On the Responsiveness of Housing Development to Rent and Price Changes: Evidence from Switzerland Simon Büchler 1 Maximilian von Ehrlich 1 Olivier Schöni 1 1 University of Bern, Center for Regional Economic Development Swiss Real Estate Research Congress 2018
Housing supply elasticity The literature has mainly focused on demand-side factors to explain rent and price increases of residential real estate. However, the supply side of housing markets is equally decisive for rent and price dynamics. If the housing supply can easily adjust to demand pressures, rent and prices will only increase moderately. If the housing supply is rigid even a small increase in demand might lead to considerable rent and price increases. The relevant economic measure is the so-called housing supply elasticity. 1/15
Motivation Why should we care about housing supply elasticity: The lower the housing supply elasticity the more demand shocks translate into higher rents and prices. Inelastic housing markets tend to be more volatile. Affordability crisis, housing is a basic necessity. However, little is known about long-term housing supply elasticities. 2/15
Swiss case 3/15
Contributions Estimate the long-run supply elasticity of Swiss housing markets with respect to rents and house prices. Investigate the role played by different levels of aggregation. Analyze the impact of: Geographic constraints. Geographic constraints Regulatory constraints on the intensive and extensive margin. Forest and FFF Evaluate the relative importance of these dimensions for the rent and price responses observed in Switzerland. 4/15
Key findings The average long-run supply elasticity in Switzerland is about 1.6 and 0.5 with respect to rent and price changes. At the municipal level, the rental (price) supply elasticity ranges from 0.2 (0.11) for Geneva (GE) to 2.49 (0.64) for Zwischbergen (VS). Extensive margin regulatory constraints decrease local supply elasticities to the largest extent, followed by intensive margin constraints and geographic ones. Geographic and regulatory constraints have a stronger impact on rent supply elasticities than on price elasticities. 5/15
Data Main data: Housing advertisements 2004 to 2016 (Source: Meta-Sys) Approximately 2.1 million postings of rental housing units. Approximately 0.8 million postings of selling properties. Federal Register of Buildings and Habitations (Source: FSO) Census of the entire residential housing stock of Switzerland. Federal Population Census and the Population and Households Survey (Source: FSO) Information on households socio-demographic characteristics. Country gird sample: Partitioning Switzerland into small square cells of 2 2 km. 6/15
Supply constraints BLN and others 7/15
Supply constraints [label=supply constraints 2] 8/15
Estimating long-run housing supply elasticities ln(yit) τ = β τ ln(q it ) + α τ s i + ɛ τ it i = unit of observation, and t = time. = time difference between 2005 and 2015. y τ = average asking rents (τ = R) or asking prices (τ = P ) per square meter. q it = total housing stock. s i = time invariant supply shifters. β τ = inverse housing supply elasticity i.e. housing supply elasticity is given by 1 β τ. Endogeneity problems Intruments 9/15
Determinants of Swiss housing price elasticities ln(yit) τ =β s,τ ln(q it ) + β hist,τ ln(q it ) q i1980 + β constr,τ ln(q it ) Λ i q i1980 + α τ s i + ɛ τ it Λ i = geographic or regulatory restriction in location i. q i1980 = total housing stock in 1980. β hist,τ β constr,τ β hist,τ and β constr,τ = heterogeneity in housing supply elasticity due to geographic or regulatory constraints. Regression results 10/15
Introduction Data Methodology Results Conclusion Literature Local supply elasticities with respect to rents 11/15
Introduction Data Methodology Results Conclusion Literature Local supply elasticities with respect to prices 12/15
Determinants of housing supply elasticity 13/15
Conclusion Swiss housing supply adapts more easily to rent than to price signals. There is considerable heterogeneity in housing supply elasticities across locations. Both regulatory and geographic constraints have a considerable effect in reducing local supply elasticities. There are trade-offs in a growing economy between restricting residential development and rent/price dynamics. The impact of policies aiming to affect housing demand will vary across space depending on the local supply elasticity. 14/15
Literature Büchler, S., Ehrlich, v. M. and Schöni, O. 2018. On the Responsiveness of Housing Development to Rent and Price Changes: Evidence from Switzerland. Study for the Swiss State Secretariat for Economic Affairs (SECO). Büchler, S., Ehrlich, v. M. and Schöni, O. 2018. Wie reagiert das Wohnraumangebot auf Preisänderungen?. Die Volkswirtschaft, Nr. 3/2018: 58-60. 15/15
Geographic constraints Go back 16/15
Regulatory constraints extensive margin Go back 17/15
Regulatory constraints extensive margin Go back 18/15
Descriptive statistics within agglomeration (n=1,167) Go back 19/15
Descriptive statistics country grid (n=2,022) Go back 20/15
Endogeneity Problems Simultaneous causality bias arises because prices influence the housing stock, but housing stock also influences the prices. Find instrument that: is a strong predictor of demand changes (instrument relevance). only affects prices through a shift in demand (instrument exogeneity). Go back 21/15
Instruments for changes in local housing demand Bartik instruments Weak or irrelevant instruments Foreign distribution Go back 22/15
Regression results Go back 23/15
Bartik instruments z it = J j=1 f ijt0 f it0 f cjt1 f cjt0 f cjt0 i = unit of observation and c = canton. j = Swiss/foreign status or main spoken language. t 0 = 2000 and t 1 = 2015. f = the number of residents of a given nationality or spoken language. Go back 24/15
Weak or irrelevant instruments Go back 25/15
Foreign distribution in 2000 Go back 26/15