Upload
albert-garriga
View
215
Download
0
Embed Size (px)
Citation preview
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 1/14
University of Utah
Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003: AFederal PerspectiveAuthor(s): Robert D. Brown and John M. BruceSource: Political Research Quarterly, Vol. 61, No. 4 (Dec., 2008), pp. 585-597Published by: Sage Publications, Inc. on behalf of the University of Utah
Stable URL: http://www.jstor.org/stable/20299762 .
Accessed: 30/10/2013 16:41
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp
.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
Sage Publications, Inc. and University of Utah are collaborating with JSTOR to digitize, preserve and extend
access to Political Research Quarterly.
http://www.jstor.org
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 2/14
Partisan-Ideological Divergence and
Changing Party Fortunes in the
States, 1968-2003
A Federal Perspective
Robert D. Brown
John M. Brace
University ofMississippi
We expand on previous literature on party competition in the American states by examining competition for both state
and national offices. We find significant differences inDemocratic Party electoral advantage within states, along with
across-state variation in changes in these differences over time. We attribute these results to movement in the partisanship and ideology of the states' citizenry. As consistency in these core political attitudes increases, parties are able
to campaign and govern on messages that are more consistent across electoral levels. The result is greater consistency
in party electoral performanceacross state and national offices.
Political Research Quarterly
Volume 61 Number 4
December 2008 585-597
? 2008 University of Utah
10.1177/1065912908317795
http://prq.sagepub.comhosted at
http ://online. sagepub.com
Keywords: party competition; ideology; partisanship; federalism
S
chattschneider's famous assertion that modem
democracy is unthinkable save in terms of the par
ties (1942, 1) rests on an implicit foundation of com
petitive parties. Appropriate to its significance, the
competitive nature of party systems has been subjectto a great deal of scholarly scrutiny. In the United
States context, work on party competition has pro
ceeded in all electoral arenas, following the contours
of the constitutional framework into examinations of
competition at federal, state, and local levels. Yet,
while commonly recognizing the often decentralized
nature of U.S. political parties, as well as the fact that
partisan elections take place within the structure of
federalism, surprisingly few analyses of party compe
tition incorporate this important contextual feature
into their analytical framework. Rather, most efforts
tend to focus overwhelmingly on party competition at
one level, either national or state, to the exclusion of
the other.1 The result is that despite a large body of lit
erature covering a significant variety of topics, state
and national electoral competition are rarely discussed
within the context of the same electoral setting.
We believe this to be a significant oversight in the
literature examining the nature of party competition
in American politics. While there is overwhelming
scholarly agreement that political parties have been
shaped by the opportunities and constraints created
by federalism, precious little work has been done to
relate these realities to what is arguably the single
most important aspect of political parties: their abil
ity to contest elections at each electoral level.
From our perspective, a key step in understanding
party competition in a federal context is to acknowl
edge the variation in political character of the states.
Even a cursory scan of the states reveals a range of
political profiles. State parties will respond to their
environment, developing messages and images that
maximize their competitiveness in that state. Their abil
ity to adjust in this fashion stems from the organiza
tion's operational boundary matching the political
boundaries of the state. In contrast, the national parties
have little ability to adjust their message or image to
match any particular region or state. Indeed, the mod
ern information age seems likely to create a very
homogenous national party image across the states.
However, the nature of how that image is received in
any particular state is going to be a function of the polit
ical character of that state. The greater the disparity
Robert D. Brown, Professor of Political Science, University of
Mississippi; e-mail: [email protected]
John M. Bruce, Associate Professor of Political Science, University
of Mississippi; e-mail: [email protected]
585
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 3/14
586 Political Research Quarterly
between the perceptions of the state and national party
(Democratic or Republican), the greater the possibility of
significant differences inparty advantage and party com
petition. Thus, we are describing not just competition
among fifty state Democratic and
Republican parties,but also among fifty potentially different perceptions of
the national Democratic and Republican parties.
In previous work (Brown and Bruce 2002), we doc
umented differences in party competition within
states, offering measures of partisan electoral compe
tition and advantage that shed light on both national
and state competition within and across the states.2 In
this article, we expand on that initial effort and exam
ine within-state differences in party advantage for
national and state offices over time (1968-2003). Our
data indicate significant change in the relative elec
toral advantage of the parties during this time period
and, importantly, variation in advantage across levels
of elections within the same general electoral context.
In addition, we offer an explanation for these changes
and differences, focusing on the shifting partisan and ide
ological environments of the states. Drawing on work
from Erikson, Wright, and Mclver, we focus primarilyon the degree of congruence or divergence in the partisan
and ideological proclivities of state electorates. In
Statehouse Democracy (1993), these authors laid out an
argument for how ideology and partisanship may factor
in voting for various levels of office. The choices offeredto the voters in a given state by the parties are already
shaped by the ideological nature of the state, with both
partiesmoving left inamore liberalstate,and to therightin a more conservative state. Erikson, Wright, and
Mclver find that ideology fails to exert an influence in
their model of state legislative balance because the par
ties have already adapted to the states' environments. As
such, we should expect party to play amore prominent
role in election outcomes at the state level. In contrast,
they argue that the national parties lack such leeway for
adjustment.
Given that constraint, Erikson,Wright
and
Mclver expect (and ind) thatideologydoes play a largerrole in presidential elections.
While our data and research question differ from
that of Erikson,Wright, andMclver (1993), the basics
of the relationship they identify are easily applied to
our work. Our thinking here is straightforward. We
argue that consistency among state electorates in
their partisan and ideological orientations should lead
the state parties (more likely to be evaluated along
party lines) to fare about as well as their national
counterparts (who are more likely to be evaluated
along ideological lines). Conversely, when state elec
torates are decidedly more liberal or conservative
than their declared partisanship, we should to see some
impact of these differential criteria for evaluation.
This greater heterogeneity in the partisan-ideological
nature of the state environment makes itmore likely
that a
party
will
experience
better electoral fortunes
at one electoral level (e.g.: state vs. national) than the
other. Thus, as the partisan and ideological nature of
the states change over time, we should expect to see
changes in the relative ability of the parties to contest
elections for both state and national offices. Importantly,
improvement in the ability to contest elections at one
level, be it state or national, does not necessarily con
tribute to similar advantages at the other level.
We proceedwith abrief discussion of the literature n
these areas and then move on to an examination of mea
surement strategies. From there, we highlight state-level
changes in partisan electoral advantage as they exist for
both state and national offices. Finally, we offer some
analyses that we believe lend insight into the nature of
these changes, as they vary across space and time.
Integrating State and National Party
Competition/Advantage in the States
Scholars in the subfield of U.S. state politics have
long been interested in the potential importance of par
tisan competition. From examinations of differencesand changes in state party competition (Patterson and
Caldeira 1983;Barrilleaux 1986), to the role of party
competition in explaining state policy outputs,3 to
issues of measurement (Holbrook and Van Dunk 1993),
the result has been a body of research that has con
tributed a great deal to our understanding of the role of
political parties in democratic governance in die states.
Despite the volume and value of the present state of
research on state-level party competition, little atten
tion has been given to cross-state variation in com
petition
for national offices. Studies of
presidentialelections often give nodding acknowledgment to these
contests as state-level phenomena, but with very few
exceptions (e.g., Rabinowitz and Macdonald 1986;
Nardulli 1995), rarely are the state electoral settings
included as important factors in understanding partisan
electoral advantage in presidential and other national
contests. Similarly, with the exception of those that
examine partisan change in a regional setting, studies
of realignment tend to treat this phenomenon as mono
lithic, proceeding at the same pace and intensity
through the nation (or region) as a whole.4
The lack of any significant integration of state and
national competition/advantage in the literature is
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 4/14
unfortunate. While the U.S. is frequently described as
having a two party system, each of those parties is
composed of any number of national, state, and local
party committees (Beck 1997). These layers of party
structure have a reasonable degree of autonomy and
legal independence. It seems very likely, therefore,
that the fate of the party in a given state may differ in
state and national political contests. One need only
consider recent history in many southern states to
find such a situation. In any number of these states
there was a significant disparity between the success
of the Democratic Party in state elections and that of
the Democratic presidential nominee.
More generally, though, we see ample reason to
believe that levels of partisan competition/advantage
for national offices and state offices deserve compar
ison at the analytical and political level of the state.Scholars have long extolled the fact that the federal
constitutional structure impacts a broad variety of
phenomena, and clearly, the partisan underpinnings
of electoral competition are no exception.5 As Jewell
and Morehouse note, the basic flavor of party com
petition in each state is itself a product of national
political forces as well as factors unique to the state
(Jewell andMorehouse 2000). The end result is that
partisan competition at the state and national levels
are related within the electoral context of the states.
To be sure, the extent of this connection will vary across
the state electoral settings depending on responses to
historical, political, and social forces. We view this as
another important reason to examine the confluence
between national and state electoral competition in
the states. Moreover, we view this as an important
reason to add a temporal component to these analyses. Not only are the state partisan environments
subject to different forces across levels of electoral
contest, but they are also likely to respond to these
forces in different ways and at different speeds. In
short, we believe that a full understanding of electoral
advantage in the states requires an explicit recogni
tion of the dual nature of party advantage as it varies
both within and across states, as well as over time.
Measuring Party Electoral
Advantage in a Federal Setting
To examine these potential relationships we first
need a measure of electoral advantage that has utility
for both state and national offices. Not surprisingly,
themeasurement
of party competition/advantage isnot without its conceptual difficulties and nuances.
Traditionally conceptualized as party dominance in
Brown, Bruce / Partisan-Ideological Divergence 587
government (Ranney 1971), our measurement strat
egy follows along more recent efforts that emphasize
actual competition for elected office (Holbrook and
Van Dunk 1993). Our main issue with party domi
nance measures is that they are, at their core, outcome
based, focusing on whether one party or the other
controls the branches of government. While certainly
a reasonable way to conceptualize competition,
outcome-based measures suffer from an inability to
distinguish between a state where one party domi
nates absolutely (i.e. there is no significant electoral
threat from the opposing party) and one where both
parties are viable but one is still winning more often
(i.e., close contests for most offices that tend to run in
favor of one party).6 We think a better approach is to
focus the degree of electoral pressure in the party sys
tem (Holbrook andVan Dunk 1993). That is, for anygiven election in a state, what is the degree that both
candidates have ameaningful chance to win the elec
tion? To what extent do elected officials feel pressure
from the electorate?
Following this distinction, Holbrook and Van
Dunk (hereafter, HVD) offer an aggregate measure of
state partisan electoral competition based on several
constituency-level indicators: the percentage of the
popular vote for the winning candidate, the margin of
victory, whether or not the seat is safe, and whether
the race was contested or not.
By using
electoral
results, this measure is much more likely to reflect
concern about electoral viability and pressure from
opponents, and thus, we believe this measure repre
sents a significant conceptual improvement.7
Yet, like Ranney, this measure is not without its lim
itations. Focusing solely on contests for state House and
Senate, one significant issue is the measure's lack of a
gubernatorial component. To the degree that Ranney
suffers for undervaluing the role of gubernatorial con
tests (King 1989), the HVD measure suffers even more
for excluding it completely. It just does not seem rea
sonable to us that a measure of state electoral competi
tion can exclude all partisan voting for the state's most
visible and (generally) hotly contested office.
Additionally, and very important for our purposes,
the components of the HVD measure make it difficult
to use when there are a small number of cases within a
category. Application of these criteria in a single race
for governor, for example, may make the safe com
ponent go to one hundred. For our purposes here?
examining both national and state party competition
within the states?this is a decided disadvantage, as the
measuredoes not work well beyond the realm of the
lower national House (and even there does not work
well in small states).
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 5/14
588 Political Research Quarterly
To examine party competition in a federal setting,we need a measure that can function within a federal
context that can be applied at both levels within the
states. Such ameasure would benefit from keeping the
best aspects of both Ranney and HVD while down
playing their more important limitations. In our view,this means retaining the focus on partisan electoral
competition offered by HVD, but including an execu
tive component in the calculation (Ranney). Ideally,
this measure would also provide both directional and
overall competition analyses, as in Ranney. Finally, it
must work well within the context of a small number
of cases to be portable to the national level.
Elsewhere (Brown and Bruce 2002), we put forth a
competition measure that is actually a simplification of
HVD (1993), but one that provides more information
about thecompetitive
nature of
parties
at different
levels of elections. We start with one component of the
HVD measure (margin of victory) andmodify it to
facilitate directionality (i.e., Democratic vs. Republican
advantage), the inclusion of the executive office in the
calculation, and greater portability to national party
competition in the states. Like HVD, we use vote
margin as the foundation of our measure because
we believe this taps most directly into the conceptual
underpinnings of partisan electoral competition?
whether or not elected officials feel pressure from the
electorate, and whether those running for office have a
reasonable chance of winning.8
Rather than using the winning candidate's margin
of victory, we first calculate vote margin so as to allow
for directionality in partisan competition. Specifically,
we use the Democratic Party candidate's vote share as
an anchor, and calculate margin of victory differently,
depending on whether the Democrat won or lost. If
the Democrat won, then the vote margin is simply: %
Democratic vote -% second place vote share. If the
Democratic candidate lost, we modify the calculation
accordingly: % Democratic vote-
% winner's vote
share.9 The result is a measure of partisan electoral
competition that focuses first on Democratic Party
advantage. The measure has a theoretical range of-100
(total Democratic failure) to 100 (total Democratic
domination).10
This information is consistent with the directional
information on party advantage provided by Ranney,
but with the constituency-level focus advocated by
HVD.11 This strategy also allows us to include an exec
utive component, meaning we can tap the most visible
state and national offices (and those that receive the
highestturnout),
adding
to the measure's face
validity.Most importantly, our approach allows us to cal
culate two sets of competition measures: one for state
level competition and one for national competition in
the states.
After settling on vote share as the indicator that is
both theoretically most appropriate and portable across
electoral levels, we proceed to constructing the actual
measure. We first determine the component parts basedon Democratic margin of victory or defeat in each con
stituency (state house, state senate, governor; U.S.
House, U.S. Senate, president) for eight year cycles
beginning in 1968 (the first year for which data are
available).12 This leaves us with six measures of con
stituency level competition for the eight-year period?one for each state and national constituency. The
margin for each separate constituency was then aver
aged by office (house, senate, executive) within states
and across years. Consistent with HVD, all initial cal
culations are made at theconstituency
level and then
aggregated to a state level average for each office.
These averages were then combined to form our base
line measures of state party advantage (state legislative
chambers and governor) and national party advantage
(U.S. legislative chambers and the presidential vote in
the state).13 We weight the three arenas of electoral
competition equally in our measure.14
To examine trends in party advantage, we also add
a temporal component to the measure. There are sev
eral methods by which this can be done.15 The first,
and perhaps most simple, would be to pick several
time periods, calculate the measures for each, and
examine the results. We find this to be a less than sat
isfactory approach, in that the arbitrary selection of
time points can force certain possible interpretations
to be more or less plausible. The second approach,
and the one followed here, is to create amoving aver
age type of measure. We simply take an eight-year
slice for one period, shift the time frame by two years,
take another slice for the next period, and repeat.
Under this approach, each point in time is included in
four different eight-year periods. The data used in this
paper cover fifteen such eight-year periods, startingwith 1968 to 1975 and ending with 1996 to 2003.16
State Democratic Party Advantage
for State and National Office
In this section, we examine broad patterns in our
Democratic Party advantage measure. We know that
individual-level partisan attachments vary and that par
ties are fluid entities, free to morph into any position
seen as most advantageous in terms of contesting office.
We also know that the political environment of each
state is influenced both by broad, national trends, as well
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 6/14
Brown, Bruce / Partisan-Ideological Divergence 589
Figure 1
State and National Democratic Advantage, by State (1968-2003)
Arizona
* ?^
8h>*t?7*v?r ***? ****
Oeorg?a
^?LlLfe*^-*j*u o ^ |A^~
:; ;*?*
f
Kentucky
Missouri
*:h*.
?*ut;v
*T7*T**^*'
?1 * *.
Soi?hCarolina
**?
?
^**** ,*;*% ******
Michigan
NewHampshire isle? ersey
**$
pJ tf*
H,j?:_?:.v_ i
Virginia Washington
...,. , ? i~,.-'-y'' '', ,t,.,,,,.,', ? T7~'rr7?'7'rr'''7'7'7 'T'7ii ... . *,*
West Virginia
?i **Wyoming
* Democratic Advantage-
State Democratic Advantage-
National
as by its own unique political, social, and economic sit
uation. Thus, we find it reasonable to expect change
over time in terms of party advantage. The interesting
question involves the nature of that change and how it
varies across both electoral levels and the states.
In figure 1,we show party advantage calculated for
state and national contests, with each state presented
separately. We recognize that the amount of data pre
sented is daunting, but we offer this first glimpse to
facilitate initial observations. Across the x-axis is time,
where the unit is the eight year periods discussed pre
viously. The y-axis represents is the party advantage
scale.17 Recall that 0 represents the point of parity
between the parties, with positive numbers indicating
an increasingly Democratic state and negative
numbers showing amore Republican state. Each graph
has two lines plotted, one for the state advantage mea
sure and one for the national advantage measure.
Across the states, the two dimensions of interest are
clearly apparent. First, it is obvious that, within most
states, the welfare of the parties varies across the level
of electoral conflict. There is no state where the mea
suresof party advantage for state and national offices
are essentially equal. There are cases where the two
lines may trend together, but they maintain different
fundamental values. Important for our purposes, there
is clear cross-state variation here, with state and
national Democratic party advantage being far more
similar inUtah and Tennessee, for example, than New
Mexico and Maryland. Thus, at least on the surface of
things, our basic assumption holds true: within states
the relative balance of party advantage is different for
state and national offices. What is interesting to us at
present is that the magnitude of these within-state dif
ferences in party advantage varies considerably across
the different state political environments.
In addition, we see clear trends in these measures
over time, as well as regional differences in these
trends. In all the southern states, both measures are
clearly trending in an increasingly Republican way.
This is hardly a surprise. Alabama, Georgia, South
Carolina, and Texas are clear cases of this pattern.
Note also that in most southern states, the loss of
Democratic advantage at the state level far exceeds
their loss at the national level. Again, this is not nec
essarily surprising given well-documented changes in
the formerly one-party South during this period, butwe find these patterns instructive nonetheless.
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 7/14
590 Political Research Quarterly
Figure 2
National and State Democratic Electoral Advantage, 1968-2003
3
LU
Democratic Advantage- State Democratic Advantage
- National
Still, we see that the South is not monolithic. In
Arkansas, for example, we see the general trend to the
GOP, but we still note significant separation between
the patterns for state and national office. Democratic
fortunes have declined both nationally and at the state
level, yet Democrats still do very well when running
for state office. While a border state, and thus not tech
nically southern, Kentucky exhibits a similar pattern.
In the nonsouthern states, the trend lines are far
less consistent. Some states have diverging lines
(Connecticut, New York), with Democrats doing better
at the national level than the state. New Hampshire is
particularly interesting here, showing substantial
swings in the level of party advantage. Other states
are shifting increasingly Democratic at both state and
national levels (Massachusetts, California). Again,
while the overall tendency is for state and national
Democratic Party advantage to track together, this
brief summary illustrates the substantial within-state
differences in these measures as well as the cross
state variation that we expected.
While the within-state trends are our primary focus,
the broaderaggregate
trends over time are alsointrigu
ing. Even though every office we consider is built
around the state or district as the electoral unit, the over
all pattern should offer some insight into the compara
tive electoral welfare of the parties. For this reason, we
offer figure 2, which shows themean values of the state
and national party advantage measures for each time
period, and is based on all forty-eight states for which
we have data.18 The two trend lines stand in rather stark
contrast. The national measure moves about the parity
line, crossing three times and ending up with the most
pro-Republican outcome in the entire series. The state
advantage line, however, shows an almost monotoni
cally decreasing level of pro-Democratic advantage.
While there has been an ebb and flow in the national
balance, the aggregate results for state returns suggest a
dramatic decline in the electoral advantage of the
Democratic Party in state elections.
To isolate regional differences in these patterns,
we replicated the previous figure, breaking the states
into South and non-South (see figure 3). Clearly, the
biggest changes have occurred for state offices in the
formerly one-party South. From a high of forty in
1968-1975 (meaning the Democrats won with an
average marginof
victoryof almost
forty points),state Democratic electoral advantage in the South has
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 8/14
Brown, Bruce / Partisan-Ideological Divergence 591
Figure 3
Democratic Party Advantage, by Region
i?i?i?i?i?i?i?\?i?i?i?i?i?r?r68-7570-772-79 48176-8378-850-87 2-89 49186-93 8-95 0-97 2-999401 6-03
i?i?\?i?i?i?i?i?i?i?i?\?i?ree-7570-772-79748176-8378-850-87 2-898491 6-9388-9590-9792994019&03
Time Frame (moving average)
State -South ?*? State -NonSouth -?4>? National -South National -NonSouth
dropped to almost parity over the thirty-five-year
time frame. While party changes in the South have
been well-documented, the visual evidence provided
by our data is still quite striking.
While obviously less dramatic, we also see interest
ing patterns for state Democratic Party advantage in the
nonsouthern states. Here, state party advantage starts
Democratic and is relatively static before dropping sig
nificantly. It is interesting to note that at the end point
of our data, Republicans carry a slight overall advan
tage in the nonsouthern states on this measure.
Turning to patterns in party advantage at the national
level, we note that while less pronounced than for state
offices, we still see substantial movement over the time
frame examined here. Among southern states, national
Democratic advantage starts off almost ten points in the
Democrats favor, then almost doubles, before proceed
ing to approximately a ten-point Republican advantage
by the last time period. We also see significant move
ment in national Democratic advantage in the non
southern states. In the aggregate, the line ebbs and
flows around theparity line, dropping during
the
Reagan/Bush era, and then rising and eventually dropping
again during the Clinton/Bush years. Interestingly, the
southern and nonsouthern national trend lines tend to
track together until around the tenth period (1986-1993),
after which the southern line moves dramatically into
the Republican realm.
Figure 3 offers simple, yet compelling evidence of
dramatic movement toward the Republican Party that
has permeated both state and national offices. Clearly
these trends are strongest for state offices in the South,
if for no other reason than these offices had the fur
thest road to travel toward competition. By the end of
the period under consideration here, this outlier has
moved into close proximity with the performance of
the other measures. At the same time, however, we
should not discount the level of change in the non
southern states. While not as dramatic as the South
(which reflects as much as anything else the change
from one party domination to competition), our data
reveal that region is not the only force driving overall
movement in Democratic electoral advantage. As a
set, the four lines drift in a Republican direction, but
in a manner that is notnearly
aspronounced
as the
shift in the advantage held by the Democratic Party in
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 9/14
592 Political Research Quarterly
state contests. At the end of the time frame, we see evi
dence that corresponds very nicely to the current era
of American electoral politics, an era marked by close
divisions and increased competition at all levels. In the
remainder of the paper, we examine the impact that
partisan-ideological divergence in state electorates
has on these patterns of partisan electoral advantage.
State Partisanship, Ideology, and the
Changing Democratic Party Advantage
To gain insight into these empirical patterns, we
examine the ideological and partisan foundations of
the state electorates. As noted earlier, our focus is
on the degree of consistency in these two attitudes
within the given state electorates. We argue that, all
things being equal, the nature of national contests is
more likely to generate the ideological messages
necessary to animate ideology as a voting force.19
Presidential contests are increasingly couched in ide
ological terms, be it liberal Democrats, or family
values as a surrogate for conservative policies and
practices. Congressional elections are increasingly
fought on ideological terms as well. As such, we
expect that the ideological leanings of state elec
torates will be an important force in determining
these national outcomes for each state electorate. Incontrast, we see partisanship
as working more promi
nently in determining state level electoral outcomes.
The relatively low information nature of these con
tests reduces their ideological tone, making it far
more likely that partisanship will guide the voting
process. Earlier work by Erikson, Wright, and Mclver
underscores these observations, finding that while
party and ideology are indeed linked (with ideology
driving partisan attitudes), party plays a stronger role
at the state level and ideology is more prominent at
the national level.
With regard to how state partisanship and ideology
account for the shifting patterns we observe in
Democratic Party advantage, more recent work by
Wright et al. (n.d.) indicates that patterns of state par
tisanship are changing as well, with state partisanship
and ideology becoming more congruent over time.
Interestingly, the patterns in these new data indicate
that state ideology had remained remarkably stable
during the 1977-2003 period they examine, with
state partisanship moving considerably.20
We note here that we are not saying that ideology
has no role in state level contests, nor are we sug
gesting that, with state ideology relatively stable, that
party is the only mechanism for change here. Rather
we follow Wright et al. in their conclusion that the
partisan change they note in state electorates is being
driven by ideology. Our belief is that to the degreethat state electorates diverge in terms of ideology and
partisanship, it is because partisanship is inconsistent
with ideology, and the movement away from these
long held ties is difficult and slow. To the degree that
we are seeing convergence in public opinion and par
tisanship, it is because ideology is driving partisan
ship, slowly and inexorably, away from its moorings
and toward a partisan orientation that ismore consis
tent with ideological foundations.
It is this combination of observations that drive
our hypothesis about differential patterns of electoral
behavior for state and national offices. To the degree
that ideology and partisanship serve different purposes
in evaluating national and state candidates, we expect
that differences in the partisan and ideological attitudes
of citizens within any given state will be related to the
differences we observe in party electoral advantage.
That is, when citizens are far more Democratic than
they are liberal, we should expect to see greater levels
of support for Democratic Party candidates at the state
level than we do at the national level. With regard to
the longitudinal patterns we observe, we follow the
same line of thinking. As overall state ideology and
partisanship becomemore
in line (that is, less divergent), we should expect to see this play out with more
consistent voting across electoral levels.
To explore this further, we calculate a measure of
state partisan-ideological divergence.21 Taken from
updated measures of the original Erikson, Wright,
and Mclver (1993) study, we calculate divergence as
simply the difference between state partisanship and
ideology in each state.22 State partisanship ismeasured
as themean of individual responses to CBS/New York
Times survey questions asking respondents if they
consider themselves, Democrat (scored 1), Independent
(scored 0), or Republican (scored -1). Similarly,
responses to the ideology question are scored: Liberal
(1), Moderate (0), or Conservative (-1). State means
are calculated on these attitudes for each state and for
each of the time frames examined in the previous dis
cussion of patterns in party advantage. It is from
these state means that we derive divergence, simply
by subtracting state ideology from state partisanship.
Our thinking here is straightforward: to the degree
that state electorates are more Democratic than they
are liberal (which is the case in every state), we should
expect to see greater differences in voting patterns forstate and national office. In states where we observe
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 10/14
Brown, Bruce / Partisan-Ideological Divergence 593
Figure 4
Democratic Advantage Differential and Partisan-Ideological Divergence
h*N g
i?i?i?i?i?i?i?i?r76-83 78-85 80-87 82-89 84-91 86-93 88-95 90-97 92-99 94-01 96-03
State-NationalAdvantageDifferential Partisan-IdeologicalDivergence
Figure 4b: South
i
<^
~i-1 i r76-83 78-8580-8782-8S
i?i?i?i?i-r~B-9388-95 90-97 92-99 94-01 96-03
Figure 4c : Non-South
,-: .
- i?i?i?i?i?i?i?i?i?i?r~76-83 78-8S 0-8782-8984-9186-93 88-95 90-9792-99 94-01 96-03
State-NationalAdvantage Differential Partisan-IdeologicalDivergence State-NationalAdvantage Differential 4> Partisan-IdeologicalDivergence
the converging of these attitudes over time, we should
expect greater balance in the party fortunes at each
electoral level.
We begin our analysis by overlaying this divergence
measure on a measure tapping state and national dif
ference inDemocratic Party advantage, which we cal
culate simply by subtracting national advantage from
state advantage. This also allows us tomatch the con
cepts more appropriately, as the party advantage dif
ference measure (state advantage-national advantage)
is constructed so that positive scores represent a state
that ismore Democratic in state races than in national
races. We show these comparisons in figure 4, which
illustrates rather clearly the relationship between
partisan-ideological divergence and within-state dif
ferences in state and national electoral outcomes.23
The first portion of the figure (4a) shows the rela
tionship between partisan-ideological divergence and
state and national differences inDemocratic electoral
advantage for all states. The relationship between the
two variables is readily apparent, and the overall cor
relation of .664 is nosurprise.
In theearly
time
frames, differences inDemocratic advantage for state
and national level offices are quite high, averaging
between ten to almost fifteen points for all states. At
the same time, we see state electorates at their most
divergent in terms of partisanship and ideology. As we
move forward in time, we see the measures move
together, dropping throughout the 1980s with a slight
resurgence in the 1990s.
Figure 4b and figure 4c repeat these patterns, split
ting the data by region.24 For the southern states, we
see the strongest evidence yet of a relationship between
divergence differences inDemocratic advantage (r = .84).
This is hardly a surprise, as we expected the relation
ship to be strongest in the South, where the difference
in Democratic prowess in state and national elections
is seen most prominently, and where partisan and ide
ological attitudes have historically been most distinct.
Yet, figure 4c shows quite clearly that the relationshipwe are positing cannot be dismissed as merely just
the South. The correlation between the two mea
sures is lower at .44, but still impressive, and the rela
tionship is readily apparent in the data illustration.
We turn now to theempirical
test of ourproposed
linkage between state partisan-ideological divergence
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 11/14
594 Political Research Quarterly
Table 1
Partisan Ideological Divergence and
Differences in Democratic
Electoral Advantage
Variable All States South Non-South
Lag of Democratic electoral 0.806** 0.641**
0.844**
Advantage (state-national) (0.074) (0.012) (0.792)
Partisan-ideological 36.14** 69.94** 27.83**
Divergence (10.08) (16.12) (10.51)Constant 8.39* 16.56** 8.71**
(4.21) (5.41) (3.04R2.86 0.86 0.83
N58 100 358
Note: Table entries are unstandardized coefficients. Panel cor
rected standard errors are in parentheses.*
significant at/? < .05.**
significant at/? < .01.
and differences inDemocratic electoral advantage. To
account for the pooled nature of the data, we employ
an ordinary least squares model with a lagged depen
dent variable and panel corrected standard errors. The
lag accounts for the propensity for serial correlation in
the data, as well as the fact that, in terms of electoral
outcomes, history matters. That is, we should expect
that differences in state and national electoral advan
tage should be related over time.25
Table 1 shows the results of our analyses. To high
light the relationship between partisan-ideological
divergence and Democratic advantage differences, we
opted to use themost simple and parsimonious approach.
We include just the lag of the dependent variable and
divergence on the right hand side of the equation. The
first column in table 1 shows the results for all states,
with the remainder of the table documenting effects
when examined by region.26
The results strongly support our hypotheses
regarding partisan-ideological divergences and dif
ference in state and national electoral outcomes. For
the entire set of states, the coefficient is positive and
significant, and indicates that a point movement in
divergence results in a thirty-six-point change in dif
ferenced party advantage. Given the actual range of
the divergence measure among the states examined
here (.578), we find that at itsmaximum, divergence
accounts for an almost twenty-one-point difference
between state at national electoral advantage. We find
these results to be particularly noteworthy, as they
occur in the face of the expected strong performance
of the lagged dependent variable.
Turningto the remainder of the table, the results
are even stronger for southern states?almost twice
as strong in fact. This fits nicely with what we know
about realignment in the South, and the ideological
underpinnings of this partisan change. As the data in
figure 3 indicate, southern states have moved from
a point of noncompetition at the state level and
relatively little competition at the national level to a
situation of strong competition for state offices and
emerging Republican dominance at the national level.
Accordingly, the within-state differences in party per
formance have declined considerably in these states.
Our results indicate that increased consistency in polit
ical attitudes, that is, declines in partisan-ideological
divergence, are playing an important role here.
Perhaps most importantly, our results confirm that
the impact of partisan-ideological divergence is not
merely a regional phenomenon. The results for non
southern states include a strong relationship between
partisan-ideological divergence and consistency in
national and state electoral outcomes. Our analyses con
vincingly show that it is the increased alignment of ide
ological and partisan attitudes that is helping to drive
this change. As both southern and nonsouthern elec
torates have changed and become more consistent in
their core political attitudes, the parties are able to cam
paign and govern on messages that are more consistent
across electoral levels. As these partisan and ideological
messages are increasingly in sync with the attitudes of
the electorate, differences in party performance for stateand national offices decline accordingly.
Discussion and Conclusion
The literature on party competition and partisan elec
toral advantage is replete with studies documenting
the importance of viable parties for the democratic
health of a nation. In the context of the United States,
these studies have tended overwhelmingly to focus
on one electoral level to the exclusion of the other,
something we find unfortunate and deficient as we
attempt to understand the nature of party competition
in a federal electoral setting. In an earlier manuscript,
we offered a new measure that allowed for the exam
ination of party competition and advantage across
electoral levels within the states. In this article, we
extend that work by adding a temporal component to
the analysis of state-level variations in both state and
national Democratic electoral advantage. First, we
find that the electoral fortunes of the parties have
converged over the thirty-five years examined in this
study.
Differences between Democratic electoral
advantage at the national and state levels have moved
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 12/14
Brown, Bruce / Partisan-Ideological Divergence 595
over time, with the overall result being one of general
parity. Overall, Democrats hold less of an advantage
at both levels, and within-state outcomes for state and
national offices have become much more consistent
over time.
Of course, the relationship between state and
national outcomes within states varies considerably
across states. In examining these differences, we find
the role of partisan-ideological divergence to be pow
erful and intuitive factor in understanding how parties
fare at different levels of electoral contests, and how
these differences have changed over time. As elec
torates become more consistent in their political atti
tudes (i.e., as ideology drives partisanship to a point
where the attitudes are more consistent) within-state
differences in electoral outcomes for state and national
offices are reduced considerably.
The patterns observed here may allow this discus
sion to be linked to other works inmass behavior that
focus on the nature of campaigns at different levels of
office. To the degree that national offices are more the
realm of ideological messages, and state offices more
prone to partisan appeals, we see here that these dif
ferences are becoming increasingly less important. In
a broader sense, our analyses bolster our theoretical
argument regarding the need to pursue examinations
of party competition with a direct recognition that
party politics takes place in a federal electoral setting,
with different contextual forces at play across elec
toral contexts. The analyses shown here highlight this
important aspect of political parties in the American
states and illustrate that even within the same state,
national, and state party fortunes may respond to dif
ferent forces at different rates over time.
Notes
1. See Gimpel (1990) and Jewell andMorehouse (2000) for
exceptions.
2. The concept being addressed here is generally referred to as
competition in the literature. However, we find it useful to dis
tinguish between advantage (a directional concept, much like
the Ranney index) and competition (a nondirectional measure,
similar in spirit to the folded Ranney index). Here, we are con
cerned only with advantage, though throughout the discussion of
literature we will refer to competition where appropriate. See
Brown and Bruce (2002) formore.
3. This is arguably the largest body of literature examining
party competition in the states. For prominent examples,see
Dawson and Robinson 1963; Cnudde andMcCrone 1969; Dye
1966, 1984; Jennings 1979; Garand 1985; Plotnik andWinters
1985; Barrilleaux 1986; Erikson, Wright, and Mclver 1989;
Barilleaux, Holbrook, and Langer 2002.
4. For an interesting exception looking at variation in the
pace of realignments across the states, see Nardulli (1995). See
also Bullock (1988) for an examination of electoral strengthacross a range of offices over time.
5. This literature is varied, and we offer but a few important
exampleshere:Wekkin
1984;Rabinowitz, Gurian,andMacdonald
1985; Lewis-Beck and Rice 1982;Garand 1988; Simon, Ostrom,andMorra 1991; Campbell 1986;Herrnson 1993;Maggiotto and
Wekkin 1992; Perkins and Guynes 1976; Barth 1992; Hadley
1985;Niemi, Wright, andPowell 1987;Clark and Lockerbie 1998;
andBruce andClark 1998.
6. See King (1989) for a more complete discussion and cri
tique of the Ranney measure.
7. Note that a recent effort by Ceaser and Saldin (2005)follows Ranney and creates a state measure using both guberna
torial election results and state legislative party dominance (mea
sured as proportion of seats). Ceaser and Saldin cite the lack of
availability of recent state legislative election returns as a princi
palreason for this choice. For theoretical reasons, we favor the
use of elections data, and, as discussed below, employ them here.
Ceaser and Saldin also argue against using district-level analyses.
We respectfully disagree and believe an elections-based measure
premisedon a constituent level focus best captures the theoreti
cal concept of competition we are trying to evaluate.
8. Focusing on margin of victory isolates the primary ingre
dient of competition and mitigates some of the problems inherent
in the HVD measure. Most importantly, this allows for measure
ment of party competition at the national level within the states.
Given the high correlation between winning vote margin and the
overall HVD measure (0.972), we believe the loss of information
created by dropping the other three components has no negative
impact and substantial positive impact.
9. In almost all cases, the measure is simply % Democratic
vote-
% Republican vote. However, this slight modification
allows for those instances when athird-party or
independentcan
didate performs better than one of the major party candidates.
While these cases arerelatively rare, they are far more likely in
state legislative races. Moreover, state legislative races where
third-party candidates tend to do well are not randomly distrib
uted across the states, but are clustered in a small number of
states (e.g., New York). Such clustering would bias the remaining
analyses of third-party votes were not included.
10. Imagine a two-party system in which Democrats lose
every race with 100 percent of the vote going to the Republican
Party: 0-100 = -100. The opposite would have Democrats win
ningevery race
by
unanimous votes: 100-0 = 100. In a situation
where all races are ties, the results would be: 50-50 =0, or per
fect competition.
11. Any focus on electoral results as a measure of party com
petition assumes that district results are (in the aggregate) unbi
ased assessments of party strength. We recognize that any type of
consistent bias in district boundaries (e.g., via gerrymandering)
would threaten that assumption. We agree that the potential
impacts of gerrymandering areimportant for measuring electoral
competition, but note that ameaningful exploration of this topicis beyond the scope of the present article.
12. The eight-year period allows for a minimum of two elec
tions for each office in each state, minimizing the influence of
specific candidate or election effects. This is also consistent with
theHolbrook and Van Dunk approach.
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 13/14
596 Political Research Quarterly
13. Our measure, like Holbrook and Van Dunk, does include
uncontested races. This type of contest, found far more com
monly at the state level, does tend to shift the resulting assess
ment of party competition. However, we do not feel that this is a
problem, but merelya fact. Our view is simply that uncontested
races reflect an utter lack of electoral party competition, and
inclusion of these races simply reflects that situation.
14. One additional advantage of this measure is that its foun
dation allows for the possibility of significant support for a
minor party or independent candidate. It is possible, for
example, that such a candidate may draw support disproportion
ately from one candidate over the other, thereby shrinking the
gap between the Democratic and Republican candidates and
shifting the nature of electoral advantage and competition. We
account for this possibility by basing the initial calculations of
Democratic advantageon total votes cast, rather than the
Democratic percentage of the two-party vote. Thus, when there
is athird-party
or independent presence, the measure of advan
tageshifts
accordingly,with the result
being
a more
competitivescore. For more on the details of how the measure is constructed,
see Brown and Bruce (2002).
15. The data used to build our measures are drawn from a number
of sources. The archives of the Interuniversity Consortium for
Political and Social Research (ICSPR)hold national returnsthrough1990 and state legislative returns rom 1968 to 1989. It is the end of
this latter data file that has precluded any efforts at updatingmea
sures of competition. We spent considerable time gathering state leg
islative election results to extend our measure of party advantage into
more recent years. At the same time, we are aware of a similar effort
undertaken by Carsey et al. We use the Carsey et al. data here, as
their data gathering effort was a National Science Foundation
funded project thatwe believe will become the standard for data on
state legislative outcomes during this time frame. Our data, collected
independently,was used to check the Carsey et al. data.
16. The fifty eight-year periods are as follows: 1968-1975,
1970-1977, 1972-1979, 1974-1981, 1976-1983, 1978-1985,
1980-1987, 1982-1989, 1984-1991, 1986-1993, 1988-1995,
1990-1997, 1992-1999, 1994-2001, 1996-2003.
17. To allow for as much information aspossible in the visi
ble presentation,we allow each state to vary on its y-axis, rather
than presenting all states on a common axis.
18. Louisiana and Nebraska, are excluded because their gen
eral elections are not traditional partisan conflicts.
19. Our references to ideologyare in the broadest possible
terms. In the context of candidate and party behavior, itmay be
merely theuse
of labels. Among voters,it is
simply self-placementon the standard ideological spectrum. We are not imposing a
more demanding notion of ideology (i.e., Converse 1964).
20. The notion of stability of ideology is consistent with the
work on ideology in parties by Gerring (1998). He argues that
ideology is along-term and stable factor in American parties. The
difficulty inparties quickly changing their ideology is in linewith
the conclusions of Kessel (1984) regarding the presidential par
ties being a sum of their pasts.
21. We appreciate the assistance of Jerry Wright in providing the
updated state opinion data so that we can calculate these measures.
22. See Erikson,Wright, andMclver (1993) for an elaboration
on the source, reliability, and validity of these data. Because of
small samples, Alaska and Hawaii are excluded from these data,
droppingour overall number of states to forty-six (after the pre
vious exclusion of Nebraska and Louisiana).
23. As in the previous figures,we use aggregate measures for
all states, by time period, to illustrate these overall patterns.
24. Because of the magnitude of southern Democratic advan
tage in the earlier years of our analysis, it is difficult to plot both
the southern and nonsouthern states on the same scale. Since our
primary concern is with the patterns of these relationships over
time and within regions,we allow the y-axis to vary to facilitate
interpretation of the data.
25. See Keele and Kelly (2005) formore for recent work on
using lagged dependent variables. Note that we tried other esti
mation strategies (e.g., fixed effects models) and came away with
very similar results.
26. Note that state dummy variables, while included in the
analysis, are not included in table 1.
References
Barrilleaux, Charles J. 1986. A dynamic model of partisan com
petition
in the American states. American Journalof
Political
Science 30:822-40.
Barrilleaux, Charles, Thomas Holbrook, and Laura Langer. 2002.
Electoral competition, legislative balance, and American state
welfare policy. American Journal of Political Science 46:415-27.
Barth, Jay. 1992. Dual partisanship in the South. Midsouth
Political Science Journal 13:487-500.
Beck, Paul A. 1997. Party Politics inAmerica, 8th ed. New York:
Longman.
Brown, Robert D., and John M. Bruce. 2002. Political parties in
state and nation: Party advantage and party competition in a
federal setting. Party Politics 8:635-56.
Bruce, John M., and John A. Clark. 1998. Segmented partisan
ship in a Southern political elite. Polity 30:627-44.
Bullock, Charles S. III. 1988. Regional realignment from an
officeholding perspective. The Journal of Politics 50:553-74.
Campbell, James E. 1986. Presidential coattails and midterm
losses in state legislative elections. American Political
Science Review 80:45-64.
Carsey, Thomas M., William D. Berry, Richard G. Niemi, Lynda
W Powell, and James M Snyder. State legislative election
returns, 1967-2003, Release Version 1.
Ceaser, James W, and Robert P. Saldin 2005. A new measure of
party strength. Political Research Quarterly 58:245-56.
Clark, John, and Brad Lockerbie. 1998. Split partisan identifica
tion. In Party activists in Southern politics: Mirrors and mak
ersof change, ed. Charles D. Hadley and Lewis Bowman.
Knoxville:University
of Tennessee Press.
Cnudde, Charles E, and Donald J.McCrone. 1969. Party compe
tition and welfare policies in the American states. American
Political Science Review 63:858-66.
Converse, Philip E. 1964. The nature of mass belief systems. In
Ideology and discontent, ed. David Apter. New York: Free Press.
Dawson, Richard E., and James A. Robinson. 1963. Inter-party
competition, economic variables, and welfare policies in the
American states. Journal of Politics 25:265-89.
Dye, Thomas R. 1966. Politics, economics, and the public.
Chicago: Rand McNally.-. 1984. Party and policy in the states. Journal of Politics
46:1097-115.
Erikson, Robert S., Gerald C. Wright, and John P. Mclver. 1989.
Political parties, public opinion, and state policy in theUnited
States. American Political Science Review 83:729-50.
This content downloaded from 158.109.174.204 on Wed, 30 Oct 2013 16:41:57 PMAll use subject to JSTOR Terms and Conditions
7/27/2019 Partisan-Ideological Divergence and Changing Party Fortunes in the States, 1968-2003; A Federal Perspective.pdf
http://slidepdf.com/reader/full/partisan-ideological-divergence-and-changing-party-fortunes-in-the-states 14/14
Brown, Bruce / Partisan-Ideological Divergence 597
-. 1993. Statehouse democracy: Public opinion and policy
in the American states. New York: Cambridge University Press.
Garand, James C. 1985. Partisan change and shifting expenditure
priorities in theAmerican States. American Politics Quarterly13:355-91.
Garand, James C. 1988. Localism and regionalism in presidentialelections: Is there a home state or
regional advantage?
Western Political Quarterly 41:85-104.
Gerring, John. 1998. Party ideologies inAmerica, 1828-1996.
New York: Cambridge University Press.
Gimpel, James G. 1990. National elections and the autonomy of
state party systems. Pittsburgh, PA: University of PittsburghPress.
Hadley, Charles D. 1985. Dual partisan identification in the
South. Journal of Politics 47:254-68.
Herrnson, Paul. 1993. The revitalization of national party organi
zations. In The Parties Respond, ed. Sandy Maisel. New York:
Longman.
Holbrook, Thomas M., and Emily Van Dunk. 1993. Electoral
competition in the American states. American Political
Science Review 87:955-62.
Jennings, Edward T. 1979. Competition, constituencies and wel
fare policies in the American states. American Political
Science Review 73:414-29.
Jewell, Malcolm E., and Sarah M. Morehouse 2000. Political
parties and elections in American states, 4th ed. Washington
DC: CQ Press.
Keele, Luke, and Nathan J. Kelly. 2006. Dynamic models for
dynamic theories: The ins and outs of lagged dependent vari
ables. Political Analysis 14:186-205.
Kessel, John H. 1984. Presidential parties. Homewood, IL: Dorsey.
King, James D. 1989. Interparty competition in the American
states: An examination of index components. Western
Political Quarterly 41:779-90.
Lewis-Beck, Michael S., and Tom W Rice. 1983. Localism in
presidential elections: The home state advantage. American
Journal of Political Science 27:548-56.
Maggiaotto, Michael A., and Gary D. Wekkin. 1992. Segmented
partisanship in a federal system.Midsouth Political Science
Journal 13:425-43.
Nardulli, Peter F. 1995. The concept of a critical realignment,electoral change, and political behavior. American Political
Science Review 89:10-22.
Niemi, Richard G., Stephen Wright, and Lynda Powell. 1987.
Multiple party identifiers and the measurement of party iden
tification. Journal of Politics 49:1093-103.Patterson, Samuel C, and Gregory A. Caldeira. 1983. The etiology
of partisan competition. American Political Science Review
78:691-707.
Perkins, Jerry, and Randall Guynes. 1976. Partisanship
in national and state politics. Public Opinion Quarterly 40:
376-78.
Plotnik, Robert D., andRichard F.Winters. 1985.A politicoeconomic theory of income redistribution. American Political
Science Review, 79:458-73.
Rabinowitz, George, and Stuart E. Macdonald. 1986. The power
of the states in U.S. presidential elections. American Political
Science Review 80:65-87.
Rabinowitz, George, Paul-Henri Gurian, and Stuart Elaine
Macdonald. 1985. The structure of presidential elections and
the process of realignment, 1944-1980. American Journal of
Political Science 29:611-35.
Ranney, Austin. 1971. Parties in State Politics. InPolitics in the
American states: A comparative analysis, ed. Herbert Jacob
and Kenneth N. Vines. Boston: Little, Brown.
Schattschneider, E. E. 1942. Party government. New York: Farrar
and Rinehart.
Simon, Dennis M., Charles W Ostrom, Jr., and Robin F. Marra.
1991. The president, referendum voting, and subnational elec
tions in theUnited States. American Political Science Review
85:1177-92.
Wekkin, Gary D. 1984. Democrat versus Democrat. Columbia:
University of Missouri Press.
Wright, Gerald C, John P.Mclver, Robert S. Erikson, andDavid
B. Holian. n.d. Stability and change in state electorates:
Carter through Clinton. Xerox.