آیا قوانین ورود به مدرسه، پیشرفت آموزشی و نتایج بازار کار را تحت تاثیر قرار می دهد؟

Age based school entry laws force parents and educators to consider an important tradeoff: though students who are the youngest in their school cohort typically have poorer academic performance, on average, they have slightly higher educational attainment. In this paper we document that for a large cohort of California and Texas natives the school entry laws increased educational attainment of students who enter school early, but also lowered their academic performance while in school. However, we find no evidence that the age at which children enter school effects job market outcomes, such as wages or the probability of employment. This suggests that the net effect on adult labor market outcomes of the increased educational attainment and poorer academic performance is close to zero.

Recently there has been substantial interest in the choice that parents face as they decide at what age to enroll their children in kindergarten. Several papers have documented the adverse effects on academic performance of being the youngest student in a classroom in the United States. Bedard and Dhuey (2006) use data from OECD to show that the youngest members of fourth and eighth grade classes have standardized test scores that are 2–12 percentiles lower than the oldest students in the same cohort. Similarly, Datar (2006) used variation in school entry cutoff dates to document that children that start kindergarten later get higher test scores.1Elder and Lubotsky (forthcoming) used the Early Childhood Longitudinal Study to document that a 1 year increase in the age at which an individual enters school reduces the probability they will be held back a grade at some point in elementary school by approximately 13%. They also find differences in test scores, but this outcome is largely driven by accumulation of skills prior to kindergarten and declines rapidly as children age. Studies focused on other countries found more mixed effects.2 Overall, these findings have lead to substantial concern among both parents and educators about the effect of age based school entry laws, and legislators in several U.S. states have changed their school entry dates in order to increase the age at which children enter kindergarten.3 All these results from the literature suggest that enrolling children in kindergarten as soon as they are eligible may be adversely affecting them. However, as we document in this paper, there is at least one positive effect of enrolling in kindergarten at the earliest age possible. The youngest students in a class complete high school at higher rates than their older peers as noted by Angrist and Krueger (1991).4 This suggests that there is an important tradeoff to consider. This paper provides estimates of the net long run impact of these opposing mechanisms on labor market outcomes in the United States. In addition to getting at the net effect of the tradeoff described above, labor market outcomes of adults are arguably of greater interest than the intermediate outcomes, such as academic performance, that are typically considered in the literature. To conduct our analysis we use the restricted access Decennial Census Long Form Data for the states of California and Texas.5 Unlike the publicly available micro-sample (PUMS), the restricted-access data has the exact day of birth for each individual for a 15% random sample of the population of each state. Our research design uses state school entry laws that regulate the minimum age at which students are eligible to enroll in school as a source of exogenous variation in the timing of school entry. The state of Texas requires that a child must be at least 5 years old by September 1st in order to enroll in kindergarten that academic year, while the threshold date is December 2nd in California for most of the age groups we examine. We take advantage of these threshold dates to implement a regression discontinuity (RD) design. The RD approach lets us estimate the long run consequences of early school entry, by comparing individuals who are similar on all dimensions, but enter school at different ages on account of the school entry laws.6 The analysis focuses on adult outcomes of individuals over the age of 30 as they are more likely to have completed their education. We find that the school entry laws have a modest effect on educational attainment: adults born right before the cutoff for school entry in Texas and California are about a percentage point more likely to complete high school. They are also about a half percentage point more likely to complete 9th, 10th and 11th grades. Evidence from contemporary cohorts shows that though school entry laws have a very pronounced effect on the timing of school entry, a substantial part of the difference is undone through retention. Data from recent cohorts also show that youngest students have lower academic performance, as measured by retention rates. Interestingly, we find no evidence that school entry laws and the additional education that results from them leads to differences in employment rates, wages, or in any of the other outcomes we observe in the Census, such as family income, house ownership, house value and marital status. We find no evidence that early school entry has an impact on adult outcomes for any of the age, gender and race subgroups we examine, not even for Hispanics who have the largest difference in educational attainment of any of the contemporary cohorts. The rest of the paper is organized as follows: in Section 2 we discuss the empirical model and data sources. In Section 3 we examine the impact of school entry laws on educational attainment. In Section 4 we present evidence on adult labor market outcomes. In Section 5 we show educational attainment and labor market outcomes by subpopulation. Section 6 presents estimates of the impact of school entry laws on contemporary cohorts. Section 7 concludes.

In this paper we documented that, though students that enter school at a younger age due to the school entry laws have poorer academic performance, on average they also have slightly higher educational attainment. When we examine the net impact of this tradeoff on adult outcomes we find no evidence that the timing of school entry affects wages or any of the other outcomes that we observe in the Census. Though the educational attainment of Hispanics is substantially more affected by the school entry laws than any of the other groups we find no evidence of any effect on labor market outcomes even in this subpopulation. These result suggests either that the increase in educational attainment induced by the school entry laws is offsetting the poorer academic performance of children who start school at a younger age or that variation in academic performance that is due purely to relative age, and not adjusted away through retention, does not affect labor market performance. We also found that contemporary cohorts of students born right before the cutoff date for school enrollment are significantly more likely to enroll in kindergarten a year earlier than similar students who were born right after the cutoff date. One third of these initial differences disappear by 9th grade since the youngest children in a cohort are held back more often than their older classmates. Minorities are more likely to comply with the law than whites and they are held back less frequently; therefore they make up a disproportionate share of the youngest students in a cohort.