"The Implications of Housing for the Design of Wealth Taxes", International Economic Review, 2022, 61, 125-159
Watch this short video of me explaining the paper
How does housing - an important asset for many households - affect the design of optimal wealth taxes? I analyze the macroeconomic and distributional consequences of wealth taxation using a model with two forms of wealth: capital used by entrepreneurs to produce goods, and housing that provides shelter services. Without housing, taxing wealth instead of capital income would improve the capital allocation and raise welfare. With housing, taxing capital and housing equally would worsen the capital allocation and lower welfare. Instead, wealth should be taxed progressively, exempting most housing, resulting in a worse capital allocation but high welfare gains.
Work in progress
Wealth inequality has elicited calls for higher taxes on capital income and wealth, but also concerns that rich households would respond by concealing their assets offshore. We use a general equilibrium model to study how taxing capital more heavily would affect offshore tax evasion and how this would affect the broader economy. Without evasion, tax revenue could be increased dramatically, inequality could be reduced, and widespread welfare gains could be achieved. After accounting for evasion, however, tax revenue would rise marginally or even fall, inequality would increase, and widespread welfare losses would result.
"Mortgage Interest Deductions? Not a Bad Idea After All", 2023, [slides] Joint with Joseph B. Steinberg, R&R @ Journal of Monetary Economics
Previous studies have found that mortgage interest deductions and other homeownership subsidies reduce welfare. However, these studies have failed to properly account for two features of the rental market: the elasticity of the rental supply curve, which determines how much rents change when rental demand shifts, and how many renters spend most of their income on housing, which determines how much changes in rents affect welfare. We develop a method to estimate the rental supply elasticity using empirical evidence on renters’ property tax incidence, and we demonstrate that subsidizing homeownership actually increases welfare when both features are accounted for.
We propose a new theory on how rents are determined in equilibrium in a city with controlled and uncontrolled rental markets operating simultaneously, support our theory with novel facts for Toronto where this rental market setup exists, and embed it in a general equilibrium model calibrated to Toronto data. In the long run, full rent decontrol would raise average rents, but reduce newborns’ rents and thus raise welfare. Results hold in transition. In the long run, full rent control would reduce rental housing supplied, raise average rents, and reduce welfare. No degree of rent control is optimal.
What would be the consequences of removing the $500,000 exemption given in the U.S. to capital gains homeowners earn upon selling their principal residence? To overcome computational limitations hindering studying this question with real house price growth, I employ the fact that inflation has represented a substantial portion of house price appreciation, and extend a general equilibrium model with constant real house prices to trace the real inflationary gains’ value homeowners make upon selling their home. Abolishing the exemption in the calibrated model to finance lump sum transfers would raise welfare and homeownership, and benefit most households.
What are the implications for welfare of misallocating residential land? I develop a methodology to calibrate a housing model to a transition path of over 50 years of Israeli data on land sales and show that Israel’s government substantially oversold land. While selling more land reduces prices and is beneficial to young households since they do not own any housing, it is determinantal to old households who rely on the value of their housing for consumption in retirement. I find that the latter outweighed the former in the benchmark.
I calibrate a housing model with aggregate uncertainty and extrapolative expectations to U.S. data, and show that expectations about income can explain 20% of the 2008 boom-bust in housing prices. During a sequence of good income shocks, households begin believing that their expected life-time income has gone up, which leads to a rapid increase in housing demand. This, in turn, increases house prices above what they would be under their true fundamentals. When a sequence of bad income shocks arrives, the process is amplified in reverse.