Return to
August 2002 Titles
This issue of
the Quarterly exemplifies the variety of legislatures that are the
subject of legislative research today. The first two articles explore aspects of
legislative institutions in central and eastern Europe, the next four deal with
the U.S. Congress, which is still by far the most studied legislature in the
world, and the last two focus on U.S. state legislatures and U.S. city councils.
As legislatures are studied in a growing number of diverse settings, the
prospects of gaining general knowledge of the institution improves.
After a decade of experience with competitively elected parliaments in central
Europe, what evidence is there of the institutionalization of these newly
democratic assemblies? Has a cadre of career politicians emerged, providing
continuity of membership, experience, and commitment to the parliamentary
institution? Goldie Shabad and Kazimierz M. Slomczynski analyze data on all
parliamentary elections that have taken place in Poland and the Czech Republic
to both houses of their parliaments from the last election before the fall of
communism to the end of the 1990s. The enormous data set they have developed
permits them to track all candidates across consecutive elections to determine
the rate at which incumbents seek election, the rate of carryover from one
election to another, and the number of times that candidates at the last
election in the 1990s had competed in previous elections. They find that there
was a sharp discontinuity between the membership of the last communist
parliaments and the newly democratic ones, despite the success of post-communist
parties. But although there have been high rates of electoral volatility and
weak party systems in these countries, parliamentary carryover rates have risen.
An increasing proportion of incumbents seek reelection and prior experience
improves their prospects of success. There is considerable variance across
parties and across countries and the process has been slower than in other
countries at comparable stages of their democratization, yet the authors provide
unmistakable evidence of stabilization in the parliamentary recruitment process
in these two important central European states. As the experienced members play
the most active role in these parliaments, continuity of membership appears to
contribute to the professionalization of these newly democratic legislatures.
Experimentation with
electoral systems in central and eastern Europe has given us an opportunity to
examine the impact of electoral systems on legislative behavior. Mixed member
systems that combine proportional representation with single-member districts
appear to provide a means to compare the effect of these two systems within the
same country. However, as Eric S. Herron points out in his article on
parliamentary elections in Ukraine, dual candidacy, which was allowed in Ukraine
and which is often permitted in mixed-member systems, as well as party influence
on nomination in both the constituencies and the party lists, make it difficult
to distinguish the effect of these two elements of a mixed member system on
legislative behavior. Herron distinguishes instead between candidates elected
from either safe constituencies or safe party list positions, and candidates
elected from more competitive positions. He tests the effect of seat safety in
the 1998 parliamentary election on the propensity of members subsequently to
vote with their party faction on the floor of the House. His conclusion is that
legislative behavior is not influenced by whether a member is elected from a
party list or from a single-member constituency, but that the deciding factor is
whether a candidate is from a safe position in either type of constituency.
Judging by this case study, mixed-member systems, which have been widely adopted
in newly democratic parliaments, may not have had the intended effect of
creating two distinct classes of legislators, those sensitive to local interests
and those to national party interests.
Four articles in this issue present new research on perennial questions related
to the U.S. Congress. The first two deal with elections to the House. Brian
Newman and Charles Ostrom, Jr. note that the outcomes of congressional elections
in the 1990s do not seem to fit previous explanations adequately. They find that
while presidential approval and major political events continue to explain
congressional election results, district level factors have gained in
importance. Critical among these is the distribution of open seats between those
that the president’s party must defend and those that must be defended by the
opposition. Midterm congressional elections are and continue to be referenda on
the president’s performance but the quality of candidates in the most
competitive districts is a newly influential factor.
Martin P. Wattenberg
and Craig Leonard Brians investigate the bias that results from differential
turnout, challenging the conventional wisdom that there is little difference
between the party preferences of voters and non-voters. That conclusion had been
based on highly visible, relatively high turnout presidential elections.
Wattenberg and Brians examine midterm congressional elections between 1978 and
1998, in which turnout is relatively low, and distinguish between the party
preferences of voters and registered non-voters. They note that in both the 1994
and 1998 congressional elections, registered non-voters were disproportionately
Democratic party identifiers. They conclude that when turnout falls below 40% of
registered voters, as it has in recent congressional elections, representation
at the polls is biased.
One of the most important changes in the behavior of the U.S. House of
Representatives in the last 25 years has been the increasing polarization
between the parties. This has usually been explained by the demise of the
conservative southern wing of the Democratic party in the House, and its
replacement by a contingent of conservative Republicans. The article by Mark
D. Brewer, Mack D. Mariani, and Jeffrey M. Stonecash supplements this
explanation with an analysis of the increasing proportion of seats won by
Democrats in the north. In this region of the country the party has been
especially successful in less affluent, urban, minority areas. Spatial
segregation between classes in these areas, income differentiation across
districts, and the growth of districts in which a considerable proportion of the
population is non-white has given the Democratic party a core of ideologically
liberal constituencies. These socio-demographic sources of party differentiation
in the House are of fairly recent origin, and are likely to become even stronger
in the future.
The continuity of membership in the U.S. House is uncommonly great in
comparative perspective. A small proportion of members do retire after each
session, and Samuel H. Fisher III and Rebekah Herrick seek to explain these
retirements. They apply a new, direct measure of members’ job satisfaction
with data form a mail survey of representatives elected after 1970 who have
since retired. They conclude that satisfaction with the job of being a member of
Congress is strongly related to career length in the case of those who retired
voluntarily but not for those who lost their reelection bid. While this is
itself not surprising, the authors are able to specify some of the elements of
job satisfaction, among which the most important is the perception that the work
is meaningful. Whatever the general public may think, financial rewards are
unimportant in keeping members in office.
The last two
articles in this issue deal with subnational legislatures in the United States.
The movement to limit the number of terms that state legislators may serve
reached a peak in the 1990s and seems to be declining. The conventional
explanation is that term limits appealed principally to Republicans both for
ideological reasons and because their party was out of power in Congress and in
many state legislatures. That conventional explanation would predict a decline
in support for term limits after the Republican return to power at the national
level and in many sates. However, using data from the 1994 National Election
Study and from their own telephone survey conducted in 1998, the authors find
that conventional explanations of support for term limits do not hold up.
Support comes instead from voters whose representative in the legislature
belongs to a different party than their own. That would suggest that support for
term limits does not have a general ideological basis in distrust of government,
does not extend to all governmental offices, and may not be a passing fancy.
The final article in this issue deals with racial polarization in U.S. city
councils. Rory Allan Austin studied this subject in six city councils, three in
the North and three in the South, whose electoral systems varied between
district elections, at-large elections, and a mixture of the two. He examined
all non-unanimous council votes in the period between 1987 and 1996, testing the
hypothesis that African-American council members tend to be on the losing side
of votes both because of polarization in the electorate and on issues within the
councils. His surprising finding in the six city councils he studied is that
racial polarization is far less evident than the conventional hypotheses
predict, and that when it occurs it does not appear to be generally correlated
with the electoral system except in specific policy areas, notably housing and
police affairs.
—Gerhard Loewenberg