Vote Choice and Vote Change in the Early U.S. House:
Evidence from the Gag Rule, 1836-1845*
Scott R. Meinke
Assistant Professor
Department of Political Science
Bucknell University
Lewisburg, PA 17837
[email protected]
http://www.facstaff.bucknell.edu/smeinke/
Abstract:
From the 24th through the 28th Congresses, the House operated under a self-imposed “gag
rule” that blocked receipt of petitions dealing with abolition and related matters. The decade-long
series of recurring votes on the gag rule represent not only a significant part of the congressional
struggle over slavery but also an excellent antebellum context in which to examine members’ voting
behavior and voting change over time. I show, first, that partisanship as well as electoral
considerations and constituency interests are related to gag rule voting. I then turn to members’
long-term voting histories, demonstrating that many members reversed their positions on this
controversial vote over time and that these reversals are explained by constituency-related and
partisan factors.
_________________
*Earlier versions of this research were presented at the 2002 Midwest Political Science Association and 2003 American
Political Science Association meetings. For helpful comments and conversations on this work, I thank Larry Baum, Jan
Box-Steffensmeier, Ed Hasecke, Kevin Scott, and Herb Weisberg. I thank Greg Yankee for his able assistance in data
collection.
On December 3, 1844, the House of Representatives voted to repeal a standing House rule
that had formally barred the chamber from even considering the abolition of slavery or the slave
trade.1 The vote reversed a policy that the House had adopted since 1836; it also signaled the
effective end of a decade-long battle that featured dozens of passionate floor speeches and many
sharply divided roll-call votes. In the course of this battle over the “gag rule,” the controversy
evolved from a chiefly partisan dispute with sectional undertones into one that displayed the
polarizing sectional forces that would drive the nation apart in the subsequent two decades.
Aside from being a fascinating episode of antebellum political history, the struggle over the
gag rule provides an illuminating case study of congressional decision making. First, since the
controversy produced a long series of similar recorded roll-call votes, this issue provides an
opportunity to explore antebellum House members’ individual decision patterns over the long term.
And second, because historical accounts suggest that the gag rule was both a partisan controversy
and a salient issue for attentive constituencies, the relative influence of the electoral connection and
partisanship should be visible, empirically, if the electoral connection operated to some degree in
the early House.
In this paper, I use the gag rule case to case to explore these two aspects of congressional
voting—the role of electoral and constituency influence and the stability of member positions over
time. I first explain members’ vote choices on the gag rule, focusing on the role of constituency
factors and partisanship in shaping decisions. Second, I analyze the long-term voting histories that
members developed on this recurring question. Despite the conventional expectation that members
adhere rigidly to a fixed position on recurring questions, I find and explain significant instability in
individual gag rule voting records. As I describe, this instability among continuing members also
played a part in the gag’s resounding defeat in the 28th Congress. Before presenting my findings,
1
however, I review theoretical expectations about vote choice and vote change and apply them to the
antebellum context, and I also provide a synoptic account of the gag rule’s history.
Choice and Change in Antebellum Voting
We know from the complementary works of Kingdon (1977, 1989) and Arnold (1990) that
contemporary members’ voting decisions result from a complex process. This process has its
origins in a hierarchy of member goals in which reelection is the primary motivation, but goals of
policy and power/influence also play a role. Since reelection is the proximate goal for members
(Mayhew 1974), calculations about the electoral consequences of vote decisions occupy center
stage in the decision process (Arnold 1990, 60; Kingdon 1977, 575) and involve assessing the
preferences of actual and potential constituencies (Arnold 1990, ch. 4). Members must make these
estimations and arrive at decisions that best satisfy their full set of goals “while constrained by
limits on time and cognitive capacity to do so without extensive study of each issue” (Kingdon
1977, 569; also Arnold 1990, 84-85). Necessarily, then, members rely on decision shortcuts in their
pursuit of electoral and other goals.
While this electorally-centered picture of congressional decision making has become
axiomatic for the contemporary Congress, its application in other contexts remains a subject of
debate. Literature on the 19th century Congress offers some contradictory perspectives on the
strength of the electoral connection and constituency influence in member decision making. It is
well-established that career patterns during Congress’ first century differ substantially from those of
its second century (Kernell 1977; Polsby 1968; Young 1966). Turnover rates in the early Congress
do not appear to support contemporary assumptions about electoral motivations. At the same time,
recent research has uncovered some evidence of an electoral connection among nineteenth century
members. Bianco, Spence, and Wilkerson (1996) demonstrate both electorally-driven voting
behavior and electoral consequences in the case of the Compensation Act of 1816, and Goodman
2
(2002) connects aggregate roll-call voting behavior with constituency characteristics in the
antebellum House (see also Goodman and Nokken [2004] on the early 20th century House).
These findings suggest a type of electoral connection despite the rarity of professionalized,
career members in the nineteenth century context. And, the trajectory of nineteenth-century
political careers provides a logic to explain electorally-connected behavior in the low-autonomy
antebellum House. Though turnover was high, members who left the House were likely to pursue
other elected offices that demanded strong constituency support; “lateral movement” was common
even though long House careers were rare (Polsby 1968, 148). Thus, even short-tenured House
members may have been sensitive to their levels of electoral support.2 Though members of the
antebellum House may not follow all contemporary patterns of decision making, there is good
reason to expect to find that behavior is related to the electoral goal.
We also know from Aldrich and Rohde (Aldrich 1995; Aldrich and Rohde 2001; Rohde
1991), among others, that party influences member behavior as a result of collective electoral goals
and party incentives for achieving those goals. Importantly for the case at hand, partisan theories of
congressional behavior emphasize that party influence is variable; the degree to which individual
electoral considerations are filtered through partisanship is conditional on the strength of
congressional parties. In the 1830s and 1840s, party power was at a high point in both electoral and
congressional politics (Aldrich 1995; Holt 1978; Silbey 1967), so we should expect to find that
members in this period pursue the goal of building and maintaining the party label.
Moreover, if these two important influences on modern congressional behavior—
electoral/constituency factors and partisanship—are to be seen in the antebellum Congress, the gag
rule presents an unusually good case in which to observe and quantify their influence. On one hand,
the gag rule and related controversies brought together a high level of grassroots constituency
activism (Zaeske 2003, 73-104), activism that sometimes looked almost like late-20th-century
3
politics with its combination of electioneering and direct lobbying (Sewell 1976, 10-23). On the
other hand, it also represented a core issue used by elites for partisan mobilization during the height
of the second party system (Holt 1978, 28-31). As I discuss below, historians disagree about the
precise reasons for the parties’ positions on the gag rule, but all agree that the parties saw in the gag
rule battle crucial implications for the meaning of the party labels. There is good reason to expect
that members of Congress felt both partisan pressure and independent constituency pressure on this
issue and felt compelled to reconcile these influences as modern legislators would.
The Vote History and Instability
The Kingdon and Arnold models emphasize shortcuts and cue-taking as a necessary part of
contemporary House voting. When a member is dealing with recurring votes on a similar question,
the vote history is itself a valuable decision cue (Arnold 1990, 87; Asher and Weisberg 1978;
Kingdon 1989, 275). The vote history is, in turn, a source of stability in members’ positions over
time.3 But this stability is bounded by the limits of the vote history. Institutional context, electoral
factors, or alterations in the content of recurring legislation all could lead members to set aside the
vote history to reevaluate and, potentially, switch their votes. Empirical work on position change
supports this general picture of voting stability and change (Asher and Weisberg 1978; Brady and
Sinclair 1984; Meinke 2005; Stratmann 2000), though the aggregate-level research of Poole and
Rosenthal (1997) leaves less room for regular, systematic instability in members’ positions.
Little research has addressed voting stability and instability in the early Congress. Several
significant works have analyzed broad patterns of voting, emphasizing the shifts in the sectional,
ideological, and partisan dimensions of conflict through the antebellum era (Alexander 1967; Poole
and Rosenthal 1997; Silbey 1967). A few historians, such as Linden (1976) and Bell (1979) have
explored long-term voting in order to explain single issues. But in general, analyzing positions on
individual issues is more difficult in the nineteenth century context since the recurring votes
4
common in the twentieth century (Asher and Weisberg 1978) were rarer in the early Congress. The
gag rule case is a clear exception, with members recording a long series of votes—votes that were
highly visible, as noted above. If the influences on issue-position stability and change can be seen
in the antebellum House, the gag rule provides an ideal case in which to do so. Before examining
the roll-call record, I provide a brief summary of the gag rule’s complex history to provide context
for the analysis.
The Gag Rule: A Brief History
The gag rule had its beginnings far from Washington, D.C., with the upsurge of the
abolitionist social movement in the early 1830s.4 Though abolitionist sentiment was not a new
phenomenon, it emerged as an organized and more widespread movement by 1833, when the
American Anti-Slavery Society was organized (Ludlum 1941, 204). The incipient abolitionist
movement began to publish and disseminate its views widely, quickly inducing violent attempts—in
the North as well as the South—to suppress abolitionism. The abolitionists accompanied their
efforts at grassroots persuasion with direct appeals to the federal government for the abolition of
slavery (Ludlum 1941; Sewell 1976; Zaeske 2003).
Relying on the constitutional right to petition, abolitionists took up a campaign of submitting
petitions to Congress, seeking redress on the specific topics of “the abolition of slavery and the
slave trade in the District of Columbia, prohibition of slavery in the territories, and the abolition of
the slave trade between the states” (McPherson 1963, 177). During the 24th Congress, House
members began to receive these petitions in mass quantities. While Congress had remained largely
detached from the slavery question for the fifteen-year period following the Missouri Compromise,
the influx of petitions—coupled with the visible anti-slavery agitation around the country—put the
issue unavoidably back on its agenda (Ludlum 1941, 205-207; Miller 1995, 105-139). When
abolition petitions arrived in the past, the House had sent them to die a quiet death in committee
5
(Rable 1975, 69), but that procedure would not be followed amid the heightened tensions of the 24th
Congress.
The events immediately leading to the first gag began with a series of petitions presented on
the House floor by New England representatives in December 1835, requesting that slavery in
Washington, D.C. be abolished (Ludlum 1941, 205-206; Rable 1975, 70-71). Though these
petitions and others that followed in the next few months were quickly disposed of, southerners
were outraged that the abolition issue had surfaced in the House chamber. The southern press and
state legislatures expressed indignation throughout the early months of 1836 as the House debated
both the immediate issue of the abolition petitions and the larger question of slavery (Rable 1975,
71-72, 74-75).
Party politics largely defined the initial reaction within the House. Depending on the
interpretation, Democrats saw the surge of petitions either as a danger to the party’s broader agenda
and intersectional coalition—and therefore as a threat that needed to be eradicated (McFaul 1975,
31-33; Miller 1995, 141; Rable 1975, 83; Riker 1982, 222)—or as useful issue for fueling interparty
conflict, defining partisan positions, and gaining strategic advantage at a time when the Whigs and
Democrats were competing to outdo each other in proslavery rhetoric, particularly in the South
(Holt 1978, 28-29).5
The House response to the petitions culminated in votes on a series of resolutions Henry
Pinckney (D-SC) presented in May. Pinckney had secured the creation of a select committee (in
January) to consider the petition issue; the committee was specifically instructed to offer reasons
why Congress could not regulate slavery in the states and should not regulate it in D.C. (Ludlum
1941, 206; Rable 1975, 75). The resolutions that the committee recommended in May included
statements to this effect, as well as the first gag rule, which resolved “that all petitions, memorials,
resolutions, propositions, or papers, relating in any way, or to any extent whatever, to the subject of
6
slavery, or the abolition of slavery, shall, without being either printed or referred, be laid upon the
table, and that no further action whatever shall be had thereon” (Congressional Globe 24 Cong 1,
26 May 1836, 505). This Pinckney gag resolution passed the House, 117-68.6
Though the resolution solved the immediate dilemma of how to deal with the influx of
abolition petitions—and advance the interests of the Democratic coalition—it did not eliminate the
slavery debate in the House. It merely refocused that debate. Opposition to the first gag rule could
be found at both extremes of slavery viewpoints, and leaders at both ends of the spectrum saw the
gag as an opportunity for vociferous position taking. In the south, a small contingent of southern
congressmen contended that the gag rule was actually a concession to the abolitionists. The gag and
its accompanying resolutions, they argued, implied that Congress had the power to abolish slavery
in D.C. but chose not to (Miller 1995, 144-146), a position that the extreme southern contingent
argued was unconstitutional on its face (Rable 1975, 75-77).
In the north, Whig members could be found on both sides of the gag rule issue in the 1830s,
but opponents were successful in attracting allies among many northern Whig congressmen as well
as some northern citizens. John Quincy Adams (W-MA) led the charge, portraying the gag as an
unconstitutional limitation on the right to petition the government (Miller 1995, 351-352). By
fostering this connection between abolitionism and the First Amendment, antislavery forces gained
allies (Barnes 1964, 111; McPherson 1963, 178-180). This may have been a politically useful
position for northern Whig leaders seeking to weaken the Jacksonian majority: the civil liberties
implications of the gag played directly into the northern “slave power” argument about southern
domination (Fogel 1989, 340-341; Riker 1982, 221).7
Those who opposed the gag found many ways to keep the issue on the front burner.
Members continued to present cleverly-worded petitions, mostly as an effort to attract public
attention (Mark 1998, 2225). Adams persistently offered petitions against the rule itself, on related
7
issues (such as Texas annexation), and against slavery (in thinly disguised form), keeping passions
surrounding the issue heated and making the gag’s renewal a critical priority for many Democrats
and southerners in the late 1830s.8 The gag rule battle became a regular feature of the first several
months of most House sessions, with gag rule adoption usually following an extended period of
conflict over receiving abolition petitions.
After several years of this recurring conflict, the first regular session of the 26th Congress
(January 1840) began with a significant shift in the politics of the gag rule. The session started, as
the others had, with attempts to introduce abolition petitions along with days of debate about a gag
rule. But the outcome would be more extreme in this case. According to McPherson, the large
“buildup of petitions at this session, the deepening economic depression, and the advent of a
presidential election combined to create an atmosphere which favored the granting of extremist
demands” (1963, 180). Those demands led to a new version of the gag rule with two notable new
features. First, the gag was incorporated into the House’s standing rules (Rule 21, or the “Johnson
gag”), and second, the wording was revised to indicate that abolition petitions would not even be
received, in contrast to the previous policy of automatically tabling the petitions (Congressional
Globe, 28 Cong 2, 28 Jan 1840, 150; Ludlum 1941, 215; Miller 1995, 357-373). In the 26th and
subsequent Congresses, then, the Adams-led gag opponents faced the taller task of attempting to
rescind the standing rule (Ludlum 1941, 215-219). In the 27th Congress, the Whig party’s ascension
to control of the Congress and the White House (1840 elections) failed to transform the issue,
despite the partisan dimensions of the initial gag rule vote.9
The 1842 elections, which yielded extremely high turnover, reversed the Whig majority. At
the start of the 28th Congress, the House narrowly defeated Adams’ traditional opening volley—the
amendment to exclude the gag from the rules being adopted. But as the House continued to
struggle through December and January over adopting a set of rules, a select committee (created by
8
an Adams motion) reported a set of rules that excluded the gag, and weeks of debate followed
during late January and February of 1844 (Ludlum 1941, 220; Miller 1995, 473). In a crucial vote,
the House refused an amendment from George Dromgoole (D-VA) that would have restored the
gag to the Adams committee’s rules.
This vote, along with several others in the first session of the 28th Congress, reflected a shift
toward majority opposition to the gag rule. However, the gag rule ultimately was not removed in
this session—the House rejected the full set of rules (without the gag) at the end of its drawn-out
battle, renewing the House rules from the 27th Congress instead. But at the beginning of the lame
duck session of the 28th Congress, Adams’ efforts at repealing the gag rule were finally successful.
The House repealed the rule by a vote of 108-80 (Ludlum 1941, 221-222; Miller 1995, 473-477),
and the proponents of the rule never successfully restored it, despite an attempt at the beginning of
the next session (Journal of the House of Representatives, 29 Cong 1, 1 Dec 1845, 11-12).
Analysis
For the analysis, I have identified twelve crucial, similar gag rule votes (Appendix B
describes the selection and coding in more detail). This section outlines expectations about
members’ decision behavior on those votes. The analysis explores two different dependent
variables—members’ vote choices and the long-term stability of those choices—so I set out
separate hypotheses for vote choice and vote change. Table 1 provides a summary of the
expectations, some of which are specific to northern or southern members since I analyze the
dynamics of these two groups separately.
[Table 1 about here]
Assuming that the opinion of attentive and latent constituencies was important to antebellum
members—and recognizing that gag rule votes presented a clear position-taking opportunity—we
should expect constituency and electoral factors to influence both vote choice and long-term vote
9
stability. Though basic sectional divisions (north vs. south) encompass some of the variation in
constituency interest, constituency differences within the north and south can be expected to affect
voting. In the analysis, I measure intrasectional variation in constituency pressures in the south by
the percentage of slaves in the district population; in the north, constituency signals are measured
using a variable that indicates whether an antislavery party candidate received votes in the district in
the previous election.10 (Details on coding and sources for all variables appear in Appendix B.)
If members are responding to constituency opinion and assessing the potential for latent
interests to affect their goals, northern members who face antislavery constituency sentiment should
be less likely to vote for the gag rule, and southern members from districts with larger slave
populations should be more likely to support the gag rule. Similarly, these members should be less
likely to change positions since the heightened electoral importance of their position gives them an
incentive to adhere to the electoral goal on the gag rule rather than diverge from the vote history to
respond to other goals.
The electoral connection also may manifest itself through an electoral marginality effect
(Bianco, Spence, and Wilkerson 1996). If the electoral goal is salient, members in less vulnerable
electoral circumstances should feel greater freedom to respond to considerations other than
constituency influence. In the north, safer members should be more likely to vote for the gag, other
things being equal, while in the south, safer members should be more likely to vote against the gag
in response to other factors. The same logic of electoral marginality predicts that safer members
will exhibit greater voting instability than members who are more vulnerable.
An even more basic expectation about gag rule voting is that party is closely associated with
members’ choices. Partisanship was nearing the high-water mark of the second party system during
this time period (Silbey 1967) and historians claim partisan politics were central to the early gag
battles in particular, as noted above. The vote choice models, then, include a dummy variable for
10
Whig affiliation. Since the gag was at least initially a Democratic party innovation, the expectation
is that Whigs will be less likely than Democrats to vote for the gag when other factors are held
constant.
[Figure 1 about here]
The interplay of party politics and growing sectional constituency influences (see Figure 1
for a descriptive illustration) point toward an important potential source of instability in member
voting. For members whose sectional constituency and party pull them in opposite directions, the
vote history should become less useful over time, and position change should be more likely. The
separate regional models of vote change, then, contain variables for party with the expectation that
southern Whigs and northern Democrats will be more likely to switch positions than their noncrosspressured counterparts. An additional party-related effect, majority control of the House, is
also controlled for in the change analysis since the partisan nature of early gag votes should make
members more likely to reevaluate their vote history when party control of the House changes.11
Within the northern and the southern sections, regional differences can also be expected to
make a difference in gag rule vote choices and vote change. Fogel (1989) and Silbey (1967) each
stress that the old northwest had strong reasons to identify politically with the south and was slow to
align with the northeast and adopt sectional politics. Northwestern members can be expected to
provide more support for the gag, but also to exhibit greater instability as gag rule politics shift over
time. In a similar way in the south, members from border states should be relatively less likely to
support the gag then their deep-south counterparts, but more likely to change positions as regional
cleavages on the issue widen.
Vote Choice
Table 2 displays the results of binary time-series-cross-sectional (BTSCS) logit models with
vote choice as the dependent variable. Because variables are operationalized differently (and are
11
expected to have different effects) in the north and south, the two sections are grouped into separate
logit models. On both models, likelihood ratio tests showed significant duration dependence, so
each includes a variable for the log of the member’s duration in a particular vote state (yea or
nay).12
Looking at behavior within the north only, the roll-call record provides some evidence for
both electoral and partisan effects on gag rule decisions. Other things being equal, northern Whigs
were less likely to support the gag then their northern Democratic counterparts, following the basic
understanding that the gag was brought about in an attempt to advance Democratic party interests.
As expected, electorally safer members were more likely than vulnerable members to support the
gag. Though this effect falls just short of the .05 level of statistical significance, the substantive
impact of electoral margin is substantively meaningful—as the size of the margin ranges from low
to high values, the predicted probability of a pro-gag vote ranges from .51 to .76 (with other
variables set at their mean/median). Meanwhile, the influence of northern constituency interest, as
measured by the presence of an antislavery vote in the incumbent’s district, is related to vote choice:
northern congressmen who experienced a threat from an antislavery constituency responded with
less support for the gag (.41 probability of pro-gag vote compared to .59, when other variables are
held at their mean/median). Regional differences within the north, which also capture some
constituency differences, have an impact on vote choice as well, with northwestern congressmen
more likely to support the gag than northeastern members. The models also include a control for
the significant change in the gag rule’s content in 1840, in the form of a dummy variable that equals
“1” for votes on the Rule 21 gag (see discussion above). This variable, as expected, indicates a
lower likelihood of northern members supporting the more sweeping gag rule.
[Table 2 about here]
12
For southern members, the hypothesized effects receive less support. One clear relationship
is between partisanship and gag rule support; southern Whig members were significantly less likely
to support the gag than southern Democrats, again supporting the notion that the gag was a partisan
measure, particularly early in its history. But unlike in the north, electoral marginality, constituency
interests, and regional effects are not apparent in the voting patterns. In part, this may reflect the
relatively low variation to explain in southern voting—while some southerners did oppose the gag
rule, the vast majority were Whigs, and anti-gag votes made up only about 7% of the total southern
votes cast during the time period. (A rare-events logit model for the southern cases yields no
significant difference in the results reported in Table 2.)
Vote Change
Constituency, electoral, partisan, and regional effects all help to explain vote choices on the
gag rule. But how did these choices vary over time—when did members choose to defect from
their established positions? Table 3 displays the results of BTSCS analyses for northern and
southern members using position change as the dependent variable: the binary dependent variable
equals “1” wherever a member’s vote at time t is the opposite of the member’s vote at t-1 and
indicating “0” when votes were cast in the same direction at t and t-1. Since the uncorrected models
exhibited duration dependence, both models include a correction for duration dependence in the
form of logged duration since the last vote change.13
In the north, partisan and regional factors are the clearest predictors of vote switching, with
some evidence for constituency-related effects as well. Democrats in the north show a higher risk
of switching positions, as do members from the northwest; these findings comport with the basic
expectation that the cross-sectional alignment of the early gag supporters broke down over time. In
addition, the negatively-signed coefficient for the antislavery variable, which falls just short of the
.05 significance level (p=.051), supports the expectation that members with a clear antislavery
13
constituency cue maintained more consistent positions over time on the gag, when other factors are
taken into account. And, as expected, the change in the gag rule’s content (Rule 21) brought about
greater risk of position change.14 Finally, the variable for duration dependence yields an interesting
substantive interpretation in this model: the negatively-signed coefficient implies that members
were at lower risk of position change as their stable vote history became more established, though it
should be noted that the coefficient is significant only at the .10 level. At the same time, the
northern vote change model does not provide any support for the electoral marginality hypothesis or
the argument that changes in House majority control would lead to party-based position shifts.
Predicted probabilities from the vote change model help to tell the story of gag rule voting
dynamics in the north. Position change was not rampant on the gag issue, so the baseline
probability of reversal (with independent variables set to means/medians) was only .15. However,
for a member from the northwest (other variables held constant at means/medians), the probability
of change rose to .31. For a northwestern Democratic member from a safe (unchallenged) electoral
district with no antislavery vote in the previous election, the probability of change rises to .44.
Northwestern Democrats, who were particularly likely to support the gag early in its history, had a
greater tendency to change positions over time as the gag rule moved from a partisan to a more
sectional issue.
Position change in the south also shows systematic components, but the explanations for
southern vote shifts are somewhat different. Notably, electoral marginality does have the expected
effects in the south, with safer members showing a higher risk of position change. As in the north,
partisanship also has a strong relationship with voting stability: here, Whig members were more
likely to change positions than Democrats, other things being equal. The correction for duration
dependence shows the stabilizing effects of a longer vote history, as in the north, but other factors,
14
including the border-state and constituency-interest effects, do not show significant effects in the
southern vote change model.
In short, we see some significant dynamics in gag rule voting over time, with electoral
status, party/section crosspressuring, legislative content, and the vote history itself each showing
some relationship with the risk of change in one or both of the regional models. Obviously, the
statistical evidence for partisan cross-pressure effects is strongest across the two change models, but
each of the models provides strong statistically significant evidence for constituency-related effects
(region in the north and electoral marginality in the south) as well. Though some of the literature on
aggregate voting behavior would lead to an expectation of no systematic position reversals, the
decade-long series of gag rule votes points toward a different conclusion. Members can be diverted
from their vote history, and factors related to their decision context help to determine whether the
vote history will be reevaluated and discarded. Though there is growing evidence for this kind of
individual-level change in the contemporary Congress, the gag rule provides a unique
demonstration of this instability in a different era.
The Gag Rule’s Demise: Replacement and Conversion
These patterns of voting instability raise a separate but equally interesting question: how did
member replacement and member conversion contribute to the House’s decision to end the gag rule
in 1844? It is possible to provide at least a descriptive picture of replacement and conversion in the
gag rule’s demise. Note first, though, that the story one tells about policy change depends on how
one defines that change. If we look at the full decade-long history of the gag rule, taking place as it
did during an era of extremely high turnover, then the shift from adopting the gag to overtly
rejecting it is by definition driven by replacement: only a handful of members who originally voted
on the gag in 1836 remained in the House in 1844. By contrast, if we focus on the makeup of the
decisive and permanent vote to kill the gag (3 December 1844, Vote 11 in Appendix A), we can see
15
the more complex dynamics of the final policy change. Here, I will look briefly at the members
who participated in this vote at the end of the 28th Congress to describe how they reflected either
replacement or conversion effects.
Jenkins and Stewart (2003), in the only other recent empirical exploration of the gag rule
case, argue that the shift can be explained largely by the replacement of northern Democrats in the
high-turnover 28th Congress, pointing out that “rookie” northern Democrats showed strong anti-gag
leanings and that they both outnumbered the returning northern Democrats and outdid the
“veterans” in gag rule opposition. As Figure 1 also illustrates, the sharp reversal in northern
Democratic gag rule support was central to the long-term policy shift, but the replacement side of
the story still needs a bit more explanation since we need to know whom the northern Democrat
“rookies” replaced (opponents or supporters) in order to specify the net change. Figure 2
categorizes the district-level replacements for the 27th and 28th Congresses that led to permanent
district position changes—that is, the new member’s voting position was the reverse of his
predecessor’s and the new position was not reversed by conversion or subsequent replacement prior
to the gag’s demise in December 1844.15 This simple descriptive analysis reveals two important
trends: district-level replacement led to a net gain of 14 anti-gag votes across these two elections,
and the most of the “anti” replacements were the consequence of northern Whigs replacing northern
Democrats or same-party replacement among northern Democrats. In addition, the south
experienced a five-seat net loss in the 1842 reapportionment (southern-to-northern replacement), a
change that essentially represented a five-vote loss for pro-gag-rule forces. Taken together, these
two sources of replacement—the 1842 reapportionment and the elections to the 27th and 28th
Congresses—appear to be sufficient to explain the anti-gag margin of victory at the end of the 28th
Congress.
[Figure 2 about here]
16
What, then, was the role of conversion? Knowing from the voting analysis that systematic
conversion did take place on the gag rule, we can examine the vote history of a key group of vote
switchers: gag proponents who converted to opposition at some point, sustained that new position,
and cast an anti-gag final vote. Sixteen members converted permanently from a “pro” to an “anti”
position on the gag rule prior to the last vote in the 28th Congress (Vote 11), in which the gag rule
was permanently removed. Among these members, fourteen were northern Democrats and the
remaining two were border-state Whigs. Meanwhile, only four members who converted in the
opposite direction contributed a pro-gag vote to the effective vote to remove the gag. This net
contribution of twelve anti-gag votes to the final outcome is not sufficient to explain the House’s
reversal—counterfactually, the gag would have been defeated very narrowly in December 1844
even if these members had retained their original position. But it is noteworthy in an era of
extraordinarily high turnover that the voting instability of continuing members changed a very
narrow outcome into a decisive victory for the anti-gag forces.
Of course, the broader story of the gag rule’s demise is as an evolutionary one. What was
originally a question of partisan power in the 1830s, with partisan vote divisions that crossed
sectional lines, became a sectional issue in the 1840s. The story of this evolution incorporated both
the long-term conversion of a small number of northern Democrats and the departure of pro-gag
northern Democrats and their replacement by anti-gag northern Whigs and northern Democrats in
the 27th and 28th Congresses. This evolutionary pattern provides additional illustration of the
relative role of constituency and party in gag rule politics. Northern Democrats became willing to
reverse their positions as the gag issue came to be seen as a question of northern interests instead of
Democratic agenda control. Moreover, significant anti-gag shifts resulted from the Democrat-toWhig replacements in the north at the start of the Whig-majority 27th House, and these shifts were
17
often then sustained in districts where northern Democrats defeated Whigs amidst the record-high
turnover and significant Whig seat losses in the 28th Congress.
Conclusion: Some Lessons from the Gag Rule
The gag rule’s obscurity to political scientists belies its relevance as an antebellum case
study in member decision making. Voting on the gag rule was driven by the powerful, shifting
forces of partisanship, section, and constituency. In members’ voting choices and in vote change
patterns, the case shows the interplay of strong party forces and powerful constituency interests.
The parties had clear interests in maintaining united positions on the gag rule, and the empirical
evidence shows that party voting was indeed strong. Yet despite the power of party in this era, and
despite record-high turnover rates, some of the evidence suggests a distinct electoral connection on
the gag rule, with voting connecting with the antislavery movement (in the north) and with electoral
circumstances. The increasing power of section and region, too, illustrates the importance of
constituency forces. The gag rule case, then, lends new weight to recent claims that an electoral
connection can be detected amidst the partisanship of the antebellum Congress; by providing a
long-term rather than cross-sectional issue for analysis, it also allows a new way to glimpse this
effect.
The gag rule also offers an additional lesson on antebellum congressional politics. On a
recurring vote choice like the gag rule, congressional scholarship typically would predict that vote
choices would be overwhelmingly stable. Among the consistent gag opponents and supporters,
though, there were many members who changed positions over time. These position reversals can
be explained by electoral circumstances, by partisan/sectional crosspressuring, by constituency
interests, and by the vote history itself. These findings provide additional illustration of the
importance of constituency and party, and they reveal the importance of understanding voting
positions over time.
18
In an even broader sense, congressional behavior during the gag rule episode can further
inform our understanding of politics during the second party system. The gag began as a tool of
party power, a majoritarian mechanism that the Democratic party used to preserve its grip on an
intersectional coalition, maintain the focus on issues of broad partisan agreement, and restrict
minority interference with its agenda. But northerners—both Democrats and the few Whigs who
helped sustain the gag rule during the Whig-controlled 27th Congress—gradually slipped away from
the pro-gag coalition through conversion and replacement. The gag became an issue of blurred
partisan lines and sharp sectional divisions by its demise in 1844. As Miller has summarized it, “we
can say as a kind of a shorthand that the gag ended when sectional identity sliced through party
identity” (1995, 484). While partisan divisions (and the second party system) would remain
generally strong for several more years, (Fogel 1989; Holt 1978, ch. 3; Silbey 1968; Silbey 1985,
ch. 5), voting dynamics on the gag rule foreshadowed in the early 1840s the divisions that would
ultimately fracture the second party system in the 1850s. On the surface, the slavery issue was
suppressed by the gag rule from 1836 to 1844, but on a deeper level, the crosscutting divisions
under the surface of the second party system played out in the controversy over the gag rule itself.
19
Appendix A. Selected Gag Rule Roll Calls, 1836-1845
a
Vote
Congress/
Session
Date of Vote
ICPSR
9822
Vote #
“Yea”
position
Rollcall
1
24/1
26 May 1836
207
Gag
117-68
2
24/2
18 Jan 1837
365
Gag
129-69
3
25/2
21 Dec 1837
72
Gag
122-74
4
25/3
12 Dec 1838
363
Gag
128-78
5
26/1
28 Jan 1840
106
Gag a
115-105
6
26/2
9 Dec 1840
639
Gag
82-58
7
27/2
6 Dec 1841
277
Anti-gag
84-87
8
27/3
12 Dec 1842
818
Gag
106-102
9
28/1
4 Dec 1843
13
Anti-gag
91-95
10
28/1
27 Feb 1844
141
Gag
85-107
11
28/2
3 Dec 1844
433
Anti-gag
108-80
12
29/1
1 Dec 1845
13
Gag
85-121
Vote to make gag rule a standing House rule (Johnson gag).
20
Position
Changes
(% of
recurring
votes)
-16
(10.3%)
17
(20.5%)
8
(5.0%)
15
(14.6%)
17
(13.6%)
5
(6.4%)
12
(7.8%)
4
(7.7%)
11
(6.7%)
5
(3.1%)
6
(5.9%)
Appendix B. Variable Operationalization and Data Sources
Dependent Variables (Roll Calls)
Roll call data for the quantitative analysis is drawn from ICPSR study number 9822. The
votes included in the analysis (see Appendix A) represent a subset of the gag-related votes cast
between 1836 and 1845. For the analyses, I have identified the votes that provide the “cleanest”
measure of member positions on the gag from each session in which the issue was debated. I have
relied on historical accounts (especially Ludlum 1941 and Miller 1995) as well as roll call
descriptions from Alexander (1967) and the ICPSR #9822 codebooks to separate crucial votes in
the gag rule battle from those that are intertwined with other matters (usually, with controversies
over the adoption of the full set of rules). I have also excluded votes taken in a spring 1841 special
session: a gag rule battle ensued, as usual, at the beginning of this session, but President Tyler had
called the session for narrow purposes, and the House eventually resolved to temporarily gag all
subjects unrelated to the economic reasons for the special session (Ludlum 1941, 216). The raw roll
call data is recoded such that “1” always represents a pro-gag rule position, though the original yea
position on some votes was in favor of rescinding the gag (see Appendix A).
Independent Variables
Party. Party coding for Whigs and Democrats is based on the corrected party variable coding in
Poole and Rosenthal NOMINATE data. For the 24th Congress, in which Poole and Rosenthal retain
Jacksonian and anti-Jacksonian party labels, I have relied on party coding in the ICPSR #9822 data
set. Conflicts between the ICPSR coding for the 24th and Poole and Rosenthal coding for later
Congresses were cross-checked and corrected with the current Biographical Directory of the United
States Congress (http://bioguide.congress.gov).
Slave Percentage. For southern districts, I have coded census-derived slave percentage data from
Parsons, Beach, and Hermann (1978) and Parsons, Beach, and Dubin (1986).
21
Antislavery Vote. For each northern House district, a dummy variable reflects support in the
incumbent’s district for a candidate from a party (or parties) associated with antislavery positions in
the preceding House election, based on election results I obtained from Dubin (1998). Most of the
districts in the data set with antislavery candidates featured Liberty party candidates. A handful of
other antislavery-coded districts included Anti-Mason candidates or candidates labeled as “AntiSlavery” or “Abolitionist.” (More prominent antislavery parties such as the Free Soil party do not
emerge until after the period of this study.)
Electoral margin. The electoral margin variable indicates the percentage of the district vote won by
the incumbent in the preceding election. The electoral data was obtained Dubin’s authoritative
compilation of congressional election results (Dubin 1998).
Measuring electoral marginality in the antebellum House poses a challenge since
multimember districts (MMD) and multimember at-large states were relatively common, especially
prior to the 1842 reapportionment. I address this problem by creating a pseudo-single-member
district for each winning candidate from a MMD. This is the method described by Niemi, Jackman,
and Winsky (1991) in the context of state legislative MMDs and applied to the antebellum House
by Bianco, Spence, and Wilkerson (1996). Simply stated, the Niemi et al. approach creates a SMD
with simulated two-party competition by pairing the top vote-getting winning candidate with the
opposite-party losing candidate with the lowest vote support; the winning candidate with the
second-highest vote total is paired with the opposite-party losing candidate with the second-lowest
total, and so forth. See Niemi, Jackman, and Winsky (1991, 97-107) for more detail.
22
Endnotes
1
Rule 21 specified “that no petition, memorial, resolution, or other paper praying the abolition of slavery in the District
of Columbia, or any State or Territory, or the slave trade between the States or Territories of the United States in which
it now exists, shall be received by this House, or entertained in any way whatever” (Congressional Globe, 28 Cong 2,
28 Jan 1840, 150). As discussed below, the gag rule was imposed as a resolution rather than in a standing rule prior to
1840.
2
Swift (1987) elaborates this point and provides evidence of lateral movement in the pre-Jacksonian House.
3
Obviously, stability is also produced by other factors, not the least of which is the continuity of members’ personal
policy preferences. I emphasize the vote history here since it is, as Asher and Weisberg have demonstrated, only
intermittently a source of stability, unlike personal policy preferences, which remain a constant for most members.
4
This quick sketch of the gag rule’s rise and fall in the House focuses on the circumstances surrounding the House’s
key gag rule roll calls. It obviously does not do justice to the full history of the gag battle and the related party politics
and social movements of the 1830s and 1840s. Readers who are interested in a more thorough historical treatment of
this issue are directed to Miller 1995, which provides the richest account of the gag rule controversy and its larger
context.
5
The petitions’ calls for federal restrictions on slavery and the slave trade tapped into the Democrats’ concern with
limiting the power of the national government, as well, and provided further opportunity to define party differences.
See Silbey 1967; 1985.
6
Note that this first gag “rule” was not a standing House rule but rather a resolution that required renewal at each new
session.
7
The Whigs faced potential sectional divisions on the issue as well (McPherson 1963), and some intraparty variation in
positions on slavery seems to have been part of both parties’ survival strategies during the 1830s and 1840s (see Holt
1978).
8
Adams’ petitions included a request for an examination of “how to make effective the constitutional guarantee of a
republican form of government, when 13 slave States had governments ‘absolutely despotic, onerous, and oppressive’
to a great part of their populations,” a suggestion for making “an amicable division of the Union by a line running
between the free and the slave States,” and a proposal for “the removal of the seat of government farther North, where
the principles of the Declaration of Independence ‘are not treated as a mere rhetorical flourish’” (Ludlum 1941, 213,
23
217). See Miller 1995 on Adams’ strategy during the entire period and Zaeske 2003 on the petitioners’ ongoing
campaign under the gag.
9
Riker (1982, 222) points out that by the early 1840s some Whigs would see a need to suppress the slavery issue in
order to maintain intersectional unity on other matters, just as the Democrats had. Holt’s (1978) contrasting
interpretation of partisan politics surrounding slavery, which emphasizes each party’s use of the gag as part of their
differential exploitation of the slavery issue by region to maintain necessary partisan conflict, implies an alternative
reason for a Whig House to sustain the gag: allowing northern and southern members alike to take useful public
positions on the issue. See also Aldrich 1995, ch. 5.
10
This measure for constituency interest follows the operationalization that Goodman (2002) has used in his study of
NOMINATE scores in the antebellum House.
11
The control variable is a dummy variable equaling “1” for the first vote after a party control change, “0” otherwise. A
t-test shows that members voted for the gag with significantly greater frequency when their own party controlled the
House (t=11.757, p<.001), though this is largely an indication of Democratic party effects since Democrats controlled
the House in all congresses in this time period except the 27th. Because of this low variation, a hypothesis about House
party control is not tested in the vote choice models.
12
This correction for duration dependence offers a better fit in both models than a linear variable for duration, and it
produces results nearly identical to models using a cubic spline while allowing more straightforward interpretation. See
Beck, Katz, and Tucker 1998 and Box-Steffensmeier and Jones 2004 (ch. 5).
13
As in the vote choice models, this correction offers a significant improvement over a linear duration form and, in this
case, slightly better improvement than a cubic spline. The substantive results for the vote change models are robust
when an alternative duration modeling approach is used, as well: Cox proportional hazards models (variance-corrected
repeated events) predicting duration to vote change produce very similar results in terms of direction and significance,
with the exception of the antislavery dummy in the north, which is not statistically significant in the event history
model.
14
Descriptive data strongly suggests this variable does reflect a reaction to the more extreme content: no northern gag
opponents switched to support on this vote, while 13 gag supporters switched to opposition (59% of those northerners
eligible for such a switch).
24
15
These replacements are predominantly within single member districts. However, where multimember districts
contained a departing and a freshman member with opposite positions in consecutive congresses, I have included them
in the count. Also, I include replacement-based reversals in the 27th in which the freshman was replaced for the 28th but
the new position was sustained through the end of the gag. Finally, the pairings in this count obviously exclude pairings
resulting from redistricting of a continuing member—i.e., incumbents who replaced other incumbents in the 28th
Congress by representing different districts are not included.
25
References
Aldrich, John H. 1995. Why Parties?: The Origin and Transformation of Political Parties in
America. Chicago: University of Chicago Press.
Aldrich, John H. and David W. Rohde. 2001. “The Logic of Conditional Party Government:
Revisiting the Electoral Connection.” in Lawrence C. Dodd and Bruce I. Oppenheimer, eds.,
Congress Reconsidered, 7th ed. Washington, DC: CQ Press.
Alexander, Thomas B. 1967. Sectional Stress and Party Strength: A Study of Roll-Call Voting
Patterns in the United States House of Representatives, 1836-1860. Nashville: Vanderbilt
University Press.
Arnold, R. Douglas. 1990. The Logic of Congressional Action. New Haven: Yale University Press.
Asher, Herbert and Herbert F. Weisberg. 1978. “Voting Change in Congress: Some Dynamic
Perspectives on an Evolutionary Process.” American Journal of Political Science 22:391425.
Barnes, Gilbert Hobbs. 1964 [1933]. The Antislavery Impulse 1830-1844. New York: Harcourt,
Brace & World.
Beck, Nathaniel, Jonathan Katz, and Richard Tucker. 1998. “Taking Time Seriously: TimeSeries—Cross-Section Analysis with a Binary Dependent Variable.” American Journal of
Political Science 42:1260-1288.
Bell, Rudolph M. 1979. “Mr. Madison’s War and Long-Term Congressional Voting Behavior.”
The William and Mary Quarterly, 3rd Series, 36:373-395.
Bianco, William T., David B. Spence, and John D. Wilkerson. 1996. “The Electoral Connection in
the Early Congress: The Case of the Compensation Act of 1816.” American Journal of
Political Science 40:145-171.
Brady, David and Barbara Sinclair. 1984. “Building Majorities for Policy Changes in the House of
Representatives.” Journal of Politics 46:1033-1060.
Box-Steffensmeier, Janet M. and Bradford S. Jones. 2004. Event History Modeling: A Guide for
Social Scientists. Cambridge: Cambridge University Press.
Dubin, Michael. 1998. United States Congressional Elections, 1788-1997: The Official Results of
the Elections of the 1st through 105th Congresses. Jefferson, NC: McFarland.
Fogel, Robert W. 1989. Without Consent or Contract: The Rise and Fall of American Slavery.
New York: W.W. Norton.
Goodman, Craig. 2002. “Where Do I Stand?: Examining the Relationship between Constituency
Preferences and Roll Call Behavior of Freshmen in the Antebellum House of
26
Representatives, 1836-1861.” Paper presented at the Annual Meetings of the American
Political Science Association.
Goodman, Craig and Timothy P. Nokken. 2004. “Lame-Duck Legislators and Consideration of the
Ship Subsidy Bill of 1922.” American Politics Research 32:465-489.
Holt, Michael F. 1978. The Political Crisis of the 1850s. New York: W.W. Norton.
Jenkins, Jeffrey A. and Charles Stewart III. 2003. “The Gag Rule, Congressional Politics, and the
Growth of Anti-Slavery Popular Politics.” Paper presented at the Annual Meetings of the
Midwest Political Science Association.
Kernell, Samuel. 1977. “Toward Understanding 19th Century Congressional Careers: Ambition,
Competition, and Rotation.” American Journal of Political Science 21: 669-693.
Kingdon, John W. 1977. “Models of Legislative Voting.” Journal of Politics 39: 563-595.
Kingdon, John W. 1989. Congressmen’s Voting Decisions, 3rd ed. Ann Arbor: University of
Michigan Press.
Linden, Glenn M. 1976. Politics or Principle: Congressional Voting on the Civil War
Amendments and Pro-Negro Measures, 1838-1869. Seattle: University of Washington
Press.
Ludlum, Robert P. 1941. “The Antislavery ‘Gag-Rule’: History and Argument.” Journal of Negro
History 26: 203-250.
Mayhew, David. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.
Mark, Gregory A. 1998. “The Vestigial Constitution: The History and Significance of the Right to
Petition.” Fordham Law Review 66: 2153-2231.
Meinke, Scott R. 2005. “Long-term Change and Stability in House Voting Decisions: The Case of
the Minimum Wage.” Legislative Studies Quarterly. Forthcoming.
McFaul, John M. 1975. “Expediency vs. Morality: Jacksonian Politics and Slavery.” The Journal
of American History 62: 24-39.
McPherson, James M. 1963. “The Fight Against the Gag Rule: Joshua Leavitt and Antislavery
Insurgency in the Whig Party, 1839-1842.” Journal of Negro History 48: 177-195.
Miller, William Lee. 1995. Arguing About Slavery: John Quincy Adams and the Great Battle in the
United States Congress. New York: Vintage.
Niemi, Richard G., Simon Jackman, and Laura R. Winsky. 1991. “Candidacies and
Competitiveness in Multimember Districts.” Legislative Studies Quarterly 16: 91-109.
27
Parsons, Stanley B., William W. Beach, and Michael J. Dubin. 1986. United States Congressional
Districts and Data, 1843-1883. New York: Greenwood Press.
Parsons, Stanley B., William W. Beach, and Dan Hermann. 1978. United States Congressional
Districts, 1788-1841. New York: Greenwood Press.
Polsby, Nelson W. 1968. “The Institutionalization of the US House of Representatives.” American
Political Science Review 62: 144-168.
Poole, Keith T. and Howard Rosenthal. 1997. Congress: A Political-Economic History of RollCall Voting. New York: Oxford University Press.
Rable, George C. 1975. “Slavery, Politics, and the South: The Gag Rule as a Case Study.” Capitol
Studies 3: 69-87.
Riker, William H. 1982. Liberalism Against Populism. Prospect Heights, IL: Waveland Press.
Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of
Chicago Press.
Sewell, Richard H. 1976. Ballots for Freedom: Antislavery Politics in the United States, 18371860. New York: Oxford University Press.
Silbey, Joel H. 1967. The Shrine of Party: Congressional Voting Behavior 1841-1852.
Pittsburgh: University of Pittsburgh Press.
Silbey, Joel H. 1985. The Partisan Imperative: The Dynamics of American Politics Before the
Civil War. New York: Oxford University Press.
Stratmann, Thomas. 2000. “Congressional Voting over Legislative Careers: Shifting Positions and
Changing Constraints.” American Political Science Review 94: 665-676.
Swift, Elaine K. 1987. “The Electoral Connection Meets the Past: Lessons from Congressional
History.” Political Science Quarterly 102:625-645.
Young, James Sterling. 1966. The Washington Community, 1800-1828. New York: Columbia
University Press.
Zaeske, Susan. 2003. Signatures of Citizenship: Petitioning, Antislavery, and Women’s Political
Identity. Chapel Hill: University of North Carolina Press.
28
Table 1. Expectations about Gag Rule Vote Choice and Vote Change
Voting for Gag
Risk of Position
Instability
+
-
+
+
+
x
x
+
+
+
x
+
+
x
+
Constituency Characteristics
Antislavery Vote (N)
Slave Population (S)
Region
Southern Border
Northwest
Crosspressuring
Southern Whig
Northern Democrat
Electoral Margin
North
South
Party (Whig=1)
House Party Control Change
(+)
(-)
(x)
Hypothesized positive relationship
Hypothesized negative relationship
No directional relationship hypothesized
29
Table 2. Gag Rule Vote Choice, 1836-1845
North
Beta
Vote Margin
Antislavery Dummy
b
-.709
RSEa
(p)
.010
(.064)
.246
(.004)
Beta
-.009
Slave Percent
-.016
Border
-.073
Northwest
a
.019
South
.526
Whigb
-3.780
Standing Rule
-1.095
ln(Stable Votes)
-2.206
Constant
.383
N
X2(df=6)
1386
174.39
(p<.001)
.227
(.021)
.356
(<.001)
.210
(<.001)
.273
(<.001)
.626
(.541)
-2.127
.406
-2.546
5.070
800
162.47
(p<.001)
Robust standard errors clustering on member ID.
Third-party members excluded from these models.
30
RSEa
(p)
.010
(.354)
.015
(.287)
.397
(.853)
.384
(<.001)
.336
(.227)
.462
(<.001)
.945
(<.001)
Table 3. Gag Rule Vote Change, 1836-1845
North
South
RSEa
(p)
.009
(.150)
.320
(.051)
Beta
Vote Margin
Antislavery Dummy
.013
-.624
Slave Percent
-.399
.992
Democratb
1.720
Standing Rule Change
1.439
ln(Stable Votes)
.240
-.505
Constant
-4.216
N
X2(df=7)
822
61.00
(p<.001)
a
b
RSEa
(p)
.011
(.001)
.018
(.655)
.484
(.410)
.269
(<.001)
Whigb
House Party Change
.035
.008
Border
Northwest
Beta
.389
(<.001)
.408
(<.001)
.492
(.626)
.295
(.086)
.611
(<.001)
1.878
.469
(<.001)
-.264
-.823
(.748)
.791
(.465)
.454
(.013)
1.098
(<.001)
.578
-1.131
-5.788
462
59.80
(p<.001)
Robust standard errors clustering on member ID.
Third party members excluded from these models.
31
Vote Number
Figure 1. Regional and Partisan Composition of Gag Rule Support over Time
1
2
3
4
5
6
7
8
9
10
11
12
0
20
40
60
80
Percentage of Total Pro-Gag Votes
S Dem Pct of Pro-Gag
NW Dem Pct of Pro-Gag
NE Dem Pct of Pro-Gag
32
100
S Whig Pct of Pro-Gag
NW Whig Pct of Pro-Gag
NE Whig Pct of Pro-Gag
Figure 2. Permanent Position Reversals Resulting from Replacement, 27th and 28th
Congresses
16
Number of Permanent Changes from Replacement
14
12
SDSD
10
SDSW
8
SWSD
SWSW
6
NWNW
4
NWND
2
NDNW
0
NDND
YtoN_27
NtoY_27
YtoN_28
NtoY_28
Direction and Congress
SDSD: Southern Democrat replaced by Southern Democrat
SDSW: Southern Democrat replaced by Southern Whig
SWSD: Southern Whig replaced by Southern Democrat
SWSW: Southern Whig replaced by Southern Whig
NWNW: Northern Whig replaced by Northern Whig
NWND: Northern Whig replaced by Northern Democrat
NDNW: Northern Democrat replaced by Northern Whig
NDND: Northern Democrat replaced by Northern Democrat
Note: Totals for the 27th Congress include one nay-to-yea replacement by special election during the 26th
Congress, and totals for the 28th include one yea-to-nay replacement during the 28th.
33