TY - JOUR
T1 - Panel data analysis of housing choices
AU - Börsch-Supan, Axel
N1 - Funding Information:
Housing choices have almost exclusively been analyzed either by using survey data gathered at a specific point in time or by using time-series of aggregate data. Panel data permits use of both time-series and cross-sectional variation in observed housing choices as well as their underlying economic and demographic causes, thereby providing a substantially superior identification and separation of the various mechanisms that are often confounded in the snapshot analysis of cross-sectional data and by aggregation in time-series analysis. Typical examples are the impossibility to distinguish age and cohort effects in a point-of-time analysis, and the bias generated by linear *I would like to thank Wolfgang Schneider and Peter Schmidt for computational assistance. I am indebted to Henry Pollakowski for supplying the housing price data and for comments on earlier versions of this paper. Two unusually helpful referee reports helped to improve the readability of the paper considerably. Financial support by the Deutsche Forschungs-gemeinschaft is gratefully acknowledged.
PY - 1990/6
Y1 - 1990/6
N2 - Housing choices have almost exclusively been analyzed either by using survey data gathered at a specific point in time or by using time-series of aggregate data. Panel data permits use of both time-series and cross-sectional variation, thereby providing a substantially superior identification and seperation of the various economic and demographic mechanisms underlying housing choice behavior which are often confounded in the snapshot analysis of cross-sectional data and by aggregation in time-series analysis. In this paper, a longitudinal discrete choice model of the choice of housing tenure and size is presented and estimated using five linked cross-sections of the Panel Survey of Income Dynamics, 1977 to 1981. We employ the conditional fixed effects multinomial logit model [FEMNL, Chamberlain (1980)] in order to account for time-invariant heterogeneity across households. These panel data results differ in many respects from results obtained from single and pooled cross-sections. First, price and income elasticities appear substantially overestimated in cross-sectional analysis as opposed to time-series and panel data analysis. Second, life-age effects are confounded by calendar-time specific effects and therefore may yield implausible results in cross-sectional analysis. In general, the influence of demographic variables appears to be understated in cross-sectional estimation.
AB - Housing choices have almost exclusively been analyzed either by using survey data gathered at a specific point in time or by using time-series of aggregate data. Panel data permits use of both time-series and cross-sectional variation, thereby providing a substantially superior identification and seperation of the various economic and demographic mechanisms underlying housing choice behavior which are often confounded in the snapshot analysis of cross-sectional data and by aggregation in time-series analysis. In this paper, a longitudinal discrete choice model of the choice of housing tenure and size is presented and estimated using five linked cross-sections of the Panel Survey of Income Dynamics, 1977 to 1981. We employ the conditional fixed effects multinomial logit model [FEMNL, Chamberlain (1980)] in order to account for time-invariant heterogeneity across households. These panel data results differ in many respects from results obtained from single and pooled cross-sections. First, price and income elasticities appear substantially overestimated in cross-sectional analysis as opposed to time-series and panel data analysis. Second, life-age effects are confounded by calendar-time specific effects and therefore may yield implausible results in cross-sectional analysis. In general, the influence of demographic variables appears to be understated in cross-sectional estimation.
UR - https://www.scopus.com/pages/publications/0025635155
U2 - 10.1016/0166-0462(90)90025-X
DO - 10.1016/0166-0462(90)90025-X
M3 - Article
AN - SCOPUS:0025635155
SN - 0166-0462
VL - 20
SP - 65
EP - 82
JO - Regional Science and Urban Economics
JF - Regional Science and Urban Economics
IS - 1
ER -