Zhang Chengliang / China Daily

The Chinese economy is at a turning point, and we need to look at why a shortage of jobs has turned into a shortage of labor.

In the 30 years since reform and opening up of the late 1970s, workers' wages have risen very little, and they have obtained work based largely on labor market forces. This is particularly true of migrant workers and their wages. This is because by the end of 2004 China was at the development stage of a typical dual economy: unlimited labor supply, a big surplus in rural labor and overstaffing in businesses, which kept a lid on wages for the unskilled. A labor shortage particularly among migrant workers appeared in 2004 and has been with us ever since. It has helped fuel the push for continued wage rises, and, in line with the definition of a dual economy, 2004 can be regarded as the year of the Lewis turning point. According to the theory of the late economist William Lewis, that point is reached when surplus labor in the subsistence sector is fully absorbed into the modern sector, and where further accumulation of capital begins to raise wages.

Marking the Lewis turning point by reference to the shortage of migrant workers is a little fuzzy, and indeed some would argue that the Lewis turning point has yet to be reached. So we have to look at another clear turning point: the working population aged 15 to 64 years old has stopped growing, a turning point that means the overall dependency ratio - which compares the number of those not in the work force with those who are - is no longer falling. As the working-age population grew and the dependency ratio fell they produced the economic and social benefits known as the demographic dividend. The disappearance of the two conditions that produced the demographic dividend takes us to what is called the demographic dividend turning point, which means the dividend has now dissipated.

According to the latest estimates, the 15-64 working-age population growth rate in China has been falling year by year. It is expected to peak next year when the total working-age population is nearly 1 billion. From then the absolute number of those in the workforce will no longer rise. It is clear then that there is no longer an unlimited supply of labor.

It is because of the short interval between these two turning points that the recruitment difficulties in China have occurred periodically. Statistics show that Japan reached the Lewis turning point in 1960, and that it reached its demographic dividend turning point after 1990, which means the country had 30 years to adjust. Even though Japan's economy has stagnated since 1990, at the same time it has become a high-income country. South Korea, which reached the Lewis turning point about 1972, has yet to reach the demographic dividend turning point. It, with China, is expected to reach that point next year. So the gap between South Korea's two turning points has been 40 years.

Using the same criteria, if we agree that China reached the Lewis turning point in 2004, and the working-age population does not keep rising next year, the dependency ratio will stop falling then, and we will have arrived at the demographic dividend turning point.

It means that the interval for adjustment between the two turning points will have been nine years at most. Such a short interval means shortage of labor will develop into a serious shortage of migrant workers or at least workers. This transition will make the Chinese labor market quite different to that of other countries.

The Chinese labor market is evolving into that of a neoclassical economy from a dual economy one, and the labor market now bears the traits of both. It is important that we understand the nature of both models.

The first thing to look at is how wages are set. In the Lewis model of a dual economy, labor is in unlimited supply, and the marginal productivity of labor in agricultural production is extremely low. Wages in agriculture are much lower than those in other industries. Because of the dearth of labor, migrant workers' pay has not improved over many years. For institutional reasons, such as China's household registration, or hukou, system, migrant workers' pay is not determined by the marginal productivity of labor. But in the neo-classical model that is how workers' pay is set, through market mechanisms and labor market institutions. As the surplus of rural labor has shrunk and given way to a shortage of labor, some industries have tended to replace it with capital, and productivity has improved significantly. This has been accompanied by a sustained, substantial increase in wages, of agricultural and nonagricultural workers alike, since 2004.

The next thing to look at is the situation of labor market clearing. In the Lewis model, because of the unlimited supply of labor, and the systemic barriers to the transfer of labor, the balance between supply and demand is not cleared by adjusting wages. That means the dual economic structure and labor oversupply will last for a long time. But in the neoclassical model the labor market can clear the supply and demand imbalance through labor mobility or wage adjustments.

Until the late 1990s, before the reform of State-owned enterprises that aimed to cut the number of staff and increase efficiency, this labor supply and demand relationship was mainly reflected in the surplus of labor in the countryside and overstaffing in businesses in cities. However, this did not show up in unemployment figures, and economic fluctuations were not reflected in changes in those numbers. But in 1997, when reform finally broke "the iron rice bowl", workers were laid off, bringing large-scale unemployment. It is estimated that at one point in 2000 the unemployment rate was as high as 7.6 percent. The government became active in implementing employment policy, and unemployment gradually fell. In 2002 the level of registered unemployment remained stable.

The biggest problem of the dual economic structure is that there is a lot of labor, but not enough job vacancies. For a long time China's employment policy was based on a surplus of labor, and the aim was to create as many jobs as possible to increase employment. But in the neo-classical model there are three main types of unemployment: cyclical, structural and frictional, the later being transitional unemployment that occurs when someone is shifting from one job to another. All these need to be solved through macroeconomic policy, namely government labor market policies. The main task in a dual economy is to increase employment and to eliminate obstacles impeding labor; but in the neo-classical economy the goal of macroeconomic policy will be clearer.

Because of the particularities of Chinese labor market restructuring and the new characteristics of the labor market, traditional policies need to be changed to eliminate decision-makers' preconceived, erroneous ideas.

First, since the total amount of the labor market is limited, structural and frictional employment difficulties need to be taken seriously. Because of industrial readjustment, new jobs are constantly being created. Some traditional jobs will inevitably disappear. If workers are not up to the new jobs, they face structural unemployment. Because labor market development in China is still in its early stages, the way human resources are allocated is poor, and workers cannot be redirected to new jobs without frictional unemployment as industrial restructuring proceeds. That means frictional unemployment will be with us for a long time.

It will take time for the labor market to absorb those entering the growing work force, including new graduates with their higher levels of education and human capital. As for those less educated and less skilled, they are likely to spend more time in frictional unemployment. It is on these two groups that the government ought to concentrate its efforts to promote employment. It needs to provide employment, entrepreneurship and on-the-job training, to help those switching jobs. Natural unemployment can be reduced in two ways: workers becoming more skilled and greater efficiency being gained in where and how these skills are deployed. Second, when unlimited labor supply gradually disappears, we should focus more on recurrent cyclical unemployment. When the market allocates resources and guides economic activity, cyclical fluctuations of the macro economy are inevitable, as is corresponding cyclical unemployment. At present, because rural migrant workers lack urban household registration and lack the job stability that others enjoy, they are generally at greater risk of cyclical unemployment.

There are two ways to deal with this type of unemployment. First, use macroeconomic levers such as monetary and fiscal policies to even out the economic cycle. For this reason, employment problems need to be given greater priority in macro economic policies, reducing the risk of cyclical unemployment. Second, cyclical unemployment, a side effect of labor market segmentation, needs to be eliminated through reform. Give migrant workers the benefits of urban unemployment insurance and help them in their efforts to find work.

The author is director of Institute of Population and Labor Economics, Chinese Academy of Social Sciences. The views do not necessarily reflect those of China Daily.