Conscription and the Returns to Education: Evidence from a Regression Discontinuity*

• Date: 01 July 2020
• Published date: 01 July 2020
• DOI: http://doi.org/10.1111/sjoe.12361

Scand. J. of Economics 122(3), 1112–1139, 2020

DOI: 10.1111/sjoe.12361

Conscription and the Returns to Education:

Evidence from a Regression Discontinuity*

Pierre Mouganie

American University of Beirut, 1107 2020 Beirut, Lebanon

pm10@aub.edu.lb

Abstract

In 1997, the French government put into effect a law that permanentlyexempted young French

men born after 1 January 1979 from mandatory military service. This paper uses a regression

discontinuity design to identify the effect of peacetime conscription policies on education and

labor market outcomes. Results indicate that conscription eligibility induces a significant increase

in years of education but has no effect on employment and wages at the ages of 30–36. Further

analysis shows that the interpretation most consistent with findings is that the average marginal

return to the additional schooling induced by conscription is low.

Keywords: Conscription; France; human capital; returns to education

JEL classification:I20; J24; J30

I. Introduction

There has been considerable interest in analyzing the consequences of

compulsory military service on a wide array of outcomes. These include

earnings (Angrist, 1990; Angrist and Krueger, 1994; Angrist and Chen,

2011), education (Card and Lemieux, 2001), crime (Galiani et al., 2011;

Lindo and Stoecker, 2014), and health outcomes (Bedard and Deschenes,

2006; Angrist et al., 2010; Conley and Heerwig, 2012). This is in part due

to the general policy interest in understanding shocks that are expected to

have large and persistent long-run effects. Indeed, conscripted individuals

are required to serve at a crucial time of their lives, usually characterized

by critical human capital investments. Moreover, the mechanics underlying

these early adulthood shocks are themselves of considerable interest. For

example, the military service and disruption caused by conscription can

*I am especially grateful to Mark Hoekstra for his guidance and support. I also thank Serena

Canaan, Jose Gabriel Castillo, Jason Lindo, Charles Louis-Sidois, Eric Maurin, Jonathan Meer,

and Steven Puller for their invaluable comments and suggestions. Thanks also go to seminar

participants at the Applied-Micro brown bag seminar at TexasA&M University and to seminar

participants at the American Universityof Beir ut, for helpful comments and discussions. Finally,

I would like to thank the staff at the Centre Maurice Halbwachs for assistance in providingme

with the data used in this paper.All er rors are myown.

C

The editors of The Scandinavian Journal of Economics 2019.

P. Mouganie 1113

have direct long-run effects. Alternatively, conscription might affect long-

run outcomes through its effect on educational attainment, as continuing

one’s education can typically allow one to defer mandatory military service.

Finally, understanding the effects of compulsory military service has direct

policy implications for countries that still have such policies in place, and

for those contemplating reinstatement.1

In this paper, I study the effects of peacetime conscription on the

education and labor market outcomes of young French men. Prior to 1997,

all French male citizens had to undergo a ten-month compulsory military

service at the age of 18. Most men preferred to defer their service, and

the best way to do so was to enroll in an educational institution. However,

in May 1996, the French government announced that individuals born after

1 January 1979 were no longer required to enlist, while those born prior

to that date were still required to do so. As there is no reason to believe

the outcomes of men born on either side of this 1 January 1979 cut-off

would have been different without the conscription policy, I use a regression

discontinuity (RD) design to compare the outcomes of individuals who

were barely subject to or exempted from mandatory military service. This

research strategy allows me to overcome selection bias in who is deemed

physically and mentally fit for military service.

Results indicate that men born just before 1 January 1979 acquire four

to six additional months of education compared with those born just after

this date; this is consistent with draft avoidance behavior. This additional

schooling does not result in increased university degree completion, though

I do find evidence of a 5–6 percentage point increase in high school

graduation for individuals from wealthier backgrounds. I find that men

subject to conscription are not more likely to be employed or to have

higher wages, but they do lose around five months of valuable early labor

market experience. I also examine several competing hypotheses as to why

conscription policy had no affect on earnings, despite evidence of increased

educational attainment for those under threat of service. To do so, requires

first knowing the extent to which military service directly affects labor

market outcomes. On the one hand, if serving in the military has a negative

effect on income, then the education induced by conscription has a positive

return, which is offset by a combination of military service and a loss of

early work experience. On the other hand, if military service has a non-

negative effect on wages, then the average marginal return to schooling

induced by conscription is lower than or equal to the loss of early labor

1In Europe, Austria, Denmark, Estonia, Finland, Greece, Norway, and Switzerland still require

their citizens to undergo mandatory military service. Additionally, Lithuania reintroduced

conscription in 2015, and France formally reinstated mandatory military service in 2018.

C

The editors of The Scandinavian Journal of Economics 2019.