Time-series analysis of economic growth.
Material type:
TextDescription: 219 pISBN: 0599640707Subject(s): Economics, General | Economics, Theory | 0501 | 0511Dissertation note: Thesis (Ph.D.)--Harvard University, 2000. Summary: Recent empirical literature on economic growth mostly focuses on cross-sectional analyses. However, the original motivation of studying economic growth focuses on the time-series and transitory dynamics of macroeconomics variables such as output, capital and wage. We present two approaches to explore the time-series evidence for economic growth. First, the traditional growth accounting analysis shows evidence that is consistent with the neo-classical growth model in the sense that investment in tangible assets play a much more significant role than productivity in explaining output growth of G7 countries for the period 1960–1995. Moreover, contrary to conventional wisdom, we find evidence that the productivity slowdown has continued instead of receded in the 1990s. Second, we develop an error correction model that allows us to test the cointegrating relationships implied by the neo-classical growth model. While the model is consistent with cointegrations, results from Granger causality tests suggest that the assumption of exogenous labor and productivity may not be applicable. Moreover, variance decomposition of output suggests that neither labor nor productivity plays a dominant role in explaining output fluctuations. This implies that the real-business-cycle models and the traditional forecasting models of potential GDP may not be consistent with time-series evidence.
Source: Dissertation Abstracts International, Volume: 61-02, Section: A, page: 0702.
Adviser: Dale Jorgenson.
Thesis (Ph.D.)--Harvard University, 2000.
Recent empirical literature on economic growth mostly focuses on cross-sectional analyses. However, the original motivation of studying economic growth focuses on the time-series and transitory dynamics of macroeconomics variables such as output, capital and wage. We present two approaches to explore the time-series evidence for economic growth. First, the traditional growth accounting analysis shows evidence that is consistent with the neo-classical growth model in the sense that investment in tangible assets play a much more significant role than productivity in explaining output growth of G7 countries for the period 1960–1995. Moreover, contrary to conventional wisdom, we find evidence that the productivity slowdown has continued instead of receded in the 1990s. Second, we develop an error correction model that allows us to test the cointegrating relationships implied by the neo-classical growth model. While the model is consistent with cointegrations, results from Granger causality tests suggest that the assumption of exogenous labor and productivity may not be applicable. Moreover, variance decomposition of output suggests that neither labor nor productivity plays a dominant role in explaining output fluctuations. This implies that the real-business-cycle models and the traditional forecasting models of potential GDP may not be consistent with time-series evidence.
School code: 0084.
There are no comments on this title.