Regional Science and Urban Economics 20 (1990) 1-4. North-Holland
SYMPOSIUM R E C E N T ISSUES IN T H E E C O N O M I C S O F H O U S I N G MARKETS Introduction
Konrad STAHL University of Mannheim, 6800 Mannheim 1, FRG
This symposium issue compiles regularly submitted and refereed papers focusing on different aspects of housing. There are three theoretical papers, by Kanemoto, Weber and Wiesmeth, and Rouwendal; and five empirical papers, by B6rsch-Supan, Goodman, Gross et al., Hutchinson, and Persky. Key properties of housing, namely indivisibility in consumption and durability of investment, are in various ways at the heart of the three theoretical analyses. These properties, coupled with incomplete information and informational asymmetries give rise to the formation of rental, or purchasing contracts of the type studied in the first paper by Kanemoto. He focuses on the performance of various contract types when the design of a complete contingent contract is unenforceable, due to transactions costs and/or informational asymmetries between demanders and suppliers of such contracts. In his model, the unenforceability relates specifically to the investment in housing (maintenance) conducted by either one of the contracting parties. In that case, a contract induces Pareto-efficient investment only if the investor can fully reap its benefits. Thus, a purchase contract, or a rental contract with the renter investing will do, as long as tenure is perfectly secure (and investment cost schedules are convex); or by contrast, a rental contact on housing with the owner investing will do, if the renter has no security of tenure. Contracts are also analyzed in Weber and Wiesmeth's paper. They consider a market in which the design of rental contracts is subject to regulation. The system of feasible contracts is thus exogenously specified. The exchange of these contracts is subject to a black market with (illegal) side payments acting as prices. Markets of this type exist in several Eastern European countries, or in cities such as Vienna and West-Berlin. The authors show the existence of a competitive equilibrium in such a contract market 01664)462/90/$3.50 © 1990, ElsevierSciencePublishers B.V. (North-Holland)
K. Stahl, Introduction
and discuss the efficiency properties of equilibrium. In particular, there may be equilibria that are not Pareto-efficient even in a constrained sense where restrictions on the space of feasible contracts enter the specification of Pareto-efficiency. The third theoretical paper by Rouwendal treats the housing market directly, rather than the market for rental contracts. In line with McFadden's seminal econometric work 1 applied by Anas both theoretically and empirically to the housing market, 2 the typical agent's demand for housing is characterized as an~outcome of the maximization of an additive random utility function. Justifying and employing gross substitutability of (expected) market demand as a key property of the theoretical housing market, the author gives a new very elegant proof of the existence of a price equilibrium, and then necessary and sufficient conditions for its uniqueness. The series of five empirical papers starts with three papers on housing demand, followed by one on housing supply, and concludes with a final paper discussing the impact of non-housing commodities on the spatial distribution of households and housing. The first of the papers on housing demand, by B6rsch-Supan, applies recently developed econometrics for analyzing panel data to housing choices. Because of the durability and irreversibility of housing investment, most adjustments in housing consumption require moving, which in turn involves considerable transactions costs. It is therefore natural to analyze housing choices in an intertemporal context. In contrast to cross-sectional data in which age and cohort effects cannot be disentangled, or of aggregate time-series data in which non-linearities in the micro relationships are confounded, the use of panel data allows for an isolation of these effects. This brings new insights, especially on the relative order of magnitude of demographic and economic influences on housing choices. In particular, the price and income elasticities customary from crosssectional data appear substantially overestimated. By contrast, the importance of demographic factors is apparently substantially understated, if discussed within a cross-sectional framework. Both of these results derived by B6rsch-Supan have strong and immediate implications on housing policy, e.g., on the effect of income subsidies on housing consumption. The second paper on housing demand by G o o d m a n also adds to our appreciation of demographic influences on housing demand. He uses crosssectional data to estimate housing demand differentiated by tenure and size. Simple demand estimates involving income and relative prices as independent variables are contrasted to estimates including in addition demographic variables such as age and sex of head of household, marital status, household size and ethnic background. Goodman's main result is that the inclusion of I F o r a survey, see M c F a d d e n (1981). 2Cf. e.g. Anas (1982).
K. Stahl, Introduction
demographic variables does not alter much price and income elasticities if evaluated at the respective means, but very much so if evaluated away from the means. This again has strong impacts on housing policy, since it is the non-average (large, non-white, low-income) household that is typically the object of interventions into the housing market. In contrast to the previous two analyses, the third and final paper on housing demand, by Gross et al., emphasizes an entirely different approach, namely the 'bid-rent' model in which housing prices are estimated as a function of housing attributes. Such estimates are ususally plagued by the problem that the influence of variations in housing and household attributes on bid prices cannot be disentangled. The authors use data on condominium transactions in which households' characteristics can be expected reasonably homogeneous, to provide new and relatively more accurate estimates on condominium bid prices. It is unfortunate that the wealth of contributions to the estimation of housing demand is not nearly matched by contributions on housing supply. The paper by Hutchinson adds to this sparse literature. The author uses a fairly general translog cost function to produce new estimates on the elasticities of substitution between the inputs land, labor and material into the production of housing, and uses these estimates to quantify the effects of a mandated above-market-level minimum wage rate. Testing for various restrictions on functional form, he first shows that the homogenous function performs best. The factor elasticities of substitution calculated from that function suggest that the land-for-labour and the land-for-material elasticities are very close to and less than unity, while that of material-for-labour exceeds unity. 3 Based on these elasticities, the author goes on to show that the magnitude of substitution of other inputs for on-side labour generated from a higher mandated wage is likely to be small. Thus, any policy-induced wage increases are likely to influence costs in proportion to the quality of labor employed for construction. In the final paper of this Symposium, Persky seeks to identify non-housing reasons for the spatial distribution of households within the suburban ring of a metropolitan area. In an attempt to explain in particular the considerable income inequality within individual suburban communities, he considers three influencing factors: the costs of access to the workplace implying, within the standard von Thiinen model, an inverse relationship between distance from the workplace and income; the consumption of local collective commodities implying, according to Tiebout's (1956) hypothesis, a self selection of equals into segregated communities; and the consumption of 'local status' according to Frank's (1985) 'local status' hypothesis. That 3The latter estimate may by blurred be the fact that labor going into prefabrication is counted as materials input.
K. Stahl, Introduction
hypothesis states that high-income families eventually desire to live amongst low-income ones, if thereby they are granted 'social status'. Conversely, lowincome families are interested, according to this hypothesis, in living amongst high-income ones because of the involved implicit or explicit subsidies, and consumption externalities. Persky rejects a modified version of the aforementioned accessibility hypothesis, by also accounting for the suburbanization of employment, with which he associates an increase in income inequality between suburban and CBD employees. He goes on to criticize the Tiebout hypothesis for its failure to account for considerable intra-suburban variations in household income, household composition and tenure status. He finally presents evidence that is at least not contradicting the (somewhat sketchy) theory of local status.
References Anas, A., 1982, Residential location markets and urban transportation (Academic Press, New York). Frank, R., 1985, Chosing the right pond (Oxford University Press, New York). McFadden, D., 1981, Econometric models of probabilistic choice, in: C.F. Manski and D. McFadden, eds., Structural analysis of discrete data (MIT-Press, Cambridge, MA). Tiebout, C., 1956, A pure theory of local expenditures, Journal of Political Economy 64, 41~424.