Does Productivity Decline After Promotion? The Case of French Academia*
| Date | 01 December 2012 |
| Published date | 01 December 2012 |
| Author | Mareva Sabatier |
| DOI | http://doi.org/10.1111/j.1468-0084.2011.00681.x |
886
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2012. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 74, 6 (2012) 0305-9049
doi: 10.1111/j.1468-0084.2011.00681.x
Does Productivity Decline After Promotion? The Case
of French Academia*
Mareva Sabatier†
†Université de Savoie – IAE Savoie Mont-Blanc - IREGE, BP 80439,
74 944 Annecy-le-Vieux Cedex - France (e-mail: mareva.sabatier@univ-savoie.fr)
Abstract
The present research examined the effect of promotion decisions on ex-post productivity
in French academia. As, once promotion decisions are known, most external incentives
vanish for promoted candidates, their productivity was expected to decrease. This
hypothesis was tested by using an original dataset and matching methods to evaluate the
impact of promotion on publication scores. The robustness of the matching estimates was
tested using sensitivity analysis. The results clearly show that the removal of extrinsic
incentives following promotion does not lead to a fall in productivity in French academia.
I. Introduction
The central role of promotions in organizations has been a major focus of the literature
on careers. Promotions have two goals (Milgrom and Roberts, 1992). First, they aim to
improve the organizations’ performance by assigning employees to the jobs that best suit
their abilities. Second, they serve as incentives [see Gibbons and Waldman (1999) and
Prendergast (1999), for a survey]. The tournament model (Lazear and Rosen, 1981) illus-
trates this second role well. Imagine a firm in which all the workers initially have the
same low-ability-level job for which they all receive the same wage. Able workers are then
promoted to jobs that demand higher levels of ability and that are rewarded with higher
wages. The greater the wage-difference between the two jobs, the greater the incentive
to work hard. Promotions can also boost performance by providing incentives for higher
human–capital investment (Carmichael, 1988; Prendergast, 1993) and by increasing job
satisfaction (Francesconi, 2001).
However, this conventional view of promotions is increasingly being called into ques-
tion, as it is recognized that promotions can also have negative consequences (Baker,
Jensen and Murphy, 1998).
ÅI would like to thank Anna Risch for her invaluable help in managing the dataset. I also thank Stephen Bazen and
everyone who took part in the Careers inAcademia research project for their helpful comments, in particular Christine
Musselin and Fr´ed´erique Pigeyre. The Ministry of Higher Education (Bureau Etudes Pr´evionnelles – DGRH-A1)
is acknowledged for giving access to data. This research was funded by the National Research Agency (ANR –
TRAJUNI Project).
JEL Classification numbers: M51, J24, J45.
Does productivity decline after promotion? 887
Numerous examples of reductions in productivity after promotion can be seen in the
real world, including in universities. For example, the productivity of American univer-
sity professors has been observed to decline after tenure is obtained (Lazear, 2004). In
order to be promoted to full professor, assistant professors must publish more than the
average; hence, the quest for tenure tends to boost productivity. However, this external,
and thus extrinsic, incentive vanishes once tenure, which is generally not reversible, has
been obtained. Conversely, assistant professors who are denied tenure must move to other
universities and publish more. Finally, promotion decisions seem to strongly influence
ex-post effort.
The present study tests whether or not productivity also decreases after being pro-
moted (or not) in the French university system. The French and American systems are very
different in that French professors are recruited as civil servants on permanent contracts.
Therefore, in France, promotions are about obtaining higher pay and extra responsibilities,
rather than gaining tenure. In addition, professors can be viewed as motivated workers,
who are influenced more by intrinsic incentives than by extrinsic ones (Besley and Ghatak,
2005). Consequently, the impact of being promoted or not on effort may be different from
what has been observed in the American system. Furthermore, France is currently experi-
encing a move towards greater evaluation of public-sector workers in all fields, including
the university sector. As a result, the work of academics is coming under increasing scru-
tiny, in terms of both teaching and research, and the Ministry of Education is expected to
introduce wide-ranging reforms in professors’ career structures and promotion procedures
(see Schwartz, 2008). However, the promotion system has never been properly evaluated
– a situation the present study remedies.
In order to carry out this evaluation, an original dataset of publication scores and pro-
motions of university professors in the field of business studies was created. Because
most promotions are awarded on the basis of a centralized promotion procedure, called
the concours d’agrégation, the study focused on this procedure. The database contained
information about promotion decisions (and their explanatory factors), and could be used
to evaluate candidates’ publication records before and after the promotion procedure. For
the purposes of the present study, publications were divided into three categories: arti-
cles in international journals, articles in national journals, and books. The quality of each
publication was also taken into account.
A major problem facing all evaluations of this type is the impossibility of knowing how
productive a person would have been if promotion had not been granted (Rubin, 1974). This
difficulty was addressed by using matching methods to obtain unbiased evaluations of the
impact of promotions (Heckman, Ichimura and Todd, 1997, 1998). The growing literature
on careers in academia was used to calculate propensity scores; that is to say, the probability
of being promoted. Because empirical studies of the determinants of promotion have high-
lighted the effects of gender (Long,Allison and McGinnis, 1993; Ginther and Hayes, 1999,
2003), productivity (McDowell, Singell and Ziliak, 2001) and networks (Combes, Linn-
emer and Visser,2008), all these factors had to be included in order to obtain robust propen-
sity scores. These propensity scores were then used to calculate counterfactual publication
scores and to compare productivity before and after promotion. In addition, a sensitivity
analysis was carried out to test the plausibility of the Conditional Independence Assump-
tion (CIA) that underlies all matching estimates (Ichino, Mealli and Nannicini, 2008).
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2012
To continue reading
Request your trial