Legislation as Insurance: A Reconsideration of Ambition Theory and the Realignment of the 1850s

John Baughman Bates College [email protected]

Abstract: Ambition Theory (Schlesinger 1966; Aldrich and Bianco 1992; Aldrich 1995) tells us that a candidate will weigh the probabilities, known in advance, that each potential affiliation will lead to winning office. Although a useful characterization of the choice candidates face, its application to the realignment of the 1850s underplays the great uncertainty candidates and office holders faced in making their decisions. I argue that the most viable tool for them to compensate for that uncertainty was to make use of new opportunities in the legislative process, specifically with the introduction of locally-oriented legislation. Although legislative attention to the district was not new, the use of bills sponsored by members was relatively new, as was their sensitivity to electoral threat in offering them. The findings suggest that Antebellum legislators were concerned about their careers and used the tools of their office in response to the partisan upheaval.

Prepared for presentation at the History and Congress Workshop at George Washington University, May 29-31, 2008. Thanks to Chris Abbott, Zach Oren and Paul Suitter for research assistance. All errors of omission or commission remain my own.

How do career-minded politicians choose a party when the party system has been disrupted? Ambition Theory (Schlesinger 1966; Aldrich and Bianco 1992; Aldrich 1995) tells us that a candidate will weigh the probabilities, known in advance, that each potential affiliation will lead to winning office. Although a useful characterization of the choice candidates face, its application to a realignment period – and especially the realignment of the 1850s – underplays the great uncertainty candidates and office holders faced in making their decisions. Gienapp, in his study of the origins of the Republican Party, writes, “Given the bewildering array of factions, alliances, issues, and movements involved, contemporaries, not surprisingly, had difficulty interpreting the significance of the 1854 results,” including leading Senators Benjamin Wade of Ohio and Stephen Douglas of Illinois, who engaged in a public dispute over their meaning (163). In the runup to the 1856 election, numerous experienced political leaders made incorrect bets about the durability of the Know Nothings, Republicans and Whigs (Holt 1999). As a result, “[m]ore than a century later, it is difficult to appreciate the Republican party’s precarious situation in the spring of 1856 ” (Gienapp 1987, 273). The problem was difficult enough with the collapse of a major party and the rise of two new party organizations, but it was compounded by the secrecy of the Know Nothings, the complex dimensionality of Antebellum politics, and the weak loyalties of most voters. The question then becomes how candidates responded to that uncertainty. This paper reconsiders Ambition Theory in light of the uncertainties surrounding the realignment of the 1850s, and investigates how members of the U.S. House of Representatives reacted in light of the threat to their own careers. I argue that the most viable tool for them at the time was to make use of new opportunities in the legislative process, specifically with the introduction of locally-oriented legislation. Although legislative attention to the district was not new by that point, the use of bills sponsored by members was relatively new, as was their sensitivity to electoral threat in offering them. The first section provides an assessment of Ambition Theory. The second section considers the realignment of the 1850s in more detail, and specifically how uncertainty was manifest in the choices and dilemmas of office seekers at the time. The third section evaluates the ways House members could compensate for that uncertainty through the use of bills. The fourth section analyzes bill introductions during the key congresses to determine whether or not they were in fact made in response to the electoral disruptions caused by the realignment. The findings show that members did make more use of legislation, and especially locally-oriented bills, when district electoral conditions were more factionalized. The concluding section suggests that this use of the legislative process prompted rules reforms to increase opportunities for individual legislators, opening the door to the member-driven legislative process identified with the modern House. Reconsidering Ambition Theory Given the multiplicity of paths – through branches and levels of government, through types of offices – how does one design a career in politics? Schlesinger (1966) characterizes the 1

question as one of ambition conditioned by opportunities provided by institutions. In his Ambition Theory, the politician’s “problem consists, first, in defining his office goal or goals and, secondly, in relating his current activity to them. To do the latter is obviously not a simple task, involving as it does a host of uncertainties; at the same time the means of advancing in politics is not a mystery” (6). Although politicians may have policy goals, his main concern is to explain the pattern of office seeking they pursue. For Schlesinger, the uncertainty they face helps explain the variety of observed career paths given a common set of institutional arrangements. Following the lead of Rohde (1979) on progressive ambition from the U.S. House, Aldrich and Bianco (1992) adapt Ambition Theory to an expected utility framework characterizing a candidate’s choice of party affiliation. They start with a pure office seeker who has a choice of two or more parties under which he or she will run for office. For each potential affiliation, the candidate receives utility from the outcome (holding office) conditional on the probability of realizing that outcome (winning election), and incurs a cost. They simplify the calculation by assuming the cost and utility of office are constant across parties, so that only the probabilities of winning with each affiliation must be compared. Aldrich and Bianco extend this simple decision-theoretic framework to consider candidates making this choice across multiple districts, which then becomes a coordination game when the probability of a given outcome is constant across districts or an assurance game when it is not.1 The actor in their model makes a straightforward forecast of the likelihood each party affiliation will produce an electoral victory. In the coordination/assurance game, these probabilities are common knowledge and all players have full information. Furthermore, the variation in probabilities across parties is conditional on the expectation that candidates in the other districts will choose the same affiliation, allowing them to gain power. Under these conditions, once the number of candidates choosing a given affiliation passes a tipping point, they all have an incentive to join. Candidates in the model by Aldrich and Bianco face a risk, whether in the decisiontheoretic framework or in the coordination/assurance game, that they will choose incorrectly and lose the election. What it does not account for, however, is the uncertainty candidates face in calculating those probabilities. As Schlesinger notes, in choosing a career path – office or party – the office seeker makes decisions with limited information. To see how this matters, let us briefly consider the problem in terms of Key’s (1964) familiar formulation, parties in the electorate, parties as organizations, and parties in government, each with a distinct set of challenges. At each stage, an office seeker faces uncertainty about the viability of his or her affiliation with a party. The element most apparent from their model is the dilemma of candidates in choosing a party, in effect their coalition partners once in office. The federal system complicates this problem because political parties with national ambitions must balance policy appeals within the state against national considerations. For the candidate, it means that even if the probability of victory under a given banner is constant across districts within a state, it may not be across 2

states. If those probabilities are common knowledge, then it turns the pursuit of federal office into an assurance game. In practice, it is unlikely to be that simple, within a district let alone across districts. As Fenno (1978) describes, even experienced incumbents do not know with precision the extent of their voter support going into the next election, even without the added problem of choosing a party affiliation and coordinating with others. Prior to the move towards candidate-centered elections, an office seeker was particularly vulnerable not just to the choices of voters, but to the local organization of party or faction loyalists who would mobilize them on election day (Altschuler and Blumin 2000). The problem of uncertainty is compounded by the need of a career-minded candidate to coordinate on a party affiliation with others. Even a cleareyed prognosticator of voter support and activist mobilization must consider whether other potential allies will see their chances the same way. As a result, the probability estimates the office seeker will make about possible party affiliations are fraught with uncertainty, from the choices of voters to those of activists to those of potential coalition partners. Aldrich and Bianco motivate their model by making reference to both the realignment of the 1850s and to contemporary party switches in Congress, and Aldrich (1995) uses it for a deeper elaboration of the realignment period. They are, in fact, two different dilemmas. Estimates of expected utility can be made with some precision when the party coalitions are stable and only a small number of office seekers (or office holders) switch. Billy Tauzin (LA) and Richard Shelby (AL) knew at the time of their switch from the Democrats to the Republicans in 1995 that they would be in the majority, and indeed that was part of the enticement (Choate 2003). Switchers are also rewarded with favorable committee assignments and seniority concessions, which can be negotiated in advance (Choate 2003; Yoshinaka 2005). The switch is not free of risk altogether, because voters may penalize the incumbent when he or she is up for reelection (Grose and Yoshinaka 2003). Nevertheless, the potential switcher evaluating the utility of changing parties under a stable two-party system has a good idea in advance whether it will benefit his or her career. With a realignment, on the other hand, and especially one as dramatic as in the 1850s and the collapse of a major party, previous patterns of voting behavior are disrupted, making forecasts by candidates more difficult. A potential office holder must evaluate the probabilities of succeeding with each of the possible affiliations, which depends also on the likelihood that other candidates will run on and win with that affiliation. 2 And rather than involving a choice between two known party organizations, in a realignment the choice may be among multiple nascent organizations, however loose, whose potential future gains vary across states. The question then becomes how office seekers respond to that uncertainty. Certainly they do not abstain from the choice, because even in the midst of American party realignments candidates nearly always have opted to run under a partisan or factional banner. Instead, I argue that they will act to compensate for that uncertainty over party choice by developing an independent appeal to voters. Not all candidates are equally situated to do so, nor does every era provide equal opportunities. To see how uncertainty matters in practice, the next section provides an overview of the disruptions caused by the collapse of the Whigs and the rise of the Republicans. The third section then suggests that institutional changes provided new 3

opportunities for House members to make direct appeals to constituents during the Antebellum era. Uncertainty in the Realignment of the 1850s After the election of 1852, Aldrich (1995) argues, the Whig Party was in clear decline. Although a few leaders such as William Seward of New York thought as late as 1856 that the party could recover – indeed, as late as 1865 some southern Whigs sought to revive the party (Holt 1999, 984-5) – and although the party held onto power in Pennsylvania into 1858, after 1854 it was no longer nationally viable. The question then became what would replace it. The rapid rise of the Know Nothings, which shocked most contemporary observers, was the immediate impetus for the Whigs’ demise. They swept into office in much of New England, and influenced elections in most of the rest of the North. At the same time, anger over the KansasNebraska Act spurred abolitionist organizing, which found greatest success in Midwest states like Michigan and Wisconsin. Aldrich’s story is that former Whigs in the North faced a strategic choice. Either they could put their lot in with the Know Nothings, whose share appeared confined to the Northeast, or they could opt for the Republicans, who held power in the Midwest and were also competitive in the Northeast. According to Aldrich, by 1855 office seekers recognized the choice of party affiliation as a coordination game, and power was tipped decidedly towards the Republicans. Though Aldrich’s characterization is correct in broad outline, the story of the decline and fall of the Whigs and the rise of the Republicans is replete with false turns, unexpected defeats and surprising alliances. Gienapp (1987) argues that however inevitable the success of the Republican Party appears in retrospect, for participants at the time it was far from certain, even after the tipping point Aldrich identifies. This section will focus on three elements of the story that had a particularly acute effect on the uncertainty felt by political leaders: The Know Nothings’ commitment to secrecy, the multiple salient dimensions of 1850s politics, and the diminished loyalties of voters. Nativism was not new to American politics in 1850. Movements of various kinds ploughed this ground beginning with the Native American Democratic Association in 1835. The Know Nothings’ success took almost everyone by surprise, however, and that was by design. The Order of the Star Spangled Banner was founded in New York in 1850, and like similar organizations of the time it sought to raise citizenship requirements, limit public office to nativeborn Americans, and protect against Catholic influence. Unlike the others, its membership was secret. The group started slowly – two years later it still had only 43 members – but by early 1854 it had acquired the nickname the “Know Nothings,” thousands of members, and a reputation for influencing elections. And by November of that year it had more than a million in its ranks and controlled most public offices in Massachusetts, among other places (Anbinder 1992). The secrecy caused two specific problems for opponents of the Know Nothings in the early years of the Whig collapse. The first involved voting. One of the challenges the major 4

parties faced at the state and local level during Antebellum period was the relative ease for insurgents to print and distribute ballots of their own, not to mention for voters to create their own hybrid ballots (Altshculer and Blumin 2000; Bensel 2004). This facilitated, for example, the competing pro- and anti-Nebraska Democratic slates in some states. With the secrecy of the Know Nothings, even loyal party members sometimes did not know whether they were being handed a strip for a Know Nothing slate under the guise of the Democrats or Whigs, and this contributed to their surprising success in 1854 (Holt 1991, 131). Until the Know Nothings dropped their commitment to secrecy in the middle of 1855, office seekers and activists outside the Know Nothings had difficulty gauging the extent of their support, and voters whether they were being duped. Second, the secrecy complicated the building of competing party organizations and of governing coalitions, even for possible allies of the Know Nothings. Combined with the increasing array of new factions, erstwhile Whigs who sought new supporters through 1855 had trouble assessing possible coalition partners without confirmation that they were members of the society (Holt 1999, 840). The dilemma become most acute when it came time to organize the House of Representatives at the start of the 34th Congress in December 1855. Democrats appeared to be relegated to the minority, but it was not yet clear who held the majority. At the time of their elections – held between August 1854 and November 1855 – about two-thirds of the anti-administration members were said to be affiliated with the Know Nothings, but by the time the session began some Know Nothings had joined the Republicans, and some Whigs the Know Nothings (Gienapp 1987, 240; Anbinder 1992, 197). The speaker chosen after two months and 133 ballots, Nathaniel Banks of Massachusetts, was nominally a member of the Know Nothings, but owed his office to an anti-Nebraska coalition for which the Republicans took credit (Anbinder 1992, 201). After early 1856, however, the problems that the Know Nothing secrecy caused for office seekers had dissipated. Slavery, and specifically support for the Kansas-Nebraska Act, was ultimately the key dimension separating Democrats and Republicans, and nativism was the catalyst for the collapse of the Whigs, but the politics of the mid-1850s was a good deal more complex than that. In fact, there were several dimensions – slavery, nativism, anti-Catholicism, temperance – which were each salient enough to mobilize activists and voters, but which were only imperfectly correlated with one another. The precarious sectional alliance that had been in place since the Missouri Compromise unraveled with the Nebraska issue, as it came to be known. Although both parties had defectors like Pennsylvania Democrat David Wilmot, not until that 1854 law were the party coalitions threatened. Candidates and factions immediately defined themselves as pro- or antiNebraska. The shorthand explanation that nativism undid the Whig Party underplays the variation in how that issue was manifest in local politics. Voters and activists responded both to the votes immigrant groups could offer and to the complaints of non-immigrants about their perceived eroding power, especially in urban centers. An explosion in immigration which peaked in the early 1850s brought several hundred thousand Irish and Germans, most of them Catholic and many unskilled, and smaller numbers of other groups. At the same time, there was growing 5

concern about Catholic influence in public life, especially in schools and patronage appointments. But not all anti-Catholic voters were nativist; newly arrived Protestants were tempted to sign on to anti-Catholic coalitions so long as the anti-immigrant messages were suppressed (Anbinder 1992). As a result, the Know Nothings – not to mention competing parties that wanted to win nativist support – shaded their message to suit the local mix of religions and ethnicities, but in so doing inhibited their ability to build organizations across states. Temperance was thrust onto the national agenda by the Maine Law, passed by that state in 1851 to ban the sale of spirits. The success of the anti-drinking movement there encouraged temperance supporters to push the issue in other states. It gained supporters for religious and health reasons, and became tied for some to nativist claims against hard-drinking immigrants (Anbinder 1992, 43). Temperance became a wedge issue for anti-Democratic leaders in the 1850s to peel away pro-temperance Democratic votes, though Whigs found that it could just as easily lose them votes with anti-temperance members of their own party (Holt 1999, 690). Whigs and Free Soilers quickly made the Maine Law a central plank in their state platforms in many parts of the country, so much so that the Free Soil Party in Maine emphasized its opposition to drinking above its opposition to slavery in the 1853 state elections (Gienapp 1987, 50). Even as many voters expressed their strong support for temperance, especially when combined with nativism, parties just as easily overreached. Coordination across states was thus challenging for parties, and even within states it was not a foregone conclusion. Democrats and Whigs had long tailored their messages to local needs, the most obvious manifestation being their respective attempts to keep slavery off the national agenda, but with unstable party organizations this became even more difficult. Some states already had competing factions within parties, like the Hardshell and Softshell Democrats in New York, a problem that became more pronounced as the Democrats and Whigs each faced internal tensions over Nebraska. Illinois illustrates the confusion across congressional districts in 1854, with an inchoate Republican Party, anti-Nebraska Democrats and Whigs, and Free Soilers, some of them with apparent Know Nothing ties, all competing for the abolitionist vote in different parts of the state. By 1854, every state outside the Deep South faced complicated multiparty elections. In some states, especially in the Midwest, political leaders attempted to negotiate fusion tickets combining multiple factions or parties under a single label, usually called a “People’s” ticket. Typically they were defined not by any clear ideology but in terms of who or what was opposed, such as the Democrats or slavery, not all of these attempts were successful, and even those that were did not last long. One newspaper in Illinois wrote prior to the 1854 election that “there will be less of fusion than confusion” (quoted in Gienapp 1987, 104). In Maine an attempt to create a statewide fusion that same year foundered on Nebraska, leading to a “multiplicity of tickets” which “made it impossible to forecast the results of the fall election” (Gienapp 1987, 131). Indiana’s fusion effort hung on through 1856 before becoming the basis for its Republican Party, as did the one in Illinois, but not before each state’s anti-Democratic coalition made difficult compromises on abolition, nativism and temperance (Gienapp 1987).

6

Weakened party ties together with the complex dimensionality of Antebellum politics meant that support from voters was unpredictable. By the early 1850s, according to Wilentz (2005), voters had become “simply fed up with mainline parties that had dropped their old ideological appeals and appeared to be nothing more than patronage and vote-gathering organizations” (683). Although some voters remained committed, most seemed to have only an intermittent and reluctant engagement with parties and the process (Altschuler and Blumin 2000). In this environment, and with one of the two major parties in rapid decline, office holders and activists had the difficult task of corralling the many disaffected eligible voters into new electoral coalitions. Political leaders saw Republican success in 1856 in particular as dependent on the degree to which the party could persuade undecided voters (Gienapp 1987). A calculation of state-level returns showed almost no correlation (r = 0.038) between the Whig vote in 1852 and the Republican vote in 1856 (Kleppner 1979, 50). Another study found that the township-level vote for Know Nothings in New York in 1854 was only weakly related to their vote just one year later (Gienapp 1987, 232). The multiple policy dimensions together with state and local variations in religious, ethnic and economic makeup meant that forecasting the vote from one election to the next or between states was highly problematic for contemporary observers. German Lutherans in parts of Pennsylvania, for example, were anti-Democratic due to the party’s affiliation with Catholics (including German Catholics), but in the upper Midwest they were drawn to the Democrats due to their opposition to temperance (Holt 1991, 78-9). Similarly, the historic links between the Scotch-Irish Protestants and the Whigs, due to the Democratic embrace of Irish Catholics, became frayed when Whig support for temperance became salient after the Maine Law (Holt 1999). In some cases, parties found expedient solutions to fight the next election. Iowa Republicans in 1856 wanted to solidify their position in the state as the chief alternative to the Democrats, but faced a dilemma between the Germans opposed to temperance and the Know Nothings in favor; they resolved the issue by ignoring it altogether (Gienapp 1987, 278-9). On other occasions political leaders underestimated the salience of an issue dimension to voters, such as New York in 1854 when Whigs mistakenly thought Nebraska and not the Maine Law would drive the election (Gienapp 1987, 156). Not all Protestants were unified from one election to the next in any case. In Indiana, Presbyterians broke from Baptists and Methodists to support the People’s fusion ticket in 1854, while the other two denominations stayed mainly with the Democrats; in fact, even though Methodist leaders were vocally pro-temperance, Methodist voters were as likely to go with the Democrats as not (Gienapp 1987, 162). The Know Nothings’ nativism and anti-Catholicism forced difficult choices on Protestant immigrant groups of all stripes, in some cases driving such groups to the Democrats, Whigs or Republicans, and in others forcing the Know Nothings to restrict their rhetoric to antiCatholicism. When the Republicans absorbed most of the former Know Nothing leaders, they faced the same dilemmas, combined with the party’s strong anti-slavery stance which further split some voting blocs. The general disaffection with party politics together with the complexity of 1850s issues caused great anxiety for the faction leaders trying to win office and grow a new party organization. 7

As a result, office-holders had an incentive to hedge their bets by building independent support in their constituency. There is evidence, albeit limited, that House members in the 1850s enjoyed a small incumbency advantage despite the dramatic reshuffling of party coalitions. Alford and Brady (1993) calculate the sophomore surge – the difference in vote between an incumbent’s first and second elections – as a bonus of about one percent at the time. Although small by modern standards, not until after Reconstruction did the surge exceed that level again, and not until the early 20th century did a net positive sophomore surge continue for more than one congress in a row. The next section considers how some did so. The Rise of Local Bills In the face of the great partisan upheaval of the mid-1850s, a career-minded legislator would want to use his office to reduce the uncertainty that voters support his candidacy in the next election. A tool common in the Jacksonian period was patronage, the distribution of offices to local allies: As Holt (1999) put it, “the currency politicians dealt in” at the time (416). Realignment presented several obstacles to its use in this context, however. The Whigs had long criticized the Democrats’ spoils system under Jackson and Van Buren, but hesitated little once winning office in making use of it themselves in order to build local party organizations, especially with William Henry Harrison’s election as the first Whig president in 1840. By the end of that decade, Whigs came to rely on the spoils of office every bit as much as the Democrats in rewarding allies and appealing to constituents. According to Holt (1999), “Patronage allocation, in short, profoundly affected the morale and performance of local activists and Whig voters in elections” (418). The early 1850s dealt a fatal blow to Whig patronage almost everywhere. The election of 1852 returned the presidency to the Democrats, who also controlled 25 of 31 governorships. The Whigs never regained the White House and never again held more than the six state houses they controlled in 1853.3 As the previous section discussed, the subsequent four years saw Whigs turned rapidly out of office in almost every part of the country, and the party organization decimated by defections to the Know Nothings, Republicans, various and sundry fusion efforts, and even the Democrats. As such, a Whig candidate no longer had an organization upon which to maintain a patronage network, nor likely prospects in office upon which to promise spoils to potential supporters. The rapid rise of Know Nothingism presented additional challenges to the use of spoils for a career-minded legislator. After having voiced accusations for a quarter century against the Jacksonians of widespread corruption in the allocation of patronage, the partisan manipulation of the postal service, and a culture of graft, the Whigs found themselves no less vilified by the public for their own perceived corruption during their years in power. The Know Nothings capitalized on these sentiments and staked their growth not just on a platform of nativism, but on a broad critique of partisan corruption (Anbinder 1992). The two threads of the movement came together, fortuitously for them, in the controversial 1853 appointment by Franklin Pierce of James Campbell as postmaster general, a prominent Philadelphia Democrat and a Catholic. An explicit appeal of the Know Nothings, therefore, was their aim to destroy existing patronage 8

networks and party organizations in the name of restoring republicanism on behalf of the (nativeborn white Protestant) people (Holt 1983). Given that Whigs saw the Know Nothings as their greatest competition to winning office, at least through 1855 in most parts of the country, the use of patronage to hold onto power became politically problematic for them. An alternative for members of Congress was to use their legislative powers in order to curry favor with local supporters. In comparison with the modern Congress, Antebellum House institutions offered limited opportunities: Committee membership was appointed by the session rather than by the congress, and with limited expectations of reappointment; subcommittees, to the extent they were used at all, were ad hoc and had no independent power; members had no staff support, apart from a single clerk employed by a committee chair; with highly malleable committee jurisdictions, there was little incentive for specialization; and there was no practice cosponsorship of legislation. Two developments offered Antebellum legislators a mechanism to use their office to appeal directly to local supporters. First, communication between representatives and their constituents became commonplace. As Cunningham (1978) documents, from the earliest days of the House members sent circulars home to the district, both distributed directly and reprinted in local newspapers. Often these letters included accounts of policymaking efforts on behalf of the district, at a rate of about two specific claims of credit per letter (Swift 1987), as well as explanations of the goings on in Washington. Expansion of the postal service through annual and biennial post road bills to every rural corner made possible, among other things, the wide circulation of political messages (Baughman 2007). Discounts for newspaper distribution through the mails and the proliferation of a partisan press meant the federal treasury heavily subsidized political communication (John 1995). Partisan use of the franking privilege by both members of Congress and by local postmasters, who were patronage appointments, allowed for broad distribution of government documents such as committee reports and members’ “bunkum” speeches to constituents, and for constituents to send letters to their representatives for free (Baughman 2006). Although regular records of the use of the frank were not kept, one postmaster general reported that 20,363 letters and about sixteen and a half tons of documents were sent by Congress over three weeks in May, June and July of 1840 (Van Buren 1840). By the early 1850s, representatives were well accustomed to communicating about politics to their constituents, including legislative matters of concern to the home district, and constituents and newspapers were well accustomed to hearing from them. Second, in 1837 representatives gave themselves the power to introduce legislation under their own name. The early Jeffersonian House was a chamber based on egalitarian principles of lawmaking, a “forum where every member was a peer and no man led” (Harlow 1917, 145). As Cooper (1988) details, the standard process involved a policy proposal being brought to the floor, the Committee of the Whole deciding on the principles, and the chamber naming a select committee to draft the legislation. The bill itself was considered “inchoate law” (Cooper and Young 1989, 69) and an expression of the will of the House. Members sought opportunities to bring forward policy proposals of their own, but found only limited avenues to do so in the form of petitions which originated from the people, and resolutions – almost always only a single 9

sentence in length – asking that committee inquire a topic of concern. In the 25th Congress, the rules were changed to allow for the introduction of bills by individual legislators without the mediation of the Committee on the Whole as in the former Jeffersonian system. Cooper and Young (1989) argue that the changes were significant in two respects. First, the rules set aside designated days to allow the introduction of bills by leave of the House, giving them a distinct space in the legislative process. Second, rather than specifying that a select committee would be named to consider the bill, the new rules omitted such mention and implicitly handed to the standing committees authority over leave bills. Conditional on a one-day advance notice and on leave being granted by the floor, individual legislators secured the right to offer their own bills for the first time. Legislators were slow to adapt to the new rules, which is not surprising given the substantially greater cost they incurred drafting a bill instead of the far more perfunctory resolutions to inquire. The first bill under the rule was H.R. 540, introduced by George Evans of Maine on February 8, 1838, to have the northeastern boundary of the United States surveyed.4 On only a dozen more occasions that congress did members request leave to offer bills under their own name, and in only ten of them was leave granted. Two decades later, it was a very different story. In the 35th Congress, members sought leave to introduce a bill on 333 occasions, and leave was granted for all but one of them.5 The legislation members introduced on leave covered a broad range of topics. During the 34th Congress Luther Hartnett, a Missouri Whig who had won election against a Benton Democrat and an Anti-Benton Democrat, brought H.R. 12 on February 18, 1856, to appropriate $500,000 to the Secretary of War for improvement of the Des Moines Rapids on the Mississippi River, one of seventeen bills he introduced that congress.6 Three days later his colleague James Denver, an Anti-Broderick Democrat from California, introduced H.R. 53 to establish railroad and telegraph links between the Atlantic and Pacific Oceans; he had defeated Whig and Broderick Democratic candidates to gain office.7 One week after that, Benjamin Leiter of Ohio offered H.R. 88 to provide a pension of eight dollars per month for a constituent, Amos Armstrong. Leiter was part of a slate of a People’s coalition of anti-Nebraskans, mainly Know Nothings and Whigs, that swept the House elections in the state against the Democrats.8 FIGURE 1 ABOUT HERE The increase in bill introductions was not linear. As Figure 1 shows, the number of bills members offered increased slowly for the first four congresses as legislators adapted before leveling off.9 Then in the 33rd Congress, they became markedly more active, reaching a new plateau maintained through the 35th. Little had changed in the House rules or organization to explain such a jump.10 The content of bills introduced between the 25th and 35th congresses provides another clue. Figure 2 disaggregates bill introductions by how locally focused the bill is – private, local, state, or general public policy.11 The graph shows that while all bill introductions increased from the 32nd to the 33rd, the bulk of that increase was driven by local and, to a lesser extent, private bills. That is, whereas general public legislation comprised 48% 10

of bill introductions through the first eight congresses, they were only 33% in the 33rd through 35th congresses, while the proportion of local bills increased from 23 to 38%. FIGURE 2 ABOUT HERE Explaining the rise in legislative activity by rank-and-file representatives, therefore, means also explaining why their efforts began to have a markedly more local cast than before. The argument here is that it was driven not by changes in the institutional structure or opportunities within the House – at least not after members had gained the ability to offer bills in 1837 – but rather by pressures and demands outside of the House, specifically the partisan upheaval in their districts.12 Analysis Let us assume, like Aldrich, that incumbents are career-minded, and that partisan affiliation is consequential for their careers.13 The question here is whether Antebellum representatives used legislation in response to electoral uncertainty, and to answer that we must first operationalize what is meant by uncertainty. The literature typically uses two measures of electoral vulnerability, the margin of victory at the last election and freshman status or length of tenure in office. The size of a winner’s plurality indicates how vulnerable he is to an electoral swing in the next election (Jacobson 2009). There is reason to believe, however, that this will not play an important role for legislators in the 1850s. Mann (1977) observed in his study of the modern House an increase in the magnitude of the swing from one election to the next, which Jacobson argues makes hitherto large margins unsafe. With the highly unstable partisan environment of the late Anebellum era, some legislators may well perceive even relatively large margins as unsafe. On the other hand, the Alford and Brady (1993) findings on the sophomore surge during that era suggest that length of tenure may have a small effect on perceptions of electoral safety. In the case of elections during a realignment like the mid-1850s, I hypothesize that the problem of uncertainty for incumbents was how unstable the party system was in the district. Due to the rapid shifts in coalitions between 1852 and 1858, and especially between 1853 and 1856, office seekers wanted to know the likelihood that the arrangement of partisans would endure at least as long as the next election. Moreover, this would be so whether or not they decided to switch parties; those deciding to switch would want to predicate that on the probability that the new party would succeed, as Aldrich notes. As a proxy for factional instability, I use the number of candidates who competed for the seat in a given election. During periods of great organizational stability – say, in the midst of the Jacksonian period, or even more so in the contemporary House – House elections would ordinarily involve two candidates, a Democrat and a Whig or a Democrat and a Republican. Many races in the mid-1850s, in contrast, had many more than two. Elihu Washburne of Illinois, for example, was first elected in the 33rd Congress as a Whig against a Democratic opponent and a Free Soiler. By the time he stood for reelection two years later he had switched to the Republicans and faced off 11

this time against a mainline Democrat and an Anti-Nebraska Democrat. That same year James Woodworth, running to serve his first term in a neighboring district also as a Republican, competed against a mainline Democrat and an Anti-Nebraska Democrat while also facing a Whig. Not every state saw such elections. Pennsylvania elections to the 34 th were contested in most cases by only a Democrat and a Whig, with the occasional Independent Whig, Independent Democrat or American Party candidate running as well. It is also important to note that the measure is not limited to the “official” candidates running under a party label in that election. In numerous locations there were multiple factions struggling for control over a party organization. The most important, but far from the only, example was the Hunker and Barnburner Democrats in New York, not to mention the split of the Silver Greys from the main Whig Party in that same state. There was broad variation in factionalism across states, and local variation within states as well, as the Illinois and Pennsylvania examples show. The ease of ballot access, noted above, underlines the usefulness of our proxy. Factional infighting and third or fourth party challenges were not unique to the mid1850s, even in comparison with the Jacksonian period. If anything, they were less common than in the height of the second party system: On average 2.7 candidates ran for every House seat in the 28th through 32nd congresses, while 2.5 candidates ran per seat in the 33 rd through 35th. Rather, because of the collapse of the Whig party, and therefore the collapse of a quarter century of stable major-party competition, the proliferation of candidates signaled something more threatening for incumbents. Whereas it was clear to participants in the second party system that the Democrats and Whigs would survive the advances of the Liberty Party in the early 1840s and the intermittent Anti-Mason challenges into the 1830s, that was not the case for the parties and coalitions that arose after 1852. Combined with the secrecy and mass popularity of the Know Nothings, the stubbornness of some leading Whigs in the face of their party’s collapse, and antiNebraska divisions among Northern Democrats, the argument here is that multiple opponents in an election in the mid-1850s indicated partisan uncertainty, and that representatives responded by investing in more locally-oriented legislation. The correlation between election opponents and legislative activity provides initial support for this claim. For the 28th through 32nd congresses, local bills and the number of opponents in the last election were not related (r = 0.02, n.s.), unlike for the period of partisan upheaval in the 33rd through 35th (r = 0.09, p < 0.05).14 The relationship becomes more pronounced when we consider the regions Aldrich identifies as beset with particularly acute partisan turmoil during those latter three congresses, New England and the Upper Midwest.15 In the New England states, the number of opponents and of local bills were correlated at r = 0.24 (p < 0.05) and for the Upper Midwest at r = 0.23 (p < 0.01). For bills on general public matters, the national and regional correlations fall well short of conventional levels of significance. In order to determine whether the relationship between electoral uncertainty and bill introductions is robust, a negative binomial regression was estimated for the number of bills a representative introduced in a congress. Three measures of electoral threat are used: The number 12

of opponents faced in the last election, the electoral margin in that election, and how many terms the legislator has served in the House. The abundant literature on legislative behavior leads us to expect that there will be a negative relationship between margin and legislative activity, though for the reasons noted above this may be less applicable to the era under study here. For tenure, two different results seem plausible. On the one hand, if tenure is associated with seat safety, then as the number of terms served increases then the amount of legislating should go down. On the other hand, because members during this era lacked staff and expert advice for drafting legislation, we may observe more activity by senior members, who with their experience see lower costs to legislating. For the key explanatory variable considered here, the hypothesis is that there is a positive relationship between the number of opponents and bill introductions. One could not introduce bills if absent from the House of Representatives. As such, as a simple control for attendance – and therefore opportunity – I use the proportion of roll call votes a member missed during the congress, as calculated from the Poole and Rosenthal roll call data.16 Dummy variables were added for the representative’s region and whether he was a member of the majority party. DW-Nominate first and second dimensions scores are included, with the first dimension during this period corresponding to standard liberalism-conservatism and the second dimension roughly reflecting North-South sectionalism (Poole and Rosenthal 1997). Finally, as a control for the demand for federal policy generically from a district, I used the number of years since the state was admitted to the union (Baughman 2008).17 To compare legislative activity before and during the realignment of the 1850s, the model is estimated for the 28th through 32nd congresses and for the 33rd through 35th. In order to evaluate whether the legislative activity was particular to district concerns, I also compare the introduction of locally-oriented bills and general public bills. The results are reported in Table 1. TABLE 1 ABOUT HERE Electoral vulnerability does appear to affect representatives’ legislative activity, largely in the manner predicted. The number of electoral opponents an incumbent faced in the last election has a positive effect on bill introductions during the realignment period, but not during the Jacksonian years. The effect is more pronounced for locally-oriented bills, but contrary to expectations it is also present for general public bills. Electoral margin does not influence legislative activity, as suspected. Tenure in office has a more subtle effect, such that both hypotheses are supported. More junior members, who may enjoy a smaller personal vote, engaged in more locally-oriented legislating during the realignment years but not during the Jacksonian. On the other hand, more senior members – those with more experience legislating in the House – introduced more bills on general public matters. Two variables have a strong and consistent effect across both periods and bill types. Nonattendance for roll call votes strongly inhibits opportunities for bill introductions. Although some missed roll calls may be for health or other idiosyncratic reasons, it may also mean that those legislators trying to exercise leverage over the majority party by withholding their attendance in quorums may also sacrifice other forms of participation. The results also show that members representing newer states are much more active in the legislative process. As 13

hypothesized, the more recent admissions seem to drive requests for policy from the federal government, like the Hartnett and Denver cases in the previous section illustrate. Finally, partisanship and ideology play a role in legislative activity by members. The late Jacksonian years saw two distinct relationships for the two types of bills. Minority party members were mildly more active in the use of locally-oriented bills than those in the majority. More notably, the DW-Nominate first dimension results show that Democrats engaged in more legislating on general public topics than Whigs. The second dimension has a more consistent effect across the two periods. Northern representatives offered more bills than their southern counterparts, consistent with the latter’s deep skepticism of federal power during the Antebellum period. Together these results show that members had two utility functions for the two classes of bills. Those for local consumption were relatively easy to author, and were used when electoral threat was greater. This supports the claim in this paper that legislation acted like insurance against the uncertainty over party affiliation during the realignment. Those bills on broader matters of public policy, in contrast, were costlier to introduce and subject to a somewhat greater degree to ideological and partisan considerations. It may therefore be a mistake to treat all legislating alike, at least in the years before the resources and institutions available to members reduced the cost of bill introductions. Conclusions Ambition Theory (Schlesinger 1966) isolates the critical institutional constraints on the career-minded calculations of office seekers, and Aldrich and Bianco (1992) extend that theory and characterize an expected utility function as it regards the calculation of party affiliation. But while that theory appears to do well when parties are stable, and therefore the utility estimates of the candidate straightforward, it becomes more problematic when they are highly unstable such as during a realignment. In such circumstances, I argue that a career-minded officeholder will use the fruits of office to reduce that uncertainty by appealing directly to voters, specifically by using locally-oriented legislation as insurance. This is a natural extension of the idea of the personal vote, which legislators cultivate in order to buffer partisan swings between elections (Fenno 1978; Cain et al. 1987). The 1850s were a period of tremendous partisan upheaval, especially in the northern and western states. With the collapse of the Whigs, coalitions splitting along multiple salient dimensions, and the rise of new party organizations, career-minded legislators were struck with a serious dilemma on choosing a partisan path. As Aldrich describes, at the most basic level former Whigs, seeking a new home, opted against more regionally limited Know Nothings and in favor of the Republicans. But even in the midst of that apparently orderly tipping game, the legislators facing that decision were burdened with great uncertainty over whether they were making the correct choice. In response to that uncertainty, they began to make use of locallyoriented legislation to win district support independent of party.

14

By the end of the Antebellum era, the interest in opportunities to introduce bills had become so great that members demanded greater access to the floor. In Alexander’s (1916) telling, this came as a direct result of the electoral value representatives had come to see in them: At first bills were not numerous. The modern habit of using them to advertise a member’s activity did not then obsess the legislative mind. . . . But when the fact developed that a bill introduced, though not passed, benefited [sic] the member, since it evidenced a disposition to serve his constituents, the House (1860) set apart each alternate Monday for their introduction and ordered them referred without debate. This radical change, encouraging members to present bills on all possible subjects, created such a Monday rush that it increased their number nearly twenty-fold in twenty years, and nullified all notices of intention and requests for leave (217). As he describes, the reform adopted in the 36th Congress guaranteed to legislators that each week they would have time allotted for the introduction of bills. Moreover, bills on leave were to be given first priority over other legislation and the referrals were not subject to a motion to reconsider (Cooper and Young 1989). Although the House for more than a decade had faced serious procedural logjams in the introduction and disposition of legislation, a proximate cause for their concern in 1860 appears to have been their recent frustration during the previous several congresses at attempting to introduce locally-oriented legislation for their districts. The reforms adopted in response to that experience opened the door to the explosion in constituency-oriented legislation after the Civil War, especially of private bills. An irony is that a chief obstacle to members’ ability to introduce bills had been their own propensity to use the legislative process for position taking. They would dispense with the call of the states and territories, the time reserved for members to bring leave bills and resolutions, in order to resolve into the Committee of the Whole so that they could make their Mayhewian appeals to constituents back home. Lawrence Branch (D-NC) remarked during the debate on the reforms that “the anxiety of gentlemen to go into committee to make Buncombe speeches is so great that they do not allow the call to be gone through with” (Cong. Globe, March 16, 1860, 1206). These “Buncombe” or “bunkum” speeches had been a common practice for at least four decades, where members would devote floor time to addressing matters of limited relevance to the topic under consideration but of great interest to readers in the district. In the standoff between two legislative tools for appealing to constituents, members had shunted aside the costlier one in favor one that was simpler – though, from a policymaking perspective, one ineffectual as well. The reform meant that making time for the Buncombe speeches would come only after bills and resolutions could be introduced. The data also provide a small counterpoint to the Holt thesis that party politics had become a struggle over control of patronage rather than substantive policy disagreements by the end of the Jacksonian period. While that may be so, those legislators who did put forward policy proposals of their own at that time seemed more concerned with matters of broader importance. When the Whigs collapsed and the party system splintered, members of the House focused their 15

legislative energies on issues of narrower, local focus. The partisan upheaval of the mid-1850s thus educated members of the House about the value and opportunity for using legislation to appeal to constituents. While they did so initially in reaction to the electoral uncertainty they faced with the collapse of the Whigs, they did not curtail the introduction of bills once the third party system stabilized.18 Rather they pushed for even greater opportunities to offer legislation, which with the end of the Civil War and the explosion of pension claims they were more than eager to seize. But while locally-oriented legislation and candidate-centered appeals did not come into full flower until late in the 19th century, the seeds were sown in the years just prior to the Civil War.

16

Bibliography

Aldrich, John H. 1995. Why Parties? The Origin and Transformation of Party Politics in America. Chicago: University of Chicago Press. Aldrich, John H., and William T. Bianco. 1992. “A Game-Theoretic Model of Party Affiliation of Candidates and Office Holders.” Mathematical and Computer Modeling 16: 103-16. Alexander, DeAlva S. 1916. History and Procedure in the House of Representatives. Boston: Houghton-Mifflin. Alford, John R., and David W. Brady. 1993. “Personal and Partisan Advantage in U.S. Congressional Elections, 1846-1990.” In Congress Reconsidered, 5th ed. Ed. Lawrence C. Dodd and Bruce I. Oppenheimer. Washington, DC: CQ Press. Altschuler, Glenn C., and Stuart M. Blumin. 2000. Rude Republic: Americans and Their Politics in the Nineteenth Century. Princeton, NJ: Princeton University Press. Anbinder, Tyler. 1992. Nativism and Slavery: The Northern Know Nothings and the Politics of the 1850s. New York: Oxford University Press. Baughman, John. 2006. “Postal Reform in the 1840s: Constituent Demand and the Limits of Party Power.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, Ill. ----------. 2007. “Building Post Roads: The Electoral Connecting and the Antebellum Pork Barrel.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, Ill. ----------. 2008. “Becoming Lawmakers: Policymaking Effort in the Antebellum House.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, Ill. Bianco, William T., David B. Spence, and John D. Wilkerson. 1996. “The Electoral Connection in the Early Congress: The Case of the Compensation Act of 1816.” American Journal of Political Science 40: 145-71. Brady, David, Kara Buckley, and Douglas Rivers. 1999. “The Roots of Careerism in the U.S. House of Representatives.” Legislative Studies Quarterly 24: 489-510. Cain, Bruce, John Ferejohn, and Morris Fiorina. 1987. The Personal Vote: Constituency Service and Electoral Independence. Cambridge, MA: Harvard University Press.

17

Carson, Jamie L., and Erik J. Engstrom. 2005. “Assessing the Electoral Connection: Evidence from the Early United States.” American Journal of Political Science 49: 746-57. Choate, Judd. 2003. Torn and Frayed: Congressional Norms and Party Switching in an Era of Reform. Westport, CT: Praeger. Cooper, Joseph. 1988. Congress and Its Committees: A Historical Approach to the Role of Committees in the Legislative Process. New York: Garland. Cooper, Joseph, and Cheryl D. Young. 1989. “Bill Introduction in the Nineteenth Century: A Study of Institutional Change.” Legislative Studies Quarterly 14: 67-105. Cunningham, Noble E., Jr. 1978. Circular Letters of Congressmen to Their Constituents, 17891829. Chapel Hill: University of North Carolina Press. Fenno, Richard F. 1978. Home Style: House Members and Their Districts. Boston: Little, Brown. Gienapp, William E. 1987. The Origins of the Republican Party, 1852-1856. New York: Oxford University Press. Grose, Christian R., and Antoine Yoshinaka. 2003. “The Electoral Consequences of Party Switching by Incumbent Members of Congress, 1947-2000.” Legislative Studies Quarterly 28: 55-75. Harlow, Ralph V. 1917. The History of Legislative Methods in the Period Before 1825. New Haven: Yale University Press. Hinds, Asher. 1907. Parliamentary Precedents of the House of Representatives of the United States. Washington, DC: Government Printing Office. Holt, Michael F. 1983. The Political Crisis of the 1850s. New York: W.W. Norton. ----------. 1991. Political Parties and American Political Development from the Age of Jackson to the Age of Lincoln. Baton Rouge: Louisiana State University Press. ----------. 1999. The Rise and Fall of the American Whig Party: Jacksonian Politics and the Onset of the Civil War. New York: Oxford University Press. Jacobson, Gary C. 2009. The Politics of Congressional Elections, 7th ed. New York: Pearson Longman. Jensen, Laura. 2003. Patriots, Settlers, and the Origins of American Social Policy. New York: Cambridge University Press. 18

John, Richard R. 1995. Spreading the News: The American Postal System from Franklin to Morse. Cambridge, MA: Harvard University Press. Kernell, Samuel. 1977. “Toward Understanding 19th Century Congressional Careers: Ambition, Competition, and Rotation.” American Journal of Political Science 21: 21-33. Kleppner, Paul. 1979. The Third Electoral System, 1853-1892: Parties, Voters, and Political Cultures. Chapel Hill, NC: University of North Carolina Press. Mann, Thomas E. 1977. Unsafe at Any Margin: Interpreting Congressional Elections. Washington, DC: American Enterprise Institute. Meinke, Scott R. 2007. “Slavery, Partisanship, and Procedure in the U.S. House: The Gag Rule, 1836-1845.” Legislative Studies Quarterly 32: 33-58. Poole, Keith T., and Howard Rosenthal. 1997. Congress: A Political-Economic History of Roll Call Voting. New York: Oxford University Press. Price, H. Douglas. 1975. “Congress and the Evolution of Legislative ‘Professionalism.’” In Congress in Change: Evolution and Reform, ed. Normal J. Ornstein. New York: Praeger Publishers. Rohde, David W. 1979. “Risk-Bearing and Progressive Ambition: The Case of the United States House of Representatives.” American Journal of Political Science 23: 1-26. Schlesinger, Joseph A. 1966. Ambition and Politics: Political Careers in the United States. Chicago: Rand McNally. Van Buren, Martin. 1840. Message From the President of the United States. House of Representatives Document No. 2. Wilentz, Sean. 2005. The Rise of American Democracy, Jefferson to Lincoln. New York: W.W. Norton. Yoshinaka, Antoine. 2005. “House Party Switchers and Committee Assignments: Who Gets ‘What, When, How?’” Legislative Studies Quarterly 30: 391-406.

19

20

21

Table 1. Negative Binomial Regression Estimates for House Bill Introductions 28th to 32nd Congresses

33rd to 35th Congresses

LocallyOriented

General Public

LocallyOriented

General Public

Margin in Last Election

0.001 (0.003)

-0.0003 (0.003)

-0.002 (0.003)

0.0005 (0.004)

Terms in Office

-0.004 (0.05)

0.19*** (0.04)

-0.11+ (0.06)

0.21** (0.07)

Opponents in Last Election

-0.0002 (0.07)

0.04 (0.06)

0.22* (0.11)

0.19+ (0.10)

Years Since State’s Admission

-0.04*** (0.006)

-0.02*** (0.005)

-0.03*** (0.005)

-0.02** (0.006)

DW-Nominate First Dimension

-0.43 (0.29)

-0.64* (0.29)

0.26 (0.28)

-0.31 (0.29)

DW-Nominate Second Dimension

0.46* (0.22)

0.09 (0.24)

0.37+ (0.19)

0.44* (0.22)

Member of Majority Party

-0.26+ (0.14)

-0.20 (0.15)

0.10 (0.13)

0.14 (0.16)

New England

-0.20 (0.63)

-0.88 (0.61)

-1.33*** (0.40)

-0.42 (0.42)

Upper Midwest

0.54 (0.35)

0.06 (0.44)

-0.26 (0.31)

0.02 (0.32)

Mid Atlantic

-0.07 (0.39)

-0.08 (0.45)

-0.59* (0.28)

0.39 (0.31)

Plains

0.86* (0.38)

-0.18 (0.47)

-0.24 (0.37)

0.83* (0.36)

South

-0.16 (0.27)

-0.48 (0.34)

-0.01 (0.31)

0.07 (0.32)

Pacific

0.55 (0.54)

0.31 (0.73)

-0.59 (0.54)

-1.11 (0.68)

Roll Call Attendance Rate

2.57*** (0.64)

2.12** (0.66)

1.77*** (0.50)

1.43* (0.67)

22

Table 1 (Continued) -1.66* (0.68)

-2.27*** (0.66)

0.008 (0.54)

-2.03** (0.71)

Log-Likelihood

-671.27

-672.90

-813.02

-571.17

Wald P2

167.52

80.31

140.28

91.26

"

1.05 (0.28)

1.37 (0.40)

1.29 (0.20)

1.29 (0.27)

Constant

N

1109

696

Notes: Hubert/White/sandwich robust standard errors, clustered on legislator, in parentheses. For the region dummy variables, border states are used as the base category. + p < 0.10 * p < 0.05 ** p < 0.01 *** p < 0.001 on a two-tailed t-test

23

Notes

1. Aldrich and Bianco also describe two versions of a game involving two candidates within a district each choosing a party affiliation to run against each other. Because the strategic problem illustrated in the game is not central to the discussion of the 1850s realignment by either Aldrich (1995) or myself, I do not consider it here. 2. Aldrich and Bianco treat candidates as pure office seekers. Although this is a strong assumption for the politics of the 1850s, let alone the contemporary period, it is a useful simplifying assumption to make the calculation of expected utilities more tractable. As such, I will not quibble with it here. Policy-based considerations did affect party choices during the realignment, and therefore did narrow the range of possible affiliations considered by most office seekers. Even so, the multiple salient policy dimensions dividing factions made the choices complex. Consequently, although candidates in the Antebellum period could not be characterized as pure office seekers, that assumption does approximate the flexibility they typically faced given the multiple coalition opportunities available to them. 3. In only one state – Massachusetts – were consecutive Whig governors elected from 1852 on. All other states saw only single-term Whigs serve as chief executive. 4. Data on bill introductions were collected from entries in the House Journal for the 25th through 35th congresses. 5. All discussion and analyses of bill introductions includes only full members of the U.S. House of Representatives and excludes delegates. 6. The Benton Democrats were named for Thomas H. Benton, an ally of Andrew Jackson and Martin Van Buren who opposed the Kansas-Nebraska Act and the repeal of the Missouri Compromise. In 1854, around the time of Kennett’s August election, Benton was widely thought to be a leading Democratic candidate for president in 1856 and someone who might draw northern anti-Nebraska votes across party lines (Gienapp 1987; Holt 1999; Wilentz 2005). In fact, it was Benton himself, seeking reelection, whom Kennett defeated. Failing to win back his House seat, and soon after also failing an attempt to return to the Senate, Benton conceded on the Nebraska issue, and with it all hope of standing for the presidency. In contrast to most other parts of the country where divisions among Whigs opened opportunities for Democrats, Know Nothings and Republicans, in Missouri a split between Democratic factions allowed Whigs to win six out of seven House seats. This illustrates why the broad approach I take to analyzing legislative responses to electoral uncertainty – including the divisions of Democrats and not just of Whigs, and including the many instances when candidates did not in fact switch parties – is appropriate. Partisan instability took many forms across states, though broadly driven by the same issues. 7. The split among California Democrats centered around David Broderick from New York, who led a faction of northern-leaning Democrats in the state against those who favored the South. 24

Seeing an opportunity to win California’s two House seats, the Whigs despite similar divisions of their own sought to remain united. The plan failed when the state Whig convention favored candidates from the southern-leaning wing. This led some Whigs to consider opting for the Broderick Democrats instead. Meanwhile, the Know Nothings of California were of a peculiarly mild flavor, actually drawing Catholics into their ranks and siding with the Anti-Brodericks. From this mix of shifting and ambiguous alliances, the Anti-Broderick candidates prevailed (Holt 1999). The strategic problem therefore mirrored Missouri in some respects, but with a different electoral outcome. 8. Here again, local politics created a peculiar coalition. Due to the electoral importance of German Protestant immigrants, the Know Nothings confined their appeals to anti-Catholicism rather than anti-immigrants generally. Also, due to the Germans’ opposition to the Maine Law, the People’s coalition did not emphasize temperance unlike Know Nothings in most other states. Due to tensions with the Free Soilers, Whigs insisted that an anti-slavery plank be narrowly construed. Finally, and most important for our purposes, in order to underline that this was just a temporary coalition of convenience, Whigs blocked an attempt to create a state committee, which would have lent greater legitimacy to a new People’s (or even Republican) Party. In their view, this arrangement should have little life after the 1854 elections (Anbinder 1992; Holt 1999). In fact, this coalition became the basis for a new and more permanent anti-Democratic alternative, though one led by a new Republican Party rather than by the Whigs. 9. These figures include only those bills for which leave was granted. When leave was not granted, a published record of the text of the proposed bill is not available, preventing a determination of whether or not it had locally-oriented content. From the 25 th Congress to the 35th, a total of 41 proposed bills were not granted leave for introduction. 10. Hinds (1907) mentions only three rulings of even tangential relevance to bill introductions in the 31st Congress: That the instructions in a motion to commit must be germane to the underlying bill (§5529 n. 4), that committees can report only on matters that fall within their jurisdictions (§4355), and that a point of order is not permitted if a committee does not follow instructions issued to it in a resolution by the House (§4689). Hinds list none of relevance in the 32nd. Nothing in the public record suggests that any of these rulings precipitated the increase in legislative activity in the 33rd Congress. 11. Bills are defined as private if they are to provide relief for some named individual(s). Local and state bills are those public bills where the benefit is directed to a named locality or state(s); it is coded as “state” if one or more states are named but no locality is specified. General public bills are all those that do not fall into one of the other three categories. 12. Economic crisis does not appear to explain the change. The country enjoyed a boom from 1849 to 1854, and not until 1857 did it suffer a widespread economic panic (Holt 1983, 110). Secular and more gradual changes like immigration and the expansion of the railroad do not explain the abrupt nature of the change in legislative behavior, either.

25

13. The first assumption is not without some dispute in the literature. Most accounts of the development of Congress identify career-minded behavior by representatives arising in the late 19th century (Price 1975; Kernell 1977; Brady et al. 1999). Others have found instances of accountability (Bianco et al. 1996; Carson and Engstrom 2005; Meinke 2007) and responsiveness (Baughman 2006; Baughman 2007) by representatives, and of attention to constituency demands (Jensen 2003). Rather than engage those controversies directly, echoing Aldrich I will use this as my starting point as well. 14. For these and all subsequent analyses, “local bills” includes all bills introduced by representatives which were coded as private, local or state. Excluded are all general public bills. 15. The Upper Midwest is defined as Illinois, Indiana, Michigan, Ohio and Wisconsin. 16. Note that this calculation of attendance does not distinguish between absence for political reasons, especially the disappearing quorum (Binder 1997), or for apolitical reasons like health or travel delays. This suggests that an indirect effect for those members choosing not to attend the House to express opposition to the majority also were foregoing the opportunity to engage in other legislative activities like offering bills and resolutions. Most members participating in a disappearing quorum did so while remaining at the Capitol, and thus fully able to participate, but an unknown number of members may have incurred additional costs to their maneuver as well by leaving the Capitol altogether. As a result, this is an imperfect measure of House attendance for the purposes of bill introduction, but one that approximates the variation of opportunity. 17. For Maine, 1820 is used for calculating the state’s age. 18. Cooper and Young (1989) calculate that more than 430 leave bills were introduced in the 36th Congress. Their count includes bills brought by both members and delegates, but even excluding the delegate legislation the total is almost certainly much higher than the 333 in the 35th. Their data suggest that the increase may have been driven by locally-oriented legislation. I find that 56 private bills were introduced by leave in the 35 th, but they count more than 80 in the 36th. They also find more than 220 public bills on constituency topics in the 36 th, up from the 170 in the 35th, although their coding of constituency legislation differs from mine. Together it means the proportion of bills that were of a general public nature was no higher than the 32% in the 35th, and probably lower.

26