Why are many workers reluctant to leave places in economic decline? This paper explores the role of housing in inter-labor market area (LMA) migration following a regional economic shock. I study the migration response of workers in the oil-exporting region of Stavanger, Norway, following the 2014 global oil price plunge which resulted in a persistent reduction in the region’s labor demand and home prices. Leveraging administrative data, I empirically document an increase in the leaving rate of renters and homeowners with less housing wealth, in contrast to a decrease among homeowners with more housing wealth. The richness of the data allows me to control for potential confounding factors and selection into housing tenure. To explain my findings, I employ a life-cycle model with endogenous home prices, and location, housing, and saving choices that replicate the reduced-form results. It shows that the adverse shock to housing wealth reduces homeowners’ value of leaving which reduces their leaving rate. For homeowners with the most housing wealth, the net effect is a reduction in the leaving rate. The model also shows that moving subsidies that are unconditional on worker characteristics influence the leaving rate of renters more than homeowners because renters are on average more liquidity-constrained. I also provide additional empirical results on the changes in the inflow to Stavanger: while the young and high-income workers avoid the region, renters and older workers of lower income and with relatives in the region arrive at the same or a higher rate than before. I.e., the population composition does not only change due to who leaves but also who arrives.
Empirical studies have estimated a big range of consumption response sizes to changes in house prices. Using a quasi-experiment, we estimate a shock of –19.4 percent to house prices in the area surrounding an airport in Stockholm after its operations were unexpectedly continued as a result of political bargaining behind closed doors. This source of price divergence is ideal for identifying housing wealth effects since it is local and unrelated to variations in macroeconomic conditions. Using a household data set with information on the location of primary residence relative to the airport, we find a short-run elasticity with respect to new car purchases of 0.39, corresponding to a one-year marginal propensity for car expenditures of 0.12 cents per dollar lost in housing wealth. Households with high loan-to-value ratios and little bank deposits respond the most. A quantitative model is consistent with the empirical findings and pinpoints important determinants of the response size, which may explain the variation in previous estimates.
Following the oil price plunge of 2014, local income and home prices fell in the oil-intense cities on the Norwegian west coast. We study the consumption of public sector workers across Norway and compare their change in expenditures in response to the fall in home prices. We argue that public sector workers are insulated from local oil revenue and that differences in changes in expenditures are due to changes to housing wealth. Using the near-universe of digital payments at a high frequency and with information on the type of store, we can decompose the change in expenditures by multiple categories and study the time dynamics of the shocks.
Does lowering corporate bond rates by QE in itself stimulate firm investment and other real outcomes? By an event-time approach with TRACE data, we try to disentangle channels of large-scale asset purchases to the firm side of the U.S. economy.
In countries with wealth and tax registries, household expenditures can be imputed using the budget constraint method (see, for example, Browning and Leth-Sorensen, 2003). However, during certain life events such as household splits and mergers, housing purchases, or other large financial transactions, or for specific types of workers such as business owners, there is a risk of measurement errors that are not well understood. In this study, we compare the budget constraint method to another source of administrative data: digital payments and transfers in Norway. The goal is to compare the two measures of household expenditures and see when they agree across households and identify situations when the methodologies differ.
Using the universe of car transactions matched to annual census data in Sweden, I study how the cash-flow channel of monetary policy affects households’ decisions to buy cars. Draft to come soon.