Parties,Preferences,and Congressional Organization

Abstract

When does Congress repeal laws enacted by prior generations? Although the substantial body of work on policy creation provides tentative explanations, we believe repeals represent an alternative way of examining the effects of congressional organization on legislative behavior. In this paper, we develop hypotheses based on both the conditional nature of party power and the location of pivot points, and test these hypotheses with a new data set of repeals from 1877 to 2012. We find that the largest effects on Congress’s capacity to repeal legislation are variation in the majority’s positive agenda control and shifts in the gridlock interval. We also find that when the majority claims control of both chambers after a long stretch in the minority, there is an increased likelihood of repeal beyond what is predicted by conditional party government alone. Because the partisan factors in our model have the largest substantive effects, and because repeals do not occur automatically in productive Congresses, we characterize repeals as long-term contests between two great “teams” over the location of the status quo.

Keywords

U.S. Congress legislative studies landmark legislation policy repeal political parties conditional party government Affordable Care Act

Introduction

When does Congress repeal legislation enacted by prior generations? Although law creation has garnered significant attention (e.g., Baumgartner and Jones 1993; Binder 2003; Cox and McCubbins 2007; Krehbiel 1992; Mayhew 1991), instances in which Congress attempts to reverse legislation have received little consideration. Given that one of the defining features of the 113th and 114th Congresses (2013–2016) has been more than fifty votes to repeal the Affordable Care Act (ACA), we believe this a neglected area in the American politics literature.

We fill this gap with an original data set of all major repeals from 1877 to 2012. In particular, we catalog major repeals following Mayhew (1991, 2005), Howell et al. (2000), Clinton and Lapinski (2006), and others who compile lists of landmark enactments. Like proposals to repeal the ACA, the repeals we identify are some of the most contentious and long-running efforts to shape national policy. In total, we cataloged eighty-nine major statutory repeals in the 135-year time span of our study (an average of 1.3 major repeals per congressional session). Examples include the repeal of multiple New Deal statutes in the 1990s, dramatic fluctuations in monetary policy in the 1890s, the repeal numerous tax statutes in the 1920s, and the repeal of the Chinese Exclusion Acts in the 1940s.

As well as being substantively important yet understudied, we believe there are both analytical and theoretical reasons to study repeals. Analytically, repeals permit a longitudinal study of Congress that is unique. Unlike measures of policy enactment, which are static, repeals allow us to compare lawmaking in two time periods and isolate the effects of long-term changes in institutional structures and coalitions. And theoretically, we argue that the causes of repeal differ from those that explain policy creation. For this reason, although we think shifts to the left or right in Congress’s membership—known in the literature as shifting pivot points—are an important cause of repeals, we focus on the ebb and flow of party strength and the majority’s electoral success as key aspects to when and why repeals happen. Indeed, we argue that with repeals, the majority needs more than the institutional position to block a bill’s passage: it requires the capacity to build a coalition, shepherd a bill to the floor, and win a repealing bill’s enactment.

But while repeals receive little attention in the literature, a handful of studies shed light on this neglected issue, finding that Congress modifies prior enactments due to a broad mix of institutional and exogenous factors (Adler and Wilkerson 2012; Berry, Burden, and Howell 2010; Maltzman and Shipan 2008; Ragusa 2010). We build on this work by offering an account of how the distribution of preferences within Congress affects the likelihood of repeal. And where the above studies examine the latter half of the twentieth century, the 135-year time span of our data coincides with a number of important changes in party organization and institutional structures that are unexamined in existing work. We then model repeals using survival analysis, finding that shifts in the size of the gridlock interval from the enacting Congress—that is, the distribution of preferences need to avoid a veto or filibuster—are a necessary but not sufficient condition for repeal. Instead, we find that repeals are most likely when the majority party has been out of power for an extended period and the parties are ideologically distinct and homogeneous.

Beyond offering new insights into when Congress will “undo” enactments, our approach informs the study of agenda control while broadening the focus beyond rule making. Indeed, Aldrich and Rohde (2001) and Cox and McCubbins (2005, chap. 10) contend that positive agenda control—the ability to create legislation—ebbs and flows over time, whereas negative agenda control—the ability to prevent legislation from passing—is invariant. Our results support this notion, finding that repeals are best explained by the tenets of conditional party government—that is, when the majority is ideologically cohesive and the parties are polarized—and are largely a product of variation in the majority’s positive agenda control (Aldrich 1995; Rohde 1991). Indeed, our main finding is that repeals do not occur uniformly over time, nor automatically in productive Congresses, but disproportionately in periods of party transitions and a homogeneous majority. We believe this is consistent with the intuition of Finocchiaro and Rohde (2008, pg. 42) in their description of the Republican takeover of the House in 1994. As they note,

Majority members during the Republican Revolution sought to reverse policies that 40 years of Democratic House control had produced . . . Negative agenda control was of relatively little use; positive agenda control was of supreme import. (Finocchiaro and Rohde 2008)

Congress and Statutory Reversal

Contrary to Kaufman’s (1976) Are Governmental Organizations Immortal? we know the government regularly adjusts or terminates existing legislation and federal programs. Adler and Wilkerson (2012); Berry, Burden, and Howell (2010); Patashnik (2008); Maltzman and Shipan (2008); and Ragusa (2010) each examine legislative adjustment across various dimensions, whereas others have focused on particular policy areas (Corder 2004; Lewis 2002) and the life cycle of federal agencies (Carpenter and Lewis 2004). Although key differences exist, each study examines legislative reversals in an “ecumenical” fashion, exploring changes in the broader political environment, variation in institutional conditions, and bill-level factors.

Among Congress-specific factors, divided government and bicameral differences have robust effects (though this is not universal, as we discuss below). Maltzman and Shipan (2008) and Ragusa (2010) find that the greater the distance between the House and Senate, the more durable the legislation (see also Binder 1999). However, there are conflicting accounts as to whether divided government leads to legal stability (Carpenter and Lewis 2004; Maltzman and Shipan 2008; Ragusa 2010) or vulnerability (Berry, Burden, and Howell 2010). Furthermore, Maltzman and Shipan find that a major amendment to a bill is more likely if it is passed under divided government, whereas Ragusa finds mixed effects: divided government creates a short-term risk of repeal but long-term stability. In addition, the literature identifies polarization, changes in majority party seat share, and shifts in chamber medians as factors that explain legislative revision as well (Berry, Burden, and Howell 2010; Ragusa 2010).

In a recent book, however, Adler and Wilkerson (2012) challenge the view that policy change is conflictual—a consequence of changing parties and pivots—but is, instead, solution driven. Unlike the studies above, they find that macropolitical conditions do little to explain the volume of policy change. Nonetheless, Adler and Wilkerson also find that macropolitical conditions do help predict change when limiting their analysis to the first instance in which a historic law was amended in a major fashion. As the focus of our paper is on major statutory change, rather than a more expansive definition of policy change, we expect to find that macropolitical conditions—including changes in parties and preferences—play a key role in this process. We elaborate these expectations in the section below.

Although not addressed in the existing literature, we note two reasons why repels are distinct from routine policy creation. First, and as we explain in more detail below, repealing legislation requires time and effort in an environment where both are scarce. Indeed, Congress exerts considerable effort to simply enact a bill, only for a subsequent Congress to exert as much or more effort to repeal that same bill at a later date. Given all the potential items on the legislative agenda, and the scarcity of plenary time (Rohde, Stiglitz, and Weingast, 2013), for members of Congress to prioritize repealing statutes is indeed unique. And second, repealing statutes goes against the grains of norms of legal stability. In other words, Congress has an institutional incentive to uphold the legal environment—thus limiting conflicting statutes—which further increases the difficulty of repealing legislation. As a whole, this suggests that repeals are not only distinct from other forms of policy change, but they may be harder as well to achieve than other forms of legislation. For this reason, we consider repeals to be a “tough test” of a governing party’s positive agenda control.¹

Parties and Agenda Control

Prominent theories of parties in Congress focus on the majority’s agenda setting capacity. Positive agenda control is exercised when the majority works to pass policies that are favored by the majority of its members. Negative agenda control, by comparison, is achieved when party leaders “gatekeep” access and prevent divisive legislation from reaching the floor. Although negative and positive agenda control are related, theories differ on when each form of control is exercised. While the former is invariant, as the majority party has relatively constant control over the movement of bills to the chamber floor, the latter depends on the ideological characteristics of the majority party.

As elaborated by Cox and McCubbins (1993, 2005, 2007), political parties behave as a “legislative cartel” that derive their authority by manipulating the “structural power of the House” (Cox and McCubbins 2005, 15). In this respect, the majority exercises negative agenda control by procedurally “stacking the deck” in its favor. Following Cox and McCubbins, the majority party’s power is (nearly) constant, with the majority party regularly able to keep any bill off the floor that a majority of the majority party is opposed to.² At the other end is a party’s ability to bring its preferred policies into effect. Unlike negative agenda power, which is largely invariant to the ideological characteristics of the majority, positive agenda power is contingent.³ When parties are homogeneous and ideologically distinct, the majority’s rank-and-file delegates greater authority to the leadership which, in turn, exploits institutional rules and powers to ensure members of the party act in a way consistent with their collective goals (Aldrich and Rohde 1997). In one study, Finocchiaro and Rohde (2008) demonstrate that the minority is more likely to be rolled in rules-related votes when the majority party is strong, but not when they majority party is weak. As a whole, while conditional party government and the cartel theory emphasize different mechanisms, they both expect that the party will occasionally allocate resources toward positive powers.

In more recent scholarship, Rohde, Stiglitz, and Weingast (2013) contend that the application of agenda setting power is contingent on plenary time. Given the persistent demands on the agenda, legislators must prioritize some issues over others. As the range of compulsory issues is expansive, members of Congress will have a limited number of discretionary issues they are able to address (Walker 1977). This “agenda scarcity” forces members of Congress to consider exactly which discretionary items to prioritize (Adler and Wilkerson 2012). Rohde, Stiglitz, and Weingast (2013) contend that homogeneous parties that have recently claimed control are more likely to centralize positive agenda control in the hands of leaders, as the rank-and-file are more likely to agree on the prioritization of discretionary items. When the parties have been in power for some time, and there is less agreement on discretionary items, agenda control will be decentralized. As a whole, therefore, this work suggests that the party’s willingness and ability to exercise positive or negative agenda control depends on the majority’s recent electoral record.

Given the above discussion, we argue that summary measures of partisanship, such as divided government and size of the majority’s coalition, do not tell the entire story about when repeals occur. With repeals, lawmakers need more than the power to prevent a bill’s passage: repeals require the capacity to build a coalition and win a repealing bill’s enactment. And because the status quo is notoriously resistant to change,⁴ we focus on party strength as critical to explaining when repeals occur. In the end, we expect that a majority’s ability to enact repeals depends (largely) on their ideological cohesion. Consistent with Aldrich and Rohde (1997), Finocchiaro and Rohde (2008), and others, we believe party leaders will be willing to loosen negative control and allow its members to bring repeal efforts forward when the party is homogeneous. When the party is large, but ideologically heterogeneous, we expect that negative agenda control will be the primary tool of majority. We also hypothesize that the effect of conditional party government is itself “conditional.” We believe that partisan transitions during periods when the parties are ideologically distinct and homogeneous are likely to yield an increase in repealing activity beyond what is predicted when there was a change in power or the new majority is homogeneous. Rather, repeals are most likely in periods when a new majority seizes control after a long stint out of power and when the new majority is homogeneous (Rohde, Stiglitz, and Weingast 2013).

Cataloging Historical Repeals from 1877 to 2012

To properly evaluate the effect of changes in parties and preferences on repeals, and given the gradual shift in legislative coalitions, a broad time period of inquiry is necessary. For this reason, we consider repeal activity over a 135-year period, from 1877 to 2012. Although several authors have complied lists of “landmark legislation” (e.g., Howell et al. 2000; Mayhew 1991, 2005), including lists dating back to 1877 (e.g., Clinton and Lapinski 2006), no publicly available database of historical repeals exists. We therefore follow those who catalog major legislation, drawing on both “contemporaneous” and “retrospective” sources of legislative activity.

We analyzed six contemporaneous volumes, focusing on legislative “session wraps” that detail the major events of a congressional session. The historical New York Times and Washington Post constitute our primary contemporaneous sources. For the period 1947 to 2012, we use the session wraps identified by David Mayhew. For the period 1877 to 1946, we extended Mayhew’s data collection effort and culled both newspapers’ archives for session wraps. We further identified session wraps in four supplementary volumes: Political Science Quarterly (1887–1925), the American Political Science Review (1919–1929), Western Political Quarterly (1949–1967), and Congressional Quarterly (1949–2012). We also analyzed nineteen retrospective volumes consisting of a mix of period-specific histories and reference volumes. In particular, we cataloged eleven period histories in the New American Nation series (see also Clinton and Lapinski 2006) as well as a series of public policy reference volumes, including Stathis’ “Landmark Legislation,” Landsberg’s “Major Acts of Congress,” and the “Oxford Encyclopedia of American Political and Legal History.”

All sources are documented in Table A1, as part of an online appendix (http://prq.sagepub.com/supplemental/). We scanned each source for mentions of repealed policies.⁵ We were conservative here, cataloging the literal word “repeal” and its closest derivations (“repealed,” “repealing,” etc.) rather than synonyms such as “abrogated,” “rescinded,” “undid,” or “killed.” Because we think repeals are conceptually unique, we exclude alternative ways Congress can reverse legislation (discussed below).

As a matter of clarification, in popular discourse, repeals are associated with a specific law (e.g., “repealing Obamacare”). In contrast, successful repeals (almost always) target major statutes within a law. Although such changes may be viewed as mere “partial repeals,” statutory repeals represent dramatic changes in the laws of the United States. Below, we take steps to validate that these statutory repeals are indeed of major consequence.

We note alternative ways Congress can “undo” existing statutes. Perhaps most importantly, legislation often contains “sunset provisions” where specific statutes are allowed to simply expire. Adler and Wilkerson (2012) note that expiring provisions are one of the most important factors that compels Congress to act in a particular policy domain, and that partisan and ideological characteristics have little to no effect. Maltzman and Shipan (2008) also show that sunsets increase the likelihood that a law will be amended after passage. Although clearly associated with policy death, we contend that these provisions are distinct from repeals on theoretical grounds. Unlike repeals, we contend that sunsets are (largely) the product of negative agenda control. Given the scarcity of plenary time and political capital resources, it is unlikely that a party would exercise positive agenda power and work to hasten the death of a bill with expiring provisions. At the same time, however, a sunset provision can alter a party’s overall agenda, as it would allow them to allocate resources toward additional items on the agenda, rather than focusing on repealing a particular statute that will expire. In the end, while sunset provisions are more closely related to negative agenda control, we think they can have implications for positive agenda control as well.

Alternatively, lawmakers may bypass statutory approaches to “undoing” legislation, preferring instead to adjust an agency’s funding or choosing to ignore procedures for the agency’s implementation of a program. Both of these actions can have important implications for the future of legislation. As with sunsets, however, we contend that defunding efforts are theoretically district. Simply put, undoing a program’s funding level requires a mix of positive and negative agenda control depending on whether the program is funded with mandatory or discretionary spending. For these reasons, coupled with the existing work on program budgets (e.g., Berry, Burden, and Howell 2010), we limit our analysis to statutory repeals as a particular type of legislative reversal.

In cataloging repeals from 1877 to 2012, we identify the enacting and repealing law. Identifying the repealing law was labor intensive but straightforward; our sources almost always provide the repealing law’s name or its bill number. Identifying the repealed law, however, posed more challenges. First, the contemporaneous and retrospective sources rarely identify the repealed law by name. For example, a 1902 Washington Post article mentions a bill “repealing the war taxes.” Finding an unnamed “war taxes bill” required considerable effort. Second, a number of repealing laws repeal various statutes enacted in different periods. For example, a 1895 New York Times article mentions that the fifty-third Congress repealed “the remaining vestiges of the reconstruction federal election laws.” The law in question, “Federal Election Laws Repeal Act,” repealed parts of three separate laws enacted from 1870 to 1871 known collectively as the Reconstruction-era “Force Acts.”

We believe our data set of eighty-nine statutory repeals indeed represents the universe of major repeals. Although journalists may use alternative terminology to describe repealing activity, these idiosyncrasies should be negated by the breadth of our source material. More importantly, although the data collection follows Mayhew’s (1991, 2005) approach, we took additional steps to validate the significance of each observation. Using the law and bill numbers of the repealed and repealing laws, we merged each observation with Clinton and Lapinski’s (2006) data set of significant legislation. Fortunately, because our sources are almost identical to Clinton and Lapinski’s, both lists should be directly comparable. We consider a repeal “significant” if either the repealing or the repealed law was among Clinton and Lapinski’s top 3,500 statutes (pp. 246–247).⁶ After examining both lists, Clinton and Lapinski’s data affirm the overall significance of our repeals data set; only eight repeals were deemed insignificant.⁷ We therefore removed these observations from the data set, leaving eighty-nine landmark repeals.

Figure 1 reveals the timing of repeal, measured as the number Congresses from passage to repeal. We find that the median time until repeal is 4.5 Congresses and the mean is about 8 Congresses. Just over half of repeals occur within ten years of a law’s enactment. The distribution of repeals presented in Figure 1 show a right skew, due to a few durable outliers: policies repealed many decades after enactment, such as the “Estate Tax,” which existed for eighty-five years before being repealed by Republicans in 2001.

Figure 1.

The timing of repeal.

A Brief History of Repeals: 1877 to 2012

Figure 2—the number of repeals per Congress—demonstrates that the frequency of repeal is not constant, but rather there are several periods of significant repealing activity. We believe the timing of these periods lend initial support to our paper’s core arguments about the interaction between parties and preferences in explaining the timing of repeal. In this section, we provide some historical context for the patterns in Figure 2 and discuss a few notable repeals in our data set. In total, we cataloged eighty-nine repeals from 1877 to 2012.

Figure 2.

Volume of repeals enacted by Congress.

An increase in repeals takes place from the fifty-first to fifty-third Congresses (1889–1895), a period marked by ideologically distinct and homogeneous parties and the implementation of “Reed Rules” in the House. During these six years, both parties engaged in significant repealing behavior. In the fifty-first Congress (1889–1891), Republicans regained control of the House and Senate for the first time in fourteen years and successfully passed two laws of note: the “Sherman Silver Purchase Act” and the “McKinley Tariff Act.” Both laws contained important repeals. However, key elements of both laws were themselves repealed in the fifty-third Congress (1893–1895), when Democrats regained power. In addition, because the fifty-third Congress was the first after Reconstruction where Democrats controlled both chambers, the Reconstruction-era “Force Acts” were repealed.

According to the data, the next spike in repeals begins in the early 1920s from the sixty-sixth to sixty-ninth Congresses.⁸ As before, the high volume of repeals occurs at a time when the two parties were ideologically distinct and homogeneous. The bulk of repealing activity starts during the sixty-sixth Congress (1919–1921) when Republicans regained unified control of both chambers after eight years out of power.⁹ What followed was a period of Republican dominance and a flurry of repeals, with many focused on taxes adopted by Democrats in the 1910s.

From 1933 to the World War II, there is another uptick in repeals. While the New Deal era is often associated with a demand for new legislation, our data show this was a period of significant statutory reversal as well. While the two parties were somewhat heterogeneous, the Democrats had controlled just one chamber over the past decade prior to the seventy-third Congress (1933–1935). For this reason, the activity fits the characterization of the “conditional” effect of conditional party government: where the transition from a long period of Republican dominance to a long period of Democratic dominance was paramount.¹⁰ Among the repeals in this period, Congress passed a resolution in 1933 that repealed the gold standard and abrogated all gold clauses.

From World War II until the 1990s, we would expect few repeals due to the Democrats’ electoral dominance and the low levels of conditional party government, and this is the broad pattern in the data. Upon closer inspection, it seems that the modest bump in the repealing activity from the eighty-ninth to ninety-second Congresses (1965–1971) is anomalous in content (mostly a response to the Cold War) and is not the product of long-term partisan conflict. For example, Congress repealed key elements of the “Subversive Activities Control Act of 1950,” enacted during McCarthy’s “Red Scare.”

And finally, we observe a modest increase in repeals coinciding with the Republican Revolution. In this period, the conditional party government (CPG) indicators are high, Republicans had not held unified control of Congress in forty years, and they targeted a number of Democratic policies. For example, welfare reform was a cornerstone of the “Contract with America,” and with the 1996 repeal of “Aid to Families with Dependent Children”—initially passed in during the New Deal—Republicans realized their goal of changing both the nature and delivery of welfare benefits in the United States. Similarly, the “Gramm–Leach–Bliley Act,” passed in 1999, repealed the New Deal-era “Glass–Steagall Act” and several provisions of the “Bank Holding Act of 1956,” both of which were Democratic policies that established barriers between banking, investment, and insurance companies.

Comparing Enactment and Subsequent Coalitions

We claim that analyzing repeals is worthwhile because, in addition to being understudied, repeals are structured by a different mix of factors compared with law creation. Furthermore, because repeals permit us to compare enactment and subsequent coalitions, we believe there is a methodological advantage as well. While research on policy creation (e.g., Baumgartner and Jones 1993; Mayhew 2005) offers significant insights, these studies have an inherent limitation given that major congressional events concern not just the layering of new policies on top of existing policies but also the efforts of two great “teams” to undo the gains of their predecessors.

In the forthcoming analysis, every measure of ideology was derived from Common Space DW-NOMINATE Scores (Carroll et al. 2015). Regarding the distribution of preferences, which is central to our theory and expectations, we follow Aldrich, Berger, and Rohde (2002) who derive a single factor score of conditional party government. We operationalize conditional party government in this manner for two reasons. First, it is endorsed by the theory’s chief proponents and, second, this measure has been used in numerous published studies (Finocchiaro and Rohde 2008; Lebo and O’Geen 2011; Miller and Overby 2014). For the variable Conditional Party Government, we calculate a one-dimensional indicator of conditional party government for both chambers using principal components analysis with a varimax rotation and then average the resulting scores into a single index (see Aldrich, Berger, and Rohde 2002). Figure 3 presents the scale scores. Analyses with individual chamber conditional party government measures produce similar results, as do tests with alternative measures (Battista 2009).

Figure 3.

Conditional party government by Congress.

We calculate the gridlock interval for each Congress, and we account for the position of the interval with respect to the interval in the previous period (Krehbiel 1998).¹¹ As the gridlock interval shifts from Congress to Congress, it exposes the enacting coalition’s laws to repeal. We construct this “exposed policy space” measure as follows. For the gridlock interval in the years before the establishment of cloture, we follow Heitshusen and Young (2006), who draw on Binder and Smith (1996), identifying the gridlock interval as the distance between the most liberal and conservative senators. In the years following the establishment of cloture, our calculation of gridlock interval is conventional: for a liberal president, the left bound is the veto pivot, and the right bound is the filibuster pivot (and vice versa for a conservative president). Using this measure, we then estimate the gridlock interval at the time of enactment, and calculate the Exposed Policy Space based on the shifts in gridlock intervals in subsequent Congresses.¹² Positive values indicate more of the gridlock interval left exposed by shifts, which should be associated with greater incidence of repeal.

We account for changes in Majority Seat Share, measured as the percentage of seats held by the majority at the subsequent time period subtracted from the percentage of seats held at enactment. We incorporate this measure to account for the varying effects of majority size: the Democrats in the seventy-third Congress saw the size of their coalition increase by 10 percent in the seventy-fourth Congress, reducing the likelihood of repeals in that time. By contrast, Democrats lost 20 percent of their seat share from the 103rd to 104th Congresses, with their laws facing a higher risk of repeal. Existing work has shown precisely this relationship (Berry, Burden, and Howell 2010; Ragusa 2010). However, Dahl and Tufte (1973) suggest that a larger party will likely be more heterogeneous. From an agenda setting perspective, this would weaken the majority’s ability to control the agenda and “protect” their laws from subsequent repeal efforts. These competing expectations offer another reason why preferences are so critical to understanding when and why repeals happen.

We also include indicators for whether there was a switch in the governing majority between the enacting Congress and each subsequent Congress. One Chamber Switch indicates a single chamber changed hands whereas the variable Two Chamber Switch indicates both chambers changed hands. We also account for changes the effect of parties with a dichotomous variable for the presence of Divided Government in the subsequent Congress. In addition, we include Binder’s (1999) measure of Bicameral Distance, where higher values indicate a larger gap between the House and Senate medians.

Finally, drawing on the importance of plenary time, we include a variable accounting for the majority party’s recent experience as minority party. As explained above, we believe parties behave differently based on their recent experiences in the minority. Parties who have only recently gained power after a long period away; those who were out of power for one session, but had been in power for much of the recent decade; and those who have not relinquished the majority for an entire decade will each have different approaches to this “agenda scarcity” and the delegation of power to party leaders (see also Barnes and O’Neill 2006; Fenno 1997). For this variable, which we label Historical Minority Status, we calculate a rolling measure of the number of chambers the minority party controlled over the previous decade. For example, in the 103rd (1993–1994) and 104th (1995–1996) Congresses, the Democratic and Republican majorities had similar levels of ideological homogeneity. Nonetheless, Democrats were the “permanent majority” in the 103rd Congress whereas Republicans in the 104th had seized control for the first time since the 1950s. So, the variable Historical Minority Status is coded “9” for Republicans in the 104th Congress (owing to Democrats near universal control of the House and Senate in the past ten years) whereas it is coded “2” for Democrats in the 103rd Congress.

Survival Analysis

Because the data set records the enactment and repeal of major laws, we estimated a series of Cox proportional hazards models where the response records whether law i had a major statute repealed in Congress t after passage.¹³ Specifically, we consider each of the eighty-nine instances of major statutory repeals, which were considered “at risk” from enactment until repeal. In total, we have 961 statute-time observations.¹⁴

As a consequence of the extended historical scope, we are unable to explore many contextual and bill-specific factors, though we do control for the specific policy domain. First, using the House and Senate committee of referral as a reference, we classify repeals as falling within six broad domains: Commerce, Taxes, Public Welfare, Foreign Relations and Internal Security, Natural Resources, and Judiciary.¹⁵ We expect tax laws to be the least durable and public welfare laws to be the most resistant to repeal. Indeed, tax laws may be the least durable, as much of the ideological conflict in the U.S. history concerns economic issues (McCarty, Poole, and Rosenthal 2006). Unlike laws in certain policy domains, taxes and economic policy are rarely considered “settled” or subject to a norm against “legal instability.” Alternatively, public welfare laws may be more resistant to repeal, as social programs are often subject to “retrenchment” (Pierson 1996). That is, expanding social programs allows politicians to claim credit—providing citizens with a material benefit—whereas contracting social programs entails considerable political costs and risks. Second, a handful of repeals are truly unique because the enacting law was passed for war purposes. We expect these War Laws will be short in duration for two reasons: either they are expressly temporary, such as the “Emergency War Tax Act” signed by Woodrow Wilson in 1914, or they were adopted in an uncertain and variable political environment, such as the late 1930s’ “Neutrality Acts.”

Main Findings

Models 1 to 5 in Table 1 contain the main results. For ease of interpretation, each model reports regression coefficients (log-relative hazards) rather than hazard ratios. Furthermore, each continuous covariate was standardized to compare the magnitude of the various effects. For this reason, each continuous covariate can be interpreted as the effect of a standard deviation change in the variable. We also report the information criterion (Akaike information criterion [AIC] and Bayesian information criterion [BIC]), where lower values indicate a better fitting model.

Table 1.

Determinants of Major Repeal (1877–2012).

	Model 1	Model 2	Model 3	Model 4	Model 5
Exposed Space	0.401**	0.371*	0.411**	0.392*	0.364*
Exposed Space	(0.15)	(0.16)	(0.16)	(0.16)	(0.16)
Divided Government	−0.257	−0.237	−0.272	−0.294	−0.260
Divided Government	(0.26)	(0.26)	(0.26)	(0.25)	(0.25)
Bicameral Distance	−0.272*	−0.316*	−0.283*	−0.277*	−0.315*
Bicameral Distance	(0.12)	(0.13)	(0.12)	(0.12)	(0.13)
Change in Seat Share	−0.118	−0.107	−0.110	−0.122	−0.114
Change in Seat Share	(0.12)	(0.12)	(0.12)	(0.12)	(0.13)
Conditional Party Government (CPG)	0.293**	0.158	0.271**	0.174	0.109
Conditional Party Government (CPG)	(0.099)	(0.11)	(0.10)	(0.11)	(0.11)
Historical Minority Status	0.207*	0.198^†	0.209*	0.175	0.177
Historical Minority Status	(0.10)	(0.12)	(0.10)	(0.11)	(0.12)
One Chamber Switch	0.549^†	0.460	0.410	0.676^†	0.557^†
One Chamber Switch	(0.30)	(0.30)	(0.36)	(0.34)	(0.33)
Full Switch	0.232	0.113	0.245	0.156	0.0808
Full Switch	(0.28)	(0.29)	(0.28)	(0.29)	(0.29)
CPG × Historical Minority Status		0.300***			0.272**
CPG × Historical Minority Status		(0.076)			(0.083)
CPG × One Chamber Switch			0.204
CPG × One Chamber Switch			(0.24)
CPG × Full Switch				0.359*	0.197
CPG × Full Switch				(0.15)	(0.18)
Commerce	0.0897	0.0964	0.0877	0.0808	0.0856
Commerce	(0.29)	(0.30)	(0.30)	(0.29)	(0.30)
Judiciary	0.444	0.341	0.421	0.394	0.307
Judiciary	(0.38)	(0.40)	(0.37)	(0.41)	(0.43)
Natural Resources	1.048***	1.015**	1.020**	1.020***	0.998**
Natural Resources	(0.32)	(0.33)	(0.32)	(0.31)	(0.32)
Taxes	0.844**	0.849**	0.870**	0.829*	0.832*
Taxes	(0.33)	(0.33)	(0.33)	(0.32)	(0.33)
Public Welfare	0.320	0.246	0.353	0.267	0.216
Public Welfare	(0.33)	(0.34)	(0.33)	(0.34)	(0.34)
Foreign Relations and Internal Security	0.673	0.708^†	0.671	0.702^†	0.712^†
Foreign Relations and Internal Security	(0.43)	(0.41)	(0.43)	(0.41)	(0.40)
War Law	0.940***	0.937***	0.953***	0.954***	0.951***
War Law	(0.23)	(0.24)	(0.24)	(0.22)	(0.23)
Observations	872	872	872	872	872
AIC	631.1	624.1	632.6	629.6	625.1
BIC	702.6	700.4	708.9	706.0	706.2

Standard errors are given in parentheses. AIC = Akaike information criterion; BIC = Bayesian information criterion.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Looking at Model 1 in Table 1, the results are as hypothesized and consistent with existing work on the durability of legislation. First, and foremost, we find that the greater the exposed space (the size of the enacting coalition’s gridlock interval exposed by shifts in preferences), the greater the likelihood of repeal. According to Model 1, a standard deviation increase in the exposed space increases the probability of repeal at time t by 44 percent.¹⁶ Consistent with Berry, Burden, and Howell (2010) and Ragusa (2010), and following our expectations, the significance and magnitude of this effect indicates that shifting pivot points are a key aspect of when and why repeals occur. We also find that bicameral distance has a significant negative effect on the likelihood of repeal (Binder 1999; Maltzman and Shipan 2008; Ragusa 2010). According to Model 1, a standard deviation increase in the distance between the House and Senate medians decreases the probability of repeal at time t by 24 percent.¹⁷ Model 1 also shows support for the argument that the longer the majority was in the minority over the preceding decade, the more likely the majority is to expedite the passage of repealing legislation (Rohde, Stiglitz, and Weingast 2013). We explore the possible interactive relationship between this variable and conditional party government in a moment. Finally, Model 1 indicates that repeals do not occur uniformly across policy domains, but are more likely in three areas: tax policy, wartime legislation, and natural resource laws. Although not central to our examination, our expectations were confirmed for tax policy and wartime legislation but disconfirmed with respect to public welfare legislation.

Although the above results are expected, our main contribution is the effect of conditional party government. Indeed, the variable Conditional Party Government is statistically significant and positive in Model 1, indicating that the majority’s ability to repeal legislation varies systematically over time as a consequence of the majority’s ideological cohesion and distance from the minority party. According to Model 1, a standard deviation increase in conditional party government increases the probability of repeal at time t by 34 percent.¹⁸ Simply put, this supports our basic contention that when it comes to “undoing” legislation, variation in the majority’s positive agenda setting power is a key piece of the puzzle. While the coefficient on the exposed space variable is larger in magnitude than the coefficient conditional party government, as a statistical matter these two effects are of equivalent in size.¹⁹ Thus, both factors matter, and they matter quite a bit.²⁰ We now turn to an examination of conditional party government’s possible interactions.

Based on the theoretical discussion, we created three interactions with Conditional Party Government capturing whether the effect of conditional party government on repealing activity is itself “conditional” on the nature of the majority’s past legislative status. The variable CPG × One Chamber Switch and the variable CPG × Two Chamber Switch test whether policies enacted by the minority in prior Congresses face a greater risk of repeal compared with policies enacted by the current majority. And the variable CPG × Historical Minority Status tests whether the volume of unsatisfactory status quo policies enacted over the prior decade has an effect on the likelihood of repeal. As noted earlier, we expect that when the majority had been out of power for an extended period, the rank-and-file are likely to delegate greater power to party leaders to expedite the passage of repealing legislation.

Models 2, 3, and 4 in Table 1 report the results. In Models 2 and 4, we find that the interaction with Historical Minority Status and Full Switch (respectively) are both significant and correctly signed. However, Model 3 shows that Conditional Party Government’s effect on repeals is unaffected by a single chamber switch.²¹ Based on these results, we estimated a final aggregate model (Model 5) with the two interaction terms included together. Regarding the substantive nature of the significant interaction effects, Figures 4 and 5 plot the marginal effects, and 90 percent confidence intervals, of a standard deviation increase in conditional party government (on the y-axis as hazard ratios) over the range of the interacting variables (on the x-axis).

Figure 4.

Marginal effect of Conditional Party Government by majority’s historical status in minority.

Figure 5.

Marginal effect of Conditional Party Government by party switch (both chambers).

Figure 4 (plotting CPG × Historical Minority Status) shows that the longer a new majority was in the minority over the past decade, the marginal effect of conditional party government increases in magnitude. In other words, repeals are more likely during periods of high conditional party government and when the majority was out of power for an extended period. For example, Figure 4 indicates that if the majority had only controlled one chamber in the previous decade (such as when Republicans regained control of both chambers in 1995), a standard deviation increase in Conditional Party Government increases the probability of repeal at time t by a whopping 208 percent. This is, of course, an extreme occurrence. If we set Historical minority status at a standard deviation above its mean (just under six chambers controlled in the previous decade), Figure 4 reveals that a standard deviation increase in conditional party government increases the probability of repeal at time t by 96 percent. Conversely, we can see that the effect of conditional party government is insignificant when the minority controlled fewer than four chambers over the past decade. As a whole, these effects support recent theoretical work on the dynamic nature of the majority’s electoral success and agenda setting capacity by Rohde, Stiglitz, and Weingast (2013, pg 13). As these authors put it,

Members of the majority party . . . often face a set of adverse status quo policies when the opposing party has controlled the legislative apparatus for a long period. . . . Under these circumstances, majority party members are likely to delegate agenda powers to party leaders, with individual members trading off legislative efficiency for control over legislative content.

Figure 5 (plotting CPG × Full Switch) shows that conditional party government has a larger effect on the likelihood of repeal following a switch in control of both chambers from the enacting Congress to subsequent Congresses. As a substantive matter, this indicates that the minority party’s policies are more likely to be targeted for repeal when both chambers change hands. According to Figure 5, a standard deviation increase in conditional party government increases the probability of repeal by 62 percent following a switch in both chambers. In contrast, the body of statutes enacted by the majority in previous eras have a much lower probability of being repealed. Indeed, Figure 5 indicates that a standard deviation increase in conditional party government increases the probability of repeal by just 25 percent when there is not a switch in both chambers. Furthermore, while the effect of conditional party government is positive absent a switch in control of both chambers, Figure 5 shows that the effect is insignificant (p = .22).

Once we consider the full scope of the two interaction effects, the estimates in Figures 4 and 5 suggest that the partisan factors in Model 5 (conditional party government and its interactions) have more consequential effects on the likelihood of repeal than the two nonpartisan factors (the exposed space and bicameral distance). However, Figures 4 and 5 treat the two interactions independently. As a consequence, the estimated increases in the probability of repeal are conservative estimates. If we consider the joint effect of a full switch in the governing majority and set the minority’s control of both chambers at a standard deviation above its mean, a standard deviation increase in conditional party government is estimated to increase the probability of repeal at time t by 127 percent.

Alternative Specifications

We examined four additional specifications to assess the robustness of our results. Because of page limits, each alternative is available in the online appendix. First, we examined whether the individual covariates are “proportional” (Box-Steffensmeier, Reiter, and Zorn 2003). We did not find evidence of temporal heterogeneity in the hazards. Second, we tested whether repeals are simply a function of congressional productivity. By controlling for productivity, we ensure that the relationship between heterogeneity and repeals is not spuriously caused by unobserved factors that explain both. We found that productive Congresses do not pass a larger volume of repeals and that the effect of conditional party government is robust alongside such a control. In addition to confirming the main finding, this further validates our claim that repealing is a distinct legislative action compared with policy creation. And third, we restricted the cases in the models to “first dimension” repealing votes (which we determined using the cutting angles from the NOMINATE algorithm). We estimated a model with only votes that had an angle greater than the absolute value of 45°, finding that our main results are unaffected by the dimensionality of the repealing vote.

Our fourth, and most important, alternative adopts an entirely distinct modeling strategy. In particular, we tested for selection bias in the results. Indeed, one possibility is that bills targeted for repeal are distinct from bills that are not targeted for repeal. In addition, the lack of a “denominator” in our analysis could pose inferential problems (e.g., Binder 2003). On both points, we want to compare the repeals we observe with the universe of all major laws. We tested for selection bias by merging our repeals data with Clinton and Lapinski’s (2006) significant legislation data set. Following Heckman’s (1979) two-step routine, we first predicted the likelihood of a bill being repealed using a restricted model while adding Clinton and Lapinski’s measure of legislative significance as a covariate. We used the results from this model to calculate the inverse mills ratio and included it as a covariate in the unrestricted duration model. We find that the inverse mills ratio is not statistically significant,²² and we thus did not find any evidence of selection bias in modeling strategy.

Repealing the ACA

In this final section, we discuss the implications of our results by applying them to Republican efforts to repeal the ACA. We note that this is not an exhaustive analysis the ACA repeal effort, as it excludes important exogenous factors, though it does serve to illustrate the causal mechanisms of our results. We also note two key caveats. First, by “repealing the ACA” we mean “repealing a major feature of the law.” Recall that the unit of analysis is a repealed statute. Indeed, laws are rarely repealed in their entirety. Rather, major statutes of the law are targeted for repeal. Second, we want to acknowledge that it is very difficult to put a precise estimate on the probability of repeal. For example, we believe norms against legal instability help explain why repeals are relatively uncommon in the first place. Indeed, laws become “institutionalized,” thus limiting the likelihood of repeal even when the enacting coalition is weakened. As a whole, while we can talk about factors that make repeal more or less likely, we do not believe our model captures the entire complexity of repeals. In the end, we are less interested in the specific predictions and more interested in how the features of the 114th and 115th Congresses make repeal more or less likely.

According to Model 5, Republicans have a 66 percent probability of enacting a major repeal in the 114th Congress.²³ The high probability of repeal is largely a function of the Republican Party’s organizational capacity, which our results identify as central to predicting repeals. Nonetheless, the model shows that Republicans are constrained by two factors. First, the gridlock interval is a constraint on Republican efforts to repeal the ACA. In fact, our results show that the filibuster pivot is the single greatest impediment to an ACA repeal. While divided government certainly matters, our model indicates that the most likely scenario is that Republican repeal bills will stall in the Senate (and thus never reach Obama’s desk for his veto). Second, Republicans are constrained by their duration in power. With Republicans controlling roughly half of the previous Congresses since 2005, our model shows that they are less likely to enact repealing legislation. Because plenary time is scarce, requiring parties to prioritize some discretionary items over others (Adler and Wilkerson 2012; Rohde, Stiglitz, and Weingast 2013), we theorized that “newly empowered” parties taking control after a long stint out of power are more likely to centralize agenda control. Although this seems to contradict the more than fifty repeal votes in the 113th and 114th Congresses, we counter that these votes—almost always “full” repeals with no chance of being enacted—are a function of electoral goals rather than policy goals (Gipson 2015). Moreover, in the 114th House, with a Republican majority for the third consecutive two-year term, there is significant disagreement among Republicans on how (and whether) to pursue an ACA repeal (in contrast to the 112th Congress, the first after Republicans regained control of the House, where Republicans were united behind proposals to repeal the entire law). Indeed, Speaker Boehner drew the ire of conservatives when he indicated a full repeal was not a major agenda item in the 113th Congress, famously declaring “Obamacare” to be the “law of the land” following the 2012 election.²⁴ As a whole, our results suggest that while Republicans have the organizational capacity to enact a repeal, they face significant veto points—divided government and the filibuster pivot—and have weakened positive agenda control on account of their recent status as the governing majority.

We also use Model 5 to investigate the implications of the 2016 presidential election on an ACA repeal in the 115th Congress (2017–2018). In this context, the results identify three key factors: the duration of the Republican’s status in majority, the presence or absence of unified government, and the location of the filibuster pivot. If a Democrat wins the presidential election and Republicans retain control of Congress, the probability of an ACA repeal drops from 66 to 54 percent. Although the institutional variables remain roughly the same (swapping one Democratic president for another), the large drop is due to the length of time Republicans have been in power. However, if a Republican presidential candidate wins the White House in 2016, unified government offsets the “penalty” for the Republican Party’s duration in the majority, and the probability of repealing a major element of the ACA remains at 66 percent.²⁵ In this case, the gridlock interval is still an impediment to repeal because a Democrat is likely to sit at the filibuster pivot. If a Republican wins the 2016 presidential election and Republicans in the Senate “go nuclear” thus shifting the filibuster pivot to Senate median, Model 5 predicts a repeal probability of 68 percent.

As a final matter, we conducted a purely speculative exercise to illustrate the importance of Republican ideological cohesion for the above estimates. In particular, we estimated the predicted probability of a Republican repeal in 114th Congress varying the conditional party government indicators to match three historical periods: the 104th Congress (1995–1996, the first after the Republican Revolution), the 97th Congress (1981–1982, Reagan’s first two-year term), and the 83rd Congress (1953–1954, the last Republican unified government before the Republican Revolution). We wanted to know: given variation in Republicans’ ideological cohesion in each of these periods, how does the probability of repeal change? Recall that in the 114th Congress, the probability of one major repeal is 66 percent. With conditional party government indicators from 104th Congress (high, but not historically high, cohesion), the probability of repeal drops to 46 percent. With conditional party government indicators from the 97th Congress (low-to-moderate cohesion), the probably of repeal is estimated to be 22 percent. And with conditional party government indicators from the 83rd Congress (historically low cohesion), the probability of repeal is just 17 percent. As a whole, the dramatic decline in the probability of repeal across these three conditions underscores just how critically important positive agenda control and variation in the majority’s ideological cohesion are for understanding when and why repeals happen.

We again emphasize that this application is motivated to demonstrate the causal mechanisms in the model, rather than to simulate the propensity for repeal while accounting for all relevant factors. A complete model would require accounting for presidential attention, public sentiment, Supreme Court cases, elite salience, and many other factors (see Adler and Wilkerson 2012; Maltzman and Shipan 2008; Ragusa 2010, among others). In other words, the precise predictions are much less important than the relative effects of the factors in our model. Nonetheless, this application illustrates the importance of accounting for the majority party’s recent experience in power, its ideological cohesion, and the influence of unified government in explaining which Congressional factors predict statutory repeals.

Conclusion

In this paper, we build on prior research by explicitly modeling repeal as a function of long-term changes in the preference arrangements within Congress. Whereas prior researchers have examined reversals in the latter half of the twentieth century, the 135-year time span of our data coincides with a number of important changes that are unexamined. In particular, this work speaks to the value of incorporating the ideological composition of the parties into explanations of when Congress will “undo” legislation. In sum, we find that the presence of strong, ideologically homogeneous parties is a critical factor in explaining how long legislation survives after enactment. Moreover, we find that the effect of conditional party government is greatest when the majority has been out of power for an extended period.

We believe these results show that repeals—some of the most dramatic efforts to shape national policy—are best characterized as long-term partisan contests over the location of the status quo. In this respect, this demonstrates that the process of “undoing” the law is distinct from the “regular” process of law creation. Indeed, while shifts in the gridlock interval play a key role in whether a party can repeal legislation, the heterogeneous nature of party power is of paramount importance when it comes to repealing the work of previous generations.

Substantively, whereas many studies examining policy reversal simply incorporate measures of party power, we focus on more nuanced approaches to studying repeals. The payoff for doing so offers an account for when party matters. Although partisan factors have large effects on the likelihood of repeal, we show that it is not sufficient to consider whether the enacting party is still in power or if there is divided government. Rather, the key underlying attributes are the ideological characteristics of the parties and the majority party’s recent history of control. In keeping with the intuition of conditional party government, party matters inasmuch as the members of the majority will allow it to, and the evidence is highly suggestive that positive agenda control is the mechanism for facilitating party influence in repealing legislation.

Footnotes

Acknowledgements

We are grateful for the comments and advice of David Rohde, Scott Adler, Larry Dodd, Michael Lynch, Jon Rogowski, Michael Crespin, Greg Koger, Jeff Harden, Charles Finocchiaro, David Bridge, and participants at the University of Colorado American Politics Workshop. We would like to thank the editors of Political Research Quarterly and three anonymous referees for their helpful feedback as well. All errors are our own.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this project came from the Initiative for Public Choice and Market Process, School of Business, College of Charleston.

Supplemental Material

Please contact the lead author for replication data availability questions.

Notes

References

Adler

E. Scott

Wilkerson

John D.

2012. Congress and the Politics of Problem Solving. Cambridge: Cambridge University Press.

Aldrich

John H.

1995. Why Parties? The Origin and Transformation of Political Parties in America. Chicago: University of Chicago Press.

Aldrich

John H.

Berger

Mark

Rohde

David

. 2002. “The Historical Variability in Conditional Party Government, 1877-1986.” In Party, Process, and Political Change in Congress, edited by Brady

David W.

McCubbins

Mathew D.

, 17–35. Stanford: Stanford University Press.

Aldrich

John H.

Rohde

David W.

1997. “The Transition to Republican Rule in the House: Implications for Theories of Congressional Politics.” Political Science Quarterly 112 (4): 541–67.

Aldrich

John H.

Rohde

David W.

2001. “The Logic of Conditional Party Government: Revisiting the Electoral Connection.” In Congress Reconsidered, edited by Dodd

Lawrence

Oppenheimer

Bruce

, 7th ed., 269–92. Washington, DC: CQ Press.

Barnes

Tiffany

O’Neill

Timothy

. 2006. “Learning to Govern: The Texas Experience.” The Journal of Political Science 34 (1): 1–36.

Battista

James Coleman

. 2009. “Why Information? Choosing Committee Informativeness in US State Legislatures.” Legislative Studies Quarterly 34 (3): 375–97.

Baumgartner

Frank R.

Jones

Bryan D.

1993. Agendas and Instability in American Politics. Chicago: University of Chicago Press.

Berry

Christopher R.

Burden

Barry C.

Howell

William G.

2010. “After Enactment: The Lives and Deaths of Federal Programs.” American Journal of Political Science 54 (1): 1–17.

10.

Binder

Sarah A.

1999. “The Dynamics of Legislative Gridlock, 1947-96.” American Political Science Review 93 (3): 519–34.

11.

Binder

Sarah A.

2003. Stalemate: Causes and Consequences of Legislative Gridlock. Washington, DC: Brookings Institution Press.

12.

Binder

Sarah A.

Smith

Steven S.

1996. Politics or Principle: Filibustering in the United States Senate. Washington, DC: Brookings Institution Press.

13.

Box-Steffensmeier

Janet M.

Reiter

Dan

Zorn

Christopher

. 2003. “Non-proportional Hazards and Event History Analysis in International Relations.” Journal of Conflict Resolution 47 (1): 33–53.

14.

Carpenter

Daniel P.

Lewis

David E.

2004. “Political Learning from Rare Events: Poisson Inference, Fiscal Constraints, and the Lifetime of Bureaus.” Political Analysis 12 (3): 201–32.

15.

Carroll

Royce

Lewis

Jeff

James

McCarty

Nolan

Poole

Keith

Rosenthal

Howard

. 2015. “Common Space DW-NOMINATE Scores with Bootstrapped Standard Errors (Joint House and Senate Scaling).” http://voteview.com/dwnomin_joint_house_and_senate.htm.

16.

Clinton

Joshua D.

Lapinski

John S.

2006. “Measuring Legislative Accomplishment, 1877-1994.” American Journal of Political Science 50 (1): 232–49.

17.

Corder

J. Kevin

. 2004. “Are Federal Programs Immortal? Estimating the Hazard of Program Termination.” American Politics Research 32 (1): 3–25.

18.

Cox

Gary W.

McCubbins

Mathew D.

1993. Legislative Leviathan: Party Government in the House. Berkeley: University of California Press.

19.

Cox

Gary W.

McCubbins

Mathew D.

2005. Setting the Agenda: Responsible Party Government in the US House of Representatives. Cambridge, MA: Cambridge University Press.

20.

Cox

Gary W.

McCubbins

Mathew D.

2007. Legislative Leviathan: Party Government in the House, 2nd ed. New York, NY: Cambridge University Press

21.

Critchlow

Donald T.

Vandermeer

Philip R.

2012. The Oxford Encyclopedia of American Political and Legal History. New York, NY: Oxford University Press.

22.

Dahl

Robert A.

Tufte

Edward

. 1973. Size and Democracy. Stanford: Stanford University Press.

23.

Fenno

Richard

1997. Learning to Govern. Washington, DC: Brookings Institution Press.

24.

Finocchiaro

Charles J.

Rohde

David W.

2008. “War for the Floor: Partisan Theory and Agenda Control in the US House of Representatives.” Legislative Studies Quarterly 33 (1): 35–61.

25.

Heckman

James J.

1979. “Sample Selection Bias as a Specification Error.” Econometrica 47 (1): 153–61.

26.

Heitshusen

Valerie

Young

Garry

. 2006. “Macropolitics and Changes in the U.S. Code.” In The Macropolitics of Congress, edited by Adler

E. Scott

Lapinski

John S.

, 129–50. Princeton: Princeton University Press.

27.

Howell

William

Adler

Scott

Cameron

Charles

Riemann

Charles

. 2000. “Divided Government and the Legislative Productivity of Congress, 1945-94.” Legislative Studies Quarterly 25 (2): 285–312.

28.

Kaufman

Herbert

. 1976. Are Government Programs Immortal? Washington, DC: Brookings Institution Press.

29.

Krehbiel

Keith

. 1992. Information and Legislative Organization. Ann Arbor: Michigan University Press.

30.

Krehbiel

Keith

. 1998. Pivotal Politics: A Theory of U.S. Lawmaking. Chicago: University of Chicago Press.

31.

Landsberg

Brian K.

2004. Major Acts of Congress. New York, NY: Macmillan Reference USA.

32.

Lebo

Matthew J.

O’Geen

Andrew

. 2011. “The President’s Role in the Partisan Congressional Arena.” Journal of Politics 73 (3): 718–34.

33.

Lewis

David E.

2002. “The Politics of Agency Termination: Confronting the Myth of Agency Immortality.” Journal of Politics 64 (1): 89–107.

34.

Maltzman

Forrest

Shipan

Charles R.

2008. “Change, Continuity, and the Evolution of the Law.” American Journal of Political Science 52 (2): 252–67.

35.

Mayhew

David R.

1991. Divided We Govern: Party Control, Lawmaking, and Investigations, 1946-1990. New Haven: Yale University Press.

36.

Mayhew

David R.

2005. Divided We Govern: Party control, Lawmaking, and Investigations, 1946-2002. New Haven: Yale University Press.

37.

McCarty

Nolan

Poole

Keith T.

Rosenthal

Howard

. 2006. Polarized America: The Dance of Ideology and Unequal Riches. Vol. 5. Cambridge: MIT Press.

38.

Miller

Susan M.

Overby

L. Marvin

. 2014. “Discharge Petitions and the Conditional Nature of Agenda Control in the U.S. House of Representatives.” Party Politics 20 (3): 444–55.

39.

Patashnik

Eric M.

2008. Reforms at Risk: What Happens after Major Policy Changes Are Enacted. Princeton: Princeton University Press.

40.

Pierson

Paul

. 1996. “The New Politics of the Welfare State.” World Politics 48 (2): 143–79.

41.

Ragusa

Jordan Michael

. 2010. “The Lifecycle of Public Policy: An Event History Analysis of Repeals to Landmark Legislative Enactments, 1951-2006.” American Politics Research 38 (6): 1015–51.

42.

Rohde

David W.

1991. Parties and Leaders in the Post-reform House. Chicago: University of Chicago Press.

43.

Rohde

David W.

Stiglitz

Edward H.

Weingast

Barry R.

2013. “Dynamic Theory of Congressional Organization.” Unpublished manuscript.

44.

Schickler

Eric

Pearson

Kathryn

. 2009. “Agenda Control, Majority Party Power, and the House Committee on Rules, 1937–52.” Legislative Studies Quarterly 34 (4): 455–91.

45.

Stathis

Stephen W.

2014. Landmark Legislation, 1774-2012: Major Acts and Treaties. Washington, DC: CQ Press.

46.

Gibson

Cary

. (2015, February 10) “Forget About Repealing Obamacare” U.S. News & World Report. Retrieved from http://www.usnews.com/opinion/blogs/opinion-blog/2015/02/10/gop-must-move-on-from-obamacare-repeal

47.

Walker

Jack

. 1977. “Setting the Agenda in the US Senate.” British Journal of Political Science 7 (4): 423–45.

48.

Woon

Jonathan

. 2009. “Change We Can Believe In? Using Political Science to Predict Policy Change in the Obama Presidency.” PS: Political Science & Politics 42 (2): 329–33.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.23 MB