Public Opinion, the Congressional Policy Agenda, and ... - SSRN papers

Public Opinion, the Congressional Policy Agenda, and the Limits of New Deal Liberalism, 1936-1945 Eric Schickler University of California, Berkeley

August 2009

PRELIMINARY DRAFT

Paper prepared for the Annual Meeting of the American Political Science Association, September 2-6, 2009, Toronto, Canada. An earlier version of this paper was presented at the Congress and History Conference, University of Virginia, May 2009. I thank Devin Caughey, Nicole Fox, Adrienne Hosek, Morris Levy, Eleanor Powell, Ruth Bloch Rubin, and Ian Yohai for their valuable research assistance. This paper draws in part on collaborative work with Brian Feinstein and Kathryn Pearson. I thank Richard Bensel, Erik Engstrom, Ira Katznelson, David Mayhew, Kathleen Frydl, Sid Milkis, Vesla Weaver, Elisabeth Sanders, Greg Wawro, Nolan McCarty, Taeku Lee, Irene Bloemraad, Cybelle Fox, Jas Sekhon, Van Houweling, Rob Mickey, and Adam Berinsky for helpful comments and suggestions.

1

The December 1936 Gallup Poll included the following question: “Is the Republican party dead?” Just over 27% of the respondents agreed that the GOP was, in fact, dead following Franklin Roosevelt’s landslide defeat of Alf Landon and Democrats’ sweeping victory in the 1936 congressional elections. The new Senate that convened in January 1937 included 16 Republicans; the House had 89 Republicans, as compared with 334 Democrats. Republicans’ dismal electoral fortunes were just one sign of the transformation in the American political landscape. Over the previous two years, Roosevelt’s “Second New Deal” program had arguably been prodded, shaped, and threatened more by attacks emanating from the left – and populist insurgencies that defy clear ideological labels – than from conservative Republican resistance.1 The most famous such challenge came with the launching of Huey Long’s “Share our Wealth” movement in 1934, but amidst all of the ferment of 1934-36 and Roosevelt’s decisive election victory, the prospects for liberal programmatic innovation seemed particularly favorable. Political observers at the time might well have been surprised to learn that the American welfare state that emerged from the New Deal era would turn out to be a laggard when compared to its more social democratic European counterparts. Yet it was during the 75th Congress of 1937-39 that a coalition of conservative southern Democrats and conservative northern Republicans began to emerge within Congress. While the first famous example of conservative cooperation occurred during the fight against FDR’s court-packing plan, southerners were by no means the only (and perhaps even the most important) Democratic foes of the bill (see Patterson 1967; Alsop and Catledge 1938). But by the time of the lead-up to the 1938 election – with Roosevelt’s purge effort, which mainly targeted dissident southerners – it had become evident that southern conservatives constituted the main obstacle to liberal advances within the Democratic party (Patterson 1967). Democrats

1

Conservative resistance did have an impact on some specific New Deal proposals – such as the wealth tax and “death sentence” provision of the Public Utilities Holding Company Bill – but much of this resistance came from nonsouthern Democrats who were not fully on board with Roosevelt (see Patterson 1967).

2

entered the 1938 midterms weakened by intraparty infighting and the brutal recession of 193738. Thus, a mere two years after their crushing defeat in the 1936 elections, Republicans began an electoral comeback by gaining 81 House seats and 7 Senate seats. This, in turn, greatly strengthened conservative prospects in the ensuing 76th Congress (1939-41) and set the stage for a long period in which conservatives and liberals would have rough parity in the U.S. Congress. Scholarship on the conservative coalition and the limits of New Deal liberalism more generally has largely focused on elite maneuvering – either in Congress or among supporters of alternative visions of economic recovery – and, appropriately, on the distinctive role of the south in America’s political economy.2 Some of these studies have noted that a rise in “popular conservatism” helped put liberals on the defensive in the late 1930s or early 1940s (e.g. Brinkley 1995), but scholars have tended to rely upon just a handful of polls and have not specified the contours of this shift. For example, how widespread was the revival in popular conservatism? Was it confined to a few issues or did it encompass many of the leading issues of the day? Was the rise in “popular conservatism” limited to particular geographic, class, religious, or ethnic groups, or did it span broadly across the American electorate? What was the relationship between the recession of 1937-38 and Democrats’ declining fortunes in 1938 and beyond?3 This paper is an initial piece of a broader project addressing the interplay of public opinion and congressional politics in shaping the consolidation and limitations of New Deal liberalism in the late 1930s and 1940s. Congress is, first and foremost, a representative 2

See Farhang and Katznelson (2005) and Katznelson et al (1993) on the distinctive role of the south. See Brinkley (1995) for a classic study of the dynamics of elite debate concerning the meaning of New Deal liberalism in this era. See also Patterson (1967), Alston and Ferrie (1999), Weir (2005), Jeffries (1990), Reiter (2001), Porter (1980), and Zelizer (2000) on related themes. Note that a more complete discussion of the relevant literature will be included in the next iteration of this paper. 3 Polenberg (1975) draws upon a handful of polls in arguing that the mass public had tired of the New Deal by the late 1930s. He argues that “the New Deal declined after 1937 because most Americans did not want to extend it much further” (1975: 255). But this leaves unanswered the contours and determinants of this mass shift; on which issues did the mass public move to the right? Which New Deal policies remained popular? Which subsets of the public moved against further extensions of the New Deal?

3

institution mediating the translation of public preferences into policy outcomes (Katznelson and Lapinski 2006). While there are many forces that make this translation far from perfect, it is nonetheless worth honing in on the interplay between public opinion and the conservative resurgence. In particular, the striking revival of Republican electoral fortunes in the north starting in 1938 and continuing into the 1940s is in part what gave congressional conservatives a majority on the House floor on many issues. Indeed, in each House election from 1938 to 1952, Republicans won a greater share of the popular vote outside the south than did Democrats (Rusk 2001: 234).4 Furthermore, conservative Republicans and southern Democrats were not just passive observers (and beneficiaries) of popular conservatism. Instead, they repeatedly teamed up to spur investigations that were intended to shape public opinion towards New Deal liberalism. A key feature of Congress as a representative institution is that its members are themselves participants in a “public sphere,” attempting to mold mass opinion (Mayhew 2000). As such, drawing linkages between the mass public and congressional politics ought to be viewed as a two-way street, rather than just as a bottom-up or top-down process. Fortunately, there is a great deal of data concerning the public’s views during this time. For one, this period coincides with the onset of the presidential use of polling data. As Eisinger (2003) and Eisinger and Brown (1998) describe, political surveys gained a respected place in national politics beginning with the Democratic National Committee’s employment of the pollster Emil Hurja to assist Roosevelt’s presidential campaign and better allocate patronage jobs. Later in his administration, Roosevelt unofficially consulted with Hadley Cantril, who commissioned and interpreted polls on behalf of the administration.5 Moreover, starting in the mid-1930s, polling companies surveyed the public about important issues on a monthly basis (see Converse 1987; Igo 2007). Together, roughly 450 national surveys were 4

By contrast, in 1936, Democrats won the popular vote in nonsouthern House elections by over 4 million votes. 5 The FDR Presidential Library in Hyde Park, NY contains a plethora of largely unstudied material devoted to the work of Hurja and Cantril.

4

conducted in 1936-45 by George Gallup’s American Institute of Public Opinion (AIPO); Elmo Roper’s polling firm; Cantril’s Office of Public Opinion Research (OPOR); and the National Opinion Research Council (NORC). The surveys include numerous questions probing respondents’ domestic and foreign policy attitudes. For example, most Gallup polls asked about presidential vote intention and/or recent presidential vote choice, with little variation in question wording. In addition, Congressional vote was included in about one quarter of the polls. Presidential approval was included in at least 88 polls from 1936-45, again with little variation in wording. Gallup also asked a party identification question at least 39 times from 1937 to 1945. There are many items probing attitudes towards labor unions, government regulation of business, taxation, redistribution, social security, relief policy, civil rights, and a variety of other issues. The polls attracted considerable attention at the time. George Gallup reported results from his polls in a regular column syndicated in over 100 newspapers by 1940 (Igo 2007). Fortune Magazine sponsored many of Elmo Roper’s surveys and presented a detailed summary of the results; Roper also popularized the results of his surveys with a syndicated column starting in the early 1940s. News coverage often referred to poll results. Members of Congress were sufficiently concerned about the implications of polling that there were several resolutions to investigate polling, and following pollsters’ underestimate of Roosevelt’s vote share in the 1944 election, a Senate committee even held hearings to explore the possibility that pollsters were biased against the Democratic President (Converse 1987). An early survey of 96 members of Congress found that while most denied relying upon polls in forming their positions, 70% believed that other elected officials look to polls, at least in part, in staking out their positions. The pollsters themselves wrote newspaper articles defending their profession against charges of bias and against the view that polls give rise to self-fulfilling bandwagon effects (see, e.g., Gallup 1939). All of this suggests that the burgeoning polling enterprise offers not just a potential window into what the public was thinking, but also had become one

5

of the sources of information looked to by both politicians and other elites in crafting their political strategies.6 While a handful of studies have made systematic use of some of this data (see Baum and Kernell 2001; Caldeira 1987; Schlozman and Verba 1979; Verba and Schlozman 1977; Weatherford and Sereyev 2000; Norpoth et al. 2009), scholars have generally been reluctant to draw upon them for two reasons. First, the data files themselves are generally in poor shape. The data contain numerous miscodings and other errors. In addition, some codebooks do not include the codes necessary to decipher important questions, such as the respondent’s occupation. Second, the data were collected using quota-controlled sampling techniques that have since been largely discredited. In quota sampling, pollsters sought to interview certain predetermined proportions of people from particular segments of the population (see Berinsky 2006 for a description of the quota sampling practices). While some pollsters used quotas in seeking a descriptively representative group of citizens (Roper 1940), others designed quotas to produce sample proportions that differed systematically from the population. George Gallup was most interested in predicting elections, so he drew samples to represent each population segment in proportion to the votes it usually cast in elections. Because Southerners, African Americans, and women turned out at low rates in this period, these groups were deliberately underrepresented in opinion polls. For example, the 1940 Census found that 50 percent of the U.S. population was women, 10 percent was African American, and 31 percent lived in the South. By contrast, a December, 1940 Gallup poll included only 34 percent women, 3 percent African Americans, and 13 percent Southerners.7 Thus, the Gallup data that scholars use to represent the voice of the mass public, in fact, comes from a skewed sample of that public. 6

Polling may have had particularly complicated implications for southern politicians. In a political system that had substantial authoritarian features yet also claimed to be democratic and representative, polls may have posed a particular challenge for political elites. Polls had the potential to construct a “public” in the south that expressed views distinct from those of the political elites who claimed to speak for the region. A key question for future research is how southern elites thought about and responded to the onset of opinion polling in the region. I thank Richard Bensel for suggesting this line of inquiry. 7 These figures are typical of the polls we have examined through the early 1940s. By the mid-1940s, however, Gallup adjusted his gender quotas to interview equal numbers of men and women. This change in the composition of the sample makes it difficult to track real changes in opinion over time.

6

The practice of quota sampling also introduced unintended distortions. Apart from having to fulfill certain demographic quotas, interviewers were given much discretion to select particular citizens to interview. Since interviewers preferred to work in safer areas and tended to survey approachable respondents, the “public” they interviewed often differed markedly from the public writ large. For example, the 1940 census indicated that about 10 percent of the population had at least some college education, while almost 30 percent of a typical 1940 Gallup sample had attended college. Similarly, polls conducted by Gallup and Roper tended to include more “professionals” than identified by the Census. The skew in these variables is not surprising, given that education and occupation were not quota categories. It is likely that the highly-educated and professionals were more willing to be interviewed, and, as a result, comprise a disproportionately large share in these samples. With support from the National Science Foundation, Adam Berinsky and I have been able to make the data suitable for analysis. We have recoded the datasets, which has involved ferreting out and correcting many errors and inconsistencies that we have identified. We have also put together a series of post-stratification weights that allow for better inferences about mass opinion during this period. In aggregate analysis, we weight the data on education levels, occupation, phone access and those quota category variables – such as gender and region – that can be matched to census data. If education levels were available, we created cell weights using education, gender, region, and race.8 If the survey did not contain a measure of education, we created cell weights using a dummy variable for “professional” occupation, gender, region, and race. If neither education nor a reliable occupation variable was available, we created raking weights based on the marginals of gender, region, phone access, and race. The necessary population counts for the 1940 census are available from the Integrated Public Use Microdata Series (Ruggles et al. 2004) and aggregate census records of phone access. 8

We included race as a separate weighting category for those surveys that had at least 20 AfricanAmericans in the sample. When race was used as a weighting variable for the purposes of creating cell weights, whites were weighted using the full stratification table (gender by region by occupation/education), while blacks were only weighted on the basis of gender due to small sample sizes.

7

Even when weighting makes only a modest difference in conclusions, it nonetheless provides more confidence that our estimates are not attributable to problematic sample design (see Berinsky and Schickler 2006; Berinsky, Powell, Schickler, and Yohai 2008). The recodes and weights will be made publicly available by the end of 2009. While the recoding is nearly complete, I have only begun to delve into the analysis of the data itself. As a result, this paper will focus on just two issues – labor policy and civil rights – and the treatment of both issues will be tentative, with the connections to action in Congress less developed than I had initially intended. At this point, the analysis will be largely descriptive, providing an overview of the terrain, rather than a causal analysis. The broader goals of the project include understanding why certain parts of the country increasingly elected Republicans to Congress starting in 1938, and whether the southern Democratic-Republican coalition was rooted in shared views at the mass level or instead reflected an “unholy alliance” that distorted the translation of mass preferences into legislative behavior. Answering such questions, however, will require close examination of other issues beyond the two addressed in this paper, along with careful analysis of vote choice.9

Labor and Civil Rights Politics The centrality of labor and civil rights to both the accomplishments and limitations of New Deal liberalism make it sensible to start with those two issues. As Farhang and Katznelson (2005) emphasize, the institutionalization of social democracy and corporatist representation in the United States would have required a labor movement that had broad strength and was widely accepted as a legitimate governing partner (see also Lichtenstein 1989, 2002).10 Instead, the labor movement that emerged from the 1940s – particularly with 9

The difficulties are compounded by changes in question wording and content over time and by the absence of panel studies. 10 Farhang and Katznelson (2005: 6) argue that the New Deal labor policies of the 1930s opened “prospects for the growth of organized labor, the emergence of corporatist interest representation, and national policies tilted in the direction of social democracy.” The “far-reaching new openness” (2005: 5) evident in the 1930s, however, gave way to a much more constricted version of liberalism in the 1940s due to congressional southern Democrats’ increased hostility to the labor movement.

8

the passage of Taft-Hartley in 1947 – was much more limited in its ambitions and occupied a more tenuous place in the post-war political economy.11 Farhang and Katznelson attribute labor’s troubles to the peculiar structure of the Democratic coalition in Congress and nationally. They characterize the Democratic Party as a hybrid of an “immigrant, labororiented, big-city, machine nonsouthern wing that looked very much like a social democratic party” and a native, rural, white supremacist southern wing that represented no blacks and just a minority of the whole south’s white population (2005: 2). While southern Democrats had gingerly backed most New Deal labor legislation in the mid-1930s – focusing primarily on seeking occupational exclusions for agricultural and domestic workers – they shifted ground in the 1940s. Wartime labor shortages and CIO unionization drives led southern elites to see New Deal labor policy as a threat to their system of racial apartheid, low-wage agriculture, and (emergent) low-wage industrialization. As a result, they joined with northern Republicans to attack labor unions and the National Labor Relations Board (NLRB), authoring a series of initiatives that put labor on the defensive and generated a more cramped and limited American labor movement.12 With labor forced into the position of a mere interest group, rather than a broad, class-based movement fusing demands for economic and racial justice, liberal dreams of a cradle-to-grave social welfare system and vigorous government intervention to limit corporate power had little or no chance of passage (see also Lichtenstein 2002). Consistent with Farhang and Katznelson’s argument, labor policy was the single issue on which conservative southern Democrats and nonsouthern Republicans united most frequently and aggressively in the late 1930s. Schickler and Pearson (2009) identify 44 11

See Plotke (1996) for a more positive assessment of labor’s place in the emergent New Deal order. As Orren and Skowronek (1998-99) point out, there is a sense in which the literature on the 1940s has had a “glass half full” / “glass half empty” aspect: the 1933-50 period encompassed both major triumphs and disappointments for labor-oriented liberalism. The political terrain had shifted with the Wagner Act and related victories; the battle now was over the meaning of those victories for future politics. In other words, to say that labor’s place in the postwar political economy did not match the aspirations of New Deal labororiented reformers of the late 1930s does not challenge the notion that the Wagner Act constituted a significant political transformation (see Orren and Skowronek 1998-99). 12 See Bensel (1984) for an early, innovative analysis of the impact of sectional political economy on congressional decision-making with respect to labor and other policy domains.

9

conservative initiatives in the Democratic Congresses from 1937-52 that were reported to the House floor by the Rules Committee over the opposition of the Democratic administration and key party leaders. Of these, 15 primarily involved labor-management relations, and several others involved farm labor (e.g. draft deferments for farm labor; incorporating farm labor costs into parity prices etc). Examples include a 1937 resolution to investigate the sit-down strikes,13 Howard Smith’s (D-VA) successful resolution to investigate the NLRB in 1939, and the series of anti-strike bills proposed by Smith and his conservative allies starting in 1941 (which culminated in enactment of the Smith-Connally Act of 1943 and Truman’s veto of the Case anti-strike bill of 1946).14 Southern Democrats and Republicans clearly viewed the status quo labor policies adopted in the 1930s as unacceptable, and they joined forces to overcome northern Democratic and union opposition to reversing those policies. If Republicans and Southern Democrats joined forces most visibly – and often – to push an aggressive agenda on labor policy, the relationship between the two groups on civil rights policy in the late 1930s and 1940s is more difficult to tease out. In civil rights, the main challenge for liberals was to force votes on bills to change the status quo (rather than to defend an existing set of labor policies, as with the Wagner Act regime). Eventually, of course, southern conservative Democrats would join forces with Republicans on civil rights, but prevailing accounts suggest that this alliance did not really take shape until the 1960s (Carmines and Stimson 1989). Recent scholarship, however, has suggested that outside the south, Republican state parties and members of Congress had already begun to adopt a less racially liberal stance than same-state Democrats by the mid-1940s (see Chen 2006, 2007, 2009; Karol 1999, 2009; Lee 2002; Feinstein and Schickler 2008; Schickler, Pearson, and Feinstein 2008; Chen, Mickey, and Van Houweling 2008; see also Jenkins et al. 2009). An important question is whether this elite shift corresponded to changes in mass opinion on civil 13

The resolution was rejected on the House floor despite the support of nearly all of the Republicans and half of the southern Democrats. 14 Roosevelt’s veto of Smith-Connally was overridden, while the Case bill veto was upheld (but provisions of the bill were incorporated in Taft-Hartley the following year).

10

rights issues. At the mass level, did New Deal liberalism come to incorporate racial liberalism in the 1930s-40s or did economic liberalism only become joined with racial liberalism much later, after the rise of the civil rights movement and the crystallization of elite positions? Were Republican elites ignoring – or responding to – their constituents as they began edging away from racial liberalism in the 1940s? Together, the issues of labor and race define crucial contours of the New Deal liberalism that emerged during the 1930s-40s. While the findings here are preliminary, I argue that the conservative resurgence in Congress was at least in part rooted in a relatively broadlybased reaction against perceived labor union radicalism that emerged in 1937 and persisted with great consistency through 1947.15 Furthermore, Republicans’ willingness to cooperate with southerners in blocking civil rights measures in the mid-1940s coincided with a shift at the mass level in which nonsouthern Republican partisans displayed markedly less support for civil rights policies than did nonsouthern Democrats. Economic and racial liberalism became aligned at the mass level by the end of the 1930s, but liberals of all stripes were fighting on a political landscape defined in part by broadly-based popular anger at arguably the key coalitional player essential for liberal success: organized labor.16

Public Opinion Towards Labor Policy The passage of the Wagner Act in 1935 signaled a dramatic shift in the relationship between the national government and labor unions. After decades in which the predominant stance of the government towards unions was one of indifference at (occasional) best and hostility at worst, the Wagner Act provided a far more favorable legal and institutional 15

See Calkins (1952) for an insightful account of Robert Taft’s 1950 reelection campaign, which emphasizes the difficulties faced by Democrats given their simultaneous reliance on CIO resources and the organization’s unpopularity with many nonsouthern swing voters. 16 As noted below, I do not argue that nonsouthern Democrats were, on the whole, intense in their support for civil rights in the late 1930s and 1940s. But they were to the left of their GOP counterparts, with economically-liberal Democrats particularly likely to support civil rights policies. The rise of the grassroots civil rights movement was a necessary precondition for this nascent Democratic liberalism on civil rights to become a potent political force – but the early mass alignments make it clear that the Democratic Party was the more likely “home” once civil rights activists forced the issue to the top of the agenda.

11

environment for union organization. While some on the left have argued that the Wagner Act tamed a labor movement that otherwise might have used collective action to force a more dramatic restructuring of the economy (Gordon 1994; Tomlins 1985; Goldfield 1989), the Act created a set of rules that were widely viewed at the time as favoring workers.17 The Act defined a series of unfair labor practices by management and created the NLRB, which was largely staffed by liberals, who at least in the early years, generally sided with workers, particularly the emergent CIO industrial unions. As Lichtenstein (2002) points out, rather than giving rise to a tame, voluntary, privatized system of collective bargaining, the labor policy regime that emerged with the Wagner Act and dominated American politics from 1935-47 was a highly politicized process that routinely put unions, capitalists, and government labor experts in the Oval Office and other government venues. Given the generally pro-labor occupants of the White House, the group most dissatisfied with this arrangement was business. In 1939, when a Roper poll sampled business executives, asking which New Deal program they disliked the most, the resounding answer was the Wagner Act (Amenta et al. 1998). The first opinion polls from 1936 include very few items on labor policy. In some ways, the absence of attention to unions is itself telling. For example, a January 1936 Roper poll asked respondents: “Do you believe that any of the following abuse their power? Bankers, Press, Radio, Pulpit, Veterans.” Labor unions were not even offered as a response option, suggesting that unions were not, in the immediate aftermath of passage of the Wagner Act, viewed as particularly powerful.18 Two years later, however, labor unions were listed as a response option when Roper asked “Which of these is most in need of reform?” Indeed, unions were by far the top choice: while 38% of respondents agreed that labor unions are “most in need of reform,” just 14% listed public utilities, 13% listed stock exchanges, and 4%

17

See Farhang and Katznelson (2005: 5-6) for a compelling critique of this strand of research. See also Zieger (1995). 18 The press narrowly beat out bankers as the group most associated with abusing power: 42% cited the press, as compared to 38% for bankers and 11% for veterans. Since we do not have individual-level data for the earliest Roper polls, these results are not weighted.

12

cited the Supreme Court.19 This theme became even more prevalent over the ensuing years. For example, in April 1944, Roper offered an open-ended question asking, “Are there any special groups that you feel had too much influence over Congress in the past year or so?” If the respondent answered yes, he or she was asked, “What groups?” Of the respondents identifying a group, 47% named labor unions (either in general or the CIO specifically), as compared to 14% that named a group associated with Wall Street, business, or finance.20 The same year, Gallup asked respondents: “which would you prefer to have influence the government more strongly – big business or labor unions?” (February 1944). Despite the use of the normally pejorative “big business,” a plurality of 41% preferred big business, as compared to 32.5% favoring labor unions. An overwhelming majority of southerners favored “big business” power and the nonsouth was split down the middle (see Table 4, item 7). What happened to lead a substantial portion of the public to view labor unions – rather than business groups – as a too-powerful influence in American politics? A key initial event was likely the wave of sit-down strikes that began in late December 1936. These strikes had notable successes – as the UAW, which was part of the recently-formed CIO, won recognition at both General Motors and Chrysler in early 1937.21 As noted above, congressional conservatives moved immediately to investigate the sit-down strikes. While the initial investigative resolution was rejected on the floor – due to the overwhelming Democratic majority in the House – the Dies Special Committee on Un-American Activities targeted the sit-down strikes in its first round of inquiries starting in 1938.22 Beyond attempting to link the strikes to communist influence, the Dies Committee sought to show that Michigan Democratic Governor Frank Murphy had tacitly supported the strikers. Murphy was defeated for reelection in 1938. Newspaper accounts at the time credited the Dies Committee investigation – and

19

We do not have individual-level data for this survey. As a result, these numbers are unweighted. Thirty-seven percent of the respondents identified a group. 21 The Supreme Court eventually ruled in February 1939 that sit-down strikes are illegal. 22 James F. Byrnes (D-SC) also unsuccessfully pushed a Senate investigation of the sit-down strikes (Patterson 1967). 20

13

more general anger over the sit-down strikes (fueled itself in part by negative press coverage) – for Murphy’s defeat (Goodman 1968; Wreszin 1974; see also Gallup 1938).23 The initial polls that indicate public ambivalence or hostility towards unions – and the CIO in particular – emerged during the sit-down strikes.24 Since the polls vary considerably in their focus and question wording, it is difficult to summarize their results in a simple way. As a first cut, Table 1 presents a series of Gallup poll items relating to the sit-down strikes. Rather than presenting detailed regional breakdowns, the table includes the overall weighted marginals, along with noting whether any regions differed substantially from the national numbers in either direction. For now, a “substantial” gap is defined as a difference of greater than 8 percentage points between a region’s margin and the national margin on an issue.25 The table also includes the breakdown among FDR and Landon voters in the nonsouth, and among all southerners.26 The latter comparison provides leverage for evaluating the south’s position relative both to the country as a whole and relative to the two partisan groups in the nonsouth.

23

Roosevelt appointed Murphy as Attorney General and, later, to the Supreme Court. As Attorney General, Murphy encouraged the Justice Department’s move towards civil rights liberalism (McMahon 2004). 24 As noted above, there were very few poll questions about unions prior to the sit-down strikes. The few exceptions are a July 1936 Roper survey, which asked: “Do you believe all wage earners should belong to a labor union?” According to Roper, 29% chose “all workers,” as compared to 8% for “most,” 23% for “some,” and 25% for “none.” We do not have the individual-level data for this survey, so these numbers are unweighted. The crosstabs presented by Roper suggest that factory workers favor all or most workers belonging to unions, as do the unemployed. The most hostile region is the Plains – just 18% favor all or most workers belonging to unions; the most friendly area is the Mountain States (68.5% favor all or most workers belonging to unions). A July 1936 Gallup survey asks: “Are you in favor of labor unions?” A health 76%-24% majority responded in favor, but again this is just before the period for which we have individual-level data. In later polls, weighting the responses tends to increase support on this item by 1 or 2 points. When the question is asked later, there is continued widespread support for unions, but the margin is smaller than in this first poll. The same survey also asked whether the respondent prefers separate craft unions or a single industrial union for all workers in an industry. Respondents preferred craft unions by a 59-41 margin. When the question was asked again in April 1937, respondents favored craft unions by a slightly greater, 64-36% margin (unweighted, so comparable to the July 1936 data). 25 This approach will be improved upon in future drafts (suggestions are welcome). Caution is required in interpreting the numbers for regions with fewer respondents, such as the Rocky Mountains and Pacific Coast (both of which typically had just 200 or so respondents in each survey). The northeast also generally had a relatively modest sample size. 26 Since party identification is not available in most of the early surveys, presidential vote choice is used in classifying Democrats and Republicans. For polls with party identification, the results are similar regardless of which measure is used.

14

Immediately after the sit-down strikes started, Gallup asked respondents: “In the current General Motors strike are your sympathies with the John L. Lewis group of striking employees, or with the employers?” In that January 1937 poll, the public was nearly evenly divided; the unweighted responses suggest a narrow plurality sympathized with GM, but the weighted numbers suggest that 37% favored the union, as compared to 32%, who favored GM, with the remainder uncertain (Table 1, item #1). The regional breakdown of support is striking in light of later developments: southern respondents supported the union at the same rate as nonsoutherners. Support for the strikers did not vary much across regions, except for the east central states, which were the least supportive (with a plurality favoring GM). These states – Michigan, Ohio, Indiana, and Illinois – were the ones most directly exposed to the strikes, of course. As the strife in Michigan wore on, opinion shifted decisively against the strikers. An identically worded question in February 1937 suggested that 42% of the public now favored GM, as compared to 32% favoring the strikers (Table 1, item #1).27 Once again, the south was similar to the rest of the country, splitting 42% to 29% in favor of GM. The only region now siding with the strikers was the Pacific Coast (34% to 30%), though the Northeast and Rocky Mountain states were also closely divided. The data also suggest a sharp class skew in the responses: professionals, business, and (to a slightly lesser extent) skilled workers opposed the union by a wide margin, while unskilled workers and the unemployed backed the strikers by a 15 point margin. Gallup also began polling respondents about their preferred policy response to the sitdown strikes in late February 1937. Initially, Gallup asked whether the respondent wanted “this state” to pass legislation “making the sit-down strikes illegal” (Gallup #71). The breakdown was 60% in favor and 28% opposed (see Table 1, item #2). Southern respondents were a bit more anti-strike than the rest of the country (65% in favor of banning the strikes, 27

The unweighted numbers – which are what politicians likely attended to – were even more one-sided: a 48% to 29% GM advantage.

15

with 22% opposed). But a healthy plurality in each region favored a ban on the strikes (e.g. 59-60% in the Northeast, Mid-Atlantic, and East Central states), with only the Rocky Mountains closely divided. When the same item was repeated in March 1937 and February 1939, the response distribution was similar. The 1939 survey shows the most lopsided margin in favor of banning the strikes (64%-22%). It is worth noting that in each case, a clear majority of Roosevelt voters in the nonsouth also favored making the strikes illegal; indeed, in the 1939 survey, nonsouthern Democrats appeared similar to their southern counterparts in their opposition to the strikes. From March to December 1937, Gallup also repeatedly asked about “using force” to remove the sit-down strikers (Table 1, item #3). The March survey revealed a slim majority in favor of using force (51% to 37%, with 12% Unsure). By December 1937, 64% favored using force, as compared to 23% opposed. Support for using force appeared strongest in the South and East Central states, with the Rocky Mountains and Mid-Atlantic states a bit less supportive. But even FDR voters in the nonsouth tended to support using force to remove the strikers (with the margin quite small in March and April, but substantial in December 1937). As one might expect, a clear majority of both AFL and CIO members opposed using force, as did a majority of unskilled workers and the unemployed (including non-union members in the latter two groups). But beyond those groups, hostility to the sit-down strikes appears to have been widespread. While we lack panel data to trace changes in individual opinion, those questions that were repeated over time suggest at least some deterioration in support for the strikes during 1937 (see, e.g, the item on using force). In addition, Gallup asked in June 1937 whether respondents’ attitude toward unions had changed during the past six months, and if so, in which direction. A little over half reported that their views had not changed, but of those whose views had shifted, those becoming more negative outnumbered the positive by more than a 2-1 margin (see Table 1, item 7).

16

Given the persistence of anti-union attitudes long after the sit-down strike wave ended, it would be wrong to credit the strikes with causing this backlash. Instead, the reaction to the sit-downs are better viewed as indicative of one of the obstacles confronting the labor movement over the long term. For all of the anger at business in the Depression years, movements that seemed to promise a fundamental reordering of authority relations faced considerable public skepticism.28 The mass public’s ambivalence towards unions provided a relatively permissive backdrop for corporations intent on fighting strikes and unions with threats, coercion, and other aggressive tactics. The CIO faced the challenge of harnessing the shopfloor energy and anger that gave rise to such innovations as the sit-down, without alienating the public as a whole (Zieger 1995). The breadth of the CIO’s agenda as it took shape in the late 1930s – incorporating civil rights and national health insurance, among other issues – accentuated this challenge. Thus, from the start, the CIO’s broad constituency and goals made it an inviting target for conservatives. From this perspective, the sit-down strikes are best viewed as just the first in a series of sparks for anti-labor drives. In addition to questions about the sit-down strikes, pollsters began asking respondents about their views towards the closed shop and union shop in the late 1930s. Table 2 suggests that a healthy majority of the public opposed both the closed and union shop from 1937 through 1945. This result holds both in questions that ask directly about these arrangements (e.g. “Are you in favor of the so-called 'closed shop'--that is, hiring only persons who are already members of the union to work in a factory or a mine?”) and questions that offer respondents a choice among the open, union, and closed shop. Opposition to the closed and union shop seems especially strong in the West Central Plains and in the South, though there is also a great deal of variability in the regions’ relative placement.29 As one might expect, Republican voters are overwhelmingly opposed to the union and closed shop, and southerners

28

Polls conducted amidst the 1936 election showed that most voters expected the Roosevelt administration to adopt a more moderate course over the next four years, and that a plurality favored such a course. 29 Again, this may in part be due to the small sample size in the Rockies, Northeast, and Pacific Coast.

17

often look more like the nonsouthern Republicans than like nonsouthern Democrats on this issue. Still, even nonsouthern Democrats are closely divided on the issue, with a majority seeming to oppose the closed and union shop but not willing to go so far as to support a constitutional amendment banning the closed shop (see Table 2, item #3). The constitutional amendment item is interesting because it anticipates a feature of Taft-Hartley. When asked in April 1943 of those respondents who had heard of the closed shop (61% of all respondents), a clear majority of southerners and nonsouthern Republicans favored the ban. The nonsouth as a whole was split nearly evenly on the question, however.30 Against this backdrop, the War Labor Board’s adoption of “maintenance of membership” clauses as a standard for wartime labor contracts can be seen as a significant boost to unions, providing a stable source of union dues even in the face of popular skepticism against such programs (see Table 2, item #4 on maintenance of membership in particular).31 Indeed, public concern about union power and tactics continued into the war years. For example, across a range of polls from 1941 to 1945, more than 70% of respondents supported banning strikes in war industries (see Table 3).32 In April 1944, 68% supported drafting strikers. Southern respondents, along with those in the Plains states (West North Central) were consistently among the most eager to ban strikes and punish strikers during the war, but even the most “pro-labor” regions had a majority in favor of banning strikes in war industries.

30

I need to explore divisions among different occupational groups more closely. A quick examination of the surveys suggests that union members supported the union or closed shop over the open shop, but that unskilled workers as a whole preferred the open shop. 31 See Atleson (1998) for a more pessimistic rendering of the War Labor Board’s impact on organized labor. While Atleson grants that maintenance-of-membership offered important benefits to unions, he charges that WLB decision-making processes and criteria gave rise to a more bureaucratic union movement that failed to capitalize on shopfloor militance. However, as Workman (2000) argues, the WLB was clearly seen by participants at the time as a union ally, and its decisions gave a much-needed boost to unions that were in a vulnerable position (see also Zieger 1995; Lichtenstein 2002). Given the political atmosphere, one can easily envision greater labor radicalism being greeted by violent suppression and draconian legislation, rather than setting the stage for a more vibrant labor movement. 32 Kornhauser (1946) rightly argues that this question is worded in a manner that may generate greater antilabor responses (since it does not mention alternative solutions to labor troubles other than an anti-strike law). One should not put too much stock in the absolute percentage opposing strikes or unions in any particular question, but the general tenor across the range of items – including open-ended questions – suggests the breadth of the popular anger towards unions.

18

Indeed, when asked in July 1945 about banning all strikes during the war, 79% of respondents agreed, including 74% of nonsouthern Roosevelt voters. While one might argue that questions about wartime strikes are particularly likely to promote anti-union responses, other items that did not reference strikes depict a similar public mood. For example, in March 1941, Gallup asked respondents: “Do you think labor union leaders are helping the national defense production program as much as they should?” (Table 3, item #4). Just 17% of respondents answered in the affirmative, as compared to 67% who said union leaders had not been helpful. Once again, even nonsouthern Roosevelt voters responded that union leaders had not done as much as they should (59% to 23%), though Republicans and southerners were even more negative. The same survey asked about “industrial (business) leaders,” with much different results: 49% responded that business leaders had done as much as they should, as compared to 32% who said that business leaders had not done so.33 It is worth emphasizing that a majority of Americans continued to support the right to strike in peacetime and were in favor of the right to organize unions. Still, the evident widespread anger over wartime strikes seems to have further eroded unions’ standing with the public. Thus, when Gallup asked in early May 1943 whether respondents’ views of unions had changed over the past year, 38% reported being less supportive, as compared to 5% who were more supportive than a year earlier (Table 4, item #1). The public dissatisfaction was reflected in letters to members of Congress, as well as the polls. Frank McNaughton, Time magazine’s Congress correspondent during the war, included several reports about the onslaught of anti-strike or anti-labor letters received by members from all sections of the country (see, e.g., McNaughton Papers, 2/27/42, 3/20/42). 33

Gallup varied the order of the two questions across the two forms of the survey. The response distributions were unaffected by the question order. The same survey asked whether employers should be able to refuse to hire “anyone because of union membership” (Form K) or “because of union activities” (Form T). A slim 46%-44% plurality agreed that employers should be able to refuse to hire union members. A slightly larger 49%-42% plurality agreed that employers should be able to refuse to hire anyone because of union activities.

19

Howard Smith and his conservative allies sought to capitalize on this perceived anti-labor tide in pushing anti-strike bills through the House (McNaughton Papers, 3/13/42). At the same time, congressional investigations of the war effort, coupled with sensationalistic press coverage, helped fuel the public anger. While Democratic leaders desperately sought to keep the lid on anti-labor measures that would challenge the party’s coalition – and potentially undermine the war effort, by alienating the vast majority of union members who were not involved in strikes – conservatives such as Smith viewed this as their opportunity to adopt restrictions that served their longer-term goal of curbing unions.

Interpreting the Labor Polls A key question is to what extent the public reaction against both the sit-down strikes and war strikes forced labor unions to temper their ambitions and accept a more limited place in American politics. Brinkley acknowledges that external pressures played an important role in encouraging unions to adopt the no-strike pledge during the war, and more generally, to moderate their demands for broad structural changes. But he emphasizes that internal splits and ideological shifts within the movement were critical in drawing unions away from a broader, reformist vision. In particular, Brinkley (1995: 225-26) observes that the landscape for labor might have looked a lot different – and potentially better in the long-term – had pragmatists such as Sidney Hillman, who were seeking to accommodate labor to the New Deal state and Democratic Party, lost out to more independent, aggressive leaders, such as John L. Lewis.34 There is little doubt that divisions within the labor movement – particularly the AFL – CIO split – were extremely damaging. But it is also worth noting the extent to which Lewis’s aggressiveness, when coupled with extraordinarily negative press coverage, made him a 34

Hillman had been one of the founders of the CIO and led the Amalgamated Clothing Workers of America. He was also the first chair of the CIO-PAC (founded in 1942). Hillman was a key link between the Roosevelt administration and the CIO. Lewis had also been a co-founder of the CIO, but withdrew his United Mine Workers from the organization in 1942.

20

negative symbol even before World War II. An early example of this is from October 1937, when Roper asked “Can you think of any well-known man in business or industry of whom you approve? If yes, who? Can you think of any such well-known man of whom you disapprove? If yes, who?” In response to this open-ended question, John L. Lewis was easily the top individual selected for “most disapproved” – with over 30% selecting Lewis (as compared to 18% for Henry Ford, 7% for J.P. Morgan, and 6% for Andrew Mellon).35 In April 1940, before the strike wave in the lead-up to the war, a plurality of 45%-33% believed that Lewis had been “harmful” rather than “helpful” to labor.36 In November 1941, amidst a major coal strike, Gallup asked: “What is your opinion of John L. Lewis?” A full 69% of respondents responded negatively to Lewis (with open-ended responses ranging from “he should be hanged” to “he’s a radical, communist”), as compared to positive responses from 12%, with 20% unsure or neutral.37 Finally, a November 1943 Roper poll asked, “Are there any prominent individuals in this country who you feel might be harmful to the future of the country unless they are curbed?” Of the 52% of respondents who named someone, Lewis was overwhelmingly the top answer (71%, as compared to 5.5% for second-place finisher Roosevelt and 5% for Henry Wallace; see Cantril and Strunk 1951: 561). Even after Lewis withdrew from the CIO, negative views of the organization persisted. A Gallup poll in May 1944 asked, “If the CIO union supported a candidate, would you be more likely to vote for that candidate, or against him?” A majority of respondents said that they would be less likely to vote for the candidate (52%), as compared to 11% who said an endorsement would make them more positive toward the candidate (Table 4, item #5). Even

35

Fewer than 1% listed Lewis as the “most approved” figure in business or industry. We do not have the individual-level data for this survey, so the results are unweighted. 36 The question read: “Which of these people do you feel have been on the whole helpful to labor and which harmful … John L. Lewis.” Source: Roper Fortune Poll, April 1940. (We have the individual-level data for this survey but have not yet implemented the weights). 37 The “negative” count excludes a handful of respondents who said that Lewis is “out to get Roosevelt” (since this could presumably be liberals angry at Lewis for endorsing Willkie in 1940). In each region, at least 60% of the responses were negative. The positive share topped out at 17%.

21

nonsouthern Roosevelt voters said a CIO endorsement would make them less likely to vote for a candidate (41-15%).38 All of this suggests that the pragmatic adjustments of individuals such as Hillman and the UAW’s Walter Reuther may well have been necessary to avoid a far more draconian governmental response. Greater labor radicalism would have likely either been met by a repressive response from the administration and Congress or by the earlier election of a Republican president and majority in Congress.39 As Zieger’s definitive history of the CIO makes clear, the organization itself was, even in its heyday, in a relatively precarious position. The considerable shop-floor resentment and militancy that generated the sit-down strikes and other grassroots actions was often not coupled with a willingness of many workers to pay union dues (Zieger 1995). This weakness made good relations with the national government essential, even if it at times required tamping down on shop-floor demands. One way to view the opinion data is that a conservative, anti-union pulse entered the electorate in 1937 with the sit-down strikes, continued in 1939-41 during the run-up to the war, and, if anything, became more intense during the war itself. The anti-labor pressure was shared across regions, though it was particularly intense in the South and Plains states. While further analysis is required examining different occupational, ethnic, and urban-rural slices of the electorate, preliminary work suggests that even urban workers and northern Roosevelt voters were not entirely immune to anti-union appeals. At the same time, two aspects of the governmental response to this public sentiment seem noteworthy. First, conservative southern Democrats proved quite willing to work with Republicans (and allies in the press) to heighten 38

The same survey asked about a National Association of Manufacturers endorsement. While the response was still negative, it was less overwhelming than in the case of a CIO endorsement (36.5% negative, 21% positive). See Calkins (1952) on how this anti-CIO sentiment complicated Democratic efforts to draw on union assistance in campaigns in nonsouthern states. 39 Indeed, while it is tempting to blame the war for the fading of liberal hopes, one could argue that the impending war was likely a major reason Democrats held the White House in 1940. It would have been more difficult for FDR to justify a third-term campaign in the absence of the conflict, and Roosevelt had not left his party with a compelling successor. The polling data also suggests that increased concerns about the European conflict contributed to FDR’s support. See Norpoth et al (2009) for an analysis suggesting that Republican Tom Dewey would have likely won in 1944 if not for the war.

22

this public anger, linking unions to communists and other subversives, and pushing for legislative action to impose drastic restrictions on labor. These southern elites were acting in ways that broadly reflected the polling data in their region.40 Second, despite the wide popular support for governmental action against the unions during the war – both within the south and the nonsouth – the Roosevelt administration and nonsouthern Democrats in Congress proved largely successful in staving off or watering down the harshest attacks. The War Labor Board – along with the Roosevelt administration as a whole and congressional leaders – understood that harsh anti-labor legislation would undermine the central goal of maximum production during the war, and thus resisted the popular attacks on unions. The WLB issued decisions that were widely viewed at the time as generally favoring unions. The anti-labor bills passed by the House often found themselves buried in the Senate Labor Committee, which had a strong prounion majority during the war. The plant seizure provisions of the Smith-Connally Act – though enacted over Roosevelt’s veto –were actually turned against management when Roosevelt seized Montgomery Ward in 1944. Roosevelt’s move sparked an angry congressional response – including an investigation that Democratic leaders unsuccessfully sought to block (Schickler and Pearson 2009). In the aftermath of the war, when another strike wave sparked the anti-labor Case bill of 1946, Truman used his veto power to defend the existing policy regime.41 Thus, while congressional conservatives fueled the anti-labor pulse among the public, they were unable to bring it to fruition in 1937-46 with major policy changes.42 Ironically, 40

An important caveat is that this electorate – and much of the earliest Gallup polling data – largely excludes southern African Americans. Additional analysis will be required to tease out how much of a gap this created between the full “public” and the section of the public represented by both polls and elected officials in the south. It is worth noting, however, that even low socio-economic status whites in the south shared in the anti-union views evident in the surveys. This may, in part, constitute a legacy of the failed textile strike of 1934. Alternatively, the anti-union sentiment may have been more intense in the south due to the union inroads during the war – and the sense that these inroads threatened both southern economic development and southern racial apartheid. 41 Truman had proposed a more balanced anti-strike bill – which also was opposed by organized labor, but was not as draconian as the Case bill. Polls suggested that the public did approve of the Case bill veto, suggesting that the public likely favored Truman’s more moderate proposal. 42 Smith-Connally did restrict political contributions by unions, but the formation of the CIO-PAC helped labor maneuver around these restrictions.

23

however, this conservative failure may have contributed to the widespread public sense that labor was too powerful in the 1940s (see Table 4). Given the structural features of American politics – and the intimate involvement of business leaders in shaping the war mobilization (Brinkley 1995) – it is difficult on its face to understand how the public could have identified labor, rather than business, as privileged with too much power during this period. But when one considers the simultaneous ability of conservatives to fuel popular outrage about unions and their failure to move policy substantially, this public sentiment may be easier to understand. From this standpoint, Republicans were able to run against labor “excesses” in the nonsouth up until they actually succeeded in passing the anti-union Taft-Hartley bill in 1947.43 The evidence that union members were more supportive than other workers of such policies as the union shop – and less supportive of laws restricting strikes – suggests the difficult strategic tradeoff facing union leaders. An aggressive campaign to organize additional industries and workers promised to bolster the number of pro-union voters and thus to keep the pressure on Democrats to withstand conservative attacks on the labor movement.44 With the onset of the war, there was the added tension between responding to shop-floor anger about working conditions and limited pay increases and yet not alienating the governmental leaders (and ultimately, general public), on whom the unions were nonetheless still dependent for favorable treatment. But the same tactics necessary to organize new workers and respond to shop-floor pressures may have reinforced the backlash that put unions on the political defensive. In that sense, unions may have been engaged in a race against time: they needed to organize a substantial enough share of the workforce to make their political position invulnerable, before their opponents succeeded in translating the anti-strike backlash into 43

Truman’s reelection in 1948 may thus owe something to the override of his Taft-Hartley veto. The GOP’s success in finally striking back against unions arguably cost the party a popular issue and allowed Democrats to charge that the Republicans had gone too far in shifting policy to the right. Still, it is worth noting that the new Democratic Congress lacked a clear majority for repeal of Taft-Hartley, suggesting that even in the aftermath of the 1948 election, a return to the Wagner Act labor policy regime may not have been a winning issue for Democrats. In other words, even if many voters believed the Republicans had gone too far, support for a return to the pre-Taft-Hartley status quo was weak. 44 In other words, worker support for unions is likely endogenous: it is reasonable to assume that membership in a union leads to more support for pro-union policies.

24

policy changes that hamper further organizing. With the ultimate passage of Taft-Hartley, unions lost that race. For liberals hoping to implement a social democratic policy regime in the aftermath of the war, however, the tenuous place of unions among the mass public may have been a key constraint on their aspirations. Elite debates about Keynesian fiscal policy, anti-monopolism, and planning took place in a context of substantial public restiveness towards the main source of organizational muscle behind liberal hopes, labor unions.45 This public sentiment was reflected most directly in Congress, where members both responded to it and, in the case of conservatives, sought to reinforce it. As discussed below, much additional work needs to be done to unpack the sources of this public reaction, but it suggests that the obstacles confronting liberals were deeply rooted in the electorate as well as Congress.

Public Opinion and Civil Rights Liberalism When congressional southern Democrats and northern Republicans joined together in advocating restrictions on labor unions, they were each reflecting a broadly held perspective among their constituents. At the same time, northern Democrats faced a more difficult situation in which their own constituents were evidently deeply split on labor issues: their core supporters included CIO members (and members of other unions) who strongly opposed restrictions on strikes, yet a substantial portion of northern Democratic voters opposed the sitdown strikes, viewed the closed and union shop with suspicion, and supported a crackdown

45

Again, this restiveness may well have encompassed other issues – which will be explored in the next draft of this paper. There also are a handful of labor survey items that I have not yet had the chance to incorporate. The most promising is a question about whether the Wagner Act ought to be kept as is, revised, or repealed. The question was repeated multiple times starting in 1938. One complication is that in some surveys, Gallup filtered out respondents who said they had not heard of the Act (while in other surveys, there was no filter question asked). Another issue is that “revised” could mean many different things (i.e. one could want it revised in order to make it more pro-labor). In a handful of surveys, respondents were asked what kind of revisions they favored; the responses suggest that a strong majority of “revised” responses were in the pro-employer direction. But the data nonetheless need to be analyzed with care. In any case, a preliminary examination suggests that a plurality favored revising the Act as opposed to keeping it unamended or repealing it.

25

against wartime strikes. How do the coalitional alignments at the mass level change when one shifts the focus to civil rights policy? As noted above, the conventional story regarding the parties’ handling of civil rights has been that nonsouthern Democratic and Republican elites occupied similar positions on the issue until the 1960s, and that partisans at the mass level only diverged after elites – such as Lyndon Johnson and Barry Goldwater – sent clear signals of their respective party’s new position (Carmines and Stimson 1989). From this perspective, the connection between racial liberalism and New Deal liberalism is quite limited: the ideological logic and coalition alignments associated with New Deal liberalism are not inherently associated with civil rights liberalism. Instead, the connection was largely an elite construction, years after the initial rise of New Deal liberalism. However, as noted above, there is considerable empirical evidence that nonsouthern elites began to polarize on civil rights in the mid-1940s, with Democrats adopting the more liberal position. Figure 1 – drawn from collaborative work with Brian Feinstein and Kathryn Pearson – shows the pattern with respect to signing discharge petitions targeting civil rights bills. While nonsouthern Republicans had been more likely to sign discharge petitions for civil rights bills than nonsouthern Democrats in the 1930s, by the mid-1940s the roles had been reversed and nonsouthern Democrats were substantially more likely to sign (Schickler, Pearson, and Feinstein 2008).46 Interestingly, evidence from the early public opinion polls suggests that this elite-level change followed, rather than led, change at the mass level. Nonsouthern white Democratic voters moved to the left of their Republican counterparts on civil rights in the late 1930s – before the elite movement became evident. In addition, support for economic liberalism became increasingly tied to racial liberalism in the nonsouth – and even, though to a lesser 46

The figure is derived from a logit model predicting discharge signatures estimated separately for each civil rights discharge petition. The model controls for seniority, holding a committee leadership position, and membership on the committee targeted by the petition. The figure shows the first difference and 95% confidence interval when one shifts from a northern Republican to a northern Democrat, holding the remaining covariates at their mean or mode.

26

extent, in the south – by the late 1930s. These initial results suggest that once the civil rights movement mobilized sufficiently to force civil rights to the top of the agenda, nonsouthern Democratic voters were by far the most likely to be responsive. In turn, this meant that nonsouthern Democratic elites would have a stronger incentive to take the lead than their GOP counterparts. This incentive would be reinforced by the presence of African Americans and pro-civil rights unions (i.e. the CIO) as core members of the Democratic coalition (Chen 2009; Karol 2009; Feinstein and Schickler 2008). Meanwhile, although nonsouthern Republicans remained much more liberal on civil rights than southern Democrats (and southern Republicans), the connection between economic conservatism and racial conservatism that emerged at the mass level helps explain the eventual “southern strategy” adopted by the Republicans. The focus in this first cut is on white respondents in order to separate out the impact of the changing racial composition of the Democratic Party in the 1930s-40s.47 That is, while northern African Americans clearly played a crucial role in changing the party’s stance towards civil rights (see Jenkins et al. 2009), it is important to determine whether there is also evidence of change among white Democrats and Republicans. The implications for our understanding of New Deal liberalism would be different if all of the Democratic mass-level change was concentrated among African American voters. I am planning a more detailed study of changes in African American partisanship during this period, but doing so is difficult due to the uneven representation of African Americans across the polls. Unfortunately, the first poll with a civil rights attitude item was not taken until January 1937. In that survey, Gallup asked respondents, “Should Congress enact a law which would make lynching a federal crime?” The question, which did not specify who would be punished and what the punishment would be, elicited an extremely positive response (see Table 5). Interestingly, white nonsouthern Landon voters were slightly more likely to support the bill

47

The “full population” numbers in Table 5 include African American respondents, of course.

27

than white nonsouthern Roosevelt voters. When one controls for demographic characteristics in a multivariate model, Landon voters are slightly more supportive of the bill than FDR voters, though the point estimates generally fall short of statistical significance.48 The survey lacks a good item measuring economic liberalism; the closest is a question asking whether government should take over all businesses in wartime. Support for a government takeover is associated with slightly greater support for the lynching bill among nonsouthern whites, but the relationship is weak (about 3-4 points). The identical lynching question was repeated in August and November 1937. In both cases, white nonsouthern Roosevelt voters were slightly more supportive than Landon voters, but the gap was quite small (2-4 points). In both cases, when one examines questions that focus on current views of FDR, as opposed to previous vote, the relationship with support for the lynching bill is a bit stronger (e.g. FDR supporters are 5-7 points more supportive of the bill than FDR opponents, as compared to 2-4 points based on the 1936 vote). When controls are added for individual demographics, voting for FDR is associated with increased support for the lynching bill among nonsouthern whites, though the relationship only reaches statistical significance (and is of greater substantive magnitude) in the November survey. In both surveys, there are also hints of a relationship between economic liberalism and support for the lynching bill, but the results were mixed.49 The December 1937 Gallup survey changed the question wording significantly, adding information about the potential punishment included in the lynching bill (Table 5, Item 1,

48

Controls for demographics (which are used in a consistent manner across the surveys discussed below) include: age, gender, occupation / class (professional, labor, poor), region (Midwest, West), urban residence, phone ownership / car ownership, and, when available, union membership and education. 49 The August survey had four potential economic liberalism measures. Two were related to support for the lynching bill among both nonsouthern whites and southern whites: support for FDR renewing the courtpacking fight and support for government ownership of the banks. But two other items were unrelated (or only very weakly related) to support for the lynching bill: attitudes towards WPA workers and support for the minimum wage. In the November survey, support for maximum hours / minimum wage legislation was related to support for the lynching bill, but other economic measures were more weakly related to support for the lynching bill.

28

variant #2).50 This question sparked a more even division among respondents, and it may be that this allowed a greater separation between weaker and stronger civil rights supporters (by essentially shifting the cutting line). In response to this item, a clear plurality of nonsouthern white FDR voters backed the bill (50% to 33%), while Landon voters split evenly (43% opposed; 43% in favor). The relationship remains strong and significant when controls are added for demographics. In addition, there is a clearer relationship with economic liberalism in this poll. An economic liberalism scale was constructed based on responses to questions concerning government ownership of the railroads and government support for those without money. About ¼ of the nonsouthern white respondents took the liberal position on both issues and ¼ took the conservative position on both. The liberals are much more supportive of the lynching bill than are the conservatives (conservatives opposed the bill 41%-47%; moderates favored it 48%-40%; and liberals backed it 57%-30%). If one narrows the focus even further to liberals who voted for FDR in 1936, the margin in favor grows to 60%-28%, while conservatives who voted for Landon opposed the lynching bill by a striking 53%-33% margin.51 In other words, by December 1937, there was a clear connection between economic liberalism, Democratic vote choice, and support for the lynching bill among nonsouthern whites. Interestingly, while southern whites are notably less liberal on the economic questions than were nonsouthern whites, it is still the case that economic moderates and liberals were less hostile to the lynching bill than were economic conservatives.52 Thus, while partisanship and civil rights support were essentially unrelated in the south, there are indications in the early data that economic liberalism was related to lower levels of racial conservatism. 50

The text read: “Congress is now considering a lynching bill which gives the Federal government power to: fine and imprison local policemen who are negligent in protecting a prisoner from a lynch mob, and also to make a county in which a lynching occurs pay a fine up to $10,000 to the victim or his family. Do you approve of this bill?” 51 One can view such individuals as “sorted partisans”: their vote choice (or partisanship) is consistent with their economic views (see Sniderman and Stiglitz 2009 on the concept of a “sorted partisan”). 52 Economic conservatives opposed the bill 66%-23%. Economic liberals favored the bill slightly (39%35%), while moderates opposed it by a 52-38% margin.

29

The same patterns occur both in the nonsouth and south in a January 1940 survey that asks a similar lynching question (see Table 5, item 1, variant #3). Once again, white FDR voters in the nonsouth are significantly more supportive of the bill than are white Landon voters. A similar gap occurs if one uses presidential approval instead of prior vote choice. Furthermore, those opposed to cuts in relief and public works spending – in both the north and south – are more supportive of the lynching bill.53 Again, the relationships hold up when controls are added for demographic variables. A December 1940 survey is the first to ask about abolishing the poll tax (see Table 5, item #2). This item is trickier to interpret, since the poll tax undermined poor whites’ voting rights in the south, as well as African Americans. But the results with the poll tax items are nonetheless consistent with the lynching bill results. Nonsouthern white Democratic voters are more supportive of banning the poll tax than are nonsouthern white Republican voters. The gap is relatively modest (8 points at the bivariate level) but holds up in multivariate models controlling for demographics. The survey also asks about 1936 vote choice. Interestingly, the 1936 vote item is less strongly related to supporting the poll tax ban than is the 1940 vote choice. Furthermore, a full 74% of new FDR voters in 1940 (after having either not voted or voted for another candidate in 1936) backed the poll tax ban, while just 60% of new Willkie voters backed the ban.54 This suggests that the relationship between support for civil rights and vote choice had tightened over the course of 1936-40.55

53

This result is striking because on one of the two forms, the question highlights that FDR was proposing the spending cuts. The relationship between economic liberalism and support for the lynching bill is only slightly greater when FDR’s position is not mentioned. The form experiment does make it trickier to summarize the results with the issue items, since the issue scales are not easily comparable across forms. However, for form K (which asks explicitly if one approves of FDR’s proposal to cut spending), FDR voters who disagree with cutting the spending are far more likely to back the lynching bill (61%-28%), while Landon voters who agree with FDR on cutting the spending are opposed to the lynching bill by a 56%-39% margin. On Form T, which does not cue FDR’s position, pro-spending FDR voters back the lynching bill 54%-36.5%, while anti-spending Landon voters oppose it 49-45%. 54 The survey did not have good economic liberalism measures. 55 In a multivariate model with controls for demographics, if one includes both 1936 vote and a variable for change in vote from 1936-40 (scored one for new FDR voters; 0 for new Willkie / GOP voters; and .5 for everyone else), the change vote variable is positive and generally statistically significant, while the 1936 vote item is substantively and statistically insignificant.

30

Unfortunately, Gallup did not ask a full national sample about the poll tax for some time after 1940. However, Gallup did ask southern respondents about abolishing the poll tax “in this state” in a handful of polls in 1941 (see Table 5, item 2, variant #1). As with many other civil rights policy questions in this era, southern Roosevelt voters were less supportive of abolishing the poll tax than were (the small number of) Republican voters. Interestingly, however, the results suggest that southerners who were liberal on non-civil rights issues were more supportive of the poll tax ban. For example, the March 1941 survey asked whether respondents approved or disapproved of Henry Ford’s refusal to recognize unions in his plans. The nearly 70% of southern whites who approved of Ford’s actions were substantially less supportive of banning the poll tax than were the 18% of white southerners who disapproved.56 A June 1942 Gallup survey asked respondents about integrating the U.S. military (see Table 5, item 3).57 While the nonsouthern white FDR voters are just six points more supportive than nonsouthern white Willkie voters, the relationship is a bit stronger when one focuses on congressional vote intention (with Democrats split 46%-46% and Republicans opposed by a 54%-38% margin).58 In addition, when one isolates economic liberals who voted for FDR and compares them to economic conservatives who voted for Willkie, there is a much bigger gap in support for integration, with liberal Democrats 15-20 points more supportive than conservative Republicans.59 Thus, six years before Truman’s executive order desegregating 56

Respondents who approved of Ford split 52-41% against abolishing the poll tax; those disapproving of Ford split 58-40% in favor of abolishing the poll tax. The February 1941 poll asked respondents whether they believe the Supreme Court is too liberal, too conservative, or about right. Very few southern whites responded “too conservative.” But those who believed the Court is “too liberal” were more likely to oppose abolishing the poll tax than were those who thought the Court is either “about right” or “too conservative.” Note that this is before the Court issued any of its landmark pro-civil rights rulings (i.e. the U.S. v. Classic ruling on primary election laws was not argued until April 1941, with a decision in late May; Smith v. Allwright, the first major Court move in favor of civil rights, was handed down in 1944). 57 The text read: “Should negro and white soldiers serve together in all branches of the armed forces?” 58 The unweighted gap between FDR voters and Willkie voters is also greater than the weighted gap in this survey (9 points, as compared to 6 points). We also compared core Democratic voters (those voting for FDR in 1940 and intending to vote Democratic for Congress in 1942 and for FDR in 1944) to core Republican voters (defined analogously). Again, core Democratic voters split evenly on integrating the military (46%-46%), while Republicans opposed integration by a substantial 57%-37% margin. 59 The precise gap varies a bit across measures. The survey has two questions that potentially tap economic liberalism: support for gas rationing and support for an income limit on corporate executives. FDR voters who support gas rationing back integration by a 55-41% margin. Willkie voters who oppose rationing

31

the military, nonsouthern white Democratic liberals backed integration at the mass level, while their economically conservative Republican counterparts opposed it. It is worth emphasizing that the integration question – like the earlier poll tax and lynching questions – made no mention of the position of any Democratic or Republican elites. The single civil rights issue that became most prominent in the mid-1940s was fair employment practices (Chen 2009). Following Roosevelt’s creation of a Fair Employment Practices Committee during the war – due to pressure from African American civil rights leaders, most notably A. Philip Randolph – there were repeated efforts to enact FEPC legislation at both the national and state level. Gallup first asked about fair employment practices in 1945. Their question focused on state laws, rather than federal legislation, which is a potential advantage since it separates out concerns about federalism. Gallup worded the question in two ways on the survey: on Form K, respondents were asked about a state law barring discrimination by employers; Form T instead asked about a state law requiring “employees to work along side persons of any race or color.”60 Respondents were more favorable towards barring discrimination by employers than requiring employees to work in an integrated setting. But in both cases, FDR voters were substantially more supportive than were Dewey voters (see Table 5, item 4). Nonsouthern white FDR voters supported a state ban on employer discrimination by a 52%-31% margin, while Dewey voters opposed the measure by a 51%-39% margin. While FDR supporters were much less supportive when the wording focuses on integrated workplaces, they still were significantly more likely to back the proposal than were GOP voters. These relationships also hold up when demographic controls are included. Economic liberalism is also strongly related to support for state fair employment practices at both the bivariate level and when demographics are included in a multivariate

oppose integration by a 57-37% margin. The relationship is a bit weaker with the income limit variable, but is even stronger when one combines the two economic liberalism measures. Support for integration is also related to opposition to banning the communist party. 60 Neither question mentions religion or ethnicity. This makes it less likely that responses were driven by attitudes towards Catholics, Jews, Italians, etc.

32

model.61 When one isolates economically liberal FDR voters and compares them to economically conservative Dewey voters, the gap is even bigger. On Form K, economically liberal FDR voters back a ban on employer discrimination by a 59%-26% margin, while economically conservative Dewey voters oppose the ban 59%-31%.62 Given that Dewey himself was a key advocate of New York state’s fair employment practices law – and that the legislation mentioned in the survey question focused on state, rather than national law – it is striking that white Republican voters were so much more opposed than white Democratic voters at this early stage in the civil rights process (see Chen 2009 for analysis of this and other survey questions on fair employment practices). Southern whites’ opposition to fair employment practices was overwhelming in the 1945 survey and did not differ appreciably between the two parties. Nonetheless, southern economic liberals were about 10 points more likely to back a ban on discrimination than were southern economic conservatives. For example, on Form K, 24% of the liberals backed the ban on employer discrimination, as compared to 14% of moderates and 13% of conservatives. On Form T, 19% of the liberals backed integrated workplaces, as compared to 11% of moderates, and 6% of conservatives. While this degree of support among white southern economic liberals is far from impressive, it does reinforce the more general message that while party and racial attitudes were not aligned in the south (and, if anything, Democrats tended to be a bit more racially conservative), economic ideology was related to racial attitudes, even among white southerners in the Jim Crow era.63 While the analysis thus far has focused on polls through 1945 – in part, because the recoding and weighting project has only covered 1936-45 – I briefly explored a Gallup survey from March 1948. The survey was taken shortly after Truman announced his civil rights 61

The two forms had different economic policy questions. Form K asked about government ownership of the railroads and helping the unemployed. Form T asked about government ownership of electric utilities and about government payments for the unemployed (the latter elicited less support than the more generic question about “helping” the unemployed on Form K). 62 The difference is smaller but still significant on Form T, where the focus is on employees rather than employers. 63 Gallup asks about FEPC again in July 1947, but I have not yet had the chance to code this survey.

33

program. As a result, it is more vulnerable to the concern that elite cues (i.e. attitudes toward Truman) are driving the responses. But this survey’s results are much the same as the earlier surveys, thus reinforcing the general message that the mass-level connection between New Deal economic liberalism and racial liberalism had become reasonably strong even as national elites were just beginning to grapple directly with the issue. The March 1948 survey asked respondents whether they had heard about Truman’s program, what they believed it included, and whether they supported specific civil rights measures. When asked about the program as a whole, the national public was closely divided, with many not registering an opinion or not having yet heard about the package (Table 5, item 5). Still, white nonsouthern FDR voters from 1944 were more supportive of the package than were white nonsouthern Dewey voters (29-13% in favor, as compared to 25%-22% opposed). The survey also asked how far the federal government ought to go “in requiring employers to hire people without regard to their race, religion, color, or nationality” (Table 5, item 6). While white nonsouthern Dewey voters overwhelmingly opposed the federal government requiring non-discrimination (59%-27%), FDR voters narrowly backed it (41%-39%).64 The 1948 survey also included an item that taps into economic liberalism, asking whether the government should “go more to the right, by following more of the views of business and conservative groups,” or “go more to the left, by following more of the views of labor and other liberal groups,” or halfway in between.65 This item nicely distinguishes respondents’ views across the range of civil rights policy items in the nonsouth. For example, while 64% of nonsoutherners who prefer moving to the left favor a ban on lynching, just 51% 64

The same survey asked about the poll tax and revealed near-universal support for a ban among nonsouthern whites in both parties (79% of FDR voters and 71% of Dewey voters). FDR voters backed a lynching bill by a 62%-38% margin, as compared to a 52%-48% split in favor among Dewey voters. FDR voters were only slightly more supportive of desegregating travel, though the wording of this item was arguably confusing (it asked about Truman’s proposal, but is worded in such a way that a response in favor of integration would be “should not” – i.e. blacks “should not” be required to use a separate railroad car). 65 This measure is not ideal since it may be that respondents are thinking of civil rights policy when they answer it. However, the question was asked before the civil rights items in the survey. Furthermore, business groups were generally not highly public in their opposition to civil rights measures (despite their hostility to FEPC). Unfortunately, the survey does not include alternative economic liberalism measures.

34

of those who want to move to the right favor the ban. The gap is even bigger for the FEPC item, with those wanting to move to the left favoring the bill by a 52%-31% margin, and those wanting to move to the right opposing FEPC by a 57%-29% margin. When one once again combines economic liberalism with vote choice, the pattern is particularly sharp: economically liberal FDR voters favor Truman’s civil rights program by a 41%-10% margin and favor FEPC by a 53%-29% margin, while conservative Dewey voters oppose Truman’s program 36%-19% and oppose FEPC by a 62%-21% margin.66 Even among southern whites, responses to the left-right item predict attitudes towards civil rights. While southern white Democratic voters are no more supportive of Truman’s program than are southern white Republicans, liberals were more likely to back most aspects of Truman’s program. For example, while southern whites who want policy to move to the right oppose the Truman program 77%-4%, those who favor moderation are less opposed (57%12%) and the (admittedly few) liberals oppose the package by a still smaller 39%-17% margin. When it comes to the poll tax, a majority of both moderates and liberals in the south back abolition, while a clear majority of conservatives want to keep the poll tax. A similar gap occurs on the lynching ban, though with each group less supportive of a ban than with respect to the poll tax. In sum, economic conservatism and racial conservatism appear to have been connected at the mass level in the south long before Barry Goldwater entered the scene (see Lowndes 2009). Interpreting the Civil Rights Polls Starting in the mid-1940s, northern Democrats became the most aggressive congressional advocates for overcoming southern gatekeeping on civil rights. This elite-level shift was preceded by noteworthy movement among nonsouthern Democratic voters. While this analysis focused solely on white voters, the movement of African Americans into the 66

Unlike most of the earlier polls, this survey includes party identification. The results tell the same substantive story if party identification is substituted for vote choice, but I use the latter item to be consistent with the earlier analysis.

35

Democratic column in the 1930s-40s widened the emergent mass-level gap between nonsouthern Democrats and Republicans. A few limitations of the analysis thus far are worth mentioning. First, the evidence with respect to racial prejudice among whites is much more mixed than is the evidence concerning civil rights policy. That is, white nonsouthern Democrats were only slightly less prejudiced than were white nonsouthern Republicans during this period.67 Second, I have not yet analyzed most of the available survey data in the late 1940s and 1950s. It is possible that the mass-level gap between nonsouthern white Democrats and Republicans receded during the Eisenhower years (though, given the growing elite-level gap in this period, I am skeptical of this). It is also worth comparing the mass-level data from the 1940s to the survey results in the “post-realignment” 1970s-80s. It is plausible that the gap between nonsouthern white Democrats and Republicans may be about the same in that latter era as it was in the 1940s. Third, I have not offered an explanation for why nonsouthern white Democrats became more supportive of civil rights policies than their GOP counterparts at the mass level. Part of the story may be related to cues provided by the social groups allied with each party (Karol 2009; Feinstein and Schickler 2008). With the 1936 elections, African Americans, CIO unions, and Jews emerged as core voters in the Democratic coalition outside the south. These same groups were prominent civil rights supporters (Kesselman 1948). Thus, even as Roosevelt kept silent on civil rights, rank-and-file voters may have taken the cue from these affiliated groups that the Democratic brand includes – or ought to include – civil rights. Yet the extent to which the mass-level Democratic-Republican gap held across societal groupings is striking. For example, nonsouthern white Democratic voters were generally more liberal on civil rights than their Republican counterparts in both urban and non-urban areas; 67

This is based on a preliminary analysis of the data. See Sugrue (2008) on the persistence of prejudice and discrimination in the north during this period. The opinion data presented above show that northern Democrats were more liberal on civil rights policy (as a general matter) than were Republicans; they do not show that most northern Democrats were racial egalitarians or that prejudice had dissipated among northern economic liberals. It is worth noting, however, that even in the contemporary era, the correlation between prejudice and partisanship (and prejudice and ideology) is relatively weak (see Sniderman and Carmines 1999).

36

among high and low socio-economic status voters; and among union and non-union members.68 This by no means refutes the group-cues explanation, but it does suggest it may be incomplete. Another, related possibility is that the meaning of New Deal liberalism itself became sharpened amidst the Democrats’ travails of 1937-38. The fierce reaction against the sit-down strikes and court-packing, along with the recession of 1937-38, may have led some members of the broad Roosevelt coalition of 1936 to peel off from the party, while sharpening the ideological division between FDR’s supporters and opponents. The emergence of a viable conservative opposition to New Deal liberalism may have generated a clearer sense among voters of what it means to be a “Roosevelt Democrat” and what it means to be anti-New Deal. Similarly, the prominent role of the CIO in the Democratic coalition presumably was most likely to alienate those voters who had a more modest view of the New Deal’s aspirations. This sharpening of lines may have helped foster a closer alignment between New Deal economic liberalism and racial liberalism. That is, the individuals who continued to support FDR and the Democrats amidst the setbacks of 1937-38 may have also been more likely to be the type of people who support the broad activist government required to safeguard civil rights.69 Similarly, the types of people who drifted away from Roosevelt and the Democrats as the meaning of New Deal economic liberalism sharpened in the late 1930s may have been the same kinds of people reluctant to support broad social change, such as civil rights. This tentative speculation leaves unanswered (for now) what it was about New Deal economic liberalism that fostered these connections. It seems safe to rule out the potential explanation that anti-civil rights attitudes themselves caused white non-southerners to become economic conservatives or Republicans in 68

In each survey, I ran a series of three-way tables comparing nonsouthern white Democrats and Republicans, controlling for a different demographic variable in each case. Nonsouthern white Democrats tended to be more liberal on civil rights than Republicans across the range of these variables. I thank Jas Sekhon for this suggestion. 69 While white Democrats clearly did not join the party in this era due to concerns about civil rights, it is plausible that the kind of people attracted to the party were also the kind of people more likely to support liberal civil rights policies.

37

this era. The low salience of civil rights on the policy agenda – and the utter absence of clear national elite cues on the issue in the late 1930s – makes it implausible to believe that the connection between racial and economic conservatism that was forged at the mass level in the 1930s reflects a causal effect of racial attitudes on economic attitudes. A more plausible alternative is that the connection between economic and racial conservatism is rooted in a more general stance towards the role of government in addressing social and economic problems. This stance could well be rooted in individual personality characteristics, as recent research has provided powerful evidence that personality traits – such as openness to new experiences – can shape a broad range of attitudes and ideological dispositions (Gerber et al 2008). Again, while this is simply speculation at this point, it may be that the New Deal economic policy “brand” resonated to the same types of people likely to view the treatment of African Americans as a policy problem, even in the absence of explicit elite cues. In any case, the results suggest that the ultimate break-up of the New Deal north-south coalition was built into the alignment that emerged in the 1930s. Nonsouthern Democrats were the group most likely to support civil rights when movement activists forced the issue to the top of the national agenda. In the north, economic liberalism, Democratic identification, and (relative) racial liberalism had already come into alignment by the end of the 1930s at the mass level (and soon after at the elite level). In the south, while racial policy views and partisanship were not aligned, there are indications that economic conservatism and racial conservatism were correlated even in the late 1930s and 1940s. When Barry Goldwater spoke of hunting “where the ducks are” in the 1960s, he was capitalizing upon a longstanding mass-level connection between economic and racial conservatism in the south – rather than creating this linkage. The willingness of congressional Republicans to help southern Democrats stall civil rights legislation in the late 1940s and 1950s was a harbinger of the future partisan alignment (see Schickler, Pearson, and Feinstein 2008). Goldwater’s nomination in 1964 thus represented the culmination of a long process of change within GOP, just as Johnson’s personal transformation on civil rights reflected deep currents in the party he sought to lead.

38

Discussion In this paper, I have attempted to provide some initial evidence concerning the relationship between the mass public and the accomplishments and limitations of New Deal liberalism in the late 1930s-40s. Much work remains to be done, however. Most obviously, the analysis thus far has focused on just two major issues, labor and civil rights. This does not allow one to assess whether the conservative aspects of public opinion toward labor were exceptions to a more general liberal public mood, or instead were emblematic of a broader shift. Similarly, the evidence thus far only barely speaks to the breadth of the anti-union sentiment. A closer examination of rural-urban, class, and ethnic cleavages is necessary.70 The connections between public opinion and congressional election outcomes and behavior also require much more careful exploration. One possibility would be to create a summary measure of liberalism (on labor policy and other issues, perhaps) by state and geographic area and examine how well this measure predicts later Republican congressional gains. For example, do anti-labor attitudes in 1937 help predict GOP gains in the 1938 midterms? Do these attitudes outperform other attitude measures (and measures of demographic composition, such as urban-rural and ethnic / religious identification)? There are numerous complications with this approach, rooted in the design of the Gallup samples, changes in question wording, and the lack of panel data. But it may be possible to use statistical modeling to gain some leverage (see Lax and Phillips 2009 on a potential approach for estimating aggregate opinion by state).

70

It will also be useful to compare attitudes towards unions to attitudes towards farmers during the war. Wartime price control policies often placed the interests of labor and farmers in direct opposition to one another, greatly undermining the chances for a “green-red” coalitional alignment. The administration viewed demands for higher farm prices as at least as great – or perhaps an even greater – threat to price controls as were wage demands from workers. Yet Congress proved far more receptive to the farmers’ claims. The rural bias in congressional (and Senate) districts no doubt helped the farmers, but it is worth exploring how the mass public viewed demands from farmers, as compared to labor. It may be that the mythology of the yeoman farmer shielded agricultural interests from the same backlash faced by labor.

39

Most broadly, however, the goal of this project is to bring the study of mass political attitudes and behavior squarely into the analysis of congressional policy-making in the New Deal era. Understanding the opportunities and constraints confronting political elites – along with the impact of institutional rules of the game within Congress itself – requires grappling with the political context presented by members’ constituents. The rise of the conservative coalition in Congress may in part be a story about changes in popular attitudes towards labor and other core elements of the emergent Democratic coalition. Conservatives both encouraged and benefited from this transformation in the political terrain – which few could have anticipated in the immediate aftermath of Roosevelt’s 1936 landslide, but which helped define the contours of American politics for decades to come.

40 References Alsop, Joseph, and Turner Catledge. 1938. The 168 Days. Garden City, NY: Doubleday. Alston, Lee, and Joseph Ferrie. 1999. Southern Paternalism and the American Welfare State. Cambridge: Cambridge University Press. Amenta, Edwin et al. 1998. “Bring Back the WPA: Work, Relief, and the Origins of American Social Policy in Welfare Reform.” Studies in American Political Development, 12: 1-56. Atleson, James B. 1998. Labor and the Wartime State: Labor Relations and Law During World War II. Urbana: University of Illinois Press. Baum, Matthew A. and Samuel Kernell. 2001. “Economic Class and Popular Support for Franklin Roosevelt in War and Peace.” Public Opinion Quarterly 65: 198-229. Bensel, Richard. 1984. Sectionalism and American Political Development. Madison: University of Wisconsin Press. Berinsky, Adam J. 2006. “American Public Opinion in the 1930s and 1940s: The Analysis of Quota-Controlled Sample Survey Data.” Public Opinion Quarterly 70: 499-529. Berinsky, Adam J., Eleanor Powell, Eric Schickler, and Ian Yohai. 2008. “Revisiting Public Opinion in the 1930s and 1940s.” Paper prepared for the Annual Meeting of the American Political Science Association, Boston, MA. Berinsky, Adam J., and Eric Schickler. 2006. “The American Mass Public in the 1930s and 1940s.” National Science Foundation, Collaborative Research Grant. http://web.mit.edu/berinsky/www/nsf.pdf Brinkley, Alan. 1995. The End of Reform. New York: Knopf. Calkins, Fay. 1952. The CIO and the Democratic Party. Chicago: University of Chicago Press. Cantril, Hadley and Mildred Strunk. 1951. Public Opinion, 1935-1946. Princeton: Princeton University Press. Caldeira, Gregory A. 1987. “Public Opinion and The U.S. Supreme Court: FDR’s CourtPacking Plan.” The American Political Science Review, 81: 1139-1153. Carmines, Edward G., and James A. Stimson. 1989. Issue Evolution: Race and the Transformation of American Politics. Princeton, NJ: Princeton University Press.

41 Chen, Anthony. 2006. “‘The Hitlerian Rule of Quotas’: Racial Conservatism and the Politics of Fair Employment Legislation in New York State, 1941–1945,” Journal of American History 92: 1238. Chen, Anthony. 2007. “The Party of Lincoln and the Politics of State Fair Employment Practices Legislation in the North, 1945-64,” American Journal of Sociology 112: 1713– 74. Chen, Anthony. 2009. The Fifth Freedom: Jobs, Politics, and Civil Rights in the United States, 1941-1972. Princeton: Princeton University Press. Chen, Anthony, Robert Mickey and Robert Van Houweling. 2008. “Explaining the Contemporary Alignment of Race and Party: Evidence from California’s 1946 Ballot Initiative on Fair Employment,” Studies in American Political Development 22: 204-228. Converse, Jean. 1987. Survey Research in the United States: Roots and Emergence. Berkeley: University of California Press. Eisinger, Robert M. 2003. The Evolution of Presidential Polling. New York, NY: Cambridge University Press. Eisinger, Robert M. and Jeremy Brown. 1998. “Polling as a Means Toward Presidential Autonomy: Emil Hurja, Hadley Cantril and the Roosevelt Administration.” International Journal of Public Opinion Research 10 (3): 237-256. Farhang, Sean, and Ira Katznelson. 2005. “The Southern Imposition: Congress and Labor in the New Deal and Fair Deal.” Studies in American Political Development, 19 (April), 1-30. Feinstein, Brian D. and Eric Schickler. 2008. “Platforms and Partners: The Civil Rights Realignment Reconsidered.” Studies in American Political Development. Gallup, George. 1939. “Public Opinion Polls Important in Democracy.” Washington Post, August 6, 1939, p. B3. Gallup, George. 1938. “Causes of the Swing to the Right Analyzed.” Washington Post, November 14, 1938. Gerber, Alan S., Gregory A. Huber, Shang Ha, and Conor M. Dowling. 2008. “Personality Traits and the Dimensions of Political Ideology.” Unpublished Manuscript, Yale University. Goldfield, Michael. 1989. “Worker Insurgency, Radical Organization, and New Deal Labor Legislation.” American Political Science Review 83 (December): 1257-1282.

42 Goodman, Walter. 1968. The Committee: The Extraordinary Career of the House Committee on Un-American Activities. New York: Farrar, Straus, and Giroux. Gordon, Colin. 1994. New Deals: Business, Labor, and Politics in America, 1920-1935. New York: Cambridge University Press. Igo, Sarah. 2007. The Averaged American: Surveys, Citizens, and the Making of a Mass Public. Cambridge: Harvard University Press. Jeffries, John W. 1990. “The ‘New’ New Deal: FDR and American Liberalism, 19371945.” Political Science Quarterly 105:397-418. Jenkins, Jeffrey, Vesla Weaver, and Justin Peck. 2009. “Between Reconstructions: Congressional Action on Civil Rights, 1890-1950.” Paper presented at the Congress and History Conference, Charlottesville, Virginia, May 2009. Karol, David. “Realignment without Replacement: Issue Evolution and Ideological Change among Members of Congress” (paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 15–18, 1999). Karol, David. Coalition Management: Explaining Party Position Change in American Politics. Forthcoming, Cambridge University Press. Katznelson, Ira, Kim Geiger, and Daniel Kryder. 1993. “Limiting Liberalism: The Southern Veto in Congress, 1933-1950.” Political Science Quarterly 108:283-306. Katznelson, Ira, and John Lapinski. 2006. “Congress and American Political Development: Missed Chances, Rich Possibilities.” Perspectives on Politics 4(2): 243260. Kesselman, Louis. 1948. The Social Politics of FEPC: A Study in Reform Pressure Movements. Chapel Hill: University of North Carolina Press. Kornhauser, Arthur. 1946. “Are Public Opinion Polls Fair to Organized Labor.” Public Opinion Quarterly 10:484-500. Lax, Jeffrey R., and Justin H. Phillips. 2009. “How Should We Estimate Public Opinion in the States?” American Journal of Political Science 53: 107-21. Lee, Taeku. 2002. Mobilizing Public Opinion. Chicago: University of Chicago Press. Lichtenstein, Nelson. 1987. Labor's War at Home: The CIO in World War II. Cambridge: Cambridge University Press.

43 Lichtenstein, Nelson. 1989. “From Corporatism to Collective Bargaining.” In The Rise and Fall of the New Deal Order, 1930-1980, Steve Fraser and Gary Gerstle, eds. Princeton: Princeton University Press. Lichtenstein, Nelson. 2002. “Politicized Unions and the New Deal Model: Labor, Business, and Taft-Hartley,” in Sidney Milkis and Jerome Mileur, eds., The New Deal and the Triumph of Liberalism, University of Massachusetts Press. Lowndes, Joseph. 2009. From the New Deal to the New Right: Race and the Social Origins of Modern Conservatism. New Haven: Yale University Press. Mayhew, David R. 2000. America’s Congress: Actions in the Public Sphere, James Madison Through Newt Gingrich. New Haven: Yale University Press. McMahon, Kevin J. 2004. Reconsidering Roosevelt on Race. Chicago: University of Chicago Press. Norpoth, Helmut, Victor Lange, and Michael Morzenti. 2009. “Voting in Wartime: The 1944 Election.” Paper presented at the Midwest Political Science Association Annual Meeting, April 2009. Orren, Karen, and Stephen Skowronek. 1998-99. “Regimes and Regime Building in American Government: A Review of Literature on the 1940s.” Political Science Quarterly (Winter 1998-99), v. 113, pp. 689-702. Patterson, James T. 1967. Congressional Conservatism and the New Deal. Lexington: University of Kentucky Press. Plotke, David. 1996. Building a Democratic Political Order: Reshaping American Liberalism in the 1930s and 1940s. Cambridge: Cambridge University Press. Porter, David L. 1980. Congress and the Waning of the New Deal. Port Washington, NY: Kennikat Press. Reiter, Howard L. 2001. “The Building of a Bifactional Structure: The Democrats in the 1940s.” Political Science Quarterly 116:107-129. Roper, Elmo. 1940. “Sampling Public Opinion.” Journal of the American Statistical Association 35(210, Part 1): 325-34. Ruggles, Steven, Matthew Sobek, Trent Alexander, Catherine A. Fitch, Ronald Goeken, Patricia Kelley Hall, Miriam King, and Chad Ronnander. 2004. “Integrated Public Use Microdata Series: Version 3.0”. Minneapolis, MN: Minnesota Population Center. Rusk, Jerrold. 2001. A Statistical History of the American Electorate. Washington, DC: CQ Press.

44

Schickler, Eric, and Kathryn Pearson. 2009. “Agenda Control, Majority Party Power, and the House Committee on Rules, 1937-1952. Legislative Studies Quarterly, Forthcoming. Schickler, Eric, Kathryn Pearson, and Brian Feinstein. 2008. “Shifting Partisan Coalitions: Support for Civil Rights in Congress from 1933-1972.” Paper prepared for presentation at the Annual Meeting of the American Political Science Association, Boston, MA. Schlozman, Kay Lehman, and Sidney Verba. 1979. Injury to Insult: Unemployment, Class, and Political Response. Cambridge: Harvard University Press. Sniderman, Paul M., and Edward H. Stiglitz. 2009. The Rational Partisan. Unpublished manuscript, Stanford University. Sugrue, Thomas. 2008. Sweet Land of Liberty: The Forgotten Struggle for Civil Rights in the North. New York: Random House. Tomlins, Christopher. 1985. The State and the Unions. New York: Cambridge University Press. Verba, Sidney and Kay Lehman Schlozman. 1977. “Unemployment, Class Consciousness, and Radical Politics: What Didn’t Happen in the Thirties.” The Journal of Politics 39: 291-323. Weatherford, M. Stephen, and Boris Sergeyev. 2000. “Thinking about Economic Interests: Class and Recession in the New Deal.” Political Behavior, 22: 311-339. Weir, Margaret. 2005. “States, Race, and the Decline of New Deal Liberalism.” Studies in APD, vol 19 (October), pp. 157-72. Workman, Andrew. 2000. “Creating the National War Labor Board: Franklin Roosevelt and the Politics of State Building in the Early 1940s.” Journal of Policy History 12: 23364. Wreszin, Michael. 1975. “The Dies Committee.” In Arthur Schlesinger and Roger Bruns, eds., Congress Investigates. New York: Chelsea House. Zelizer, Julian. 2000. “The Forgotten Legacy of the New Deal: Fiscal Conservatism and the Roosevelt Administration, 1933-1938.” Presidential Studies Quarterly 30: 332-59. Zieger, Robert H. 1995. The CIO: 1935-1955. Chapel Hill: The University of North Carolina Press.

3. Use force to remove sit-down strikers

2. Sit-down strikes made illegal?

1. Sympathy in GM Strike

Question Wording

56% Use force 34% Do not force 10% Don’t Know 59% Use force 30% Do not force 11% Don’t Know

3/37 (Gallup 74)

3/37 (Gallup 75)

64% Make illegal 22% Do not ban 13% Don’t Know

2/39 (Gallup 147)

51% Use force 37% Do not force 12% Don’t Know

55% Make illegal 32% Do not ban 13% Don’t Know

3/37 (Gallup 72)

3/37 (Gallup 73)

60% Make illegal 28% Do not ban 12% Don’t Know

32% Pro-union 42% Pro-GM 25% Don’t Know

2/37 (Gallup 67)

2/37 (Gallup 71)

37% Pro-union 33% Pro-GM 30% Don’t Know

National #s

1/37 (Gallup 64)

Date

South & Pac Coast

East Central

South

East and West Central

South

South

West Central

East Central

Relatively Anti-Labor

Table 1: Attitudes toward Sit-Down Strikes (and Immediate Aftermath)

Rocky Mtn & West Central

Mid-Atlantic

Rocky Mountain

Rocky Mountain & Pac Coast

Rocky Mountain

Rocky Mountain

Pac Coast

Mid-Atlantic Rocky Mountain

Relatively Pro-Labor

49-40.5 use force

47-44 use force

49-40 do not force

63-24 illegal

49-39 illegal

52-36 illegal

40-35 prostrikers

46-23 prostrikers

North Dems.

71-19 use force

76-18 use force

74-18 use force

80-11 illegal

67-22 illegal

75-17 illegal

65-18 pro-GM

52-18 pro-GM

North Repubs.

67-21 use force

56-30 use force

58-26 use force

62-20 illegal

59-24 illegal

65-22 illegal

43-30 proGM

39-33 prostrikers

South

45

6/37 (Gallup 87)

2/37 (Gallup 69)

12/37 (Gallup 107)

6/37 (Gallup 87)

5. Want CIO to succeed in organizing Ford?

6. Sympathy in Ford Dispute?

7. Union view change, last 6 months

East Central & Pac Coast

East Central

South

East Central, South, Pacific Coast

South

Relatively Anti-Labor

33% Less favorable West Central 54% No change 13% More favorable

28% Pro-union 46% Pro-Ford 27% DK

46% Fail 25% Succeed 29% Don’t Know

52% Militia 37% No militia 10% Don’t Know

59% Use force 24% Do not force 17% Don’t Know

12/37 (Gallup 105)

4. Should militia be called in?

57% Use force 33% Do not force 10% Don’t Know

4/37 (Gallup 76)

Use force to remove sit-down strikers

National #s

Date

Question Wording

Mid-Atlantic

Mid Atl, Pac Coast

Pacific Coast

Mid-Atlantic & Northeast

Rocky Mountain & West Central

Relatively Pro-Labor

Table 1, CONTINUED: Attitudes toward Sit-Down Strikes (and Immediate Aftermath)

28-18 less favor

42%-36% pro-union

38-37% fail

46-46% split

51-33% use force

47-44 use force

North Dems.

47-5 less favor

72-11% pro-Ford

73-7% fail

63-29% for militia

77-13% use force

73-19 use force

North Repubs.

27-10% less favor

40-21% pro-Ford

38-19% fail

57-29% for militia

61-15% use force

63-24 use force

South

46

Question Wording 1. “In the current General Motors strike are your sympathies with the John L. Lewis group of striking employees, or with the employers?” (Gallup 64, 67) 2. “Do you think this state should pass legislation making sit-down strikes illegal?” (Gallup 71, 72, 147). 3. “Do you think that state and local authorities should use force in removing sit-down strikers?” (Gallup 73, 74, 76, 105) 4. “Should the militia be called out whenever strike trouble threatens?” (Gallup 87) 5. “Would you like to see John L. Lewis succeed in organizing the Ford Motor Company?” (Gallup 69) 6. “In the present dispute between Henry Ford and the Automobile Workers union, are your sympathies with Ford or with the Union?” (Gallup 107). 7. “Has your attitude towards labor unions changed any during the last six months?” [if yes:] “Are you more in favor, or less in favor of labor unions at this time?”

47

10/41* (Gallup 249)

11/41* 77% Oppose (Gallup 252K) 15% Favor 8% Don’t Know 11/41* 80% Oppose (Gallup 252T) 13% Favor 7% Don’t Know 7/45* (Gallup 351)

5/39 (Gallup 158)

10/41* (Gallup 249)

Variant #2

Variant #2

Variant #1

Variant #2

2. Union shop

Variant #1

Mid-Atlantic

Northeast & Rocky Mtn

Mid-Atlantic

Northeast

Northeast & Mid-Atlantic

Relatively Pro-Labor

52% Oppose West Central, Mid-Atlantic & 35% Favor South, Rocky Mtn East Central 12% Don’t Know

65% Oppose West Central, 25% Favor South, Pac Coast 10% Don’t Know

72% Oppose West Central 15% Favor 13% Don’t Know

Rocky Mtn & Northeast

Rocky Mtn & South

75% Oppose Pac Coast 13% Favor 12% Don’t Know

61% Oppose Pac Coast 22% Favor 17% Don’t Knows

5/39 (Gallup 158)

Variant #1

Relatively Anti-Labor

56% Oppose West Central & 29% Favor South 15% Don’t Know

National #s

7/37 (Gallup 89)

Date

1. Closed shop

Question Wording

Table 2: Attitudes toward Union Shop / Closed Shop

45-42 support

51-42 oppose

68-23 oppose

76-17 oppose

72-21 oppose

70-17 oppose

50-31 oppose

46-40 oppose

North Dems.

64-25 oppose

78-12 oppose

84-7 oppose

87-8 oppose

85-10 oppose

82-7 oppose

74-11 oppose

80-9 oppose

North Repubs.

59-28 oppose

80-10 oppose

70-10 oppose

82-10 oppose

84-7 oppose

77-11 oppose

63-20 oppose

59-24 oppose

South

48

4-5/43

4-5/44* (Gallup 319)

4-5/44* (Gallup 319)

11/41 (Gallup 254)

7/45 (Gallup 351)

3. Favor State Const amend. banning closed shop (subsample only)

4. Maintenance of membership

Variant 1

5. Which favor in factories / mines

6. Which of these plans prefer?

53% Open shop Northeast 17% Union shop 9% Closed shop 21% Don’t Know

60% Open shop South 23% Union shop 4% Closed shop 13% Don’t Know

59% Oppose Northeast, 24% Favor South & 17% Don’t Know Rocky Mtn

58% Oppose South & 27% Favor East Central 15% Don’t Know

47% Favor ban South 40% Oppose ban 13% Don’t Know

57% Oppose West Central, 28% Favor Northeast & 14% Don’t Know South

National #s

Relatively Anti-Labor

Mid Atlantic & East Central

Pac Coast

Mid Atlantic & Pac Coast

West Central & Pac Coast

Mid-Atlantic, East Central, Rocky Mtn, Pac Coast

Relatively Pro-Labor 68-20 oppose

North Repubs.

67% open 17% union 3% closed

69-14 oppose

75-15 oppose

44% open 67% open 24% union 13% union 15% closed 2% closed

52% open 28% union 6% closed

50-32 oppose

49-37 oppose

51-36 54-31 oppose ban favor ban

50-41 oppose

North Dems.

53% open 10% union 7% closed

71% open 15% union 1% closed

68-17 oppose

65-17 oppose

62-26 favor ban

63-14 oppose

South

Question wording: 1. “Are you in favor of the “closed shop”—that is, requiring every worker to belong to the union?” (Gallup 89) Variant 1: “Are you in favor of the so-called 'closed shop'--that is, hiring only persons who are already members of the union to work in a factory or a mine?” (158, 252T)

7/45* (Gallup 351)

Date

Union shop, Variant #2

Question Wording

Table 2, Continued: Attitudes toward Union Shop / Closed Shop

49

Variant 2: “Are you in favor of the Closed Shop--that is, requiring every worker in a company to belong to a union before he can be hired?” (Gallup 249, 252K, 351K) 2. Are you in favor of so-called 'union-shop'--that is, requiring every worker in a factory or mine to join the union? (Gallup 158) Variant 1: “Are you in favor of the Union Shop--that is, requiring every worker in a company to join a union after he is hired?” (Gallup 249) Variant 2: “Are you in favor of the Union Shop – that is, requiring every worker in a company where there is a union to join that union AFTER he is hired?” (351, Form K) 3. Would you favor or oppose adding a law to the Constitution of this state forbidding the 'closed shop'? (Gallup 294) NOTE: just asked of those who answered that they knew what is meant by “closed shop” (61% of all respondents). 4. “Do you think a person who joins a union should have to continue to belong to that union in order to hold his job?” (Gallup 319, Form K) Variant 1: “Do you favor or oppose 'maintenance of membership'--that is, requiring a person who joins a union to continue to belong to that union in order to hold his job?” (Gallup 319, Form T) 5. “Which of these arrangements do you favor for workers in factories and mines...open shop, union shop or closed shop?” (Gallup 254, Form T) 6. “Which one of these three plans do you prefer?... The closed shop--that is, requiring every worker in a company to belong to a union before he can be hired. The union shop--that is, requiring every worker in a company where there is a union to join that union after he is hired. The open shop--that is, not requiring any worker to join a union but letting each one decide whether or not to join.” (Gallup 351, Form T)

50

4-5/43 (gallup 294) 10/43 (Gallup 305) 3/45 (Gallup 342) 7/45 (Gallup 351) 4-5/44 (Gallup 319) 3/41 (Gallup 232)

Variant #1

Variant #1

2. Forbid all strikes in wartime

3. Draft Strikers

4. Have labor leaders done as much as should to help production program?

Northeast, West Central & South

2% Yes! South & 15% Yes Rocky Mtn 28% No 39% No! 16% Don’t Know

68% Draft South & 22% Don’t Draft West Central 10% Don’t Know

79% ban strikes 15% Do not ban 6% Don’t know

75% ban strikes Northeast, 20% right to strike West Central & 6% Don’t Know South

70% ban strikes South & East 22% right to strike Central 8% Don’t Know

82% ban strikes South 13% right to strike 5% Don’t Know

74% ban strikes South 18% right to strike 7% Don’t Know

11/41 (Gallup 254)

Relatively Anti-Labor

76% ban strikes South, West 18% right to strike Central, & Rocky 6% Don’t Know Mtn

National #s

11/41 (Gallup 252)

Date

Variant #1

1. Forbid strikes in war industries

Question Wording

Table 3: Attitudes toward Wartime Strikes and Mobilization Effort

Mid Atlantic

Pac Coast & Mid Atlantic

Pac Coast & Rocky Mtn

Mid-Atlantic

Northeast, MidAtlantic, Pac Coast

Pac Coast

Rocky Mtn

Northeast & Mid-Atlantic

Relatively Pro-Labor

4% Yes! 19% Yes 28% No 31% No! 18% DK

62-29 draft them

74-22 ban strikes

66-29% ban strikes

60-33 ban strikes

78-18 ban strikes

71-21% ban strikes

74-22% ban strikes

North Dems.

1% Yes! 10% Yes 30% No 46% No! 13% DK

71-20 draft them

83-13 ban strikes

80-17% ban strikes

77-14 ban strikes

87-10 ban strikes

76-18% ban strikes

78-16% ban strikes

North Repubs.

1% Yes! 10% Yes 25% No 52% No! 12% DK

83-10 draft them

83-7% ban strikes

83-9% ban strikes

79-10% ban strikes

90%-5% ban strikes

85-9% ban strikes

85-9% ban strikes

South

51

1. Should the government forbid strikes in defense industries or should the workers in those industries continue to have the right to go on strike? (Gallup 252, 254) Variant 1: Should Congress pass a law forbidding strikes in war industries or should the workers in war industries continue to have the right to go on strike? (Gallup 294, 305, 342). [note in AIPO 305 union members are split about evenly on the question) 2. “Do you think Congress should pass a law forbidding strikes in ALL industries during war-time?” (AIPO 351) 3. “Would you favor or oppose a law which would permit local draft boards to draft any man between the ages of 18 and 45 who goes on strike?” (Gallup 319) 4. “Do you think labor union leaders are helping the national defense production program as much as they should?” (Gallup 232).

Question Wording:

52

Date 4-5/43 (Gallup 294)

10/41 (Gallup 249)

10/40*

10/40*

11/43 (NORC 217)

5/44 (Gallup 319)

Question Wording

1. View of Unions change, past year

2. Want more or less union power?

3. More or less labor regulation next four years

Compare: More or less business regulation next four years

4. After war, regulate unions more or less than before war (no weights yet)

5. CIO endorsement impact?

52% Vote against 20% No difference 11% Vote in favor 18% Don’t Know

61% More 15% Same 9% Less 15% Depends / Don’t Know

23% More regulated 17% Same 37% Less regulated 24% Don’t Know

41% More regulated 14% Same 14% Less regulated 31% Don’t Know

48% Less Power 25% Same 9% More Power 18% Don’t Know

38% Less support 57% Same support 5% More supportive

National #s

South, West Central & Rocky Mtn

West North Central & Mountain

Pac Coast & Northeast (want less business reg)

Pac Coast

Pac Coast

South

Relatively Anti-Labor

Mid-Atlantic

Mid-Atlantic

North Repubs.

South

39% More 19% Same 9% Less 33% DK

40% Less 31% Same 11% More

41-15% Against

61% more 17% same 9% less

70-5% Against

69% more 14% same 8% less

12% More 7% Same 68% Less 13% DK

46% More 10% Same 21% Less 23% DK

62% Less 22% Same 4% More

58-6% Against

61% more 11% same 8% less

26% More 23% Same 25% Less 26% DK

41% More 13% Same 10% Less 36% DK

52% Less 19% Same 7% More

36% less vs. 43% less vs. 44% less vs. 7% more 3% more 3% more

North Dems.

South (want 31% More more business 22% Same reg – split evenly) 17% Less 30% DK

West Central

Mid-Atlantic

Pac Coast

Relatively Pro-Labor

Table 4: Attitudes toward Union Power, Regulating Unions, and CIO Endorsement

53

Relatively Anti-Labor

Relatively Pro-Labor

41% Prefer big bus. 32.5% Prefer unions 27% Don’t Know

South

Mid-Atlantic & Rocky Mtn

7% Unions do excellent job 32% Some mistakes, but good on whole 46% Gone much too far but some good in past 7% Do away with entirely 9% Don’t know

National #s

50-25 pro-union

54% pos. 38% neg.

North Dems.

34% pos. 52%neg.

South

55-21 52-17 pro-business pro-business

28% pos. 68% neg.

North Repubs.

Question Wording: 1. Has your attitude toward labor unions changed in any way during the last year?. If 'Yes' on Q10A: Are you more in favor or less in favor of labor unions than you were a year ago? (Gallup 294) 2. “Do you think labor unions should have more power, or less power, than at present?” (Gallup 249) 3. “During the next four years, do you think there should be more regulation or less regulation of labor unions by the Federal Government, than at present?” (AIPO 215; just includes people who claim they are “able to vote in November”; use presidential vote intention 1940). Compare: “During the next four years, do you think there should be more regulation or less regulation of business by the Federal Government, than at present?” (AIPO 215) 4. “After the war, do you think the Federal Government should regulate Gas and Electric Companies more or less than it did before the war started (say 1938)? How about Labor Unions?” (note for gas / electric, distribution is: 34% more, 30% same, 15% less regulation) (NORC 217) 5. “If the CIO union supported a candidate, would you be more likely to vote for that candidate, or against him?” (Gallup 319). Note: survey also asks about NAM endorsement; also viewed negatively, but less than CIO 6. “Which one of these four statements comes closest to expressing your feelings? 1. Labor unions have done an excellent job for this country, and they should be given more power than they now have 2. Labor unions have made a number of mistakes, but on the whole they have done much more good than harm and the public should continue to support them 3. Although labor unions have done some good in the past, they have gone much too far and should be prevented by law from doing many of the things they now do 4. Labor unions are a bad thing for the country and they should be done away with entirely.” Note that code 1 or 2 as “positive” and 3 or 4 as negative. (Roper 36) Same question is asked in other Roper surveys in this period (need to add--xx) 7. “Most people believe that no one group should have too much influence on the government. However, if you HAD to choose, which would you prefer to have influence the government more strongly--big business or labor unions?” (Gallup 312)

2/44 (Gallup 312)

10/43 (Roper 36)

6. General union evaluation (no weights)

7. Union vs. Big Business power

Date

Question Wording

Table 4 (Continued) : Attitudes toward Union Power, Regulating Unions, and CIO Endorsement

54

12/37 (Gallup 106) 1/40 (Gallup 181)

12/40 (Gallup 226)

Variant #2

Variant #3

2. Abolish poll tax

37% Yes! 26% Yes 11% No 15% No! 12% DK

24% Approve! 25% Approve 19% Disapprove 23% Disapprove! 8% No opinion

44% Support 40% Oppose 16% DK

61% Support 25% Oppose 14% DK

62% Support 24% Oppose 13% DK

11/37 (Gallup 102)

11/37 (Gallup 104)

59% Support 24% Oppose 17% DK

8/37 (Gallup 96)

Variant #1

60% Support 23% Oppose 17% DK

1/37 (Gallup 63)

1. Lynching bill

Full Population

Date

Question Wording

34% Yes! 26% Yes 11% No 16% No! 13% DK

19% Approve! 24% Approve 23% Disapprove 27% Disapprove! 7% No opinion

43% Support 43% Oppose 14% DK

60% Support 27% Oppose 13% DK

66% Support 23% Oppose 11% DK

59% Support 22% Oppose 19% DK

63% Support 22% Oppose 14% DK

40% Yes! 28% Yes 10% No 11% No! 11% DK

26% Approve! 26% Approve 20% Disapprove 19% Disapprove! 9% No opinion

50% Support 33% Oppose 17% DK

61% Support 25% Oppose 13% DK

68% Support 19% Oppose 12% DK

63% Support 21% Oppose 16% DK

59% Support 22% Oppose 14% DK

NONSOUTHERN WHITES Landon / FDR / GOP voters Democratic

Table 5: Civil Rights Policy Views and Partisan Vote Choice, By Region

32% Yes! 17% Yes 12% No 34% No! 6% DK

5% Approve! 15% Approve 25% Disapprove 49% Disapprove! 6% No opinion

28% Support 68% Oppose 4% DK

64% Support 27% Oppose 9% DK

44% Support 39% Oppose 16% DK

37% Support 47% Oppose 16% DK

59% Support 24% Oppose 16% DK

25% Yes! 25% Yes 14% No 25% No! 9% DK

16% Approve! 12% Approve 21% Disapprove 43% Disapprove! 8% No opinion

30% Support 57% Oppose 13% DK

60% Support 28% Oppose 12% DK

38% Support 43% Oppose 18% DK

44% Support 39% Oppose 18% DK

54% Support 28% Oppose 18% DK

SOUTHERN WHITES Landon / FDR / GOP voters Democratic

55

Date 2/41 (Gallup 230)

3/41 (Gallup 232)

6/42 (Gallup 270)

6/45 (Gallup 349K)

6/45 (Gallup 349T)

Question Wording

Poll Tax, Variant 1: (Asked only in south)

Poll Tax, Variant 1: (Asked only in south)

3. Integrate the military

4. State Fair Employment Practices

Variant 1:

34% Favor 57% Oppose 9% DK

44% Favor 44% Oppose 13% DK

41% Favor 51% Oppose 7% DK

NA

NA

Full Population

32% Favor 61% Oppose 7% DK

39% Favor 51% Oppose 10% DK

38% Favor 54% Oppose 8% DK

41% Favor 48% Oppose 10% DK

52% Favor 31% Oppose 16% DK

44% Favor 48% Oppose 8% DK

NONSOUTHERN WHITES Landon / FDR / GOP voters Democratic

Table 5, Continued: Civil Rights Policy Views and Partisan Vote Choice, By Region

8% Favor 82% Oppose 10% DK

15% Favor 80% Oppose 5% DK

33% Favor 65% Oppose 2% DK

38% Yes! 13% Yes 14% No 23% No! 12% DK

39% Yes! 14% Yes 12% No 26% No! 10% DK

12% Favor 85% Oppose 4% DK

13% Favor 80% Oppose 7% DK

19% Favor 76% Oppose 5% DK

26% Yes! 18% Yes 18% No 33% No! 5% DK

25% Yes! 19% Yes 10% No 39% No! 6% DK

SOUTHERN WHITES Landon / FDR / GOP voters Democratic

56

34% Favor 7% Partway 48% Oppose 11% Don’t Know

27% Favor 7% Partway 59% Oppose 7% DK

22% Favor 25% Oppose 23% No opinion 30% Not heard 41% Favor 8% Partway 39% Oppose 11% DK

29% Favor 13% Oppose 21% No opinion 38% Not heard

10% Favor 2% Partway 81% Oppose 8% DK

13% Favor 62% Oppose 9% No opinion 16% Not heard

9% Favor 7% Partway 73% Oppose 11% DK

8% Favor 66% Oppose 12% No opinion 14% Not heard

SOUTHERN WHITES Landon / FDR / GOP voters Democratic

Question wording: 1. “Should Congress enact a law which would make lynching a federal crime?” (Gallup 63, 96, 102). Variant 1: “If a local peace officer is negligent in protecting a prisoner from a lynch mob, should the federal government have the right to punish this peace officer?” (Gallup 104) Variant 2: “Congress is now considering a lynching bill which gives the Federal government power to: fine and imprison local policemen who are negligent in protecting a prisoner from a lynch mob, and also to make a county in which a lynching occurs pay a fine up to $10,000 to the victim or his family. Do you approve of this bill?” (Gallup 106). Variant 3: Congress is now considering a bill against lynching which would punish lynching in these two ways: Will you please read this card and tell me whether you approve or disapprove of this bill against lynching? Code for hand card: 1. Fine and imprison local police officers who fail to protect a prisoner from a lynch mob 2. Make a county in which a lynching occurs pay a fine up $10,000 to the victim or his family (Note this is Gallup 181; seem to have same item on Gallup 182 and get same pattern but only asked of 1/3 of sample roughly—missing data for 2/3 – unclear why but seem like real problems with that dataset so omit) 2. Some Southern states require every other voter to pay a poll tax amounting to about a dollar a year before they can vote. Do you think these poll taxes should be abolished? (Gallup 226). Variant 1 (asked only in MN and south): Should the poll tax be abolished as a requirement for voting in this state? (Gallup 230, 232) 3. Should negro and white soldiers serve together in all branches of the armed forces? (Gallup 270). 4. Do you favor or oppose a law in this state which would require employers to hire a person if he is qualified for the job, regardless of his race or color? (349K) Variant 1: Would you favor or oppose a state law which would require employees to work along side persons of any race or color? (349T) 5. “Have you heard or read about the ‘civil rights’ program recently suggested by President Truman?” If yes: “How do you feel about Truman’s civil rights program. Do you think Congress should or should not pass the program, as a whole?” (Gallup 414) 6. “One of Truman’s proposals concerns employment practices. How far do you yourself think the federal government should go B?” (note: treat “leave up to states” volunteered response as missing here; results same if treat that as anti). Note that did not weight yet. (Gallup 414)

3/48 (Gallup 414)

6. Federal Fair Employment Practices (how far to go)

Full Population 22% Favor 20% Oppose 21% No opinion 36% Not heard of it

Date

5. Pass Truman Civil Rights Program? 3/48 (Gallup 414)

Question Wording

NONSOUTHERN WHITES Landon / FDR / GOP voters Democratic

Table 5, Continued: Civil Rights Policy Views and Partisan Vote Choice, By Region

57

FIGURE 1:

58