Only a few years ago the contrast between a dynamic, high-employment American labour market and stagnant, low-employment labour markets in Europe was a major theme in international political economy and labour market economics. While the low wages, inequality and precariousness associated with the American model were recognised and widely deplored, the U.S.’s low rates of unemployment and swift recovery from recessions during the 1980’s and 1990’s led even those on the left to query aspects of the European social model. European labour market performance was increasingly evaluated and criticised through the prism of the ultra-flexible, minimum social protection American system.

The ultimate expression of the contrast- the withdrawal of minimal federal safety net provision for the poor under the Clinton welfare reforms of the 1990’s- even came to be taken seriously as a model for welfare reform among some social democrats. Thus in a promotional piece for Tony Blair’s ‘legacy’ programme of welfare reform in 2006, Will Hutton cited the falls in welfare receipt following the Clinton reforms and made the case for tough policy choices: ‘New Labour has tried hard, but has never felt able to reproduce the robustness of Clinton's measures in a British context........ Too many British live on benefit for no better reason than they don't want to work....Part of the problem is that too many in progressive Britain still do not want to come to terms with the facts.’

Similar sentiments are still voiced today- underpinned by a belief that failing to take a sufficiently aggressive approach to welfare in its last term contributed to Labour’s 2010 election defeat. Whether or not that is the case- and those who assert it have as yet to specify any plausible counterfactual in terms of which the claim might be assessed- the advantages previously claimed for the American labour market model are now looking distinctly threadbare. Since the late 1990’s the performance of the U.S. labour market has been remarkably weak, even allowing for the impact of the recessions of 2001 and 2008. The high rates of prime-age male employment of the late 1990’s have never even been approached in the new century, and uniquely among wealthy nations, the U.S. shows a long-term decline in employment and economic activity for prime-age women. These trends contrast sharply with robust growth in employment for men and women in comparable European countries, including the U.K., up to the 2008 financial markets crisis.

Even for welfare ‘doves’ unsympathetic to the case for 'hawkish’ Clinton style reforms, these findings come as something of a surprise. Economic theory predicts that minimal regulation and low social protection, other things being equal, tend to lead to higher levels of employment and more rapid adjustment to changes in demand for labour. At the end of the twentieth century American and European labour markets seemed in their different ways to be behaving in exactly the way textbooks predicted. But since the turn of this century, the American advantage in employment over Europe has disappeared.

We concentrate on ‘prime age’ workers, aged 25-54, as their employment is less affected by education and retirement policy variables than other age groups. For most of the 1980’s and 1990’s, the U.S. had higher rates of employment for men aged 25-54 than the European economies, and the gap grew over time, reaching its widest point in the mid-1990’s. However, with the collapse of the dotcom bubble, U.S. employment rates for prime-age men rapidly converged downward with Europe. Employment recovery in the U.S. was mediocre even before the cataclysmic impact of the recent recession while European employment showed strong growth from mid-decade, meaning that over most of the last decade it was on par with the U.S., something which had not occurred over an extended period since the early 1980’s.

For women aged 25-54, the story is one of more gradual erosion of the American employment advantage. In 1981, just under half of European women in this age group were employed compared to over 60% in the U.S. Despite these very different baselines, trends in women’s employment were extraordinarily similar in terms of the rate and tempo of growth up to the mid-1990’s, at which point growth the U.S. first levelled off, and then went into reverse. Meanwhile women’s employment continued its secular rise in Europe, converging with the U.S. in 2008.

The disappearance of the American employment advantage over Europe over the last decade seems to have received little comment to date. This may be because it occurred over a period in which the U.S. experienced two recessions in relatively rapid succession, making it difficult to assess the balance between conjunctural and longer term influences. However the pattern shown by employment rates is mirrored in data on economic activity, which is somewhat less sensitive to the economic cycle. This shows gradually falling activity for prime age women from the late 1990’s in the U.S., a surprising finding given the welfare reforms and strong employment growth of that period. Breaking the prime age group into five year age bands shows moderate falls in activity from the mid- to late-1990’s at all ages except those aged 50-54, with some differences in timing. In general the fall is more marked for younger age bands (1 to 2 percentage points) but for all bands except those over 50 the overall pattern is that the increases in activity seen during the first half of the 1990’s might as well never have happened. For men the pattern by age is different, with 30-44 year olds showing little change from the mid-1990’s but substantial falls (of 2-3 percentage points) for younger (25-29) and older (45-54) men.

At the same time, comparing the U.S. with a narrower sample of wealthier European nations dispels any suspicion that this change in comparative performance is an artefact of syncopated business cycle movements. For prime age men, the CWE nations (= Continental Western European: France, Germany, Austria, Belgium, Netherlands) had already converged with the U.S. by the time the bubble peaked, while the Nordic countries (=Denmark, Finland, Norway, Sweden) emerged mid-decade with far stronger employment growth than the U.S. For prime age women, CWE nations have exceeded stagnant American employment rates since the turn of the century, while in the Nordic countries, which have higher rates for women in any case, employment showed much more convincing growth with the mid-decade upturn. Taking cycles into account strengthens rather than weakens the sense that American employment performance has been weak compared to that of European countries since at least the turn of the century.

What are the lessons from all this? We need to know more about the drivers of this reversal in comparative performance before any firm conclusions can be drawn. But there is one obvious lesson for those in Labour who continue to believe Labour should take inspiration from the brutal bi-partisan welfare reforms of the Clinton era: cutting benefit rolls is not the same thing as raising employment. More tentatively, the abysmal U.S. employment performance since the turn of this century makes the previously unthinkable possibility that welfare reform had no positive economic impact in the medium term not only possible but plausible. If that turns out to be the case, economic failure would be added to political failure (as convincingly argued here ) and social disaster in the long-term judgment on the Clinton reforms.

Data comes from a study I've been working on on comparative employment performance: employment from OECD labour market database, activity from ILO Economically Active Population and Projections dataset.

Acknowledgment: to Paul Gregg for numerous discussions on this, and to Jane Waldfogel for contextual detail and references. Usual disclaimer applies.