Studies On Existence And Formation Mechanism Of Economic Growth Convergence Hypothesis

Research results on convergence hypothesis of economic growth have long been controversial. In this dissertation, researches on convergence hypothesis of economic growth in recent half century are summarized; formation mechanism of four convergence hypotheses including conditionalβ- convergence, club convergence,σ- convergence and absoluteβ- convergence are elaborated; and empirical research methods in convergence hypothesis of economic growth are comprehensively overviewed.Based on the above, in this dissertation, nature and connotation of convergence hypothesis of economic growth are extended, and some new concepts and new types of convergence are proposed, which include:"vertical"convergence,"horizontal"convergence,"strong"σ- convergence and"weak"σ- convergence."Vertical"convergence means output per capita of a single closed economy converge to their own long-term steady-states, with other economic structural parameters unchanged;"horizontal"convergence is a dynamic convergence process in which multi-opened economies catch up with each other through flow of production factors such as capital and labor, etc;"strong"σ- convergence refers toσ- convergence index series of N economies follows a steady time series process such as I (0)integration; and"weak"σ- convergence refers toσ- convergence index series of N economies follows a fractional integrated long memory time series process such as I ( d )0 < d< 1. The above new concepts of growth convergence help deepen our recognition and understanding of convergence hypothesis of economic growth, and they also help enrich and complement theoretical concepts and empirical researches of it. From this perspective, in this dissertation, some leading econometric methods and time series methods including autocorrelation function method, spatial filtering, regression testing procedures, unit root test method and fractional integration test technology are adopted to make a thorough and systematic test and analysis on"vertical"convergence,"horizontal"convergence,"club"convergence,"strong"σ- convergence, and"weak"σ- convergence.Test results of"vertical"convergence and"horizontal"convergence show that: in China,"vertical"convergence trend exists only in 13 provinces and municipalities including Beijing, Shanxi, Liaoning, Heilongjiang, Shanghai, Fujian, Hunan, Guangdong, Shaanxi, Qinghai, Ningxia, Xinjiang and Tibet autonomous region; and"horizontal"convergence trend exists in 13 provinces and municipalities including Beijing, Liaoning, Shanghai, Fujian, Henan, Hubei, Guangdong, Chongqing, Shaanxi, Qinghai, Ningxia, Xinjiang and Tibet autonomous region. And it takes different time period for different provinces and municipalities to complete their own convergence process, which means a nationwide"horizontal"convergence does not exist. Test result of club convergence based on endogenous grouping method shows that: except Heilongjiang, Yunnan and Guizhou provinces, three convergence groups has taken shape for the rest 27 provinces. The first convergence group consists of 7 provinces and municipalities, which includes: Shanghai, Tianjin, Jiangsu, Zhejiang, Guangdong, Fujian, and Inner Mongolia; the second convergence group covers 14 provinces and municipalities including Beijing, Liaoning, Ningxia, Shanxi, Shandong, Hebei, Hubei, Jilin, Tibet autonomous region, Henan, Shaanxi, Chongqing, Anhui, and Sichuan; and the third one comprises 6 provinces and municipalities such as Qinghai, Gansu, Xinjiang, Jiangxi, Hunan, and Guangxi. What's more, the researches on convergence characteristics of these three convergence groups indicate boundaries between adjacent convergence groups have been vaguer in spite of the club convergence characteristic of China's inter-provincial economic growth being more obvious; and a convergence trend among provinces and municipalities in adjacent convergence groups has taken shape, which specifies that in the future, there will be more uncertainties in inter-provincial economic growth path in China. Provinces and municipalities now in an upper convergence group may be downgraded into a lower one, and that in a lower convergence group may be upgraded into an upper one. The complicated"upgrading"trend and"downgrading"trend in inter-provincial economic growth path may occur frequently. Test result ofσ- convergence finds the nine tested sample areas show obvious"weak"σ- convergence characteristic,except"Pan-Pearl River Delta"; and for"strong"σ- convergence characteristic,test results of different areas and different convergence indices in same area are mixed, which makes it impossible for us to draw an unique deterministic conclusion. On the whole, a"strong"σ- convergence does not exist in western area, Pan-Pearl River Delta and the third convergence group. But if we only examine one convergence index, that is, natural logarithm of real per capita GDP, we find a strongσ- convergence trend in central area and the first convergence group. The uncertainty of test result in"strong"σ- convergence coincides with controversies in current regionalσ- convergence researches by and large.Finally, convergence and divergence formation mechanism of inter-provincial economic growth path are discussed within a dynamic analysis framework. Firstly, Bayesian model averaging approach is used to objectively select and identify effectiveness and stability of influential factors of China's economic growth from model uncertainty perspective. Result shows: within 18 explanatory variables selected in advance, 6 of them have close relationship with economic growth, which including numbers of college students per 10,000 people; dummy variable of east area; average growth rate of industrial output. Secondly, we choose the five most explainable and most stable influential factors (location factors deleted for not being able to be quantified), and adopt boundary test method which can be compatible with I ( 0)和I (1 ) series to examine dynamic relationship between differences in inter-provincial economic growth and that of long-term influential factors with the result specifies that within five long-term influential factors, only differences in three of them including high education, industrialization, and degree of opening up have long-term relationship with that of economic growth. And finally, when we use ARDL model to test marginal effect of inter-provincial differences in those three long-term influential factors for that of economic growth, we find differences in inter-provincial foreign trade are playing a leading role, i.e., it is differences in inter-provincial foreign trade that result in gap between inter-provincial economic growth. Besides, the imbalance of industrialization and high education are also widening the gap in inter-provincial economic growth.