The name is absent

but the ‘injury theory’ a negative δ1 .

The covariates natteamit and natteamit (1 - natteamit) also enter the regression equa-
tions interacted with the variable euro_n , indicating whether n is a Euro 2008 participant,
and after_t, which indicates the time period after the official qualification date (November
21, 2007). These interactions are introduced to make sure that the estimates of δ1 and δ2
capture the effect of national team exposure in combination with Cup qualification rather
than differential effects for different nationalities or in different time periods.

We see two main reasons why natteamit may affect player performance even in the
absence of concerns related to participation in an international Cup. On the one hand,
playing for the national team may wear out a player, but on the other hand, the national
team may provide a player with additional training or experience useful for his club
matches. When using natteamit as an explanatory variable, however, we face the following
potential problem: although we control for time-invariant quality differences between
players, both natteam_it and Y_it may be driven by unobservable variation in i’s ability
(temporary ups and downs in his physical shape for example). What makes the problem
less severe conceptually is that natteamit is based on past observations, while Yit depends
only on the current shape of the player. The results in the next section indicate that we
have no such simultaneity problem: the estimated coefficient of natteam_it turns out to be
insignificant in all regressions.

Another prediction of the nomination contest theory is that the precision about a
player’s ability should affect his effort to win the contest. Roughly speaking, the incentive
effect should be low for both very unknown players and well-known players, but high for
players about whom the national trainer has some yet limited information (see Figure 2
for this in an example). However, we would expect a decreasing relationship if there is
sufficient information about all professional soccer players so that the area of precision
close to zero in the theory can be ignored. In addition, post Euro Cup career concerns
should be stronger for less known players, which implies that those players should have a
higher incentive to make it into the national team.

We use age as a proxy for the precision of beliefs about a player’s ability. A possible
alternative measure would be seniority, a good measure of which would probably be the
number of years a player has played in one of the top soccer leagues (essentially the Italian,
English, Spanish, French, and German leagues).³⁰ We decided to use age instead, since
³⁰ Using a player’s seniority in the German league alone would not be very meaningful, since players
often move between different leagues, and the number of years played in the German league does not

25

<<   28   29   30   31   32   33   34   >>

More intriguing information