Authors’ Response

Abstract

It is a special honor to publish our work alongside rigorous critiques from Dorothy Sue Cobble, Kristoffer Smemo, Eric Schickler, and Devin Caughey. We are grateful for the opportunity to debate our research in the pages of Labor with such an esteemed and interdisciplinary group of labor historians and political scientists. In this memo we will restate our argument, address three common concerns raised by the commentators, respond to specific questions from individual commentators, and reflect on the potential implications of our argument for contemporary American politics.Our article, “Rewarded by Friends and Punished by Enemies,” argues that the CIO's political action committee (CIO-PAC) contributed to an ongoing anti-labor backlash from the Republican Party. The CIO-PAC was founded in 1943 and quickly formed a de facto alliance with the Democratic Party. Although the CIO-PAC repeatedly claimed to be nonpartisan, 94 percent of its endorsements in 1944 went to Democrats. While the CIO-PAC's partisan political engagement helped to attract prolabor support from Democratic members of Congress, it simultaneously pushed the Republican Party further toward anti-labor legislation.The CIO-PAC's opposite effects on the two parties meant that CIO power was associated with increasing polarization; the difference between Democratic and Republican voting on congressional labor legislation was small in districts and states where the CIO was weak, but this difference grew as CIO strength increased. The result was that Republicans were most likely to support anti-labor legislation, such as the Taft-Hartley Act of 1947, when they represented constituencies in which the CIO was at its strongest. As Schickler and Caughey note, this dynamic cannot be explained by most theories of representation, which would expect Republican support for unions to increase with the strength of the CIO in their constituency. In other words, the CIO faced its fiercest enemies in places where we might otherwise expect it to have attracted friends.This polarizing effect of the CIO-PAC, as we stress throughout our article, was not the only cause of the passage of Taft-Hartley. Southern Democrats almost unanimously supported Taft-Hartley on the grounds that the CIO's efforts to unionize the South posed a threat to white supremacy. Around the country, public opposition to the CIO strike wave of 1945–46 encouraged many Democrats and Republicans to support Taft-Hartley. And the rising conservatism of the Republican Party, which had only increased since its late 1930s opposition to the New Deal, made Republicans more likely to support Taft-Hartley than Democrats were. Independent of these important dynamics, we assert that the CIO-PAC's political engagement further galvanized Republican opposition to organized labor in a way that may have contributed to the passage of Taft-Hartley.Our argument is therefore as nuanced as it is controversial, and we are excited to see it inspire difficult questions and critiques. Throughout the commentators’ excellent responses, three main issues were raised repeatedly. First, what was the relationship between the CIO-PAC and the anti-labor animus that motivated large parts of the Republican Party? If one caused the other, our commentators point out, then surely the 1943 creation of the CIO-PAC cannot explain a Republican anti-labor backlash that began in the 1930s. We believe that this common critique is based on a misunderstanding of our argument, and we welcome the opportunity to restate and clarify our claim that “the American labor movement's political strategy coevolved with the labor policy positions of the Republican and Democratic Parties.”Second, how did the CIO-PAC influence Republicans beyond our brief example of New York's 1944 Senate election? While our quantitative analysis does demonstrate a general relationship across the United States, we understand the desire for more qualitative evidence that the CIO-PAC's de facto alliance with the Democratic Party pushed the Republican Party in an anti-labor direction. Below, we briefly discuss how the CIO-PAC's 1946 decision not to endorse progressive, prolabor Republicans in Wisconsin and Minnesota likely contributed to the election of anti-labor Republicans (Joseph McCarthy and Edward John Thye), who voted in favor of Taft-Hartley the following year.Third, was the CIO-PAC responsible for the passage of Taft-Hartley in 1947? Cobble suggests that we unfairly blame the CIO-PAC for the anti-labor conservatism of the Republican Party and the passage of Taft-Hartley—“mighty weights,” she writes, “for the CIO-PAC to bear.” Schickler and Caughey, similarly, write that our article suggests that the CIO-PAC “cost unions more than it gained them” and therefore represented a strategic mistake by the CIO. However, we have no intention of holding the CIO materially responsible for the passage of Taft-Hartley, which was ultimately the result of numerous crosscurrents in American political life. Our hope is merely to draw attention to important negative consequences of the CIO-PAC (increased Republican backlash) that have been overlooked in previous research. Whether or not this negative effect was outweighed by the positive effect of increased Democratic support requires additional research. Similarly, more investigation and debate is needed to determine whether or not an alternative political strategy existed by which the CIO could have captured the benefits of the CIO-PAC without suffering all of its negative consequences.We believe that the CIO's political strategy coevolved with the labor policy positions of the Democratic and Republican Parties. This means that (1) the CIO's decisions regarding the PAC were influenced by the labor policies advocated by the two parties and that (2) the CIO-PAC's informal alliance with the Democratic Party influenced the subsequent labor policy positions of both parties.Regarding the first point, our article clearly explains how the labor policies supported by Democrats and Republicans in the late 1930s and early 1940s influenced the CIO's decision to create a PAC and almost exclusively to endorse Democrats. The prolabor legislation supported by President Roosevelt and the Democratic Party during the New Deal of the 1930s made a CIO alliance with the Democrats an appealing option. As we note in the article, the roots of the CIO-PAC's alliance with Democrats can be traced back to the CIO's Non-Partisan League of 1936, a short-lived organization that campaigned for FDR's reelection after realizing that “a willing Roosevelt administration offered labor a chance for a new kind of political advocacy.”Just as Democratic support for prolabor policies attracted the CIO, growing Republican opposition to organized labor made a CIO alliance with the GOP all but impossible. As we note in our article, the CIO's 1942 report on the possibility of creating a political action committee concluded that unions “should not pretend that there is the slightest possibility of our achieving genuine influence in the Republican Party” given that “the Republican Party had reconstituted itself on anti–New Deal grounds, making it an unlikely partner.” We also explicitly noted that before the CIO-PAC, “the GOP joined forces with the business community in an increasingly aggressive backlash against the New Deal in the late 1930s and 1940s. This renewed alliance paved the way for an ideological assault on labor rights in the name of economic freedom.” In these ways, we agree with the commentators and the conventional wisdom that “the conservative counterreaction against labor was already well underway in Congress by the late 1930s, with overwhelming support from Republicans.” As we write, “Of course, the Republican Party's move toward anti-unionism was not solely driven by the CIO-PAC's informal alliance with the Democratic Party.”However, the Republican Party's general shift toward anti-labor policies before 1943 is only part of the story of Republican voting on Taft-Hartley in 1947. Most important, this rising conservatism cannot explain the variation that continued to exist among Republicans on questions of labor legislation. As Smemo discusses, the vote on Taft-Hartley exposed a rift within the Republican Party; while some “yearned to gut New Deal labor law,” other members of the GOP “could not imagine a direct confrontation with the organized labor movement.” Yes, the GOP was broadly moving to the right of the Democratic Party on labor issues, but many Republicans still voted against anti-labor legislation throughout the Eightieth Congress (1947–48).This is where the second part of our coevolutionary argument comes into play: the CIO-PAC's informal alliance with the Democratic Party influenced the subsequent voting behavior of both political parties. For Democrats, support for unions (voting against Taft-Hartley) increased with the strength of the CIO in their constituency. For Republicans, by contrast, support for unions decreased with the strength of the CIO in their constituency. As our interlocutors emphasize, the overall impact of the CIO-PAC on congressional voting was therefore to increase polarization between the two parties.Thus, we fully agree with Smemo that “the anti-labor reaction in Congress began shortly after the 1936 elections” and with Schickler and Caughey that “the CIO-PAC formed in 1943 largely in reaction to the consolidation of this new conservative bloc.” This earlier conservative backlash contributed to the large difference between Republican and Democratic voting on Taft-Hartley; model 4 in table 1 shows that Democrats, on average, were less likely than Republicans to support Taft-Hartley. However, such general partisan differences tell us nothing about why Democrats and Republicans voted similarly to one another in states and districts where the CIO was weak, but voted so disparately in places where the CIO was strongest. We can better understand this polarization, we argue, if we explore the ways in which the CIO-PAC made friends of Democrats and enemies of Republicans.Caughey and Schickler note that the core qualitative evidence for the mechanism presented in our article is an analysis of the New York Senate race in 1944, in which the Republican Party, confronting the CIO-backed Democrat Robert Wagner, decided not to nominate the relatively prolabor candidate Irving Ives in favor of avowed anti-communist Thomas Curran. The reasoning for this decision, we contend, was significantly rooted in the party's reaction to the presence of the CIO-PAC, which had made it difficult for prolabor Republicans to rely on traditional nonpartisan union support and had convinced party leaders to instead shift toward a more explicit anti-union (in this case, anti-communist) stance.As Caughey and Schickler point out, that was not the end of the story, nor did this particular moment signify a permanent break between Republicans and labor. Ives was nominated in 1946 and defeated former governor Herbert H. Lehman in the Republican wave of that year, signifying that the party as a whole was not quite ready to fully commit to a strategy of polarizing the electorate.Yet as Smemo's response to our article makes plain, this partial retreat from a stridently anti-labor platform did not necessarily imply a concomitant shift toward labor. Instead, it reflected a complicated negotiation among the factions of the Republican Party to take advantage of the splits in the labor movement as well as find a way to reconcile hardline business conservatism with the more accommodationist interests in the party. But this move almost inevitably involved shifting the priorities away from explicit alliance with the unions themselves and instead toward making labor legislation more favorable to management.1 After all, despite the endorsement of the AFL, Irving Ives did indeed vote for the Taft-Hartley Act. Similarly, Smemo points out that Ives's Fair Employment Practices Commission was “notoriously weak” and that he championed an “industrial pluralism to keep industrial unions subordinate to management”; by comparing Ives to Curran, our article may have overstated Ives's prolabor credentials.More broadly, however, Cobble points out that the class of 1946 included conservative Republican senators from states without a particularly strong CIO presence, some of whom (like Harry Cain and James Kem) would also vote for Taft-Hartley.2 We do not dispute that the Republican Party as a whole would have “jumped at the chance to eviscerate the Wagner Act in 1947.” What we contend is that in places where the CIO was strong, the Republican response to the PAC was particularly negative. For instance, twelve Republicans either won open seats or beat incumbent Democrats in the 1946 elections; of these, seven were in states with significant CIO membership. At the same time, rather than accommodating labor or moderating their party's pro-Taft-Hartley position, all seven voted for the bill.3 To conventional theories of interest group politics, such an outcome is very puzzling: it appeared that by 1946 in many of those states we would expect the CIO's voice to be loudest and most persuasive, strongly anti-labor Republicans were instead elected. While we agree that a general conservative backlash in public opinion (as well as Truman's more general unpopularity and the onset of the Cold War) partly explains this pattern, we contend that the threat posed by the CIO-PAC to Republicans seems to have played a key polarizing role.Take, as another example, the case of Senator Joseph McCarthy, who defeated the progressive incumbent Robert La Follette Jr. in the Wisconsin Republican primary in 1946 and went on to secure one of the state's Senate seats. Cobble is correct that Wisconsin was not the most favorable state for CIO political activism (although it was in the 70th percentile in terms of CIO density among US states) and that McCarthy's conservatism seems particularly ideological. However, the CIO was particularly active in the Wisconsin's urban centers and actively worked to undermine La Follette's campaign by attacking him for being insufficiently committed to the New Deal.Many historians have attributed the CIO's animosity toward La Follette to a communist faction of the union active in Milwaukee, which was ideologically opposed to his stance against the USSR. But historian David Oshinsky convincingly demonstrates that even “the anti-Communists” in the CIO “were committed firmly to the Democratic party and pictured La Follette's return to the GOP as the Waterloo for liberal unity in Wisconsin.”4 Thus, Howard McMurray, the Democratic nominee for Senate in Wisconsin, engaged in a calculated effort to appeal to CIO partisanship by tying La Follette—a consistent supporter of many of the New Deal labor efforts, even though lukewarm at times in his support for Roosevelt—to “‘calculated subversion’ of the liberal forces in Wisconsin.”5 Throughout the campaign, McMurray and the labor press worked in tandem to attack La Follette and shift the labor vote toward the Democrats, a strategy that largely reflected the national CIO-PAC's own priorities.Moreover, the attack was successful. La Follette's vote share plummeted in labor wards (particularly those with strong CIO presence) in cities like Kenosha and Milwaukee during the primary in 1946, while the Democrat's rose relative to 1940, when La Follette had last been on the ballot,6 even though the AFL had directly endorsed La Follette's candidacy. As the New York Times opined in 1946, “The CIO Political Action Committee forces contributed somewhat to Mr. La Follette's defeat,” as the CIO-PAC “urged the workers to vote in the Democratic primary.”7What lessons, then, might Joseph McCarthy have learned from La Follette's defeat at the hands of the CIO? Oshinsky notes that at the beginning of the Republican primary, “McCarthy employed a relatively moderate posture in relation to labor-management problems.”8 He even made a few appeals to the labor vote himself, arguing that “I do not blame labor for strikes and stoppages” and decrying industry leaders who had “not yet learned that labor unions are as much a part of the American way of life as industry itself,” pinning some of the disorder on them.9 As the campaign dragged on, however, and he witnessed the effect of the CIO-PAC's campaign against La Follette, McCarthy seems to have become increasingly convinced of the political benefits of tying labor directly to concerns about communism. Thus, he claimed that the union strategy of concerted labor action “played along with the Communists who know that their theories will grow best in an economy of industrial discord and strikes,” and he accused McMurray of being “little more than a megaphone . . . being used by the Communist-dominated Political Action Committee in Milwaukee.”10 In other words, though McCarthy was never particularly sympathetic to the interests of organized labor, it appears that he also saw no reason to make any real concessions to the demands of industrial unionism. Instead, he too went on the attack.Other progressive Republicans also suffered from the CIO's shift toward partisanship in the 1946 election, even when they avoided the direct ire of the PAC. Henrik Shipstead, a former member of the Farmer-Labor Party in Minnesota who had bolted to the Republicans in protest of Roosevelt's increasing internationalism in 1940, had long been considered a friend to labor and had been a key member of the state's burgeoning liberal coalition.11 In 1935, Shipstead had voted in favor of the Wagner Act. However, by 1946 the CIO had largely abandoned Shipstead for partisan reasons, paving the way for his conservative Republican rival, Governor Edward Thye, to secure the party's nomination and defeat Theodore Jorgenson in the general election. In his nationally syndicated column, respected journalist Thomas Stokes noted that despite the fact that “old-line labor” groups such as the “railway brotherhoods and the AFL” would back Shipstead in the state's primary, the CIO “probably will vote in the Democratic Farmer-Labor primary where there is a better fight for control among the CIO, left-wing and Irish Democratic elements.”12 In this case, the partisan shift induced by the CIO-PAC created less incentive for Republicans to cater to labor precisely because they increasingly realized that fewer industrial unionists were willing to participate in Republican politics at all.In other words, we agree with our interlocutors that the story did not end with Irving Ives's failed renomination in 1944. But the partisan dynamics unleashed by the CIO-PAC had a deep effect on a series of crucial Senate races in the following years and undoubtedly made it much easier—and, in some cases, desirable—for the Republican Party to shift even further away from an alliance with organized labor.It is worth noting that the McCarthy (R-Wisconsin) and Thye (R-Minnesota) examples also help to illustrate a slightly different mechanism through which the CIO-PAC may have contributed to the Republican backlash against organized labor. In both cases, the CIO-PAC withheld its support from progressive Republicans senators and helped to tilt the outcome of Republican primaries toward anti-labor candidates, both of whom went on to win general elections in 1946 and vote in favor of Taft-Hartley in 1947. If the CIO-PAC had supported prolabor Democrats while also endorsing prolabor Republicans such as La Follette and Shipstead, the composition of the Republican Party in 1947 may have been less anti-labor.At the very least, it is easy to imagine that CIO-PAC support for La Follette (R-Wisconsin) and Shipstead (R-Minnesota) could have secured their primary and general election victories, as well as two Senate votes against Taft-Hartley. We are therefore grateful to our commentators for pushing us to further explore the effect of the CIO-PAC on Republicans outside New York. There is ample room for new research on the 1946 election of anti-labor Republicans in Massachusetts, Missouri, Ohio, Pennsylvania, and Utah—five states in which the CIO was well organized yet Republicans won Senate campaigns and subsequently voted in favor of Taft-Hartley. Scholars should examine whether and, if so, how the CIO-PAC influenced Republican primaries and the subsequent labor policy positions of Republicans in these states.At the same time, although we argue that the CIO-PAC played an important and previously overlooked role in the politics of Taft-Hartley, we certainly do not contend that the CIO should be blamed for the entirety of the GOP backlash or the act's final passage in 1947. The argument we present in our article is more nuanced and less radical than the interpretation presented by our commentators. We wrote that the CIO-PAC “contributed to a backlash from the Republican Party that culminated in the passage of Taft-Hartley.” As explained above, we recognize that the Republican backlash against organized labor began before the CIO-PAC, hence our argument that the CIO-PAC merely contributed to a backlash. Similarly, we believe that it was this Republican backlash (with numerous causes, including but not limited to the CIO-PAC) that culminated in the passage of Taft-Hartley. To be clear, we do not believe that the creation of the CIO-PAC and its informal alliance led, on its own, to the GOP backlash or the passage of Taft-Hartley.That being said, we do argue that the partisan polarization associated with the CIO-PAC raises important and unresolved questions about the risks and rewards of electoral politics. The commentators offer fascinating cost-benefit analyses of the CIO-PAC's alliance with the Democratic Party and conclude that the CIO gained more from Democrats than it lost from Republicans. Cobble argues that the CIO-PAC “energized the left more than the right” and that “the ‘frontlash’ effect [from Democrats] of the CIO appears greater than the ‘backlash’ [from Republicans].” Similarly, Schickler and Caughey argue that “it was only because the CIO was so closely tied to the Democratic Party that it avoided anti-labor legislation as long as it did.” In these ways, the commentators appear to accept our main argument about the negative aspects of the CIO-PAC (i.e., it contributed to the Republican anti-labor backlash), while maintaining that these costs were clearly outweighed by the positive benefits.Perhaps. But we believe that the debate is not quite this simple. Even if the benefits of the CIO-PAC outweighed its costs, it remains possible that an alternative political strategy could have captured most of these benefits without suffering all of the same costs. As Schickler and Caughey note when discussing our finding that Republicans were increasingly anti-labor as CIO strength in their constituency increased, the only analogous dynamic in American political history is Key's “racial threat” hypothesis: that whites become increasingly racist when the local percentage of Blacks increases. Could the CIO have found a way to support prolabor candidates (even if most of them were Democrats) without triggering a backlash reminiscent of the South's defense of white supremacy? In other words, just because the benefits of the CIO-PAC outweighed its costs does not mean it was the best possible strategy.One possible alternative strategy was the AFL's nonpartisan political engagement—Gompers's famous call to “reward our friends and punish our enemies” regardless of their political party. Could the CIO have continued this nonpartisan tradition in a way that garnered the benefits of the CIO-PAC without suffering all of its negative costs? In our article, we offered two reasons to doubt the efficacy of the AFL's approach. First, a large secondary literature concludes that the AFL had little impact on the formation or passage of prolabor legislation during the early twentieth century.13 Second, our own analysis of congressional voting on the Wagner Act found that AFL union density was not associated with support or opposition from Democrats or Republicans. As we concluded, the AFL “neither garnered support nor inspired opposition from either party.”Cobble disagrees, however, and thereby opens space for exactly the kind of debate we hope our work will inspire. She argues that “the AFL turned to national politics with a vengeance” in the early twentieth century and “had some success” with the 1932 prolabor Norris-LaGuardia Act, thus “paving the way for the Wagner Act.” But if the AFL's nonpartisan politics contributed to such important legislative successes, perhaps the CIO could also have built support for prolabor legislation without forming a partisan alliance with the Democratic Party. Was there a political strategy between the AFL's nonpartisan politics and the CIO-PAC that could have garnered support from Democrats without simultaneously generating backlash from Republicans?Such a debate about the political strategy of the CIO would mirror a well-known debate about the CIO's economic strategy (e.g., strikes) during the 1940s. Seidman, Schickler and Caughey, and others have argued that the strike wave at the end of World War II led to a public backlash against the CIO that contributed to the passage of Taft-Hartley in 1947.14 The CIO was too confrontational, the argument implies, and a less-aggressive strategy might have captured important economic gains without triggering the backlash that culminated in Taft-Hartley. In sharp contrast, Lichtenstein (2003) argues that organized labor's eventual decline was driven by a mistaken lack of CIO militancy during World War II. Rather than strike for social democracy and a better society when the opportunity arose, CIO leaders accepted a no-strike pledge in exchange for union security measures that secured the union shop and built large unions. In a narrow sense, this bargain may have benefited the CIO more than it cost them, but Lichtenstein suggests that an alternative strategy may have garnered many of the same benefits (or more) without suffering the same costs.Beyond the three common concerns discussed above, the commentators posed a number of questions that demand replies. This section reflects briefly on these additional issues.We were very pleased to read that Schickler and Caughey found our state- and district-level measures of CIO density to be both innovative and clever. They are correct that the only other data on district-level CIO membership comes from William Riker's 1948 doctoral thesis. Riker's data is available for only 159 congressional districts, whereas our approach generates estimates for 429 congressional districts. The two measures are relatively similar; for the 159 districts in Riker's data, the correlation with our measure of CIO density is 0.69.Our regression results are also similar when using Riker's data, despite the relative decrease in variation available (we code Riker's missing districts as 0). When we examine House voting on Taft-Hartley (replicating model 1 from table 1), we find a negative and statistically significant interaction term between Democrat and CIO. This means that the difference between Democratic and Republican voting on Taft-Hartley (partisan polarization) increased along with the district-level strength of the CIO. The main difference we find using the Riker data is that the constitutive effect of CIO is no longer positive and statistically significant. That is, when one uses the Riker data, CIO density is not associated with greater Republican support for Taft-Hartley.With this in mind, it is worth noting that our article may have set too high a bar for testing our argument about the CIO-PAC and the Republican backlash. Should we only conclude that the CIO-PAC had negative consequences if it led Republicans to support Taft-Hartley? What if the CIO-PAC only resulted in a neutral relationship between CIO density and Republican voting on Taft-Hartley? Even that null finding (as we mention above, using the Riker data) suggests that Republican members of Congress did not respond to the CIO in the way most theories of representation would predict; Republicans in districts with stronger CIO constituencies were no more likely to vote against Taft-Hartley then were Republicans in districts with weak CIO constituencies.Although Taft-Hartley was supported by the overwhelming majority of Republicans in Congress, divisions within the party continued to exist; thirty-one Republican members of the House and four Republican members of the Senate withheld their support from the final vote on Taft-Hartley. In his insightful commentary on the dialectical relationship between Taft-Hartley and the Republican Party, Smemo notes that this division between “old-guard” and “liberal” Republicans extended into the subsequent decades. Smemo explains that “while the GOP ‘old guard’ yearned to gut New Deal labor law, the party's ‘liberal’ wing could not imagine a direct confrontation with the organized labor movement.”With this welcome call for a longer time horizon, Smemo highlights the need for more research on the dynamic relationship between the CIO-PAC and the Republican Party. Our article demonstrates that in 1947 Republicans with strong CIO constituencies were more likely to support Taft-Hartley than Republicans with weak CIO constituencies. Did this relationship continue into the 1950s, driving a wedge between Republicans? By contrast, Smemo suggests that the divisions within the GOP may have been the exact opposite: it could be that “Republicans in heavily unionized states” were the least supportive of anti-labor legislation such as the right-to-work laws permitted by Taft-Hartley.Of course, answering this question will require further empirical research on congressional voting patterns in the 1950s.15 If the dynamic we highlight in our article holds during this period, then the polarizing effect of the CIO-PAC (attracting Democrats but repelling Republicans) may help to explain partisan polarization over a longer period of time. By contrast, if Smemo is correct that the relationship reversed in the 1950s (as Schickler and Coughey note that most theories of representation would expect), then this would represent a fascinating puzzle for future research.How did the CIO-PAC go from repelling Republicans in 1947 (similar to Key's “racial threat” hypothesis) to attracting them in the way we would expect organized interest groups to do? Did the CIO learn from 1947 (when many Republicans from strong CIO states voted in favor of Taft-Hartley) and recalibrate its political strategy in a way that maintained Democratic support without antagonizing Republicans? If so, understanding such strategic changes would further inform debates about the creation of the CIO-PAC and the passage of Taft-Hartley: Could the CIO have pursued such a political strategy earlier and thus avoided the passage of Taft-Hartley?Exploring the effects of the CIO-PAC on both the Democratic and Republican Parties, as well as the more general consequences of labor's shift away from nonpartisanship, is essential to understanding the historical development of electoral politics in the postwar United States. Taft-Hartley was not only a setback for labor; it also heralded a relatively novel political role for the unions as part and parcel of the Democratic coalition. Moreover, we believe it is crucial to situate organized labor and party leadership in a coevolutionary relationship. This reminds us that actions taken for one goal (exploiting opportunities within a party organization; mobilizing constituents in partisan rather than interest-based terms) can have inadvertent by-products (polarizing opponents and creating backlash). The serious and engaged responses of our interlocutors to our argument about the centrality of political strategy on the part of the CIO also reveals the methodological advantages of cross-disciplinary exchange and the value of exploring specific events to shed light on longer-term developmental trajectories.16Our work on the passage of Taft-Hartley in 1947 may even hold lessons for contemporary American politics. Joe Biden's promise to be “the most pro-union president you've ever seen” is clearly meant to echo FDR and the New Deal, when organized labor first began its evolving alliance with the Democratic Party. Today as in the 1940s, the Democratic Party is clearly the one with which unions have the strongest support. Yet our research suggests that labor should be careful with its partisan commitments, mindful of the Republican Party inevitably returning to power in the coming years. And while today's GOP no longer contains anything like the progressive Republicans of the 1930s, there still may be some potential for Republicans to support organized labor.In 2020, five Republican members of the House of Representatives voted in favor of the Protecting the Right to Organize (PRO) Act, the most sweeping prolabor legislation debated in Congress in a generation. And while the ethno-nationalism of the Trump era proved disastrous for the increasingly diverse labor movement, there is potential for the party's populist faction to support labor unions moving forward. Oren Cass, a Republican strategist and self-styled “pro-union conservative,” frequently claims that several Republican senators share his view that “America needs a conservative labor movement.” Although labor unions would be justified in working hard to defeat most Republican senators, it is also easy to imagine that Cass's Senate allies could lose primaries to more-conservative Republicans who would then prove to be worse enemies of organized labor.Allying with one party can make the successful pursuit of prolabor policies subject to the same feast-or-famine dynamics characterizing other policy domains in contemporary American life. A robust labor movement must be capable of defending its gains regardless of who holds power in Washington. This may only be possible, however, if unions are willing to remain independent from a Democratic Party that itself has increasingly abandoned the language of class and solidarity among working people. Our story of the CIO-PAC and Taft-Hartley suggests that today's labor movement would be wise to find ways to embrace Democrats without alienating potential allies in the GOP.If such a balancing act proves difficult—as it did for the CIO-PAC in the 1940s—it may speak more to the constraints of the two-party system than to any failure on behalf of the labor movement. As Marx wrote, “Men make their own history, but they do not make it as they please; they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past.”