【原创】Has Okun’s Law Become Invalid in China?

Has Okun’s Law Become Invalid in China?

- Estimation of Potential GDP Growth Rates and Natural Rates of Unemployment Based on Dual Structure

Zou Peijiang  (邹沛江)

Institute of Economics and Management, Southwestern University of Finance and Economics, Chengdu, China

Abstract: This paper calculates the potential GDP growth rates and the natural rates of unemployment in China using the dual decomposition of the simple Okun’s law. The results show that every 1% increase in China’s urban registered unemployment rate leads to a 1.98% drop of GDP, the potential growth rate of GDP of China is about 10%, and the average natural rate of unemployment is about 3.2% in China in the past three decades. The paper finds that Okun’s law is still valid in China in the long run, but the Okun coefficient is about -3.79, overestimating the impact of the registered unemployment rate on GDP growth, and needs to be corrected based on China’s dual economic structure.

Keywords: Okun’s Law, dual structure, natural rate

JEL Classification: E24, E23, E32

1.Introduction

Arthur Melvin Okun proposed the famous Okun’s law in his published thesis in 1962, which described the stable relationship between the change in the unemployment rate and economic growth. According to Okun (1962), for every 1% increase in the unemployment rate in the U.S., the GDP decrease by about 3%. This stable relationship has been proved by many studies, such as the study by Freeman (2001) on the panel data of ten industrial countries, only that the study reduced the coefficient from 3% to 2 %.

But Okun’s law seems to be having trouble in China, as lots of empirical studies indicate that the law is invalid in China. For example, the study of Jiang and Liu (2005) shows that a typical Okun’s law relationship does not exist between the urban registered unemployment rate and economic growth in China. However, their study only uses a simple regression to analyze the changes of urban registered unemployment rate and economic growth without any further discussion on the model of Okun’s law. The study of Zou (2003) also shows that there is no typical and stable Okun’s law relationship between unemployment and economic growth in China.

Cai (2007) states in his study that “the existence of Okun’s law requires strict conditions, namely sound market environment and market mechanism”, and that “China’s unemployment are mainly natural one due to the structural changes in the reform and labor market imperfections. The change of unemployment rate in this process is the impact of restructuring and imperfect labor market on the natural unemployment rate. On the other hand, the Okun’s law analyzes the deviation the unemployment rate from the natural unemployment rate, i.e. the impact of cyclical unemployment on economic growth,” therefore, the Okun’s law in the traditional sense cannot be found in China.

As Okun’s law describes the relationship between the potential GDP growth rate and the natural rate of unemployment, Knotek (2007), an economist with the Federal Reserve Bank of Kansas, has pointed out in his paper, “the Okun’s law plays a significant role in Fed’s policy-making.” The estimations of these two natural rates can better help policy makers determine whether the economy is overheated and whether there is full employment in the labor market. The validation of the law in China can help better formulate macroeconomic policies and employment policies to facilitate China’s economic development. There are different approaches for the measurement of potential GDP growth rate and the natural unemployment rate.

In his estimations of the natural rates in Europe (potential GDP growth rate, the natural rates of unemployment, core inflation rates), Berger (2008) separates the GDP growth rate into the long-term portion and the periodical portion before estimating it with Okun’s law and the Phillips curve. The premise of the study is that the simple Okun’s law is established in the region, as the data on European countries used by Berger (2008) show that there is an obvious Okun’s law in Europe. However, if the simple Okun’s law is not valid in China, it is not feasible to estimate the GDP growth rates and the natural rates of unemployment using the simple Okun’s law.

The Institute of Economics of Chinese Academy of Social Sciences (2005) uses vector auto-regression and Kalman filtering to estimate the potential GDP growth rate in China, and the result is about 8%. The institute (2006) uses HP filter and the result is about 9%. Yang (2007) uses the production function and gets a result between 10.54% and 11.56%.

One of the most popular methods of estimation of the natural rate is the method based on the decomposition of unobservable components. It divides the GDP growth rate and the unemployment rate into the long-term stable portion and the cyclical portion. Appropriate analysis and measurement can then calculate the natural rate based on the long-term stable rate.

The above literature review shows that the simple Okun’s law is not significantly established in China because of China’s imperfect market system and dual economic structure. Therefore, this paper uses Cobb-Douglas production function to deduce the Okun’s law in a dual economy before estimating the potential GDP growth rate. Then it removes the unobservable portion of the unemployment rate to estimate the natural rate of unemployment. This is how the paper calculates the potential GDP growth rate and the natural rate of unemployment in China.

2. Simple Okun’s Law and China’s Reality

Okun (1962) defined the Okun’s law in three forms, one being:

The above comparison seems to say that the simple Okun’s law is invalid in China, as there is no significant relationship between unemployment and economic growth. However, we should note that the U.S. has been a mature market economy for a long time, and its economic data can better demonstrate the characteristics of a market economy. China only officially started to introduce the market economy system in 1978 with reform and opening-up, and its economy has its own unique characteristics in many ways. Most of China’s labor force is still located in rural areas, and China has a dual economic structure featuring difference between urban and rural development. Therefore, the dual economic structure in China has to be considered when studying China’s unemployment.

In addition, a large number of rural workers came into the cities as China’d sped up its urbanization process, a group known as migrant workers. The employment status of these migrant workers is not included in China’s urban registered unemployment statistics, therefore the statistics are inaccurate and only reflect the unemployment situation of the urban citizens.

Therefore, in the next section, this paper uses Cobb-Douglas production function to establish the model of Okun’s law under China’s dual economic structure before using China’s real data to estimate the potential GDP growth rates and the natural rates of unemployment to determine whether the law is valid.

3. Model Setup

3.1 Potential GDP Growth Rate Based on Douglas Production Function

(1) Workforce decomposition of the two sectors

This paper assumes that the production function is Cobb-Douglas production function, namely:

of China (1979-2010)

中文GDP growth rates     HP-filtered potential growth rates

Figure 3 shows the trends of real GDP growth rates, as well as the filtered potential GDP growth rates in a more smoothed curve (please refer to Appendix Table 1 for HP-filtered potential GDP growth rates and the real rates of that year).

We can see that the real GDP growth experienced violent fluctuations over past 30 years. The real GDP surged from 1980 to 1985 because of the overheated economy at the initial stage of reform and opening-up. The real rate was as much as 15% at the peak, significantly higher than the potential GDP growth rate. Due to political causes, GDP fell sharply from 1988 to 1990 to below the potential growth rate, but it quickly returned to the initial level. China regulated the overheated economy from 1992 to 2000, and real GDP growth fell from the peak to achieve a soft landing for the economy. We can see that the potential GDP growth rate also declined due to the tightening policy. Due to the SOE reform and a series of other factors after 2000, the improved production efficiency and easy monetary policy have caused a rapid growth from 2000 to 2007, and the potential GDP growth rate also started to rise slowly. The real rate fell below the potential rate in 2008 and 2009 because of the international financial crisis. The economy bottomed out in 2009 and began to recover in 2010 due to a series of government policies to stimulate the economy, and the real rate and the potential rate gradually moved closer.

Descriptive statistics show that the filtered average potential GDP growth rate has been 10% for the past 30 years. As can be seen from Figure 3, the filtered GDP growth is very stable, indicating the stable nature of potential GDP growth rate. There are also some ups and downs, due to the impacts of specific macroeconomic conditions. For example, the potential rate decreased around 2000 because of deflation and started to rise after 2000 because of the economic growth.

4.1 Estimates of the Natural Rate of Unemployment

First, we use the formula (18) in Section 3 to estimate the cyclical unemployment rate and formula (17) to estimate the natural rate of unemployment. We need the cyclical unemployment rate for formula (18). As we have the potential GDP growth rate from previous regression analysis, we need to find how the observed GDP growth rate relates to the HP-filtered potential GDP growth rate, so that the cyclical GDP growth rate can be included in the regression residuals. After that, we can obtain the cyclical GDP growth rate with extraction of regression residuals. The regression results are in Table 7.

As can be seen in Figure 5, Okun’s law was valid in China before 2000. For example, the unemployment rate was higher than the natural rate of unemployment in 1990, because the GDP growth rate was below potential GDP growth. Because the unemployment rate was lower than the natural rate in 1985, the GDP growth rate was higher than the potential rate.

However, at around 2000, the GDP growth rate was not higher than the potential rate when the registered urban unemployment rate was below the natural rate, and when the registered urban unemployment rate was higher than the natural rate, the GDP growth rate went up instead of dropping. But such a phenomenon gradually disappeared after 2005. It is because of the deviation at that period that many researchers such as Zhu (2009), Yi and Zhou (2010) believed that the Okun’s law was invalid in China.

This paper believes the deviation from Okun’s law in China around 2000 is mainly caused by the SEO reform. The reform laid off a large number of workers who enjoyed the government unemployment benefits for the first three years. A lot of such unemployed workers did not register as unemployed, which is why there was no significant increase in the registered urban unemployment rate. Because of the restructuring and other reasons during the reform, the production efficiency was reduced and the real GDP growth dropped to below the potential rate. Therefore, the GDP growth rate was lower than the potential rate even as the registered unemployment rate was below the natural rate of unemployment.

When the three-year government unemployment benefits ended, the unemployed workers were registered, and the registered urban unemployment rate rises above the natural rate of unemployment. With the completion of the SOE reform and improved production efficiency at around 2002, the GDP growth rate became higher than the potential rate.

As the failure of Okun’s law only appeared in the large-scale SOE reform period, this paper believes that Okun’s law is still valid in China in the long run, just that the simple Okun coefficient may overestimate the impact of changes in the unemployment rate on GDP growth rate.

5. Conclusions

Based on the previous model deduction and empirical analysis, this paper argues that every 1% increase in China’s urban registered unemployment rate will lead to a 1.98% decrease in the real GDP growth rate. Over the past three decades, the potential GDP growth rate is approximately 10% and the average natural rate of unemployment is about 3.2%. Meanwhile, we believe that the Okun’s law is still valid in China in the long run, just that the value of -3.79 of Okun coefficient overestimates the impact of the registered urban unemployment rate on GDP growth, and needs to be modified in accordance with China’s dual economic structure.

The conclusions show that the various economic policies to reduce unemployment in China can promote rapid GDP growth and have a positive impact on the national economy.

References:

(1) Cai, Fang. 2007. “Why does Okun’s Law Fail in China: Review on the Relationship between Economic Growth and Employment.” Macroeconomics, 1: 11-15.

(2) Edward, C. P, and Robert. J. H. 1980. “Business Cycles: An Empirical Investigation.” Camegie Mellon University discussion paper.

(3) Freeman, Donald. G. 2001. “Panel Test of Okun’s Law of Ten Industrial Countries.” Economic Inquiry, 4(39): 511-523.

(4) Huang, Xinfei, and Yuan Shu. 2010. “Empirical Analysis of Sacrifice Ratios with Long-lived Effect Based on HP Filter.” The Journal of Quantitative & Technical Economics, 27(6): 119-146.

(5) Institute of Economics of Chinese Academy of Social Sciences. 2005. Applications of VAR and Kalman Filter.

(6) Institute of Economics of Chinese Academy of Social Sciences. 2006. The Report of China Business Cycle Analysis during the Eleventh Five Year Plan.

(7) Jiang, Wei, and Shicheng Liu. 2005. “Okun’s Law and Its Empirical Study in China (1978-2004).” Statistics and Decision, 4: 7-9.

(8) Knotek E. S. 2007. “How Useful Is Okun’s Law.” Economic Review, Q IV: 73-103.

(9) Okun, A. M. 1962. “Potential GNP: Its Measurement and Significance.” American Statistical Association, Proceedings of Business and Economics Section.

(10) Peng, Zhaoqi, and Chao Sun. 2011. “Research on China Economic Growth: Based on HP Filter.” Journal of ShanXi Finance and Economics University, 33(1): 15-17.

(11) Rober, J.H, and Edward C. P. 1997. “Postwar US Business Cycles: An Empirical Investigation.” Journal of Money, Credit and Banking, 1(29): 1-16.

(12) Tino, B. 2008. “Estimating Europe’s Natural Rates.” Empirical Economics, 40(2): 521-536.

(13) Yang, Xu, Juan Li, and Zhehao Wang. 2007. “The Calculation of China’s Potential Growth Rate: Based on Okun’s Law of Dual-economy.” The Journal of Quantitative & Technical Economics, 24(10): 14-23.

(14) Yin, Bibo, and Jianjun Zhou. 2010. “China’s High Economic Growth and Low Employment: Empirical Study of Okun’s Law in China.” Finance and Economics, 262(1): 56-61.

(15) Zhang, Liancheng, and Bei Han. 2009. “Analysis of Potential Growth Rate in China: The Choice and Application of Parameters for HP Filter.” Research on Economics and Management, 3: 22-28.

(16) Zhang, Xinghua. 2006. “Research on China’s Rural Labor Migration.” Contemporary Economics, 9: 4-7.

(17) Zhu, Xiao. 2009. “Empirical Analysis of Okun’s Law in China: Coexistence between High Growth Rate and High Unemployment Rate in China.” Business Economy, 319(2): 16-17.

(18) Zou, Wei. 2003. “Okun’s Law and China’s Unemployment.” The Journal of World Economy, 6: 40-47.