In the latter stages of the 21st century, China's government undertook a fundamental reform of the nation's economy. The gradual transition to a market economy resulted in extremely high levels of output being sustained over the course of nearly three decades. A considerable body of work has been devoted to exploring this trend outcome. However, far less attention has been paid to exploring the nature, behaviour, causes and effects of output growth volatility. Understanding output growth volatility is important from a policymaking perspective, as well as serving to increase our understanding of the workings of the Chinese economy. This thesis assesses four key questions related to output growth volatility.
The first relates to the nature of output growth volatility. The existing literature focuses on business cycle volatility and in doing so ignores the fact that volatility can also occur at both higher and lower frequencies. An analysis of the nature of output growth volatility provides insight into whether the majority of this may be attributed to fiscal and monetary policymaking. That is, if the majority of output growth is present at business cycle frequencies. However, through use of spectral density analysis it is determined that a significant proportion of output growth volatility is present at lower than business cycle frequencies. This indicates that volatility is likely to have been driven by changes emanating from the supply-side of the economy.
The second question relates to the behaviour of China's output growth volatility over time. This is particularly interesting given that a wide range of countries experienced significant declines in the volatility of their macroeconomic measures during the 1970s and 1980s. Using a wide range of analytical tools it is shown that China experienced a significant decline in its volatility in the early 1990s. It is established that there is instability in the variance of the real GDP growth and a structural break in the variance of the series in 1994Q3. It was also determined that the mean rate of growth and the variance of the inflation series declined during the reform period.
The third question relates to the causes of the decline in volatility. These causes are examined through use of spectral density analysis. Given data constraints, this emerges as the most appropriate analytical technique and it also has the benefit of being able to nest several different hypotheses that might explain the volatility decline. The results indicate that the decline in volatility may predominantly be attributed to an improvement in policy and practices, such as changes in the implementation of monetary policy.
Finally, the effect of volatility on mean growth is examined through use of ARCH-M analysis. It is found that there exists a positive but statistically insignificant relationship between these two measures, which to a large extent confirms a priori expectations.