(0
presidential support. Why shouldn’ t presidents make use of all the tools at their disposal,
especially since peace events would probably be less risky, less costly, and possibly less
damaging if the illegitimate motives became known. One explanation, of course, is that uses of
force during recessions may not be diversionary: the observed correlation could be a
coincidence. Alternatively, there could be a missing factor that is a common cause of both, such
as shared connections to the electoral calendar. The relative effectiveness of peace and war may
also vary under different economic conditions – uses of force might be relatively more popular
when economic conditions are poor, for example. Future research is planned that will investigate
these possibilities.
22
Conclusion
Dramatic presidential peace activity has resulted in increases in presidential approval.
On average, major peace events – those receiving a story per day or more in the New York Times
for two week – give presidents about a 5% increase. This compares to 6% for similarly
publicized uses of force. This confirms Hypothesis 1. In addition, there is a systematic variation
between the effects of difference peace events: more coverage means larger gains, and elite
opinion can have a profound influence on the response. Presidential actions which are heavily
criticized might actually cause the president to lose support. This generally confirms Hypothesis
2, although success did not matter as expected.
Taken together, these results suggest that the media priming model has greater
explanatory power than the conflict-cohesion theory. Similar changes in approval result from
//
:
$ $ 6
$
6
$
$
$
5
$
$
5
$
=
1
$
#$
>1=:
$ 00) >1 !
#
5 6
$
5
5
$
;
5%
5
1=
/++(>
Convention