WORK IN PROGRESS
Estimating the insurance value of public policies
Many government policies redistribute income between individuals, but also provide income insurance by redistributing income from periods of high income to periods of low income within the life-cycle of an individual. Such policies include for example different social insurance programs, income taxes, and differentiated sales tax/VAT rates for different goods. This paper provides a general framework for characterizing and estimating the insurance and redistributive effects of such policies. I provide a novel Slutsky-style decomposition of individual-level welfare impacts of policy changes when there is uncertainty about the state of the world. This decomposition allows me to characterize the insurance and redistributive effects of government policies. The insurance value of a policy is a function of unobservable marginal utilities. Building on the unemployment insurance literature in the Baily (1978) – Chetty (2006) -tradition, I develop three methods to estimate the insurance value. Using a consumption based approach together with data from the Panel Study of Income Dynamics, I estimate the insurance and redistributive effects of multiple government policies.
Payroll taxes, incidence, and input choices of firms (with Youssef Benzarti and Jarkko Harju)
This paper studies the incidence of payroll taxes and the effects of payroll taxes on the input choices of firms. We exploit the abolition of a size-based capital depreciation threshold in Finland above which employer-level payroll tax rates were higher, creating a tax notch. We report large impacts on firm distribution and dynamics that extend far from the threshold. We show that standard difference-in-differences estimators are biased in such a setting and develop a new method to provide meaningful estimates of the impacts of the reform. Using our novel methodology, we find large effects on a variety of outcomes such as employment, revenue and investment, suggesting sizeable scaling effects. Our incidence estimates suggest a 37--63 split between firm owners and workers, respectively.
Difference-in-Kinks design (with Petri Böckerman, Ohto Kanninen and Ozan Yanar)
This paper introduces the Difference-in-Kinks (DiK) design, an econometric framework that extends the regression kink design to settings where the slope of a policy rule changes over time. By combining features of regression kink design and difference-in-differences, the DiK design identifies causal effects from time-series variation in kink intensity. We formalize sharp and fuzzy versions of the estimator and provide identification conditions under a parallel trends assumption. An application to Finland’s 2011 guarantee pension reform shows that changes in marginal incentives significantly affect retirement behavior. The DiK design offers a new tool for policy evaluation in dynamic, non-linear settings.
When are sin taxes effective? From substitution to elasticities (with Tuomas Kosonen and Riikka Savolainen)
This paper studies when are sin taxes and other targeted consumption taxes effective in affecting consumption. We organize the paper through a simple model that characterizes how consumer preferences between two goods are related to elasticity of consumption. The model yields as a result a pattern where demand elasticity is very small otherwise, but when the two goods are very close substitutes the elasticity suddenly explodes to a large level. Empirically we analyze a Finnish sin tax scheme for sweets, soda and ice cream providing us with quasi-experimental variation through multiple reforms. We have product and store-level data on sales and prices containing hundreds of millions of observations. We also provide survey evidence on substitution preferences across categories of goods. Our estimated consumption elasticity is close to zero for sweets and ice cream that have intermediate non-taxed substitute: cookies. In a stark contrast, when the tax rate was doubled for sugary soft drinks but not for their very close substitute non-sugary soft drinks, consumption elasticity is close to unity. These estimates align well with our theory framework. We also provide a meta-analysis to literature analyzing consumption elasticities and show that the elasticity estimates align with our theory framework.