illegal and terrible news for parties since it reduces them to serving as financial conduits
(Krasno and Sorauf 2003). Parties may, on the other hand, have welcomed these two
races exactly because they gave them the excuse to advertise in the two largest media
markets in the country to spread the good news about themselves or, more likely, the bad
news about the other side. The problem with this is that the ads that the parties broadcast
in those races show no sign of being anything but the standard candidate-centered fare.
Most important, none of them even mentioned the party that sponsored them, except in
the barely discernable disclaimer that flashed onscreen near the end of each spot.
In the end, I believe that the most plausible interpretation of these findings is also
the most straightforward, namely that the parties respond to one another. There is
evidence that candidates do the same thing, adjusting their media buys as their
opponent’s advertising picks up volume. What is different about parties is the degree to
which they appear to follow one another (the coefficient is nearly twice as large as is
candidates’) and their seeming lack of attention to a host of other factors that affect
candidates (from the closeness of the race to the structure of the media market). In fact,
the results in Table 5 suggest that the other party is practically the only thing that parties
consider when they allocate their media dollars. That is an exaggeration – someone had
to first in any case – but it nonetheless suggests these political institutions may be more
than a bit myopic, focusing narrowly on each other. It is an image wholly consistent with
Shea’s (1995) account of legislative campaign committees in the states as highly
centralized operations, lacking sensitivity to local concerns and relations with the parties’
grassroots activists. While some of the advertisements captured by the CMAG system
were officially sponsored by state party organizations, the money for them and the
19
Convention