Search

We formulate an economic time use model and add to it an epidemiological SIR block. In the event of a pandemic, households shift their leisure time from activities with a high degree of social interaction to activities with less, and also choose to work more from home. Our model highlights the different actions taken by young individuals, who are less severely affected by the disease, and by old individuals, who are more vulnerable. We calibrate our model to time use data from ATUS, employment data, epidemiological data, and estimates of the value of a statistical life. There are qualitative as well as quantitative differences between the competitive equilibrium and social planner allocation and, moreover, these depend critically on when a cure arrives. Due to the role played by social activities in people's welfare, simple indicators such as deaths and GDP are insufficient for judging outcomes in our economy.

Atkinson et al. (2018) propose a measure of the glass ceiling exploiting that top incomes are approximately Pareto distributed. We clarify how this glass-ceiling coefficient describes the increasing scarcity of women further up in the income distribution and show how it relates to the top-income gender gap. If interpreting top income gender differences as caused by a female-specific income tax, the gender gap and glass ceiling coefficient measure its level and progressivity, respectively. Using Danish data on earnings, we show that the top gender gap and the glass-ceiling coefficient evolves across time, the life cycle, and educational groups.

Private consumption demand falls in response to increased unemployment risk during a recession, as households increase their precautionary savings and postpone irreversible durable investments. The postponement effect is seven times as large as the precautionary-savings effect in a calibrated buffer-stock savings model. In consequence, anticipation of future unemployment risk is more important than realized unemployment shocks in accounting for durable expenditure dynamics during recessions, while the opposite is true for nondurables. The importance of anticipation of future unemployment risk also means that having many ’hand-to-mouth’ households, who do not respond to changes in income risk, significantly dampens the demand response for durables to an adverse labor market shock. We find that the model elasticities of durable and nondurable expenditures with respect to unemployment risk are close to what we estimate in micro survey data.

Consumption Dynamics under Time-Varying Unemployment Risk We study the response of households' demand for durable goods to fluctuations in unemployment risk. First, using survey data, we document that household durable expenditures react strongly to unemployment risk, while the effect on nondurable expenditures is indistinguishable from zero. Second, we construct a buffer-stock savings model that includes adjustment frictions for durable goods. We show that although not targeted in the calibration, the model reproduces the semi-elasticities of expenditures to unemployment risk estimated in the data. Third, using the model, we find that the inclusion of adjustment frictions raises the aggregate demand response of durable goods to fluctuations in perceived unemployment risk by approximately 200 percent. Moreover, the aggregate consumption dynamics are state dependent. Upon experiencing an adverse risk shock, the responsiveness of durable goods demand to the interest rate and income changes is dampened, thus constraining monetary and fiscal transfer policies in stabilizing consumption during recessions. The Labor-Market Origins of Cyclical Income Risk We use Danish administrative data 1980-2013 to study the underlying mechanisms generating fluctuations in income risk. We partition the population into 37 narrowly defined educational categories and document the cyclicality of labor income risk for each category separately. For the individual educational categories, mean income growth is strongly correlated with income growth skewness, with an average correlation of 0.87-0.88. We show that the connection between income growth skewness and mean income growth is not only strong in the time dimension, but also in the cross section. Across the 37 educational categories, the correlation between mean income growth and income growth skewness is 0.93-0.96. We show that labor-market frictions together with variations in productivity growth generate the relationship between mean income growth and income growth skewness. In a quantitative job-ladder model, variations in productivity growth quantitatively capture both the time-series and cross-sectional relationship. In contrast, variations in the job-finding rate, the job-separation rate and the offer-arrival rate for employed fail to generate the relationship between mean income growth and income growth skewness in our framework. A Pareto-Distribution Perspective on Top-Income Gender Disparities We propose a novel decomposition of top-income gender disparities into a top-income gender gap coefficient, capturing the absolute underrepresentation of women in the top, and a glass ceiling coefficient, capturing the relative underrepresentation of women at the very top given the representation of women at the top. The decomposition uses that the top of both the male and female income distributions are well approximated by Pareto distributions. We apply our decomposition to Danish labor income data 1980-2013. We find that the gender gap coefficient is slowly and steadily falling over the whole period while the glass ceiling coefficient has been stable since 1995. We perform heterogeneity analysis along three dimensions. First, we perform the decomposition for different age groups. The glass ceiling coefficient has been largely stable across both time and age groups since 1995. Second, we perform the decomposition for parents and non-parents. The glass ceiling coefficient is larger for parents, but this stems from the different income distribution of fathers and non-father men. Mothers and non-mothers have similar glass ceiling characteristics. Third, we perform the decomposition for the two most represented educational degrees in the top one percent, medical doctors and lawyers. Whereas the glass ceiling coefficient is small for medical doctors, it is much larger for lawyers. A Note: The Effect of Assortative Mating on Income Inequality I provide a theoretical upper bound on the effect of assortative mating on income inequality by comparing perfect assortative mating with random mating. The percentage drop in the Gini coefficient from perfect assortative to random mating is bounded by 1− ≈ 29%. Furthermore, I compare the Gini coefficient of the income distribution of actual households with the Gini coefficient under random mating using US census data. Under all specifications, the effect of randomization on the Gini coefficient is never larger than 0.015.