post-autistic economics review
Issue no. 40, 1 December 2006
article 1



      issue 40 contents                            PAE Review index                               home page



Is the U.S. a Good Model for

Reducing Social Exclusion in Europe?

John Schmitt and Ben Zipperer   [Center for Economic and Policy Research, USA]





            Sustained, high levels of unemployment in the majority of Europe's largest economies have led many Europeans to look to the United States as a possible alternative economic model. The political right and center in Europe have emphasized what they see as the flexibility and dynamism of the U.S. economy. Much of the left, meanwhile, have argued that high unemployment in Europe, which is often concentrated in specific geographic regions or demographic groups, is the driving force behind "social exclusion" in Europe today. This has led many Europeans – even some in the continent’s social democratic parties – to the reluctant conclusion that the United States may be a good model for reducing social exclusion there.


            This paper reviews several international indicators of social exclusion to assess how well the United States has done in using its apparently greater flexibility and dynamism to reduce social exclusion. On most measures of inequality, poverty, health, education, crime, and punishment, the United States does not fare well compared to the much-better-funded welfare states in Europe. The gap between U.S. and European performance in many of these dimensions is striking, and not fully acknowledged in the current debate around promoting U.S.-style reforms in Europe. What is more surprising, however, is that the United States, in fact, performs poorly in two areas where U.S. superiority is usually simply taken for granted: incorporating traditionally disadvantaged groups into the paid labor force and providing opportunities for economic mobility.



Income inequality


            We start with what is probably the most basic indicator of social exclusion – household income inequality. Table 1 presents data on income inequality for 28 OECD countries in various years during the 1990s and the year 2000 from Smeeding (2004). (All tables are located here.) The final column of the table, which reports data on the Gini coefficient1, the most common measure of income inequality, shows that the United States (0.37) had the second highest Gini coefficient among the countries with available data – only Mexico (0.49) had higher income inequality by this measure. The United Kingdom (0.35) was the European country with the next highest level of income inequality, followed by Ireland and Italy (both 0.33), with most of the remaining countries in Europe below 0.30. The countries with the lowest Gini coefficients were Denmark (0.24), Belgium (0.25), Finland (0.25), Germany (0.25), the Netherlands (0.25), Norway (0.25), and Sweden (0.25).2


            Another basic measure of income inequality is the distance between the 10th, the 50th, and the 90th percentiles of the national income distribution. The greater the distance between points in the distribution, the greater the overall inequality. The first column of Table 1 demonstrate that, in the United States, the 10th percentile household earned about 39 percent of what the median household earned, while the 90th percentile household (see column two) earned about 210 percent of the median. The 10th percentile earner in the United States was further below the median than was the case in every other country in the table except Mexico (28 percent). In every European country except Italy (44), Ireland (46), and the United Kingdom (47), the 10th percentile household made at least 50 percent of median earnings. Among the major OECD economies, 10th percentile households fared best in Norway (57), Sweden (57), and the Netherlands (56).




What is Social Exclusion?


The term social exclusion has had a prominent place in the European debate on social problems and policies. The term grew out of a desire to encourage a richer discussion ofeconomic and social inequality and deprivation, which had traditionally focused onincome-based measures of inequality and poverty.


The British government, which has established a Social Exclusion Unit, states that socialexclusion is "...about more than income poverty. Social exclusion happens when people orplaces suffer from a series of problems such as unemployment, discrimination, poor skills, low incomes, poor housing, high crime, ill health and family breakdown. When such problems combine they can create a vicious cycle."



According to Ruggeri Laderchi, Saith, and Stewart (2003): "The first use of the term[social exclusion] has been attributed to Lenoir, French Secretary of State for SocialAction in Government in 1974, referring to people who did not fit into the norms of industrial societies, were not protected by social insurance, and were considered socialmisfits." (p. 21) The term grew to encompass "...processes of marginalisation anddeprivation which can arise even within rich countries with comprehensive welfare provisions." (p. 21) Today, they note, "[t]he concept now forms a central aspect of[European Union] social policy." (p. 22)




Meanwhile, the 90th percentile household in the United States (210) was further above the median than in almost every other country in the table. Only Mexico (328), Luxembourg (215), and the

United Kingdom (215) had larger gaps between the 90th percentile and the median. Incomes at the top were closest to the median in Denmark (155), Slovakia (162), Finland (164) , and the Netherlands (167).


            The third column in the table calculates the ratio of the 90th and 10th percentile earnings, as an additional measure of income inequality (see Figure 1). Mexico (11.55) had, by far, the highest



 FIGURE 1. Household income inequality (ratio of 90th to 10th percentiles)

  Source: Smeeding (2004)


inequality using this simple gauge of inequality. The United States (5.45) was next, well ahead of the United Kingdom (4.58), Australia (4.33), and Canada (4.13). The countries with the lowest "90-10" gap were Norway (2.80), Denmark (2.85), Slovakia (2.88), Finland (2.90), and the Netherlands (2.98).


            By most measures, the United States is the most unequal of the major OECD countries, with a higher Gini coefficient, lower relative incomes at the 10th percentile, and a bigger gap between the incomes of rich and poor households than in any of the countries in Western Europe. Whatever capacity the United States might have for using its labor-market flexibility and dynamism to create jobs and channel potential workers into employment (which we examine below), this capacity has not avoided the emergence of substantial levels of income inequality with the resulting potential for heightened levels of social exclusion.





            Income inequality is, in and of itself, a cause for social concern,3 but poverty – extreme relative or absolute deprivation – is generally seen as a more important indicator of potential social exclusion. As Townsend (1979) argues: those in poverty have “resources... so seriously below those commanded by the average family or individual that they are in effect excluded from ordinary living patterns, customs and activities.”4


            Table 2 presents data from Scruggs and Allan (2005) on relative and absolute measures of poverty at different points in time over the years 1990 to 2000 for a subset of the countries in the earlier figures on income inequality. The first column of Table 2 contains data on the relative poverty rate, defined as the share of the population in households with incomes below 40 percent of the median (which is obviously closely related to income inequality). Consistent with the earlier results for income inequality, the United States (10.7 percent) had the highest rate of relative poverty, followed

by Ireland (8.0) and Italy (7.3). Relative poverty was lowest in Finland (2.1), Norway (2.8), Belgium (3.2), France (3.3), and Sweden (3.6).



    FIGURE 2. Absolute Poverty Rate (percent of population)

     Source: Scruggs and Allan (2005)



With respect to absolute poverty (see column two of Table 2 and Figure 2), defined here as earning at least 40 percent of the inflation-adjusted 1986 median income in the United States (converted to local currencies using purchasing power parity exchange rates), the United States, which has a much higher GDP per capita than most of the other countries in the sample,5 does substantially better. About 8.7 percent of the U.S. population was living in poverty by these criteria, well below rates in Italy (18.8), Australia (16.4), Ireland (15.4), and the United Kingdom (11.8). The United States also does somewhat better than France (10.0). The rest of the European countries in the table, however, have lower absolute poverty rates, despite also having income levels that are 70 to 80 percent of U.S. levels. Norway (which has a GDP per person close to that of the United States) had an absolute poverty rate of only 2.6 percent; the rate in Switzerland was 3.5 percent.6





            Education is arguably the single most important tool available to combat social exclusion. Table 3 shows the educational attainment rates, standardized by the OECD, for our sample of OECD countries for 2003. The first two columns examine the share of the adult population with at least an upper-secondary education (roughly the equivalent of a high-school degree in the United States). The first column gives the figures for all adults age 25 to 64. The United States had the highest share of high-school-equivalent graduates, with 88 percent. Norway (87) and Slovakia (87) trailed close

behind. In most of the rest of Western Europe between 60 and 80 percent of 25-to-64 year olds had completed the equivalent of high school. The biggest exceptions in Europe were Portugal (23), Spain (43), Italy (44), and Greece (51).


            European countries do considerably better, however, when we focus on just 25-to-34 year olds (see the second column of Table 3). High-school completion rates for this younger group are generally much higher than for the full 25-to-64 year olds, while rates are almost identical across the two age ranges in the United States (87-88 percent). Nevertheless, the United States generally still does better than European countries do. The exceptions are Finland (89), Sweden (91), Czech Republic (92), Slovakia (94), and Norway (95); while Austria (85), Germany (85), and Denmark (86) do not lag far behind the United States.


            The last two columns of the same table show the share of the adult population with roughly the equivalent of a four-year college degree or more. Once again, the United States, with 38 percent of 25-to-64 year olds with college degrees (see column three), does well compared to Western Europe. Only Denmark (31), Norway (31), and Sweden (33) have at least 30 percent of their adult populations with college degrees. Most Western European countries fall in the 20-30 percent range, with several in the teens.


            When we look just at 25-to-34 year olds (see column four), many European countries do almost as well or better than the United States (39 percent) with respect to college graduates: Denmark (35), France (37), Ireland (37), Spain (38), Belgium (39), Finland (40), Norway (40), and Sweden (40). Several Western European countries, however, still lag far behind the United States: Italy (12), Austria (15), Portugal (16), Germany (22), and Greece (24).


            Attainment rates are only one way to measure the potential for educational outcomes to contribute to social exclusion. Table 4 presents results tabulated by the OECD from an international standardized test of mathematics administered to 15-year-olds. In Western Europe, only Greece (445), Italy (466), and Portugal (466) scored, on average, lower than the United States (483) (see Figure 10). Switzerland (527), Belgium (529), the Netherlands (538), and Finland (544) did the best in Western Europe (see Figure 3).


            For purposes of social exclusion, however, we may be particularly interested in the scores of the poorest-performing students. The first column of Table 4, therefore, also shows the 10th percentile test scores in each country. In Western Europe, only Greece (324), Italy (342), and Portugal (352) scored lower than the United States (356). The best performers in Western Europe with respect to students at the 10th percentile were Ireland (393), Denmark (396), Iceland (396), Switzerland (396), Netherlands (415), and Finland (438). (For completeness, the last column in the table displays the results at the 90th percentile.)


            Table 5 demonstrates that the United States does poorly at both the mean and the 10th percentile7 despite spending substantially more on education at the primary ($8,049 per student) and secondary ($9,098) level than almost every other country in the OECD. Only Luxembourg spends more at both levels ($10,611 for primary and $15,195 for secondary), and Norway more at the secondary-school level ($10,154). (The data in the next-to-last column demonstrate that at the tertiary level, the United States does spend substantially more per student per year ($20,545) than all other European countries except Switzerland ($23,714). These expenditures, of course, have no direct impact on test scores of 15-year-olds.) As Table 6 makes clear, the vast majority of these expenditures at the primary and secondary level in the United States are in public schools (3.8 percentage points of U.S. GDP in 2002), not in private schools (only 0.3 percentage points of GDP in the same year).8



  FIGURE 3. Mathematics performance among 15-year olds, 2003 (PISA mathematics
  scale scores)


                  Source : OECD





            The United States spends much more on health care than any other country in our sample. Table 7 lists total expenditures on health care in 2003, separately for the public and private sectors, based on calculations by the OECD. The first three columns express expenditures as a share of national GDP. The United States spent 15.0 percent of its GDP on health care in 2003 (see Figure 4). The next closest countries were Switzerland (11.5) and Germany (11.1); only three other countries spent more than ten percent (Iceland, 10.5; Norway, 10.3; and France, 10.1). Since U.S. GDP per capita is substantially higher than most of the countries in our sample, the gap between U.S. expenditures and those in other countries are even greater when we express health-care costs in terms of expenditures per person per year, which we do in the last three columns of the table. On average, the United States spends about $5,635 on health care per person per year. Of the remaining countries, only four others spend more than $3,000 per person per year: Norway ($3,807), Switzerland ($3,781), Luxembourg ($3,705), and Canada ($3,001).



  FIGURE 4. Annual health-care expenditures, 2003 (percent of GDP)

              Source: OECD



            Table 7 also breaks down health-care expenditures by whether they are in the public or private sector. The United States is the only country, except Mexico, in which expenditures in the private sector (8.3 percent of GDP) exceed those in the public sector (6.7). Greece and Switzerland are the only other countries where private-sector health expenditures exceed 40 percent of the total. Even though private expenditures represent the bulk of health expenditures in the United States, public-sector health costs in the United States still fall in about the middle of the range for public expenditures in Western European countries. Denmark (7.5), France (7.7), Sweden (8.0), Norway (8.6), and Germany (8.7) spend more in their public sectors, but Austria (5.1), Finland (5.7), Greece (5.1), Ireland (5.8), Italy (6.3), the Netherlands (6.1), Portugal (6.7), Spain (5.5), Switzerland (6.7), and the United Kingdom (6.4) all spend the same or less than the United States does.


            The data in Table 7 establish that the United States spends considerably more on health care than other rich countries do, but other data suggest that the United States nevertheless suffers from high levels of social exclusion with respect to health care. The most obvious element of this exclusion is the high share of the U.S. population without health insurance. The United States and Mexico are the only countries in Table 7 that do not provide essentially universal health-care coverage. In 2003, 15.6 percent of the U.S. population (about 45 million people or roughly the population of Spain) was without any form of health insurance, public or private, throughout the entire year.9 An additional 12 percent of the U.S. population lacked health insurance for any part of the year.10


            Data on many of the most common health indicators also suggest that the U.S. health-care system is highly inefficient, yielding poor outcomes despite high levels of expenditures. Table 8 provides details on several broad measures of health outcomes compiled by the OECD. Only Mexico and the transition economies of Eastern Europe have a lower overall life expectancy than


   FIGURE 5. Life expectancy (years)

   Source: OECD



the United States (77.2 years, identical to Denmark, see column three of Table 8 and Figure 5.) On average, residents of Spain (80.5), Switzerland (80.4), and Sweden (80.2) – the three countries with the longest life expectancies in our sample – live three full years longer than residents of the United States. Among the major OECD economies, the United States also has the highest rate of infant mortality (7.0 per 1,000 live births, see column four). The next-highest rate in Western Europe is in the United Kingdom (5.3), while Norway (3.4), Finland (3.1), and Sweden (3.1) have rates that are less than half of those in the United States. The United States also fares poorly with respect to maternal mortality (see column five). At the turn of the century, the United States had 9.1 maternal deaths per 100,000 births, the fourth- highest rate in the table behind Mexico (70.7), Denmark (11.1), and Luxembourg (10.9).11 As with infant mortality, many Western European countries had maternal mortality rates that were less than half those in the United States: Ireland (3.1), Italy (3.1), Austria (3.6), Greece (3.9), Spain (4.2), Sweden (4.2), and Germany (4.3).


            The United States also has a much higher share of its population that exceeds the medical standard of obesity (a body mass index, BMI, of 30 or greater). Just over 30 percent of adults in the United States are obese, compared to 23.0 percent in the United Kingdom, the Western European country with the highest rate of obesity; meanwhile, Switzerland (7.7), Norway (8.3), Italy (8.5), Austria (9.1), France (9.4), Denmark (9.5), and Sweden (9.7) all have obesity rates below ten percent.


Public-health campaigns against smoking, however, have apparently been much more successful in the United States than they have been in most of Europe. Only 17.5 percent of U.S. adults smoke cigarettes daily (see the last column of Table 8). In Western Europe, only Sweden (17.5) has a rate as low. Most of Western Europe has smoking rates around 25 percent, with rates above 30 percent in the Netherlands (32.0), Greece (35.0), and Austria (36.3).


The United States spends markedly more on health care (as a share of GDP or in dollars on a per person basis) than any other country in the world. Yet, more than 15 percent of its population typically finds itself without health coverage – private or public – throughout the entire length of any given year, with 27 percent lacking coverage at some point during the year. The additional U.S. expenditures on health care are also associated with substantially worse outcomes for basic health indicators including life expectancy, infant and maternal mortality, and obesity. The United States, however, has succeeded in lowering rates of adult smoking to the lowest level among the rich, industrialized countries.



Crime and Punishment


Another potential dimension of social exclusion is crime. Table 9 summarizes some basic indicators of both the prevalence of criminal activity, as well as the associated incarceration rates.

The most reliable crime data are for murders, since murders are generally reported and accurately recorded. The first column of the table gives the murder rate for our list of countries, based on data compiled by the UK Home Office. The United States, at 5.6 murders per 100,000 people, has by far the highest murder rate in the sample of countries in the table. Finland (2.9) is next, followed by Slovakia (2.6), the Czech Republic (2.5), and New Zealand (2.5). The U.S. murder rate is about five times higher than the rate in the safest Western European countries: Austria (1.2), Germany (1.2), Portugal (1.2), Spain (1.1), Sweden (1.1), Switzerland (1.1), and Denmark (1.0).


The United States does substantially better with respect to self-reported victimization rates, falling near but not at the top of the countries in Table 9. The second column of the table shows criminal victimization rates, expressed as reported offences per 100 people, from the 2000 International Crime Victims Survey.12 In Western Europe, Switzerland (42.6 per 100 per year), Sweden (45.6), the Netherlands (48.1), and the United Kingdom (54.5) had higher victimization rates than the United States (39.5), while Denmark (35.1), France (33.9), Belgium (33.3), Austria (31.4), Finland (28.6), and Portugal (25.8) were all below the U.S. rate.


Given that the United States has high, but not the highest overall, victimization rates, all else constant, we might expect the United States to fall somewhere near the top, but not at the top of the sample of countries when it comes to the portion of its population that is incarcerated. The last two columns of Table 9, which report prison-population rates from the International Center for Prison Studies, demonstrate however, that the United States has a prison-population rate (724 per 100,000) that is five to ten times higher than rates in Western Europe, where incarceration rates range from 68 in Norway to 143 in Spain and Luxembourg and 144 in the United Kingdom. Most of Western Europe, in

fact, has incarceration rates below 100, including Finland (75), Denmark (77), Sweden  (78),



   FIGURE 6. Prison population rate (number of prisoners per 100,000 people)

  Source: International Centre for Prison Studies (2006)



Switzerland (83), Ireland (85), France (88), Belgium (90), Greece (90), Germany (97), and Italy (97) (see Figure 6).


            The magnitude of the incarcerated population in the United States is sometimes difficult to comprehend. In 2004, U.S. prisons and jails held 2.1 million inmates, about 90 percent of whom were men.13 Given that the adult male workforce age 16 and older in the same year was about 78.7 million,14 this implies that a staggering 2.3 percent of the adult male population of the United States was in prison or jail in 2004.



Labor Market


            Based on the evidence reviewed so far, the U.S. economic and social model appears to generate a considerable degree of social exclusion, with high levels of income inequality, high relative and even absolute poverty rates, poor and unequal educational outcomes, poor and unequal health outcomes, and high rates of crime and incarceration. The U.S. model maintains its appeal in the face of poor performance in these areas, however, because supporters believe that the United States offers two compensating advantages: a flexible economy that yields high employment rates, and high income mobility that, in principle, compensates for greater inequality.


            As the first column of Table 10 demonstrates, the U.S. experience with overall unemployment (5.6 percent in 2004) is good, and certainly far better than in Germany (9.9), France (9.6), and Spain (11.0).  At the same time, several Western European countries, with decidedly less “flexible” labor markets in the usual sense of that term, had unemployment rates in 2004 that were the same or lower than the United States: Ireland (4.4), Switzerland (4.4), Norway (4.5), the Netherlands (4.7), the United Kingdom (4.7), Austria (5.3), and Denmark (5.3).


            Despite the alleged superiority of U>S.-style flexibility, the United States does not do much better when it comes to unemployment rates for typically marginalized groups such as young people and those with less education, the kinds of groups most likely to benefit from greater wage flexibility, for example. The third column of Table 10 reports the unemployment rate for 15-to-24 year-olds. The rate in the United States (11.8 percent) is well below rates in France (21.3), Italy (23.5), and Spain (22.0), but above rates in Switzerland (7.7), Denmark (7.8), the Netherlands (8.0), Ireland (8.1), the United Kingdom (10.9), Austria (11.0), Germany (11.7), and Norway (11.7). (The unemployment rate, and even the employment rate, for youth does not necessarily paint an accurate picture of how well the labor-market is performing for young people, since many young people are probably best off in school. We will examine this issue below.) The fourth column shows a similar pattern for those with the equivalent of less-than-a-high-school education. The U.S. unemployment rate for this group (in 2002) was 9.9 percent, higher than the corresponding rates in Norway (3.9), Portugal (5.7), Sweden (6.1), Switzerland (6.1), Ireland (6.3), Greece (6.6), United Kingdom (6.9), Denmark (7.2), and Austria (7.9).


            The unemployment rate, however, is not the only measure of labour-market performance.  The next four columns of Table 10 give the employment-to-population rates for different demographic groups. Among 15-to-64 year olds, the United States does manage to incorporate more of the population into jobs (71.2 percent) than is the case in several major European economies, most notably France (62.8), Germany (65.5), Italy (57.4), and Spain (62.0) (see Figure 7). Nevertheless, many smaller, "less flexible" Western European economies have higher employment rates than the United States: the United Kingdom (72.7), the Netherlands (73.1), Sweden (73.5), Norway (75.6), Denmark (76.0), and Switzerland (77.4).15


            The United States has done well in incorporating women into the paid labor force.  But, the data in column six show that many Western European countries have also succeeded in this respect. In 2004, 65.4 percent of U.S. women ages 15-to-64 were employed. This was substantially higher than the corresponding rates in Italy (45.2), Spain (49.0), France (56.9), and Germany (59.9). The U.S. rates, however, are not as high as those in many European economies: Finland (65.5), the Netherlands (65.7), the United Kingdom (66.6), Switzerland (70.3), Sweden (71.8), Denmark (72.0), and Norway (72.7).


            Employment rates for youth (column seven) repeat the now familiar pattern.  The United States does better than the large, high-unemployment economies, but not as well as a host of smaller European economies. For youth, employment rates in the United States were 53.9 percent in 2004, well above the rates in Italy (27.2), France (29.5), Spain (38.4), and Germany (41.9), but not as high as rates in Norway (54.4), the United Kingdom (60.1), Denmark (61.3), Switzerland (62.0), and the Netherlands (66.2).



      FIGURE 7. Employment-to-population rate (percent employed, all individuals ages 15-64)

   Source: OECD



            With respect to employment rates for the less-educated, the United States actually underperforms when compared with much of Western Europe. In 2003, 58 percent of the less-educated population in the United States was in work. This rate was near or below rates in Ireland (57), Spain (57), Finland (58), Greece (58), France (59), Denmark (61), Norway (62), Switzerland (66), Sweden (68), and Portugal (72).


             Earlier, we mentioned that using the unemployment rate (and even the employment rate) to measure social exclusion among youth may be misleading. From a societal perspective, we may be just as concerned about whether young people are in school as we are about whether they are in work. The last three columns of Table 10, therefore, report OECD data for 2002 on the share of young people ineach country that were neither in work nor in employment. The United States does not do particularly well among either 15-to-19 year olds or 20-to-24 year olds. For the younger group, only Hungary (8.0), the United Kingdom (8.6), Italy (10.5), and Finland (14.8 percent) had a higher share of young people out of both work and school (the U.S. rate was 7.5 percent). For the next-older age group, the United States (15.6) does better than some Western European economies – Germany (15.9), Belgium (17.4), Finland (18.8), Greece (22.0), and Italy (24.3) – but not as well as Denmark (7.3), the Netherlands (7.9), Norway (9.7), Switzerland (9.7), Ireland (10.8), Sweden (11.2), Austria (11.7), Portugal (12.0), France (14.4), Spain (15.1), and the United Kingdom (15.3).


              The review of these data suggests that U.S. labor-market performance is generally – though not always – better than that of the four, large, high-unemployment European economies (France, Germany, Italy, and Spain). Nevertheless, the United States consistently underperforms relative to many of the smaller Western European economies whose labor markets are conventionally seen as much more rigid than those of the United States.



Economic Mobility


            Advocates of the U.S. model also maintain that the country's economic dynamism produces a level of economic mobility that compensates for high levels of inequality and poverty. Economic and social distances may be much greater in the United States than they are in Europe, but, the argument goes, those at the bottom have a much greater chance to get ahead than they do in Europe. In this final section, we briefly review some international evidence on economic mobility both within and across generations.


            Table 11 and Figure 8 present OECD data on short-term income mobility for a subsample of 14 countries. The table gives the share of low-income families (where low-income was defined as earning less than half of the national median income) that managed to escape from low-income status over a three-year period in the mid-1990s.16 Contrary to the view that the United States offers substantial mobility, the United States has the lowest share of low-income workers that exit their low-income status from one year to the next (29.5 percent). The corresponding rates in several European countries are greater than 50 percent: Ireland (54.6), the Netherlands (55.7), the United Kingdom (58.8), and Denmark (60.4).



  FIGURE 8. Income mobility, late 1980s-mid 1990s (percent of low-income families
  exiting low-income status each year)

              Source: OECD



            Table 12 summarizes the results from three separate studies of longer-term intergenerational mobility across countries. In all three cases, the studies investigated the degree of correlation between fathers' and sons' incomes at different points in time. These intergenerational income coefficients quantify the economic advantage conferred by parents to their children: the higher the coefficient, the more likely that children born to poor parents are to remain poor later in life.


            Panel (a) summarizes Blanden's (2004) findings for Canada, Germany, the United Kingdom, and the United States. Blanden found the lowest level of correlation between fathers’ and sons’ incomes – therefore, the highest degree of economic mobility-- in Germany (0.12), followed by Canada (0.18) and the United Kingdom (0.27). Intergenerational economic mobility was lowest, by a substantial margin, in the United States (0.45).


            Panel (b) presents similar correlation coefficients from a review of international studies by Solon (1992).17 The 0.40 coefficient for the United States is Solon's estimated average based on research in the United States. According to these data, only South Africa (0.44) and, in one of two studies, the United Kingdom (0.57), had lower rates of mobility than the United States (0.40) did. Canada (0.23), Finland (0.13 and 0.22), Germany (0.11 and 0.34), and Sweden (0.13, 0.14, and 0.28) all appear to have substantially greater economic mobility across generations than the United States does.


Corak's (2004) review (see panel (c)) reaches similar conclusions. The United Kingdom (0.50) and the United States (0.47) have the least economic mobility. France (0.41), Germany (0.32), Sweden (0.27), Canada (0.19), Finland (0.18), Norway (0.17), and Denmark (0.15) all offer greater economic mobility than the United States.


            What appear to be small differences in intergenerational income coefficients actually imply substantial differences in economic mobility. Take, for example, the case of a family with earnings that are half of the national average. Other factors held constant, if a country has a correlation coefficient for parent-child earnings of 0.20, we would expect that descendants of the poor family would reach the average national earnings in less than two generations, or about 25 to 50 years.18 In countries with a coefficient of 0.45, a typical level in the estimates for the United States (and, in some cases, for the United Kingdom), however, descendants of the poor family would not, on average, close the income gap with the average family for more than three generations, or about 75 to 100 years.





            The U.S. economic and social model is associated with substantial levels of social exclusion, including high levels of income inequality, high relative and absolute poverty rates, poor and unequal educational outcomes, poor health outcomes, and high rates of crime and incarceration.


            At the same time, the available evidence provides little support for the view that U.S.-style labor-market flexibility dramatically improves labor-market outcomes. The U.S. labor market appears to fare consistently better than the four large, high-unemployment economies in Europe – France, Germany, Italy, and Spain – but the U.S. does no better and often noticeably worse than many smaller European economies that have labor markets that are highly regulated relative to the United States and even relative to the labor markets in the large, high-unemployment countries.


            The data also appear to contradict the belief that greater economic mobility in the United States can somehow compensate for greater levels of inequality and "social exclusion." Despite popular prejudices to the contrary, the U.S. economy consistently affords a lower level of economic mobility, both in the short-term (from one year to the next) and in the longer-term (across generations), than all the continental European countries for which data are available. Given the high direct levels of social exclusion in the United States and especially the low levels of economic mobility across generations, the United States, therefore, stands as a poor model for a Europe seeking to combat social exclusion.






1. The Gini coefficient varies from zero to one. A Gini coefficient of zero would indicate perfectly equal distribution of income across all households; a Gini coefficient of one indicates that all income is concentrated in one household.


2. The Gini coefficients in the text are calculated using net disposable income, which subtracts taxes and includes transfer benefits. When measured using pre-tax income, the United States is not such an outlier. Using pre-tax income the Gini coefficient in the United States (0.45) lies well within the European range of market income inequality (0.39 to 0.50). Progressive taxes and especially benefits and transfer payments dramatically reduce inequality in most European nations, with only relatively modest effects in the United States.


3. See, for example, Groningen Growth and Development Centre and the Conference Board, Total Economy Database, May 2006,


4. Townsend (1979), p. 31.


5. See, for example, Groningen Growth and Development Centre and the Conference Board, Total Economy Database, May 2006,


6. Smeeding (2006) defines poverty as half of national median income and finds the pattern of poverty remains largely the same in the analysis by Scruggs and Allan (2005).


7. The relative performance of the United States is only marginally better at the 90th percentile, as Table 4 also shows.


8. In the United States, private educational expenditures are more important at the tertiary level, where the country spends about 1.2 percentage points of GDP on public higher education and 1.4 percentage points on private higher education.


9. See Carmen DeNavas-Walt, Bernadette D. Proctor, and Robert J. Mills, “Income, Poverty, and Health Insurance Coverage in the United States: 2003,” Washington, DC: U.S. Census Bureau (August 2004), p. 14.


10. The data refer to 2002, from Boushey (2004).


11. Since only a very small share of women die in childbirth, the data for maternal mortality, which are typically presented per 100,000 births, can vary substantially from year to year. As a result, Table 8 presents maternal mortality data averaged over the five most recent (available) years. For small countries with few births per year, even a small number of relatively bad years can have a relatively long-lasting impact on maternal mortality rates.


12. Total of ten crimes: car theft; theft from car; motor-cycle theft; bicycle theft; burglary; attempted burglary; robbery; personal thefts; and assaults or threats.


13. See, U.S. Department of Justice, Office of Justice Programs, Bureau of Justice Statistics, Bulletin, "Prison and Jail Inmates at Midyear 2004," April 2005.


14. Bureau of Labor Statistics, Current Population Survey home page,, customized tables, series LNS11000001Q, for second quarter 2004, which corresponds most closely to the mid-year 2004 prison and jail estimates.


15. Schmitt and Baker (2006) find that the declining coverage rate of the Current Population Survey (CPS) in recent decades may lead the CPS, which is the source of the U.S. unemployment and employment rate figures cited here, to overstate employment in the United States by about 1.4 percentage points, with the largest biases for more marginalized groups, especially young black men and young Hispanic women. To the extent that European surveys do not suffer from similar problems, the comparison here would overstate the U.S. performance relative to Europe.


16. The data for the United States refer to 1987-1989. The OECD notes that: "The time periods used to study poverty dynamics in the different countries are not fully comparable. The most important instance of non-comparable time periods is that poverty dynamics for the United States are studied for an earlier period ... than that studied for the other countries, due to data consistency problems in the American data for more recent years. Although the periods chosen are those for which business cycle conditions in the United States approximated those in the other countries studied, this difference means that the results do not reflect the impact on American poverty dynamics of recent reforms in welfare programmes and more generous in-work benefits (i.e. expansion of the Earned Income Tax Credit). On the other hand, the PSID data for income years after 1992 show greater poverty incidence and persistence in the United States, so that the use of these data would reinforce the comparative results for the United States. Exclusion of these data can be regarded as representing a somewhat conservative approach to the assessment of American poverty."


17. Some countries have more than one study.


18. Intergenerational mobility coefficients are determined by the regression: ln Yi,t = α + βln Yi,t-1 + εi,t, where generations are indexed by t. If Gt = Y1,t/Y2,t and non-parental income influences are ignored (ε = 0), the income gaps between two sets of parents, G0, and their respective children, G1, satisfy G1 = G0β. Similarly, Gn = G0β^n, which implies that n = ln (ln Gn/ln G0)/ln β. The calculations above assume G0 = 2 (the 200 percent gap between the mean and half the mean), Gn=1.05 (only a five percent gap) , and a generation equals 25 years.






Blanden, Jo. 2004. "International Evidence on Inter-generational Mobility," Centre for Economic Performance, unpublished paper.

Boushey, Heather. 2004. “Analysis of the Upcoming Release of 2003 Data on Income, Poverty, and Health Insurance,” Center for Economic and Policy Research Briefing Paper, (January).

Corak, Miles. 2004. "Do poor children become poor adults?" Statistics Canada, unpublished paper.

DeNavas-Walt, Carmen, Bernadette D. Proctor, and Robert J. Mills. 2004. “Income, Poverty, and Health Insurance Coverage in the United States: 2003,” Washington, DC: U.S. Census Bureau.

International Centre for Prison Studies. 2006. "Entire World - Prison Population Rates per 100,000 of the national population,", downloaded, May 2006.

Gordon Barclay and Cynthia Tavares. 2001. "International comparisons of criminal justice statistics 2001," London: United Kingdom Home Office, Research Development and Statistics Directorate (October 24).

Kesteren, J.N.van, P. Mayhew, and P. Nieuwbeerta. 2000. "Criminal Victimisation in Seventeen Industrialised Countries: Key-findings from the 2000 international Crime Victims Survey," The Hague: Ministry of Justice.

Mauer, Marc. 2003. "Comparative International Rates of Incarceration: An Examination of Causes and Trends," paper presented to the U.S. Commission on Civil Rights, Washington, DC: The Sentencing Project.

Navarro, Vicente. 2002. The Political Economy of Social Inequalities: Consequences for Health and Quality of Life. Amityville, New York: Baywood Publishing Company.

OECD. 2005. Education at a Glance 2005, Paris: OECD.

Schmitt, John and Dean Baker. 2006. "Missing Inaction: Evidence of Undercounting of Non-Workers in the Current Population Survey (CPS)," Center for Economic and Policy Research Briefing Paper. Washington, DC: CEPR.

Smeeding, Timothy M. 2004. "Public Policy and Economic Inequality: The United States in Comparative Perspective," Luxembourg Income Study Working Paper Series No. 367 (February).

----------. 2006. "Poor People in Rich Nations: The United States in Comparative Perspective," Journal of Economic Perspectives, vol 20, no. 1 (Winter), pp. 69-90.

Solon, Gary. 2002. "Cross-Country Differences in Intergenerational Earnings Mobility," Journal of Economic Perspectives, vol 16, no. 3 (Summer), pp. 59-66.

Townsend, Peter. 1979. Poverty in the United Kingdom. Harmondsworth: Penguin.