EU-US convergence ? — Crooked Timber

The NYT ran yet another round in the long-running EU vs US series a week or so ago. Although it’s not covered explicitly in the NYT, there is actually some news to report here, in addition to rehearsal of the same old themes.

For quite some time, the US and the leading EU countries have been fairly comparable in terms of output per hour worked. The US has had higher output per person for two reasons: a relatively high employment/population ratio and very high average hours worked per person. The first of these is important because it raises the possibility that EU countries performing well on productivity measures are benefiting from the “Thatcher effect” . If low-skilled workers are excluded from employment, for example by restrictive macro policy, as in Thatcher’s case, or by labor market sclerosis, as claimed by critics of European institutions, then productivity measures are artificially boosted.

This issue is now moot. As a result of the crisis, the US employment/population ratio has dropped sharply, to the point where the US is now little different from the EU. The difference in GDP per person between the US and leading European countries is driven primarily by differences in average hours worked by employed people.

To get the data on this, I’ve had to combine Eurostat and OECD info (always a little problematic, but neither had all the info I wanted).

From Eurostat, the E/P ratio (total employment/pop 15-64) for the euro area was 58.5 in 1997 and rose to 64.8 by 2009 (France 64.2 , Germany 70.0). Over the same period, the US ratio has fallen from 73.5 to 67.6, with the bulk of the decline in the last couple of years. The remaining difference is entirely due to the higher US employment-population ratio for women – the ratios for men are virtually identical.

Turning to the OECD for information on productivity and GDP per capita, these tables shows that relative to the euro area as a whole, the US still has a substantial lead in productivity (about 15 per cent). But for the leading European economies, like France, Germany and the Netherlands, the productivity gap is below 10 per cent, which is well within the margin of error associated with PPP conversions[1]. Particularly for the latter two, the big difference is in annual average hours worked (1681 for the US, 1390 for Germany, 1378 for the Netherlands). The difference in average hours almost entirely explains the gap in GDP per person between Germany and the US, and more than explains the gap for the Netherlands.

As is well known, Europeans tend to offset their lower hours of paid work by doing more household labor. Taking this into account properly would diminish the gap in both directions – relative to the US, European hours of work would rise, and so would output per person.

I was hoping for a good exposition of this from Peter Baldwin whose book The Narcissism of Minor Differences: How America and Europe are Alike has a promising title (I haven’t read it yet). Unfortunately, he only gets half of the story, saying

Americans work 23 percent more than Germans in the marketplace. However, once we factor in household labor, the drudgery that allows us to function in the world, the difference in total work drops to 12 percent. And interestingly, the figures for time actually spent at leisure are almost precisely the same for the two nations.

That Americans work 12 percent more than Germans seems to be the hard kernel that emerges from the statistics. Considering that for that 12 percent investment the American G.N.P. per capita is 32 percent higher than the German, this seems a defensible trade-off. Perhaps Americans have collectively decided to work somewhat harder to be substantially better off.

The problem here is that Baldwin has missed the point that household labor is productive.

Coming to my own take on all this, it seems that the European and US systems yield roughly equal productivity, and roughly equal labor market performance (as measured by E/P ratios). Higher European taxes mean more and better public services (at the cost of reduced private consumption) and they are also (along with social preferences) reflected in lower hours of work and more household labor. I know which looks more appealing to me, but there’s no obvious way of saying which is best.

Rather more clear-cut is the price paid by the US in terms of greater inequality. Compared to the European case, and to the US in the past, the top percentiles of US households collect a much larger share of total income, and there doesn’t seem to be any net economic payoff for this.

fn1. (Very wonkish note) Although PPP numbers are often treated as if they are are raw facts, they are index numbers which are fundamentally imprecise (even if the underlying data is perfectly accurate, which it isn’t). From work I did with Steve Dowrick in the 1990s, I estimate the difference between upper and lower bounds at around 10 per cent. It’s likely that any bias in PPP numbers favors the US. That’s because they are a generalized kind of Laspeyres index, and (as I understand it) the base data is derived largely from Europe.

20 thoughts on “EU-US convergence ? — Crooked Timber

  1. Thank you Prof Quiggin.

    I am curious as to why you didn’t consider your own country in this comparison though. Australia would make for an interesting divergence from both systems.

    I’d also be very interseted in your thoughts on the changes in wealth distribution in Australia over recent years/decades. Our unrelenting real estate boom must have had some impact on that, in constrast to the US.

  2. Thanks Prof, interesting analysis of a very interesting issue. However, one issue which you seem to overlook is the likely substantial direct effect on welfare of having a job. There are considerable differences in structural unemployment rates in the US and Europe.

  3. john, good post on an important subject

    you say that “As is well known, Europeans tend to offset their lower hours of paid work by doing more household labor.”

    Do americans have exceptionally untidy houses? What do you mean by this?

    do europeans like to have choice about this additional household labour.

    who in the market supplies these household services in the USA. Is it a low skilled?

    david autor found rising employment and wages in low-skilled in-person service occupations in the USA.

    he found that the share of U.S. labor hours in service occupations grew by 35 percent between 1980 and 2005, after having been at or declining in the three prior decades. Accompanying their rising employment, real wage growth in service occupations substantially outpaced that in other low skill occupations, averaging seven percent per decade between 1980 and 2005.

  4. John,

    Bob Gordon has written extensively on the Trans-Atlantic income and productivity gap.

    I hope the following is a fair summary of his work:

    1. After fifty years of catching up to the U. S. level of productivity, since 1995 Europe has been falling behind.
    2. The growth rate in output per hour over 1995-2003 in Europe was just half that in the United States, and this annual growth shortfall caused the level of European productivity to fall back from 94 percent of the U. S. level to 85 percent.
    3. Fully one-fifth of the European catch-up (from 44 to 94 percent) over the previous half-century has been lost over the period since 1995.
    4. Gordon’s explanation is that Europe’s previous catch-up to the U. S. productivity level before 1995 was artificial – by making labor expensive – Europe raised its average product.
    5. After 1995 as labor market reforms have helped to make EU labor cheaper, so the average product of labor has been pushed down and its growth rate has faltered.
    6. The falling behind of Europe is quite simple to explain; European regulations defied the laws of labor market equilibrium for decades, and a slight leak in the dam holding back market forces has already led to faster growth in European employment and slower growth in productivity.
    7. Gordon’s explanatory variables for both the revival of EU employment growth and the slowdown in productivity growth include six policy and institutional variables.
    8. These are the tax wedge, employment policy legislation, product market regulation, the average replacement rate of unemployment insurance, union density, and an index of “high corporatism”.
    9. Gordon found that that several of these variables have significant negative effects on employment per capita, with policy changes that raised labor costs reducing employment before 1995, and those that reduced labor costs increasing employment after 1995.

  5. Jim, the point of the post is that virtually all of Gordon’s analysis is now obsolete. It depends, almost entirely, on the factual claim that the US economy does a better job of mobilising available labor. As of 2010, that claim is longer consistent with the data.

    The massive shift in relative E/P ratios, in favor of Europe has been accompanied by a small gain in relative US productivity. But almost all the benefits of this have gone to the top percentile of the income distribution.

  6. @jquiggin
    a sample of two and one-half is a significant break in a trend that is 50 years long? are we comparing the cycle with the trend or the other way around.

    You also say that “Higher European taxes mean more and better public services (at the cost of reduced private consumption) …’

    Any social expenditure comparisons between the USA and EU are distorted because of no GST in USA! Which side of 20% is the GST in European countries is paid on almost every dollar spent by the poor to fund middle class welfare? Little wonder that Rawls disliked welfare states because of their treatment of the poor.

    As another example, the Nordics collect income taxes on social welfare cash payments at rates that are four to five times the rates paid by American recipients. Then these Nordic welfare recipients go out to but anything, they pay consumption tax on their purchases that are 4 to 5 times the rate paid by the poor in America.

    The USA is a safety-net society: workers and people with adequate incomes purchase directly, or receive through their employer, private life insurance, private health insurance, and private pension plans. There is no churning through the tax system, which discourages work, savings and investing in human capital.

    The U.S. also offers a wide of tax breaks such as the earned income tax credit for low-income workers with families and generous tax breaks for medical expenses and for private pensions. Unfortunately many of these poor miss out on tax credits because they do not file tax returns. Medicaid and CHIPs targeted poor and working poor families.

    The Nordic countries are more focused on providing the same basic benefit to everybody, while the U.S. focus is on providing benefits once someone is in trouble.

    Director’s law may well rule in the EU: the bulk of EU public programs are designed primarily to benefit the middle classes but are financed by taxes paid primarily by the upper and lower classes. The middle class will always be the dominant interest group in a modern democracy. The middle class will use its majority voting power to maximise the state benefits it receives and minimise the portion of the costs it bears.

    Any country admired by progressives, and especially those in the Northern Europe, should be examined closely through the lens of Director’s law of public expenditure.

  7. “a sample of two and one-half is a significant break in a trend that is 50 years long?”

    Umm, the 50-year trend you mention is EU-US convergence, as in the title of this post. Gordon is the one claiming a break in trend – it now appears that claim was premature.

  8. @jquiggin
    thanks, but what EU convergence, post-war reconstruction and common market aside?

    Ed Prescott is keen on the productivity gains from economic unions such as the common market, but he may not be your natural first port of call.

    To quote from Gordon again:

    “Starting from barely half the level of productivity and per?capita income as the United
    States in 1950, Europe began a rapid catch?up. While Europe’s level of productivity almost
    reached the US level in 1995, its income per person never exceeded 75 percent and has since fallen below 70 percent.”


    “The problem is that Europe has failed to catch up to the US level of output per capita since 1975, with the EU?US ratio of output per capita roughly fixed at around 70 percent of the US level.”

  9. @Jim Rose

    Jim rose – I take it you are referring to a recent paper by Price Fishback , which uses OECD net social expenditure data (google it).

    What the OECD does is look at aggregate spending components, not the distribution of benefits.

    Social spending includes social security and welfare spending, health care spending and community services (for example, child care and nursing homes.) However, they do include private social spending as well as public spending, so private health spending in the USA gets counted as social spending.

    The figures also take account of the impact of direct and indirect taxes both in reducing benefits and in substituting for benefits or subsidising benefits. For example, as you point out Nordic countries tax benefits rather heavily and indirect taxes are much heavier, so when you take these into account “net” spending is much less than gross spending. This does not occur anywhere to the same extent in the USA.

    In addition, the US provides support for families through the tax system in the form of the EITC and child tax credits, whereas in the Nordic countries, virtually all family support is in the form of direct benefits or services.

    For example, Denmark spends nearly 27% of GDP on the welfare state compared to 16% of GDP in the USA on a gross basis, but on a net basis (after subtracting taxes on benefits and addiing tax expenditures), Denmark spends 20.2% of GDP and the USA spends 20.1% – almost exactly the same.

    However, net private social expenditure in Denmark is only 1.3% of GDP while in the USA it is 9.4% of GDP – which is almost entirely due to the very high level of private health care spending in the United States.

    So overall the USA spends much more as a society on social welfare than Denmark, mainly due to heavy taxes on benefits in the Nordic world, but also because of the extraordinarily high US private health care spending.

    But this does not mean that the USA is more generous to the poor than Denmark. All of the Nordic countries direct a higher proportion of their spending to the poor than does the USA. Tax expenditures in the USA tend to be less progressive than direct expenditures. And private health care spending is not a benefit to the poor.

    For example, calculations i have made and contained in
    show that after taking account of the level of spending on benefits, the degree to which it is targeted to the poor and the direct tax clawbacks that you refer to that the poorest quintile in the USA receive net transfers of 1.9% of household disposable income, whereas Sweden and Denmark – and in passing also Australia – transfer nearly 6% net of household disposable income to the poorest 20%. Indirect taxes will claw back more – about 20% of these levels in the Nordic countries and 5% in the USA – but even after these effects the USA spends a lot less on the poor than most other developed countries, and in fact more middle class welfare – it is just that it is delivered through the tax code in the USA.

  10. As I think is obvious even to the non-economists here (i.e. me), the point is, most economic measures that economists choose to care are at best unhelpful, and more often than not harmfully misleading.

    To what extent does the above debate help to determine which is the more ecologically sustainable society, the more just, the more caring, the more human? Hard to tell (but I know where I’d put my money).

  11. @Peter Whiteford
    Thanks for your considered and thoughtful posting.

    It is not usual to conduct comparison at this level of distributional details, although it should be at this level of detail, when people kneel and worship at the feet of European welfare states.

    Only when caught out about the high level of middle class welfare and churning in the old EU are these important details explored under the full glare of the public eye, such as after Price Fishback’s recent writings and respones to him.

    The slightly better amount of welfare transfers in the EU to buy the political silence of the very bottom of the income distributions is surrounding by a large amount of labour supply and investments depressing taxation, social insurance and employment protection.

    Some of the EU governments mitigate the burden on income per capita the best they can of funding an extensive transfer society – albeit with scraps from the table for the poor – by using:
    • efficient taxes such as a GST;
    • lower tax rates on capital income;
    • relatively light levels of product market in some cases, and even light labour market regulation in also some cases according the various indexes of economic freedom – Denmark is ranked higher then NZ! and
    • the competitive and regulatory disciplines and scale economies of a large common market.

    Relative European income per capita has been stagnant since the mid-1970s because of welfare states and excessive labour market regulation and high taxes. Hours worked per working age person has collapsed in many European countries relative to countries where you get to keep a much larger share of the fruits of your labours such as the USA and Australia.

    A recent papers by Rogerson show that between 1956 and 2003, hours worked in Europe decline by almost 45% compared to the US. This change is almost an order of magnitude larger than the effects associated with the increase in unemployment over this time period.

    The decline in hours worked in Europe is almost entirely accounted for Europe developing a much smaller service sector than the US.

    The services sector is a major source of jobs for low-paid and low-skilled workers

    The poor pay a high price in Europe for middle class welfare, tax churning and employment protection for incumbent prime age middle class workers. Do the EU welfare states pass muster under Rawl’s difference principle?

  12. @Jim Rose

    European welfare states are very diverse and I think that your characterisation of the poor paying a high price for European welfare states applies only in some cases.

    Have a look at which gives figures from the 1998 OECD report Growing Unequal, which shows the PPP adjusted level of household income at different points in the income distribution, i.e. these are absolute income levels and they take account of consumption taxes in determining purchasing power. (They don’t include government non-cash benefits which would tend to favour the poor in Europe more than the USA.)

    Now the US comes out as having the second highest average income level in the OECD, but the average for the poorest decile in the USA is lower than in every European country north of Spain and west of the Czech republic. The average incomes of the second decile is higher in absolute terms in 12 European countries, and the average for the third decile is higher in 8 European countries, and all of these countries have extensive welfare states, so the argument that these welfare states have made the poor worse-off is difficult to sustain.

    The nordic welfare states also have had higher employment levels than the USA for the past decade.

    At the same level of spending, you can have badly designed welfare states or well-designed welfare states.

  13. @Peter Whiteford
    as an interim response, the rawlsian concern is absolute levels of primary goods, not relativities.

    if the bottom decile in the USA is 20% richer in absolute terms than the bottom decile in europe the compariion needs correcion. there may be an empsy set for the USA

    the poor in the USA may be as weathy as the lower middle class in europe, the middle class in europe may be a rich as the upper middle class in the old EU and so on.

  14. “As a result of the crisis, the US employment/population ratio has dropped sharply, to the point where the US is now little different from the EU. ”

    Your whole argument (as in your comment #5) depends on this point. But (as you have pointed out in the past) employment in the US is much more volatile because of lower hiring & firing costs. Do you think this shift is permanent?

  15. @Declan Trott
    yes, good point.

    It took the worse U.S. post-war recession to get U.S. unemployment rates and U.S. employment/population ratios down to the levels that the EU experienced for every year of most of the last 20 years!

    does that mean that the EU has been in a deep recession or depression for 20 years plus? another victory for social democratics politics?!

  16. “System design” causes measurement and comparison difficulties.

    For countries that deliver health, education and other social benfits through non-market mechanisms (government or private non-profit institutions) the contribution to output (and ultimately GDP) in nominal terms is measured at cost, the major component of which is labour costs.

    Further, in assessing “real” contributions to output, the “price” of non-market services is an oxy-moron, so nominal output is deflated by the price of INPUTs.

    Under these circumstances there can be no measurable productivity gain…. more effective use of labour results in no change of $ value of output per $ value of input.

    So countries with a redistributive model for delivery of health and education (the Nordic model?) should by definition underperform an country that delivers these services via a market model.

    Adjusting for this effect may result in the EU overtaking the US.

    A similar adjustment for unpaid household work may make a very large difference if housework and child rearing via market mechanisms is substantial in one of the economies being compared.

    Finally, from my subjective view of biases in PPPs, they tend to favour rich nations over the poorer, and I doubt that there is substantial bias for US/EU comparisons. When in doubt use exchange rates for advanced nation comparisons. They are probably a better predictor of expectations than PPP outcomes.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s