tailieunhanh - The Economic Returns to an MBA∗

The results for women are presented in Table 6. Unlike Montgomery and Powell (2003), the estimated returns to an MBA are consistently lower for women. The return to an MBA for women is estimated to be with no ability controls and falls to with ability controls. The fixed effect estimate is a little under 4%. The return to math ability is higher for women than for men, with again no return to verbal ability. This finding is consistent with the twins’ literature that finds that OLS estimates of the return to schooling are biased upward compared to twin fixed effects (See Card (1999) for. | The Economic Returns to an MBA Peter ArcidiaconoJ Jane CooleyỊ Andrew Hussey January 3 2007 Abstract Estimating the returns to education is difficult in part because we rarely observe the coun-terfactual of the wages without the education. One of the advantages of examining the returns to an MBA is that most programs require work experience before being admitted. These observations on wages allow us to see how productive people are before they actually receive an MBA and to identify and correct for potential bias in the estimated treatment effect. Controlling for individual fixed effects generally reduces the estimated returns to an MBA and especially so for those in top programs. However for full-time MBA students attending schools outside of the top-25 the estimated returns are higher when we control for individual fixed effects. We show that this arises neither because of a dip in wages before enrolling nor because these individuals are weaker in observed ability measures than those who do not obtain an MBA. Rather there is some evidence that those who take the GMAT but do not obtain an MBA are stronger in dimensions such as workplace skills that are not easily measured. Including proxies for these skills substantially reduces the gap between the OLS and fixed effects estimates. Keywords Returns to education ability bias panel data JEL J3 I2 C23 We thank Brad Heim Paul Ellickson Bill Johnson Margie McElroy Bob Miller David Ridley Alessandro Tarozzi and participants at the Duke Applied Microeconomics Lunch for valuable comments. Suggestions by the editor and three referees substantially improved the paper. We are grateful to Mark Montgomery for generously providing the data. - Duke University Department of Economics psarcidi@. University of Wisconsin-Madison Department of Economics jcooley@ University of Memphis Department of Economics ajhussey@ 1 1 Introduction While it is generally accepted that more education leads to an .