Research
Publications
This paper explains the cyclical behavior of the fluctuations in unemployment and vacancies by demand externalities. Adding such externalities to an otherwise standard search and matching model reduces the need for exogenous shocks in explaining these fluctuations. Under plausible parameter values, the equilibrium dynamics include a stable limit cycle that resembles the empirically observed counterclockwise cycles around the Beveridge curve. Calibrated to the duration of the business cycle, these endogenous `Beveridge cycles' are as persistent as the data, without losing any of the amplification of the standard model.
Awarded the Hicks-Tinbergen Award 2022 by the European Economic Association
Journal of the European Economic Association, 19 (1), 2021 [journal] [online appendix] [replication package]
Housing transactions by moving homeowners take two steps—buying a new house and selling the old one. This paper argues that the transaction sequence decisions of moving homeowners have important effects on the housing market. Moving homeowners prefer to buy first whenever there are more buyers than sellers in the market. However, this congests the buyer side of the market and increases the buyer–seller ratio, further strengthening the incentives of other moving owners to buy first. This endogenous strategic complementarity leads to multiple steady state equilibria and large fluctuations, which are broadly consistent with stylized facts about the housing cycle.
Summary in professional journal ESB (in Dutch):
Moen, E., P. Nenov, and F. Sniekers (2015). `Eerst kopen of eerst verkopen op de woningmarkt', Economisch Statistische Berichten, 100(4706), 169-199.
I present the macroeconomic effects of COVID-19 using the Keynesian cross. I show that the rest of the economy suffers from a lack of demand once one sector of the economy is shut down, and that the government spending and tax multipliers are smaller than usual. Fully insuring workers in the sector that is shut down cannot prevent a recession, but for the same aggregate transfers, such targeted income transfers do more to restore aggregate output than unconditional transfers. I extend the analysis to the IS-curve, and argue that a lockdown results in disinflation. These insights can be taught in an introductory or intermediate macroeconomics course.
In most OECD countries, unemployment rates show no trend, which is puzzling if advancements in ICT decrease labour market frictions. We show, both analytically and quantitatively, that accounting for the secular decline in self-employment rates solves the puzzle. While declining labour market frictions can theoretically explain these trends, we provide contradictory causal evidence that the rollout of broadband Internet increased self-employment and decreased unemployment rates. We reconcile these observations with a new model featuring frictions in both labour and goods markets. We explain falling self-employment and non-trending unemployment quantitatively by labour market frictions declining relatively more than goods market frictions.
Working papers
We study the effect of borrowing constraints in an assignment model of the housing market. When constraints apply symmetrically to all households, these lead to lower prices but unchanged housing consumption. When households can invest their own wealth and may differ in tastes, borrowing constraints will in general result in lower house prices and higher housing consumption for unconstrained households, while housing consumption of constrained households may fall. Binding borrowing constraints result in profitable arbitrage possibilities for buy-to-let investors. They can buy houses that are preferred by constrained households unable to finance them, and make them available as rental housing. In an equilibrium with free entry of such investors, house prices and the allocation of houses to households is the same as without borrowing constraints.
We propose a theory of self-employment which, contrary to earlier studies, does not rely on ex-ante heterogeneity across agents. Households face a choice between self-employment and searching for a job at a firm. Both labor and product markets have search frictions. In the mixed strategy equilibrium, households trade the risk of not meeting a customer when self-employed against the risk of unemployment and the rent extraction in a guaranteed wage contract. The equilibrium is inefficient under risk-aversion due to excessive vacancy creation and self-insurance by self-employed, who attract too many customers by low prices. Type of employment dependent differentiated taxes and unemployment insurance benefits can restore efficiency under a balanced budget.
Work in progress
Household income and self-protection by the self-employed (new draft coming soon)
We document that when self-employed workers have worse access to other sources of income within their household, they earn less per hour. Using annual British panel data, we regress hourly self-employed earnings on `residual household income' - a combination of spousal income, income from savings and investment, income from a second job, etc. - in an otherwise standard Mincer equation. Controlling for individual fixed effects and marriage, there is some evidence of positive assortative matching in the marriage market, but the effect remains: moving up one standard deviation in the distribution of residual household income increases hourly earnings by 2 to 6%. We interpret this effect as self-protection: risk-averse self-employed workers' attempt to increase the likelihood to earn income by decreasing their prices. Consistent with this interpretation, the effect is absent for payroll employees, who cannot readily choose their own wages.
Buying first or selling first? An empirical analysis of moving strategies on the housing market
We merge the universe of housing transactions with the universe of residential registration spells to analyze the relationship between moving behavior and housing market conditions in the Netherlands. We document a substantial shift in the fraction of moving owner-occupiers that buys first versus that sells first since the Great Recession. The fraction that buys first falls from about 80% to less than 50%, only recovering slightly in the most popular cities. We find a strong relationship between the choice to sell first and time-on-market. Hedonic price analysis shows that the decision to buy first or sell first affects the bargaining position of buyers and sellers and thus sales prices.
Sleeping
Despite strong strategic complementarities to buy first whenever the majority of other owner-occupiers buys first, and sell first whenever the majority sells first, the fraction of households that buys first moves slowly over time, and never approaches zero or one hundred percent. This paper explains this lack of coordination by informational frictions: the ratio of buyers to sellers is unknown to moving owner-occupiers. Assuming that these owners switch between buying first and selling first by learning about other moving owner-occupiers can result in asymmetric steady states. Moreover, when owner-occupiers take into account the presence of speculating real estate firms that buy up houses when they are cheap, a limit cycle can exist. This limit cycle closely matches the aggregate dynamics of the housing market.