By Caijing Chief Economist Shen Minggao

Eight economists were invited by Premier Wen Jiabao and the State Council to a closed-door meeting between July 8 and 11 on China’s macroeconomic conditions and employment outlook. Cai Fang, director for the Institute of Population and Labor Economics, Chinese Academy of Social Sciences, revealed to Caijing his message to China’s decision makers.

Rising Unemployment

CJ: Please tell us what you and the other economists said at the meeting.

Cai: The discussion focused on current economic conditions. We talked about whether economic growth will slow, how to contain inflation and stimulate growth, and whether China should maintain its tight monetary policy.

Economists are split on the two major macro tasks: fighting inflation and stimulating growth. However, we generally agreed that monetary policy should remain as it is. We should neither loosen it nor tighten it further.

One thing that concerns officials at the State Council is how much the slowdown of GDP growth will hurt employment. Should we retain growth momentum to ensure high employment rates? I mainly talked about this issue.

CJ: How do you evaluate current employment conditions?

Cai: I’m quite worried about the economic growth. Last year, GDP grew by 11.9 percent, and in the first of half of 2008, it was 10.4 percent. Now the growth rate is flattening out, while unemployment is climbing up.

The registered urban unemployment rate in the first quarter of 2008 was 4 percent, while for the first six months it was 4.05 percent. That means that in the second quarter, it was somewhere around 4.1 percent. Clearly the unemployment rate is rising. Before this year, China’s unemployment rate was declining.

Some have questioned the legitimacy of using the registered urban unemployment rate as a valid indicator for the labor market. I think it works. The International Labor Organization has established the survey unemployment rate as the major indicator instead of the registered unemployment rate. The U.S. uses both. When economic conditions are stable, the registered unemployment rate can reflect the fluctuation of actual unemployment, at least for urban formal sector.

A few years ago, China’s registered unemployment rate didn’t reflect the real situation because it excluded laid-off workers (the real unemployment rate should be calculated on the basis of the registered unemployment rate plus laid-off workers). But the number of laid-off workers has largely been reduced in recent years due to the government’s reemployment efforts and even disappeared in some places. Now, the registered unemployment rate is close to what it is in reality.

CJ: So you think slower economic growth has affected the unemployment rate. But will high growth ensure high employment rates?

Cai: Not necessarily. People often apply Okun’s Law – an inverse relationship between economic growth and unemployment – to the study of the labor market. But economic growth and employment are not closely related in China.

One reason is that Chinese policy favors large size companies. The preference is even more obvious when the government adjusts its macro-economic policies recently. The selective policy tends to support some companies while leave others on their own. Those receiving support are usually enterprises with high profits, low emissions, low rates of pollution and less reliance on resources. In reality, they are big companies, especially state-controlled ones, equipped with better technologies, but usually not the one that necessarily fit the above criteria.

We did a survey of 17 industries and found that capital-intensive companies, most of which are large enterprises, contribute a lot to GDP growth but are not so impressive in terms of creating new jobs.

A tight monetary policy was put into effect last year. The control of credit was implemented in two ways: quantity and quality. In terms of quality, better risk controls and higher earnings were required for lending, and that situation diverted loans toward larger companies and away from small ones – because lending to them became even riskier.

Another reason is the structure of China’s unemployment rate. The biggest part of unemployment is the non-accelerating inflation rate of unemployment (NAIRU), which includes structural and frictional unemployment. Neither is affected by China’s economic cycle. That’s why this part of economic growth is not closely linked to jobs. My personal estimate is that China’s real unemployment rate (the survey unemployment rate) is about 5 percent, of which 4 percentage points are contributed by NAIRU and 1 percentage point from cyclical unemployment.

Cyclical unemployment is caused by economic slowdowns. Policies to stimulate growth can only reduce cyclical unemployment. So to reduce unemployment across China, we have to focus on NAIRU. The overall unemployment rate has been easing in recent years, but NAIRU is relatively stable. Coupled with the low employment rate of college graduates, the rate even shows signs of rising.

The government can make a difference in natural unemployment. In the U.S., the Clinton administration once boasted the success of low inflation and a high employment rate because it drove down NAIRU. China launched an employment promotion act this year. If that can reduce job discrimination, promote employment-related services and increase training and information provision to job candidates, it might be able to reduce NAIRU. If smaller export growth increased cyclical unemployment rate by 1 percentage point, then we can offset it by reducing NAIRU by 1 percentage point. That will ease the pressure of slower economic growth on employment.

CJ: When the economy slows, where will we see the most serious impact on cyclical employment?

Cai: Nobody wants to see a steep downturn of the growth. It would be great if the yearly growth rate reaches 10 percent in 2008. But if 1.9 percentage points of the growth rate from last year are wiped out, labor-intensive industries are hardest-hit.

Looking back, massive layoffs from state-owned factories started in the 1990s. Small and medium-sized enterprises (SMEs), private companies and even some informal employment opportunities played an important role in absorbing labor. The current economic slowdown, however, hit them hard. Actually, the registered unemployment rate possibly underestimates real situation of unemployment, because statistics tend to miss unemployment in these sectors. For instance, migrant workers won’t register.

In terms of urban income, part of that also falls out of our data. When the economy rises fast, the undocumented income jumps fast. But it drops equally rapidly when the economy is going downward. The result is, it looks like urban residents’ income hasn’t changed much, but the unreported group can be much worse than that.

Caijing: You said most economists agree on continuing tight monetary policies, but an economic slowdown will definitely affect the situation of SMEs. How can macro-policy help them?

Cai: Tight monetary policy is not good news for SMEs. The influence is mostly indirect. Historically, it’s hard for these companies to borrow from banks, and they turn to the market for financing. Now that banks have window guidance from the central bank as a measure to tight lending, it becomes even harder for SMEs to borrow from formal channels. That leads to skyrocketing interest rates for private borrowers, which is a heavy blow to smaller companies.

To help them, I think reducing taxes is a good method. Tax revenues rose rapidly in the first half of 2008, and it’s widely agreed among scholars that we can cut tax rates a bit. But there are needs for more government spending – natural disasters hit China one after another and we just hosted the Olympics. The government has to spend heavily, and it might want to set aside some money for a rainy day. Another difficulty is that no one speaks for SMEs. When it comes to agriculture, education and employment, there are relevant government organs to guard specific interests, but nobody is speaking on behalf of SMEs and advocating low taxes.

The New Labor Law, Good or Bad

CJ: Quite a few companies, especially SMEs, complain that the new labor law increases costs. Employees don’t seem to like the labor law very much either. Why?

Cai: Debate over the new labor law has never ceased, even after it became effective. Some think the law is a burden to enterprises while not benefiting workers.

One problem is a flawed social security net for migrant workers. There is no nationwide social security system; some provinces even don’t have a province-wide social security net. That leads to many migrant workers withdrawing from the social security system. Why? For instance, in the pension system, workers pay 8 percent of their salary and companies match it. In some seasonal, labor-intensive industries, workers finish up their terms and leave the job for good. But their social security benefits can’t go with them. So they have to withdraw from the system, taking back their own contributions, but the part from the company stays with the system. So there is no upside for workers to join the pension system, and for companies it creates a financial burden. Officials from inland provinces complain that coastal provinces have seen a fast increase of social security funds because they not only siphoned laborers from inland but social security funds as well.

In fact, by withdrawing from the social security system, migrant workers send a message that the system is flawed and they are not happy with it. The government should react to that positively and try to create a mechanism that can meet the needs of migrant workers and is capable of merging with other social security systems in the future.

I think several principles should be upheld. First, migrant workers must be included in the basic pension system. The labor law stipulates that, and the employment promotion act prohibits any discrimination. Second, basic insurance for migrant workers will be gradually shifted to a nationwide, portable mechanism. Third, when workers move, they should take their own contributions to the social security funds. But when retire, they should be able to receive both their own contributions and those from their employers, so that they can clearly anticipate what they will have.

China’s development has reached a stage where labor shortages are occurring and the labor supply-demand equation is changing. That requires a rise in salary and other benefits. The ideal result should be that by raising salaries 10 percent, companies are able to hire suitable workers. So the intention of the legislation is good: to provide a social safety net tailored for migrant workers. We should come up with an equally good design to meet the goal.

CJ: Do you think the labor law was enacted in a rush? Should we continue using it?

Cai: I think we should stick to the law. There are problems with enforcing this law, which were not created by the law itself but by a lack of support measures. Companies feel overburdened, partly because of the inadequate social security system. This is not the fault of the labor law.

I think it is right for the labor law to raise the cost of labor. If a company can’t bear a modest rise, it is not competitive except as a sweatshop. We should let such companies die if they have to. So I hope we don’t see the enforcement of the labor law scaled back when we talk about amending the law and adding support measures. Of course, we have to acknowledge that this year is unusual, with a number of adverse factors (such as natural disasters and slowing external demand). The pressure from rising labor costs is salient.

In fact, at the center of labor economics discussions is the question of whether such legislation is necessary. Now the popular view is that workers benefit from the minimum wage system in developed countries, and companies are not heavily affected. But in developing countries, sometimes when laws are made to protect workers, the result is higher unemployment. The unlimited supply of labor in developing countries is to blame.

India is one good example. India has a very strict labor law structure and each state fills in the details. That creates a variety of labor laws, some loose, some strict. Research shows that economic development levels in different parts of India are directly related to their labor policies, and those who have tight labor regulations often lag in economic development.

In India and some Latin American countries, overly strict labor laws resulted in a zero increase in formal employment. If the overall employment rates are rising in such situations, the real engine is informal hiring. And informal employment can be as large as 60 percent of total employment or even more. That high percentage will hurt formal employment. Yet informal employment can’t create truly competitive industries.

The reason I think the new labor law should be implemented is the change in the labor supply from a surplus to a relative balance. There must be some kind of incentive to spur labor supply and attract workers. Higher pay is a useful measure.

CJ: Many companies, especially textile manufacturers, are complaining about how hard it is to find employees, let alone skilled ones. Is that the labor shortage you mentioned?

CIA: What I mean is the shortage of ordinary workers, or unskilled workers. The need for skilled workers is a widespread problem everywhere – in America, in China and in Europe. In Germany, it is said that highly skilled workers are the backbone of the country. So the lack of skilled workers is eternal. What is worth our attention is the lack of unskilled ones.

No More Demographic Dividend？

CJ: Does China still have an endless labor supply? When will we need incentives to stimulate labor supply?

Cai: When the Lewis Turning Point comes. The shift from labor surplus to labor shortage is known as the Lewis Turning Point. Before we reach that point, there is an endless supply of labor and a so-called “demographic dividend.” The current birth rate in China is 1.7, much lower than the replacement rate of 2.1, which means China has entered an era of low birth rates, low death rates and low population growth.

It’s wrong to say that China has an endless labor surplus. Recent years have seen a rapid expansion of the economy and employment, but the growth of the population is going down after 30 years of imposing family-planning policies. The reduced birth rate will be reflected in the labor market, and a steep slowdown in labor growth is almost certain to occur.

CJ: When will the Lewis Turning Point hit China? Do we still have demographic dividend?

Cai: The Lewis Turning Point is not a moment, but a period or time span. A typical dual-sector economic model (labor transition between a traditional agricultural sector and a modern industrial sector) features an endless labor supply, and companies don’t have to use incentives to lure employees. But when you have to increase wages to retain workers, that means the Lewis Turning Point has arrived. There are actually two Lewis Turning Points. The first involves raising salaries to hire staff; the second is the integration of urban and rural China to the point where their marginal productivity are on par. The first turning point has arrived. But it might take 20 to 30 years for the second to come.

The idea of a demographic dividend is somewhat murky. When we calculate a demographic dividend, we use the dependency ratio as the key variable. The dependency ratio divides populations into three age groups: under 15 years, 16 to 64, and above 65. The young and elderly groups together, divided by the middle group, yield the dependency ratio. The higher the ratio, the heavier the burden. According to our estimate, the dependency ratio has been decreasing since the mid-60s, but the downturn will come to a stop in 2013 and start to rise. That’s what we call aging.

Our conclusion is, when the dependency ratio drops by 1 percentage point, its contribution to GDP is 0.115 percentage point. That means the dependency ratio contributed 27 percent of China’s per capita GDP growth during the last 30 years. If the conclusion is valid, then after 2013 when the dependency ratio rises by 1 percentage point, our per capita GDP growth will suffer a 0.115 percentage point loss.

Even without the economic cycle, some companies will have to close because the Lewis Turning Point is taking effect. That would be a gradual, slow process, but inevitable. Companies used to operate without raising salaries for as long as a decade, but now they have to raise salaries by double-digit percentage points. That will be too much to some companies, and they will be swept away by the trend.

Productivity and Total Factor Productivity

CJ: Can we say that rising salaries in today’s China will help drive up the labor supply? Won’t that increase unit labor costs? And can the growth of labor productivity catch up with the rise of labor costs?

Cai: Rising incomes will certainly increase the labor supply and ease the shortage. Generally speaking, China’s labor is very cost-effective. Most workers have at least a middle-school education, thanks to China’s basic education system. Labor productivity in China also rises fast. We conducted a survey between 2000 and 2004 and found out that the rise of manufacturing salaries was less than one third of the rise of marginal productivity of labor. The two numbers should be close under normal circumstances. In 2004, the salary level was about one fourth of marginal productivity of labor. That means the current rise of labor costs is backed by the increase in productivity, and the former hasn’t exceeded or even caught up with the latter. So our real labor efficiency and competitiveness are still rising.

I have one concern about labor productivity. When we conducted the survey, companies provided only the numbers of permanent staff. If all the sales and revenue data is accurate, the number of workers might be artificially low because many of the newly created jobs are informal ones with no contract, and therefore they are not included in our survey. So productivity might be inflated. In that case, we should lower the average salary level as well as productivity.

But the ratio of household income divided by national income continues to rise, showing the effect of increasing wages. China’s household income/GDP ratio was 40.5 percent in 2006, which was not very low. Of course, in countries like the U.S., the ratio is above 50 percent. But most middle-income nations have a ratio lower than China’s. For example, Argentina, Mexico and Venezuela are 35.9 percent, 32.6 percent and 31.1 percent, respectively. In Thailand and the Philippines, it’s 28.5 percent and 27.2 percent.

Again, there are methodology problems that cause an under-estimated labor income ratio. In other words, the employment rate from the urban labor survey (conducted by Cai) is higher than the labor statistics survey (released by the National Bureau of Statistics) by 53 percent. The income of informal employment might also be missing from GDP. In addition, not all wages of migrant workers are calculated. There is also informal employment. If you add all these together, we should have a lower average wage level, but a higher total amount.

CJ: Does that mean now is the time to raise total factor productivity to spur economic development?

Cai: Economists think that return to capital (ROC) has to diminish as a economic law, and that we should shift the growth from a reliance on inputs of capital and labor to increasing total factor productivity. That’s the only way to maintain increasing ROC. In the U.S., there is an increasing ROC. Economist Paul Krugman has said that the fast development of some Asian countries such as Singapore was not miracle, because there was no increase in total factor productivity behind high growth and such the growth is not sustainable.

The problem with his conclusion is that he didn’t realize the contribution of a demographic dividend to economic growth. The reason for a diminishing ROC is a labor shortage, and that’s a basic assumption of neo-classical economics. If there is no labor shortage, there won’t be diminishing ROC. Capital and labor income can rise hand in hand.

Not every country has a demographic dividend, especially some developed countries. The U.S. used to have it. The so-called transatlantic economy relies on a constant flow of immigrants. Work forces migrate from the old world to the new world, that’s a very good structure in a labor force and creates a demographic dividend. But it’s in fact a small dividend, significant only compared with the old world.

The effect of a demographic dividend is more obvious in developing countries. Demographic transition happen much more rapidly in developing countries due to medical services and family planning. Japan and the four Asian dragons are good examples of high growth with constant ROC. Their high growth relied heavily on capital and labor investment. Because there is a demographic dividend, that didn’t cause diminishing ROC.

When systematic reforms were carried out in China, a demographic dividend started to take effect. The freed-up productivity also led to an increase in the total productivity rate. But much research shows that total factor productivity is a residual. And according to the World Bank, two factors driving it are technological efficiency and allocation efficiency. Our interpretation is, reallocation efficiency, or labor readjustment, contributed 21 percent of China’s GDP, and technological efficiency contributed only 3 percent. That means China can raise total factor productivity, but it’s not necessary at this moment as China has not entered into diminishing ROC.

In terms of total factor productivity in China, the general view is that its rise was faster in the 1980s, but the growth rate slowed in the 1990s.

CJ: Does that mean we don’t need technological upgrading to drive GDP growth before the arrival of the Lewis Turning Point?

Cai: Yes. That’s why although we’ve been talking about changing the development model for so long, no fundamental changes have taken place, and a cheap, low-cost development model prevails. When the Lewis Turning Point comes, the demographic dividend disappears, and what that says for our economy is, now is the time to change. Because when the labor surplus ends, the ROC will go down. If we don’t turn to total factor productivity, we can’t sustain our development, and that exactly meets Krugman’s expectation.

CJ: What are the major obstacles and difficulties in carrying out the change?

Cai: Changing the development model means shifting from a reliance on capital and labor to the growth of total factor productivity. Some research shows that in the late years of the era of the “Four Asian Dragons,” total factor productivity rose sharply. But the Soviet Union suffered a constant labor shortage. It should have transformed to a growth pattern relying on productivity rise. But it didn’t.

Since the 1980s, China has promoted a shift from an extensive economy to an intensive one. That was interpreted by many as a shift from less technology to more, from low capital intensity to more. But that’s not the key. In the Soviet Union, there was a heavy concentration of technology and capital, but total factor productivity was negative.

In terms of obstacles, the major one is a distorted price mechanism, particularly for factors of production. The more prices are distorted, the slower the transition. Indeed, we don’t know how to raise total factor productivity. It’s a task left to entrepreneurs. What the government has to do is to make the price of inputs right.

Agricultural Subsidies

CJ：You are also a leading agricultural expert. Did you talk about agricultural policy during the meeting?

Cai: I made some suggestions about direct grain subsidies. Statistics shows that subsidies to farmers are at a record high this year, at 103 billion yuan. That means the direct subsidies to grain farmers have doubled every year in the past few years. The subsidies have increased farmers’ income, and we should continue.

But some research shows that direct subsidies increase income, not productivity. The subsidies are allocated to land, regardless of production and sales. These subsidies are allowed because they fall in the “green box” category of the World Trade Organization (WTO). They are not against the WTO since they have little effect on sales, production, or price. Therefore, they have limited ability to boost grain production.

How do the subsidies work? If a piece of land has yielded taxes previously, it would be eligible for subsidies today. But many farmers are laboring elsewhere and lease their land for others to cultivate. The owners of the land receive subsidies, but the ones who are cultivating the land don’t. In some labor-exporting regions farming is reduced from two seasons a year to one and much land is left idle. That’s the real hazard to food security.

Food prices have been rising fast globally, but not in China. In northeastern regions, the price of corn has even dropped, and some farmers are complaining that it’s hard to sell food crops. Some economists said at the meeting that some higher food prices will spur farmers to produce, as well as generating more income. But direct subsidies bring few results at high costs.

We should let food prices rise and subsidize consumers. If the low-income urban population totals 100 million, we should subsidize their food costs, which might rise 200-300 yuan per year, or 20 to 30 billion yuan for the whole country. That’s much cheaper than the current direct subsidies to producers.