Mirror, mirror on the wall
Jun 17th 2004
From The Economist print edition
America is widely admired as the beauty queen of the economic world. But the euro area's figures are more shapely than its reputation suggests
AS AMERICA'S economy has bounced back, the economies of the euro area still seem to be crawling along. This perception has reinforced pervasive gloom about continental Europe's economic future. A great deal has been written about America's superior performance relative to the euro area. But wait a minute: the widely held belief that the euro area economies have persistently lagged America's is simply not supported by the facts.
America's GDP surged by 5% in the year to the first quarter, while the euro area grew by only 1.3%. Europe's GDP growth has consistently fallen behind America's over the past decade: in the ten years to 2003 America's annual growth averaged 3.3%, compared with 2.1% in the euro area. Yet GDP figures exaggerate America's relative performance, because its population is growing much faster. GDP per person (the single best measure of economic performance) grew at an average annual rate of 2.1% in America, against 1.8% in the euro area—a far more modest gap.
Furthermore, all of that underperformance can be explained by a single country, Germany, whose economy has struggled since German reunification in 1990. Strip out Germany, and the euro area's annual growth in GDP per person rises to 2.1%, exactly the same as America's. Germany does represent around one-third of euro-area GDP, but still the fact is that economic statistics for the 11 countries that make up the other two-thirds look surprisingly like America's (see chart 1). (Were Britain part of the euro area, this effect would be even more striking.)
The most popular myth is that America's labour-productivity growth has outstripped that in the euro area by a wide margin. America's productivity has indeed quickened in recent years, but the difference between productivity growth in America and the euro area is exaggerated by misleading, incomparable figures. In America the most commonly used measure of productivity is output per hour in the non-farm business sector. This grew by an annual average of 2.6% over the ten years to 2003. For the euro area, the European Central Bank publishes figures for GDP per worker for the whole economy. This shows a growth rate for the period of only 1.5%. But unlike the American numbers, this figure includes the public sector, where productivity growth is always slower, and it does not adjust for the decline in average hours worked.
Using instead GDP per hour worked across the whole economy, American productivity has risen by an annual average of 2.0% since 1994, a bit faster than the euro area's 1.7% growth rate. However, a study* by Kevin Daly, an economist at Goldman Sachs, finds that, after adjusting for differences in their economic cycles, trend productivity growth in the euro area has been slightly faster than that in America over the past ten years. Since 1996 productivity growth in the euro area has been slower than America's. But it seems fairer to take a full ten years.
But has not America combined rapid productivity growth with strong jobs growth, whereas continental Europe's productivity growth has been at the expense of jobs? This may have been true once, but no longer is. Over the past decade, total employment has expanded by 1.3% a year in America against 1% in the euro area. Again, excluding Germany, jobs in the rest of the euro area grew at exactly the same pace as in America. And since 1997 more jobs have been created in the euro area as a whole: total employment has risen by 8%, compared with 6% in America.
It is true that, during the past decade, productivity growth has accelerated in America, but slowed in the euro area. Alan Greenspan, chairman of America's Federal Reserve, blames Europe's rigid labour and product markets. Structural barriers to laying off workers or to new methods of work may have prevented firms from making the best use of IT equipment.
However, there is another, less worrying reason why productivity growth has slowed in continental Europe. Reforms to make labour markets more flexible have deliberately made GDP growth more job-intensive. Firms now have more incentive to hire new workers, thanks to lower labour taxes for low-paid workers and a loosening of rules on hiring part-time and temporary workers, which allow firms to get around strict job-protection laws. The flipside is slower productivity growth for a period, as more unskilled and inexperienced workers enter the workforce. This is exactly what happened in America in the 1980s. In the longer term, more flexible labour markets should help to boost growth.
Another popular misconception is that the return on capital is much lower in the euro area than in America, because European business is inefficient and hobbled by high wage costs and red tape. This argument is often given in defence of America's large current-account deficit. America's higher return on capital, it is argued, attracts a net inflow of foreign money, so it has to run a current-account deficit. But according to calculations by Goldman Sachs, the return on capital in the euro area has actually been roughly the same as in America in recent years. The total return on equities over the past decade has also been broadly the same—which is what you would expect given their similar pace of productivity growth.
Nonsense in, nonsense out
So far we have established that, based on official statistics, productivity growth over the past decade has been virtually the same in the euro area as in America, and although GDP per person has grown a bit slower, the gap is modest. However, using official statistics can be like comparing apples with pears, because of differences in the way that GDP is measured in different countries. For example, American statisticians count firms' spending on computer software as investment, so it contributes to GDP. In Europe it is generally counted as a current expense and so is excluded from final output. As a result, the surge in software spending has inflated America's growth relative to Europe's.
A second important difference is the price deflator used to convert growth in nominal spending on information technology equipment into real terms. In America, if a computer costs the same as two years ago, but is twice as powerful, then this is counted as a 50% fall in price. Though logical, this is nevertheless a contentious issue among economists. Most euro area countries do not allow fully for improvements in computer “quality”, so again official figures probably understate Europe's growth (in both GDP and productivity) relative to America. This reinforces the argument that the euro area has not been doing that badly.
Despite such statistical quibbles, however, it is undeniable that the average person in the euro area is still about 30% poorer (in terms of GDP per person measured at purchasing-power parity) than the average American, and this gap has barely changed over the past 30 years. Thus even if income per person is growing at almost the same pace as in America, Europeans are still stuck with much lower living standards than Americans.
Olivier Blanchard, an economist at the Massachussetts Institute of Technology, offers a more optimistic view†. The main reason why the income gap has not narrowed, he argues, is that over time Europeans have used some of the increase in their productivity to expand their leisure rather than their incomes. Americans, by contrast, continue to toil long hours for more income. Who is really better off?
In fact, Europe's GDP per person is no longer lower than America's because its economies are much less productive. Average GDP per hour worked in the euro area is now only 5% below that in America; 30 years ago it was about 30% lower. GDP per hour in Germany and France now exceeds that in America. Income per person is higher in America largely because the average person there works more hours. In the euro area, fewer people work and those who do hold a job work shorter average hours. By one estimate the average American worker clocks up 40% more hours during his life time than the average person in Germany, France or Italy.
The narrowing of the productivity gap between America and the euro area over the past 30 years has not been reflected in a catch-up in the euro area's GDP per person because hours worked have fallen sharply. Compare France with America. Between 1970 and 2000 America's GDP per hour worked rose by 38% and average hours worked per person rose by 26%, so GDP per person increased by 64%. French GDP per hour rose by a more impressive 83%, but hours worked per person fell 23%, so GDP per person only increased by 60%. Chart 2 shows for the whole of the euro zone how its improvement in productivity relative to America has also been fully offset by a fall in hours worked.
If leisure is a normal good, then it is surely appropriate that demand for it increases in line with income. A broader analysis of living standards based on economic welfare rather than GDP should place some value on longer leisure time. The tricky question is whether the decrease in hours worked is due to employees' preference to take more leisure rather than more income, or due to distortions from maximum working hours, forced early retirement or high taxes.
Mr Blanchard's analysis finds that most of the fall in hours worked in Europe has been due to a decline in average hours per worker (thanks to longer holidays or shorter working weeks), rather than a rise in unemployment or a fall in the proportion of the population seeking work. Furthermore, most of the reduction in average hours worked was due to full-time workers putting in shorter hours, not because of an increase in part-time workers who might not have been able to get full-time jobs. Mr Blanchard concludes that the fall in hours worked is mostly voluntary.
But that does not settle the matter. Perhaps Europeans choose to work fewer hours because of high taxes. Marginal tax rates have indeed risen by more in Europe than in America over the past 30 years. Taxes reduce the incentive to work an extra hour rather than go home, once a reasonable standard of living has been reached.
This is a hotly debated issue. A study** by Edward Prescott, an economist at the Federal Reserve Bank of Minneapolis, claims that virtually all of the fall in hours worked in the euro area can be blamed on higher taxes. But the flaw in this theory, says Mr Blanchard, is that within Europe there is little correlation between the fall in hours worked and the increase in taxes. Ireland has seen a 25% fall in average hours worked since 1970, despite an even smaller increase in tax rates than in America. Other studies have found that taxes have played a more modest role, accounting for about one-third of the fall in hours worked per person.
Mr Blanchard concludes that most, but not all, of the fall in hours worked over the past 30 years is due to a preference for more leisure as incomes have increased. Europeans simply enjoy leisure more. Americans seem more obsessed with keeping up with the Jones's in terms of their consumption of material goods. As a result, they may work too hard and consume too little leisure. Their GDP figures look good, but perhaps at a cost to their overall economic welfare.
Robert Gordon‡, an economist at Northwestern University, agrees that GDP comparisons overstate America's living standards, but he goes even further. America has to spend more than Europe, he says, on both heating and air conditioning because of its more extreme climate. This boosts GDP, but does not enhance welfare. America's higher crime rate means that more of its GDP is spent on home and business security. The cost of keeping 2m people in prison, a far bigger percentage of its population than in Europe, boosts America's GDP, but not its welfare. The convenience of Europe's public transport also does not show up in GDP figures. Taking account of all these factors and adding in the value of extra leisure time, Mr Gordon reckons that Europe's living standards are now less than 10% behind America's.
Flexing the macro-muscles
But even if the euro area has not lagged far behind America, does not its pathetic growth over the past couple of years bode ill for the future? Surely America's stronger rebound since the global economic downturn in 2001 is proof of greater flexibility in its economy? In fact, both suggestions are questionable. The main explanation for America's more rapid recovery is that it has enjoyed the biggest monetary and fiscal stimulus in its history. Since 2000 America's structural budget deficit (after adjusting for the impact of the economic cycle) has increased by almost six percentage-points of GDP. Meanwhile, the euro area has had no net stimulus (see chart 3).
American interest rates were also cut by much more than those in the euro area. Without this boost, America's growth would have been much slower over the past three years. In other words, America's much faster growth of late may mainly be the result of looser (and unsustainable) fiscal and monetary policies, rather than greater flexibility.
While this might have been the right policy to support America's economy, it means that America's recent growth rate says little about its likely performance over the coming years. Indeed, the super-lax policies of the past few years have left behind large economic and financial imbalances that cast doubt on the sustainability of America's growth. From a position of surplus before 2000, the structural budget deficit (including state and local governments) now stands at almost 5% of GDP, three times as big as that in the euro area. America has a current-account deficit of 5% of GDP, while the euro area has a small surplus. American households now save less than 2% of their disposable income; the saving rate in the euro area stands at a comfortable 12%. Total household debt in America amounts to 84% of GDP, compared with only 50% in the euro zone.
America's recent rapid growth has been driven partly by a home-mortgage bubble. As interest rates fell and house prices rose, people took out bigger mortgages and spent the cash on a car or a new kitchen. House prices have also been lively in some euro-zone countries, with house prices rising at double-digit rates over the past year in France, Italy, Spain and Ireland. But in general, households have not borrowed to the hilt against those capital gains. Some European policymakers hope that America's bubble will soon burst and that Europe could then sprint ahead. That may be wishful thinking: a sharp slowdown in American consumer spending is also likely to dent Europe's growth rate. It is true, however, that the eurozone's consumer finances are in much better shape.
So, America's superior economic performance over the past decade is much exaggerated. Productivity has grown just as fast in the euro area; GDP per person has grown a bit slower, but mainly because Europeans have chosen to take more leisure rather than more income; European employment in recent years has grown even faster than in America; and America has created some serious imbalances which could yet trip the economy up badly.
“Bullish on America”
Indeed, one might say that the economic performance of the euro zone and America has not been hugely different over the past decade, but that American optimism has disguised this. European policymakers are forever fretting aloud about structural rigidities, slow growth, excessive budget deficits and the looming pensions problem. In contrast, American policymakers love to boast about America's economic success while playing down the importance of its economic imbalances.
This does not mean that the euro area can be complacent. It still needs to push ahead with structural reforms. Its average jobless rate of 9%, against 5.6% in America, is too high. Contrary to the beliefs of many Americans, there has been labour- and product-market reform in continental Europe over the past decade, which is why employment has perked up. But unemployment remains a problem and, sadly, economic reforms now seem to have stalled in France and Germany.
The biggest snag, of course, is that because of its less favourable demographics, Europe has an older economy than America. With lower birth and immigration rates and an ageing population, Europe's labour force will soon start to shrink as a share of the population. That will make it harder for Europe to maintain its current pace of growth in GDP per person—and thus harder for governments to pay pension bills. Without faster growth, Europe will be unable to afford its welfare system.
If Europeans do not want to slip down the rankings of GDP per person in future, then they will need to work longer hours during their lifetimes. Alternatively, they may continue to attach more value to leisure and the quality of life, rather than hard cash. That is their choice. A truer picture of their economy might help them make it in an informed way.
* “Euroland's secret success story”. Goldman Sachs Global Economics Paper No. 102, January 2004.
† “The economic future of Europe”, NBER Working Paper No. 10310, March 2004.
** “Why do Americans work so much more than Europeans?” Federal Reserve Bank of Minneapolis Research Staff Report 321, November 2003.
‡ “Two centuries of economic growth: Europe chasing the American frontier”, October 2002