China has experienced a rapidly growing economy for more than three decades since reform and opening. Now China becomes the second largest economy in the world and its growing speed is still very high. Though China achieves great successes, it faces a lot of problems. Regulation is one of the tough issues. Recent years, skewed distribution of income has become a sensitive issue in China. Because it is related to how to deepen the reform, it causes a wide public concern. More and more people blame society equality on economic growth and always say income gap widening is the inevitable product in the process of developing. As a result, how to enhance regulation now is a hot issue attracting more and more people to talk about in China.

Get Help With Your Essay
If you need assistance with writing your essay, our professional essay writing service is here to help!
Essay Writing Service

Nearly all countries face the growing income disparities when they develop from poor countries to the developed and China is not an exception. Different groups get different gains, the rich become richer and the poor become poorer in the process of economic development. The economy grows with a skewed distribution of income. Genii coefficient is an acknowledged indicator to show the degree of inequality. With the economy grows, China’s Genii coefficient also increases rapidly, from 0.315 in 1980 to 0.438 in 2010, which has already surpassed the world’s recognized cordon, 0.4. This demonstrates the income distribution imbalance in China is serious and there is a potential trend that the income gap will become larger and larger.
I collect time series data from 1980 to 2010 in China and most of them are from China’s National Bureau of Statistics. There are two variables. For economic growth, I use per capita GDP (G) as logarithmic form to represent it. Here I choose Genii coefficient (GN) to measure inequality (Though not comprehensive, it’s a common practice). China’s government does not publish Genii coefficient before, but it can be calculated.
Now I get all the data I need and then I will test the stationary property of these time-series data. I utilize five unit-root tests to measure the stationary property, which will also help to set up a VAR model later. Here are the five tests: (1) augmented Dickey and Fuller (1979) (ADF), (2) Phillips and Perron (1988) (PP), (3) Elliott (1996) came up with Dickey-Fuller GLS detrended (DF-GLS). (4) Kwiatkowski et al.(1992) (KPSS), (5) Ng and Perron (2001) -NP. I can see from the result that nearly all the two variables are not significant in level, which means they are not stationary. But after differencing, all G and GN are significant. So the integration order cannot exceed 1 and the highest integration order d=1.
Then I will set up a level VAR model by using TY procedure, and test Granger causality between the variables. TY procedure is attractive because it has the follow three advantages: (1) TY procedure does not need to know variables’ cointegration property so I do not have to make a cointegration test before modeling. (2) TY procedure is allowed to any level of integration order of all variables. (3) Utilizing TY procedure can make us have a level VAR model, which can retain all data information (differencing may cause some information loss).
TY procedure test is in fact a Wald test. Employing TY procedure means we will Change a level VAR (k) model to a level VAR (k+d) model. Then test for whether the first k parameters are significant, which will follow an asymptotic chi-square distribution with freedom degree k. According to the previous part, I get highest integration order d=1 and in terms of Schwarz criteria, I get optimum lag length k=1. So I set up a VAR (2) model. The formula is as follow:
(1)
Where , is a column vector of constant, and are both coefficient matrixes. is a white noise process.
From the result, I can see that per capita GDP unidirectionally Granger cause Genii coefficient (Though per capita GDP is significant at 10% level). This means per capita GDP can help to forecast the variation of Genii coefficient and the influence is positive. In addition to it, we should know that Genii coefficient does not Granger cause any other variables.
I can get the follow conclusions from the outcome above: In the long run, China’s per capita GDP can help forecast the variation of Genii coefficient and there is logic that economic growth may cause more serious equality, but this influence is not so significant (only in 10% level). There is no evidence to show that Genii coefficient can affect other variables. The Granger causality is unidirectional between them.
Then, I will utilize impulsive response and variance decompositions. The Granger causality test can reflect the long-term relationship of these variables and impulsive response and variance decompositions can be employed to test variables’ short-run relations, so they can provide some valuable views on the relations in the short run and make us have a more comprehensive understanding of the relationship.
Impulsive response shows how the aim variable will react to other variables’ impulse at first and how long the effect will last, whether it will disappear quickly. Variance decompositions show what proportion of forecast error variance of the aim variable can be explained by the change of other variables.
From the outcome of impulse response, I can see that Genii coefficient has a positive reaction in the start of per capita GDP’s impulse and the effect will last for a few periods before disappear. Here I can get a conclusion that: in the short run, economic growth has a positive effect on Genii coefficient, which will cause income gap increasing and more serious inequality in the short run.
Then let’s see the condition in variance decompositions. From the outcome of variance decompositions, I can see that at start, more than 30% of Genii coefficient’s error variation can be explained by per capita GDP, but this influence disappears in the long run. Of course, Genii coefficient itself has always been the main reason for error variation. These outcomes match with Granger causality test. Genii coefficient cannot help to forecast the error variance of per capita GDP all the time. This is also similar to the conclusion in the Granger causality test that the relation is unidirectional.
Conclusion and Regulation Suggestions
Through setting up a level VAR model by using TY procedure, this paper investigates the relationship of inequality and economic growth by employing Granger causality test, impulse response and variance decompositions. I get follow conclusions:
In the short run, economic growth has positive effect on Genii coefficient, which means skewed income distribution will be more serious with economic growth. But in the long run, this effect is weak and can be nearly ignored and so economic growth won’t increase income gap. The other way round, inequality does not affect economic growth no matter in the short run or long run. As a result, from a long-term view, there is no evidence to show we need to weigh economic growth and inequality.
Though GDP is the second largest in the world, China’s per capita GDP is still in a low level. From my model, I can see that in the future, economic growth shouldn’t be a main reason for society inequality. Economic growth will not for sure lead to inequality. China’s government is supposed to implement and enhance some redistribution measures (such as tax reform and social insurance) to adjust and improve income distribution, not be worried about that may hinder economic growth.
References
[1] Alesina, Alberto, and Dani Rodrik (1998). Distributive Politics and Economic Growth, Quarterly Journal of Economics, 109, 465–490.
[2] Deininger, Klaus and Lyn Squire (1996). A New Data Set Measuring Income Inequality, World Bank Economic Review, 10, 3, 565–591.
[3] Edin, Per-Anders, and Robert Topel (1997). Wage Policy and Restructuring: The Swedish Labor Market since 1960, in R. B. Freeman, R. Topel and B. Swedenborg (eds.), The Welfare State in Transition, Reforming the Swedish Model. Chicago: University of Chicago Press, pp. 155–201.
[4] Fishcher (2001)．Evolution of inequality after trade liberalization，Journal of Development Economies，66, 555—579．
[5] Forbes, Kristin J (2000). A Reassessment of the Relationship Between Inequality and Growth, American Economic Review, 90, 869–887.
[6] Guillermo Perry, Marcelo Olarreaga (2006), Trade Liberalization, Inequality and Poverty Reduction in Latin America, World Bank, Washington, DC
[7] Kakwani, N. and H. H. Son (2006), How costly is it to achieve the millennium development goal of halving poverty between 1990 and 2015? Working Papers, No. 19, UNDP International Poverty Centre, Brazil.
[8] Li, Hongyi, and Hengfu Zou (1998). Income Inequality is Not Harmful for Growth: Theory and Evidence, Review of Development Economics, 2, 318–334.
[9] Silva J A，Robin M L (2003)．Regional income inequality and international trade．Economic Geography，80, 261-286．
[10] Toda, H.Y., Yamamoto, T (1995). Statistical inference in vector autoregression with possibly integrated processes, Journal of Econometrics, 66, 225–250.
[11]Ugur Soytas, Ramazan Sari and Bradley T. Ewing (2006), Analysis Enegy consumption, income, and carbon emissions in the United States, Ecological Economics, 62, 482-489
[12] Wood A，Ridao-Cano C (1999)．Skill, trade and international inequality, Oxford Economies Papers，51, 89—119．