ArticlePDF Available

Divided Government and Delay in the Legislative Process: Evidence From Important Bills, 1949-2010

Authors:

Abstract and Figures

Despite a robust history of studies examining legislative outputs, little is known about how divided government affects the policymaking process. This article examines these dynamics by analyzing the relationship between divided government and delay in the consideration of important legislation. We also introduce a more nuanced measure of divided government—the strength of the president’s party in Congress—that measures both the presence and magnitude of inter-branch conflict. Using a Cox proportional hazards model to analyze delay of important legislation from 1949 to 2010, the results indicate both divided government and the strength of the president’s party in Congress are significantly related to legislative delay. Moreover, presidential party strength significantly interacts with partisan polarization. When the parties are moderately or highly polarized, there is a significant relationship between the strength of the president’s party and legislative delay; this relationship is insignificant at lower levels of polarization. Taken together, these findings enhance our understanding of how inter-branch conflict affects the policymaking process in Congress.
Content may be subject to copyright.
American Politics Research
2015, Vol. 43(5) 771 –792
© The Author(s) 2015
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1532673X15574594
apr.sagepub.com
Article
Divided Government and
Delay in the Legislative
Process: Evidence From
Important Bills,
1949-2010
Tyler Hughes1 and Deven Carlson2
Abstract
Despite a robust history of studies examining legislative outputs, little is
known about how divided government affects the policymaking process.
This article examines these dynamics by analyzing the relationship between
divided government and delay in the consideration of important legislation.
We also introduce a more nuanced measure of divided government—the
strength of the president’s party in Congress—that measures both the
presence and magnitude of inter-branch conflict. Using a Cox proportional
hazards model to analyze delay of important legislation from 1949 to 2010, the
results indicate both divided government and the strength of the president’s
party in Congress are significantly related to legislative delay. Moreover,
presidential party strength significantly interacts with partisan polarization.
When the parties are moderately or highly polarized, there is a significant
relationship between the strength of the president’s party and legislative
delay; this relationship is insignificant at lower levels of polarization. Taken
together, these findings enhance our understanding of how inter-branch
conflict affects the policymaking process in Congress.
1California State University, Northridge, CA, USA
2University of Oklahoma, Norman, OK, USA
Corresponding Author:
Tyler Hughes, California State University, 210 Sierra Hall, 18111 Nordoff Street, Northridge,
CA 91330-8254, USA.
Email: tyler.hughes@ou.edu
574594APRXXX10.1177/1532673X15574594American Politics ResearchHughes and Carlson
research-article2015
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
772 American Politics Research 43(5)
Keywords
Congress, divided government, delay, polarization, legislative process
Introduction
On August 2, 2011, President Obama signed the Budget Control Act of
2011—universally considered an important piece of legislation—into law.
This legislation passed through both a Republican-controlled House of
Representatives and a Democratic-controlled Senate and, on the surface,
could reasonably be viewed as little more than the latest in a long line of
important legislative outputs from a divided government context. Such a
view, however, fails to recognize the considerable acrimony and rancor that
preceded the passage of the Budget Control Act. Indeed, congressional lead-
ers negotiated with the president for months on the issue, and numerous com-
peting bills were introduced, debated, and voted on in the two chambers of
Congress. The political spectacle surrounding the Budget Control Act made
it difficult for legislators to focus on other aspects of the legislative agenda,
and the divisive political environment—coupled with the perceived ineffi-
ciency of the federal government—drove public approval of Congress to a
historic low of 13%, according to Gallup.
A long line of prior studies analyzed the effects of divided government, but
existing work focuses almost exclusively on how partisan control of our gov-
erning institutions affects legislative outputs (e.g., Binder, 1999, 2003;
Coleman, 1999; Edwards, Barrett, & Peake, 1997; C. O. Jones, 1994; Krutz,
2001; Mayhew, 1991, 2005). The case of the Budget Control Act of 2011 dem-
onstrates that, even if important legislation is ultimately passed, divided gov-
ernment also has the potential to affect the legislative process. In light of this
demonstration, this article aims to shift the focus from the effects of divided
government on legislative outputs to its effects on the legislative process.
Drawing on a data set containing the universe of important legislation from
1949 to 2010—defined as bills in the top 25% of coverage in the CQ Almanac
we analyze the relationship between divided government and legislative delay.
Delay is measured as the number of days from initial introduction of a piece of
legislation to its final passage or to the end of the 2-year congressional session
if the bill does not become a public law. Moreover, we examine this relation-
ship using a new, more nuanced measure of divided government—a continuous
measure of the strength of the president’s party (or lack thereof) in Congress.
Together, these advances provide significant insights into the relationship
between divided government and the legislative process.
This article proceeds in three parts. First, we present a brief overview of
existing studies on the effects of divided government, focusing most heavily
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 773
on the theoretical implications of these studies for our analysis of legislative
delay. This overview of the literature provides valuable context for the sec-
ond stage of our study, where we develop our measures of legislative delay
and divided government. Finally, using the measures we develop in the sec-
ond section of the article, we estimate a series of survival models to examine
the relationship between divided government and legislative delay. The
results demonstrate important legislation encounters more delay in a divided
government context than under unified government. The results further indi-
cate divided government interacts with party polarization in a meaningful
way. Specifically, the interaction demonstrates divided government only has
a significant effect on delay at moderate and high levels of partisan polariza-
tion—This interaction is masked in analyses using the traditional dichoto-
mous measure of divided government.
These findings have implications for several important issues, including
the legislative agenda and productivity of Congress, the content of legisla-
tion, and public trust of government. For example, lengthy consideration of a
particular piece of legislation—as occurred with the Budget Control Act of
2011—can result in that topic monopolizing the legislative agenda, and may
ultimately result in enactment of a smaller proportion of the congressional or
presidential policy agenda over the course of the legislative session. Similarly,
delay in the legislative process may be symptomatic of inter-party or inter-
branch negotiation over the substance of the bill in question. If this is the
case, then the content of delayed legislation is unlikely to reflect the prefer-
ences of any single actor or group in the process—The legislation is the prod-
uct of compromise. In contrast, the content of legislation that passed with
relative expediency may be likely to adhere closely to the preferences of a
single stakeholder, most likely the president or the majority party in Congress.
Overall, this article illustrates the utility of efforts to further understand insti-
tutional and political influences on congressional performance, and high-
lights how inter-branch conflict affects the policymaking process even when
we look beyond legislative outputs.
Divided Government and Legislative Delay
Analysis of the effects of divided government began in earnest with Mayhew’s
(1991, 2005) seminal work testing the hypothesis that divided government
affected legislative productivity within Congress. Drawing on a data set of
significant legislation passed by each Congress since 1946, Mayhew found no
significant differences in the amount of significant legislation passed in
divided versus unified government contexts. Mayhew’s work sparked a series
of studies on the topic, a number of which reached the same general
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
774 American Politics Research 43(5)
conclusion that divided government had little effect on legislative productivity
(e.g., Fiorina, 1996; C. O. Jones, 1994; Thorson, 1998). However, other stud-
ies—critiquing Mayhew’s approach on both conceptual (Binder, 1999, 2003;
Edwards et al., 1997) and methodological (Howell et al., 2000) grounds—
reached starkly different conclusions.
Conceptually, several studies noted Mayhew’s analysis overlooked failed
legislation. These studies argued the best test for the effects of divided gov-
ernment on legislative productivity was not the amount of important legisla-
tion passed into law but rather the proportion of the legislative agenda that
was enacted (Binder, 1999, 2003; Edwards et al., 1997). Conceptualized in
this manner, these studies concluded divided government inhibits legislative
productivity (e.g., Binder, 1999, 2003; Edwards et al., 1997). Similar find-
ings emerged from studies using models accounting for a wider range of fac-
tors than Mayhew considered, such as party responsiveness and intra-party
conflict (Coleman, 1999), although there is some evidence that controlling
for polarization results in findings that mirror Mayhew’s (D. R. Jones, 2001).1
Methodologically, Mayhew’s study was critiqued for presenting the data as
an aggregated time series, which compressed important variation, and Howell
et al. (2000) demonstrate how the application of appropriate time-series tech-
niques results in the conclusion that divided government has a substantial
negative effect on the enactment of landmark legislation. Considered as a
whole, the preponderance of evidence from previous studies suggests divided
government has at least some negative effect on legislative productivity.
To this point, existing work on divided government focused almost exclu-
sively on its effect on some variant of legislative output, typically the number
of pieces of legislation passed or the proportion of the legislative agenda
enacted. As such, prior work provides little information on how divided gov-
ernment might affect the legislative process. To the extent previous studies
provide any relevant information on this issue, the evidence indicating
divided government inhibits legislative productivity (e.g., Binder, 1999,
2003; Edwards et al., 1997) could reasonably support an expectation that
divided partisan control of our governing institutions also affects the legisla-
tive process. However, such a hypothesis has never been the subject of a
direct empirical test. In an effort to better understand the effects of divided
government on the legislative process—and not just outputs—we explore its
relationship with legislative delay.
Although largely overlooked in the literature, legislative delay is an
important phenomenon because it has the potential to produce a variety of
negative consequences for a democratic government. For example, long
delays in the passage of legislation may represent an inefficiency of govern-
ment. Such delays have the potential to usurp the congressional agenda and
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 775
monopolize the time and energies of lawmakers. These implications are espe-
cially important considering the scarcity of time and attention as policymak-
ing resources (B. D. Jones & Baumgartner, 2005) and the limited agenda
space available in legislative institutions, such as Congress (Adler &
Wilkerson, 2012). Delay in the creation of public laws can also affect public
trust of government institutions, as long, drawn-out political battles over
important legislation can cause the public to become jaded with the demo-
cratic process, which further complicates matters of efficacy and legitimacy
(Hibbing & Theiss-Morse, 1995).
A small body of work has analyzed the effect of divided government on
delay in other political processes, including the confirmation of judges
(Binder & Maltzmann, 2002; Shipan & Shannon, 2003) and executive branch
nominations (Bond, Fleisher, & Krutz, 2009), the budget process (Binder,
2003; Woon & Anderson, 2012), and treaty-making (Krutz & Peake, 2009;
Peake, Krutz, & Hughes, 2012). Although each of these studies analyzes
delay only among a very specific piece of congressional action—each pro-
cess takes place within a single chamber in Congress—together, they contrib-
ute evidence that divided government has implications for the legislative
process, particularly with respect to delay in that process. Indeed, almost
without exception, these studies conclude—relative to unified partisan con-
trol—divided government introduces delay into the actions of Congress.
Considered as a whole, previous studies examining the effect of divided gov-
ernment on both legislative productivity and delay in specific legislative pro-
cesses provide a basis for the expectation of divided government causing
delay in the consideration of important legislation—an aspect of congressio-
nal policymaking yet to be explicitly considered. In the following sections,
we describe the approach used to test this expectation empirically.
Measuring Delay in the Legislative Process
The data set underlying our analysis of the relationship between divided gov-
ernment and legislative delay contains information on 2,207 separate pieces
of important legislation considered by Congress between 1949 and 2010—
Importantly, our data contain both bills that were signed into law and bills
that failed to pass. We use the CQ Almanac, which contains an article on
every major piece of legislation lawmakers considered throughout a
Congressional session, as the basis for identifying the pieces legislation
included in our data. More specifically, we draw on the Policy Agenda
Project’s2 (PAP) compilation of all CQ Almanac articles from 1949 and 2010
and then—using the PAP’s measure of article length—identify all bills in the
top 25% of coverage over this time period for inclusion in our data (see Krutz,
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
776 American Politics Research 43(5)
2001 for a similar procedure for identifying important legislation from CQ
Almanac articles). We restrict our sample to bills in the top 25% of coverage
in the CQ Almanac because it is these bills for which the potential conse-
quences of delay—for example, monopolization of Congress’ policy agenda,
effects on public trust and efficacy, and others—are most pronounced. Our
approach to identifying important legislation results in our data containing
observations across the full-time period.
Most policy ideas undergo several iterations prior to reaching a final leg-
islative form—even within a given session of Congress. We account for this
reality in our data by taking advantage of the fact that each article in the CQ
Almanac lists the full slate of bills pertaining to the legislation covered in the
article. The introduction date of the legislation in our data is that of the bill
with the earliest introduction among the slate of related bills listed in the
Almanac. After identifying the 2,207 pieces of important legislation for inclu-
sion in our data and specifying their introduction date, we drew on multiple
sources of information—including the PAP, the Congressional Bills Project,3
the CQ Almanac, and Poole and Rosenthal’s (1997) work—to attach several
characteristics to the legislation in our data. We describe the specific charac-
teristics attached to the bills later in the article.
For the pieces of legislation in our data that became law, we operationalize
the concept of legislative delay by calculating—within a given session of
Congress—the number of days between the date a bill was introduced and the
date the bill was signed into law. For legislation that did not become law, we
measure delay as the number of days between the bill’s introduction and the
final day on the legislative calendar in that congressional session. We
acknowledge many—perhaps most—policy ideas are first introduced several
congresses prior to enactment (or ultimate failure) and our measure of delay
is constrained to a maximum of 2 years. However, we believe this approach
to measuring delay is most appropriate, given the institutional structure of
Congress. Specifically, our approach to measuring delay is consistent with
the institutional rules requiring the work of Congress to begin de novo every
2 years—Policy ideas from previous congresses may be reintroduced in a
new session, but this reintroduction requires a new draft of legislation with a
new bill number.
Table 1 summarizes the legislation in our data by Congressional session and
passage status. Specifically—for both passed and failed legislation—the table
presents the number of bills in our data in each session of Congress, as well as
the median number of days to exit from our data—either final passage or fail-
ure induced by the end of the legislative session. The table makes clear our data
contain 1,663 pieces of passed legislation and 544 pieces of failed legislation.4
It is also clear a substantial proportion of the legislation in our data—about
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 777
Table 1. Summary of Time Passage or Failure for Significant Legislation, by
Congress (1949-2010).
Congress
Passed legislation Failed legislation
Divided/
unified
government
Presidential
party
strength
Number
of bills
Median time
to exit
Number
of bills
Median
time to exit
81st (1949-1950) 38 130 16 677 Unified −4
82nd (1951-1952) 28 115 4 557 Unified −10
83rd (1953-1954) 26 63 2 544 Unified −10
84th (1955-1956) 36 66 16 661 Divided −15
85th (1957-1958) 34 91 12 610 Divided −17
86th (1959-1960) 30 93 10 652 Divided −65
87th (1961-1962) 68 99 8 541 Unified 4
88th (1963-1964) 66 162 18 525 Unified 7
89th (1965-1966) 119 138 22 562 Unified 8
90th (1967-1968) 92 150 24 667 Unified 4
91st (1969-1970) 97 169 25 396 Divided −26
92nd (1971-1972) 81 97 43 468 Divided −38
93rd (1973-1974) 100 133 33 648 Divided −26
94th (1975-1976) 95 158 39 627 Divided −74
95th (1977-1978) 110 179 35 614 Unified 1
96th (1979-1980) 73 198 24 636 Unified −2
97th (1981-1982) 43 167 12 425 Divided −26
98th (1983-1984) 41 164 16 529 Divided −52
99th (1985-1986) 33 241 13 534 Divided −36
100th (1987-1988) 46 165 22 545 Divided −41
101st (1989-1990) 62 132 27 540 Divided −43
102nd (1991-1992) 60 124 20 596 Divided −51
103rd (1993-1994) 58 126 14 444 Unified −3
104th (1995-1996) 36 185 17 575 Divided −14
105th (1997-1998) 43 136 20 630 Divided −15
106th (1999-2000) 48 173 34 584 Divided −15
107th (2001-2002) 24 144 7 525 Divided −10
108th (2003-2004) 20 155 2 596 Unified −9
109th (2005-2006) 21 159 4 278 Unified −5
110th (2007-2008) 19 240 1 655 Divided −20
111th (2009-2010) 16 141 4 552 Unified −3
41% of observations—are from the 1960s and 1970s. Such a result is consistent
with other efforts to identify important legislation over time. For example,
approximately 32% of Mayhew’s (2005) list of important enactments came
from this time period. The fact that this time period disproportionately contrib-
utes to our data is not surprising, as the late 1960s and early 1970s saw Congress
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
778 American Politics Research 43(5)
debating (and passing) a number of large, significant pieces of legislation.
Finally, Table 1 demonstrates considerable heterogeneity in time to enactment
(or failure) across the years our data span. While there are relatively short
median times to enactment in earlier periods of the data, there does not seem to
be a clear pattern throughout the entirety of the data.
Determinants of Delay
As reviewed above, a number of studies focused on how inter-branch con-
flict—in the form of divided government—affects legislative outputs.
Although these studies reach competing substantive conclusions, each study
measures inter-branch conflict using a simple dichotomous variable indicat-
ing divided government. Following this work, we also use a dichotomous
measure of divided government in our analysis. However, we also develop an
additional measure of divided government, one that accounts for not only the
presence of inter-branch conflict but also the magnitude.
To measure the magnitude of the legislative barriers the president’s party
must overcome to pass a piece of legislation through Congress, we subtract
the number of presidential partisans in each chamber from the voting thresh-
old required to pass legislation in that chamber—a simple majority in the
House and the three fifths required for cloture in the Senate.5 The larger of
these differences is the largest barrier to the passage of legislation and repre-
sents the strength of the president’s party in a given Congress. This measure
builds on the traditional measure of divided government by not only captur-
ing partisan control but also identifying the magnitude of the largest partisan
hurdle—in either the House or the Senate—the president’s party must over-
come to pass legislation. We multiply the measure by negative one to make
the direction of the variable intuitively match its label. Consequently, nega-
tive values of our measure indicate substantial legislative barriers for the
president’s party; any value above zero indicates a Congress in which the
president’s party holds both a majority in the House and a supermajority in
the Senate.
Table 1 displays both the traditional measure of divided government as
well as our new measure of presidential party strength over time. Presidential
party strength clearly captures the presence of divided government and allows
for an assessment of the magnitude of this inter-branch conflict. Table 1 also
indicates presidents typically deal with substantial legislative barriers—even
during times of unified government because of the supermajority require-
ment of the Senate. For example, President Truman enjoyed unified govern-
ment during the 83rd and 84th Congresses (1951-1954), but the party strength
variable remains well below zero because the Democrats only held a very
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 779
slim majority in the Senate that did not come close to the two thirds required
to invoke cloture at the time. Table 1 also suggests a weak relationship
between median time to enactment and the presidential party strength vari-
able—There is considerable variation across political environments. A simi-
larly weak relationship exists between divided government and time to
enactment. Despite the weak relationship between divided government and
median time to enactment in Table 1, the extant literature on divided govern-
ment leads to an expectation that legislative delay should decrease as presi-
dential party strength increases. Similarly, the presence of divided government
should increase legislative delay.
Other Variables
In addition to divided government and party strength, there are several addi-
tional political and institutional factors that may contribute to legislative
delay; we include these factors as control variables in our analysis. The first
such factor is the level of ideological polarization between the two major par-
ties. Some studies suggest party polarization slows down the legislative pro-
cess by increasing conflict and decreasing the likelihood of compromise
(Binder, 1999, 2003; Gilmour, 1995). These ideological gaps can also affect
inter-branch conflict indirectly—Larger ideological gaps may make it more
difficult for the two parties to reach a compromise. Using Poole and
Rosenthal’s (1997) common-space DW-NOMINATE scores, which allow for
the comparison of ideologies across chambers (Chiou & Rothenberg, 2007;
Lewis & Poole, 2004), we measure party polarization as the distance between
the median members of each party in Congress.
Some scholars present a more nuanced argument regarding the effects of
polarization on the legislative process. These arguments note polarization is
a necessary condition for conditional party government and thus may serve to
actually overcome legislative conflicts (Aldrich, 1995; Aldrich & Rohde,
2001; Rohde, 1991). This line of thinking holds that polarization induces
individual legislators to follow the party line, which should increase the abil-
ity of the majority party to push legislation through Congress. Similarly,
scholars note rank-and-file legislators ceded significant power to party lead-
ers during the contemporary period of a polarized Congress and contend this
increased power—coupled with more homogeneous party caucuses—allowed
majority party leaders to better control floor debate through the use of special
rules and thus achieve their desired legislative outcomes (Lee, 2009; Aldrich
& Rohde, 2001; Theriault, 2008).
Given this information, the effects of party polarization on legislative
delay may be conditional on the partisan make-up of Congress. Under
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
780 American Politics Research 43(5)
periods of unified government and high polarization, the majority party has
extensive control over the congressional agenda and floor activity, so legisla-
tion should move through Congress with relative expediency. However, leg-
islation should encounter more delay during times of divided government
and high polarization when either a single party is unable to control the
entirety of the congressional agenda or the president may threaten to veto
legislation pushed through Congress by the opposition party. There is no
expectation about how this relationship plays out at lower levels of polariza-
tion when there is a great deal of ideological overlap between the parties. We
include an interaction term between our measure of presidential party strength
and polarization in the analysis to capture the potentially conditional effects
of party polarization.6
Beyond party dynamics, there are a number of potential institutional hur-
dles to the passage of legislation. For instance, substantial conflict between
the House and Senate may result in a bill passing in one chamber but failing
in the other (Binder, 1999, 2003). To measure the level of conflict between
the two chambers, we use DW-NOMINATE scores as the basis for a variable
measuring the absolute value of the difference between the median legislator
in the House and Senate. The overall health of the budget represents another
possible hurdle to the passage of new policy programs; it is often difficult to
pass legislation that increases spending or decreases revenue in times of high
budget deficits. Consistent with previous research, budgetary health is mea-
sured as the size of the federal surplus or deficit as a percentage of federal
outlays—Positive values represent budgetary surpluses and negative values
represent deficits (Binder, 1999; Edwards et al., 1997; Mayhew, 1991, 2005).7
The workload of Congress may also affect legislative productivity. It is pos-
sible increased workload in Congress slows down the legislative process, as
a greater number of agenda items require a larger time commitment for the
institution. To capture this potential dynamic, we use the raw number of hear-
ings held in a session of Congress to measure the congressional agenda.8 This
measure of the agenda is not the same as the “systematic agenda” described
by Cobb and Elder (1983), but Baumgartner and Jones (1993, 2009) success-
fully demonstrated the use of congressional hearings to capture the political
agenda. The final institutional variable included in the model accounts for the
institutional reforms of the mid-1970s, which caused dramatic changes to
lawmaking in Congress with the implementation of multiple referral, com-
mittee sunshine laws, and the decentralization of the committee system
(Sinclair, 2011). A dichotomous variable indicating the post-reform
Congress—beginning in 1975—is included to account for these changes.
We also include a number of bill-specific variables in the model. Some
of the important bills in our data were passed as smaller pieces of a larger
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 781
omnibus package. Such packages are indicative of strategies leaders use
when faced with contentious congressional policy making (Krutz, 2001).
Consequently, we include a variable in our analysis indicating whether
the bill was passed as part of an omnibus package.9 We measure delay for
these bills as the amount of time between the introduction of the initial
piece of legislation and passage of the omnibus package, rather than as
the number of days between introduction and passage of the omnibus
legislation.
Salience and the number of legislative actions on a bill also potentially
affect legislative delay. We include a measure of the number of lines dedi-
cated to each bill in the CQ Almanac to account for these dynamics—More
article space is inherently dedicated to bills with multiple legislative actions
and bills garnering more attention from the general public. Finally, we take
into account the policy content of each bill using the 20 major-topic policy
categories detailed in the Policy Agendas Project’s coding scheme. The mea-
sure appears in the model as a series of dichotomous variables, with “eco-
nomic policy” being the excluded category.10
Given the nature of our dependent variable—a variable measuring the
time until a piece of legislation exits the data set—we use a series of Cox
proportional hazards models to test the article’s expectations (Box-
Steffensmeier & Jones, 2004).11 In this case, the “survival” of a bill refers to
the length of time between the bill’s introduction and the bill’s subsequent
exit from the data set (when we stop observing the survival-time of an obser-
vation). A “failure” in the model refers to the date on which a bill is passed
into a law. Bills not passed into law exit the data set on the final day of the
2-year congressional session in which the bill was introduced—These bills
are “censored.” For this reason, the models assess risk as it relates to a bill
passing into law; positive coefficients represent decreases in delay, and nega-
tive coefficients represent increases in delay.12
With the independent variables described in this section, we estimate three
Cox proportional hazards models to examine the relationship between inter-
branch conflict and legislative delay. The first of these models regresses our
measure of delay on the standard, dichotomous measure of divided govern-
ment and the set of control variables described above. Our second model is
similar in structure to the first but replaces the standard measure of divided
government with our measure of party strength—The set of control variables
remains identical. The final model contains both our measure of party strength
and an interaction between party strength and party polarization to assess the
conditional nature of the relationship between polarization and legislative
delay. The third model also contains the same set of control variables present
in the first two.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
782 American Politics Research 43(5)
Results
In addition to displaying the expected relationship between legislative delay
and the variables in our model (other than policy topic), Table 2 presents the
results from estimating each of the three Cox proportional hazards models
described above. Results from the first model—containing the standard,
dichotomous measure of divided government—indicate split partisan control
of the presidency and at least one chamber of Congress decreases the hazard
rate of important legislation. Put differently, the results indicate the presence
of divided government slows the legislative process for important
legislation.
As described above, the second model in Table 2 replaces the standard,
dichotomous measure of divided government with our continuous measure of
presidential party strength—The remaining contents of the second model are
identical to those of the first. The results from estimating this model are pre-
sented in the third column of Table 2. As expected, the strength of the presi-
dent’s party in Congress has a statistically significant, positive effect on the
hazard rate of important legislation. As the party strength variable measures
the magnitude of the largest legislative hurdle the president’s party must
overcome to pass legislation, the results clearly demonstrate larger hurdles—
a greater number of legislators of the opposite party—result in significantly
greater levels of delay in the legislative process.
The results for control variables are similar across models, and largely
adhere to expectations. For example, the results indicate chamber distance
has a statistically significant, negative relationship with legislative delay—
Greater ideological distance between the median legislator in each chamber
increases the likelihood a bill encounters delay. Likewise, the results suggest
bills introduced in the post-reform Congress are more likely to encounter
delay than bills introduced prior to 1975. This finding lends further evidence
to the idea of the reforms of the 1970s ushering in a new era of lawmaking in
Congress, one that slowed the process of legislating (Sinclair, 2011).
Budgetary health also has a statistically significant effect in the models, but
not in the expected direction. According to the results, important bills are
more likely to encounter legislative delay in eras of budgetary health, relative
to years with larger budget deficits. Finally, the workload of Congress—mea-
sured as the number of annual hearings—does not appear to be related to
delay in the legislative process.
Turning to the bill-specific measures in Table 2, the results indicate bills
included in an omnibus package encounter significantly more delay than
non-omnibus legislation. This finding may seem unintuitive—and it is at
odds with our a priori expectation—but the results are consistent with a
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 783
Table 2. Cox Proportional Hazards Models Predicting Survival Time of Important
Legislation (1949-2010).
Expected
relationship Model 1 Model 2 Model 3
Party control of government
Divided government −0.183* (0.057)
Presidential party strength + 0.003* (0.001) 0.021 (0.020)
Presidential party strength/party
polarization interaction
+ 0.029 (0.033)
Institutional factors
Party polarization 0.0363 (0.561) −0.323 (0.524) −0.620 (0.626)
Chamber distance −0.832* (0.474) −0.920* (0.474) −1.091* (0.513)
Budgetary health + −0.008* (0.003) −0.010* (0.003) −0.010* (0.003)
Congressional workload (1,000s) −0.050 (0.075) −0.064 (0.071) −0.073 (0.072)
Post-reform Congress −0.370* (0.139) −0.239* (0.130) −0.247* (0.131)
Bill-specific factors
Omnibus package + −0.553* (0.202) −0.577* (0.201) −0.598* (0.203)
Article length (100s) 0.004 (0.003) 0.004 (0.003) 0.004 (0.003)
Civil rights and liberties policy −1.107* (0.210) −1.112* (0.210) −1.114* (0.210)
Health policy N/A −0.603* (0.164) −0.600* (0.164) −0.601* (0.164)
Agricultural policy N/A 0.026 (0.153) 0.216 (0.152) 0.024 (0.153)
Labor policy N/A −0.820* (0.176) −0.822* (0.176) −0.826* (0.176)
Education policy N/A −0.625* (0.173) −0.630* (0.178) −0.623* (0.178)
Environmental policy N/A −0.895* (0.178) −0.900* (0.178) −0.898* (0.178)
Energy policy N/A −0.389* (0.162) −0.390* (0.163) −0.386* (0.163)
Immigration policy N/A −1.048* (0.371) −1.055* (0.371) −1.055* (0.372)
Transportation policy N/A −0.202 (0.153) −0.200 (1.53) −0.197 (0.153)
Crime & family policy N/A −0.885* (0.188) −0.881* (0.188) −0.881* (0.188)
Social welfare policy N/A −0.230 (0.174) −0.233 (0.174) −0.323 (0.174)
Housing and development policy N/A −0.311* (0.174) −0.311* (0.174) −0.307* (0.174)
Domestic commerce policy N/A −0.860* (0.164) −0.861* (0.164) −0.859* (0.163)
Defense policy N/A 0.377* (0.130) 0.382* (0.130) 0.385* (0.130)
Science and technology policy N/A −0.221 (0.183) −0.223 (0.183) −0.230 (0.183)
Trade policy N/A −0.501* (0.195) −0.513* (0.195) −0.519* (0.195)
Foreign affairs policy N/A 0.115 (0.150) 0.115 (0.150) 0.116 (0.150)
Government operations policy N/A 0.066 (0.128) 0.062 (0.128) 0.066 (0.128)
Public lands policy N/A −0.476* (0.169) −0.479* (0.168) −0.480* (0.169)
BIC 23,356.52 23,361.21 23,368.14
AIC 23,202.64 23,207.33 23,208.55
Log-likelihood −11,574.32 −11,576.66 −11,576.28
N2,207 2,207 2,207
Note. Numbers in parentheses are standard errors. BIC = Bayesian information criterion; AIC = Akaike
information criterion.
*p < .05 (one-tailed).
scenario where bills are introduced and sit in committee for a long period of
time before being attached to an omnibus bill in an effort to strategically
move the bill toward passage. As described earlier, we measure delay for bills
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
784 American Politics Research 43(5)
passed as part of an omnibus package as the amount of time between the
introduction of the initial piece of legislation and passage of the omnibus
package, rather than as the number of days between introduction and passage
of the omnibus legislation. At least some bills ultimately included in omnibus
packages languished in committee without advancement through the legisla-
tive process. Such bills likely contribute to the negative coefficient estimates
on the omnibus variable in Table 2.
The results in Table 2 also make clear our measure of salience—the num-
ber of lines dedicated to a bill in the CQ Almanac—is not significantly related
to legislative delay. Finally, with respect to substantive policy area, the find-
ings suggest bills addressing issues of macroeconomic policy—the excluded
category in the policy issue factor variable—encounter less delay than bills
within nearly every other policy domain. Defense policy is a notable excep-
tion to this pattern—Bills in this domain experience even less delay than
macroeconomic legislation.
Table 3 displays the predicted change in the hazard rate across the full
range of the significant covariates in Models 1 and 2.13 According to the pre-
dictions, the hazard rate of important legislation considered under periods of
divided government is 15.4% lower than that of important legislation consid-
ered under unified government. Similarly, the hazard rate for important legis-
lation increases by 11.6% when moving from the minimum to the maximum
values of the party strength variable. These findings make it clear the sub-
stantive impacts of a dichotomous divided government variable and the con-
tinuous measure of presidential party strength are very similar. Moreover, the
substantive impact of the ideological distance between the two chambers is
similar to that of the measures capturing inter-branch conflict; the hazard rate
of important legislation decreases by 13.5% across the range of the chamber
Table 3. Predicted Change in Hazard Rates for Significant Covariates.
Variable Change in X (from, to)
Percentage change
in hazard rate
Divided government (0, 1) −15.40
Presidential party strength (−74, 8) 11.56
Chamber distance (0.002, 0.222) −13.56
Budgetary health (−40.15, 15.50) −38.77
Post-reform congress (0, 1) −31.01
Omnibus package (0, 1) −35.72
Note. Change in hazards rates are calculated over the full range of significant covariates.
Predictions are based on the mean values of other independent variables in the model.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 785
distance variable. The variables capturing legislation introduced in the post-
reform Congress and bills included as part of an omnibus package also have
sizable substantive relationships with legislative delay. The hazard rate of
post-reform important legislation is 31.0% lower than that of pre-reform
important legislation. Similarly, the hazard rate of bills included as part of an
omnibus package is 35.7% lower than non-omnibus legislation.
The final model in Table 2 retains our measure of party strength but also
includes an interaction between the measure of party strength and party
polarization. Although the coefficient on the interaction between these vari-
ables is not statistically significant, a simple interpretation of the coefficient
does not tell the whole story—The significance of the marginal effect of party
strength on delay may vary across different values of party polarization
(Brambor et al., 2006). Figure 1 presents the plot of the interaction between
party strength and polarization. Importantly, Figure 1 demonstrates the rela-
tionship between party strength and legislative delay is statistically signifi-
cant across a segment of the full range of party polarization.14 Specifically, at
lower levels of polarization, the relationship between party strength and leg-
islative delay is insignificant. At medium to high level of polarization, how-
ever, the relationship between party strength and legislative delay becomes
significant before returning to insignificance at the highest levels of polariza-
tion. Even at this point, though, the point estimate is large and positive, but
Figure 1. Simulated effects of the interaction between presidential party strength
and party polarization on survival time of important legislation.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
786 American Politics Research 43(5)
the relatively small number of observations at these values result in less pre-
cision and correspondingly large confidence intervals. In other words, the
strength of the president’s party in Congress has a statistically significant
relationship with legislative delay when the parties are moderately to highly
polarized, but no discernible relationship exists when levels of party polariza-
tion are lower.
It is difficult, perhaps even impossible, to completely disentangle the
effects of time and polarization when examining the contemporary Congress.
Party polarization shows a linear increase starting in the mid-1970s (Theriault,
2008). For this reason, the interactive effect in Figure 1 may simply be an
artifact of delay increasing over the same time period.15 The inclusion of the
post-reform Congress variable helps account for this potentially confounding
problem because the reforms took effect at the same time the parties began
moving away from one another. Regardless, the defining characteristic of the
contemporary Congress is partisan polarization, so it may not be possible to
entirely separate time and polarization.
The interactive effect is intuitively appealing because low levels of polar-
ization imply substantial ideological overlap between the parties, making
legislative voting barriers easier to overcome, in theory. However, greater
levels of polarization between the parties create a contentious legislative
environment, which is likely to make compromise far more difficult. This
relationship is relevant to the concept of “conditional party government”
(Aldrich, 1995; Aldrich & Rohde, 2001; Rohde, 1991) wherein party loyalty
votes become far more likely as inter-party differences are increased and
intra-party conflict is decreased. Under such conditions, greater levels of
polarization result in fewer party defections on roll call votes, making it far
more difficult for the president’s party to overcome voting deficits when
attempting to pass legislation. It is important to note we also estimated a
model where we interacted the traditional, dichotomous divided government
variable with polarization, but the interaction was insignificant.16 Taken
together, these findings further suggest the party strength variable adds a
level of nuance overlooked by a dichotomous divided government variable.
Discussion
There is a robust history of studies examining the relationship between
divided government and congressional productivity (Binder, 1999, 2003;
Coleman, 1999; Edwards et al., 1997; C. O. Jones, 1994; Krutz, 2001;
Mayhew, 1991, 2005). Although these studies are inconsistent with respect to
their conclusions about the nature of this relationship, they are fully consis-
tent in their focus on analyzing legislative output. In doing so, these studies
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 787
provide valuable insight into how institutional context shapes the final prod-
uct of policy making, but they provide less information on how contextual
factors might shape the legislative process more generally—especially issues
of legislative delay. Although delayed legislation may ultimately be passed
into law in some cases, a drawn-out and acrimonious policy debate may pro-
vide the ultimate undoing of legislation in other cases. Regardless of whether
important legislation ultimately becomes law, delay in the legislative process
can have very real consequences for democratic governance. Specifically,
delay in the legislative process can constrict the working agenda of Congress,
and such constraints can render the institution unable to address pressing
public problems. Moreover, contentious politics has been shown to draw the
ire of the public, which can undermine the efficacy and legitimacy of
Congress (Hibbing & Theiss-Morse, 1995). Motivated by these implications,
this article sought to provide evidence on the relationship between divided
government and delay in the legislative process.
In addition to shifting the focus from legislative outputs to legislative pro-
cesses—specifically legislative delay—we also developed a more refined
measure of divided government in this article. This new measure calculates
the strength of the president’s party in Congress, and although highly corre-
lated with the traditional measure of divided government, this “party strength”
variable represents a valuable innovation measuring both the presence and
magnitude of inter-branch conflict. Together, these advances facilitate a more
comprehensive understanding of the effects of divided government on the
legislative process.
At a basic level, our analysis demonstrates divided government—mea-
sured both traditionally as well as by our new measure of party strength—
increases the amount of time important bills are immersed in the legislative
process. However, we take advantage of the nuanced nature of the party
strength variable to further demonstrate an interactive effect of presidential
party strength and partisan polarization on legislative delay. Specifically, we
demonstrate presidential party strength in Congress has no effect on delay at
low levels of party polarization. However, the effect becomes significant as
party polarization reaches moderate levels. In other words, voting deficits for
the president’s party are easier to overcome when the parties are ideologically
similar; the president’s party has greater difficulty reaching compromise with
the opposition as the parties begin to polarize. This interaction between presi-
dential party strength and partisan polarization is just one of a number of
possible insights our more nuanced measure of divided government can
reveal. Our ability to uncover this effect underscores the benefits of using a
measure that captures both the presence and magnitude of inter-branch
conflict.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
788 American Politics Research 43(5)
Considered as a whole, this article adds to our understanding of the effects
of divided government on the legislative process by analyzing how divided
partisan control of government—and the magnitude of that division—affects
the amount of time important legislation is immersed in the legislative pro-
cess. The insights generated by this analysis are important given their impli-
cations for congressional productivity, the efficacy of policymaking
institutions, public trust in government, and others. For example, lengthy
consideration of a particular piece of legislation can result in the issue
monopolizing the legislative agenda, and may ultimately lead to enactment of
a smaller proportion of the congressional or presidential policy agenda over
the course of the legislative session. Future work could examine whether this
theoretical relationship is corroborated empirically. Similarly, delay in the
legislative process may reflect the occurrence of inter-party and/or inter-
branch negotiation over the substance of the bill in question. In such cases, it
seems likely the content of delayed legislation will not reflect the preferences
of any single actor or group in the process—It typically takes time to develop
a bill deemed acceptable to legislators with a broad range of policy prefer-
ences and views. In contrast, the content of legislation passed with relative
expediency seems likely to adhere more closely to the preferences of a single
stakeholder, most likely the president or the majority party in Congress. Of
course, the relationship between legislative delay and alignment of legisla-
tive content with the preferences of various actors in the policymaking pro-
cess is ultimately an empirical question, the investigation of which would
serve as a valuable extension to the results presented in this article.
Finally, future work could usefully build on our investigation of the rela-
tionship between party control, polarization, and legislative delay by analyz-
ing whether we observe heterogeneity in this relationship across policy
topics. For example, it seems likely some policy topics are inherently more
conflictual than others, and divided government would result in substantially
more delay for legislation in these areas, relative to bills addressing less con-
flictual topics. Similarly, it is possible legislation on a particular topic experi-
ences more or less delay depending on partisan control. For example,
legislation on a topic may experience less delay when party in power “owns”
the issue. Empirical investigation of these and other similar issues would
further improve our understanding of the interrelationships between party
control, polarization, and legislative delay.
Acknowledgement
This article would not have been possible without the generous support of the faculty
and staff of the Carl Albert Congressional Research and Studies Center at the
University of Oklahoma.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 789
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.
Notes
1. Additional evidence indicates divided government does not affect productivity
directly but does affect the presence of party unity votes, which indirectly affects
productivity in Congress (Thorson 1998).
2. The data used here were originally collected by Frank R. Baumgartner and
Bryan D. Jones, with the support of National Science Foundation (NSF) Grants
SBR 9320922 and 0111611, and were distributed through the Department of
Government at the University of Texas at Austin. Neither NSF nor the original
collectors of the data bear any responsibility for the analysis reported here.
3. E. Scott Adler and John Wilkerson, Congressional Bills Project (1949-2010),
NSF 00880066 and 00880061.
4. As a point of comparison, approximately 80% of Mayhew’s (1991, 2005)
updated list of important enactments is covered by the operationalization of
important bills used in this article.
5. The number of votes required for cloture changed from two thirds of the chamber
to three fifths in 1975. The data accurately reflect reforms to the cloture rule. For
an explanation of the 1975 cloture reform, see Kroger (2010).
6. The presidential party strength variable is included in the interaction, rather than
the divided government measure, because an interaction between two continu-
ous-level variables presents a more nuanced explanation of this relationship than
including an interaction between the dichotomous divided/unified government
variable and party polarization.
7. Calculated using data from http://www.presidency.ucsb.edu/data/budget.php.
8. As measured by the Policy Agendas Project.
9. Omnibus legislation is identified using data from the Policy Agendas Project.
The coding scheme refers to the final bill that the original policy idea is attached
to. In most cases, this means the introduction of the original bill took place long
before the introduction of the larger omnibus bill, which is different from the
analysis conducted by Krutz (2001). The data set contains 106 bills passed as
part of a larger omnibus package.
10. We specify “Economic Policy” as the excluded category because it is the modal
policy topic in our data. Because a good number of the bills in this category are
appropriations bills, which have historically been considered “must pass” legis-
lation and are subject to special procedural rules, it is reasonable to expect these
bills to experience less delay than bills in other policy areas. That said, we do not
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
790 American Politics Research 43(5)
present any formal hypotheses regarding the relationship between policy topic
and legislative delay.
11. A non-parametric functional form is used because we have no a priori expecta-
tions about the distribution of the hazard function.
12. For ease of interpretation, models displayed in the article use coefficients rather
than hazard rates.
13. The policy-specific factor variables are excluded to save space. These findings
are available on request.
14. The insignificance of marginal effects is characterized by the 90% confidence
interval containing zero.
15. The inclusion of a linear time variable in the models produces nearly identical
results as the party polarization variable. However, it is impossible to include
both variables in the same model due to problems with collinearity.
16. This model was omitted for considerations of space, but the results are available
on request. A slope–dummy interaction—an interaction between a dichotomous
variable and a continuous variable—does not need to be plotted; such an interac-
tion simply measures the difference between two coefficient slopes.
References
Adler, E. S., & Wilkerson, J. D. (2012). Congress and the politics of problem solving.
New York, NY: Cambridge University Press.
Aldrich, J. H. (1995). Why parties? The origin and transformation of party politics in
America. Chicago, IL: University of Chicago Press.
Aldrich, J. H., & Rohde, D. W. (2001). The logic of conditional party government:
Revisiting the electoral connection. In L. Dodd & B. Oppenheimer (Eds.),
Congress reconsidered (7th ed., pp. 269-292). Washington, DC: CQ Press.
Baumgartner, F. R., & Jones, B. D. (1993). Agendas and instability in American poli-
tics. Chicago, IL: University of Chicago Press.
Baumgartner, F. R., & Jones, B. D. (2009). Agendas and instability in American poli-
tics (2nd ed.). Chicago, IL: University of Chicago Press.
Binder, S. A. (1999). The dynamics of legislative gridlock, 1947-96. The American
Political Science Review, 93, 519-533.
Binder, S. A. (2003). Stalemate: Causes and consequences of legislative gridlock.
Washington, DC: Brookings Institution Press.
Binder, S. A., & Maltzmann, F. (2002). Senatorial delay in confirming federal judges,
1947-1998. American Journal of Political Science, 46, 190-199.
Bond, J. R., Fleisher, R., & Krutz, G. S. (2009). Malign neglect: Evidence that delay
has become the primary method of defeating presidential appointments. Congress
& the Presidency, 36, 226-243.
Box-Steffensmeier, J. M., & Jones, B. S. (2004). Event history modeling: A guide for
social scientists. New York, NY: Cambridge University Press.
Brambor, T., Clark, W. R., & Golder, M. (2006). Understanding interaction models:
Improving empirical analyses. Political Analysis, 14, 63-82.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
Hughes and Carlson 791
Chiou, F.-Y., & Rothenberg, L. S. (2007). Comparing legislators and legislatures: The
dynamics of legislative gridlock reconsidered. Political Analysis, 16, 197-212.
Cobb, R. W., & Elder, C. D. (1983). Participation in American politics: The dynamics
of agenda-building. Baltimore, MD: Johns Hopkins University Press.
Coleman, J. J. (1999). Unified government, divided government, and party respon-
siveness. The American Political Science Review, 93, 821-835.
Edwards, G. C., Barrett, A., & Peake, J. (1997). The legislative impact of divided
government. American Journal of Political Science, 41, 545-563.
Fiorina, M. P. (1996). Divided government (2nd ed.). Needham, MA: Allyn & Bacon.
Gilmour, J. (1995). Strategic disagreement: Stalemate in American politics.
Pittsburgh, PA: University of Pittsburgh Press.
Hibbing, J., & Theiss-Morse, E. (1995). Congress as public enemy: Public attitudes
toward American political institutions. New York, NY: Cambridge University
Press.
Howell, W., Adler, S., Cameron, C., & Riemann, C. (2000). Divided government and
the legislative productivity of Congress, 1945-94. Legislative Studies Quarterly,
25, 285-312.
Jones, B. D., & Baumgartner, F. R. (2005). The politics of attention: How government
prioritizes problems. Chicago, IL: University of Chicago Press.
Jones, C. O. (1994). The presidency in a separated system. Washington, DC:
Brookings Institution Press.
Jones, D. R. (2001). Party polarization and legislative gridlock. Political Research
Quarterly, 54, 125-141.
Kroger, G. (2010). Filibustering: A political history of obstruction in the house and
senate. Chicago, IL: University of Chicago Press.
Krutz, G. S. (2001). Hitching a ride: Omnibus legislating in the U.S. Congress.
Columbus: Ohio State University Press.
Krutz, G. S., & Peake, J. S. (2009). Treaty politics and the rise of executive agree-
ments: International commitments in a system of shared powers. Ann Arbor:
University of Michigan Press.
Lee, F. E. (2009). Beyond ideology: Politics, principles, and partisanship in the U.S.
Senate. Chicago, IL: Chicago University Press.
Lewis, J. B., & Poole, K. T. (2004). Measuring bias and uncertainty in ideal point
estimates via the parametric bootstrap. Political Analysis, 12, 105-127.
Mayhew, D. R. (1991). Divided we govern: Party control, lawmaking, and investiga-
tions, 1946-1990. New Haven, CT: Yale University Press.
Mayhew, D. R. (2005). Divided we govern: Party control, lawmaking, and investiga-
tions, 1946-2002 (2nd ed.). New Haven, CT: Yale University Press.
Peake, J. S., Krutz, G. S., & Hughes, T. (2012). President Obama, the senate, and
the polarized politics of treaty making. Social Science Quarterly, 93, 1295-1315.
Poole, K. T., & Rosenthal, H. (1997). Congress: A political-economic history of roll
call voting. New York, NY: Oxford University Press.
Rohde, D. (1991). Parties and leaders in the post-reform House. Chicago, IL:
University of Chicago Press.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
792 American Politics Research 43(5)
Shipan, C. R., & Shannon, M. L. (2003). Delaying justice(s): A duration analysis
of Supreme Court confirmations. American Journal of Political Science, 47,
654-668.
Sinclair, B. (2011). Unorthodox lawmaking: New legislative processes in the U.S.
Congress. Washington, DC: CQ Press.
Theriault, S. M. (2008). Party polarization in Congress. New York, NY: Cambridge
University Press.
Thorson, G. R. (1998). Divided government and the Passage of Partisan Legislation,
1947–1990. Political Research Quarterly, 51, 751-764.
Woon, J., & Anderson, S. (2012). Political bargaining and the timing of congressional
appropriations. Legislative Studies Quarterly, 37, 409-436.
Author Biographies
Tyler Hughes is an assistant professor of political science at California State
University, Northridge. His research focuses on policymaking dynamics within both
Congress and policy subsystems.
Deven Carlson is an assistant professor of political science at the University of
Oklahoma. His research analyzes the operations and effects of public policies, as well
as the politics surrounding policy issues.
at CALIFORNIA ST UNIV NORTHRIDGE on May 10, 2016apr.sagepub.comDownloaded from
... We reduce these complexities to two numbers that reflect the overall costs accruing to the U.S. administration from pursuing its desired policy outcome via either bilateral aid or multilateral aid. 10 The administration's cost of bilateral aid depends on the level of congressional resistance. Gridlock and inaction when the president's party does not control Congress suggests higher costs under divided government (Binder 1999;Hughes and Carlson 2015). We focus on this mechanism in part because it has well defined temporal variation and is straightforward to measure. ...
... A number of alternative measures have been suggested to operationalize the concept of Bdividedness.^More sophisticated alternatives we experimented with include the measure of strength of the president's party suggested by Hughes and Carlson (2015), the ratio of the number of failed bills to the number of total bills, and various measures 12 The point that Congress has more influence (and correspondingly the administration has less influence) in the case of divided government is also made by Raunio (2014, 546) and Raunio and Wagner (2017, 4). 13 We make the weaker claim that the results of previous research are driven primarily by years with divided U.S. government, not the stronger claim that there is no U.S. influence in years of undivided government or that U.S. influence is significantly stronger in years of divided U.S. government than in years of undivided U.S. government. ...
Article
Full-text available
Do U.S. presidential administrations exert more informal influence over international financial institutions when they face an uncooperative Congress and thus have less control over bilateral aid? Reexamining four empirical studies of the World Bank, we demonstrate that U.S. informal influence is driven by years with divided U.S. government. This provides a richer picture of when and why the U.S. exerts influence in multilateral settings and an alternate explanation to persistent questions about the role of international organizations in the international political economy.
... 27 Also on measurement, some scholars have looked at delay rather than the number of laws passed in congress. There is some recent compelling evidence that divided government causes delay in important legislation (Hughes and Carlson 2015) and, more specifically, in the budget process at the state level in the US (Kirkland and Phillips 2018). However, looking specifically at welfare reforms, Bernecker (2016) finds that divided US state governments are more likely to implement reforms than unified governments. ...
Article
Full-text available
Separation of powers with checks and balances (SP) is usually regarded as a key institution complementing elections in the control of elected officials. However, some analysts and many politicians also warn that excessive checks on the executive in the presence of polarization may lead to political inaction. We analyze the interaction between elections and SP, and study under what circumstances they complement each other. We first introduce a political agency model with ideological parties where citizens and politicians care about rents (a valence issue) and policy (a positional issue). Then, we analyze the impact of SP on the effectiveness of elections to discipline and select politicians. We demonstrate that SP unambiguously raises a majority of voters’ welfare in highly polarized non-competitive political environments, because it strengthens both discipline and selection without causing political gridlock. SP also raises voters’ welfare if elections are very effective at disciplining first period incumbents. Nevertheless, SP may reduce voters’ welfare if most rents go undetected and reform is not a first-order issue.
... executive and congressional branches are controlled by different political parties) legislate important bills as quickly as a cohesive government, and therefore government governs regardless of the unity of the government-a more constrained institution has no effect. However, later research from Edwards et al. (1997) and Howell et al. (2000) provide a rebuttal of Mayhew's thesis, whereas, subsequent findings are still mixed on the effect of divided government (Adler & Wilkerson, 2012;Baumgartner et al., 2013;Hughes & Carlson, 2015). By and large, a slower policymaking process is expected in a divided government as there is greater bargaining cost. ...
Thesis
This thesis considers how political liberalisation, as a dynamic process, affects policy processes in authoritarian regimes. Prior studies observed either the bargaining process or the information exchange process but not both. These studies consider them as two distinct and competing theoretical perspectives. First, the bargaining perspective asserts that the process of political liberalisation, with the introduction of more inclusive and competitive elections, increases the bargaining costs and make policymaking more difficult. Second, political liberalisation increases the social and political freedom and it enhances information exchange that facilitates policymaking. However, these two processes contradict with each other and create a theoretical puzzle that requires a systematic theoretical and empirical investigation. Building on the punctuated equilibrium theory, this thesis offers a novel bargaining/information exchange hybrid theoretical model to explain policy processes during a period of political liberalisation. Rather than treating the two perspectives as competing explanations, this thesis integrates them by recognising the duality of the electoral system—it is a source of political bargaining as well as a source of information exchange. The liberalisation of the electoral system increases the likelihood of transforming the authoritarian regime from a one-party or one-party-dominant system to a multiparty system. An increased number of political parties in the policy processes intensifies the bargaining process but also increases the information exchange. As such, the changes in bargaining and information exchange processes happen simultaneously as regimes liberalise. The hybrid model contributes to advancement in the field with this refined way of thinking about policymaking during a period of political liberalisation. It captures the complexity and non-linearity of the social and political world. Using a novel and original time-series dataset of legislative bills of the Hong Kong Legislative Council, the analysis adopts the policy content coding system of the Comparative Agendas Project (CAP) to measure legislative attention from 1975 to 2016. Policy processes are measured by legislative speed (i.e., the duration of the legislative process), distribution of policy change (i.e., changes in policy contents in the legislative bills) and issue diversity of the policy agenda (i.e., the concentration of policymakers’ attention across different policy issues). Political liberalisation is measured by the degree of inclusive and competitive elections, while the number of political parties is measured by the effective number of political parties. This thesis uses stochastic process methods to measure distributions of policy change, determining whether policymaking is characterised by a pattern of mostly stable and occasional periods of rapid and radical policy change. Further, it uses time-series analysis and event history analysis to estimate the impact of the number of political parties on policy processes. This thesis has a number of important findings. First, an increase in the number of political parties brings in more information and greater bargaining costs. This effectively slows down the legislative process. Second, it finds that the distribution of policy change is punctuated—characterised by long periods of stasis but also a more frequent occurrence of radical changes—rather than incremental change in Hong Kong’s legislative agenda and it confirms the expectation of the punctuated equilibrium theory. However, as the regime liberalises, the distribution of policy change becomes less punctuated, meaning that there are more frequent and moderate policy adjustments. It suggests that a greater number of political parties facilitates policy change and can be explained by the information model. However, comparing Hong Kong’s distribution of policy change with other advanced democracies provides support for the bargaining model. These puzzling findings show a more inclusive bargaining/information exchange hybrid model is needed to explain policymaking in countries with different degrees of political liberalisation. Third, the study of issue diversity of the policy agenda shows an “inverted-U” curve as the number of political parties increases. It implies that autocrats pay more attention to a wider range of policy topics when the number of political parties increases from a low level whereas the attention to different policy topics shrinks when the number of political parties escalates to an extremely high number. The empirical evidence provides unambiguous support for the bargaining/information exchange hybrid model. Overall, this thesis builds on the punctuated equilibrium theory by offering a novel bargaining/information exchange hybrid model to understand policymaking and tests it systematically with new empirical evidence. It contributes to the understanding of policymaking in authoritarian regimes. It also provides an important insight into the duality of political and other social processes in comparative politics.<br/
... A substantial body of research investigates the ways in which political parties and governments respond to the public (e.g., Adams, Clark, Ezrow, & Glasgow, 2004;Lax & Phillips, 2009). Other studies examine why government becomes gridlocked, that the likelihood of changing an existing policy is markedly reduced or eliminated (e.g., Hughes & Carlson, 2015;Klarner, Phillips, & Muckler, 2012). Still other scholars suggest that the existence of punctuated equilibrium policy dynamics; i.e., policies experiencing long periods of stability might be interrupted occasionally by a short period of large changes and then switch to a new equilibrium (e.g., Baumgartner & Jones, 1993). ...
Article
Full-text available
What factors cause policies experiencing long periods of stability to be interrupted occationally by a short period of large changes? This study argues that electoral incentives might influence the search, supply, and processing of information on constituency issues, as well as the associated cognitive or institutional frictions, and thus determine the presence and variation of punctuated policies. This article develops and evaluates this claim within a systemic framework consisting of policy transparency, political institutions, and electoral incentives. For the purpose of identifying policy punctuations, this research uses the Generalized Pareto Distribution in the Extreme Value Theory. This study analyzes budget spending data collected from FY 1988 to FY 2008 for all 50 American states. This study finds that greater policy transparency is associated with larger spending stability. By contrast, greater gubernatorial competition is more likely to produce extreme spending changes. Electoral incentives shaped by public preference and political term limits have a profound impact on nonincremental policy changes. The impact of policy transparency is conditional on public preference, while that of electoral competition and legislative professionalism is moderated by political term limits. Particularly, a transparent policy consistent with public preference and legislative professionalism with term limits are more likely to give rise to punctuated policies, while gubernatorial (legislative) competition leads to less punctuated changes when governors (legislators) are subject to term limits. 哪些因素导致长期稳定的政策偶尔被短期的重大变动所打断?本研究认为,选举激励可能会影响人们对选区问题信息的搜索、提供和处理,以及相关的认知性或制度性摩擦,从而决定间断性政策的发生和变化。本文采用由政策透明、政治制度和选举激励三者组成的系统框架来发展并评估这一论断。本研究使用极值理论(EVT)中的广义帕累托分布(GPD)来识别政策间断点。我们分析了美国50个州从1988到2008财政年度的预算支出数据。研究发现,较高的政策透明度与较大的支出稳定性是相关联的。相比之下,激烈的州长选举竞争更有可能带来支出的急剧变化。选举激励受到公共偏好和政治任期的影响,进而对政策的非渐进性变化造成很大影响。政策透明对政策变化的影响取决于公众偏好,而选举竞争和立法专业化对政策变化的影响则受任期限制的影响。具体来说,符合公共偏好的透明政策以及具有任期限制的立法专业特性更有可能带来间断性政策,而当州长(立法者)受限于任期制时,州长(立法者)竞争会带来较少的间断性政策变化。
... We do not re-examine Kilby 4 There is a plethora of alternative approaches. More sophisticated measures we experimented with include the measure of strength of the president's party suggested by Hughes and Carlson (2015), the ratio of the number of failed bills to the number of total bills and various measures of polarization and/or ideological distance based on Poole and Rosenthal's NOMINATE scores (see Poole and Rosenthal 1997). The results using these more complicated measures are similar but less clear, leading us to prefer the simplest option. ...
Working Paper
Full-text available
Do U.S. presidential administrations exert more informal influence over international financial institutions when they face an uncooperative Congress and thus have less control over bilateral aid? Reexamining four empirical studies of the World Bank, we demonstrate that U.S. informal influence is driven by years with divided U.S. government. This provides a richer picture of when and why the U.S. exerts influence in multilateral settings and an alternate explanation to persistent questions about the role of international organizations in the international political economy.
... On the other hand, the Israeli Knesset is almost the mirror image of the American Congress when it comes to gridlock. The American federal legislative process (with features such as bicameralism, presidential veto, and the filibuster) is particularly cumbersome and veto-gate laden (Eskridge 2012;Oleszek 2014), and susceptible to gridlock especially in times of divided government and polarized parties (Davis et al. 2014;Hughes & Carlson 2015;Levinson & Pildes 2006). In contrast, the Israeli unicameral legislature, characterized by a lack of constitutional restrictions on its legislative procedure and very lax parliamentary rules (which, for example, do not even include a minimal quorum requirement and are very permissive in allowing procedures that facilitate lawmaking even further, such as expedited omnibus legislation), creates relatively low hurdles on the ability of the coalition government to pass its legislative agenda (Bar-Siman-Tov 2016). ...
Article
Full-text available
This article presents the findings of an extensive multi-method empirical study that explored the relationship between temporary legislation, better regulation, and experimentalist governance. Temporary (or “sunset”) legislation, statutory provisions enacted for a limited time and set to expire unless their validity is extended, is often hailed as a key tool for promoting experimental and better regulation. Despite the importance of temporary legislation and the burgeoning theoretical scholarship on the subject, there is still a dearth of empirical studies about how temporary legislation is used in practice. The lack of empirical evidence creates a lacuna in at least three areas of theoretical scholarship, concerning temporary legislation, better regulation, and experimentalist governance. This paper is a first step in filling this gap.
Article
I consider a dynamic model of bargaining where alternatives to the status quo arrive stochastically during the bargaining process, the proposer can bundle multiple alternatives into a single proposal, and a forward-looking voter elects the agenda-setter. I show that the prevailing wisdom that policy bundling reduces gridlock—by facilitating compromise across different policy areas—is incomplete. Policy bundling can also increase gridlock: a player may veto or delay a bipartisan alternative, which is unanimously preferred to the status quo, so that in the future they can bundle this same alternative with a divisive alternative that otherwise would not pass. Gridlock of this form is more likely to occur during periods of economic or political stability and, when it occurs, suggests that traditional measures of legislator ideology will overstate polarization. From the voter’s perspective, gridlock occurs at an inefficiently high frequency. This state of “excess gridlock” is driven by the voter being forward-looking and lacking commitment power.
Article
This paper examines the impact of governmental unity on corporate performance. Calendar years are classified as united when the president, Senate majority, and House majority are all from the same political party. Results indicate that firms have higher accounting return on assets when the federal government is united than when it is not. Firms have higher accounting return on assets when federal and state governments are both united under the same party, though the structures of federal and state governments may differ in the same time episode. The positive effect of government unity on corporate performance is stronger when the firm's sales growth rate is greater and profitability more volatile. The results are consistent with united government promoting legislation efficiency and reducing uncertainty in political environments.
Article
This study examines whether and how legislative gridlock in Congress affects legislative uncertainty. Results show that legislative gridlock increases cross-sectional dispersion in returns on individual stocks and portfolios over short windows around roll-call votes on Congressional bills. The effect of gridlock is beyond political polarization, Presidential ideology, unified versus divided governments, economic recession, policy uncertainty (Baker et al. 2016), and macroeconomic uncertainty (Jurado et al. 2015). Greater legislative uncertainty is associated with higher bond yield and lower likelihood of seasoned equity offerings and analyst recommendations. The effect of gridlock depends on industry and firm characteristics, as well as political polarization and Presidential ideology. Overall, legislative gridlock increases legislative uncertainty as reflected in cross-sectional dispersion in stock returns.
Article
Full-text available
This article investigates how parties affect legislative gridlock - the inability of government to enact significant proposals on the policy agenda. Conventional accounts suggest that divided party control of government causes such stalemate. I offer an alternative partisan model of gridlock that incorporates party polarization, party seat division, and the interaction between these two factors. Using an original data set of major legislative proposals considered between 1975 and 1994, I find that divided government does not affect gridlock once party polarization and party seat division are taken into account. Instead, I find that higher party polarization increases the likelihood of encountering gridlock on a given proposal, but that the magnitude of this increase diminishes to the extent that a party is close to having enough seats to thwart filibusters and vetoes.
Article
Presidents traditionally have had great success when nominating justices to the Supreme Court, with confirmation being the norm and rejection being the rare exception. While the confirmation process usually ends with the nominee taking a seat on the Court, however, there is a great deal of variance in the amount of time it takes the Senate to act. To derive a theoretical explanation of this underlying dynamic in the confirmation process, we draw on a spatial model of presidential nominations to the Court. We then employ a hazard model to test this explanation, using data on all Supreme Court nominations and confirmations since the end of the Civil War. Our primary finding is that the duration of the confirmation process increases as the ideological distance between the president and the Senate increases. We also find evidence that suggests that the duration increases for critical nominees and chief justices and decreases for older nominees, current and previous senators, and nominees with prior experience on state and federal district courts.
Book
Here is an accessible, up-to-date guide to event history analysis for researchers and advanced students in the social sciences. The foundational principles of event history analysis are discussed and ample examples are estimated and interpreted using standard statistical packages, such as STATA and S-Plus. Recent and critical innovations in diagnostics are discussed, including testing the proportional hazards assumption, identifying outliers, and assessing model fit. The treatment of complicated events includes coverage of unobserved heterogeneity, repeated events, and competing risks models. The authors point out common problems in the analysis of time-to-event data in the social sciences and make recommendations regarding the implementation of duration modeling methods.
Article
For the past twenty-five years concluding essays in volumes like this one have addressed the prospect of electoral realignment. No more. Realignment — at least as classically conceived — appears to be a dead concept. Since 1964 we have waited for the realignment; and each time the academic consensus has announced, ‘No, that wasn’t it’. But, classical realignment or not, American politics have been transformed in the past quarter century. In particular, we seem to have settled into a persistent pattern of divided government. Republican standard-bearer George Bush won an easy victory in 1988, carrying forty states with 54 per cent of the popular vote. At the same time, Democratic candidates took 260 of the 435 seats in the House of Representatives, and won 18 of the 33 Senate races, leaving that body too under Democratic control. This split outcome surprised no one; three of the previous eight presidential elections (1956, 1968, 1972) produced the same split, and two more (1980, 1984) placed the Presidency and Senate at odds with the House of Representatives. Of the past nine presidential elections, only three (1960, 1964, 1976) have given one party control of the Presidency and both Houses of Congress.
Article
In this prize-winning book, a renowned political scientist debunks the commonly held myth that the American national government functions effectively only when one political party controls the presidency and Congress. For this new edition, David R. Mayhew has provided a new Preface, a new appendix, and a new concluding chapter that brings the historical narrative up to date.
Article
Political observers have remarked on a staggering increase in the length of time it takes for the Senate to confirm presidential appointees to the lower federal bench. Here, we focus on the duration of the confirmation process for presidential appointees to the United States Circuit Courts of Appeal between 1947 and 1998 and explain the variation over time in the length of the confirmation process. With newly collected data on the fate of all appellate nominees during that period, we show how ideological incentives and institutional opportunities combine to affect the timing of Senate confirmation of judicial appointees.