Revealed Preference and Time Series Analyses of U.S. Macroeconomic Aggregates

Show full item record

Title: Revealed Preference and Time Series Analyses of U.S. Macroeconomic Aggregates
Author: Maia Filho, Luiz Flavio
Advisors: John J. Seater, Committee Chair
John S. Lapp, Committee Member
Douglas K. Pearce, Committee Member
Walter N. Thurman, Committee Member
Abstract: This research extends the literature on the revealed preference analysis of macroeconomic aggregates in multiple ways. The relevance of recent methodological changes in data construction is our first topic, as Varian's (1982, 1983) nonparametric tests are run on U.S. consumption series built under NIPA's old and new methods. The results indicate that previous conclusions on the overall GARP-consistency of data and on weak separability of particular aggregates are affected by the methodological changes in data. Additionally, test results are observed to be sensitive to the adoption of series at different frequencies. The issue of temporal aggregation is examined in two ways. We initially show that those changes do not seem to have significantly altered the univariate time-series properties of aggregates or previous conclusions about the impacts of temporal aggregation on those properties; therefore, the aggregation of economic flows into annual figures is once more found to involve significant losses of information about the dynamic behavior of higher-frequency data. The power of the GARP test in datasets of different frequencies is then investigated from analytical and empirical standpoints. Time aggregation is found to reduce the power of the GARP test. Finally, we apply Varian's tools to study for the first time a dataset including the value of nonmarket services produced inside the household. The modification involves a more detailed picture of consumers' allocation of time, alternatively a source of utility (leisure) or a resource in household production. We observe that the changing number of hours spent on average in household production — due to the increasing participation of women in the civilian labor force over recent decades — can be characterized as a rational decision made by the representative agent in a standard utility maximization model.
Date: 2004-08-17
Degree: PhD
Discipline: Economics

Files in this item

Files Size Format View
etd.pdf 773.2Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record