Labor Productivity: Are Diverging Trends between Developed Countries Durable?

Date December 5, 2008
Speaker Michel FOUQUIN(Deputy Director, Centre D'Etudes Prospectives et D'Informations Internationales (CEPII) / Associate Professor, University of Paris I (Pantheon- Sorbonne))
Commentator / Moderator FUKAO Kyoji(Faculty Fellow, RIETI / Professor, Institute of Economic Research, Hitotsubashi University)


While events like the current financial crisis occur with regularity in the economic cycle, long-term trends also exist in productivity. The maintenance and preservation of these long-term trends is a primary economic policy objective. However, controversy often develops when productivity trends in the European Union and the United States diverge from each other. Similar divergences also occur from time to time between the productivity trends of the EU and Japan, as well as between those of Japan and the U.S, all of which brings up two important questions: How should breaks in these trends be defined and how can they be measured?

Two common themes in recently published European Commission economic papers exploring new labour policies on productivity have been increased R&D spending and balanced employment ratios between younger and older workers. To some extent, however, these goals may be contradictory.

From the 1950s up until the oil shock in the early 1970s, a convergence in the productivity of developed countries was clearly in effect. Figures show that the lower the initial productivity in a country, the higher the rate of productivity growth. The converse was also true, with higher initial levels of productivity corresponding to lower productivity gains per worker. However, this relationship held true only for OECD countries and certain East Asian countries including South Korea, Taiwan, Hong Kong and Singapore; it was not seen in the countries of the developing world. But the convergence process weakens after the first oil shock and almost disappears in the last period 1995-2003. Nevertheless the gap between the highest productive country and the lowest one has been reduced from a factor five in 1950 to a factor two in 1995. These results might indicate that the difficulty increases with the passage of time for trailing economies attempting to catch up to the leading economy.

Econometric tests can be used to determine whether or not breaks exist in productivity trends. Several studies have tried to develop a special methodology for that purpose such as Maury and Pluyaud (2007) who found only three breaks in the U.S. economy productivity trends during the 20th century: 1922 or 1933, 1967 or 1973, and 1995.

Our test show that for the developed countries, 11 out of the 19 OECD countries registered their first break at the time of the 1974 oil shock when productivity dropped by 60%. The next break occurred in the 1990s when increases were registered by the U.S., Sweden and Greece, while the rest of the EU and Japan experienced declines. It is clear that for the first oil shock there is a macroeconomic explanation for these breaks in productivity trends. On the contrary the second breaks around 1995 are of a different nature some countries (the USA in particular) improve their productivity record while others lose ground. The new technology argument, which can explain part of the US performance, cannot explain the EU Japan decline. As all these countries have raised their investments in new technology they get a positive impact from that effort even if it was less than in the USA.

Will the gap between the US and EU remain? In a up to date analysis using quarterly data for the 1995-2008 period we found that productivity trends were still lower for all countries except Germany after year 2001 but a gap of 0.6% per year between the EU and the U.S., in favour of the U.S. remains.

How do we explain the gap? What is striking is that the U.S. recovery is less and less intensive in job creations. Job creation after the recession of the 1970s was higher than that after the recession of the 1990s, and both were higher than the rate of job creation after the 2002 recession. The low rate of job creation has had a major impact on the U.S.'s higher labour productivity growth rates.

Using EU KLEMS data, a major change can be seen in the period before 1995 as opposed to the period after 1995. Labour contribution was much higher in the U.S. than in the EU before 1995, which pointed to intense job creation in the U.S. However, this trend reversed itself after 1995 when the EU showed signs of higher job creation rates.

From a technological standpoint, the U.S. clearly holds an advantage with its higher levels of IT utilization. However, the labour productivity growth gap with Europe is currently much smaller than the job creation gap. U.S. growth was 0.9% higher than European growth, but labour contribution was -0.8%.

Through an econometric analysis of the impact of the growing gap in unemployment, it can be ascertained that with greater employment rates comes slower growth in productivity. Highly skilled workers are at nearly full employment in most countries, but young, old and non-skilled workers are facing lower employment rates. If the employment rate of the economy on a whole increases, most new jobs will go to workers with generally lower productivity than their highly skilled counterparts. This causes overall productivity to drop in the short to medium term.

In regard to changes in working hours, OECD data was used to analyze the differences between OECD countries. From 1995-2006, rates of employment growth in France were higher than the 1985-1995 period, with the government's drive to reduce unemployment cited as the reason behind the increase. On the contrary, working hours in France have continued to decline. The U.S. saw a decline in rates of employment that caused an increase in overall productivity.

In an analysis of the main sectors of the economy, the U.S. saw increases in productivity in both the non-electric/electronic manufacturing sector and the finance and business services sector. There was also a decline in the number of hours worked in each of these sectors. France experienced the opposite in both the non-electric/electronic manufacturing sector and other manufacturing and distribution sector, which saw decreases in productivity and increases in hours worked.

Decreases in both hours worked and productivity in Japan have caused a steep decline in value-added growth. Electric and electronic firms as well as providers of distribution services are now experiencing large reductions in the amount of total hours worked. Also, government-funded personal and social services programs reflect efforts that are being made to soften the effects of low economic growth.

Manufacturing industry breaks were analyzed by comparing average annual growth rates from 1975-2005. Finland, Sweden, Greece and the U.S. enjoyed strong improvements in productivity with gains of around 6%, while six European countries, Japan and Australia saw breaks that resulted in decreases in productivity, and South Korea and six other European countries did not experience breaks in their rates of productivity growth.

Analysis indicates that the emergence of information technology can be discounted as a potential factor responsible for the breaks in labour productivity growth. Clearly, the U.S., Japan and South Korea have invested heavily in new technologies that have improved productivity in those countries, but comparable though slightly smaller investments in Europe have been followed by decreases in productivity on that continent. Technology on the whole has been a net benefit for Japan, the U.S. and the EU, but this cannot explain the diverging trends in the rates of labour productivity growth.

As for breaks in the business and financial sector, gains are strong, though lower than those seen in the manufacturing sector. Five countries including the U.S. and Ireland made significant progress after having negative rates in the past. Many Continental European countries and Japan have experienced negative trends. Tests were simultaneously run for labour productivity and hourly productivity, and they yielded roughly similar results.

In conclusion, diverging trends clearly exist between the U.S. and northern European countries such as Sweden and Finland on one side, and central European countries on the other. Labour input appears to play a larger role in the period after 1995, with technology playing an important but secondary role. The results indicate that other factors, such as intangible investments and how they are included in the national account framework should be taken into consideration.

To summarize the findings of Professor Fouquin's paper, many European countries and Japan experienced slower labor productivity growth in the 1990s, while the U.S. saw an increase. Europe's slowdown was mainly caused by slowing multi-factor productivity growth and not by decreasing capital accumulation in areas such as IT and communication technology investments. The slowdown in Europe can also be explained by the creation of jobs, which led to decreasing overall productivity. It is important to analyze these findings and ascertain whether the same logic can be applied to Japan.

From 1980-1995, the contribution of labor input growth in Europe generally did not decline by a substantial amount. While total factor productivity (TFP) growth declined in certain European countries, as already stated, labor contribution increased. In Japan, TFP growth experienced a similar slowdown, but so did capital input and labor contribution. From the viewpoint of capital accumulation and labor input, Japan and Europe had different experiences.

As for the contribution of capital input growth from 1995-2004 in the U.S., Japan and major EU economies, France and the U.S. encountered large increases in IT and communications technology capital, while Japan experienced a decline in this area. About 0.5% of the gap in capital input growth between Japan and the U.S. can be explained by Japan's lack of IT capital. In the same period, Japan experienced a very large decline in hours worked due primarily to demographic changes.

The slowing labor productivity growth in Japan was caused not only by lower multi-factor productivity growth, but also by a slowdown of capital accumulation. The creation of new jobs cannot explain the decline in multi-factor productivity growth in Japan, though intangible investments and low investment in IT and communications technology are probably to blame. Furthermore, Japan's ratio of intangible investments to output is much lower than that of the U.S. and EU, even though Japan's tangible investments remain high.

Japan's situation regarding labor productivity growth is inherently different from that of the EU. Most famously, the EU did not confront an economic crisis anywhere near the proportion of Japan's "lost decade" experience. Large public investments in real estate are seen as inefficient by economists, and because of this reason, the results of this study did not focus on comparisons involving Japan due to Japan's unique circumstances and the lost decade.

Michel FOUQUIN and Kyoji FUKAO

Questions and Answers

Q: Are there any specific figures available comparing the absolute productivity levels of Japan, the EU and the U.S.? Also, how accurate are reports that Japan's overall labor productivity stands at around 70% of the U.S. figure? How has the state of real labor productivity among the nations in the developed world changed in the last 20-30 years?

While the exact numbers are not on hand, detailed sectoral information on the levels of productivity up to 2005 can be found for free on the EU KLEMS database. Japan's productivity standing at 70% of the U.S. figure does not sound too far from reality.

Recently EU KLEMS published a paper supporting the claim that Japan's TFP level is about 70% of the U.S.'s. However, international comparisons of TFP are very difficult. For example, transportation productivity is measured by kilometer-ton per worker, but variables such as quality and timeliness of service, which are important factors in the business in Japan, are not fully taken into account. With that being said, however, there is much Japan can do to increase its productivity in comparison to other developed countries.

Q: It seems significant that the country with the world's largest trade deficit, the U.S., has the highest productivity. Can you please explain this phenomenon in laymen's terms?

When thinking about globalization in the context of Britain in the 19th century, it is surprising to find that the trade deficit of the UK was very high during the period of primacy of the British Empire. Great Britain was experiencing trade deficits of 6%, a seemingly impossible number. This was sustainable because the UK invested in Australia, Canada, New Zealand and other overseas destinations. The revenue from international investment was balancing the trade deficit. Nowadays, the so-called "imperial United States" similarly has a very large trade deficit that is compensated by the revenue of international investment.

Q: What was British labor productivity at that time?

Labor productivity at that time was at peak levels. Up to the end of the 19th century, the U.S. and Germany were catching up to Great Britain, but before the 1870s the British were well ahead of all other countries. After that, Germany and the U.S. began to catch up very fast. At the beginning of World War I, it was believed that Germany and the U.S. had overtaken British productivity in industry.

Q: Up until the first oil shock, the trend in labor productivity was convergence. At this time, it could be seen as "catching up" to the U.S. model. At times of technological revolution there may be opportunities for divergence. Furthermore, U.S. productivity in manufacturing is increasing. This reflects what you have said, that the U.S. is realizing globalization. In particular, the division of labor between the U.S. and China must be taken into account. What do you think about this?

The fact that the U.S. faced a strong dollar and intense competition from China had a negative effect on the trade balance. It also made it much harder for the competition within the U.S. to eliminate weak industries and promote stronger ones. The improvement in productivity was the result of increased competition from outside as well as inside.

Q: Regarding the increase of U.S. productivity and the decrease in its labor force versus Europe, the changes in Europe are more unusual than those in the U.S. Why did this occur?

The intensity of job creation in the U.S. was lower in the last business cycle. It is difficult to clearly explain why this is, but on the European side there were clearly policy measures to reform the labor market in order to improve or reduce mass unemployment. Europe and France in particular have experienced massive unemployment. Since the first oil shock unemployment has been growing, and it became very high after the 1980s. It is clear that while some prefer leisure to work, Europeans in general prefer shorter hours at work, and all want to have a job.

The reforms were aimed at lowering the cost of low-skilled workers. The indirect cost of labor in France was around 50% higher than direct income. If this cost is reduced by 10% or 20%, it is easier for small and medium-sized enterprises to create low-skilled jobs. These jobs, however, are low on productivity, and this condition led to the rise in employment but drop in productivity.

Q: The service sector seems to be the focus of many government policies in terms of labor productivity. What methods can you suggest for the service sector to develop better efficiency? What are the important elements in the service sector that must be targeted? While branding is important to attract customers in the service sector, particularly the tourism industry, how do you think productivity can be increased in this particular sector?

As for France, very successful big retailers use advanced technologies in their businesses and thus achieve a very high level of productivity. When compared to the U.S., generally speaking, the French and Germans have lower productivity in retail. This is probably because small shops in France are rather late in adapting to new technology. Also, the jobs required of retailers in the U.S. and in France are different. There are many specialized shops in the U.S. that are less-diversified than stores in France. Nevertheless, when looking at the level of employment in retail and retail distribution, the level of employment was higher in the U.S. compared to France. One of the reasons is that it is an activity where people with low qualifications are numerous, and the price for such workers in France is very high. The level of employment in retail in France is relatively low. It will probably increase in the future, and it will again be detrimental to productivity.

Q: What is your opinion on the financial crisis?

After seeing the impact of the financial crisis in Asia, first in Europe as in Asia there is a feeling that European countries, with the exception of the UK, should have escaped the financial turmoil. The reason is that the exposure of French and German banks to the subprime crisis was not very significant. Nevertheless, contagion came to Europe and even to Asia. Europeans believed that Asia was not exposed to the financial crisis.

The stock market was the first place where the effects were felt. A problem in New York will immediately lead to problems in Europe. While the reason for this is not clearly understood, it can be posited that bad news for the U.S. generally means bad news for Europe. The declines of the stock markets in Europe are now as bad as those in the US.

Banking problems also developed not only in Europe, but also in Japan. Japanese banks had financial portfolios that were quickly losing value as assets devalued. If banks saw that asset values were declining, they started selling more assets in order to recapitalize and improve their prudential ratios.

In Asia, while the stock market problems were similar, there was a more important real impact through trade. In South Korea, exports have dropped more than 20% from last year's figures, and that may continue. China's net export contribution to growth was around 2% for the last five years, but will now drop to 0%. Japan's exports are also declining very rapidly, primarily due to China.

Also, there has been a withdrawal in financial flows as U.S. companies try to get their money back into the U.S. This can be seen as a movement toward "de-globalization," which adds to the expansion of the crisis for the whole world. This is the first truly global crisis. It was expected that Asian economies would not be included, but they are.

Q: Is there a possibility that this financial crisis will bring a new break year into your analysis?

I do not know if that may happen. At the moment, developed countries are making efforts to increase the flexibility of their labor markets. More flexibility may not be such a good thing. Unemployment is rising dramatically, especially in the U.S. and UK. This flexibility may lead to increased social issues and public deficit problems. It may be time to focus less on labor flexibility and more on security, stability, and the development of a system capable of ensuring stability on the financial side.

Q: Why, particularly in European countries and especially in France, has such a substantial amount of public money been injected into the banking system? The banking system in France has not yet suffered very much. Nevertheless, the French and German governments have injected substantial amounts of money into their banking systems.

It should be mentioned that the first government to inject money into the banking system was the British government. The city of London is an essential part of the British economy. That country, at the forefront of globalization and the free movement of capital, was the first to inject public money into the system. It is paradoxical as they are the free-traders and the liberals of the world economy. Compared to that, the French government has done less than the British.

Gordon Brown was the mastermind of the European rescue program for the banking system. Between the French and the Germans there are always conflicts, but both the French and Germans were able to agree with Gordon Brown's program. The French and Germans basically followed the British example.

The stimulus package story is one that is not yet finished. There have been discussions about how to implement the most efficient stimulus package. What this package will be is not at all yet clear. I believe that the fact that Gordon Brown chose to reduce the value-added tax is probably not a good idea. We are now in a deflationary situation in terms of prices. Prices are declining and a value-added tax decrease will increase the downward pressure on prices. It is inefficient, it will cost a lot of money for the public budget, and it will be politically very difficult to return to the former rate. These are three reasons not to lower the value-added tax.

President Sarkozy announced his program, and it is a classic stimulus program under which high-speed trains will be built. Regarding small and medium-sized enterprises, their problems are being felt around the world. Japan has placed much emphasis on them, while in South Korea large companies are doing well at the expense of the smaller enterprises. It is from such enterprises that unemployment is stemming. Every country needs to undertake measures in order to provide some fiscal stimulus for small and medium-sized enterprises. Also, some programs for the working poor need to be enacted.

*This summary was compiled by RIETI Editorial staff.