Reconsidering the “Southern Veto”: The Two-Thirds Rule at Democratic National Conventions, 1832–1936

Abstract

Beginning from 1832, the Democratic Party required a two-thirds majority at national conventions for the nomination of presidential candidates. Despite assessments that this “two-thirds rule” produced excessively long and disruptive nomination battles and low-quality presidential candidates, the rule survived until 1936. The rule's longevity is generally attributed to it functioning as a “Southern veto”: while the Democratic Party performed strongest electorally in the South in this period, the region's representation at conventions was small in comparison. By setting the bar for presidential nominations high, the South was given the ability to block unacceptable candidates. However, while the “Southern veto” argument is pervasive, there are little data and few concrete examples of Southern delegates blocking Democratic nominations through the two-thirds rule. In this paper, I reassess the two-thirds rule's history and appliance and show that Southern states barely had enough votes to block nominations and generally would need to vote against a candidate at a rate of nearly 90 percent to do so. As a result, the South almost never vetoed candidates: in only one case (Martin Van Buren in 1844) was Southern opposition pivotal in preventing a candidate with majority support from winning the nomination. Additionally, the two-thirds rule was generally accepted by broad majorities in the party (both Southern and non-Southern) and, while Southerners were among the defenders of the rule, representatives of the region were also among those opposing it. These findings suggest that the two-thirds rule rarely functioned as a Southern veto—not because the South had no power in the Democratic Party but because the necessity of maintaining intraparty consensus applied regardless of the existence of the two-thirds rule.

At the first Democratic National Convention—held in 1832—the Democratic Party set a surprisingly high bar for nominating its presidential and vice-presidential candidates. Unlike other political parties at the time, or the Whigs and Republicans later on, Democrats required a two-thirds majority of delegate support, thereby allowing a minority of delegates to block nominations. Scholarly and contemporaneous assessments of the effects of this “two-thirds rule” have largely been negative. The high bar to nomination has been blamed for causing extraordinarily long nomination fights—including the record-breaking 1924 convention, which required 103 ballots to come to a nomination. These extensive fights, as a delegate to the 1928 convention wrote, subjected the Democratic Party to “a terrific nervous strain and often leaves it on the verge of a presidential election in a state of spiritual exhaustion.”¹ Additionally, both contemporary political actors and scholars blamed the rule for producing low-quality candidates: as the San Francisco Daily Evening Bulletin noted in 1876, the rule's effect “for a great many years has been to bring on the stage mediocre men”² since more compelling (but, therefore, more controversial) candidates were believed to be unable to win two-thirds majorities.

Combined, a clear sentiment exists that the two-thirds rule was a net negative for the Democratic Party. Yet, despite these assessments, the two-thirds rule survived for a full century before it was finally abolished in 1936. This long survival of the rule is particularly surprising when considering that all other decisions at Democratic conventions—including whether to maintain the two-thirds rule—required only a simple majority. If a majority of delegates could have ended a (broadly considered) bad rule, why did the two-thirds rule survive for so long?

In answering this question, scholars have long argued that the two-thirds rule served a particular purpose within the Democratic Party. Throughout most of its existence, but particularly after the Civil War, the South was the area of the country where Democrats performed strongest electorally—largely due to Southern Democrats’ disenfranchisement of Black voters. However, the size of each state's delegation at Democratic national conventions in this period was linked to states’ shares of the electoral college vote, regardless of whether they voted for the party or not. As a result, the South's representation at national conventions was smaller than its position within the party arguably warranted. The two-thirds rule provided a solution: because the bar to nomination was so high, the South was effectively granted a veto over Democratic presidential selection. While Southern states could not single-handedly nominate a presidential candidate, they could veto candidates they found objectionable. As a result, under the two-thirds rule, any Democratic nominee would “have to be acceptable to the South insofar as he pledged not to interfere with the southern social and economic caste system.”³

But while the idea of the two-thirds rule as a Southern veto is pervasive, scholarly literature presents little data and even few concrete examples of the South successfully vetoing a presidential candidate in this way. To be sure, at least one clear example of such a Southern veto does exist: in 1844, Martin Van Buren received support from a majority of delegates but failed to win a two-thirds majority. Southern opposition to Van Buren was pivotal in his defeat: even if all non-Southern delegates had supported Van Buren, he would still have been unable to win the nomination. But while the Van Buren case is important, it is unclear how common such Southern vetoes were otherwise. Nor is there much known about how cohesive the South's voting would need to be to veto candidates at Democratic conventions in this period.

The goal of this article is to assess whether the two-thirds rule was indeed systematically used by the South as a veto to control presidential selection at conventions between 1832 and 1932. To test the “Southern veto” thesis, I focus on three questions that should clarify the extent to which the South could and did use the rule to dominate presidential selection within the Democratic Party in this period. First, I look at whether Southern delegates had enough votes at Democratic national conventions to block nominations if they should choose to. Second, I look at how often nominees failed to win the Democratic presidential nomination due to Southern opposition. Finally, I identify how frequently the two-thirds rule was challenged—and, when it was—whether support and opposition for the rule fell along regional lines.

The answers to each of these questions suggest that the two-thirds rule largely did not function as a Southern veto. First, the “South” as a region had enough votes to prevent a nomination only if we rely on a relatively broad definition of the South as existing of both the eleven states that made up the Confederacy and all possible Southern border states. Even then, internal cohesion within these states needed to be high to block a nomination: from 1876 onward, Southern delegates needed to vote together in opposition to a leading candidate by at least 89 percent or more. And this number increased over time: by 1932, a Southern veto required 99 percent of Southern delegates to vote against a leading candidate. Thus, a truly Southern veto required a coalition of “broad” Southern states to be in near-complete consensus that a candidate was unacceptable.

Second (and likely related), actual Southern vetoes were exceptionally rare. Across all conventions, by far most candidates who led on the first ballot eventually won their nomination. Looking at ballots on which a candidate received the most votes but failed to win the nomination between 1832 and 1932, in only two instances was Southern opposition so unified that the candidate could not have won the nomination even if they had received full support from all non-Southern delegates: Van Buren in 1844 and Al Smith in 1924. However, Smith was altogether far removed from the nomination, with roughly 50 percent of non-Southern delegates also opposing him. Thus, the only case of a “true” Southern veto by way of the two-thirds rule is the failed Van Buren nomination in 1844.

Third, challenges to the two-thirds rule were rare. At by far most conventions, the two-thirds rule was reconfirmed by substantial majorities of both Southern and non-Southern delegates. In the limited cases where the rule was debated and challenged, Southern delegates often supported the rule, but that support was rarely unanimous. Indeed, in some cases (such as in 1876 and 1932), opposition to the rule largely came from Southern party leaders and delegates who were concerned that their favored candidates would be unable to meet the high delegate support bar necessary for a nomination. Meanwhile, when challenges did occur, the rule's survival always depended on robust support from the non-South. And, in each case prior to 1936, delegates from outside of the South supported the rule in substantial numbers.

These findings suggest that the two-thirds rule was not, in fact, used as a Southern veto. This does not mean that the South had no meaningful influence over the selection of Democratic presidential candidates in this period. Instead, the two-thirds rule may have largely been a symbolic one. Throughout the period in which the rule applied, the Democratic Party existed of a coalition of different groups with conflicting preferences. Nominating a presidential candidate at a convention represented only a part of the puzzle of maintaining such a party: even if a small majority would have been able to force a preferred candidate on the party at a convention, the candidate would be unable to win without support from the party's minority factions in the general election. This was particularly true for Southern electoral support: while the South alone could not elect Democrats to the White House, a Democratic nominee without Southern support would need to radically reinvent the electoral map to have any chance at all at winning. Prior to the New Deal realignment, the South's support was thus practically nonnegotiable. This meant that, throughout the period in which the rule was maintained, Democratic leaders faced the political reality that broad intraparty consensus was a necessity independent of whether a two-thirds majority was actually required for presidential nominations or not. The long survival of the two-thirds rule, then, may have been less an attempt at providing the South with a weapon it needed to control the party, but rather a de-facto public acknowledgement of what was obvious to all participants involved: that any nominee would need to be acceptable to the South, as well as other relevant constituencies within the party.

1. Democratic National Conventions and the two-thirds rule

In the late eighteenth and early nineteenth centuries, newly formed American political parties largely nominated presidential candidates through Congressional caucus votes. Starting in 1796—the first presidential election without George Washington as a candidate—John Adams, Thomas Pinckney, Thomas Jefferson, and Aaron Burr were “nominated” by “informal caucuses of the congressional members of the respective parties.”⁴ In subsequent years, parties took mostly similar approaches. The Democratic-Republicans selected their candidates through Congressional caucuses between 1800 and 1820. The Federalists relied on conventions in 1808 and 1812⁵ though “these meetings were […] quite unlike the conventions subsequently introduced in the Jacksonian period” as they existed of “small groups of Federalist leaders designated in widely different ways” who “met in secrecy.”⁶ By 1824, there no longer was a coherent party process of selecting nominees for the Democratic-Republican party, and candidates were instead “selected” through a variety of different approaches—including nomination by state legislatures.

After Andrew Jackson's victory in the presidential election of 1828, multiple parties began organizing national conventions in the run-up to the 1832 election. The Anti-Masonic party nominated William Wirt at a convention in September 1831, while the National Republicans nominated Henry Clay at their national convention in December 1831.⁷ The Democratic Party saw little disagreement about renominating Jackson. However, a major schism existed within the party between supporters of Vice President John C. Calhoun and former Secretary of State Martin Van Buren. Over the course of his administration, Jackson had become estranged from Calhoun and had determined that Van Buren should be his next running mate and (over time) successor as president. To ensure success in this regard, Jackson required a nomination method that would not allow for multiple candidates to claim the “title” of Democratic vice-presidential candidate, which nomination by state legislatures would make extremely likely. Indeed, by spring of 1831, several names (aside from Van Buren and Calhoun) were in circulation in this regard. Thus, the Democrats also organized their first national convention, which, as Chase argues, “originated in the endless list of names advanced for the vice-presidency and the consequent need to solidify the party behind a single ticket.”⁸

At this first Democratic convention—which met in May 1832 in Baltimore—the delegates decided to require a two-thirds majority for the nomination of the vice-presidential candidate. Specifically, the convention described the process as follows:

Resolved, That the delegates from each State in this convention, be entitled to as many votes in selecting a suitable person for the office of Vice-President, as such State will be entitled to in the Electoral College […] and that two-thirds of the whole number of the votes given be required for a nomination, and on all questions connected therewith.⁹

While the rule agreed on in 1832 only concerned the question of a vice-presidential nomination (and there was no vote on Jackson's presidential renomination) subsequent Democratic conventions applied the two-thirds rule to the selection of the entire presidential ticket. Notably, all other decisions at these conventions—including settling disputes regarding the seating of delegates, the selection of officers of the convention, determining a party platform, and setting rules for the convention itself—required only a simple majority. From the start, the two-thirds rule was a unique element to Democratic conventions: neither the Anti-Masonic party nor the National Republicans relied on a two-thirds rule to select their nominees for the 1832 election. And, later, after the formation of the Whig and Republican parties, they too relied on simple majorities in selecting their presidential tickets. Why did the Jacksonian Democrats decide to set such a high bar for selecting a presidential ticket?

One argument—presented by Senator William R. King (AL) who chaired the rules committee at the 1832 convention—was that “a nomination made by two-thirds of the whole body would show a more general concurrence of sentiment in favor of a particular individual, would carry with it a greater moral weight and would be more favorable received than one made by a smaller number.”¹⁰ Others argued the two-thirds rule was intended to block attempts at nominating Calhoun—though, given that the Jacksonians aimed to use the convention to nominate Van Buren, increasing the bar for a nomination would have also complicated their own efforts.¹¹ Chase notes that an alternative explanation is that setting a high bar for a nomination could help “meet criticism leveled at the convention prior to its meeting”: specifically, opponents of the national convention had argued a nominee would “not represent the whole party, since each state was to cast the same number of votes as it had in the Electoral College without reference to its Democratic strength.”¹² As a result, states that were unlikely to vote for Jackson in the presidential election could theoretically nominate the vice-president. Romulus Saunders, a delegate from North Carolina, explained how the two-thirds rule provided a solution: “the question of what States would be allowed to vote, threatened a […] serious disturbance to the harmony of our proceedings” since blocking states from voting entirely would be “unjust and impolitic” but allowing those same states to nominate a vice-president by themselves was unacceptable as well. The two-thirds rule would “obviate this difficulty, and […] guard against the objection of suffering those non-effective States to decide the question.”¹³ By setting a two-thirds majority for nominations, the 1832 convention thus tried to solve a problem of legitimacy and balancing power dynamics within the party on the basis of geographic electoral strength. On the one hand, states where the party underperformed still were allowed a major say in intraparty matters. On the other hand, a two-thirds majority made it so that those states could not single-handedly determine the make-up of the presidential ticket.

From 1832 onward, each Democratic convention until 1936 applied the two-thirds rule in some form.¹⁴ But while there may have been good logic for introducing the rule in 1832, by most accounts, the continued reliance on the rule had considerable negative effects. For example, Aldrich notes that the rule was “controversial and at times costly” to the party since it ensured lengthy nomination ballot fights—including the 1924 convention which required a record 103 ballots to result in a presidential nomination—“and such rancorous and lengthy divisions could only hurt the party at the polls.”¹⁵

Additionally, the two-thirds rule was blamed for producing low-quality nominees. As Nichols has argued, the original aim of the Democratic convention system was to effectively ratify decisions made by the national party leaders (that is, nominate Van Buren as the vice-presidential nominee in 1832 and, later, as the presidential nominee in 1836).¹⁶ However, in practice, the new convention system meant that—as Ceaser notes—

the power of choosing the nominees came increasingly into the hands of state and local politicians, men who may have been less inclined than Congressmen to view problems from a national perspective and more disposed, at least in the long-run, to introduce narrow considerations of local patronage in the selection process.¹⁷

To be sure, this was not a problem unique to the Democratic convention: battles over federal patronage would become a defining characteristic of Republican Party convention politics as well.¹⁸ But the kind of negotiations and haggling these local party leaders engaged in certainly could have been further complicated with the added hurdle of having to produce a two-thirds majority. David, Goldman, and Bain argue the rule “was a major factor in producing the weak executive leadership of the period, since its normal effect was to eliminate anyone with sufficient character and record to have aroused serious opposition.”¹⁹ This assessment was not limited to retroactive analysis from scholars: Gilchrist B. Stockton, a delegate at the 1928 convention and the author of an extensive study published that year on the history of the two-thirds rule, described it as

a cancer continually sapping the vigor and unity of the Party. Whenever an outstanding and generally accepted candidate has gone into a convention, there has been no efficacy in this rule, for such a candidate is usually nominated by an overwhelming vote on the first ballot […]. On the other hand, whenever there have been a number of candidates of almost equal merit and almost equal popular support, it has been a tedious proceeding to obtain two-thirds for any of them.²⁰

Contemporaneous newspaper coverage was also often critical of the rule. In 1876, for example, the New York Herald argued that the rule

tends to imbecility in the government by additional facilities for foisting weak and obscure men into the highest position. It is hard enough to unite the majority of a convention on a really good candidate, but when one-third of the members are permitted to override such a majority weakness and obscurity have every advantage over established character and recognized merit.²¹

That same year, the San Francisco Daily Evening Bulletin made a similar argument, claiming the rule's “usual effect has been to kill off prominent men” as “no one can be long in public life without contracting more or less enmity. Under the two-thirds rule that enmity is always sufficient to shelve him.”²²

The common perspective that the two-thirds rule hurt the Democratic Party raises a puzzle regarding the rule's century-long survival. After all, while nominating a presidential ticket required two-thirds majorities, the specific rules each convention applied to the nominating process were determined by simple majorities. This meant that at any convention between 1832 and 1936, a majority of Democratic delegates could have ended the rule. Why did the two-thirds rule survive so long given both its downsides and this (seemingly) easy solution?

2. The two-thirds rule as a Southern Veto

The common explanation for the longevity of the two-thirds rule has focused on the South and its role in the Democratic Party. For much of the period in which the rule was maintained, the South was the region where Democrats performed strongest electorally, largely because of Southern Democrats’ disenfranchisement of Black voters. With the Democratic Party performing inconsistently in other parts of the nation in the first decades after the Civil War, and with Republicans dominating the non-South during the System of 1896, the South had a particularly important position within the party. Indeed, during the first decades of the twentieth century, it was not uncommon for the region to provide the party with a majority of its seats in the House and Senate and in electoral votes in presidential elections.

But while the South dominated the Democratic party-in-office, the region had no such strength at Democratic conventions since the number of delegates awarded to each state was still linked to their electoral votes, regardless of whether the states voted for Democratic candidates. As a result, the South's representation at Democratic conventions was lower than its contribution to the party's electoral performance arguably warranted.²³ The South's position in the party in relation to its delegation size represented a problem from the perspective of intraparty management. For example, similar to the arguments made in 1832, delegates at the 1844 convention noted that the delegate apportionment system meant that “delegations from states not representing a majority of the Party” could nominate presidential and vice-presidential candidates, “which might jeopardize the Party's interests at the polls.”²⁴ This was—at least theoretically—a particular concern to the South as it introduced the possibility that non-Southern states could nominate a presidential candidate hostile to Southern interests—that is, maintaining slavery and (later) segregation.

Scholars have long argued that the two-thirds rule functioned as a solution to this problem since it provided the South with an ability to veto such undesirable potential nominees: while the division of delegates did not allow the South to select a nominee, the two-thirds rule could allow it to prevent selection when it believed candidates were in opposition to Southern interests. As Aldrich notes, requiring two-thirds of delegate support during the Jacksonian era meant “the South held an effective veto over nominations in the Democratic party for president and vice-president in the face of a free-state delegate majority.”²⁵ By placing a higher threshold for nominating a presidential candidate the

two-thirds rule effectively ensured that the South would have to agree to any nominee, made certain that no extremist, whether pro- or antislave, could be nominated, helped produce balanced tickets, and effectively attained and maintained the intersectional alliance in the Jacksonian Democratic Party.²⁶

Other political scientists and historians have made the same argument. For example, Bass observes that the rule throughout its existence was defended by “chiefly southern-based advocates” who believed “its retention as protective of their philosophical and regional interests”²⁷ and that criticism of the two-thirds rule only increased as “the Democratic Party began to expand from its southern base into the north and west.”²⁸ Milkis describes the rule as originating “in the South, which regarded it as vital hallowed protection against the nomination of candidates unsympathetic to its problems” and that, even after the Civil War, the rule protected the South “against the imposition of an unwanted nominee by the […] North, East, and West.”²⁹ Rae claims the rule institutionalized a “southern veto” which ensured that any Democratic nominee would “have to be acceptable to the South insofar as he pledged not to interfere with the southern social and economic caste system.”³⁰ And Landis has argued that the rule meant “Southerners, though a numeric minority, could, with the aid of a few willing Northerners, dictate policy and candidates.”³¹ Meanwhile, other scholars have argued that the abolition of the two-thirds rule in 1936 meant the South was no longer able to exert the kind of control it previously had. For example, Boyd wrote that the nomination of Harry Truman in 1948 (in the face of considerable Southern opposition and a Dixiecrat walk-out at the convention) came about “because of the abolition of the two-thirds rule in the Convention.”³²

To be sure, some scholars have been cautious about describing the rule as centering on the South's ability to veto candidates. Nonetheless, they too generally identify the two-thirds rule as a mostly Southern tool used to try and control the Democratic Party. For example, Stockton referred to it as a Southern “weapon,”³³ while Black and Black portrayed the two-thirds rule as allowing the region “to exert considerable leverage over the policies and candidates of the party.”³⁴ And Reiter described the two-thirds rule as one of the South's “major weapons”³⁵ in intraparty politics.

The scholarly argument for the survival of the two-thirds rule is thus that the South—due to its special position within the Democratic Party—was underrepresented at Democratic conventions and that it required some level of protection against other parts of the country nominating candidates it strongly opposed. The two-thirds rule provided a solution: if Southern delegates found a potential nominee (backed by non-Southerners) to be objectionable it could block their nomination by collectively voting for other candidates.³⁶

3. Reassessing Southern reliance on the two-thirds rule

It is self-evident that, when the threshold for a convention nomination rises, it becomes easier for any minority group of delegates to try and prevent a candidate from being nominated. Since the South represented a minority interest at Democratic national conventions in this period, the two-thirds rule on that basis would have assisted its efforts to prevent any potential unwanted nominations. But this logic applies to any minority combination of delegates—whether they be Southerners or non-Southerners. To what extent did the South in particular rely on the two-thirds rule to veto potential Democratic presidential nominees it found objectionable?

There is one very notable and important example of the South doing exactly that. At the 1844 convention, former president Martin Van Buren attempted a political comeback and received a simple majority on the first ballot. However, he remained short of the two-thirds majority on the first and each subsequent ballot (see Figure 1). As balloting continued, Van Buren's vote dropped and the convention nominated James K. Polk. Crucially, the failed renomination of Van Buren was the product of Southern opposition: on the first and strongest ballot, Van Buren was thirty-one votes short of the nomination and the bulk of opposition came from Southern delegates. Even if Van Buren had received unanimous support from the non-Southern states, Southern opposition was so unified that it would not have been enough for him to win the nomination.

Sources: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition); “Proceedings of the Conventions,” Baltimore Sun, May 29, 1844.Notes: On the ninth ballot, as it became clear Polk would win the nomination, several states shifted their original votes. The original ninth ballot votes are presented as 9a, the votes after these switches are presented as 9b. The horizontal line denotes the number of votes required to win the nomination.

Figure 1. Candidate Performance by Ballot at the 1844 Democratic National Convention.

But while 1844 is an important example, it is not clear how common Southern vetoes were otherwise. Nor is it apparent that the South was particularly more likely to use such a veto in comparison to other groups within the party. Indeed, the two-thirds weapon was sometimes used against a candidate with considerable support from the South. In 1924, William G. McAdoo received majority support of Southern delegates but failed to win the nomination due to non-Southern opposition—leading Stockton to conclude that the rule had now been “effectively turned against”³⁷ the South. Meanwhile, at the 1932 convention, Franklin Delano Roosevelt won comfortable simple majorities with considerable Southern support on the first three convention ballots but remained below the two-thirds bar. After the third ballot, there was concern in the Roosevelt camp that delegates could begin abandoning him unless a breakthrough was forged on the fourth ballot.³⁸

Relatedly, some scholars have noted that the South's position on the two-thirds rule was more complex than we might expect if the rule was understood by all to be a Southern veto. For example, David, Goldman, and Bain acknowledge that “the South has frequently been credited with responsibility for preventing the rules amendment until 1936” and “provided most of the active leadership in defending the old rule.”³⁹ But they also point out that Southern delegations were split on maintaining the rule when it was challenged at the 1932 convention. And, while, when the two-thirds rule was abolished in 1936, opposition was strongest among delegates of Southern states, a majority of Southern states supported ending the rule.⁴⁰

Combined, then, there is reason to question to what extent the two-thirds rule really represented a Southern veto and how often it was applied as such. Answering these questions is complicated in no small part because the existence of the two-thirds rule possibly also affected strategic behavior before and during conventions. That is, throughout this period, Democratic politicians likely functioned under the assumption that any presidential nomination at an upcoming convention would need a two-thirds majority. As a result, the persistence of the rule may have had a preemptive effect on the kind of strategic choices these actors made. In particular, the rule could have functioned as a form of a “second face of power,” in which a group of actors under the control of a more powerful group decides not to engage in actions challenging that power because they know they cannot win a potential confrontation.⁴¹ If a majority of convention delegates from outside of the South believed a potential candidate would never be able to receive the necessary two-thirds majority to actually be nominated due to Southern opposition, they may have decided simply not to introduce the candidate or to not vote for them to avoid confrontations. However, it is worth remembering that the non-Southern states needed only a simple majority to end the two-thirds rule. While the “non-South” was not a monolithic political region within the party, if the argument is that the persistence of the two-thirds rule reflected a Southern stranglehold over the interests of other regions, we might expect to see at least some evidence of attempts by non-Southerners to end it and of Southerners to defend the rule.

To assess to what extent any of this occurred, I look at three types of data. First, I establish whether Southern states had enough votes at Democratic conventions to block nominations, and at what rate they would need to vote together against a candidate to successfully do so. Second, I look at the extent to which the South was actually pivotal in preventing nominations from occurring. To do so, I look at both successful nominations and at failed candidates, the extent to which they were opposed by the South, and whether Southern opposition was pivotal to failed nominations. Finally, I look at how often attempts were made to end the rule and any evidence of consistency in regional support and opposition to such attempts.

4. Defining the South

Before doing any of this, it is first necessary to define what combination of states actually counts as “the South.” While it is common to refer to the South as a comprehensive region with particular relevance in American political history, there is no universally agreed upon definition for what states make up its membership. As Bullock and Rozell note in The Oxford Handbook of Southern Politics, “various suggestions have been made concerning which states constitute the South” but the “most common definition considers the South to be the eleven states that seceded from the Union and formed the Confederacy”⁴²—that is, Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas, and Virginia. Indeed, Confederacy membership is the definition used by a number of scholars of the South, including Lublin, Black and Black, and others.⁴³ Alternatively, scholars have considered the South to consist of these eleven states but on a different definitional basis. For example, V.O. Key Jr., in Southern Politics in State and Nation, focuses on the same states but based on their support for Democratic candidates in presidential elections after they were brought back into the union.⁴⁴

Focusing on the Confederate South as a comprehensive region during the Civil War and in the subsequent periods of Reconstruction, Redemption, and Jim Crowism (and beyond) is sensible—though it is worth noting that even within this region relevant political differentiation existed.⁴⁵ However, relying on the Confederacy as the core definition of what “counts” as the South introduces the problem that the set of states that ended up seceding from the union after the 1860 election were not predetermined to be that exact set of states. That is, while from 1861 onward, viewing the South as the (former) Confederacy may be reasonable, prior to the start of the Civil War there were other states that also could have ended up seceding but (for various reasons) did not. Even after the Civil War a subset of Southern border states shared the Confederate South's general support for segregation. On this basis, other, more expansive, definitions of the South also exist. In addition to the Confederate states, the U.S. Census Bureau considers Delaware, Kentucky, Maryland, Oklahoma, West Virginia, and the District of Columbia to be Southern. However, of those additional “Census states,” the Southern Political Report considers only Kentucky and Oklahoma to be Southern states. Alternative definitions include all former slave states not part of the Confederacy, which would also add Missouri.⁴⁶ Meanwhile, Bateman, Katznelson, and Lapinski define the South as the seventeen states “in which racial segregation in schools was mandated by law before Brown v. Board of Education was adjudicated by the Supreme Court in 1954.”⁴⁷ And Price relies on a scale of “Southerness” incorporating (amongst others) support for Strom Thurmond during the 1948 presidential election, the size of Black population in states in 1950, whether states allowed slavery in 1860, and states requiring school segregation as of 1954.⁴⁸

Determining which definition of the South is the “correct” one across American history is, perhaps, impossible, and certainly outside of the scope of this paper. For the purposes of this study, the goal is to rely on a broad measure of the South (so as not to miss any Southern veto efforts through the exclusion of relevant states) while also remaining cautious about the fact that a more expansive definition of the South runs the risk of overestimating the impact of these states as a comprehensive regional political force. To cover all bases, I, therefore, rely on a combination of definitional approaches and present data for the Confederate South, the border states, and the combination of the two, which I will refer to as the “broad South” (see Table 1).

Table 1. Defining the South

5. Southern delegate shares at Democratic conventions

To what extent were Southern delegates able to act as a veto on presidential nominations through the two-thirds rule? To answer this question, we can identify the percentage of delegates Southern states held at each convention in comparison to the total number of delegates and see whether Southern states had enough votes to block a nomination. Doing so is possible for all conventions with one exception: at the 1840 convention, incumbent president Martin Van Buren was renominated by acclamation. While this is not uncommon, 1840 is a unique case since there were no other votes at the convention.⁴⁹ As a result, there are no ballots indicating the exact number of votes each state had. While a list of delegates does exist, the number of individual delegates each state sent to a convention did not necessarily align with the number of votes the states were given. For example, in the proceedings for the 1832 convention, only one delegate is listed representing Massachusetts even though the state had fourteen votes. As a result, it is not possible to identify the actual division of votes each state had in 1840.⁵⁰ For all other conventions, we do have these numbers. Figure 2 shows the percentage of Confederate, border, and broad Southern delegates for each convention except 1840 between 1832 and 1936. In the figure, the horizontal line reflects the two-thirds majority bar necessary for a nomination to be successful.

Sources: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition); Proceedings of the Democratic National Convention held at Baltimore, June 1-5 1852 (Washington DC: Robert  Armstrong, 1852); Official Proceedings of the National Democratic Convention held at New York, July 4-9, 1868 (Boston, MA: Rockwell & Rollins Printers, 1868); Official Proceedings of the National Democratic Convention held in St. Louis, Missouri, June 5-7 1888 (St. Louis, MO: Woodward & Tiernan Printing Co., 1888); Official Proceedings of the Democratic National Convention held in Chicago, Ill., July 7-11, 1896 (Logansport, IN: Wilson, Humphreys & Co., 1896); Official Report of the Proceedings of the Democratic National Convention held in Baltimore, Maryland, June 25-July 2, 1912 (Chicago, IL: The Peterson Linotyping Co., 1912); Official Report of the Proceedings of the Democratic National Convention held in San Francisco, California, June 28-July 6, 1920 (Indianapolis, IN: Bookwalter-Ball Printing Company, 1920); Official Report of the Proceedings of the Democratic National Convention held in Madison Square Garden, New  York City, June 24-July 9, 1924 (Indianapolis, IN: Bookwalter-Ball-Greathouse Printing Company, 1924); Official Report of the Proceedings of the Democratic National Convention held at Philadelphia, Pennsylvania, June 23-27, 1936 (Washington DC: Democratic National Convention, 1936).

Figure 2. Percentage of Southern Delegate  Vote at Democratic National Conventions, 1832–1936.

These data immediately underline the importance of how to define the South. The Confederate South at no convention had enough delegates to veto a presidential nomination. Indeed, over the entire period 1832–1936, the Confederate South, on average, represented just 25.6 percent of convention votes.⁵¹ This average reflects two broad periods of time: before and after the Civil War. Between 1832 and 1860, the states that would go on to form the Confederacy on average held 29.4 percent of delegate votes. At the 1864 Civil War convention, no Confederate delegates were present. After delegates from the now former-Confederate states were again represented in 1868, the vote share for their region began to decline: between 1868 and 1932, as more states were admitted and territories were also granted convention votes, the Confederate South's delegate vote share dropped to an average of just 24.2 percent.

But including the border states produces a considerably stronger position for the South. In 1832, the broad South controlled 42 percent of convention delegates—more than enough to veto presidential or vice-presidential candidates. And while this share of the total number of delegates also declined over time, it consistently remained above the number necessary to function as a (broad) Southern veto. On average, the broad South through the period 1832–1860 (arguably a more correct way of thinking about the South as a political region since the existing border states at the time also practiced slavery) held 40.2 percent of convention votes, while for the period 1868–1932 the average was 35.9 percent. To be sure, even the broad South saw its delegate share drop over time: by 1932, the broad South held 33.8 percent of convention votes, just barely enough to prevent an undesirable candidate from winning the nomination without non-Southern support in a veto-effort. This means that, throughout the century-long existence of the two-thirds rule, for the South to block nominations it would need to vote in opposition in a relatively unified manner across both Confederate and border states.

Table 2 presents the necessary number of votes to block a nomination at each convention and the percentage of broad Southern delegates that needed to vote in opposition together to meet that threshold. These numbers show Southern consensus needed to be very high in by far most years. With the exception of the convention of 1864 (in which the border Southern states had nowhere near enough votes to block an undesirable candidate), the broad South at each convention prior to 1936 had the theoretical possibility to block a nomination. But to do so, Southern states needed, on average, a 90.7 percent level of delegate opposition to the leading candidate. And, as the South's share of convention votes decreased, that level of required consensus went up. By the 1920s, a Southern two-thirds veto required roughly 97 percent of Southern Confederate and border delegates to oppose a candidate. By 1932, that number was 99 percent. And, had the 1936 convention maintained the two-thirds rule, even complete consensus could not have blocked a nomination: while 368 votes would have been needed to deny a nomination, the broad South by then only had 366 delegates.⁵²

Table 2. Southern Consensus Required to Veto Nominations

Sources: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition); Official Proceedings of the National Democratic Convention held in St. Louis, Missouri, June 5-7 1888 (St. Louis, MO: Woodward & Tiernan Printing Co., 1888); Official Report of the Proceedings of the Democratic National Convention held at Philadelphia, Pennsylvania, June 23-27, 1936 (Washington DC: Democratic National Convention, 1936).

To be sure, Southern delegates need not have relied exclusively on their own region to block a nomination: if non-Southerners also opposed a candidate, some variation in Southern voting would still allow for a two-thirds veto. Given the high level of delegate consensus necessary to achieve a nomination, a sizeable Southern bloc in combination with opposition from other regions could very well sink a nomination. But the more such blockades would have relied on mixed regional opposition, the less realistic it would be to interpret the rule as reflecting a “Southern veto.” And the numbers presented here suggest that, while a Southern veto was possible (and, as we saw in the case of 1844, not just in a purely theoretical way), to achieve such a veto without substantial non-Southern support required considerable regional consistency. While Confederate and border states would not have to agree on which candidate to support, they would need to vote together against a candidate at a remarkably high level to produce a Southern veto.

6. Southern opposition to presidential contenders

To what extent then did Southern delegates actually vote collectively to block presidential nominations? To assess this, I collected data from presidential nominating ballots at each convention between 1832 and 1932. While many Democratic conventions in this period were able to nominate a candidate on the first ballot, in other cases additional rounds of voting were required. To provide a manageable oversight, I focus on two types of ballots: (1) the deciding ballot on which a candidate was nominated and (2) any ballots on which a candidate received the most votes and came closest to winning the nomination but failed to do so. Combined, this provides a universe of candidates who either were actually nominated or who had a reasonable chance of winning the nomination but did not.

Table 3 shows the results of each deciding ballot at Democratic conventions across this period. In 1832, 1840, 1888, and 1916, no actual vote was held since the nominee (in each case an incumbent president running for reelection) was nominated by acclamation. Aside from these years, the nominated candidates generally received considerable support from both Southern and non-Southern delegates.⁵³ Half of Democratic nominees won with over 80 percent of the vote and with little to no Southern opposition. Even in closer races, five of the nominees (Cass in 1848, Tilden in 1876, Bryan in 1896, Parker in 1904, Cox in 1920, and Davis in 1924) won majority support from Confederate and border Southern delegates.

Table 3. Southern Opposition to Democratic Presidential Nominees on Final Convention Ballots, 1832–1932

Source: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition).

Notes:

^a Nomination by acclamation.

^b Numbers reflect votes after switch within the same ballot.

^cThe convention chair interpreted the two-thirds rule to apply to the percentage won of the number of votes cast not the total number of delegate votes.

^d The 1860 convention was split into two conventions: the first took place in Charleston and ended after fifty-seven ballots without a nominee. The second reconvened—without many Southern delegates—in Baltimore and nominated Douglas on its second ballot. Opposition percentages include votes not cast.

^eNo delegates from Confederate states present at this convention.

Of course, focusing on the actual nominees is not a great test of the Southern veto thesis as these candidates were the ones who were able to overcome the two-thirds hurdle. Nonetheless, the data in Table 3 do provide a small number of cases in which the South as a region was “rolled” and failed to veto candidates a majority of Southern delegates opposed. In two convention years, the eventual candidate was opposed by a majority of broad Southern delegates (1860 and 1928), while in 1892 the nominee was opposed by a majority of Confederate Southern delegates. The most peculiar of these concerns the 1860 nomination of Stephen Douglas: Democratic delegates had gathered in Charleston, South Carolina, in April 1860, but were unable to come to a nomination. After fifty-seven ballots, the convention disbanded and reconvened again in Baltimore in June. While the two-thirds rule applied at both conventions, its interpretation shifted. At Charleston, the rule was interpreted to apply to the total number of votes as a concession to Southern delegates after a number of Southerners had left the convention. However, after more Southern defections, the Baltimore version of the convention decided to interpret the two-thirds rule as applying to just the number of votes cast. As a result, Douglas was nominated with a little less than 63 percent of all possible delegate votes.⁵⁴

The two other cases concern Grover Cleveland's third nomination in 1892 and Al Smith's nomination in 1928. In both cases, the failure to veto largely was the product of a lack of unity in Southern opposition. Of the two, Cleveland was the least controversial: while a small majority of delegates from the Confederate South opposed his candidacy, a considerable majority of delegates from border Southern states supported him. As a result, the Southern opposition consensus percentage was just 44 percent—well below the nearly 94 percent needed to veto the nomination. Smith saw opposition of two-thirds of Confederate Southern delegates and 41 percent of border state delegates. Combined, 58.7 percent of broad Southern delegates opposed Smith, also well below the now even higher threshold of consensus necessary. Indeed, the fact that the South was unable to block Smith—whose Catholicism and opposition to Prohibition made him particularly unappealing to Southern Democrats—underlines just how hard it was for the South to have enough internal consensus to block a candidate.

The more important question is whether Southern delegates were pivotal in blocking any of the failed nominees. To produce a comprehensive list of failed potential nominees, I identified every candidate who at one point during the nomination process received the most votes on a ballot, but still failed to win the nomination. For each of those candidates, I look at Southern opposition and whether it was pivotal in sinking their nomination.⁵⁵ Table 4 presents the complete list of these candidates. The first thing to note is that the number of failed possible nominees by this definition is surprisingly small. Indeed, at by far most Democratic conventions the candidate who eventually became the nominee (regardless of how many rounds of voting it took for them to hit a two-thirds majority) always was the highest performing candidate: in 1832, 1836, 1840, 1848, 1856, 1860, 1864, 1872, 1876, 1880, 1884, 1888, 1892, 1900, 1904, 1908, 1916, 1928, and 1932, the candidate who was nominated also received the most votes on every single ballot or was nominated by acclamation. At the remaining seven conventions, only fourteen candidates at some point in the balloting process received the most votes on at least one ballot but were not nominated.

Table 4. Southern Opposition to Failed Democratic Presidential Contenders, 1832–1932

Sources: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition); “Proceedings of the Conventions,” Baltimore Sun, May 29, 1844; Proceedings of the Democratic National Convention held at Baltimore, June 1-5 1852 (Washington DC: Robert Armstrong, 1852); Official Proceedings of the National Democratic Convention held at New York, July 4-9, 1868 (Boston, MA: Rockwell & Rollins Printers, 1868); Official Proceedings of the Democratic National Convention held in Chicago, Ill., July 7-11, 1896 (Logansport, IN: Wilson, Humphreys & Co., 1896); Official Report of the Proceedings of the Democratic National Convention held in Baltimore, Maryland, June 25-July 2, 1912 (Chicago, IL: The Peterson Linotyping Co., 1912); Official Report of the Proceedings of the Democratic National Convention held in San Francisco, California, June 28-July 6, 1920 (Indianapolis, IN: Bookwalter-Ball Printing Company, 1920); Official Report of the Proceedings of the Democratic National Convention held in Madison Square Garden, New York City, June 24-July 9, 1924 (Indianapolis, IN: Bookwalter-Ball-Greathouse Printing Company, 1924).

How important was the South in blocking these failed nominees? To assess this, I identify whether Southern opposition was pivotal in denying the candidate the nomination. That is, if the candidate could have won the nomination with non-Southern support, Southern opposition was not by itself decisive in the outcome. For example, at the 1912 Democratic convention, Speaker of the House Champ Clark was the leading candidate on the first twenty-nine ballots. He came closest to winning the nomination on the tenth ballot, on which he received 556 votes. With 1,094 total votes available, Clark needed 729 votes to surpass the two-thirds majority required, meaning that, on his strongest ballot, Clark was 173 votes short. Clark received 41.5 votes (out of 252) from delegates from the former Confederacy and 110 votes from Southern border states (out of 126). This means there were 226.5 broad Southern delegate votes against his nomination on this ballot. However, from outside of the South, Clark lost 311.5 votes. This means there were enough non-Southern delegate votes available to achieve a two-thirds majority. While Southern opposition was part of the reason for Clark's defeat, it was not pivotal.

As shown in Table 4, by this metric, only two candidates lost their nomination due to Southern delegates refusing to support them: Van Buren in 1844 and Smith in 1924. As noted above, Van Buren was thirty-one votes short of winning the nomination with only twenty-seven non-Southern delegates opposing him. Even if Van Buren had won unanimous support from the non-Southern delegates, he would have failed to win a two-thirds majority and the nomination. The Smith case in 1924 is a little more complicated: on the 88th ballot Smith received 362 votes, slightly less than 33 percent of the total of 1,098 delegate votes, and quite far removed from winning the nomination. Nonetheless, a large part of Smith's failure to break through was due to near unanimous Southern opposition: Smith received just one vote from the Confederate South and zero from the border states. As a result, while non-Southern opposition was robust as well (359 votes in opposition), even if each non-Southern delegate had switched their vote, Smith would still have been eleven votes short of winning the nomination.⁵⁶

While these two cases are notable, the large majority of failed nominees were not the victim of a Southern veto. To be sure, most failed candidates (ten out of fourteen) saw a majority of delegates from the broad South oppose them, but six of those candidates (Buchanan, Douglas, Cass, and Marcy in 1852; Bland in 1896; McAdoo in 1920) also had a majority of non-Southern delegates oppose them. And, crucially, unlike with Van Buren and Smith, in a number of cases where a majority of Southern delegates opposed a candidate, they still received a non-negligible level of Southern support. Buchanan, Cass, and Marcy (all in 1852), Bland (1896), and Clark (1912) each received over 40 percent of Southern delegate votes, indicating considerable internal regional division and non-Southern opposition. Only four candidates who received majority support from non-Southern delegates failed to win the nomination. Of those, only two received a majority of all delegate votes: Van Buren in 1844 and Clark in 1912. And while Clark won only 16.5 percent of Confederate state delegates, he received a large majority of Southern border state votes (87.3 percent). On the other hand, four of the failed candidates (Cass in 1844; Pendleton and Hancock in 1868; McAdoo in 1924) received majority support of Southern delegates but lost their nomination because of non-Southern opposition, proving that the two-thirds rule cut both ways.

These data suggest there is little evidence that the South systematically used the two-thirds rule to veto undesirable candidates. By far, most nominees were always the leading candidate over however many ballots it took them to reach a two-thirds majority. For candidates that failed to win the nomination but were genuinely in the running, Southern opposition was rarely pivotal to their defeat. Indeed, in many cases, the candidates were quite far removed from even a simple majority, let alone a two-thirds one. In only two cases, did any of these candidates win a simple majority, but in the case of Clark in 1912 the broad support he received from Southern border states undermines the argument that his defeat was the product of a Southern veto. Across the entire century of two-thirds rule Democratic conventions, only one candidate received majority support and was still blocked from the nomination due to Southern opposition: Martin Van Buren in 1844.

7. Southern defense of the two-thirds rule

Of course, it remains possible that the South did not need to veto candidates because those known to be unacceptable to the region were never seriously considered. That is, if non-Southern delegates knew a candidate could not reach a two-thirds majority due to unified Southern opposition, they may have never nominated the candidate or chose to not vote for them if they were. But if this were a regular occurrence, we might also expect those non-Southerners to have attacked the rule. And we would likely expect a strong, unified defense of the rule from the South in response.

However, based on secondary literature assessments of the lifespan of the two-thirds rule (particularly Stockton and Bass⁵⁷), the two-thirds rule was readopted with little to no debate in the large majority of conventions: in 1840, 1852, 1856, 1864, 1872, 1880, 1888, 1892, 1896, 1900, 1904, 1908, 1912, 1916, and 1928, delegates either readopted the two-thirds rule with overwhelming support (and majority backing from both Southern and non-Southern delegates) or by acclamation. In some cases, the rule was simply applied without even having a vote: at the 1920 convention, the two-thirds rule was not explicitly part of the set of convention rules adopted, but the chair of the convention nonetheless applied it without challenges throughout the forty-four ballots necessary to nominate James Cox.⁵⁸

At the remaining conventions, debate about the two-thirds rule roughly fell into two categories: at some conventions, there was general consensus that a version of the two-thirds majority requirement would apply but disagreement existed about what counted as a two-thirds majority. Specifically, this concerned the question of whether a candidate needed to win two-thirds of all possible delegate votes or two-thirds of the votes cast. This was relevant particularly in cases where delegates boycotted a vote or bolted the convention entirely. In 1848 and 1860, the rule was (eventually) interpreted as applying to votes cast, resulting in the nominations of Cass (with support of the South) and Douglas (overriding a Southern boycott) without a two-thirds majority of the total delegate vote.

The more relevant debates concerned the question of whether the two-thirds rule should be ended entirely. Such attempts at challenging the rule directly were rare. In some cases, delegates called for a vote on the rule after balloting on the presidential nomination had already begun. In each of these cases, such challenges were ruled as out of order by the convention chair or dismissed by a majority of delegates. For example, in 1912, the two-thirds rule was readopted without debate at the outset of the convention. After the fifth ballot (on which Clark received nearly 41 percent of the vote), a delegate from Oklahoma proposed discontinuing the rule. The proposal was deemed out of order by the chair of the convention and the rule remained in place.⁵⁹ Similarly, during the 1924 convention, multiple attempts were made to reconsider the rule as the number of failed presidential nomination ballots began to add up: on the 11th day of the Convention (after seventy-three ballots), a delegate from Texas proposed dropping the lowest scoring candidate in each subsequent ballot until two candidates remained and that should neither candidate win a two-thirds majority on the ballot after, both the unit-rule and the two-thirds rule would be abrogated and a simple majority would suffice. After extensive debate, the proposal was voted down by voice vote.⁶⁰ Another attempt was made to end the rule after the 97th ballot—this time by a delegate from Oklahoma—but the resolution was again rejected by voice vote.⁶¹

In the limited number of cases where the rule was extensively debated, Southern states were often (though, not always) among the core defenders of the rule. Yet, that support was rarely unanimous within the region and drew support from many non-Southern states as well. The rule was most frequently challenged in the first years of its existence. In 1836, a majority of delegates in fact rejected the two-thirds majority. However, in response to the surprise defeat of the rule (231 delegate votes in favor to 210 against), a second vote was immediately called to reconsider, and the two-thirds rule was reinstated by voice vote.⁶² In the run-up to the 1844 convention, former president Martin Van Buren appeared to be the most likely candidate to win the nomination. However, in April, Secretary of State John C. Calhoun signed a treaty for the annexation of Texas. Many Democrats—including former president Andrew Jackson—publicly supported immediate annexation of the territory, but Van Buren argued against doing so. In response, Southern Democrats (including Jackson) “[commenced] a search for another candidate.”⁶³ At the convention in May, a long and rowdy debate ensued over whether to apply the two-thirds rule, which was decided with a 148 to 116 vote in favor.

As Lambert notes, strong support for the rule came from the South but—given that the decision to apply the rule required a simple majority and the broad South held only 42 percent of delegates—support from other regions was required. And the South was not even unanimous in its support for the rule: “sectionally, the South, including Missouri, Kentucky, Delaware and Maryland, cast ninety votes for the rule and fourteen against it.” As a result, “the key to the adoption of the rule was that it was voted for by over one-third of the delegates from the Northern states. […] Severe defections in Pennsylvania, Indiana, Illinois, and the New England states accounted for the margin of victory by which the rule was adopted.”⁶⁴ These non-Southern defections indicated the relatively weak support Van Buren may have had among a subset of delegates who did end up voting for him on the nomination ballots: since it was clear to the delegates present that applying the two-thirds rule would result in Van Buren losing the nomination, “the vote on the two-thirds rule is a far more valid test of Van Buren's real strength […] than that cast for him on the first ballot for candidates.”⁶⁵ Still, in 1844, nearly 86 percent of broad Southern delegates supported the rule while only 36 percent of non-Southern delegates did.

Four years later, in 1848, another debate on the rule occurred though this time “more moderate in expression.”⁶⁶ That convention eventually decided in a 176 to 78 vote to maintain the rule. As in 1844, support was strongest in the South: every single delegate from the future Confederate states supported the rule, and over 93 percent of border state delegates did as well. In contrast, a small majority of non-Southern delegates (52.6–47.4 percent) opposed the rule.⁶⁷ However, as in 1844, the non-South had a majority of delegate votes and could have ended the rule if it wanted to. In both 1844 and 1848, the rule survived because of strong Southern support but in combination with non-Southern backing.

With the exception of a brief debate in 1868, Democratic conventions consistently approved some version of a two-thirds rule between 1852 and 1872.⁶⁸ However, in the run-up to the 1876 convention, there was speculation in newspapers that the two-thirds rule could be abrogated. At the meeting of the Democratic National Committee (DNC) in February of that year, a DNC member from Nebraska offered a resolution recommending delegates to abolish the rule, though the resolution was not actually voted on by the committee.⁶⁹ Newspapers noted that different presidential hopefuls would push for the rule to be abolished, though San Francisco's Daily Evening Bulletin warned that “the Democracy is always a great stickler for ‘time honored custom.’ The two-thirds rule is old enough to be entitled to that designation.”⁷⁰ Notably, there were widely divergent theories of how the continuation of the rule would affect the different candidates. Several papers confidently predicted New York governor Samuel J. Tilden would only be able to win the nomination without the rule in place.⁷¹ But others predicted the two-thirds rule would instead end the presidential aspirations of Indiana governor Thomas Hendricks, Tilden's main competitor. And the New York Herald claimed the two-thirds rule would effectively end all possible candidacies as it would “not only kill off Mr. Tilden, [Ohio senator Allen G.] Thurman and [Delaware senator Thomas F.] Bayard, but also Mr. Hendricks and [former vice-presidential nominee George H.] Pendleton.”⁷²

These newspapers generally had a negative view of the two-thirds rule: the New York Times described it as “an embarrassing feature of the Democratic Convention,”⁷³ while the Herald called it “absurd.”⁷⁴ But, at least at this moment in time, they rarely connected the rule to Southern influence. One exception concerned an article published in the Richmond Dispatch, explaining that the survival of the rule “was formerly considered by Southern statesmen as ample guarantee that the interests of the South would be safe in their hands;” however, the “South has now no slaves to protect, and no property which its people cannot carry with them into any of the territories of the United States” and, as a result, the rule “has now no reason for being.”⁷⁵ Crucially, the Dispatch connected the need to abolish the rule to its preference of nominating Tilden: “Southern Democrats must go to St. Louis and, refusing to readopt the two-thirds rule, render the nomination of Tilden certain beyond a doubt.”⁷⁶ Despite this attention on the possible abolishment of the two-thirds rule, state conventions largely sidestepped the issue. When they did engage it—for example in Iowa and Illinois—motions to pledge state delegations to ending the rule were defeated or withdrawn before a vote.⁷⁷ However, some of the rare exceptions of state delegations supporting ending the rule came from the South: the Arkansas and Mississippi delegations were both pledged to vote for repealing the rule.⁷⁸

Despite all this build-up, at the convention itself, the debate was settled quickly: the committee on organization proposed applying the rules and regulations of the 1872 convention—including the two-thirds rule—and delegates accepted this proposal without debate.⁷⁹ However, after the presidential and vice-presidential nominations had been concluded a delegate from Michigan introduced a resolution describing the two-thirds rule as “unwise and unnecessary” and called on delegates to the convention in 1880 to determine “whether it be desirable to continue the two-thirds rule longer in force.”⁸⁰ In a subsequent vote on whether to table the resolution—that is, whether to remove it from formal consideration by the convention—a majority of delegates voted no and the resolution was subsequently adopted without further vote or debate.⁸¹ While a majority of Southern delegates voted in favor of tabling the resolution (see Table 5), the numbers were very close: ninety-six of the delegates representing the Confederate South voted in favor of tabling, but ninety-four voted against. Among the border states, support for the rule was stronger, but across the broad South only a small majority of delegates (53.3–46.7 percent) voted to protect the two-thirds rule. Crucially, the vote came after the nomination of Tilden was assured. Thus, Southern delegates who voted against tabling the resolution were not voting to try and help their preferred candidate win the nomination. Instead, delegations from Arkansas, Delaware, Louisiana, Maryland, Mississippi, South Carolina, Virginia, and a considerable subset of delegates from Missouri (seventeen Ayes and thirteen Nays) and Texas (two Ayes and fourteen Nays) voted to undermine a rule supposed to protect Southern interests.

Table 5. Southern and Non-Southern Voting on Tabling Anti-Two-Thirds Rule Resolution at the 1876 Democratic National Convention

Source: Official Proceedings of the National Democratic Convention held in St. Louis, Mo., June 27-29, 1876 (St. Louis, MO: Woodward, Tiernan & Hale, Printers and Binders, 1876), 167.

Notes: The votes by state listed in the proceedings incorrectly show Delaware casting twelve delegate votes, while the state only had six. The total number of votes reported in the proceedings is correct and counts Delaware as having just six votes. The numbers presented here reflect the corrected votes.

However, with Tilden barely losing the contested election of 1876, by 1880, the mood on the two-thirds rule had shifted back to acceptance. While there was some speculation in the year leading up to the convention that maintaining or rejecting the rule was part of a larger strategy to either keep Tilden from getting renominated or winning the nomination, by late spring 1880, there was general consensus at state conventions that the rule should be maintained. Indeed, Democratic state conventions across all regions (including many whose delegations had supported the 1876 instruction to end the rule) now pledged their delegates to support it, including Texas, Vermont, Connecticut, Pennsylvania, California, New Hampshire, Ohio, Minnesota, Virginia, Delaware, Missouri, Indiana, Georgia, and Illinois. In New York, where a split between Tilden and the Tammany Hall machine resulted in two competing state conventions, both voted to support the rule.⁸² The New York Herald concluded that the two-thirds rule “is evidently regarded as a greater preventive of acrimonious quarrelling and of log rolling and an important help to good work.”⁸³ At the convention, the rule was readopted without debate.⁸⁴

The last challenge to the two-thirds rule prior to its abrogation in 1936 was at the 1932 convention. While New York governor Franklin Delano Roosevelt was in a strong position throughout the months leading up to the convention, it was unlikely he would command a two-thirds majority of delegates on the first ballot, in part due to a series of primary losses in states like Massachusetts and California. As early as December 1931, there was speculation that FDR's campaign would attempt to end the rule at the convention. However, there was concern that attempting to do so would “meet with opposition which might react against the cause of the Governor.”⁸⁵ Indeed, the New York Times suggested a longer term plan to use the 1932 convention to convince the 1936 convention to drop the rule was more likely, as any strategy at abrogating the rule can “not be delayed until the meeting of the national convention, when changes are almost certain to be rejected because of the special advantage they might yield to some candidate.”⁸⁶

Despite this, several state party conventions—including in Wisconsin, Washington, and Georgia—instructed their delegates to oppose the two-thirds rule at the convention. Additionally, Democratic politicians such as William G. McAdoo, former DNC chair Homer S. Cummings, Arizona Senator Henry F. Ashurst, and Louisiana Senator Huey P. Long all publicly called for the rule to be abolished.⁸⁷ The actual attempt at ending the rule appears to have been mostly uncoordinated. At a gathering of FDR supporters in the days prior to the start of the convention, a number of state party leaders reportedly criticized the two-thirds rule. Notably, many of those calling for the end of the rule were Southerners: according to press reports, Long, South Carolina governor Ibra C. Blackwood, South Carolina Democratic Party chair Claude N. Sapp, Tennessee senator and former DNC chair Cordell Hull, and other politicians from North Carolina, Alabama, and Arkansas all pushed for the FDR campaign to try and abolish the rule.⁸⁸

Roosevelt himself was reportedly kept largely unaware of many of the specifics but the meeting activated his campaign to try and act on ending the rule at the 1932 convention. Initially, these efforts looked like they could pay off: internal estimates of delegate opinion from inside the FDR campaign published by the New York Times suggested that a majority of delegates (60.4 percent) were in support of ending the rule. Crucially, this included support from the South: according to these numbers, 61.3 percent of broad Southern delegates were in favor of ending the rule (70.6 percent of the Confederate Southerners and 42 percent of the border state delegates).⁸⁹ However, as predicted, news coverage of the campaign's plan to abolish the rule caused major uproar. Al Smith (FDR's main challenger) argued that “a new rule should not be adopted in the heat of political battle. The spirit of American fair play will not tolerate any eleventh-hour, unsportsmanlike attempt to change the rules after the game has been started.”⁹⁰ Similar criticism arose from former presidential nominees James M. Cox and John W. Davis and from other 1932 candidates Newton D. Baker and Maryland governor Albert Ritchie.⁹¹ Meanwhile, Mississippi senator John Sharp Williams argued it would be “idiotic”⁹² for the South to turn against the rule. The backlash quickly complicated matters: just 2 days after the FDR campaign's internal polling suggested a comfortable majority on board with ending the rule, new delegate polling showed this majority had all but dissipated and support for ending the rule was now down to just 50.6 percent of delegates, with broad Southern support at 49.5 percent.⁹³

With passage of an abrogation of the two-thirds rule now doubtful, FDR's allies on the rules committee decided to pass a compromise of sorts: in the new plan, the two-thirds rule applied but only in the first six ballots. However, if no candidate won the nomination after the sixth ballot, the seventh (and any necessary subsequent ballots) would require just a simple majority. On the convention's rules committee, Southern delegates representing Arkansas, Delaware, Florida, Georgia, Louisiana, Kentucky, Mississippi, Tennessee, and West Virginia supported this rule change, while those representing Alabama, Missouri, Maryland, North Carolina, Texas, and Virginia were opposed.⁹⁴ But while a majority on the rules committee voted in favor of the change, the Roosevelt campaign at the last minute decided to drop the effort: FDR himself was reportedly (once again) not involved in the decision to attempt the compromise effort and later ordered his campaign manager James Farley to call off the attempt in a telegram. Subsequently, Farley burst into a meeting of rules committee members who had voted against the compromise to inform them that the campaign would drop its efforts after all and accept the original rule. While the committee still recommended reconsidering the two-thirds rule in 1936, there was now consensus the rule would apply in 1932.⁹⁵

After the uncoordinated failure of ending the rule in 1932, Roosevelt (now safely in the White House and preparing his first renomination) and Farley began planning their next attempt well in advance. In 1935, Farley began the “working up of sentiment for revocation of the two-thirds convention rule”⁹⁶ and at a January 1936 DNC meeting, he ensured the call for the convention included the request for state delegations to take positions on whether to maintain the rule.⁹⁷ Given that FDR's renomination was effectively guaranteed, in 1936, there would be no problem of being branded a spoiler by changing the rules “during the game.” Yet, the convention would still need to vote in favor of abrogating the rule. Notably, unlike in 1876 and 1932, in 1936, the debate on the rule was framed in newspapers as a clash between “leaders in Northern states”⁹⁸—who favored ending the rule—and Southern party leaders who favored saving it. Yet, as Bass notes “such reports rarely specified individuals or actions resisting the reformers.”⁹⁹ Indeed, while a small number of Southerners—like Virginia Senator Harry F. Byrd—defended the rule, the Los Angeles Times reported that “there is nothing like unanimity” among “Bourbon bigwigs”¹⁰⁰ about how to respond to the two-thirds rule challenge and the Washington Post noted that “other Southerners have remained silent.”¹⁰¹ Meanwhile, some Southerners such as Senator Josiah W. Bailey and Rep. Robert W. Doughton (both NC) and Senate Majority Leader Joseph T. Robinson (AR) indicated they backed ending the rule.¹⁰²

One of the strongest supporters of this goal was border state senator Bennett C. Clark (MO). The son of Champ Clark, Bennett Clark told reporters that, while he “could have had the chairmanship of the national convention,” he chose to chair the rules committee “because it will give me an opportunity to help defeat the two-thirds majority rule”—which he described as “the undemocratic rule which deprived my father of the Presidency.”¹⁰³ As Senator James F. Byrnes (SC)—a supporter of maintaining the rule—commented, Clark “pursued his goal with all the energy of an avenging fury.”¹⁰⁴ By May 1936, Clark reported that a large majority of delegates supported ending the rule.¹⁰⁵ Still, as delegates began to gather in Philadelphia in June, Farley and other convention organizers remained cautious. In fact, the convention schedule was extended to allow for a possibly time-consuming floor fight on the rule.¹⁰⁶ Newspapers noted opposition was mostly centered in Virginia and Texas with additional support from “a sprinkling of delegates from other states.”¹⁰⁷

On the first day of the hearings held by the rules committee—at which states were represented by one delegate each—opposition to changing the rules came from Southern delegates from Georgia, Alabama, Texas, and South Carolina.¹⁰⁸ On the second day, a compromise of sorts was reached: the two-thirds rule would be ended but the convention would also recommend a reconsideration of the allocation of delegates to states for future conventions. In response, a large majority (thirty-six to thirteen) of committee members supported ending the two-thirds rule. Notably, a majority of confederate Southern states opposed the change: Alabama, Georgia, Mississippi, South Carolina, Tennessee, Texas, and Virginia all voted against ending the rule. But a majority of the broad Southern states supported abrogation: Arkansas, Louisiana, North Carolina, and every single Southern border state backed Farley, Clark, and FDR. Meanwhile, among the minority opposing the change were northern (New York, Massachusetts), western (Colorado, New Mexico), and midwestern (Indiana) states as well.¹⁰⁹

This discussion of challenges against the rule spotlights two main findings. First, the rule was rarely disputed: in the large majority of conventions, no meaningful debate was held on whether to apply the rule or not. In the cases where the rule was at some point debated, it generally happened after balloting had already begun—at which point the rule was already in place and the convention consistently supported maintaining it. Second, Southern support for the rule was inconsistent. In the 1840s, large majorities of Southern delegates voted in favor of the rule, but survival of the rule required considerable non-Southern support, which it received. And in 1876 and 1932, Southerners were among those opposing the rule. Meanwhile, in 1936, while much of the media coverage focused on Southern opposition to ending the rule, a majority of broad Southern states supported the effort in the rules committee. While there certainly were Southern leaders who interpreted the two-thirds rule to be a protectionary device for their region's interest, this sentiment was hardly universal across time or Southern states. And, while non-Southern delegates often voted against the rule when it was brought up, many supported the rule—either by never challenging it in the first place or voting in its favor when its survival was brought up for a vote.

8. Conclusion

This article argues that the interpretation of the two-thirds rule as a Southern veto is (at best) much more complicated than the term would imply. Southern states—broadly defined—may have had enough votes at each Democratic convention between 1832 and 1932 to block presidential nominations, but they could only do so as a region if they voted together against a candidate at an extremely high rate. From 1876 onward, the broad South required a level of consensus against a candidate of nearly 90 percent or more to block a nomination without support from non-Southern delegates. By 1932, that number had risen to 99 percent—meaning that even the smallest level of dissent among delegates of the seventeen states and the District of Columbia that combined can be considered the broad South would undermine the region's ability to veto a nomination. To be sure, a subset of delegates from the South could still block nominations if they collaborated with non-Southern delegates. And a large bloc of Southern opposition votes could effectively doom a candidacy if it convinced other delegates that a path toward a two-thirds majority would be extremely difficult. But the issue remains that if a candidate was vetoed by a broad cross-regional coalition of delegates it would be hard to argue that their candidacies were ended by the South specifically.

Perhaps unsurprisingly then, the number of cases in which the South vetoed a candidate is extremely small. Throughout the period in which the two-thirds rule was in place, the large majority of candidates who won the most votes on any single ballot also eventually won the party's nomination. While some candidates—such as Smith in 1928—inspired robust opposition from the South, even then it proved difficult for Southern delegates to unite at a consistent enough level to block the nomination. Indeed, while Smith had many defects from the perspective of the South, he still received 41 percent of broad Southern delegate support, many times more than the mere 2.4 percent of Southern delegate support “allowable” for a pure Southern veto to prevent his nomination. The sole example of a candidate who received majority support from delegates but was unable to win the nomination due to Southern opposition was Van Buren in 1844.

This is not to say that Southern delegates played no role in preventing candidates from winning a Democratic nomination in this period. Out of (only) fourteen candidates who at some point led a ballot but failed to win the nomination, ten saw a majority of Southern delegates oppose their candidacy on their strongest ballot. But of those candidates (with the exception of Van Buren) every single candidate also saw at least 40 percent of non-Southern delegates vote against them. While it remains possible that some part of this opposition reflected strategic decisions to not support a candidate non-Southern delegates understood were highly unlikely to ever reach the two-thirds majority needed, it is also quite possible that a good number of the failed candidates simply lacked party-wide support. Regardless, out of every convention year, 1844 remains the only clear example of a “true” Southern veto of a Democratic presidential nomination.

These findings leave a variety of questions related to the two-thirds rule unanswered and raise several new ones as well. The most obvious is why the two-thirds rule did survive if not because of its role as a Southern veto? One part of an alternative explanation for the rule's longevity is that the negative effects attributed to it were limited throughout most of its existence.¹¹⁰ This is particularly true in terms of the complexity of nominating a Democratic presidential candidate. Figure 3 shows the number of ballots at every Democratic, Whig, and Republican convention between 1832 and 1932. Prior to the Civil War, the Whigs and Democrats had largely comparable numbers—despite the fact that the Whigs required only simple majorities for their presidential nominations. For example, in 1852, Democrats took forty-nine ballots to nominate Franklin Pierce. But that same year, the Whigs went through fifty-three rounds of voting before they were able to nominate a candidate. Additionally, the extraordinarily long nomination fights that we associate with the two-thirds rule largely were the product of the 1910s and 1920s: between 1856 and 1908, Democratic conventions needed, on average, eight rounds of voting to nominate a candidate, while Republicans needed five. Between 1912 and 1932, however, the Democratic average was thirty-three ballots while the Republican was just three. This difference is produced practically entirely by the conventions of 1912 (46 ballots for the Democrats), 1920 (44 ballots), and 1924 (103 ballots). While the question of ending the two-thirds rule, therefore, became much more relevant in the 1920s and 1930s, the topic was likely less relevant in prior decades. To many delegations, prior to 1912, the two-thirds rule (having become one of the “time honored customs” of the Democratic Party) may simply not have been worth fighting over in comparison to other, more salient issues on the table.

Source: CQ Press Guide to U.S. Elections (Washington, D.C.: CQ Press, 2010, 6th edition).

Figure 3. Ballots Necessary to Complete Presidential Nominations at the Democratic, Whig, and Republican National Conventions, 1832–1932.

A related consideration is that, as the failed efforts at ending the rule in 1876 and 1932 show, building consensus to end the rule (particularly as its application became an implied part of the conventions’ design) required considerable cross-state organization and coordination before and during the convention. The collective action problem this produced was exacerbated by the fact that, in most convention years, attempts to end the rule were linked to the fortunes of specific presidential hopefuls. As a result, efforts to put the two-thirds rule on the convention agenda were not seen independently of the presidential contest at hand. In proposing an end to the rule, proponents of a presidential contender thus ran the risk of opening themselves up to accusations of foul play (as FDR's experience in 1932 shows) and/or signal weakness to party leaders by suggesting that their candidate would be incapable of building a broad enough majority to win the nomination if the rule remained in place. Strategically then, avoiding debate about the two-thirds rule may have been a wiser move in many convention years. In this regard, the 1936 convention was the perfect storm for ending the rule, as it came at a time when the negative consequences of the two-thirds rule were more evident given the many lengthy ballot fights of the 1910s and ‘20s, with an incumbent president whose renomination was unchallenged, whose authority in his party was considerably stronger than his predecessors had been, and who was willing to put political capital into ending the rule.¹¹¹

Finally, while imperfectly designed,¹¹² the two-thirds rule may have served a real function within the party. As Jaenicke has argued, the Democratic Party from its founding represented “an umbrella organization that embraced a plethora of groups with conflicting ultimate purposes.” That is, many of the groups that made up the Democratic Party never truly agreed on fundamental issues. For the party to function, members needed to respect their “shared negative desire to be left alone in their separate pursuits and to avoid any debate concerning their actual ways of life.”¹¹³ Nominating a presidential candidate at a convention represented only a part of the ongoing puzzle of maintaining the Democratic Party as a political organization: even if a small majority in the party was able to force a nomination on a minority, this candidate would still need the support of the opposition in a general election. This was particularly true for Southern electoral support: while the South alone could not elect Democrats to the White House, a Democratic nominee without Southern support would need to radically reinvent the electoral map to have any chance at all at winning. While the election of 1932 undermined the power of the South within the party—providing further space to 1936 as the year in which the rule could be ended—prior to the New Deal realignment the South's support was practically nonnegotiable.

By presenting the two-thirds rule as a Southern veto, it (perhaps inadvertently) suggests that the South—unless it could use the rule as a “fail-safe”—faced the very real risk that non-Southern delegates would nominate candidates hostile to its pro-slavery and segregationist interests. But actual challenges to these Southern policy preferences appear to have been limited for most of this period. Crucially, non-Southern delegates could always have passed platforms that included policies hostile to Southern interests with simple majorities, yet nearly always refrained from doing so. What exact considerations pushed non-Southern party leaders to not confront the South at Democratic conventions—the South's dominant position in the Democratic Congressional caucus? Its contribution to the party's electoral vote? A basic lack of non-Southern Democratic interest in racial liberalism prior to the incorporation of Black voters into the party?¹¹⁴—is worth further and more systematic investigation, but the risk inherent in nominating a candidate against the wishes of the South was not purely theoretical. Indeed, in the cases of Stephen Douglas (in 1860) and Al Smith (in 1928), the nomination of a candidate against the wishes of the South resulted in electoral punishment: in 1860, Southern Democrats bolted the party and ran their own candidate, and in 1928 many Southern Democrats voted for Republican nominee Herbert Hoover.¹¹⁵

Combined, what this meant was that throughout this period, Democratic leaders faced the political reality that broad intraparty consensus was a necessity independent of whether a two-thirds majority was required for presidential nominations or not. The long survival of the two-thirds rule, then, may have been less an attempt at providing the South with a weapon it needed to control the party, but rather a de-facto public acknowledgement of what was obvious to all participants involved: that any nominee would need to be acceptable to the South, as well as other relevant constituencies within the party. In this regard, the value of the continued use of the two-thirds rule was perhaps similar to Aldrich's assessment of the custom of running regionally balanced party tickets in the pre–Civil War era:

Balanced party tickets were better understood as symbolic, public affirmations of the party's continuing commitment to maintain its intersectional alliance for electoral purposes than as actual assurances that a southerner would be able to cast the tie-breaking vote in the Senate or to veto legislation in the White House. The genuine commitment was to maintain the alliance to ensure that neither strongly antislavery nor diehard proslavery leaders would be chosen.¹¹⁶

The two-thirds rule's survival then was less about producing a political institution to allow a group within the party to have a direct veto over presidential selection and more about maintaining a symbol of the underlying political agreements that bound the party together.