society.
In a new paper, I try to shed light on these and related issues by examining the development of wealth-income ratios in Sweden over the past two centuries (Waldenström 2015a). Sweden is the archetypical ‘small, open economy’ and it industrialised much later than the larger economies studied by Piketty and Zucman. Furthermore, during the 20th century, Sweden developed a social democratic welfare state that came to be one of the world’s most extensive in terms of egalitarian ambitions. Contrasting aspects of these welfare state institutions against those found in Continental European and Anglo-Saxon countries, the case of Sweden can offer insights into how political institutions contribute to the accumulation and composition of national wealth.
Historical wealth-income ratios: Was Sweden ‘the US of Europe’?
The basis for the analysis is a newly constructed historical database of Swedish aggregate national and private wealth and savings covering the period 1810–2014. The database is constructed in accordance with current national accounting standards as well as the database of Piketty and Zucman (2014).1
Figure 1 depicts the trends in aggregate private wealth-income ratios in Sweden and the four large European and American countries. Both common and distinct patterns emerge. Differences are starkest in the era up to WWI, when Sweden’s wealth-income ratio was only about half of that seen in the rest of Europe. A most striking feature though is that it was almost exactly the same as the US series, raising the question if Sweden, which also was scarcely populated and relatively late to industrialise, was in fact ‘the US of Europe?’ An examination of the underlying factors of the two countries, however, suggests otherwise. Sweden’s low wealth-income ratio was mainly due to very low net (and gross) saving rates of around 1–2%. This effectively means that the Swedes were, in the agrarian era, so poor that they had to consume basically everything that they earned, preventing any substantial accumulation of new assets. By contrast, US saving rates were in line with those observed in France, Germany, and the UK, and the low wealth-income ratio was instead mainly due to income growth being even higher.
During the 20th century, the patterns are more alike between Sweden and the other countries. After the dramatic falls in European wealth-income ratios during WWI, wealth-income ratios hovered at levels of between two and four years of national income. In the most recent decades, a rise is documented in all countries, and using statistical structural break techniques, the study establishes that these increases are statistically significant in both country-specific and cross-country dimensions.
Figure 1. Wealth-income ratios in five countries, 1810–2010
How could poor Sweden finance its industrialisation?
Sweden was thus a relatively capital-scarce economy during the 19th century. But if this is true, how could the country manage to industrialise with such a small amount of domestic capital? One possible answer is capital imports, and the extent to which this was used to finance the Swedish industrialisation has been debated by Swedish economic historians.
Figure 2 sheds new light on the issue by showing the ratio of net foreign assets (defined as all claims on foreigners’ net of foreign claims on citizens at current market prices) to national income. A fascinating pattern emerges – Sweden was indeed a net importer of capital during the entire industrial take-off, with foreign wealth representing approximately one seventh of the total private capital stock. Most of these foreign funds came from German and French creditors who bought Swedish bonds floated on the Hamburg and Paris markets. Compared with the other countries in the figure, Sweden was the largest net debtor and only the US also had a negative foreign position, though at a smaller level. France, Germany, and the UK were instead net creditors on international capital markets. Gauging the quantitative importance of these foreign funds for Swedish industrialisation is difficult, but it represented approximately 80% of total commercial bank credit and 150% of central government debt around the year 1900.
Figure 2. The role of foreign capital
Do institutions matter?
Although considerable research shows how political and economic institutions matter for several of society’s economic outcomes, relatively little is known about their importance for wealth accumulation. Comparing Sweden’s special political context and welfare-state institutions with those in Continental European and Anglo-Saxon countries offers an opportunity to address this issue.
My study documents an extraordinarily large growth of Swedish government wealth in the post-war era, precisely around the time of the most intensive expansion of the social democratic welfare state. Of particular importance appears to have been the emergence of the Swedish pension system. Reforms in 1947 and 1960 implied large increases in the public sector’s involvement in the pension system. Most of the pensions were unfunded, based on defined-benefit schemes, and largely financed through the tax system. However, since these unfunded pensions are not equivalent to other forms of marketable assets, they are often excluded in wealth analyses (including those of Piketty and Zucman). This exclusion may be motivated from a pure accounting perspective, but from an economic perspective it can be questioned since people expecting to receive these (unfunded) pensions may save less for retirement and thus have less private wealth than if they did not possess any unfunded pension entitlements.2
Figure 3 presents a cross-country comparison of private wealth-income ratios in Sweden, the UK and the US, when private wealth includes and excludes these unfunded pension assets in public and private schemes.3 The figure shows a striking contrast between Sweden and the other two countries over the post-war era. Just after the war, unfunded public and private pension wealth was insignificant in all countries. But by the 1960s, after the Swedish pension reforms, unfunded pension wealth had grown in Sweden to reach the same value as all of private wealth, whereas it remained at only a small fraction in the UK (about one third) and the US (one tenth). Furthermore, virtually all of the pension wealth expansion in Sweden was in the form of public pensions, whereas in the Anglo-Saxon countries the public schemes represented less than half, while the private schemes stood for more than half.
Figure 3. Private wealth-income ratios with and without unfunded pension wealth
Summing up, a new dataset on Swedish historical national wealth is used to study whether the patterns in wealth-income ratios previously found by Piketty and Zucman (2014) carry over to small countries that were historically backward and that developed a different set of political and economic institutions during the 20th century. The findings highlight both similarities and differences. In the pre-industrial era, Swedish wealth levels were low and the main explanation is that Swedes were too poor to save. During the 20th century, Swedish wealth-income ratios are more similar to the rest of Europe, but they still stand out in terms of the structure of national wealth. In Sweden, government wealth grew much faster and became more important, not least through its relatively large public pension system. This points to an explicit role of economic and political institutions for the long-run evolution of national wealth-income ratios.
Blake, D, and J M Orszag (1999) “Annual estimates of personal wealth holdings in the United Kingdom since 1948”, Applied Financial Economics, 9(4): 397–421.
Feldstein, M (1974) “Social security, induced retirement, and aggregate capital accumulation”, Journal of Political Economy, 82(5): 905–926.
Investment Company Institute (2015) “The US retirement market, first quarter 2015”, June, Washington, D.C., 2015-09.
Piketty, T (2014) Capital in the 21st century, Cambridge, MA: Belknap.
Piketty, T and G Zucman (2014) “Capital is back: Wealth-income ratios in rich countries, 1700–2010”, Quarterly Journal of Economics, 129(3): 1255–1310.
Waldenström, D (2015a) “Wealth-income ratios in a small, late-industrializing, welfare-state economy: Sweden, 1810–2014”, CEPR Discussion Paper 10708.
Waldenström, D (2015b) “The national wealth of Sweden, 1810–2014”, CEPR Discussion Paper 10882.
Wolff, E N and M Marley (1989) “Long-term trends in US wealth inequality: Methodological issues and results”, in The Measurement of Saving, Investment, and Wealth, Chicago, IL: University of Chicago Press.
2 Indeed, evidence on such a crowding-out effect has been documented in some countries’ pension systems (see, e.g., Feldstein 1974).
3 The series for Sweden are presented for the first time and are based on calculations using data from a wide array of sources and past research. The calculations for the UK were made by Blake and Orzag (1999) and for the US by Wolff and Marley (1989) and Investment Company Institute (2015).