Return to
February 2003 Titles
The first two articles in this issue examine the impact of legislative
institutions on political behavior in two different non-American settings. In
each case the authors employ theories and concepts derived from research on U.S.
legislatures. Paul Chaisty asks why the Communist party that won a near majority
in the election to the second Russian Duma in 1995 did not reverse the
power-sharing arrangements that the first Duma had established. What explains
this early preference for the institutional status quo? His careful analysis of
voting behavior in the Duma reveals that both the Communists and their potential
allies were not sufficiently disciplined to make use of majority rights.
Instead, the existing power-sharing procedures enabled the Communist faction to
bargain effectively not only with other groups and factions but also with the
executive branch. It turned out that even after the Communist election victory
of 1995, the Duma remained an assembly lacking stable majorities. The
power-sharing procedures it had developed in its first term proved to be
preferable to majoritarian procedures in its second term, and these procedures
have in fact proven tenacious into the Putin era.
Regarding impeachment as a legislative process that can be used to hold
presidents accountable, Naoko Kada examines that process in two Latin American
countries, focusing on the role of investigating committees in determining
impeachment outcomes. The author employs an informational theory of
investigative committees to compare the impeachment of President Fernando Collor
in Brazil in 1992 with the failure to impeach President Ernesto Samper of
Colombia in 1996. While the impeachment processes in the two legislatures are
similar, Kada shows that the investigative committees operated differently. She
explains the differences in the impeachment outcomes in the two countries by the
differences between the two committees’ control over information. In Brazil,
the opposition parties were able to use information strategically to mobilize
the supporters of impeachment, while in Colombia the majority was able to
control information flow to protect the president. Kada shows that the
differences between information control in the two cases is the result of how
committee members are chosen, the distribution of investigative authority within
them, and the selection of the principal investigator.
In the stable American two-party system, it is rare for a member of Congress to
switch parties while in office, although there is a body of research on the
reasons why a few members do switch and how their voting behavior changes after
their switch. Christian R. Grose and Antoine Yoshinaka fill a gap in research on
this subject by investigating how party switching affects an incumbent’s
subsequent electoral support both in primaries and in general elections. They
use data on the electoral consequences for 25 members of the U.S. House and
Senate who switched parties while in office between 1947 and 2000. The authors
test the models of party affiliation developed by Aldrich and Bianco, which
predict that members’ choice of party is based on a calculation that it will
bring them electoral success. They find that switching lowers an incumbent’s
electoral margin in the primaries but that this disadvantage diminishes over
time. Switching also lowers the margin that incumbents had in the general
election and does so for all subsequent elections. This leaves open the question
of whether switching occurs as a result of a miscalculation of its consequences
or is an attempt to forestall a loss of popularity that has other reasons.
A different aspect of the relationship between political elites and the
electorate is the subject of the article by John Frendreis, Alan R. Gitelson,
Shannon Jenkins, and Douglas D. Roscoe. They are interested in determining why
candidates’ issue positions do not converge on the position of the median
voter, as Downs’ original model of electoral competition predicts. To that end
they examine the relative influence of voters and elites on the issue positions
of candidates in eight U.S. states in the 1994 and 1996 elections to the two
houses of the state legislature. Their data consist of the issue positions of
county party chairs and major interest groups, the positions of legislative
candidates in eight states during the 1994 and 1996 elections, as well as the
demographic characteristics of their constituencies. They demonstrate what the
extended Downsian model has assumed, that candidates’ positions are influenced
by party and interest group elites, whose positions are more extreme than the
positions of the median voter and of the candidates. This is understandable in
view of the poor level of information that voters have about candidates’ issue
positions, and the reliance of candidates on the resources that party and
interest group elites command.
A distinguishing characteristic of U.S. legislative elections is the use of primary elections for the nomination of candidates. Research on primary elections is, however, relatively rare. Robert E. Hogan examines how competitive these elections are, with data for elections to the lower house of the state legislature in 25 states in the elections of 1994 and 1996. The factors that determine competitiveness are of course different in primaries, where voters have no party cues, than they are in general elections. Hogan finds that only one-fifth of all primary elections are contested. The level of competitiveness in contested elections is influenced by how winnable a contest seems and by the level of professionalism of the legislature. This interesting conclusion shows that while professionalism may help to protect incumbents from competition, as has often been shown, it may also stimulate competition for open seats by making a legislative seats more attractive.
—Gerhard Loewenberg