Umeå universitets logga

Chapter [I] analyses the dynamic properties of life-cycle earnings of men and women in Sweden by fitting a model to the covariance structure of earnings. We find that the financial crisis of 2008 did not have any major impact on the variability of earnings and that the upward trend in permanent earnings inequality observed in Sweden during the 1990s does not continue in the 2000s. Using the model to simulate the accumulation of income pension entitlements, we find that college-educated workers have smaller variations in pension entitlements than workers with less education.

Chapter [II] presents a life-cycle earnings dynamics model that includes endogenous employment and job change. The model is estimated on Swedish register data using indirect inference. By simulating data from this model, we study the macroeconomic consequences of transitory shocks to unemployment risk, how unemployment at different ages affects the accumulation of pension entitlements, and analyze how different factors contribute to earnings inequality. We find that transitory aggregate shocks to unemployment risk have long-lasting negative effects on employment and earnings, that becoming unemployed at age 40 has a large negative effect on pension accumulations, and that unobserved individual heterogeneity contributes substantially to the observed life-cycle earnings inequality for both men and women in Sweden.

Chapter [III] presents a model of earnings dynamics that includes transitions in and out of employment and business cycle fluctuations. The model is estimated using indirect inference and a mix of Swedish register, survey and macro data. We find that the transitions from unemployment to employment are more sensitive to business cycle fluctuations than the probability of remaining employed. By simulating data from the model, we find that the business cycle has a relatively small impact on earnings inequality in Sweden and that young workers are more sensitive to business cycle fluctuations than older workers.

Chapter [IV] deals with optimal nonlinear taxation of income and profits in a general equilibrium model with frictional unemployment. We find that the government can achieve redistribution of income through taxation without distorting production efficiency. This outcome is possible if the government uses two nonlinear tax instruments, taxing profits and labor income separately. The results also show that including involuntary unemployment creates an incentive to tax entrepreneurial income at lower marginal rates and labor income at higher marginal rates than otherwise.

Paper [I] analyzes the dynamic properties of life-cycle earnings in Sweden using microdata. We study the evolution of permanent and transitory earnings inequality over the period 2002--2015. Our data comes from administrative records gathered in the ASTRID database. We find that some features of the data do not match the predictions of the heterogeneous or restricted income profile models commonly applied in the earning dynamics literature. Instead, we estimate an alternative permanent-transitory (PT) error components model. Analyzing the covariance structure of both male and female earnings, controlling for educational background, we find that the upward trend in permanent earnings inequality observed in Sweden during the 1990s does not continue in the 2000s, and the financial crisis of 2008 did not have any major impact on the variability of earnings. We further simulate the accumulation of income pension entitlements and find that variations in pension entitlements are smaller among college-educated workers.    

Paper [II] studies the life-cycle effects of favorable marginal tax treatment on older workers' optimal life-cycle labor supply, retirement timing, and savings. I develop a structural model in continuous time where the life cycle of a representative agent is divided into three distinct phases: pre-treatment, post-treatment, and retirement. Solutions for consumption and savings, labor supply and leisure, and retirement timing are then obtained by solving the model as a salvage value problem. I then calibrate the model to Swedish earnings data and find that the increased extensive margin labor supply is partially offset by a reduction in hours worked during the pre-treatment period. The total effect, however, is an increase in life-cycle labor supply and consumption. \\

Paper [III] studies the implications of the structure of public pension programs for the trade-offs determining economic behavior over the life cycle. The economy is modeled as a continuous-time overlapping generations model with endogenous labor supply, savings, and human capital formation. Individuals differ in ability and are free to choose how much to work at each period in time and when to enter and exit the labor market. Numerical simulations provide qualitative insights that a redistributive pension system introduces opposite effects on the incentives for retirement for high- and low-skilled individuals, which leads to increased earnings inequality. This effect can, in turn, dominate reduced pension inequality such that lifetime and population-wide income inequality increase. Ultimately, the equity–efficiency trade-off is found to be difficult to characterize. \noindent \\

Paper [IV] explores the effects of pension illiteracy on aggregate labor supply and the redistributive performance of public pension systems. I consider an OLG model in continuous time populated with individuals who differ in labor productivity and pension literacy. Agents suffering from pension illiteracy fail to fully account for the structure of the pension system when planning their economic behavior over the life cycle. I find that pension illiteracy can negatively impact aggregate labor supply and increase earnings inequality and lifetime income inequality. This suggests that pension illiteracy may limit the efficiency gains of an increased correlation between individual contributions and benefits, making the equity–efficiency trade-off difficult to characterize in the context of pension reforms.