« Back to Results

Determinants of Academic Achievement

Paper Session

Sunday, Jan. 5, 2020 10:15 AM - 12:15 PM (PDT)

Marriott Marquis, Newport Beach
Hosted By: American Economic Association
  • Chair: Andrew Hill, Montana State University

Better Parents or Richer Parents: Understanding Intergenerational Transmission of Human Capital

Aiday Sikhova
Cornerstone Research


There are two essential mechanisms in the canonical model of the transmission of human capital across generations - parental income and parental education. We provide novel empirical evidence to disentangle the significance of these two factors in determining children's human capital. Two reforms in Sweden provide us with natural experiments to separately identify the effects of parental income versus parental education: an educational reform that exogenously changed the level of compulsory schooling and quality of education of the parent generation; and a tax reform that exogenously altered parents' net income. Using Swedish administrative data, we find that a 1,000 SEK increase in parental human capital leads to a 190 SEK increase in children's human capital. Exploiting the tax reform of 1991, we show that a 117 SEK increase in children's human capital - that is, slightly over 60% of the overall effect - is due to the parental education channel.

Poor Sleep: Sunset Time and Human Capital Production

Maulik Jagnani
Yale University


This paper provides the first evidence on the long run importance of child sleep for learning outcomes. I show later sunset reduces children's sleep: when the sun sets later, children go to bed later; however, children fail to compensate by waking up later. Sleep-deprived students decrease time spent studying and increase time allocated to sedentary and compensatory leisure activities. Overall, sustained sunset-induced sleep deficits translate into fewer years of education and decrease primary and middle school completion rates among school-age children.

School Bus Diesel Emissions and Academic Performance: A National Study

Garrett Wes Austin
Georgia State University


Prior work shows that air pollution affects cognitive performance, and school bus diesel emissions meaningfully contribute to this exposure for school-age children. I exploit variation in the timing and location of 28,000 school bus engine retrofits or replacements across the US from 2008-2016 to test how reducing diesel emissions affects student test scores. I also incorporate satellite-based PM 2.5 measurements from the Atmospheric Composition Analysis Group to provide the first evidence that these engine retrofits significantly reduce ambient atmospheric pollution levels, suggesting potentially large spillover benefits from the retrofits separate from the student test score improvements. Finally, I ask how the cost of retrofitting these busses, approximately $233 million over 209 unique grant cycles, compares to the monetized benefits of the observed test score improvements.

Teach Me in My Mother Tongue: Experimental Evidence on Mother Tongue Education from Primary Schools in Malawi

Hyuncheol Bryant Kim
Cornell University
Seonghoon Kim
Singapore Management University


Most sub-Saharan African countries use international languages, such as English and French, as the medium of instruction from primary school education onward. However, it could impose a learning barrier for children who do not have fluency in those languages. To study the impacts of learning in the mother tongue, we randomized the language of instruction during the six-week-long summer school program that provides social studies and mathematics lessons for 4th and 5th graders in Malawi, where the language of instruction changes from Chichewa to English from Grade 5. We first find that students learned in Chichewa have higher mathematics and social science test scores during the summer school, compared to those learned in English. The summer school program also improves final exam scores by average 0.25 standard deviation in the following semesters in which exams were administered in English. Impacts on test scores are higher in the Chichewa group than that in the English group, especially in mathematics, but not in social studies. Given that fluency in the international language is not required for most people’s daily lives in developing countries, our findings imply that delaying the transition of instruction language could help improve learning outcomes of primary school children in developing countries.

What Explains the Gender Gap Reversal in Educational Attainment?

Laurent Bossavie
World Bank
Ohto Kanninen
Labour Institute for Economic Research


The reversal of the gender gap in educational attainment is becoming a global phenomenon. Its drivers, however, are not well understood and remain largely untested empirically. This paper develops a unified theoretical framework that allows to formulate and test two main hypotheses for the reversal. It introduces the tail hypothesis, which builds on the lower dispersion of academic achievement among females observed globally. It also formulates the mean dynamics hypothesis, based on previous literature, which states that females’ school performance has increased over time relative to males. The paper theoretically shows that both hypotheses can explain the reversal in our framework. The parameters of the two hypotheses derived from the model are then estimated using educational attainment data from 1950 to 2010 in over 120 countries. We find that both hypotheses strongly predict the gender gap dynamics in educational attainment when estimated separately. When jointly estimated, however, the predictive power of the tail hypothesis is more robust, while the mean hypothesis plays a weak role in secondary school completion dynamics.
JEL Classifications
  • I2 - Education and Research Institutions