This dissertation with the title Four Essays in the Economics of Education studies a broad spectrum of research questions. The topics covered range from the documentation of earnings over the working life to school grading, school starting age and discrimination in the labor market for apprenticeships. Chapter 1 deals with the lifetime perspective of education and earnings whereof little is known since a large part of existing literature is concerned with estimating the returns to education. Therefore the purpose of this chapter is to calculate and analyze how age-earnings profiles and work-life earnings differ for educational groups. Work-life earnings describe the accumulated earnings from labor market entry to retirement. The main contributions are using longitudinal data with precise educational information to calculate work-life earnings and introducing a measure of work-life earnings that includes periods when individuals have no earnings due to not being employed. Foregone earnings during education are also calculated. A German dataset is used, i.e. survey data from the adult cohort of the National Educational Panel Study (NEPS). The survey data is linked to social security records (NEPS-SC6-ADIAB) to attain data on employment spells and earnings subject to the social insurance contribution. The findings are that, on average, high skilled men earn 2.3 million EUR over the course of their working life. This is about 0.7 million EUR more than a medium skilled individual earns (1.7 million EUR). Foregone earnings of high skilled men during education as a fraction of work-life earnings are 3.9 percent. High skilled women earn, on average, only 0.3 million EUR more than medium skilled women. The results show that including periods where individuals are not employed makes a substantial difference in work-life earnings. At the same time foregone earnings during education can also add up to a large fraction of work-life earnings. Chapter 2 investigates whether school reports should include grades or verbal assessments in the first years of primary school. A large part of existing pedagogical literature is concerned with this question. Very little is known, however, about the effects of actual reforms that led to later grading on education opportunity. The research question for this project is therefore: what effect does later grading in primary school have on educational transitions and educational achievement for students with different parental background. A reform is analyzed for Germany which meant federal states replaced grades in school reports by verbal assessments in the first classes of primary school starting in the 1970s. A difference-in-difference estimator is employed for the empirical strategy. This allows using variation given by the different timing of the implementation of the reform in federal states to identify a causal effect. Although results show that the coefficients are not significant the estimates still provide some evidence for an effect of later grading on the transition to upper secondary school. Chapter 3 studies whether school starting age (SSA) differentials continue into adulthood or fade away after leaving school. The previous literature has shown that children who enter school at a more advanced age outperform their younger classmates on competency tests taken between kindergarten and Grade 10. This chapter thereby contributes to this literature but also complements the literature investigating the long-run effects of SSA on individuals' wages and employment. The identification of the effect of SSA on competencies relies on an instrumental variable strategy that exploits the state- and year-specific rules given by the cut-off dates. Results show that the impact of SSA on math and text comprehension measured in adulthood are considerably smaller than what the literature has shown for children in school. Further, both estimates are statistically insignificant. In contrast, the effect of SSA on receptive vocabulary is sizable in adulthood, with a one-year-higher SSA increasing competency by around a third of a standard deviation. Chapter 4 is about discrimination in the labor market for apprenticeships. Since individuals without German citizenship have only half the apprenticeship entry rate compared to those with German citizenship it is important to understand what role discrimination plays. In this chapter a correspondence experiment is used to study ethnic discrimination in the hiring market for apprenticeships in Germany. We send fictitious applications with German and Arabic-sounding names to real vacancies, augmenting applications with a short motivational video. These video introductions by applicants serve as a simple way of extending employer insight into the personality and motivation of applicants and therefore solving the problem of discrimination in the hiring market for apprenticeships. This is done with the intention of decreasing potential statistical discrimination. We find that applications with German-sounding names are on average twice as likely to receive an invitation for a job interview. Providing a video in the application that gives employers more insight into the personality and motivation of applicants increases invitations to job interviews for Arabic-sounding names. However, since these conditional effects are insignificant we interpret the results as weak evidence of videos increasing invitations for Arabic-sounding names.