Two important issues in open economy macroeconomics are explored in four separate essays. The first essay is an empirical examination of the importance of money in the determination of economic activity in the United States. This is followed by a suite of three essays that develop and apply a methodology for measuring the extent of international capital mobility using non-differential, time series, saving and investment data of eight OECD countries Essay 1. Recent literature has concluded that money no longer plays a fundamental role in the determination of real and nominal economic activity in the U.S., especially when data from the late 1980s are included in the analysis. This paper reexamines the issue using data sets ranging back to the Civil War. Cointegration tests show that an equilibrium relationship holds between real money and real income in all data samples of thirty-five years or longer. While cointegration may not hold for some of the short samples, failure is equally likely in the pre-war and inter-war periods as in the recent past. Furthermore, when normal lags between money supply changes and resultant changes in income are taken into account, cointegration frequently obtains in samples of twenty years or less. Vector Autoregressions with annual data reveal that money is the most important variable in explaining real output for both the full (post-1875) sample period as well as for the post-1952 sub-sample. Quarterly analysis supports these results for all but the 1952-1982 sub-sample, a period during which the Fed was targeting interest rates. Finally, using monthly data, by adjusting the VAR specification to capture recent structural changes in the U.S. economy, it is shown that money still Granger-causes economic activity in the post 1982 data set Essay 2. A benchmark definition of capital mobility is established. It is argued that the Feldstein-Horioka (1980) estimation strategy provides information about capital mobility that is not available via tests of interest rate parity. A literature review then illustrates that main empirical pitfalls identified by other researchers Essay 3. Evidence is provided that log-levels of U.S. saving and investment are I(1) and cointegrated over the 1947:1-1995:2 sample period. However, a sub-sample analysis reveals an I(0)-sub-sample-I(1)-full-sample inconsistency in the unit root tests. As a check, Zivot and Andrews' (1992) trend break ADF procedure is employed. Both series are shown to test I(1) even when a trend break is incorporated into the alternative hypothesis. Phillips and Hansen (1990) Fully Modified OLS yields a saving-investment coefficient of approximately 0.8, which is significantly different from both zero and one. Parameter stability tests verify that the saving-investment relationship is stable throughout the post-war sample Essay 4. Time series properties of saving and investment are explored and time series estimates of saving-investment correlations are obtained for eight OECD countries. All regressions are checked for cointegration and for parameter stability. Results are broadly supportive of Feldstein and Horioka's original contention that long-term capital is highly immobile, but reject the hypothesis of complete immobility / acase@tulane.edu
| Identifer | oai:union.ndltd.org:TULANE/oai:http://digitallibrary.tulane.edu/:tulane_24417 |
| Date | January 1996 |
| Contributors | Davis, Mark Steven (Author), Tanner, J. Ernest (Thesis advisor) |
| Publisher | Tulane University |
| Source Sets | Tulane University |
| Language | English |
| Detected Language | English |
| Rights | Access requires a license to the Dissertations and Theses (ProQuest) database., Copyright is in accordance with U.S. Copyright law |
Page generated in 0.1452 seconds