Abstract:
Energy price shocks are large and persistent relative to other price shocks. How do these shocks affect households in different income groups? Using the Consumer Expenditure Survey (CEX), I document a negative relationship between household income and the expenditure share on energy for both commuting and residential utilities. Households in the lowest income quintile spend 12.5% of their budget on energy, while the highest spend 7.8%. To quantify the welfare effects of an energy price shock across income groups, I develop a heterogeneous-agent incomplete market model featuring non-homothetic consumption preferences, costs of commuting to work, and energy as a factor of production for non-energy goods, taking energy price as exogenous. A calibrated version of the model successfully reproduces many salient features of the United States data, including the cross-sectional distribution of income, wealth, and expenditure shares on energy consumption for both commuting and residential utilities. An energy price shock in the model unevenly impacts households in different income groups. Specifically, a one-standard-deviation inflationary shock results in four times greater welfare losses for those in the lowest income quintile than those of the highest quintile on impact. The paper also shows that while work-from-home opportunities exacerbate inequality, targeted transfers help to mitigate it.
Publications
Progressive Income Taxation and Consumption Baskets of Rich and Poor
Journal of Economic Dynamics and Control, December 2023, 157: 104758
Abstract: In this paper, I analyze the implications of differences in consumption baskets across income groups to evaluate the effects of redistributive taxation on efficiency and inequality. To this end, I develop a static multi-sector general equilibrium model incorporating a parametric tax function, non-homothetic consumption preferences, and endogenous labor supply, with varying compositions of skilled and unskilled labor in production across sectors. A calibrated version of the model captures the cross-sectional differences in the compositions of households' consumption baskets in the United States. I find that considering the differences in consumption baskets between high- and low-income households leads to a lower optimal choice of income tax progressivity compared to the conventional approach that ignores this feature of the data.