Causal inference and American political development: the case of the gag rule

See also Sheingate (2014), which we will refer to specifically in our conclusion.

On the method of process tracing, see Campbell (1975), George and McKeown (1985), King et al. (1994), Gerring (2007), Mahoney (2010) and Collier (2011). On “critical junctures”, see Pierson (2004) and Capoccia and Kelemen (2007). Our use of the term “critical juncture” departs from that of Pierson somewhat, to the degree that much of the politics surrounding the gag rule was about whether to set the House on a path that was then difficult to alter. As the narrative makes clear, while the gag remained on the House rulebook, it did so on increasingly narrow margins. In less than a decade of its initial passage, the gag rule was removed, never to reappear.

Throughout the present paper, party labels for individual members of Congress are taken from Martis (1989). Beginning in the 25th Congress, the main parties were the Democrats and the Whigs. In the 24th Congress, however, the partisan breakdowns would include Jacksonians, Anti-Jacksonians, Anti-Masons, and Nullifiers. The Jacksonians would become Democrats in the subsequent Congress, while the Anti-Jacksonians would become Whigs. The Nullifiers would eventually become Democrats in the 26th Congress.

See Second Annual Report of the American Anti-Slavery Society (1835); Fifth Annual Report of the American Anti-Slavery Society (1838); and Wesley (1944).

More extreme measures also were supported. For example, several southern communities posted a $50,000 reward for the capture and delivery of Arthur Tappan, dead or alive. See Richards (1970, pp. 50–52).

See Article I, Sect. 8, Clause 17.

According to Miller (1996, pp. 111–112), the select committee to which the petitions eventually were referred (the Pinckney committee)—one that “was not sympathetic to the petitioners”— claimed that 176 petitions were sumitted, with around 34,000 signatures. The Massachusetts Anti-Slavery Society, on the other hand, reported larger figures: at least 500 petitions, containing between 75,000 and 100,000 signatures.

For a rundown of various attempts to identify precisely the number of petitions and signatories, see Miller (1996, pp. 305–309) and Carpenter and Moore (2014).

Under the House Rules operating at the time, each day during the first thirty days of a session were petition days. After that, petitions were received every Monday. An important procedural accident is worth noting: the order for the call of states to receive petitions was geographical, starting in the North and proceeding South. Thus, the most anti-slavery part of the country got to lead off on each petition day. Because of that rule, wrangling over anti-slavery petitions served to obstruct action by southern House members who themselves wanted to be seen as helpful to constituents back home, by presenting their own petitions.

Register of Debates, 24th Congress, 1st Session (18 Dec. 1835, p. 1967). Also, see Freeman (2018) on the idea of “honor” as understood by southern members of Congress, who tended to conflate attacks on the region’s institutions with attacks on the region’s legislators. In that context, Hammond’s use of the word “assault” was anything but metaphorical.

Similar questions also were being debated throughout January and February 1836 in the Senate. Finally, on March 9, 1836, John Calhoun offered a motion very similar to Hammond’s, which called for peremptory rejection of anti-slavery petitions. Calhoun’s motion was defeated 36-10. Register of Debates, 24th Congress, 1st Session (9 March 1836, p. 779). Five days later, James Buchanan (Jacksonian-Penn.) proposed instead that anti-slavery petitions be received, but that the accompanying prayers for abolition immediately be rejected without consideration, which passed 34-6. Register of Debates, 24th Congress, 1st Session (14 March 1836, p. 810).

The delay owed to two related factors: (1) the Whigs had closed the gap between themselves and the Democrats considerably in the prior midterm elections and (2) five of New Jersey’s six House seats were being contested.

Congressional Globe, 26th Congress, 2nd Session (9 Dec. 1840, pp. 11–12). Adams spent the remainder of the lame-duck session arguing the Amistad case before the Supreme Court.

Congressional Globe, 27th Congress, 1st Session (16 June 1841, p. 63). It has been argued that the declining fortunes of the anti-gag forces in this series of votes was explained by the slow arrival of House members at the start of the Congress — the arriving stragglers changed the mix of sentiments in the House, tipping the balance in the direction of support for the gag rule. While literally true on the reconsideration vote, that view ignores the larger pattern of votes—the significant shifting in sentiments between the Adams resolution and the Fornance reconsideration motion among House members who had arrived on time. Nearly one-quarter of the supporters of the Adams resolution later voted to reconsider their previous support. However, that change in sentiment was not sufficient to force reconsideration. Thus, the ten members who voted on the Fornance reconsideration motion but had not voted on the Adams resolution were pivotal, supporting reconsideration 8–2 and providing the margin necessary for passage. However, it also turns out that the House members who were absent for the Adams resolution, but who voted on the final Stuart resolution, actually opposed the resolution 10–3. Of course, had the Fornance reconsideration motion not passed, the Stuart resolution wouldn’t have even been possible. Still, the early absentees would not have been pivotal had a quarter of Adams’s erstwhile support not abandoned him.

For coverage of these interactions, see The New York Herald, December 26, 1843, p. 2. Miller (1996, p. 473) offers various reasons why Wise dropped his support for the gag. One fairly plausible explanation is that he was up for a ministerial post to Brazil and was trying to curry favor with northern senators.

To what extent was the repeal of the gag rule a function of regional divisions within the Democratic Party stemming from its presidential nominating convention in May 1844? In the convention, Martin Van Buren garnered a majority of delegates, but was denied the nomination because of a two-thirds’ voting rule that had been instituted to placate southern delegates. In the end, southerners would not support Van Buren because of his lukewarm view of Texas’s annexation, leading eventually to the choice of James K. Polk (Tenn.) as the nominee. That southern “veto” created resentment within the northern ranks over the disproportionate influence of the southern “Slave Power”. See Silbey (2005). Little evidence exists, however, suggesting that such resentment was the critical factor in the gag rule’s repeal, as a result of northern Democrats “punishing” southern Democrats for opposing Van Buren. In fact, as Miller notes, the vote breakdown on the roll call that repealed the gag rule closely mirrored the vote breakdown on the roll call to add the gag back into the House’s rules in the first session of the 28th Congress. See Miller (1996, pp. 481–484). In effect, Miller argues that the gag rule was all but finished in the first session of the 28th Congress, long before the convention, but survived (temporarily) on a extremely close vote (88–87) because of a combination of factors, such as deliberate absences, procedural fatigue, and complicated parliamentary maneuvering. The repeal in the second session of the 28th Congress was, therefore, not related to a post-convention backlash, but rather was inevitable from a purely preference-based perspective.

House Journal, 29th Congress, 1st Session, (1 Dec. 1845, pp. 1–12). The record of Chapman’s motion is contained in the Journal, but no mention of it occurs in the Congressional Globe’s proceedings of the day.

The Liberty Party’s James Birney’s best showing in 1844 was just over 8% of the presidential votes cast in Massachusetts, New Hampshire, and Vermont See Dubin (2002, pp. 88, 89–90, 94–95).

Interestingly, the passage of the rule changes and the subsequent organization of the House were held up by the Whigs’ efforts to repeal the gag rule. For 2 weeks, the Whigs and Democrats battled over the gag, but the Whigs could not muster enough support to overcome the Democrats’ opposition. Finally, Whig party leaders decided that the organization of the chamber and the pursuance of their legislative agenda could be put off no longer, and they agreed to postpone the fight to repeal the gag rule until the following session.

We thank Joe Cooper for pointing this out to us.

Such streamlining went hand-in-hand with the minority-right restrictions passed earlier in the Congress.

A total of eleven roll-call votes were associated with the Pinckney resolution.

The exception is the first session of the 28th Congress, in which we examine two “key” votes.

Why did the various gag rules become more extreme over time? Freehling (1990) argues that the stringency of the gag became a “litmus test” for southern House members. That is, southern Whigs painted the southern Democrats’ support for the initially less stringent gags as evidence of their “softness” on the slavery issue, and made electoral hay on the claim in subsequent congressional elections. As a result, southern Whigs were forced to push for more stringent gags (like what became Rule 21) to prove their mettle.

One example of the preference shift was the vote in the 27th Congress (1st session) to reconsider the previous passage of the House rules. That is the vote that led, through a long string of parliamentary maneuvers, from the short-term repeal of the gag rule back to its reinstatement.

For more on the D-NOMINATE procedure, see Poole and Rosenthal (1997). For examples of APD research that incorporates D-NOMINATE scores, see Carpenter and Sin (2007), Jenkins and Nokken (2008) and Valelly (2009).

The correlations are -0.44 (24th Congress), − 0.48 (25th), − 0.48 (26th), − 0.49 (27th), − 0.37 (28th) and − 0.34 (29th).

The cutting line angles are calculated as follows. Define β₁ as the probit coefficient associated with the first dimension D-NOMINATE score, likewise for β₂. The tangent of the cutting line is defined as − β₁/β₂. The angle is the arctangent of that ratio.

The following argument contrasts with the work of Scott Meinke (2007), who posits a predominantly conversion-based explanation of the voting changes depicted herein.

Such high turnover between the 27th and 28th Congresses, driven largely by the economic downturn during the early 1840 s, occurred amidst reapportionment of seats. The Apportionment Act of 1842 reduced the size of the House from 242 to 223. Of the 19 seats lost, ten were from states that later would comprise the Confederacy.

We consider the latter vote to have been “easier” for northern Democrats because the Atherton “states’ rights gag” contained a declaration of principles to accompany the suppression of anti-slavery petitions. The latter vote merely suppressed the petitions without any declaration of principles. The removal of the inflammatory principles should have resulted in a few northern Democrats supporting the gag rule, as a party-building matter; the opposite in fact occurred.

It would be illuminating to identify the vote switchers in this sequence of votes more precisely; for instance, explaining vote switching in terms of local constituency pressure, party, D-NOMINATE scores, or electoral insecurity. The number of vote switchers is so small, however, that all of our efforts to throw the data at multivariate statistical techniques came to naught.

For evidence of an “electoral connection” between Members of Congress and their constituents during the antebellum era, see Bianco, Spence and Wilkerson (1996) and Carson and Engstrom (2005).

Professor Daniel Carpenter at Harvard University has undertaken a project focused on anti-slavery petitions during the era and has drawn our attention to a similar argument that makes the episode we eamine of interest to students of American politics and history more generally. See Carpenter (2003, 2016). One interpretation of the anti-slavery movement’s strategy is that abolitionists attempted to use petitions to create and mobilize mass public opinion—arguably a first in American politics. Heretofore, petitions to Congress, as was generally true of petitions to legislators in Anglo-American politics, were almost entirely confined to the redress of particular grievances—pressing foreign spoilation claims, requesting pensions, claiming a patent for an invention, requesting relief from a tariff, and so on. Rarely did constituents mention general issues, such as the state of the economy, and the continuation of the Bank of the United States, unless those issues impinged materially on an individual’s well-being. On the role of petitions in the creation of the “public sphere”, see Zaret (1996).

The framing evokes Riker’s (1982, 1986) idea of “heresthetics”. Ours is not strictly a case of successful heresthetic maneuvering, however, because the shift in dimensionality on the gag rule appears not to have happened through the agency of any one political actor, but rather through the follow-on effects of electing a growing number of northern Democrats whose electoral success depended on pleasing anti-slavery voters.