Economic growth

Lane Kenworthy, The Good Society
May 2018

Economic growth is valuable, partly because it boosts living standards and partly because other aims are more readily achieved in an affluent, growing economy.1 What are the chief contributors to economic growth? Why have some countries grown more rapidly than others?

CORE SOURCES OF ECONOMIC GROWTH

A capitalist economy features extensive private ownership of property and markets in goods, services, and labor. Markets are helpful in two respects. One is resource allocation. The scale and complexity of a national economy make effective planning and coordination via commands very difficult. Market prices turn out to work better.2 The second is innovation and improvement. Competition creates incentives to develop new products and services, to improve existing ones, and to increase efficiency.

But sustained growth requires more than markets. According to Daron Acemoglu and James Robinson, “Secure property rights, the law, public services, and the freedom to contract and exchange all rely on the state, the institution with the coercive capacity to impose order, prevent theft and fraud, and enforce contracts between private parties. To function well, society also needs other public services: roads and a transport network so that goods can be transported, a public infrastructure so that economic activity can flourish, and some type of basic regulation to prevent fraud and malfeasance.”3 For the most part, it is governments that provide these things. Economic growth thus depends in part on effective government.

A third key is science — more precisely, the application of the scientific method and the means needed to measure and analyze in order to improve products and the process of producing them. Without science, market incentives are likely to yield some advance in productivity, but probably a very limited amount.

Economic growth may also require a cultural shift. In The Protestant Ethic and the Spirit of Capitalism, Max Weber asked why early business owners would reinvest profits in order to accumulate wealth. At the time of his writing, around 1905, the capitalist economies in much of western Europe and the United States featured substantial competition, which encourages capitalists to reinvest profits in order to raise productivity and lower costs, in order to avoid being underbid and pushed out of business.4 But why did early entrepreneurs continuously reinvest profits, especially in conditions of considerable uncertainty? As Weber noted, this orientation runs counter to “traditional” rational economic behavior: “A man does not ‘by nature’ wish to earn more and more money, but simply to live as he is accustomed to live and to earn as much as is necessary for that purpose.” The reason, Weber suggested, is that many early industrialists felt there was no choice in the matter; they believed it their duty to reinvest and accumulate. A disproportionate number of these entrepreneurs were Protestant, affiliated with Calvinism or one of several related sects, and in Calvinist doctrine attainment of wealth through hard labor in pursuit of a calling — the calling of entrepreneur — was viewed as a sign of God’s blessing.

We can see the impact of these institutions, policies, and attitudes in long-run economic growth patterns. Figure 1 shows estimates of gross domestic product (GDP) per person over the past two thousand years in France, and over recent centuries in Germany and the United States. There was little if any economic advance prior to the late 1800s. Then, once this configuration was in place, economic growth took off.

Figure 1. GDP per capita over 2000 years
Adjusted for inflation and converted to 2011 US dollars using purchasing power parities. “k” = thousand. The line that extends back to 1 AD is for France. The data begin in 1500 for Germany and in 1650 for the United States. Data source: Maddison Project Database 2018, rug.nl/ggdc.

Figure 2 highlights the importance of markets. It shows the available historical data on GDP per capita in the United Kingdom and China. In the UK this full configuration existed by the mid-1700s, and that’s when economic growth began to surge. In China, by contrast, markets were severely constrained until economic reforms were introduced beginning in the late 1970s. China had hardly any economic growth prior to these reforms but very rapid growth thereafter.

Figure 2. GDP per capita in the UK and China
Adjusted for inflation and converted to 2011 US dollars using purchasing power parities. “k” = thousand. Data source: Maddison Project Database 2018, rug.nl/ggdc.

ECONOMIC GROWTH IN RICH DEMOCRATIC NATIONS SINCE THE EARLY 1800S

Policy makers tend to care about short-term economic growth, but growth rates in many of the world’s affluent longstanding-democratic nations have been remarkably constant over the past century or more. Countries that do especially well or poorly during a particular decade or business cycle often subsequently revert back to their long-run growth rate.

Figure 3 shows GDP per capita in these countries since the early 1800s, according to estimates compiled by Angus Maddison. The data are shown in logarithmic form, along with a regression line that begins in 1870. If GDP per capita grows at a constant rate, the data points will fall along this line. That’s exactly what we see for a striking number of the countries. For a handful of others there is a clear break in growth patterns that occurs around World War II, with growth accelerating after the war. In those cases two lines are shown in the figure.

gdppclog-21countries-1820to2016-country1to8b

gdppclog-21countries-1820to2016-country9to16b

Figure 3. GDP per capita (log)
Natural log of inflation-adjusted GDP per capita. A log scale is used to focus on rates of change. “k” = thousand. The vertical axis does not begin at zero. The lines are linear regression lines; they represent a constant rate of economic growth since 1870. For Ireland, Italy, Japan, South Korea, Portugal, and Spain the data suggest a break in growth patterns, so there is one line for 1870-1945 and another for 1945-2016. Data source: Maddison Project Database 2018, rug.nl/ggdc.

DIFFERENCES IN ECONOMIC GROWTH ACROSS RICH DEMOCRATIC NATIONS SINCE THE LATE 1970S

All of the world’s affluent longstanding-democratic countries have capitalist economies with well-developed markets, government provision of necessary laws and public goods, access to modern science, and widespread attitudes conducive to investment and accumulation. And all of them have had fairly stable long-term rates of growth — some over the very long run and others since the mid-1900s. However, they differ in their growth rates. The horizontal axis in figure 3 shows average rates of growth in GDP per capita since 1979. (Three countries that could in principle be included here are not: Ireland, South Korea, and Norway. Each had much faster growth rates than the other nations, and for idiosyncratic reasons — Norway’s oil, Ireland’s tax haven status — that aren’t likely to tell us anything useful about growth patterns in the other nations.) Some of these countries have grown at a rate nearly twice that of others.

I focus on the period since the late 1970s because this is the era of globalization, central bank independence and tight money, and neoliberalism; because growth in the 1940s, 1950s, and 1960s was heavily influenced by post-World War II rebuilding; because the 1970s were an odd decade dominated economically by two oil shocks and stagflation; and because Portugal and Spain didn’t become democratic until the mid-to-late 1970s.

The cross-country differences in economic growth since the late 1970s owe partly to “catch-up”: countries that begin with a lower per capita GDP are able to grow more rapidly by borrowing technology from richer nations. The horizontal axis in figure 3 shows unadjusted economic growth rates since 1979. The vertical axis shows growth rates adjusted for this catch-up process. A number of the countries move around in the rank-ordering.

Figure 4. Economic growth and catchup-adjusted economic growth
Average annual rate of growth in inflation-adjusted GDP per capita. The “catchup-adjusted” growth rates are residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD.

What accounts for the cross-national variation in post-1970s economic growth that isn’t due to catch-up? Social scientists have offered many hypotheses. Let’s consider some of the most influential and see how they square with the data.

Education

Technological advance is the core source of economic growth, and a better-educated populace is likely to yield more innovation. It’s also likely to be more productive.5

The standard measure of education, years of schooling completed, isn’t very helpful when comparing across countries because in some nations a significant segment of the population enrolls in apprenticeship programs, which can yield highly skilled workers even though they won’t have finished as many years of schooling. A better measure comes from test scores. Eric Hanushek and Ludger Woessmann have compiled scores that cover the period 1965 to 2003. For the years 2006 to 2015 we have even better data, from the PISA tests administered every three years by the OECD, though since these come late in the period of interest they can only predict growth if the cross-country differences in test scores haven’t changed much over time. Figure 5 suggests a mild positive association between both test score indicators and catchup-adjusted economic growth rates, but the correlation drops to nearly zero if Italy is excluded.

Figure 5. Education and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Test scores 1965-2003: average test score in math and science, primary through end of secondary school. Data source: Eric A. Hanushek and Ludger Woessmann, “Do Better Schools Lead to More Growth? Cognitive Skills, Economic Outcomes, and Causation,” Journal of Economic Growth, 2012. Test scores 2006-2015: average student score on Program for International Assessment (PISA) reading, math, and science tests, 15-year-olds. Data source: OECD. The lines are linear regression lines. The dashed lines are with all 18 countries; the solid lines are with Italy excluded. In the first chart, the correlation is +.33 with all 18 countries and +.14 with Italy excluded. In the second chart, the correlation is +.25 with all 18 countries and +.01 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

Modest economic regulation

Since Adam Smith, a stream of analysts and commentators have argued that the institutions most conducive to national wealth in a capitalist economy are those that facilitate the ability of economic actors to form companies, hire and fire employees, and reap the rewards of innovation.6 Such institutions include strong and clear protection of property rights, effective antitrust enforcement, limited protection of market position for firms and job security for employees, a financial system that provides capital at reasonable cost, stable and consistent macroeconomic policy, government support for research and development, and wide-reaching provision of high-quality education.

The Fraser Institute has a measure of economic regulation that takes into account a country’s legal system and property rights, credit market regulations, labor market regulations, business regulations, and freedom to trade internationally. Scores are available in each country beginning in 1980. Figure 6 shows that this index is positively correlated with catchup-adjusted economic growth rates, but the association isn’t especially strong and hinges on Italy’s inclusion.

Figure 6. Modest economic regulation and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Modest economic regulation: average score for legal system and property rights, credit market regulations, labor market regulations, business regulations, and freedom to trade internationally. Higher scores indicate less regulation. 1980-2015. Data source: Fraser Institute, Economic Freedom database. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with Italy excluded. The correlation is +.35 with all 18 countries and +.07 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

International trade

Most economists believe that trade increases economic growth, by increasing competition and by allowing access to less expensive and/or higher quality inputs.7 Other voices over the past two centuries have argued that limits on trade, or at least on imports, will boost a nation’s economic health by shielding domestic firms from foreign competition and thereby giving them more opportunity to grow. Figure 7 shows there there is no association across the rich democracies between imports (as a share of GDP) and catchup-adjusted economic growth.

Figure 7. International trade and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Imports: share of GDP. 1979-2015. Data source: OECD. The line is a linear regression line. The correlation is -.02. “Asl” is Australia; “Aus” is Austria.

Government taxing and spending

As noted earlier, some government expenditure is necessary for a well-functioning market economy. But government spending may waste funds that could otherwise boost economic growth. And if taxes are too high, they can weaken financial incentives for innovation, investment, and work effort. Figure 8 suggests little reason to think there is a causal association between government revenues and catchup-adjusted rates of economic growth across the rich democracies.8

Figure 8. Government size and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Government revenues: share of GDP. 1995-2015. Data source: OECD. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with France and Italy excluded. The correlation is -.20 with all 18 countries and -.08 with France and Italy excluded. “Asl” is Australia; “Aus” is Austria.

Government debt

Governments that persistently run a deficit can accumulate a large debt, the repayment of which diverts resources away from productive economic activity. Figure 9, however, suggests little or no association between public debt size and catchup-adjusted economic growth.

Figure 9. Government debt and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Government net debt: share of GDP. Data source: OECD. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with Italy excluded. The correlation is -.37 with all 18 countries and -.14 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

State guidance

Proponents of limited government point to the potential for government failings. Governments have limited information. They are buffered from the competitive pressures. Their ability to flexibly adapt may be hindered by bureaucracy. They are subject to pressure from rent-seeking interest groups who aim for their own benefit rather than the most productive allocation of investment. Yet while markets are good at allocating resources to productive use, they are far from perfect in doing so. Private investors may have short time horizons, emphasizing near-term profits over productivity, long-run returns, market share, or export competitiveness. They may have limited information. They may be unconcerned about spillover benefits from particular firms and industries. They may benefit from, and thereby accentuate, limits to competition. In these ways, market allocation of resources may be growth-inhibiting.

Governments can help to remedy this sort of market failure via proactive steering of capital toward particular firms or sectors.9 They can do this via subsidies, favorable loan terms, assistance with coordination, export help, import protection, and public ownership. This type of state guidance sometimes is referred to as “industrial policy.”

Most discussions of state guidance or industrial policy have not attempted a thorough empirical assessment of the hypothesis. Many single out Japan and France as the exemplary cases. A dichotomous scoring in which Japan and France are coded as “high” and all other countries “low” isn’t correlated with catchup-adjusted economic growth, as figure 10 shows.10

Figure 10. State guidance and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. The line is a linear regression line. The correlation is -.12. “Asl” is Australia; “Aus” is Austria.

Government effectiveness

Even if active state guidance of the economy isn’t needed for a healthy economy, having an effective government might be. There are various ways to measure the quality of government.11 A common one is the World Bank’s “government effectiveness” indicator, which attempts to gauge public and expert perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies.

The dashed line in figure 11 suggests a strong positive association between government effectiveness and catchup-adjusted economic growth. However, the strength of the association owes almost entirely to a single country, Italy, which has very low measured government effectiveness and the poorest growth record since the 1970s. The solid line in the chart shows that with Italy excluded there is little if any correlation.

Figure 11. Government effectiveness and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Government effectiveness attempts to capture perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. The data set includes most of the world’s countries, so “moderate” and “high” government effectiveness are relative to this larger group. 1996-2015. Data source: Stefan Dahlberg, Sören Holmberg, Bo Rothstein, Anna Khomenko, and Richard Svensson, Quality of Government Basic Dataset, version Jan16, Quality of Government Institute, University of Gothenburg, series wbgi_gee, using data from the World Bank. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with Italy excluded. The correlation is +.50 with all 18 countries and +.05 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

Interest group organization

In democratic societies, individuals tend to organize in interest groups. These groups can contribute to well-being and happiness. But such groups also have an incentive to lobby government for special favors and impede market functioning, creating what Mancur Olson termed “institutional sclerosis.”12 This can reduce economic growth. Olson argued that interest groups tend to accumulate over time, though wars may weaken or destroy them. The longer a country has experienced a period of uninterrupted democracy, therefore, the greater the sclerosis and the slower the rate of economic growth. Olson noted, though, that “encompassing” interest groups — ones that represent a relatively large share of the population — have an incentive to act differently from smaller ones. Encompassing groups are more likely internalize the costs of rent-seeking and thus to engage in behavior that is good, rather than bad, for economic growth.

Olson thus offered two predictions about the impact of interest groups on economic growth. First, the number of years of uninterrupted democracy should be negatively related to economic growth. Second, the relationship between interest group encompassingness and economic growth should be U-shaped, with both low and high levels conducive to healthy performance. For this second hypothesis, I’ll use the unionization rate. Olson saw unions as the principal example of an interest group that may impede market functioning, both by influencing wages and employer behavior and by lobbying government for special privileges. The prediction is that high (encompassing) and low unionization rates should be better for economic growth, so I’ll use the absolute value of the unionization rate minus 50.

The two charts in figure 12 suggest no association between either of the interest group organization measures and catchup-adjusted economic growth over the 1979-2016 period.

Figure 12. Institutional sclerosis and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Uninterrupted democracy: years of uninterrupted democracy as of 1979. Data source: Mancur Olson, “The Varieties of Eurosclerosis: The Rise and Decline of Nations since 1982,” in Economic Growth in Europe since 1945, 1996. Unionization (distance from 50%): absolute value of union density over 1979-2016 minus 50. Data source: Jelle Visser, “ICTWSS: Database on Institutional Characteristics of Trade Unions, Wage Setting, State Intervention, and Social Pacts,” Amsterdam Institute for Advanced Labour Studies, version 4, 2013, series ud, ud_s. The lines are linear regression lines. In the second chart, the dashed line is with all 18 countries and the solid line is with Italy excluded. In the first chart, the correlation is +.04. In the second chart, the correlation is +.21 with all 18 countries and +.04 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

Corporatism

Regularized dialogue among organized interest groups and between interest groups and government — often referred to as corporatism or corporatist concertation — can produce less conflictual behavior and greater policy coherence. Government policy choices are likely to be based on more and better information, to be better coordinated across policy areas, and to be less subject to dispute and resistance once implemented. Corporatism has been linked with healthier economic performance in some studies.13

Some measures of corporatism focus on the degree or centralization of union and/or employer organization, while others concentrate on the degree or type of input organized interest groups have in policy making.14 I use a measure, from Alan Siaroff, which is a composite of 23 corporatism indicators. Figure 13 suggests a moderately strong correlation between this indicator and catchup-adjusted economic growth rates, but the association shrinks to nearly zero if Italy is excluded.

Figure 13. Corporatism and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Corporatism: composite of 23 measures. Data source: Alan Siaroff, “Corporatism in 24 Industrial Democracies: Meaning and Measurement,” European Journal of Political Research, 1999. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with France and Italy excluded. The correlation is +.35 with all 18 countries and +.12 with France and Italy excluded. “Asl” is Australia; “Aus” is Austria.

Left government

Political parties will tend to pursue policies that serve the real and/or perceived economic interests of their chief constituencies. For left parties this has tended to mean the working class and the poor; for right parties it has meant owners of capital and higher-paid employees. Douglas Hibbs and others have argued that left parties are therefore likely to implement macroeconomic policy strategies that aim for low unemployment (and thus rapid economic growth), while right parties are more likely to favor low inflation (and thus slower growth).15 Hibbs found support in a cross-country analysis and in an examination of the United States. Carles Boix found support in analyses of cross-country patterns, though he argued that the relevant policies are public investment and education rather than monetary and fiscal. Larry Bartels has updated the analysis of the U.S. case, again finding strong support.16

The standard indicator of left government is leftist parties’ share of parliamentary seats within the governing coalition. Figure 14 shows that there is no association between left government and catchup-adjusted economic growth across the rich democracies.

Figure 14. Left government and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Left government: share of seats in parliament held by leftist parties in the most recent government as a percentage of all seats held by the government. Data source: David Brady, Evelyne Huber, and John Stephens, Comparative Welfare States Dataset, 2016, series leftcab. The line is a linear regression line. The correlation is +.05. “Asl” is Australia; “Aus” is Austria.

Interest group-government coherence

Peter Lange and Geoffrey Garrett proposed a modification of the corporatism and left government hypotheses.17 They suggested that when unions are centralized and strong, left governments tend to pursue expansionary fiscal and social policy, knowing that doing so will not lead to wage militancy. When unions are fragmented and weak, right governments are able to implement free-market policies. Each of these scenarios, they argued, is conducive to economic growth. In the other two possible configurations — strong unions with right government, weak unions with left government — they predicted incoherence between government action and union wage behavior, and hence slower growth. Their analyses of patterns of economic growth in affluent democracies in the 1970s and 1980s suggested support for the hypothesis.18

I use Lange and Garrett’s “labor organization index” to create the union encompassingness component.19 Countries that score high on the index are Sweden, Austria, Denmark, Finland, and Belgium.20 Low-scoring countries are Japan, France, the United States, and Canada. New Zealand and Switzerland were not included in their scores; I add them to this low labor encompassingness group.21 For countries with high union encompassingness, I calculate the number of years since 1979 in which left party parliamentary seat share was 50% or more. For countries with low encompassingness, I calculate the number of years in which right party parliamentary seat share exceeded 50%. The measure is a count of the number of years in a country in which there were strong unions with left government or weak unions with right government. Higher scores reflect more interest group-government coherence.

There is no indication of the expected positive association between this measure and economic growth, as we see in figure 15.

Figure 15. Interest group-government coherence and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Interest group-government coherence: see the text. The lines are linear regression lines. The dashed line is with all 18 countries; the solid line is with Italy excluded. The correlation is +.10 with all 18 countries and -.11 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

Cooperation-promoting institutions

Long-term relationships and formal organization can promote cooperation within and between firms and between interest groups and the state. For example, long-term partnerships with suppliers potentially offer companies advantages relative to short-term, market-based supplier relationships or vertical integration. Such partnerships enable firms to reap the benefits of low fixed costs and supplier expertise while encouraging suppliers to invest in long-term improvements and to communicate extensively both with the purchaser and with other suppliers. Within the firm, functional specialization of the stages along the production chain research, design, development, production, and so on can render companies slow and ineffective at moving from creation to production. The process becomes disjointed, leading to less coherence, more delays, and higher costs as errors are discovered late in the process. Cross-functional project teams, whose members represent different departments and stay with a project as it moves through the various stages, can enhance coordination and continuity and thereby reduce the time and costs involved in bringing new products, or product improvements, to market.

Alexander Hicks and I identified cooperation-inducing institutions in nine spheres.22 Some apply to macro-level actors (government, union confederations, employer confederations), some to the “meso” level (between firms), and some to the “micro” level (within firms). They are: (1) relations among firms across industries; (2) relations among unions; (3) relations between the state and interest groups; (4) relations among firms and investors; (5) relations among firms and suppliers; (6) relations among competing firms; (7) relations between labor and management; (8) relations among workers; (9) relations among functional departments within firms. We hypothesized that countries with greater prevalence of cooperation-inducing institutions are likely to grow more rapidly.

We scored each nation on the prevalence (low, medium, or high) of cooperation-promoting institutions in each of these nine economic spheres. A factor analysis suggested two dimensions of cooperation: “firm-level cooperation” and “neocorporatism.” We found the former to be positively associated with economic growth in the 1960s, 1970s, and 1980s.

However, as the first chart in figure 16 shows, the average scores for this firm-level cooperation index do not correlate positively with economic growth over the years 1979-2016. Nor do the average cooperation scores for all nine economic spheres, as we see in the second chart.

Figure 16. Cooperation-promoting institutions and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Firm-level cooperation: see the text. Data source: Alexander Hicks and Lane Kenworthy, “Cooperation and Political Economic Performance in Affluent Democratic Capitalism,” American Journal of Sociology, 1998. Overall economic cooperation: see the text. Data source: Hicks and Kenworthy. The lines are linear regression lines. In the first chart, the correlations is +.07. In the second chart, the correlation is +.18. “Asl” is Australia; “Aus” is Austria.

Institutional coherence

Peter Hall and David Soskice’s “varieties of capitalism” perspective emphasizes five economic “spheres”: (1) industrial relations (bargaining over wages and working conditions); (2) vocational training and education; (3) corporate governance (relations between firms and their investors); (4) inter-firm relations (between firms and their suppliers, clients, and competitors); (5) relations with employees (information-sharing, work effort incentives).23 Their core hypothesis is that political economies tend to be characterized by “institutional complementarities,” whereby the presence of one institution increases the efficiency of another. For instance, “long-term employment is more feasible where the financial system provides capital on terms that are not sensitive to current profitability. Conversely, fluid labor markets may be more effective at sustaining employment in the presence of financial markets that transfer resources readily among endeavors thereby maintaining a demand for labor.”24

Hall and Soskice found that institutional complementarities do indeed tend to be present in the affluent OECD economies, and they suggested that these economies fall into two groups. Coordination is market-based in six “liberal market economies”: Australia, Canada, Ireland, New Zealand, the United Kingdom, and the United States. Coordination is based largely on nonmarket or extramarket institutions in ten “coordinated market economies”: Austria, Belgium, Denmark, Finland, Germany, Japan, the Netherlands, Norway, Sweden, and Switzerland.

Hall and Soskice hypothesized that economic performance is a function of institutional coherence. Both nonmarket- and market-oriented institutions can work well provided they are coupled with complementary institutions in other spheres. Institutional configurations that more closely correspond to either of the two pure types tend to promote growth.

Hall and Daniel Gingerich developed a “coordination index” that aims to gauge the degree to which countries rely on nonmarket economic institutions.25 The index is created via factor analysis of six indicators, each measured as of the early- or mid-1990s: (1) shareholder power; (2) dispersion of control; (3) size of the stock market; (4) level of wage coordination; (5) degree of wage coordination; (6) labor turnover. The factor analysis yielded a single factor, which is highly correlated with each of these six indicators. I have rescored the Hall-Gingerich factor scores to create an indicator in which more institutional coherence is scored high and less coherence is scored low.26

Figure 17 suggests no association between this institutional coherence measure and catchup-adjusted economic growth.

Figure 17. Institutional coherence and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Institutional coherence: see the text. The line is a linear regression line. The correlation is +.03. “Asl” is Australia; “Aus” is Austria.

Income inequality

Income inequality may be good for economic growth. If the financial payoff to innovation, hard work, or business success is sufficiently great, people might strive harder and thereby boost productivity. Moreover, the rich tend to save a larger portion of their income than those in the middle-class and below. To the extent this increases the supply of funds available for investment, this may increase economic growth.27

On the other hand, income inequality might be bad for economic growth. Because the rich spend a smaller fraction of their income than the middle class and the poor, greater inequality may reduce consumer demand. And people may not work as hard if they perceive the distribution of pay and income to be unfair.28

We have two main indicators of income inequality within countries. One is inequality between those at the top and everyone else, commonly measured via the top 1%’s share of income. The other is income inequality within the bottom 99%, which usually is measured using the Gini coefficient.29 Figure 18 suggests no noteworthy correlation between either of these indicators and economic growth rates in the rich democracies.22

Figure 18. Income inequality and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Top 1%’s income share: pretax income, excluding capital gains. Data source: World Inequality Database. Bottom 99% Gini: posttransfer-posttax income, adjusted for household size . Data sources: Luxembourg Income Study; OECD. The lines are linear regression lines. The dashed lines are with all 18 countries; the solid lines are with the United States (first chart) or Italy (second chart) excluded. In the first chart, the correlation is +.12 with all 18 countries and -.16 with the United States excluded. In the second chart, the correlation is -.14 with all 18 countries and -.02 with Italy excluded. “Asl” is Australia; “Aus” is Austria.

Ethnic diversity

Heterogeneity in a country’s population may reduce the likelihood of mutually beneficial economic cooperation and impede the formulation of stable and effective public policy. On the other hand, diversity in norms, customs, experiences, and knowledge may increase the supply of new ideas and enhance competition.30

Most studies of the impact of population diversity on economic performance have focused on ethnic diversity. Figure 19 offers little indication of an association between ethnic diversity and catchup-adjusted economic growth in the affluent democratic nations.

Figure 19. Ethnic diversity and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Ethnic diversity: Herfindahl index of ethnic group shares. 1990. Data source: Alberto Alesina, Johann Harnoss, and Hillel Rappoport, “Birthplace Diversity and Economic Prosperity,” Working Paper 18699, National Bureau of Economic Research, 2013, table A4. The correlation is -.17. “Asl” is Australia; “Aus” is Austria.

Trust

Trust lubricates economic relationships. It enables economic actors to cooperate in prisoners-dilemma-type circumstances, in which each has an incentive to act selfishly but both would benefit if each acts in a cooperative fashion.31 Much actual economic cooperation, however, can be traced to incentives rather than to trust.32 In any case, the pattern across the rich nations, shown in figure 20, suggests little reason to conclude that trust is an important determinant of economic growth.

Figure 20. Trust and economic growth
Catchup-adjusted economic growth: residuals from a regression of 1979-2016 growth rates on 1979 level of GDP per capita. Data source: OECD. Trust: share of adults responding “most people can be trusted.” Question: “Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people?” Data sources: World Values Survey and European Values Survey. The line is a linear regression line. The dashed line is with all 18 countries; the solid line is with France and Italy excluded. The correlation is +.25 with all 18 countries and +.08 with France and Italy excluded. “Asl” is Australia; “Aus” is Austria.

Conclusion: We don’t know why some rich democracies grow faster than others

The world’s rich democratic nations differ in their medium-to-long run rates of economic growth. Part of this owes to catch-up. But none of the other hypothesized causes examined here appears to be a robust contributor to the cross-country variation. The same holds in multivariate statistical analyses (not shown here).33

Another perplexing piece of the growth puzzle is the tendency of countries to do well for a while and then falter. In the past half century, any number of national models have gone in and out of fashion, first surging and then falling back. “Modell Deutschland” and “Japan Inc.” in the 1970s and 1980s, the “great American jobs machine” in the 1980s and 1990s, the “Dutch miracle” in the 1990s, and the “Celtic tiger” in the 1990s and 2000s are among the more prominent examples.

Though not widely appreciated, this hole in our understanding isn’t unknown. Paul Krugman wrote in 1994 that “There are many economic puzzles, but there are only two really great mysteries. One of these mysteries is why economic growth takes place at different rates over time and across countries. Nobody really knows why….” Two decades later his assessment hadn’t changed: “The reasons some countries grow more successfully than others remain fairly mysterious….”34 Patterns in the post-1970s era suggest that Krugman is correct. When it comes to the rich democratic nations, we have very little clue about what yields faster economic growth over the medium to long run.


  1. Benjamin M. Friedman, The Moral Consequences of Economic Growth, Knopf, 2005; Ronald Inglehart, “Modernization, Existential Security, and Cultural Change,” in Handbook of Advances in Culture and Psychology, volume 7, edited by Michele J. Gelfand, Chi-yue Chiu, and Ying-yi Hong, Oxford University Press, 2018; Lane Kenworthy, “Happiness,” The Good Society. 
  2. Friedrich Hayek, “The Price System as a Mechanism for Using Knowledge,” American Economic Review, 1945. 
  3. Daron Acemoglu and James Robinson, Why Nations Fail: The Origins of Power, Prosperity, and Poverty, Crown, 2012, pp. 75-76. 
  4. “The capitalistic economy of the present day is an immense cosmos into which the individual is born, and which presents itself to him, at least as an individual, as an unalterable order of things in which he must live. It forces the individual, in so far as he is involved in the system of market relationships, to conform to capitalistic rules of action. The manufacturer who in the long run acts counter to these norms, will … inevitably be eliminated from the economic scene.” Max Weber, The Protestant Ethic and the Spirit of Capitalism, 1905. 
  5. Eric A. Hanushek and Ludger Woessmann, “Do Better Schools Lead to More Growth? Cognitive Skills, Economic Outcomes, and Causation,” Journal of Economic Growth, 2012. 
  6. Adam Smith, The Wealth of Nations, 1776; Douglass C. North, Institutions, Institutional Change, and Economic Performance, Cambridge University Press, 1990; Michael Porter, The Competitive Advantage of Nations, Free Press, 1990; André Sapir et al, An Agenda for a Growing Europe: The Sapir Report, Oxford University Press, 2004; William J. Baumol, Robert E. Litan, and Carl J. Schramm, Good Capitalism, Bad Capitalism, and the Economics of Growth and Prosperity, Yale University Press, 2007. 
  7. Lane Kenworthy, “Trade,” The Good Society. 
  8. See also Lane Kenworthy, “Is Big Government Bad for the Economy?,” The Good Society. 
  9. Andrew Shonfield, Modern Capitalism, Oxford University Press, 1965; Chalmers Johnson, MITI and the Japanese Miracle, Stanford University Press, 1982; Ira C. Magaziner and Robert B. Reich, Minding America’s Business, Vintage, 1983; John Zysman, Governments, Markets, and Growth, Cornell University Press, 1983; Peter A. Hall, Governing the Economy, Oxford University Press, 1986;Ronald Dore, Taking Japan Seriously, Stanford University Press, 1987; Alice Amsden, Asia’s Next Giant: South Korea and Late Iindustrialization, Oxford University Press, 1989; Joseph E. Stiglitz, “On the Economic Role of the State,” in The Economic Role of the State, edited by Arnold Heertje, Basil Blackwell, 1989; Dani Rodrik, One Economics, Many Recipes, Princeton University Press, 2007. 
  10. Harold Wilensky and Lowell Turner (Democratic Corporatism and Policy Linkages, Institute of International Studies, 1987, p. 32) attempted to rank-order eight of the countries on degree of industrial policy intervention, based on their reading of the secondary literature. This ranking too is negatively associated with economic growth. 
  11. Bo Rothstein, The Quality of Government, University of Chicago Press, 2011. 
  12. Mancur Olson, The Rise and Decline of Nations, Yale University Press, 1982; Olson, “The Varieties of Eurosclerosis: The Rise and Decline of Nations since 1982,” in Economic Growth in Europe since 1945, edited by N. Crafts and G. Toniolo, Cambridge University Press, 1996. 
  13. Peter J. Katzenstein, Small States in World Markets, Cornell University Press, 1985; Wilensky and Turner, Democratic Corporatism and Policy Linkages. 
  14. Alan Siaroff, “Corporatism in 24 Industrial Democracies: Meaning and Measurement,” European Journal of Political Research, 1999; Lane Kenworthy, “Quantitative Indicators of Corporatism,” International Journal of Sociology, 2003. 
  15. Douglas A. Hibbs, Jr., “Political Parties and Macroeconomic Policy,” American Political Science Review, 1977. 
  16. Hibbs, “Political Parties and Macroeconomic Policy”; Hibbs, The American Political Economy: Macroeconomics and Electoral Politics, Harvard University Press, 1987; Carles Boix, Political Parties, Growth, and Equality, Cambridge University Press, 1999; Larry Bartels, Unequal Democracy, 2nd edition, Princeton University Press, 2016. See also Lane Kenworthy, “Do Election Outcomes Matter?,” The Good Society. 
  17. Peter Lange and Geoffrey Garrett, “The Politics of Growth,” Journal of Politics, 1985. 
  18. Lange and Garrett, “The Politics of Growth”; R. Michael Alvarez, Geoffrey Garrett, and Peter Lange, “Government Partisanship, Labor Organization, and Macroeconomic Performance,” American Political Science Review, 1991; Nathaniel Beck, Jonathan N. Katz, R. Michael Alvarez, Geoffrey Garrett, and Peter Lange, “Government Partisanship, Labor Organization, and Macroeconomic Performance: A Corrigendum,” American Political Science Review, 1993. 
  19. Alvarez, Garrett, and Lange, “Government Partisanship, Labor Organization, and Macroeconomic Performance,” p. 553. 
  20. Norway also scores high, but it is an economic growth outlier and so omitted here. 
  21. For justification, see Kenworthy, “Quantitative Indicators of Corporatism.” 
  22. Lane Kenworthy, “Is Income Inequality Harmful?,” The Good Society. 
  23. Peter A. Hall and David Soskice, “An Introduction to Varieties of Capitalism,” in Varieties of Capitalism, edited by Peter A. Hall and David Soskice, Oxford University Press, 2001. 
  24. Hall and Soskice, “An Introduction to Varieties of Capitalism,” p. 18. 
  25. Peter A. Hall and Daniel W. Gingerich, “Varieties of Capitalism and Institutional Complementarities in the Macroeconomy: An Empirical Analysis,” Discussion Paper 04/5, Max Planck Institute for the Study of Societies, 2004; Hall and Gingerich, “Varieties of Capitalism and Institutional Complementarities in the Political Economy: An Empirical Analysis,” British Journal of Political Science, 2009. 
  26. Lane Kenworthy, “Institutional Coherence and Macroeconomic Performance,” Socio-Economic Review, 2004. 
  27. Arthur Okun, Equality and Efficiency: The Big Tradeoff, Brookings Institution, 1975; Milton Friedman and Rose Friedman, Free to Choose, Harcourt Brace Jovanovich, 1979. 
  28. Lane Kenworthy, Egalitarian Capitalism, Russell Sage Foundation, 2004; Sarah Voitchovsky, “Inequality and Economic Growth,” in The Oxford Handbook of Economic Inequality, edited by Wiemer Salverda, Brian Nolan, and Timothy M. Smeeding, Oxford University Press, 2009; Joseph Stiglitz, The Price of Inequality, W.W. Norton, 2012; Thomas I. Palley, “Inequality, the Financial Crisis, and Stagnation: Competing Stories and Why They Matter,” Working Paper 151, Macroeconomic Policy Institute, 2015. 
  29. For details, see Lane Kenworthy, “Income Inequality,” The Good Society. 
  30. Alberto Alesina, Johann Harnoss, and Hillel Rappoport, “Birthplace Diversity and Economic Prosperity,” Working Paper 18699, National Bureau of Economic Research, 2013; Vincenzo Bove and Leandro Elia, “Migration, Diversity, and Economic Growth,” World Development, 2017. 
  31. Robert D. Putnam, “The Prosperous Community: Social Capital and Public Life,” The American Prospect, 1993; John Helliwell and Robert D. Putnam, “Economic Growth and Social Capital in Italy,” Eastern Economic Journal, 1995; Stephen Knack and Philip Keefer, “Does Social Capital Have an Economic Payoff? A Cross-Country Investigation,” Quarterly Journal of Economics, 1997. 
  32. Lane Kenworthy, “Social Capital, Cooperation, and Economic Performance,” in Beyond Tocqueville: Civil Society and the Social Capital Debate in Comparative Perspective, edited by Bob Edwards, Michael W. Foley, and Mario Diani. University Press of New England, 2001. 
  33. See, for example, Lane Kenworthy, “Institutions, Wealth, and Inequality,” in Oxford Handbook of Comparative Institutional Analysis, edited by Glenn Morgan, John L. Campbell, Colin Crouch, Ove Kaj Pedersen, and Richard Whitley, Oxford University Press, 2010. 
  34. Paul Krugman, Peddling Prosperity, W.W. Norton, 1994, p. 24; Krugman, “The New Growth Fizzle,” The Conscience of a Liberal, August 18, 2013. See also Ryan Avent, “Economists Understand Little about the Causes of Growth,” The Economist: Free Exchange, 2018.