Slavery or Sheep?
The Antebellum Realignment in Vermont 1840-1860
Christopher H. Achen
Department of Politics
Prepared for presentation at the Midwestern Political Science Association Annual Meeting, Chicago, IL, April, 2012. An earlier version was given at the Center for the Study of Democratic Politics weekly seminar, Princeton University, May 26, 2011. Copyright by the author.
Political scientists and historians have long debated the causes of the dramatic partisan shifts occurring in realignments. Antebellum Vermont provides an excellent test case. In the 1850s, Vermont become the most Republican state in the Union, a position it held for a century. Vermonters were anti-slavery, and in the conventional view, this honorable stance led them to the GOP. Yet in this period, Vermonters were also anti-Catholic, anti-liquor, and above all, pro-tariff on behalf of their enormous flocks of sheep. These issue positions, too, were shared by the Republicans. Using population data, agricultural censuses, referendum results, and election returns from Vermont, this paper dismisses temperance and anti-Catholicism as important influences on the Vermont vote. It then carries out a test of whether the pre-Civil War realignment in Vermont was derived from abolitionist sentiment or wool prices. It turns out that neither economics nor moral views alone sufficed. Instead, it was their dynamic interaction, with economic interest making anti-slavery sentiment an effective force at the polls, that led to the Republican realignment.
Antebellum Political Historiography1
The rise of the Republican Party in the 1850s had immediate and profound consequences for American politics. “The surge of Republican power,” culminating in the election of Lincoln in 1860, was the proximate cause of the American Civil War. (The phrase is Silbey’s 1985, chap. 9.) Thus disentangling the Republican voting coalition is central to understanding why the war came.
Antebellum historians have identified several groups as constituting the initial GOP electorate (for example, Gienapp 1987, chap. 11). The first and best-remembered are the foes of slavery, who came to supply the popular understanding of the terrible war that followed. But anti-immigrant and anti-Catholic sentiment also mattered in many places. Indeed, the American Party (“Know-Nothings”), whose central interest was anti-Catholicism, bid fair in many states to outcompete the Republicans as the Whigs’ replacement. Temperance advocates, too, often became prominent GOP spokesmen. Lastly, North-South divisions over economic policy, including the tariff, internal improvements and other disputes generated by the divisions between free and slave labor, helped fuel the rise of the purely sectional Republican party.
By emphasizing one or another aspect of the GOP coalition, historians arrive at different causes of that cataclysm. In the decades after the war, prominent historians focused on slavery: “The meaning of the  election was that the great and powerful North declared slavery an evil” (Rhodes 1906, 502). In the interwar period, Progressive historians like the Beards (1930) argued that the differences in Northern and Southern economic systems, exacerbated by tariff battles and financial panics, were more central. Others argued that the war was avoidable, brought on by sectional fanaticism and blundering politicians (Randall 1940; Craven 1942 and 1953). Postwar revisionists pointed to local ethno-cultural divisions, especially along religious lines, as providing a more powerful explanation of antebellum voting, with the rhetoric about slavery and tariffs less important (for example, Holt 1969). Historical research suggests that no one answer fits every state, and that somehow all these causes must be consequential. However, especially since the civil rights movement of the 1960s, the historical consensus has moved back to the view that attitudes toward slavery were the central aspect of the Republican appeal (for example, Sewell 1976, chap. 14). Thus McPherson’s (1988, 158; chap. 7) widely respected study of the Civil War period says of the 1856 presidential campaign that “the salient issues were slavery, race, and above all Union”; and his discussion of the 1860 campaign is similar. The best known political science study of party realignments takes the same view (Sundquist 1983, chaps. 4, 5).
Most historians of the antebellum era rely on the qualitative analysis of documents—party platforms, speeches, newspaper editorials, and private letters. Their sure-handedness at sorting reliable witnesses from the forgetful, partisan, or dishonest, and then entering the minds of people whose conceptual world may be quite different from that of the current era, results in a depth of sympathetic understanding unattainable with other methodological tools. This approach is the foundation of most historiography on antebellum elections and the primary source of the current near-consensus about why the Republicans won.2
Yet the classic historical method has important limitations when election outcomes are central to the argument. The documents tell us primarily what political elites thought. But between activists and ordinary citizens often falls a chasm. The last half century of survey research has demonstrated unequivocally what thoughtful observers like Lippmann (1922) had spelled out earlier, namely that what appears in party platforms, newspaper editorials, and the letters of politicians and political operatives often has little or no resonance in the minds of the voters, even in times of crisis (Converse 1964; Delli Carpini and Keeter 1996). The citizenry may adopt the high-flown rhetoric as their own, but their motives for party choice are often quite different. And if that is true of the relatively well-educated voters of the twenty-first century, then it is at least a plausible hypothesis for the isolated, mostly rural, and often only partially literate people of the mid-nineteenth century.
Altschuler and Blumin (2000, chap. 4) show that a selection of antebellum men’s diaries exhibit little engagement with politics. One third never mention it at all (p. 143), and others report voting without mentioning for whom or what the issues were, even when they go on at length about other topics closer to home. As Michael Holt 1992, 319) has argued, when a distinguished historian writes, “Congress would become for fifteen years the arena of a continuous battle watched by millions of aroused sectional partisans” (Potter 1976, 49), he is making a strong and doubtful claim. Bassett (1952, 19, cited in Bigelow 1970, 74) more plausibly concludes from his study of the antebellum Vermont farmer:
His communication was as simple and limited as his travel….Perhaps one out of three managed to pay for a weekly local newspaper….Yet farmers did not take much time to read and were suspicious of indirect secondhand booklearning.
Even Gilmore’s (1989, 126) detailed study of Vermonters’ reading in this period, which takes a more optimistic view, nevertheless notes that the reading that did occur consisted mostly of
the family’s yearly almanac, some of the contents of a rural [weekly] newspaper, and other basic books and pamphlets, especially the Bible, hymnals, songbooks, devotional tracts, schoolbooks, elementary geographies, histories, some light fiction, poetry, short essays, and popular tracts of various sorts.
Hence to resist the dangers of a misguided “Civil War synthesis” based solely on the thinking of elites, we need not just the close analysis of documents, but detailed study of disaggregated election returns as well, as the “new political historians” have long emphasized (Benson 1957, 1967-1968; Silbey 1964). Yet even now, it is striking how small a role the modern study of election returns has played in historical explanations of the antebellum period. For example, Sewell’s (1976) impressive study of anti-slavery parties, entitled Ballots for Freedom, makes only passing mention of actual ballots, and it undertakes no detailed analysis of them. Even for 1860, perhaps the most consequential election the country has ever faced, the literature is thin, and many states have never been studied at all. If one asks, “Who voted for Lincoln?” or “Why?” the only defensible answer for most places is that we have only one or two initial studies, and more often no studies at all.3
In an attempt at partial remediation, this paper takes up the electoral analysis of one small state. Vermont has been so little studied that it is usually given only glancing mention, or even omitted entirely, in antebellum political historiography. Gienapp’s (1987) magisterial, multi-state textual and statistical analysis of the rise of the GOP mentions Vermont on just four pages, and very briefly in each case. Yet the state was unsurpassed in the speed and enthusiasm of its conversion to the Republicans in the 1850s. Moreover, as I hope to show, the state had certain unusual social and economic characteristics that illuminate how ideological and economic factors came together to generate the dramatic Republican realignment. Thus this paper attempts to speak to antebellum historiography, but also to the literatures on realignments and on the development of partisanship. To do so, it proposes a unified conceptual framework for thinking about the electoral forces that brought on the war.
Politics in Antebellum Vermont
The Vermont Republican Party came into being at a Montpelier convention in the summer of 1854 (Hand 2002, 6-7). That fall, Republicans captured the governorship with 63% of the vote. Two years later, the state voted Republican for president by a 78-21 margin, the most one-sided Republican victory in any state. From then until the gubernatorial election of 1962, Vermont never voted for a single Democrat for either office, a span of more than a century. Famously, Vermont and Maine were the only two states to vote for Alf Landon in 1936, but lopsided GOP victories in Vermont extended throughout the period. For example, in 1896, McKinley carried Vermont 80-17, his largest margin. In 1924, Calvin Coolidge took the state 78-16, also his largest margin. And in 1956, Eisenhower won Vermont 72-28, his largest margin as well. Vermont Democrats were strictly an afterthought until Philip Hoff squeaked into the governor’s chair with 50.5% of the vote in 1962.
Nineteenth century Vermont is an attractive laboratory for studying the rise of a new partisanship. The Republican conversion was rapid, widespread, and long lasting, so that trivial aspects of contemporary politics can be set aside as explanations. For this was not a state where longstanding partisan loyalties slowed the absorption of a new alignment. The Antimasonic Party carried Vermont in the presidential election of 1832, the only state they won. The Antimasons then won the governorship through 1835, when they were absorbed into the Whigs. That party then dominated Vermont politics at the national and state level for the next decade, although gubernatorial elections were usually very competitive. After that, the successive Liberty and Free Soil insurgencies took votes from both Whigs and Democrats, routinely leaving the major parties below 50% in gubernatorial elections and throwing the decision into the legislature (McCormick 1966, 69-76; Carter 1989, 240-246).4
By the 1856 election, when the Republicans first entered the presidential lists, the Whig and Free Soil parties that had captured 70% of the Vermont presidential electorate just four years earlier no longer existed. Even the ongoing (Jacksonian) Democrats were creatures of less than 30 years’ standing, while the Antimasonic, Whig, Liberty, and Free Soil parties had been born, lived, and died in even less time. The rapid turnover of parties and the ease with which new parties attained substantial vote shares suggest strongly that, for most Vermonters, partisanship had only shallow roots before the rise of the Republicans. Turnout rose rapidly in this period after the 1830s, too, so that many voters were relatively new to politics in the 1840s and 1850s.5 Certainly, generation-to-generation party loyalties, which provided a powerful brake on later realignments in the U.S., did not exist in 1850s Vermont. The Republicans could, if not quite write on a blank slate, at least get an immediate hearing for their party’s point of view.
What caused the abrupt conversion in the North to Republican hegemony? One frequently cited cause was anti-Catholicism, which was rampant in this period (Billington 1938), and which had the clear effect of producing GOP votes in some locales (Holt 1969; Silbey 1985, chap. 9). Heavily Protestant Vermont was undoubtedly no more enthusiastic about Catholic doctrine than other states, but the issue was much less relevant there than elsewhere. Vermont had only modest economic opportunities and thus few immigrants, Catholic or otherwise; indeed, several counties and many towns in this period were losing population (Arnold 1980). Catholics from Quebec and Ireland were present, but only in a handful of towns and in small numbers. Often they had no priests (Feeney 2006).
The result was that the Know-Nothings ran only once for governor, in 1855, receiving just 8% of the vote. Their areas of relative strength bore no relationship to Free Soil or Republican sentiment, or even to the presence of Catholics. In Addison county, the most Republican in the state, the Know-Nothings took 12 votes out of nearly 3000. 6 In Lamoille county, the principal hotbed of Liberty Party and Free Soil sentiment, the Know-Nothings received 37 votes among more than 1600. In Chittendon county, home of the major city, Burlington, and the place with the most Catholics in Vermont (Feeney 2006, 108-113), the Know-Nothings received a grand total of 4 votes among more than 3000. Fillmore’s American Party candidacy for the presidency in 1856 was an even bigger failure: He attracted 21.5% nationwide, but just 1% in Vermont. In the near-absence of Catholics, anti-Catholicism was too abstract an issue to become politically potent in the Green Mountain State.7
Similar remarks apply to the impact of temperance on the rise of Vermont Republicanism. Widespread drunkenness made temperance a powerful issue in the U.S. during this period (Rorabaugh 1979; Tyrrell 1979). Vermont was no exception. Like abolitionist parties before them, Republicans were anxious to capture the votes of temperance advocates. Undoubtedly, some moralizing voters saw temperance, anti-slavery, and Republican loyalty as all of a piece. However, it turns out that overall, there is no relationship between temperance votes and support for the Liberty, Free Soil, or Republican parties.
Vermont voted five times on temperance between 1845 and 1853, generally at town meetings. Only the last of these decisions, a complete ban on consumption and production of alcohol, was put to a conventional statewide referendum. It drew the largest turnout, and thus is the most representative measure of Vermont opinion. It passed by a very narrow margin.
Figure 1 shows the county-level plot of the Republican presidential vote in 1860 against 1853 temperance support. The absence of a relationship is apparent (r = -.06).8 The plots for the 1844 Liberty Party vote, the 1848 and 1852 Free Soil votes, and the 1856 GOP vote are similar. The corresponding correlations are small and erratic: -.16, .18, .09, and .20, respectively. As usual, with more disaggregated data, the relationships are even weaker. Bigelow (1970, 84) studied the earlier temperance votes and their relationship to Liberty Party votes, using town-level returns. The correlation coefficients were always near zero, and never more than .02.9
Thus whatever the relationship between temperance and abolitionism among political elites, the evidence is strong that the ordinary Vermonters who supported temperance were no more likely to vote for anti-slavery parties than their tippling neighbors. Whatever forces drove the temperance movement, they bore little relationship to abolition sentiment or to GOP voting among Vermonters as a whole.10 Indeed, after being burned by the issue initially, Republican leaders nationwide were generally anxious to soft-pedal temperance in favor of a focus on their central issue, anti-slavery, in hopes of attracting both drinkers and abstainers to their cause (Gienapp 1987, 190, 206-208, 227-228).
Thus in assessing the causes of the 1850s Republican surge, Vermont presents a simpler inferential problem than most states. Temperance and anti-Catholicism can be set aside as major forces. That leaves to be considered the two powerful factors that dominate most treatments of Civil War causation—slavery and economics.
Anti-Slavery and Tariffs in Early Antebellum Vermont
In Vermont memory, the cause of its fondness for the Republican party is simply a deep aversion to slavery. In the main hall of the capitol in Montpelier, engraved on granite, one may read this inscription, taken from a Vermont Senate Report of 1855:
Born of a resistance to Arbitrary Power—her first voice a declaration of equal rights of man—how could her people be other than haters of slavery—how can they do less than sympathize with every human being and every community which asserts the rights of all men to blessings like their own?
Condemnations of slavery and the South from religious leaders and other activists were a commonplace of Vermont life, beginning in the 1830s (Ludlum 1939, chap. 5). Indeed, so flagrant and frequent were Vermont denunciations that the Georgia state senate reportedly passed the following motion at the time of Kansas-Nebraska agitation (Graffagnino 1977, 31):
Resolved, That His Excellency, President Pierce, be requested to employ a sufficient number of able-bodied Irishmen to proceed to the State of Vermont, and to dig a ditch around the limits of the same, and to float “the thing” into the Atlantic.
No doubt overcome by the emotions of the moment, the senators failed to explain how “the thing” was to float across New Hampshire and Maine.11
General histories of Vermont invariably recount the earliest-in-the-nation prohibition of slavery in the first state constitution in 1777, the abolition societies, the enfranchisement of African-Americans, the anti-slavery petitions to Congress (Carpenter 2011), the prominence of their elected representatives in anti-slavery battles in the nation’s capital (Miller 1995, chap. 6 and passim), and the comparatively large vote for the Liberty and Free Soil parties as evidence that Vermonters focused their antebellum politics around opposition to slavery and a love of liberty for all. “Vermont enjoyed the reputation of being the most antislavery state in the nation” (Johnson 1979, 258). Attachment to the Republican Party then seems to have followed in the natural course.
The difficulty with this argument is that northern whites were comfortably racist in this period (Litwack 1961, Woodward 1962), and that Vermonters were no exception (Graffagnino 1977). Abolitionist speakers were often shouted down in 1830s Vermont, for example, and after very rough treatment during an 1843 speaking tour, Frederick Douglas remarked that Vermont was “surprisingly under the influence of the slave power” (Lovejoy 2001, 61). Much of antebellum Vermonters’ sympathy for blacks “was paternal or designed to make a point, and it was far overshadowed by the malicious attentions of other white Vermonters” (Roth 1987, 272-273). As late as the Civil War, Vermont soldiers’ letters home were peppered with negative, even vicious remarks about African-Americans, often the first they had ever seen (for examples, see Marshall 1999, 5, 142; of course, some soldiers had the opposite view--see Manning 2007). In this respect, Vermont was no different from other states that were soon to become hotbeds of Republican sentiment. Both Wisconsin in 1847 and Michigan in 1850 voted down enfranchisement of blacks by 65% or more (Smith 1897, 332-335).
Vermonters had lived with Southern slavery from the time of their entry into the Union, and without worrying much about it. The Liberty Party received just 319 Vermont votes in the 1840 presidential election. During the next dozen years of constant agitation, neither Liberty nor Free Soil ever came close to winning a statewide Vermont election.
Like other northerners, what Vermonters opposed was not slavery, but the Slave Power (Davis 1969; Gara 1969; Richards 2000). Antipathy toward the South had early roots in Vermont. Southerners opposed Vermont statehood in the late eighteenth century, since it would add two additional anti-slavery senators (Richards 2000, 46). Vermont sent unofficial delegates and observers to the Hartford Convention during the War of 1812, a convocation called to consider New England secession as a response to an economically inconvenient war conducted by a Virginian president, “Mr. Madison’s war” (Horsman 1969, 211-214; Sherman et al. 2004, 163-164).
Thus what gave anti-Southern outbursts resonance was not the depth of positive feelings toward the slaves themselves, but rather the impact, or imagined impact, of the South on the economic fortunes of Vermonters. Initially, the focus was the tariff, which was, apart from slavery, the central issue in American politics throughout most of the 19th century. The federal government drew the bulk of its revenues from the tariff, so that debates about the level of duties were also a debate about the size of government and its role in the domestic economy. When times were bad, out-parties blamed the state of the economy on the in-party’s tariff policies (Taussig 1889; Stanwood 1903; Hofstadter 1938; Bolt 2009).
In Vermont, tariffs were closely related to the sheep-growing industry. George Washington at one point owned 800 sheep, and he once wrote in a letter that he had “no doubt as to the good policy of increasing the number of sheep in every State” (Carman 1892, 54, 57). Vermont took him at his word. Beginning in the late 1820s, a “sheep mania” overtook the state. By 1836, Vermont had one million sheep; by 1840, the state contained just under 1.7 million head, nearly six per capita. Addison County had eleven sheep for every man, woman, and child.12 Other New England states joined the sheep-raising industry, but no other state had the same density of the wooly creatures. (For overviews of the wool industry in this period, see Wright 1910; Cole 1926; Wilson 1935; Wilson 1936). Vermont’s governor remarked in his annual message of 1842, “Our citizens have become so dependent upon the growing of wool that this article may be said to be the staple of the state” (Wilson 1936, 81).
Unfortunately for Vermonters, competition from British woolens was stiff. Thus Vermonters all over the state had a special reason to care about the tariff on wool. This uniformity of economic interest makes it easier to assess economic effects on politics in Vermont than in virtually any other states in this period.
Alexander Hamilton as Secretary of the Treasury had reported to Congress as early as 1791 to encourage a tariff on imported wool and other products to help support American industry (Carman 1892, 119-120). Tariff agitation began in earnest in the mid-1820s, resulting in the Tariff of 1828--the “Tariff of Abominations” to its opponents (Wright 1910, 47-51). It was modified slightly by the Tariff of 1832.
The South was an exporting area, and it had no interest in paying duties on imports, which it regarded as a subsidy to the North and to New England in particular.13 In addition, when American tariffs were high, the British retaliated by buying less cotton, further damaging the South. Feelings ran high. Newspapers and books were filled with dramatic essays condemning the tariff and its Northern supporters (notably, Turnbull 1827) . It is sometimes forgotten that John C. Calhoun’s development of nullification theory was directed at the 1828 tariff. His arguments on behalf of nullification mention only the tariff, not slavery. Tensions rose high enough that in November of 1832, South Carolina’s special Nullification Convention declared the Tariffs of 1832 and 1828 unconstitutional. South Carolina came close to mobilizing troops over the issue and seceding (Freehling 1965). Nor were Northern interests in the tariff merely perfunctory: “The debates over the tariff were not trivial or subsidiary to northerners, but went to the heart of northern concerns over the nation’s future” (Grant 2000, 44).
The tariff also funded “internal improvements” such as roads and canals. The South had little need of either, while the Northeast needed tariffs and the “Northwest” (that is, what is now the upper Midwest) needed internal improvements. The parties divided. Whigs favored tariffs and improvements, while Democrats opposed. In the early antebellum period, party lines held and were usually strong enough to overcome sectional forces (Silbey 1985, chap. 3). But with time, the sectional division over economic policy came to line up with attitudes toward slavery, with fateful consequences.
Vermont knew where its economic interests lay. Its newspapers in this period are filled with intense discussions of tariff questions. By 1832, early in the sheep mania, the correlation of county Democratic vote with sheep per capita was already -.4. The correlation grew to -.6 by 1836 and 1840. (The Democratic vote is used because of the third-party vote in this period, and because the Democrats held the presidency from 1828-1840.) Figure 2 gives the relationship for 1840, a high turnout election more representative of Vermont sentiment than the earlier years. In the wake of the Panic of 1837, wool prices had fallen.14 No part of the state was very enthusiastic about the Democrats, but their tariff policies were blamed for declining wool prices, so that Democrats became hopelessly uncompetitive in the counties most reliant on sheep.15 Put the other way, the more sheep, the better the pro-tariff Whigs did at the polls.
The 1840 election brought the Whigs to power (“Tippecanoe and Tyler, too”), and with them went the hopes of Vermonters for better wool prices. “No group celebrated Harrison’s victory more optimistically than Vermont’s sheep farmers. They hoped to be rewarded for their votes with tariff relief” (Sherman et al. 2004, 199). However, the 1833 Compromise Tariff Act, arranged by Calhoun and Kentucky Senator Henry Clay, had replaced the Tariff of 1828. It had a sunset clause that led to rapidly declining rates in 1842. For that or other reasons, wool prices fell dramatically. The Whigs raised rates again in 1842 over Democratic opposition, but recovery was slow. Sometime in this period, John Saxe (ca. 1841) of Burlington, a Democratic party activist, wrote “The Whig’s Lament”:
In old Vermont-mont-mont
We’re in a dreadful state.
Instead of fifty cents for wool
We can’t get thirty eight.
They promised if we’ed vote for Tip
That wool would surely rise;
But all they’ve done with wool has been
To pull it o’er our eyes.
By 1844, only a weak recovery had taken place, and on average, Vermont moved a few percentage points against the incumbent administration and toward the Democrats, particularly in counties with relatively few sheep. Only in two counties did voters continue to shed Democratic votes. (See Figure 3.) Addison County, buried in sheep and evangelical fervor (Potash 1991), did so. Strongly abolitionist Lamoille also dropped a bit in its support for the Democrats for reasons unrelated to wool.16
As Figure 3 shows, the changes in 1844 voting strengthened the already strong relationship between sheep densities and presidential voting into even closer alignment in Vermont. (The zero-order county-level correlation with the 1844 Democratic vote is -.7, higher than ever before; graph not shown.) Thus by 1844, Vermont had become a Whig state where the Democrats were competitive only in about half the counties. Voting patterns were closely tied, not to slavery issues, but to the presence of sheep and their need for tariff protection.17