Loyce Bagwell

Publication Date


Advisor(s) - Committee Chair

Robert Pulsinelli, Stephen Lile, John Wassom

Degree Program

Department of Economics

Degree Type

Master of Arts


Many studies approach the housing-income relationship by calculating the income elasticity of demand for housing. This study employs this approach in order to (1) test the implications of the permanent income hypothesis by using data which provide both biased and unbiased elasticity coefficients and (2) test the effects which the explanatory variable, family composition, may have on the elasticity coefficients.

The methodology involves regressing the log of housing expenditure on the log of disposable income for seven different family compositions in order to obtain the income elasticity coefficients. This approach is applied to data (cross section 1960-61 BLS consumer surveys) which are disaggregated into urban and rural samples. The data are then classified into income classes and occupational groupings. The income class data are referred to as biased data because they represent current measured income which contains transitory components. The occupations are referred to as unbiased data because they provide a measure of the group average or permanent income, since transitory effects are presumed to cancel out.

There are three major hypotheses concerning the calculated elasticity coefficients (1) the biased coefficients should be significantly smaller than the unbiased coefficients in the individual groupings (2) the biased group coefficients should be smaller than the unbiased group coefficients and (3) the coefficients should differ across the differing family compositions within the biased and unbiased data if family composition is a significant determinant of the income elasticity of demand for housing.

Hypotheses (1) and (2) are implications of Friedman’s PIH, the first one being the more important in this analysis. The third hypothesis is the major hypothesis of this analysis and is based on the assumption that families with younger children will have higher income elasticities of demand for housing than families with older children.

Analysis of variance is the statistical test which is used to determine if there are significant differences in the elasticity coefficients and, if so, whether these differences are due to the first hypothesis, is biased-unbiased effects or the third hypothesis, family composition.

The results were as follows:

(1) The rural, data, in the main, are not supportive of either the Friedman hypothesis, that the use of measured income rather than permanent income tends to cause a downward bias in the coefficients of the income elasticity of demand for housing, or our major hypothesis, that family composition is significant determinant of the elasticity coefficient.

(2) The urban data, in the main, are supportive of the Friedman hypothesis and our hypothesis regarding housing expenditures and family composition.

(3) It appears that the overall group elasticities for the urban and rural data support the Friedman hypothesis because the biased coefficients are smaller than the unbiased coefficients.


Business | Economics | Macroeconomics | Real Estate | Social and Behavioral Sciences