Congress may enact a law on a controversial matter, but will the decision be smoothly implemented? Routinely, the answer is no. Further iteration or dispute may ensue. In cases of intense opposition, the reaction may even entail a repeal drive. In this paper, we consider three American policy controversies that have entailed memorable repeal drives. Those drives have targeted the Eighteenth Amendment (Prohibition), the 14b right-to-work clause of the Taft–Hartley Act, and the Affordable Care Act (the ACA, or Obamacare). We take pains to justify these selections.
We do not address the actual conducting of these repeal drives—that is, who did what, when, and where in terms of process. Our focus is broader. We examine the policy controversies in which these drives have figured. Components in that focus might include relevant historical background, the processes in enacting the statutes that the drives have targeted, the political contexts that underpinned these enactment processes, the cleavage contours of the controversies, and the relevant political byplay (not just the drives) that has extended on in time following the original enacting of the measures. For each case, we also consider the effects on society that the combination of enacting a statute and strongly reacting to doing so has brought. Our guiding premise is that a policy controversy that engenders a resort to the extreme weaponry of a major repeal drive is a phenomenon that merits special examination. It is politics at an extreme.
From three cases, what, in principle, can be learned more generally? This question is especially apt if the cases are extreme, which ours are. But perhaps there can be payoff. Our entry into the matter has been to search for commonalities across the three cases of controversy. In principle, commonalities on a matter may be of interest in themselves, yet they may also radiate implications beyond their immediate domains. Commonalities are plausible platforms for more general illumination or explanation. Accordingly, this paper offers a detailed presentation of commonalities across our three cases. Our enterprise is very much an inductive one. We invite readers to examine the commonalities we report and to reflect on what to make of them. We present our own reflections and conjectures. In historical, theoretical, or practical terms, what lessons seem to be available about policymaking in the American system? Also, how well have existent scholarships been addressing the kinds of questions that we raise?
Our search for commonalities took us through an assortment of academic literatures. The yield may be seen in our citations, which include lists of general works as well as scattered references to special topics. We also spent time reading congressional hearings and debates—on Prohibition in 1914, 1917, 1918, and 1919; on the Wagner Act in 1935; on the Taft–Hartley Act in 1947 and 1949; on the ACA in 2009 and 2010.Footnote 1
Distinctive in their own terms our controversies may be, they plainly share characteristics with other historical cases. Ready at hand are those having to do with race, slavery, and civil rights, or with North versus South. For one thing, the eventual fate of the Eighteenth Amendment bears more than a passing resemblance to the generations-long non-enforcement of the Fifteenth Amendment. But we leave aside this set of topics. It has exhibited its own pattern of geography, it has drawn on constitutional rights as a major consideration, it has not engendered repeal drives (the politics has worked differently), and a vast scholarship has addressed it. Besides race and section, several other U.S. policy controversies since 1789 have exhibited intensity, prominence, longevity, geographic cleavage, or some combination of these features. Certain of the insights in arguments of this paper may thus pertain elsewhere.
The selection of our three repeal drives requires justification. But first, a factual refresher on the drives and the enactments that triggered them. All three drives and their surrounding controversies have been associated with especially far-reaching regulatory designs that Congress has voted since 1900—a date that seems a decent proxy for the launch of the modern administrative state. The histories are complex. The Eighteenth Amendment to the Constitution banning the production, transport, and sale of intoxicating liquors was debated in the House in 1914, approved by Congress in 1917, ratified by the states in 1919, supplemented by an enforcement measure, the Volstead Act, in 1919, and formally repealed via the Twenty-First Amendment in 1933. The plan was to be jointly enforced by the national and state governments.Footnote 2 Congress enacted the Taft–Hartley Labor Management Relations Act in 1947. A full-steam move to repeal the entire act failed in 1949, but that drive soon settled into a thus far always losing campaign to repeal the act’s 14b clause, which guarantees the states the authority to enact right-to-work laws, a headache for labor unions. But to address Taft–Hartley requires also an inspection of the earlier regulatory measure against which it was itself a reaction—the National Labor Relations Act of 1935 (i.e., the Wagner Act). We conduct that inspection here.Footnote 3 Enactment of the ACA in 2010 brought a repeal drive through 2017 ending in a showdown failure at that time.Footnote 4
Many political scientists or historians might readily endorse our choices of these repeal drives. But surer ground is needed. One good supportive source is Jordan M. Ragusa and Nathaniel A. Birkhead, Congress in Reverse: Repeals from Reconstruction to the Present.Footnote 5 These authors’ topic is repeals of congressional statutes. Repeals are not our interest. Repeal drives (whether or not they have succeeded) are our interest. But, in passing, these authors present a time series of use here.Footnote 6 Tracking citations in the historical New York Times, their work counts newspaper articles in 6-month blocks from 1877 through 2017 that exhibit jointly the terms “Congress” and “repeal.” Leaving aside the entries from before 1900, there are five tall spikes. They jut up distinctively. None of the smaller spikes in their Figure 1.1 reaches half as high as any of these five. One of them can be set aside—that for 1939 involving repeal of the Neutrality Act, a showdown over foreign policy triggered by war in Europe. The other tall spikes are for 1930, 1949, 2017, and for an 18-month span covering 1932 and the first half of 1933. The authors credit the 1930 entry to debates over repealing the Volstead Act and the Eighteenth Amendment; 1949 to the failed repeal of Taft–Hartley; 2017 to the failed repeal of the ACA.Footnote 7 We checked out the 1932–33 entry ourselves in the New York Times source. That juncture’s tall spike draws from more than 700 articles reporting on the repeal of the Volstead Act or the Eighteenth Amendment, or the legalizing of “near beer,” or the passage of the Twenty-First Amendment.Footnote 8 (No other topic receives more than a dozen or two mentions.) Also, these authors report that one of their smaller spikes, that for 1965, draws partly from that year’s Taft–Hartley repeal effort. In sum, our selection of cases receives help from the Ragusa–Birkhead data. Accepting a restriction to domestic policymaking, their data help isolate our threesome of instances as an exhaustive set of major repeal drives undertaken since 1900.
Another justification for our case selection is as follows. In each of the three cases, one of the two major parties has called for repeal in one or more of its quadrennial platforms. The Democrats targeted the Eighteenth Amendment in 1932, and also the Taft–Hartley Act in whole or in part nine times between 1948 and 1980, and the Republicans targeted the ACA in 2012 and 2016. Several dozen repeal calls regarding other measures have appeared in major-party platforms since 1900 (see Appendix A), but the bulk of these have been forgotten, and none have figured in drives as prominent—as suggested, for example, in the Ragusa–Birkhead data—as the ones taken up here. The count of platform mentions on Taft–Hartley is especially remarkable.
The politics surrounding these three drives has differed a good deal in its partisan texture and its rhythms. The Eighteenth Amendment had a nonpartisan origin, but the Democrats weighed in at its demise in 1932 and 1933. The Republicans engendered the Taft–Hartley Act, but most House Democrats (notably, the bulk of the party’s southerners) voted to pass it. The politics of the ACA has been resolutely partisan—the Democrats generated the measure; Republicans have opposed it.
As for rhythm, sentiment to repeal the Eighteenth Amendment grew gradually during the 1920s, then soared in the early 1930s. After 1949, the drawn-out campaign to repeal 14b featured a congressional showdown in 1965, then a lapse. The Republicans’ anti-ACA drive proceeded at a steady temperature until 2017, peaked in prominence then, but fell away after the drive’s Capitol Hill failure. Still, the ACA story remained unfinished as of 2025, given President Donald Trump’s stated aims.Footnote 9
See Figure 1 for the timelines of the three repeal histories. We use ProQuest counts of mentions in media articles that include the term “repeal” plus a relevant topic focus.Footnote 10
Number of Articles with Repeal and Terms Related to Prohibition, Section 14b, and the ACA.
The flow of the paper is as follows. We present a series of five observations drawing on the commonalities that we have found to obtain across each of our controversies. We comment as well as describe as we go along. (The second of these observations is especially long and is divided into three parts.) The topics of the observations are geography-inflected conflict, Congress’s deliberative practices, crisis-driven decision-making, the states’ reaction to federal regulation, and the impacts of the controversies on society. We preface each of the five observations with a brief statement in bold that provides the gist of what follows.
1. All three cases have featured severe, high-publicity conflict that broke along trademark geographic lines
Geography plus intensity can invite a signal pattern of politics in a federal system where the lower constitutional units enjoy leeway as centers of authority and voice. Consider Texas in this country, or Alberta or Quebec in Canada.Footnote 11 The American states may legislate, bring lawsuits, foot-drag bureaucratically, stage kickback referendums, send members to Congress who invigorate repeal drives, or otherwise oppose national policies. Also, interacting with the relevant constitutional structures, the size and heterogeneity of a country may serve as commotion magnifiers. The United States seems a decent match to these ideas.
Conflicts over Prohibition, right-to-work, and the ACA have divided the U.S. geographically, and, beyond that, they have divided it in a similar configuration. One rendition of those divisions is the set of maps in Figure 2.Footnote 12 Each of the three policy controversies is recorded here in binary fashion. Colored in gray for Prohibition are the thirty-two states that had enacted their own dry (i.e., prohibition) statutes (they were free to do that) at some point before the necessary pivotal state ratified the Eighteenth Amendment bringing it into effect in January 1919. This is a plausible indicator of enthusiasm for the dry cause at that time.Footnote 13 Colored in gray for right-to-work is a measure whose values draw from an expanse of time—the twenty-six states that had right-to-work laws on their statute books as of 2023.Footnote 14 Colored in gray for the ACA are the thirty-five states where majorities of the public responded against the ACA in Kaiser tracking surveys that asked for up-or-down judgments just after the enactment in 2010–2011. New York and Utah supplied the extremes of these judgments at 40.5 percent and 65.0 percent unfavorable, respectively.Footnote 15
State Positions on Questions of Prohibition, Right-to-Work (RTW), and the ACA.
This is one work-up of maps among a possible many. In each of these cases, we aimed for a cut-point in the data that is binary and cleanly documentable (as in yes versus no), and that affords a picture of the overall shape of each controversy. These were persisting conflicts that, allowing for ups and downs, could trend either way. Beyond our selections, certain alternative cut-points are possible. Some options exhibit lopsidedness—for example, in the case of the ACA, the twelve jurisdictions (including the District of Columbia) that operated their own state-based marketplaces as of 2018, versus the thirty-nine that did not; or, on the opposite side, the ten states that held out against Medicaid expansion as of February 2024 versus the forty-one jurisdictions that had signed on.Footnote 16 But an aim for overall balance seems to make sense.
For our data, we sought renditions of public views. Those might include decisions by elected officials (an indirect rendition of views through party representation), decisions in state referendums, and responses in opinion surveys. We saw all these sources as usable, if unevenly available. It would be pleasing, for example, although impossible, to have opinion data on Prohibition in 1918 or so. These reaches into public views all have their wrinkles and question marks. There can be dissonances across them, as seen in, for example, the decisions by partisan majorities in state legislatures (and in Congress) versus the views of median voters in their domains. (More on this later in the paper.) But all these sources are windows into public views.
Finally, the binary measure for right-to-work poses an awkwardness. No readily intuitive cut-point works for the early years of Taft–Hartley. Although the bulk of the eventual sign-ons to right-to-work occurred before 1960, some have occurred more recently. We use all of them through 2023. That makes for a contrast with the Prohibition and ACA measures, which site all their data at the times of their statutes’ original enactments. We see this awkwardness as an acceptable price to pay for gauging right-to-work in a cleanly intuitive way (yes or no).Footnote 17 In fact, all three controversies have seen constant maneuvering for years downstream of their origins—notably, by states that have newly signed on or signed off at various times to this or that feature of an original program for this or that reason.
There is intricacy in Figure 2’s maps. Certain entries in the Prohibition map, for example, likely owe to the views of German Americans in the Midwest who did not favor drinking curbs and to the presence of liquor industries in Kentucky, Wisconsin, and Missouri. New Orleans was notoriously wide open. But there is commonality across the maps. In the country’s life, it comes as no surprise when Massachusetts, New York, Illinois, and California align on one side of things and Alabama, the Dakotas, Texas, and Utah on the other. Why might that be? For reasons of culture, or whatever matters it is that generate the consistency across the maps, an urban-versus-rural dimension is evident in each of them.Footnote 18 Signs of that dimension appeared in our Congressional Record canvasses—as in pro-drys sneering at cities, a pro-wet sneering back at yokels, and anti-ACA members complaining of antirural discrimination. In each case, to report relevant calculations, the rural percentage of a state’s population in a preceding national Census (those of 1910, 1940, and 2000) correlates at a .01 level of significance with the policy expression exhibited in a map.Footnote 19 The gray side on the maps is the rural side. In the case of right-to-work, one multivariate cross-state study offers a grainier account. Through 1991 (the study necessarily misses a few later adoptions), presence of state right-to-work laws is said to have owed appreciably to three variables: percent African American population (generally high in the South), percent of small as opposed to large businesses, and conservative “citizen ideology” as measured in opinion surveys.Footnote 20
What does this discussion say about our controversies? As we see it, there is a likely interaction effect. The country’s system of formal vertical federalism, which is rich in political options, has interacted with the country’s urban-versus-rural differentiation to help spur and shape these three prolonged controversies. Generally speaking, the urban-versus-rural cleavage in American politics could use more emphasis. Culture and conditions of life are entailed. Across time, urban-versus-rural may deserve a rank in overall importance somewhere near Democratic-versus-Republican or North-versus-South. The relations among these three cleavages have varied. It is worth noting that party does not seem to have played any role in Congress’s original handling of Prohibition.
The geography-inflected patterns on view here invite certain questions. What went on in the congressional deliberations? What connections were drawn? What kind of attention did Congress give to decentralization, as opposed to centralized standardization, as a possible way of handling policy? Given the powers and resources of the states, how deft was Congress in addressing the ins and outs of vertical federalism in its deliberations? We mean “address” in the specific sense of congressional hearings and debates, the forums where, at least in principle, proposals are presented, explained, tested, and defended before the public.Footnote 21 An observation accordingly follows.
2. Generally speaking, Congress’s addressing of decentralization and vertical federalism in these cases has been sparse and ragged
Many warrants exist for making policies that are nationally uniform: election mandates, nation-centered democracy, science, moral imperatives, destiny (“history is with us”), efficiency, clarity, crisis, constitutional rights (as with the Reconstruction amendments), constitutional grants of authority (as with the Commerce Clause), negative externalities caused by state governments jostling against each other, not to mention common sense.
But locating policymaking at lower levels of government can have advantages, too. They range from the theoretical to the practical. A list might include the following. Democracy—in the Tocquevillian sense that public participation and choice at lower levels can define and energize a society.Footnote 22 States’ rights—the traditional American doctrine. Information—in the Hayekian sense that central decision-making can lack the nuance needed to address local circumstances. It can be “irrational when applied to particular places,” as Martha Derthick has argued.Footnote 23 Experimentation—the “states as laboratories” idea associated with Justice Louis Brandeis.Footnote 24 Maximizing the satisfaction of people’s policy tastes—the idea that one-size-fits-all policymaking can work out worse in satisfying an overall schedule of individual views than does an allowed variation in local policies. In a way, this idea is simple but it has theory behind it.Footnote 25 Cooperation and compliance—the difficulty that courts, lower governing bodies, or the public itself may idle, block, or balk in reaction to national policy decisions.Footnote 26 Conflict reduction or avoidance—a premise sewn into this country’s states’ rights origins, and basic to the governing of geographically cleaved countries like Canada and Belgium.
We take up one by one the enactments that spurred the three repeal drives. In the debates and hearings, how were questions of decentralization and vertical federalism addressed? What do we see in the way of attention to cleavages—as in urban versus rural? Other themes are touched on.
Prohibition. In the Prohibition debates, it is easy to see urban-versus-rural dissonance. At one extreme, certain pro-dry members spoke contemptuously of “the slums of our great cities,” “some semicivilized foreign colony in New York City,” the “gamblers, hoodlums, and harlots” in an antidry parade in Chicago, and the foreign-born population of Detroit failing to “bow to American ideals.” A Kentuckian spoke favorably of “rural communities and … the central, southern, and western parts of our country, with the population made up almost entirely of Americans.” On the other side, an antidry member pointed to “the little jack rabbit, coyote and sage-brush State[s] of the far West.”Footnote 27
As for decentralization or vertical federalism, it cannot be said that the argumentation offered by the antidry side was sparse or ragged. It was robust. Almost all the boxes were checked. Under threat was self-government or states’ rights, although the states’ new initiative and referendum processes of the Progressive era brought little mention as exercises of democracy. Over a dozen antidry members, chiefly southerners, drew on states’ rights doctrine.Footnote 28 Varying state conditions, an ingredient of the “information” idea, drew interest.Footnote 29 Long historical experience with the dry laws of the states drew attention (the wets said they failed; the drys said they worked), with occasional references to the term “experiment.”Footnote 30 Conflict avoidance made an appearance. Enactment of the Eighteenth Amendment “would light the fires of a controversy which would probably rage for a generation.” It would plunge the country into a “vortex of internal strife.”Footnote 31
But the centerpiece of the antidry case was compliance. Could Prohibition be enforced? Some three dozen members of Congress said no. “[T]o enforce it in the cities of the country would end in failure.” “How will you discover the manufacture and sale of these beverages?” “I have seen in Western States whole towns wide open in the face of the State constitution.” Local juries would not convict.Footnote 32 Nine senators or House members noted that the Fourteenth or the Fifteenth Amendment or both had proven unenforceable. Why expect better of the Eighteenth?Footnote 33
Certain arguments tending toward the abstract did not seem available to the members. In several hundred pages of debate on Prohibition, we did not come across a clean statement of the principle of maximizing the satisfaction of individual preferences. Also, what is the logic of experimentation? Notwithstanding a good deal of testimony about experiences in Kansas, Maine, and elsewhere, no one seems to have addressed the question: How can the experience or experimentation of any state be extrapolated in theoretical or practical terms to a case for policy uniformity across a large, heterogeneous continent? That is a parts-versus-whole question.
On to the dry side. Leaving aside the specifics of problems and remedies, the crux of any policy debate, what warrants did the members express for Prohibition action? There are recurring themes.Footnote 34 Nation-centered democracy kept appearing: “Shall the people rule?” “[T]here is a Nation-wide demand for it.” “All power is inherent in the people.” “Over 6,000,000 American citizens have petitioned on the subject ….” A pro-dry scroll 150 feet long was stretched across the congressional chamber.Footnote 35 Also, Prohibition was a moral imperative: “Alcohol is a liquid poison.” “[T]his terrible evil in our land.” “It creates moral leprosy ….” “[O]f the same importance as slavery was in that other day.”Footnote 36 There were touches of teleology or destiny: “Its day has come.” “The truth about alcohol is rapidly bringing men and nations out from under the alcoholic anesthesia of history.”Footnote 37
But on most facets of decentralization or vertical federalism, the drys offered silence broken by occasional flourishes of loose argument. On states’ rights, there was what seems like casuistry: “The proposition is simply to give the States an opportunity to say whether the Federal Government shall be granted additional power over this particular subject matter…. [T]his is not an invasion of State rights”—a logic hard to sell in New York.Footnote 38 Senator Robert La Follette (R-WI), the era’s icon of Progressivism, although “very certain” that an enactment like this would “fail of enforcement,” voted for the Eighteenth Amendment anyway (technically, to send it to the states) based on his “convictions in support of democracy.”Footnote 39 There were confident assurances on shared authority: “If power is given [to] both the Nation and the States to enforce the prohibition, either independently or concurrently, there can be no conflict.” “[N]o conflict can arise between the Federal and State acts.”Footnote 40 Congressman Edwin Webb (D-NC), a floor manager for the Amendment in 1917, saw no problems ahead: “I think it is all right. I am not afraid to trust the States about that…. I never saw one that went counter to the United States Constitution, or whose law officers failed to enforce the law.”Footnote 41
Right-to-work laws. Right-to-work laws are a different kind of story. Some background is in order. The Wagner Act of 1935, a capstone victory for the labor union cause, did not cover agricultural workers, railroad workers, or public-sector employees (these omissions might be kept in mind), and it did not bar state right-to-work laws.Footnote 42 Before the Republicans passed the Taft–Hartley Act in the Eightieth Congress in 1947, the states of Florida, Arkansas, Arizona, Nebraska, and South Dakota had already voted for right-to-work in public referendums.Footnote 43 Through its 14b clause in 1947, the Taft–Hartley Act in effect firmed up an already-existent state leeway into a formal congressional guarantee.
The clause’s wording goes: “Nothing in this act shall be construed as authorizing the execution or application of agreements requiring membership in a labor organization as a condition of employment in any State or Territory in which such execution or application is prohibited by State or Territorial law.”Footnote 44
Following their election victory in 1948, the Democrats tried to repeal Taft–Hartley, and, beyond that, reaching past the Wagner Act, to enact a substitute measure that would ban state right-to-work laws. The latter achievement would have been a first. They failed on all counts. A hindrance to these aims was that majorities of both senators and House members in the new Congress in 1949, even though the Democratic Party now enjoyed sizable majorities, had voted for Taft–Hartley in 1947 and they stayed put.Footnote 45 And so, Taft–Hartley along with its 14b clause remained in place.
In the relevant hearings and debates in 1935 as well as 1947 and 1949, geography and business size figured in various ways. For example: “There is a great difference in the degree of industrialization ….”Footnote 46 Heavy-handed national regulation was said to be a bad fit for small towns and rural areas: “Industry in South Dakota with a few exceptions has not gone much beyond the stage where the employer calls his men by their first names.”Footnote 47 In 1949, most of the senators from New England, whose mills had been drifting to the South where unions were weak (a long-standing problem), backed a national standardization of law regarding the unions’ membership security, but they lost.Footnote 48 On the question of business size, the deliberations brought complaints that smaller firms suffered special harm from strikes and other union activities.Footnote 49 Senator Robert Taft (R-OH) in 1947: “[L]arger employers can well look after themselves, but throughout the United States there are hundreds of thousands of smaller employers, smaller businessmen, who, under existing statutes, have come gradually to be at the mercy of labor-union leaders ….”Footnote 50 A business testifier in 1947: “But in the case of the average smaller employer, a strike is a quick way of inviting the unions, the bankers, the sheriff, or one’s competitors to take the remains. Either of the four is willing to step in.”Footnote 51 In fact, there is evidence that smaller businesses have been special fans of right-to-work laws.Footnote 52
How did the members of Congress justify acting at all in 1935, 1947, and 1949? Events underlay certain warrants. In 1935, the industry codes of the National Industrial Recovery Act of 1933 were running out and needed replacement. Also, in light of the nationwide strikes of 1934, rife with violence, it is no surprise that Senator Robert Wagner (D-NY), stage-managing the new legislation, made a pitch “to promote industrial peace.”Footnote 53 In 1947, the Republicans dwelt on the postwar strike surge of 1945–46.Footnote 54 In 1949, the Democrats dwelt on the party’s election verdict of 1948 as a mandate to repeal Taft–Hartley.Footnote 55
On matters intrinsic to decentralization or vertical federalism, we did not find a great deal in these deliberations to report. In the committee hearings, several business representatives made the local-conditions case, as did Senator George (“Molly”) Malone (R-NV) in an extended floor speech: When we try to “settle every dispute in the country, hundreds of thousands of miles away, we are placing upon the National Labor Relations Board a duty which it is impossible for it to fulfill.” “It is like averaging the length of a lot of pairs of pants; the result does not fit anybody.”Footnote 56 States’ rights was there, but no one on either side seems to have considered the values of experimentation, close-to-the-ground democracy, maximal catering to voter tastes, or conflict avoidance.Footnote 57 These ideas did not come up.
Specifically on right-to-work and the 14b clause, the deliberations are sparse. In the Congressional Record, they appear on parts of a few pages and occasion not more than one or two percent of the Taft–Hartley wordage in 1947 and 1949.Footnote 58 Other considerations drew more attention. When they said anything at all, how did the backers of 14b support it? They noted that state leeway on the matter was the status quo, that many states had already enacted right-to-work laws or other union constraints. These facts, stated without adornment or elaboration, were their message. It strewed through the deliberations. The implication was: Why interfere with this state authority?Footnote 59 As an add-on, they did occasionally argue that the leeway already resting in the states could be cemented against any tampering, by courts or others, that might later come along by writing it into the specific 14b guarantee.Footnote 60 As an analog to the 14b clause, they occasionally pointed to state control of private-insurance systems, formally guaranteed by congressional statute.Footnote 61 Three southerners dwelt on states’ rights. That was about it.Footnote 62
Required are national standards, the pro-union opponents of 14b argued. The Commerce Clause cries out for them. Surrender to a hodgepodge of state laws brings inconsistency and confusion. “The industrial strength of this nation flows over state lines…. When local diversity of control constitutes a burden on that interstate flow and effort, it must give way to central control.” Conflicting rules can cause troubles for both unions and employers. Employers in union-weak states can enjoy an unfair competitive edge. This package was the chief pitch of the opposition side.Footnote 63 But it faced an embarrassment. The Wagner Act had operated for a dozen years without enjoying the requisite national standardization.Footnote 64
There were silences on both sides. At stake, everyone knew, whatever else, was the economic and political power of the labor unions. A few liberal senators pressed the value of strong unions. Senator Wayne Morse (R-OR): “Unfair to the legitimate rights of labor.” Senator Paul Douglas (D-IL): “In the decade from 1937 to 1947, the period in which the Wagner Act was in effect, great progress was made.” Senator Hubert Humphrey (D-MN): “The truth is that those who sermonize on the sacred ‘right to work’ really are thinking of the sacred right to starve.”Footnote 65 But there was modest take-up. A national Gallup poll of January 1947 cited in the deliberations brought no help to the cause. Asked to evaluate union-security agreements, 8 percent of respondents chose “closed shop” (join a union to get a job), 18 percent “union shop” (join a union and pay dues to hold a job), and 66 percent “open shop” (neither of the above).Footnote 66 Senator Morse conceded that the “closed shop” if put to a national referendum would lose.Footnote 67 The unions’ membership security formulas seem to have been political poison.
Given the subsequent economic and political role of the 14b clause, and the persisting controversy regarding it, it is hard to read these deliberations of 1947 and 1949 without reflecting that the treatment of 14b was rather cursory. Aside from nods to state authority, the Republicans said virtually nothing to justify the 14b clause. There was little discussion of the place of labor unions in the economy or the society. On the Democratic side, there were occasional eloquent interjections. But the anti-14b case was sparse and its arguments seemed incomplete or vulnerable. Why did the Commerce Clause need a rev-up in application? What was the justification for imposing union membership security rules onto a national public that, on various evidence hard to contest, was deeply hostile to them?
Generally speaking, the enactment of Section 14b just flew past. It flew past the New York Times, too, whose coverage of it at the congressional decision days of 1947 was sparse to nonexistent.Footnote 68
The ACA. Why enact universal health insurance in 2009? Two warrants infused the deliberations. Destiny was one. Present was “the call of history,” “the right side of history.” After “waiting for 100 years,” a “hope,” “aspiration,” “vision,” “challenge,” “dream,” “promise,” or “goal” was finally to be met.Footnote 69 There would be a “culmination.”Footnote 70 We counted some sixty references to earlier promises or efforts, including thirty to Theodore Roosevelt—the notable original advocate in 1912. Urgency was another warrant. It was not just a matter of problems; it was immediacy and urgency. Across the country, troubles were said to be rising to a boil—health insurance premiums “spiral[ing] out of control,” “eating into family budgets faster and faster.”Footnote 71 Senator Dianne Feinstein (D-CA) reported: “So over the past 2 years, California has lost insurance for 2 million people, bringing the total of people up to 8 million who have no insurance whatsoever.”Footnote 72 There were dozens of such claims (including some by Republican members), which have a ring of credibility. An analogy is the focus on soaring strike activity in the Taft–Hartley debates of 1947.
Scarce, however, in the ACA debates, was any action warrant based on the 2008 election. In hundreds of pages, we found just three claims of an election mandate on health policy: “health care reform,” “make it possible for every American to afford to live a healthy life,” and “improving health care for all Americans.”Footnote 73 There is nothing here specifically about the ACA as it evolved. This reticence may owe partly to the fact that Barack Obama in his 2008 nomination drive had attacked, not supported, the idea of an “individual mandate,” a key component of the ACA. The Obama team “had polled a mandate and thought it was a loser.”Footnote 74
In the hearings and debates, health-care reform at the state level received abundant attention. At the time, state reform action was ongoing and prominent, and taking account of its designs could help supply legitimacy to a national design.Footnote 75 At least eighteen states drew comments in the ACA deliberations about their success, failure, or shying from action. A Senate committee conducted a special panel on Massachusetts, California, Vermont, and Utah.Footnote 76 In general in the deliberations, easily the leader in recognition of action was Massachusetts with its fresh “Romneycare” program (an individual mandate, 97 percent insurance coverage).Footnote 77 Second in level of notice was Utah (market reforms, low costs, a new market exchange).Footnote 78 Progress was reported in Minnesota (92 percent coverage, 90 percent enrollment in nonprofit health maintenance organizations), Maine (guaranteed issue, a ban on cancellations), and Texas (costs down through tort reform).Footnote 79 TennCare in Tennessee had failed.Footnote 80 Single-payer efforts drew testimony.Footnote 81 The language of models, experiments, and laboratories appeared, although it lacked discussion about extrapolating from state to full nation. Differences in local “conditions” appeared as an argument against national standardization, yet, except in one area, the case was ill-developed and vague on alleged hurdles.Footnote 82 We did not come across a substantial, coherent anti-ACA argument based on conditions. The exception was rural-versus-urban contexts. Repeatedly, the ACA design was seen as a threat to hospitals and other health services in rural areas. Republicans from Maine, New Hampshire, South Dakota, Nebraska, Kansas, Oklahoma, Texas, Wyoming, Montana, Arizona, and Alaska voiced this alarm, as did two Democrats from Virginia.Footnote 83 This list has a rural, although not a southern, flavor.
Otherwise in the deliberations, the logics that might apply to decentralization or vertical federalism drew little attention. There was silence on both sides. The idea of maximizing the satisfaction of individual views made no appearance, nor did the pluses or minuses of decentralized democracy. In the hundreds of pages of debate, we came across no references by either side to conflict avoidance as a public good independent of all else. Portents from the 1990s did not figure. In 1996, Haynes Johnson and David S. Broder entitled their account of the Clintons’ unsuccessful health-care drive, The System: The American Way of Politics at the Breaking Point.Footnote 84 The countermobilization to this drive grew fierce in 1994. At a climactic moment, the Clinton White House launched pro-health-reform bus caravans to proceed in a flourish of publicity across the country. “Every one of the caravan routes—which set out from Portland, Dallas, Independence, and Boston—became an expedition into enemy territory…. The longer they traveled, the greater the signs of opposition they encountered.”Footnote 85
As for compliance or cooperation, the ACA would enjoy smooth sailing once it was enacted. That was the assumption in the deliberations. Neither the pro nor the anti side saw otherwise. Once empowered to do so, the states would set up their own insurance exchanges (the plans for which received little attention in the deliberations).Footnote 86 That confidence was shared by the Obama administration and by Democratic congressional aides. Medicaid expansion, the second major thrust of the measure, would proceed smoothly.Footnote 87 Nowhere in the debates did we come across any alertness that the courts or the states might balk at these performances. Dr. Raymond C. Scheppach, Executive Director of the National Governors Association, testifying in the ACA hearings at an early stage, did voice caution. On the exchanges: “The bottom line is, given the rigidity of the administrative rules here, I question at this time whether a substantial number of states would actually opt into the system.” On Medicaid expansion: “my sense is that [the governors] would question the necessity of increasing the eligibility of childless adults and parents over 100 percent of poverty [an aim of Medicaid expansion].”Footnote 88 But both the Democrats and the Republicans ignored Dr. Scheppach. None of the states’ governors testified in the hearings.
In theory, how might health insurance best align with vertical federalism? One line of argument owes to Nicholas Bagley, who offers a two-track model, separating the fiscal side from the regulatory side.Footnote 89 On the one hand, the American states are in a poor position to fund expensive health-insurance programs. Blocking that option are their annual balanced-budget constraints as well as Congress’s ERISA preemption that walls off employee-benefits plans from state interference. Testimony in the ACA deliberations notes this state fiscal incapacity.Footnote 90 Thus, in Bagley’s argument, as a practical matter, gushers of funding from Washington, D.C., are called for. On the other hand, the case for tight regulation of the states is less compelling. In Bagley’s argument: “if federalism means anything, it is that national judgment should not supersede state judgment, absent a good reason for federal intervention. Yes, federal money might be squandered in a state that adopts stupid insurance rules.… But that's an issue between the state and its voters.”Footnote 91 “Roughly, the states should retain control over regulation while passing to the federal government responsibility for money…. The country can easily accommodate a patchwork of state insurance laws. Indeed, it already does.”Footnote 92 We did not come across this two-track idea anywhere in the ACA deliberations. Canada’s way of handling health insurance seems to entail such a two-track plan. Ample funding from Ottawa joins with, on the regulatory side, chronic catering to Quebec.Footnote 93 So far as we could tell, Canada’s experience with vertical federalism did not come up at any place in the ACA deliberations.
We hope we have made a convincing argument that congressional deliberation in our three cases has been sparse and ragged. What can be said about this sequence of performances? At a theoretical level, a typical list of reasons for having Congress or elected assemblies in general might include the following: “[A]ssemblies serve as arenas for the perpetual fine-tuning of conflicts.”Footnote 94 “[T]horough discussion in Congress makes policies more legitimate.”Footnote 95 In the American system, questions bearing on centralization and vertical federalism can present third rails. Given this, we were struck in our three cases by how haphazardly Congress rose to a role on these fronts as fine-tuner of conflict or legitimizer of result.
At a more practical and empirical level, does it make any difference what congressional debate amounts to? Or how a national audience reacts to it? Was it of no consequence that the progenitors of Prohibition spoke dubiously about enforcing it, brushing off, for one thing, queries about the precariousness of joint federal-state enforcement? On Taft–Hartley’s 14b in 1947, was it of no consequence that its progenitors scarcely defended it at all? Was it of no consequence that the progenitors of the ACA spoke on for months saying little about the blueprints for the proposed new “exchanges” and exhibiting scant awareness that their plans for both the exchanges and Medicaid expansion might stumble once they encountered the courts or the states?
The American public does not glue itself to congressional debates.Footnote 96 But some attentive readers, listeners, or watchers are out there. In earlier times, newspaper journalists sat in on congressional debates and reported what was said, often verbatim. Live radio once had a role in Congress.Footnote 97 Now C-SPAN is available. But the question of how the public reacts to congressional proceedings is greatly understudied. So, we end in conjecture. In theoretical terms, a conditional argument seems to make sense. The bulk of congressional activity goes unattended. But perceived quality of debate might make a difference at least in rare circumstances where a significantly large audience witnesses and cares. In such circumstances, low perceived quality might invite lasting controversy. That would need to be the story. To take the positive side, it does not seem to be unknown in congressional history for content and style of floor presentation to confer a legitimizing effect, as in the textbook cases of Henry Clay and Arthur Vandenberg.
Next, we make a detour. In all our three controversies, an additional feature of the American policymaking environment has played an explanatory role.
3. National crises—the management of them, the policy opportunities afforded by them, and the downswings from them—have been key contributors to these long-lasting controversies
Prominently in the United States, where, owing to the mechanics of separation of powers it is hard to move the government to act, crises have rivaled elections as spurs to policy change.Footnote 98 The rhythm is familiar. A crisis occurs. “Policy windows” open. There is a logic of remedies. In some such instances, policies designed to be crisis remedies, such as the Agricultural Adjustment Act of 1933, far outlast their crises. But there is also a logic of opportunity. Momentary fears or excitements, or temporary congressional majorities heaved up by crises, may help along policy initiatives that are not politically feasible in normal times. Bonus options thus accrue to those in power. But then a crisis ends, the tide recedes, political disequilibrium gives way to a new equilibrium, and reaction or regret comes to flourish. This dynamic figured significantly in all three of the enterprises analyzed in this paper, with altogether five crisis nodes contributing to the stories.
In this country and elsewhere, prohibition of alcohol owed its success to World War I.Footnote 99 Before that, the American drive had been vigorous and was accelerating but had stalled. Now, a long-sought policy goal morphed into a crisis remedy. A spirit of sacrifice and abstemiousness was called for. Grain, a source of alcohol, became a precious wartime food commodity. The Anti-Saloon League and its allies pounced. “The prohibitionists draped their political pressure tactics in patriotic bunting. The war made it easy for drys to portray the predominantly German American brewers as subversives, if not traitors.” No congressional hearings were held. The Eighteenth Amendment sailed through Congress easily, and the states took only 13 months to ratify it.Footnote 100
Hence national Prohibition. Yet, a half-generation later, the Great Depression brought another crisis node. Support for the national dry regime had eroded in the 1920s, but now it collapsed. The causal story on this is cloudy but it is multiply subscribed to and seems compelling. “The growing malaise of the Great Depression introduced new political and social as well as economic circumstances, greatly accelerating the revolt against prohibition ….”Footnote 101 Available was a new
window for policy change…. In addition to concerns over increases in crime and infringements on states’ rights, in the midst of economic depression, a persuasive economic argument was put forth—that a resurrected, legal liquor industry would provide a respectable source of government tax revenues and produce badly needed jobs nationwide.Footnote 102
The 1932 election played a role: “As the liquor issue was one of the few issues that clearly divided the two parties, the results were interpreted as a popular mandate for repeal ….”Footnote 103 Once Congress recommended formal repeal in February 1933, it took the states only 9 months to ratify the Twenty-first Amendment—faster than they had ratified the Eighteenth Amendment.
In labor–management relations policy, there were two crisis nodes. The Great Depression brought fortune to the labor unions as the Democrats won the White House and swelled their congressional ranks. The party swept to victory in 1932 yet rose even higher in its remarkably positive midterm of 1934. The party in 1935, having gained in both chambers, now held 322 House seats and 69 Senate seats. New urban Democrats abounded. One consequence was: “in the wake of the November elections labor strength in Congress was greater than ever.”Footnote 104 “Wagner and his allies could be more confident in forcing their labor bill to a vote in 1935 even in the face of presidential reluctance to endorse it.”Footnote 105 The resulting Wagner Act of 1935 established the unions’ right to organize. The bill passed with scant floor debate and little opposition.
Was there a crisis in 1945–1946? The voters seemed to believe so. Economic troubles wracked the country as wartime gave way to peacetime—inflation, wrangling over price controls, food and housing shortages, a cascade of militant nationwide strikes. “In fact, 1946 turned out to be the most strike-ridden year in American history.” The year’s election season overlapped a nationwide meat shortage. In a flash November victory, the Republicans gained fifty-five House seats and twelve Senate seats to win control of the incoming Eightieth Congress. There, in 1947, in the Taft–Hartley Act, they overhauled labor–management relations to their taste (including 14b) and maneuvered it past a White House veto.Footnote 106 Thus, labor–management relations, like Prohibition, drew action connected to one crisis node, but then reaction connected to another.
Rahm Emanuel, Chief of Staff of the incoming Obama administration in 2009, a scary and memorable crisis juncture due to the banking collapse of 2008, mused: “Never allow a good crisis [to] go to waste. It’s an opportunity to do the things you once thought were impossible.”Footnote 107 In such circumstances, spasms of creativity and drive may seize a leadership class.Footnote 108 But also, as in 1935 and 1947, leeway in numbers of congressional seats can help. In the 2008 election campaign, Obama and the Democrats enjoyed at least decent prospects through the summer, but then came the financial meltdown in September, which was bad news for the White House party (the Republicans) and also exposed the presidential candidates to glaring scrutiny of how they would handle the crisis. As can be seen in tracking polls, Obama then soared from probably a slight favorite to his seven-point November victory.Footnote 109 The Democrats, already in control of Congress, gained twenty-one House seats and eight Senate seats. How did Obama’s late-season surge relate to the party’s results at the congressional level? There is no easy way to tell. But, absent it, it seems unlikely that the party would have had enough Senate slack—that is, sixty votes—to enact as ambitious a bill as the ACA.Footnote 110 At the margin, a statistically necessary Minnesota seat accrued to the Democrats in mid-2009 by way of a 0.01 percent victory edge only after a laborious 7-month overtime vote recount.Footnote 111 On the question of health-care reform, we do not know what would have happened absent the 2008 financial crisis and its election echoes.
The argument in these three cases is straightforward. Drifts back to equilibrium after “seize the day” lawmaking can entail understandable countermobilizations, lasting conflict, and, sometimes, repeal drives. A deeper cause is the American separation-of-powers system, which tends toward legislative stasis. Crises can break deadlocks, but their disequilibrating breakthroughs can be followed by lasting controversy.
On to the next observation.
4. In these three cases, the states have played a key constitutional role in supplying voice, force, shape, and duration to the cabining of congressional decisions
We begin with a discussion of congressional lawmaking in general. In a recent work, Susan Rose-Ackerman points up the uniqueness of U.S. national policymaking as contrasted with that in the British, French, and German systems. Tight, centralized party leaderships in league with government bureaucracies are the European story. The parliaments taken alone lean toward being passive. The American Congress, given its constitutional standing and its processes of deliberation and decision, is more powerful, busier, and more transparent than the parliaments, thus in some ways rendering the U.S. system more democratic. On Capitol Hill, anyone can make a contribution, and the media take notice of the proceedings.Footnote 112
This American design has met a test of time. Yet difficulties arise. Congressional processes can be ragged. The images are familiar: the 535 cooks, partisan hardball, position taking, dashes to 218 or 51 or 60 votes, blitzes while opportunity allows. Partly in consequence, the contents of the deliberations or the enactments themselves may be ragged or ill-thought-out. “[L]egislation typically involves compromises that generate vague language and inconsistent provisions.”Footnote 113 Concerns for administrability may take a back seat.Footnote 114 So may concerns for constitutionality, as in the case of the ACA: “Twenty-two hearings tied to health care legislation were held in each chamber of Congress …, though none meaningfully considered the constitutionality of the ACA.”Footnote 115 Grasp of a measure’s post-enactment future may be cloudy, as also instanced in the ACA: “Key staff, by and large, neither contemplated nor recognized the significance of the ACA's post-enactment years.”Footnote 116
Such raggedness is of course not always the case. For an instance in this paper, it is hard to beat Senator Wagner’s skill in maneuvering the Wagner Act into law, not to mention that act's design of taming industrial conflict by corralling it into government-monitored elections in the workplace—an exercise of engineering ingenuity.Footnote 117
Even so, raggedness does occur. Yet there are curbs on it. At the level of lateral federalism—that is, Congress, the executive branch, and the courts—the system offers a kind of encasement around such congressional difficulties. Think of a statute as a first draft. A statute emerges, and then institutions elsewhere go to work cleaning it up—clarifying it, polishing it, fleshing it out, cabining it. In this task, Rose-Ackerman emphasizes the role of bureaucratic rule-making, which the United States has pioneered as a post-enactment device—notably in the operations of the Administrative Procedure Act, a congressional creation. Also, the courts are on watch. U.S. policymaking, approached in this comprehensive sense, consists of far more than the enactment of statutes: It draws in multiple players and can go on for quite a while.
Our contribution here is to bring in the states.Footnote 118 In vertical federalism, we see an expansion of this constitutional encasement across space and time. The states also react to congressional statutes, also supply follow-up players as well as elongated discussion and disputation. Yet, two special points are worth making in this regard.
First, Congress’s neglect of the states—see the earlier discussion in this paper—can be a special spur to state reaction. Neal Devins has written, “There is no federalism constituency in Congress that pushes lawmakers to take [vertical] federalism into account when enacting legislation.”Footnote 119 This is a peculiar absence. Vertical federalism does not enjoy much of a niche in the argument repertoires of the members of Congress. This is notwithstanding a clear and growing public favor, in recent times at any rate, for state or local as opposed to national decision-making on policies. Support for the latter has been on the slide.Footnote 120
Second, vertical federalism juxtaposes the whole of the country to parts of the country. Thus is supplied a readymade vehicle for expressing differences between the country’s center and its periphery, and across its periphery. This formal structure contrasts with lateral federalism, where the presidency, the federal courts, and Congress taken as a whole serve the same nationwide constituency. Bring in the states, and the resulting geography-related differences in policy tastes can be deep. In the three cases in this paper, they have been. This second point is obvious, but we embroider on it later.
Yet we begin with a rollout of historical experience—the records of the three policy enterprises under study here as they entailed—or sometimes did not entail—interaction between the national government and the states.
Taken in the round as a policy enterprise, the prohibition cause went on for 115 years. The states began voting themselves dry in the 1850s. Dry or wet victories took place all over the country, in some cases states going dry and then backing off, in a mix of spirited legislative and referendum processes lasting until World War I.Footnote 121 Then came the national triumph. Of key interest here is the span of 1918 through 1933. The wartime enthusiasm carried on for a while. In a spirit of concurrence, nearly every state enacted its own enforcement law, a little Volstead Act.Footnote 122 Even so, “[t]hat the federal government had any role at all proved an excellent excuse for states to abdicate their law enforcement responsibilities.”Footnote 123 Soon, a response of, so to speak, “leave it to the feds” became the norm.Footnote 124 State funding for enforcement never appeared, fell off, or went away. “In 1926 the state legislatures allocated eight times more to implement the fish and game laws than to enforce prohibition.”Footnote 125 But the national government sagged, too: “While the states’ enforcement of prohibition varied from lackadaisical to nonexistent, the national government’s efforts ranged from inefficient to corrupt.” Enforcement enjoyed success in rural areas, but scored a poor record in New Orleans, Memphis, St. Louis, Richmond, Detroit, Chicago, Cleveland, New York, and elsewhere, where politicians ran for office opposing it and juries would not convict.Footnote 126
During the 1920s, the states took the lead in dialing back as support for prohibition waned. The New York legislature, a pioneer mover, passed its own little Volstead Act in 1921 (yet appropriated no money for it), but then broke precedent by repealing that act in 1923, and the state’s voters added a pro-wet victory in a referendum in 1926.Footnote 127 Any state, by legislative or referendum action, could slacken or repeal its own laws, thus setting a tone or an example, or could formally petition the national government to back away, and many states followed one such course or another. Following New York’s example, Nevada, Montana, and Wisconsin repealed their dry laws in the late 1920s. The wet side won referendums in Massachusetts, Illinois, and Rhode Island in 1930. Eleven states piled on in referendums in 1932. From there, it was easy sailing to the state conventions ratifying the Twenty-first Amendment in 1933.Footnote 128 Six states nonetheless kept their dry laws. Mississippi was the last to repeal, in 1966.Footnote 129 In brief, alcohol prohibition amounted to an over-century-long enterprise that the national government, thanks to world war, launched a brief, shaky intrusion into, then backed away from.
The story of right-to-work is more involved. It begins with the Wagner Act in 1935. Congress enacted that measure at a time of high crisis, catering to a constituency—the unions—favored temporarily by the off-normal Democratic congressional majorities of the time. Also, it might be noted that Congress in considering the measure addressed the country’s economic geography as it then existed. In the Senate hearings, over 80 percent of the witnesses representing local or regional entities listed a connection with firms in the Northeast or the Rust Belt.Footnote 130 That was the country’s industrial base at the time.
But much was to happen during the next decade. World War II brought “a great wartime expansion of industry into the South and Southwest,” creating a larger and more diverse industrial map. Think Arizona and California.Footnote 131 Also, union membership skyrocketed nationwide thanks to the Wagner Act and the war.Footnote 132 In sync with this growth, labor union strikes and other union activities, beginning with the Michigan sit-down strikes in 1937 and later becoming a signature of the war years, stirred an immense backlash in public opinion.Footnote 133 Finally, or relatedly, in a trademark onset of political homeostasis, the New Deal impulse and the congressional leeway that accompanied it gave way in the late 1930s.
Against this background, the states swung into action. Five of them, bearing the spirit of 1935, enacted “Baby Wagner Acts” in 1937. But then the thrust shifted. Starting in 1938, roughly half the states, year after year, by way of legislative or referendum processes, took steps to restrain unions. Targeted were picketing, boycotts, sit-down strikes, union dues plans, internal union processes, alleged intimidation, and other practices, culminating in the right-to-work campaigns of the mid-1940s. By the fall of 1947, “[c]losed shops, and usually all other types of union-security agreements, were banned by constitutional amendment or statute or both in thirteen states, all of them predominantly agrarian, nonindustrialized states ….” Eight other states constrained such agreements.Footnote 134 Congress, in enacting the Taft–Hartley Act in 1947, was stepping into a fast-moving stream. Then, shielded by the new 14b clause, anti-union interests kept pressing. The Supreme Court upheld Arizona’s right-to-work law in 1949.Footnote 135 Since 1947, the drive for right-to-work laws has gone on for three-quarters of a century. In state after state, through a mix of state legislative and referendum decisions, many choices have been made. As with prohibition, some states have opted in, then reneged.Footnote 136 Labor interests have racked up decisive referendum wins in certain states with historically strong unions—for example, Ohio in 1958, California in 1958, and Missouri in 2018.Footnote 137 But the list of right-to-work states has kept growing, reaching a total of twenty-six as of 2023, and the jousting does not seem to be over.Footnote 138
For the ACA, the story told here begins in 2009–2010. Many of us remember, perhaps falteringly now, the aftermath of the measure’s enactment. There were expectations: “The vast majority of states were expected to embrace the chance to craft their own exchanges [today called “marketplaces”], nor did the reliance on Medicaid appear problematic since all states were expected to expand it given the alternative of losing all of their federal Medicaid funding.”Footnote 139 In fact, only fourteen states plus the District of Columbia signed on to run their own exchanges. The rest opted out—some likely in a spirit of “leave it to the feds” echoing the 1920s, but many in a spirit of nullification that seemed to echo back to the 1820s.Footnote 140 As with right-to-work in 1949, the Supreme Court backed up the states’ authority. The Court ruled in 2012 that the ACA’s mandate for Medicaid expansion was unacceptably coercive—“a gun to the head”—and, as of 2015, twenty-two states had passed up participating.Footnote 141 Speedily after the ACA’s passage, voters in Missouri, Oklahoma, and Arizona amended their state constitutions to bar any “individual mandate.”Footnote 142 Officials in twenty-six states filed lawsuits arguing that the measure was unconstitutional.Footnote 143 Once the ACA was in place, a number of hostile state governments took steps to impede the “navigators” sent by national officials to ease people into the exchanges, which were now, by default in those states, federally operated.Footnote 144
Yet, like a river cutting new channels, or as comparably seen in the policy paths of prohibition and labor–management relations, the ACA has kept evolving at the state level. Medicaid expansion has increasingly caught on. The giant states of Florida and Texas remain holdouts, but many more states have signed on, thanks to thirsty state budgets and their publics’ desires for benefits.Footnote 145 Referendum victories in Idaho, Nebraska, Utah, and Maine have occurred.Footnote 146 But a surprising secondary policy regime has appeared. In many states, the early conflict over Medicaid expansion soon spurred new enthusiasm for “work requirements” as conditions for benefits, and that drive came to spread its force into even the non-expansion states, threatening to make participation there even tougher than before. Work requirements became a cause. A decade or so after the ACA’s enactment, the work-requirements question had resolved into a kind of waivers warfare in which a Republican president (Trump in his first term) said yes to states asking to waive the ACA’s national standards, and a Democratic president (Biden) said no.Footnote 147 In 2025, this dissensus over work requirements persisted.Footnote 148 It belongs to a family of such geography-related expression. Thus, much as Missouri, Oklahoma, and Arizona rushed to ban the individual mandate around 2010, the states of Vermont and New Jersey and the District of Columbia, after a Republican Congress repealed the ACA individual mandate in 2017, rushed to enact their own, so to speak, “baby individual mandates.”Footnote 149
Across this paper’s three enterprises, what does the history of state action signify? Given, notably although not only, their juxtaposition of intensity to an inflection of geography, it has been difficult for Congress to make a major policy move and have it decisively accepted. The states have joined into the processes (or were in already). As a practical matter, the policymaking process in these cases has enrolled a multiplicity of constitutional actors, including the states as well as the public and it has extended across time. All this activity has arguably constituted the policymaking process. In these instances, this is the way the country has been making its decisions—jaggedly and extendedly. For their part, given this way of construing things, the three repeal drives have served as components of policymaking, not responses to it.
It is important to see the role of the states. In these three policy enterprises, decisions by the state legislatures or electorates have figured repeatedly. We found the referendum processes especially interesting. They are not just creations of the Progressive era. Popular votes to amend state constitutions date back more than two centuries and are used frequently. Roughly half the states make the holding of referendums relatively easy, and all but Delaware allow occasional direct voter choice in some fashion.Footnote 150 Direct voter democracy is widely available. “Missouri strikes again,” for example, has been a story for this paper—decisive no votes on prohibition in 1918, right-to-work in 1978, the individual mandate in 2012, and right-to-work again in 2018.Footnote 151 The system has a Swiss-like referendum flavor at the bottom that figures in its overall checking and balancing.
Finally, what can happen when vertical federalism confronts party polarization? Comparing across countries, G. Bingham Powell, Jr., has written, “distance between [nationwide] median voter and [national] government” varies with degree of party polarization.Footnote 152 As an empirical matter, a wider gap between parties in ideological positioning translates into a wider gap between median voter and governing coalition in that respect. True, it is no surprise that, regardless of levels of party-versus-party polarization, victorious parties may move to overshoot the median voter once in office, thus among other things pleasing their activists.Footnote 153 But a context of high party-versus-party polarization seems to bolster the overshoot option. A ruling party's executable “bias bonus” so to speak, goes up.Footnote 154 In the United States, such an off-center majority-party legislative bias, increasing or not, may be leaving traces in certain recent controversies in the states. In Kansas on abortion restrictions, a result was: Republican state legislature yes, referendum voters no.Footnote 155 In California on affirmative action, a result was: Democratic state legislature yes, referendum voters no.Footnote 156 These have been striking dissonances. Of relevance here is: In the current condition of U.S. party polarization, what can occur when one party controls Congress and the other party controls a selection of state legislatures? This juxtaposition seems to invite a double whammy of “bias bonuses” spoking out in opposite directions and inviting special tension nationwide as policy is made.
In causal terms, our suggestion is: In our three cases, the states, as in Rose-Ackerman’s argument although extending it, have in effect participated in encasements of Congress as they have reacted to the passage of laws. This seems to be a constitutional, or quasi-constitutional role. It can help keep controversies persisting extendedly at high volume.
5. Enactments have consequences
In recent times, concerns about “implementation” and “sustainability” have led the scholarship on the impacts ensuing from congressional enactments. For the three enterprises here, the reports in that regard might go as follows. On Prohibition, vexed implementation was followed by collapse. For labor–management relations, the central thrust of the Wagner Act has been a success. Besides its workplace mechanisms, the act has offered a comfort zone for the political parties, a legitimizing flexibility: the staggered appointments to the National Labor Relations Board have allowed one party to edge the Board’s rulings one way, the other party, once the wheel turns, the other way.Footnote 157 But the right-to-work regimes in the states have proven sustainable, too. Right-to-work politics has brought a split territorial settlement that the country is accustomed to living with, although its boundaries change.
The ACA, minus certain of its original provisions and still evolving as discussed above, has proven sustainable. It could not be repealed in 2017. Accordingly, bolstered is a canonical insight in the welfare-state literature regarding retrenchment. That is, whatever such policy slippage may come about indirectly or in small doses (it can be ample), a direct, large-scale rollback of social provision is a difficult assignment.Footnote 158 A year after the repeal of the ACA failed, the midterm of 2018 ratified the result. In a telling detail, the (twenty) Republican House members who had voted against repeal fared statistically better in the midterm, net of all else, than did the Republican incumbents who had voted in favor of it.Footnote 159 Thus Medicaid expansion and a collection of constraints on insurance companies emerged from 2018 as likely “entrenched.”Footnote 160
But concern for implementation in these cases should not end the account. Statutes can induce consequences of many kinds. In this paper’s cases, for example, Prohibition brought on celebrity gangsters, corruption, bootlegging, and a dynamic new illegal alcohol industry extending down the Appalachians.Footnote 161 The Wagner Act, by empowering the labor unions, also enabled indirectly the widespread strike activity of ensuing years. In 1945, when asked in a national survey “who you feel might be harmful to the future of the country unless they are curbed,” respondents chose John L. Lewis, the president of the United Mine Workers. No one else ran a close second.Footnote 162
We close with comments on the political parties. One way or another, congressional enactments can impinge on the parties. Both of today’s parties were born partly as reactions to enactments—the Democrats to Alexander Hamilton’s economic program, which came into existence through congressional enactments; the Republicans to the Kansas–Nebraska Act in 1854. No such births have occurred for a long time. But there can be leadership teams and factions. Reactions to this paper’s statutes have sparked or been closely associated with some of these. Alfred E. Smith rose to power in New York and in the national Democratic Party as the symbol of the antidry cause.Footnote 163 Barry Goldwater cut his political teeth on right-to-work controversies in Arizona in the 1940s.Footnote 164 In the case of the ACA, there is the associated rise of the Tea Party.Footnote 165 Why did the “conservative coalition” emerge in the late 1930s? It is possible that the public's icy reaction to the strikes, as recently documented by Eric Schickler and Devin Caughey, was a larger factor than has been assumed.Footnote 166
There are the fortunes of the parties. Does right-to-work make a political difference? One recent work suggests that it does: “we find that right-to-work laws reduce Democratic Presidential vote shares by 3.5 percentage points. We find similar effects in Senate, House, and Gubernatorial races, as well as on state legislative control.”Footnote 167 Figuring as a mechanism is whether unions can be strong enough, as in levying and using dues, to exercise influence in the country’s elections to public office. Policymaking effects may ensue from election victories.Footnote 168 This is a familiar logic. But this reported 3.5 percent figure is remarkably large. How many elections are decided by less than that? It is small wonder that Section 14b rose to its high standing as a repeal target.
Did the Democrats pay a political price for their ACA policy victory? Apparently, they did. After passage, the 2010 midterm bore out the measure’s hostile standing in that time’s polls. The (thirty-nine) Democratic House members who had cast roll call votes against the ACA fared significantly better in the 2010 midterm, controlling for other factors, than did those Democrats who had voted in favor of it. Only in districts over 72 percent for Obama in 2008 (read, especially: the cities) did a pro-ACA roll call vote prove an electoral plus rather than a minus.Footnote 169 After that midterm, “Democrats held fewer elected offices nationwide than at any time since the 1920s.”Footnote 170 “[B]acklash to the ACA cost the Democrats congressional seats and likely majority status in the House for eight years.”Footnote 171 The Republicans could draw the legislative maps for the decade. Collateral damage to the Democratic Party included four new right-to-work states—West Virginia, Kentucky, Indiana, and Wisconsin—as well as a disempowerment of the public employees’ unions in Wisconsin.Footnote 172 Aside from adverse reaction to the ACA itself, the conflict erupting during its passage likely contributed to the situation. Voters do not relish conflict on Capitol Hill.Footnote 173 Possibly, the experience proved a drag for a while against large policy enterprises. For whatever cluster of reasons, the public’s survey-measured temperature on trust in Congress, confidence in Congress, and favorable toward Congress plummeted during Obama’s first term and has stayed low.Footnote 174
Acknowledgments
Paper prepared initially for presentation at the Congress and History Conference, Washington, D.C., May 16–17, 2024. We are grateful for the assistance we have received on this paper. Thanks to Elizabeth Rigby for sharing a dataset; to Judith Swanson for thoughts, sources, and copyediting; to Frances Lee for triggering the idea for the paper, and to Sarah Binder, Steven Calabresi, John Dearborn, David Karol, Eric Patashnik, Susan Rose-Ackerman, Sam Rosenberg, Wendy Schiller, and Ian Shapiro for comments.
Competing interests
The authors declare none.
Appendix A: National party platform calls since 1900 to “repeal” congressional enactmentsFootnote 175
Open access
