Journal Issue

Growing Pains

International Monetary Fund. External Relations Dept.
Published Date:
June 2004
  • ShareShare
Show Summary Details

Dissecting the causes of Europe’s lagging economic performance

FOLLOWING World War II, European economies entered a 30-year period of rising prosperity. This golden age was one of rapid catch-up with the United States in terms of GDP per capita. The pattern of growth was consistent with the conventional “convergence” view: less advanced economies grow faster than more advanced ones. During the 1980s, however, this catch-up process paused before unraveling during the 1990s as per capita output grew more slowly in large European economies than in the United States. This European setback came as a surprise. In an ever more integrated world, with low barriers to trade, global financial markets, declining obstacles to foreign direct investment, and rapid technological diffusion, one would have expected convergence paths to accelerate.

These developments prompt a number of questions. What is behind the European slowdown relative to other countries in the 30-member Organization for Economic Cooperation and Development (OECD)? What policies and reforms would help rekindle economic growth? And how should these changes be implemented?

Divergent trends

Sizable differences in economic growth have materialized over the past decade within the OECD area. Looking beyond cyclical developments, growth in the United States and a number of other, mostly English-speaking economies, including Australia, Canada, and New Zealand, has averaged 3 percent or more. In contrast, long-term growth is estimated at about 2 percent in Europe as a whole and 1 percent in Japan. But nearly half the 1 percent growth gap between the United States and the European Union (EU) reflects differences in population growth. Moreover, the size of the differences is expected to widen in the years ahead because demographic decline is more advanced in the EU and Japan.

Since countries cannot do much about demography in the immediate future, comparisons of economic performance in terms of GDP per capita are more informative. These too reveal marked divergences. Over the 1990s, annual per capita output growth in the United States was half a percent higher than in the European Union and almost 1 percentage point above what was achieved in the three large euro area economies—France, Germany, and Italy. As a result of these trends, in 2002 per capita incomes converted using purchasing power parities (PPPs) were almost 30 percent lower in the EU than in the United States (see Chart 1). If, on the other hand, the rate of convergence recorded during the 1970s had been maintained, the three major euro area economies would by now have almost the same levels of output per capita as the United States.

Chart 1Wrong convergence

Real per capita GDP in the European Union is no longer catching up with that in the United States.1

Citation: 41, 2; 10.5089/9781451952957.022.A006

Source: Organization for Economic Cooperation and Development, Annual National Accounts, 2003.

Notes: EU-15: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, and United Kingdom.

Euro-3: France, Germany, and Italy.

EU smalls: EU countries excluding France, Germany, Italy, and the United Kingdom.

1Trend indices, based on 1995 purchasing power parity and 1995 prices. Trend calculated using a Hodrick-Prescott filter (smoothing parameter set to 100) over a period that includes projections through 2010.

Such broad-brush comparisons between Europe and the United States, however, conceal a wide disparity of economic performances within Europe. In fact, there have also been extraordinary success stories. For example, in Ireland, average per capita output during the 1990s expanded by almost 6½ percent, the fastest pace recorded in any OECD country. As a result, Irish GDP per capita has risen from a level well below the OECD average to one of the highest. Greece (albeit from a low rate), Luxembourg, and the Netherlands also managed to boost per capita output growth over the 1990s, as did Finland and Spain over the second half of the decade.

Short-run performances have also diverged in unexpected ways, following the bursting of the bubble in high-technology investment spending. Even though the adverse demand shock had its epicenter in the United States, its effects hit continental Europe with as much vigor as the United States. Other OECD economies resisted better, especially English-speaking countries and, to a lesser extent, the Nordic ones.

Sources of divergence

All in all, these statistics suggest that continental Europe has suffered from relatively weak performance, in terms of both long-term growth and short-term resilience in the face of conjunctural shocks. The OECD is investigating why some economies are resilient and others are not and has examined in depth the sources of long-term growth (OECD, 2003). What singles out the fast-growing economies is a rare ability to combine a high degree of labor utilization with strong productivity. In contrast, the two principal sources of weakness in Europe are the trend decline in labor utilization and, to a lesser extent, in labor productivity.

Low labor utilization rates. Overall, about 30 percent of the working-age population in the European Union is neither employed nor seeking work, compared with less than 25 percent in the United States. Moreover, in many European economies, especially the larger ones, the employment rate has fallen over the past couple of decades while it has remained broadly steady in the faster-growing OECD economies. This dichotomy explains around 85 percent of the gap in Europe’s GDP per capita relative to that of the United States (see Chart 2).

Chart 2Not working enough

Poorer labor utilization helps explain differences in per capita GDP

(percentage point differences in trend, per capita GDP with respect to the United States, 2000; based on purchasing power parity)

Citation: 41, 2; 10.5089/9781451952957.022.A006

Source: Organization for Economic Cooperation and Development.

1Based on the ratio of working-age population (15–64 years) to total population. The demographic effects are not shown (less than 5 percent for all countries).

2Based on employment rates and average hours worked.

3GDP per hour worked.

4Lowest OECD refers to Norway; the highest OECD refers to Mexico.

The declines in the European employment rate have affected older workers more. For instance, in Belgium, France, and Italy, around one-third or less of the population aged 55–64 has a job while the employment rate in Europe as a whole for this group is just under 40 percent. Other contributing factors include a slower trend rise in female labor force participation—especially of women with limited affordable access to child care services—high marginal effective tax rates, lower average hours worked per employed person, and higher structural unemployment.

A key issue, then, is what has led to this low level of labor utilization? At the risk of oversimplifying, two polar explanations are feasible. First, an argument could be made that Europeans have a greater preference for leisure. In other words, when they become wealthier, Europeans prefer to trade higher incomes for more leisure time. But it could equally be that low labor utilization reflects European labor market and tax policies that have the effect of weakening the incentives to work and of reducing the demand to hire employees.

Rewards of better policy

Growth regressions help policymakers evaluate the potential effect of a given change in policy. Adjusting a variable in the left column can have a significant impact in the long run.1

Effect on output per working-age person

VariableThrough economic efficiencyThrough investmentOverall
Rate of inflation fall of 1 percentage point0.4 to 0.50.4 to 0.5
Variability of inflation, fall of 1 percent in the standard deviation of inflation2.02.0
Tax burden, increase of 1 percentage point3–0.3–0.3 to –0.4–0.6 to–0.7
Business research and development intensity, increase of 0.1 percentage point31.21.2
Trade exposure, increase of 10 percentage points34.04.0
Human capital, additional year of education4.0 to 7.0
Source: Organization for Economic Cooperation and Development.

The values reported are the estimated long-run effects on output per working-age person of a given policy change. The range reported reflects the values obtained in different specifications of the growth equation.

The direct effect refers to the impact on per capita output over and above any potential influence on the accumulation of physical capital. The indirect effect refers to the combined impact of the variable on the investment rate and, through that channel, on per capita output.

Percentage of GDP.

Source: Organization for Economic Cooperation and Development.

The values reported are the estimated long-run effects on output per working-age person of a given policy change. The range reported reflects the values obtained in different specifications of the growth equation.

The direct effect refers to the impact on per capita output over and above any potential influence on the accumulation of physical capital. The indirect effect refers to the combined impact of the variable on the investment rate and, through that channel, on per capita output.

Percentage of GDP.

Which explanation fits better is a matter of empirical analysis. Although we cannot say much about the relative preferences for leisure on the two sides of the Atlantic, we can say, based on OECD and other evidence, that institutions and structural policies matter for labor utilization (OECD, 1999). This is the case, for example, with labor market policies that subsidize preretirement schemes and financially penalize workers willing to stay at work beyond the legal retirement age. Such policies were implemented in many European economies based on the mistaken belief that, by removing workers from the labor force, unemployment would be reduced. The evidence shows that these policies had a negative impact on the employment rate, particularly of older workers.

We have also found that low labor utilization in Europe is connected to limited labor market flexibility. Restrictive employment protection laws, for instance, tend to dampen both hiring and firing and lengthen average unemployment spells. Through time, some long-term unemployed become discouraged and permanently exit the labor market. Similarly, a high minimum cost of labor provides a disincentive to recruit unskilled and inexperienced workers. And, in many countries, the interplay of tax and benefit systems results in unemployment and poverty traps.

Labor productivity differs. Lower labor productivity is a second principal reason why Europe’s GDP per capita lags behind that in the United States. It accounts for about 15 percent of the gap between European and U.S. GDP per capita, though the variation between countries is large. In more than one-fourth of the EU economies, labor productivity is higher than in the United States. On the face of it, therefore, there is less scope to improve output in Europe by boosting productivity than there is by raising the employment rate.

This interpretation may, however, be too simplistic because European statistics are somewhat flattering. Indeed, in Europe many people who have below-average productivity are not employed and, thus, are not included in the measurement of labor productivity. Across OECD countries, there is a strong negative correlation between employment rates and labor productivity, although such cross-country comparisons do not amount to a full-fledged explanation of productivity differences. From this perspective, it is reasonable to conclude that many EU countries lag behind the United States in terms of labor productivity and that one challenge is to rekindle productivity growth.

While governments cannot fine-tune productivity, OECD empirical research stresses the importance of good policies and institutions. In this context, it is important that inflation be low and stable and that fiscal policy not be procyclical. Macroeconomic stability, however, is a necessary, though not sufficient, condition. Structural policies that boost flexibility and sharpen an economy’s adaptability to shocks are also required. Policies that promote educational achievements, well-designed incentives for research and development, and highly competitive and open markets are some of the levers governments can use to boost productivity. Together, these policies promote efficiency and, ultimately, innovation and the adoption of new technologies.

All these influences can be estimated and put together in a coherent analytical framework using, for instance, crosscountry panel data analysis. The results of such a quantitative exercise suggest that the benefits from sound policy can be large. For example, it is estimated that if those sectors that lag in terms of productivity were to modernize their regulatory framework and align it on best practice, their productivity could increase by as much as 10 percent. Similarly, one additional year of education raises output in the long run by about 6 percent, bringing high returns to individuals and society as a whole. Other examples of the estimated long-run effect of given changes in policy and institutional variables are reported in the table.

EU economies are less resilient. Growth decompositions traditionally abstract from short-run cyclical developments. This partly reflects the fact that there has not been much to separate the cyclical profile of business cycles. Nonetheless, what may be less appreciated is that those economies that have experienced high long-term growth have also generally been less prone to accidents and have recovered quickly when accidents have occurred. Such economies could be said to be resilient. Indeed, there is prima facie evidence suggesting that countries with stronger long-term growth may also have gone through less protracted economic downturns during the world slowdown that started in 2001. This, in turn, suggests that the structural policy settings that foster good trend growth may also be conducive to good short-run performance.

To be sure, while structural policies cannot immunize an economy from the business cycle, they can limit the length and magnitude of deviations in output from potential. The channels of operation, however, are less well understood than they are for long-term economic performance. The OECD is investigating further why some economies are resilient, and why others, mostly in the EU area, are not. It is not an easy exercise because many factors are at play. Initial findings, not surprisingly, underscore the role of macroeconomic stabilization policies. But that does not offer a complete explanation. After all, many of the fast-growing OECD economies had relatively conservative macroeconomic policies, suggesting that other factors played a part. Indeed, the weaker resilience appears to be largely the consequence of structural policies and institutions. For example, highly restrictive employment regulations and, in some countries, the interaction of tax and benefit systems are important sources of economic sclerosis that prolong unemployment spells and delay labor reallocation and wage adjustment.

Implementing reforms

Europe has had some success in implementing structural reforms over the past five years or so. Many European economies have managed to generate a “labor deepening process” resulting in stronger job creation and lower unemployment. Labor market policy initiatives have operated both from the demand side of the market by, for instance, cutting nonwage labor costs and promoting wage moderation and, later, from the supply side through in-work benefits and tax credits designed to fight poverty and unemployment traps. A number of governments have also tightened access to early retirement schemes and introduced greater flexibility through an expansion of temporary and part-time employment contracts.

But progress has been uneven, and, in many instances, the initiatives taken, while very welcome, have been the easiest to implement. As we have learned from the OECD’s best-performing economies, structural policies that promote competition in product markets and enhance the flexibility of an economy, combined with good macroeconomic policies, are essential for fostering both short- and long-term success in economic performance. A lot of work remains to be done to restore healthy incentives to work and encourage entrepreneurship. For instance, tax and benefit systems should properly balance safety net and incentive considerations, employment protection arrangements should not inhibit hiring, and more focused active labor market programs are often needed.

These OECD policy prescriptions have a broad consensus among economists. Yet it has often been politically difficult to translate this professional consensus into concrete policy reform. The challenge facing policymakers is how to overcome this resistance. There is no self-evident or sure-handed way of doing this. The best one can do is observe the conditions and methods adopted by those countries that have been successful in implementing structural reforms and attempt to distill the lessons learned.

On this basis, and in recognition of the diversity of country experiences, there appear to be two common elements. First, it is always important to identify, quantify, and disseminate estimates of the costs of maintaining existing policies and how they are likely to evolve. This process helps to put into perspective the short-term adjustment costs against the long-term costs associated with inaction. The OECD can play a supporting role in this process through its reports, based on impartial cross-country analyses. However, even though the benefits of reform are greater than the costs, often resistance to change is large and well organized. This is partly because of vested interests and the dilemma of weighing short-term pain (borne by a small fraction of the community) against the long-term gains (which are widely dispersed).

“A lot of work remains to be done to restore healthy incentives to work and encourage entrepreneurship.”

The second common element of successful implementation of structural reforms is the bundling of reforms. Such an approach has several advantages. It allows for the design of a coherent reform package, including, where considered appropriate, measures to compensate groups for the impact of changes that conflict with policy objectives in other areas. Furthermore, a broad-based reform package is more likely to result in a more balanced distribution of the adjustment costs and benefits of reform, thereby softening resistance to change.

The best time to implement reforms is open to debate. It is sometimes argued that reforms are more easily introduced when an economy is growing fast. This may help to ensure that displaced labor is more quickly reallocated, but a strong economy also tends to make the need for reforms more difficult to sell in the first instance. In contrast, if it is widely recognized that slow growth reflects, in part, the structural problems faced by an economy, the costs of inaction come to the fore, leading to a general acceptance that something must change. Examples are New Zealand and Australia in the early 1980s, when these countries saw their ranking, in terms of per capita GDP, decline and initiated reforms in response. It may also currently be a dynamic operating in Germany, where a wide-ranging and ambitious structural reform program has recently been outlined.

Jean Philippe Cotis is Chief Economist at the OECD.


Other Resources Citing This Publication