is strongest (and positive) for players with an intermediate nomination chance, which is
consistent with our nomination contest theory. Moreover, our results also lend support
to what we call the ‘injury theory’: for several output measures, the effect of nomination
contest participation is negative for players with an expected nomination probability close
to one. This makes sense given that professional soccer carries a high injury risk, which
can be reduced by a more passive style of play. This effect should be strongest for top
players, for whom Cup participation is basically a certainty but conditional on avoiding
an injury prior to the Cup. For players with no past national team exposure, we find no
significant impact of the Euro Cup treatment on performance in club matches.
Next, we test for differential effects of the Euro Cup treatment by age, which we in-
terpret as a measure of the accuracy of information national team coaches have about a
player’s skill. Our nomination contest theory predicts that very young players (possibly
younger than any player in the League) should exert less effort than somewhat older play-
ers. Beyond a certain age threshold, however, effort to win the nomination contest should
be lower the older a player. There is another reason to believe that older players will
have lower incentives: the job market aspect of a Euro Cup should be more important for
younger players who are less well-known and have longer careers ahead of them. Again us-
ing a differences-in-differences-in-differences analysis, we find evidence that participation
in a nomination contest affected younger players more strongly than older players. This
is compatible with the nomination contest theory for sufficiently high levels of precision,
but also with predictions based on post Euro Cup career concerns.
Related Literature On the theoretical side, our paper introduces signal jamming as
in Holmstrom’s (1982) career concerns model in the theory of rank-order tournaments of
Lazear and Rosen (1981). Hoffler and Sliwka’s (2003) basic model is similar to ours, but
we allow for more asymmetries between agents and derive several additional comparative
statics results that we then test empirically.
The theoretical predictions concerning the relation between anticipated winning prob-
abilities and effort incentives would be the same in tournaments, where relative observed
performances instead of perceived abilities determine payoffs. The empirical literature
on tournaments focuses largely on the impact of higher prizes or prize differentials how-
ever; see Ehrenberg and Bognanno (1990) as well as Orszag (1994) for evidence from
golf tournaments, Becker and Huselid (1992) for evidence from auto racing, Knoeber and
Thurman (1995) for an examination of the impact of tournament-style contracts in the