Working Papers

Cross-country Variation in Educational Attainment: Structural Change or Within Industry Skill Upgrading?

Educational attainment varies greatly across countries and within countries over time. This paper asks whether the variation in education is primarily due to structural change or to within industry skill upgrading. The main finding is that within industry variation accounts for at least 2/3 of the cross-country and the time series variation in educational attainment. The within industry education gaps are broadly consistent with a model of industry neutral cross-country differences in skilled labor productivity. These results suggest that theories of educational development should focus on skill upgrading within industries rather than structural change.

[Paper] [Mapping from country specific to U.S. industries] [First draft: Jun-2007] [This draft: Feb-2008] [SED slides]

Educational Attainment in U.S. Cities

This paper proposes a theory of educational attainment differences across U.S. metropolitan areas. The theory is motivated by the finding that employment in business services predicts more than 70% of the observed cross-city variation in education. In the model, agglomeration economies in the production of business services, which are complementary with skilled labor, account for cross-city variation in education. The theory makes a number of testable predictions which find strong support in U.S. data.

Key words: Education, human capital, cities, agglomeration economies.

[PDF paper] [Slides] [Version: Sep-2007] [First draft: December 2006]

Why Does Education Differ Across Countries?

The purpose of this paper is to investigate the reasons why average years of schooling differ across countries. A version of the Ben-Porath model is used to decompose observed schooling differences into the contributions of demographic factors, distortions to investment in schooling and physical capital, and the skill bias of technology. The main finding is that about 80% of schooling gaps are due cross-country variation in skill bias. Distortions to school investments account for roughly 15% of the observed variation. Demographic factors and distortions to investment in physical capital make only small contributions.

Key words: Schooling, education. JEL: I21, J24.

[PDF paper] [First draft: March 2005] [Slides]

Why Do Hours Worked Differ Across Countries? Evidence from U.S. Immigrants

Hours worked differ substantially across countries. Two types of explanations have been proposed: Either local institutions offer different incentives for work, or countries differ in their social norms or "industriousness" (Leamer 1999). This paper offers evidence that sheds light on the relative importance of these explanations. If social norms are important, then hours worked by U.S. immigrants should be correlated with hours worked in the source countries. The data show that this is the case for women, but not for men. Hours worked by female immigrants are significantly correlated with hours worked and with the religious composition of the source countries. On the other hand, if hours worked are determined by local incentives, then U.S. immigrants should work roughly the same hours as do U.S. natives. To a first approximation, this is the case for male immigrants. One interpretation is that social and, in particular, religious norms exclude women, but not men, from the labor market in some countries.
Key words: Hours worked, labor force participation, immigrants, social norms.
JEL: J21, J61, Z13.

[PDF paper] [First draft: November 2004]

Why Does Educational Attainment Differ Across U.S. States?

College attainment differs nearly two-fold across U.S. states. This paper shows that highly educated states employ skill-biased technologies, specialize in skill-intensive industries, but do not pay lower skill premia. A theory based on agglomeration economies is developed to account for these observations.

[PDF paper] [Version: September 2004] [First draft: November 2003]

Intended and Accidental Bequests in a Life-cycle Economy

This paper studies quantitative importance of accidental versus intended bequests. The main finding is that accidental bequests account for at least half, and perhaps for all of observed bequests.

[PDF paper] [First draft: August 2001]  [Slides]

Bequests and Retirement Wealth in U.S. Data

This is a background paper for "Intended and Accidental Bequests in a Life-cycle Economy." It documents bequests and retirement wealth in the SCF and the PSID.

[PDF paper] [PDF tables]