Gerald A. Whitney

Nonparametric Tests of Utility Maximization and Weak Separability for Consumption, Leisure and Money

Some of the most fundamental assumptions of economics are utility maximization and weak separability of the arguments in the represen tative consumers utility function. This paper contains results from nonparametric tests of these assumptions about consumer behavior. The authors find that quarterly per capita data on consumption goods, le isure, and monetary assets are consistent with utility maximization. Further, consumption goods and leisure meet necessary and sufficient conditions for weak separability. Additionally, one grouping of relat ively liquid monetary assets meets the necessary conditions for weak separability. They also test and find no evidence of homothetic prefe rences. Copyright 1987 by MIT Press.

A revealed preference test for weakly separable utility maximization with incomplete adjustment

Abstract Functional structure is an important issue in many areas of economics. We suggest a nonparametric revealed preference test for weakly separable utility maximization that permits category incomplete adjustment within the period under study. This test is an extension of the necessary and sufficient condition for weakly separable utility maximization developed by Varian (1983). We implement this test and identify a quarterly economic monetary aggregate. This is the first implementation that we know about of any revealed preference necessary and sufficient condition for weakly separable utility maximization. We also perform sensitivity analyses of the test and find it to be reasonably robust.

A New PC-Based Test for Varian's Weak Separability Conditions

This article develops a new method to evaluate revealed preference separability conditions. In contrast to previous studies, our results generally find weak separability, even when datasets have some measurement error. In addition, revealed preference and weak separability appear robust to measurement error, different price distributions, and alternative preference settings. Measurement error generally results in relatively few violations of revealed preference or weak separability.

A Comparison of Nonparametric Tests of Weak Separability for Annual and Quarterly Data on Consumption, Leisure, and Money

Among the most fundamental assumptions made in economics are utility maximization and the separability of the arguments in the representative consumers utility function. These assumptions are important for theoretical and empirical applications of economics. In this article, we present results from nonparametric tests of these assumptions of consumer behavior. We find that utility maximization generally obtains with either annual or quarterly per capita data on consumption goods, leisure, and relatively liquid monetary assets. Annual data on consumption goods, leisure, and all monetary assets are consistent with utility maximization. There is some evidence in support of using partial adjustment models when estimating quarterly data models of the demand for monetary assets. Further, annual data on consumption goods and leisure and on liquid monetary assets meet the necessary and sufficient conditions for weak separability. These results support the notion of a monetary aggregate more broadly based than cu...

The unbiased forward rate hypothesis re-examined

Abstract This paper examines the unbiased forward rate hypothesis using an Error Correction Model (ECM). Previous tests have utilized mainly two specifications—differences and levels. The advantage of the ECM is that the parameters of the level specification appear in the error correction term, and hence the ECM provides a direct link to the levels specification. In this paper we fully exploit this link and demonstrate that both specifications yield essentially the same results when the ECM is explicitly derived from a levels specification and the levels parameters are estimated simultaneously with other parameters of the ECM.

Testing for the Significance of Violations of Afriat's Inequalities

Two new nonparametric procedures are developed to evaluate the significance of violations of revealed preference found by standard nonstochastic tests. Our tests with high probability correctly detect utility maximization for data generated with measurement error. The procedures are not very sensitive to misspecifying the amount of error that could have caused the data to violate revealed preference. The tests have power against an alternative of random behavior. Both tests fail to reject the null of rational utility maximization from a monetary dataset that has violations of revealed preference.

Bounding an Economic Monetary Aggregate under Nonhomothetic Preferences

We construct nonparametric bounds on true-quantity indexes for the realistic case of nonhomothetic preferences. These bounds do not require assumptions about the form of the underlying aggregator function. We construct these bounds using monetary assets that meet necessary conditions for the existence of an aggregate. We find that a Tornqvist index of these monetary assets falls within the nonparametric bounds on true-quantity indexes almost all of the time, whereas a simple sum of these monetary assets falls outside these bounds more than one-third of the time.

The Changing Empirical Definition of Money: A Comment

In a recent issue of this Journal, Belongia and Chalfant (1989) suggested that several groups of financial assets formed monthly economic monetary aggregates from January 1983 through February 1986. They based these findings on nonparametric revealed preference tests developed by Varian (1982, 1983). They implemented these tests using nominal assets data. Implementing these tests with nominal data is inconsistent with a neoclassical framework of consumer choice based on real variables. We find that their use of nominal data led Belongia and Chalfant to obtain results different from those they would have obtained if they had used real assets data. We reran their tests using real data and found that none of the monetary aggregates they identified can be shown to be consistent with the existence of an economic aggregate.

Estimating demand elasticities under rationing

Eliminating rationing in the United Kingdom following the Second World War was a concern for policy-makers because of potentially large fluctuations in post-war prices and the impact on unrationed goods. This study shows that in using virtual prices, elasticities can be estimated from a ‘free’ demand system consistent with observed consumer choices. Substitution estimates without accounting for rationing are misleading. In contrast, using virtual prices and estimating a ‘free’ market system yield results similar to those of the pre-war period. Results show that food rationing affected expenditure across unrationed goods. Rationing on other services had little effect on expenditure across unrationed goods.

Revealed Preference Tests of Utility Maximization and Weak Separability of Consumption, Leisure and Money with Incomplete Adjustment

Swofford and Whitney (1987) investigated the validity of two types of assumptions that underlie the representative agent models of modern macroeconomics and monetary economics. These assumptions are utility maximization and weak or functional separability that is required for an economic aggregate to exist. To reinvestigate the structure of the representative consumer’s preferences we develop a mixed integer programming revealed preference test with incomplete adjustment. We find that both a narrow official US monetary aggregate, M1, and a broad collection of assets are weakly separable. We further find that a modern analog of money as suggested by Friedman and Schwartz (1963) is also weakly separable. We also find that consumption goods and leisure are separable from all monetary goods. We find no evidence that official US M2 or MZERO are consistent with utility maximization and weak separability. That is, the assets in these measures do not meet the requirement for forming an aggregate over goods that is consistent with economic theory. Finally, we find that three broad categories of consumption goods, durables, nondurables and services, do not meet the weak separability conditions required for forming a consumption aggregate. However, a consumption aggregate of nondurables and services is weakly separable.