The Bizarre Red-Blue Politics of Election Consolidation

In a 2012 state legislative hearing, the lead proponent of a bill to consolidate local elections in November of even-numbered years said: 

This bill would do one thing and one thing only. It would make Election Day uniform throughout the state…[it] ought to be a non-controversial topic. …This bill saves money. It increases voter turnout. …If we believe in representative democracy…we should support this bill. 

Was the speaker progressive or conservative? A Republican or a Democrat? 

What about the champion of a similar bill in a different state who said this in 2015? 

“There is one major contributing factor to low voter turnout—the timing of elections—that could be addressed with a relatively simple policy change.” 

And how about the legislative sponsor of a 2023 bill in yet another state who proposed to move “every single type of election in the state…to our regular even-year elections” because “doubling turnout—that’s all for the good”? 

Bizarre-partisanship 

The first speaker was the Arizona arch-conservative Clint Bolick, co-founder of the libertarian Institute for Justice. The second quote is from the arch-liberal interest group California Common Cause. The third is from of Montana state representative Mike Hopkins (R). 

Is election consolidation (moving local elections to the same November ballot as national elections) a rare political case, then? Is it a reform where the left and right work together? 

Not at all. To date, it’s been more bizarre-partisan than bipartisan. In these states and others, proponents and opponents recite the same arguments for and against election consolidation. Indeed, if you go online and watch hearings on these bills (as I have done for five states) or comb through media coverage from a half dozen other states considering the idea, you’ll learn that the scripts are almost verbatim but the parties keep trading parts. 

Swapping scripts 

An example from the pro side: 

“It’s better to have 60 percent of the people rather than 30 or 40 percent of the people choosing,” Kansas state senator Damon Thayer (R) said in 2020 as he argued against unified Democratic opposition for consolidated elections. Three years later, New York state senator James Skoufis (D) argued incredulously for election consolidation against a phalanx of Republican opponents. “You have 20 or so percent of voters deciding the outcome for the entire jurisdiction,” he said. “Why are you so afraid of 50, 60, 70 percent of voters determining who should hold these local positions?” 

And one from the con side: 

“[Democrats] will stop at nothing to manipulate the system to rig themselves into total and permanent power,” state Republican party chair Nick Langworthy of New York complained in 2022. A year later, Tennessee House Democratic caucus chair John Ray Clemmons called a Republican’s election consolidation bill a way to “manipulate the democratic process for the sole purpose of consolidating even more power.” 

Watch enough of these hearings and you’ll experience a singular combination of déjà vu and whiplash. 

Role reversals 

Leaders’ actions matched their words. Almost every state election consolidation proposal in living memory has split legislators along party lines. In red states, Republicans vote yea and Democrats vote nay. In blue states, vice versa. It’s an unusual pattern that’s possibly unique. 

Often, the proposals were almost identical, even borrowed. In 2012, for example, the Arizona legislature passed the bill Bolick was testifying for (HB2826 2012). Championed by the conservative Goldwater Institute and supported by GOP mainstays like the Arizona Chamber of Commerce & Industry, the bill won with three-fourths of Republican senators and four-fifths of Republican representatives voting yes. Republican governor Jan Brewer signed the bill, which almost all Democrats opposed, into law. 

Three years later, in 2015, neighboring California considered a similar bill (SB 415). One innovation, included to avoid legal troubles with state constitutional guarantees of local control for certain cities, was a trigger provision that allowed cities to run their elections off-cycle if they could keep voter participation within 25 percentage points of what they achieved in on-cycle state elections. This bill passed. All but one Republican in the legislature voted no, and all but two Democrats voted yes or, in five cases, abstained. 

Three years later, in 2018, Arizona’s legislature returned to the question after the state supreme court invalidated parts of the 2012 law for overstepping local authority. Hoping to render the court order moot, the legislature imported the California trigger provision. As in 2012, the vote split on party lines. Arizona House Bill 2604 (2018)—literally copied from the California Democrats’ law—passed without the support of a single Democrat in either house of the legislature and without the opposition of a single Republican. 

The pattern repeats elsewhere: in recent years in red Idaho and Montana and in Kansas, Kentucky, and Tennessee, Republicans have favored election consolidation and Democrats have opposed it; in blue New York and Washington, the polarization is inverted. 

Election consolidation is a political puzzle. Why does it, perhaps alone among reform proposals, follow this weird red-blue pattern from state to state, with support coming only from Republicans in red and Democrats in blue states. 

The policy itself 

As policy, of course, election consolidation is not puzzling at all. This is the best-kept secret of democracy upgrades: election consolidation boosts turnout more than any other reform, often doubling participation in local elections. Indeed, it raises turnout more than all other common reforms combined. It also improves government accountability, electing officials who more faithfully reflect local sentiment. It brings out a more representative electorate, especially by age but also by ideology, race, and ethnicity. It saves public money. And it’s wildly popular; its poll numbers would make any political consultant salivate, and virtually every public vote on consolidated elections passes by a huge margin. 

By dramatically expanding participation, election consolidation also gives elected leaders greater legitimacy and a stronger mandate. In doing so, it may even lay the foundation for increased trust in local government—a rare bright spot in the bleak landscape of declining confidence in public institutions. Still, election consolidation is the ugly duckling of electoral reforms; it’s rarely even mentioned in the raging, furious, nationally polarized partisan debates over election integrity, voter fraud, and voter suppression. 

Accidents of history 

There is no relationship between when local elections take place and the familiar pattern of red, purple, and blue states. In seven states, including deep-red Arkansas and Kentucky, purple Nevada, and deep-blue Hawai’i and Oregon, the law requires on-cycle local elections in every city. In 18 states, including red Alaska, Texas, and Wyoming, purple Florida and Michigan, and blue Delaware, Maryland, and Minnesota, state law lets cities decide when to hold local elections. And 25 states require all cities to hold their elections off-cycle, including red states such as Idaho, Kansas, and Montana; purple states such as Ohio, Pennsylvania, and Wisconsin; and blue states such as New Mexico, New York, and Washington. 

This crazy quilt is an artifact of history. It’s a sign of the relative state-by-state influence of the Progressive movement more than a century ago, but it does not remotely resemble current politics. What, then, explains today’s bizarre-partisanship? Why do Democrats and Republicans swap scripts on the self-same bills depending on what state they are in? Why do Democrats support election consolidation in California and New York but oppose it in Arizona, Montana, and Tennessee, while Republicans do the opposite? 

The reflexive answer is partisan self-interest: calendar reform must help Republicans in red states and Democrats in blue states. That explanation may be right, in which case we can expect steadily more red and blue states to consolidate their local elections. But that’s not the only (or even the best) explanation for the available evidence. 

A surprising alternative hypothesis is that bizarre-partisanship is a byproduct of other dynamics—that it’s almost an accident. And if this hypothesis is right, it follows that not only polarized partisan action but also bipartisan action for election consolidation is possible. The happy result of this reasoning is that election consolidation has a bright future, with prospects for progress in many states. 

Mile-wide, inch-deep 

To explain bizarre-partisanship, it helps to begin with five facts: 

1. Election consolidation’s support among the public is a mile wide

Ordinary voters prefer fewer, better elections. They would rather fill out one ballot per year, even if it’s long, than three or five ballots. The United States is an outlier among its peer countries around the globe. Voters in the state of Washington, for example, face four election days in a typical year, while in neighboring British Columbia, Canada, voters usually have one.

In 2008 Democrats, Republicans, and independents all separately favored synchronized elections by roughly two to one in an academic survey with an enormous sample; the 2020 version of the same survey showed the same ratio of support, as did a smaller 2021 poll in Washington, a 2023 Sienna College poll in New York State, and a 2024 Manhattan Institute poll in New York City. Finally, a 2024 Secure Elections Project survey found that Republicans and conservative-minded independents supported the policy by about 80 percent in Arizona, Georgia, and Wisconsin.

These polls line up with actual elections too; virtually every public vote (21 of 22 in the past decade) on a proposal to move local elections to November of even years has passed overwhelmingly.

2. Election consolidation’s support among the organized right is widespread

The American Enterprise Institute, which typifies the business-establishment branch of conservatism, writes, with emphasis on school boards:

The conservative case for on-cycle elections is straightforward and powerful: Greater local democratic participation will empower local control by yielding school board members whose values more closely reflect their constituents and who will make policy and fiscal decisions that more closely reflect parents’ preferences.

The libertarian-minded Koch family’s marquee advocacy network Americans for Prosperity writes, “Benefits of one election in November: increase in voter turnout, more citizens engaged on local issues, increased accountability on local officials, reduced cost for local taxpayers.”

The Trump-aligned America First Policy Institute picks up the theme: “Placing elections at obscure times does not promote democracy—it inhibits it.” The Foundation for Government Accountability agrees, as does the Honest Elections Project, an affiliate of the right’s flagship think tank the Heritage Foundation. And so do the urban conservatives of the Manhattan Institute; they even drafted a model policy.

3. Support for election consolidation is equally widespread on the American left

The Brookings Institution, the quintessential think tank of the liberal establishment, supports on-cycle elections. So does the next-generation left-of-center think tank New America. Progressive advocacy organizations that support on-cycle voting include many state and local chapters of the American Civil Liberties Union, California Common Cause, and the League of Women Voters. 

The historic progressive movement began about 130 years ago, exemplified by the organization the National Civic League. Back then, the league led the charge against on-cycle local elections. Now its modern counterparts have turned 180 degrees and endorsed them wholeheartedly. Its New York affiliate, Citizens Union, is a prominent proponent. Other progressive state-based nonprofits, such as MassINC and Policy for Progress in Massachusetts and More Equitable Democracy and the Northwest Progressive Institute in Washington, also energetically advocate for election consolidation. 

The progressive bona fides of synchronizing elections are also apparent in local campaigns. To take just one example, voters in Boulder, Colorado, cast ballots in November 2022 on consolidation, and the “yes” side won endorsements from a who’s who of liberal groups: the ACLU, NAACP, and Sierra Club, along with political parties ranging from Democrats to the Working Families Party, and AOC’s Democratic Socialists of America.

4. For all the breadth of support for election consolidation, that support is an inch deep

Even among the all-star roster of right and left organizations that endorse it, only a handful lift a finger to advance it. It’s almost no one’s priority. 

Most legislative consolidation bills have drawn little public notice and have failed more for lack of enthusiasm than because of organized opposition. Sean McMorris, a program manager at California Common Cause, which sponsored that state’s 2015 reform, told me, “There was no opposition to the bill.” At the bill’s senate committee hearings, no one showed up to testify. Although other legislative fights, including ones in New York and Washington in the past year, have stirred at least a trace of controversy, the issue remains surprisingly low profile. 

Meanwhile, virtually every public vote on the question has been a sleeper, with hardly any public attention. In King County, Washington, home of Seattle and the most populous jurisdiction to vote on the question in recent years, election administrators could not even find anyone to write a “con” statement for the official voters’ guide. 

5. The nonpartisanship of local elections is a charade

Local elections are nominally nonpartisan in about 85 percent of US cities—another consequence of the Progressive movement more than a century ago. But the nonpartisanship of such elections is superficial. Parties routinely endorse candidates in allegedly nonpartisan races, and candidates routinely advertise their party allegiance. Expectations of partisan advantage weigh heavily in legislators’ thinking about local election reform, as in all things. 

Scratching deeper for an explanation 

Again, the presumptive explanation for bizarre-partisanship is self-interest. Legislators, you would assume, know what’s best for them and their team, so in some way, election consolidation must benefit Republicans in red states and Democrats in blue states. Perhaps, but the best studies suggest that election consolidation does little for both liberals and conservatives. 

Consider: In March 2024 Justin de Benedictis-Kessner of Harvard’s Kennedy School of Government and Christopher Warshaw of George Washington University released an impressive best-in-class study that applies sophisticated statistical techniques to a new database on local elections they had assembled. Their data covers almost 18,000 individual city-council-seat elections across the entire United States over three decades. It also draws in vast quantities of other information about candidates’ ideology, partisan affiliation, race and ethnicity, and more. Their analysis looks closely at hundreds of cities before and after election consolidation, and the findings are unequivocal: “Moving local elections on-cycle significantly increases overall voter turnout and the participation of younger and less wealthy voters. But it has negligible effects on the partisan composition of the electorate or the partisan and ideological outcome of elections.” 

The same month, Katherine Levine Einstein and three colleagues at Boston University released another best-in-class study that looked at more than 500 cities to compare the electorates in recent on-cycle and off-cycle mayoral races. Although the authors did not look at effects in the same cities before and after consolidation (instead comparing on-cycle and off-cycle cities), their study had the advantage of national access to arguably the best commercial “voter file” currently available, from the firm L2. The authors found much higher turnout in on-cycle cities among working-age voters and renters, somewhat more turnout among Latinos, and equal turnout among African Americans. On-cycle city electorates also included a larger share of those cities’ independent voters, while off-cycle city electorates overrepresented both partisan Democrats and Republicans, in roughly equal measure. 

These findings mostly match but differ somewhat in partisanship from the previous field-leading study, published in 2021 by Zoltan Hajnal of the University of California, San Diego; Vladimir Kogan of Ohio State University; and G. Agustin Markarian of Loyola University Chicago. They applied statistical analysis to the massive trove of information on voters maintained by the political data firm Catalist, and showed that California cities that had consolidated their elections saw not only turnout surges but also shifts in the composition of the electorate. Mainly, the electorate got much younger: voters age 55 and older made up almost half of the electorate in city elections before those cities switched to on-cycle elections. After the switch, far more young voters participated in city elections, diluting elders’ votes; older voters’ share of the electorate shrank by 22 points in presidential elections and 13 points in midterms. 

Given that younger cohorts are more racially and ethnically diverse throughout the United States and especially in populous states such as California, it’s no surprise that on-cycle elections attracted more voters of color. In cities that switched to on-cycle elections, Latinos’ share of voters grew by 7 points in presidential years, and Asian Americans grew by 2 points; both grew half as much in midterms. Less-affluent voters showed up more in on-cycle elections, but the Black vote share was unchanged. Finally, the Democratic vote share grew by 4 points in presidential elections and 2 points in midterms. 

This research, unlike the newer papers by de Benedictis-Kessner and Warshaw and by Einstein and colleagues, suggested a leftward advantage to on-cycle elections. But California is an extremely blue state. Is it an outlier? Do states’ city voters skew left of the state averages in blue states and right of the state norms in red states? 

What about purple and red states? 

Drs. Hajnal and Markarian subsequently applied the same methods to analyze the increasingly red state of Florida and shared their unpublished findings with Sightline. They again compared local elections in the same cities, before and after those cities consolidated elections. Over the past decade, on-cycle municipal elections in Florida brought out vastly more voters, nearly doubling turnout in midterms and nearly tripling it in presidential years. Just so, the on-cycle city electorate in Florida included far more working-age people. Those under the age of 40 made up more than 16 percent of city voters in presidential elections, for example, but less than 7 percent in off-cycle elections. 

Some other Florida effects went the same direction as California but less dramatically. On-cycle elections induced more Latino votes but only about half as much as in California (a 3.7 percentage point increase during presidential years). In Florida, Asian Americans’ vote share rose by only 0.3 points. As in California, Florida saw no effect on the Black vote share. It’s important to note that, unlike in California, Florida’s on-cycle city electorate was unchanged in its partisanship; neither Democrats nor Republicans gained from on-cycle elections. The Florida results line up with those of de Benedictis-Kessner and Warshaw and of Einstein et al. 

Curious to understand these dynamics better, Dr. Markarian applied the same methods to Montana, a red state, and Georgia, a purpling state. These states did not allow cities to switch their election schedules during the past decade, so Dr. Markarian instead compared off-cycle local elections, such as city council, with on-cycle state and national elections within the same boundaries. The findings in Montana were of a piece with Florida: twice the turnout and a much younger voting population. Interestingly, the on-cycle electorate was unchanged from the off-cycle electorate in Montana in partisanship but included more people that data provider Catalist marked as moderate in ideology. Otherwise, the changes were tiny or statistically insignificant. 

In Georgia, Dr. Markarian’s analysis ran into data problems. Still, the evidence suggests the same patterns as Montana’s: compared with off-cycle voting, on-cycle local voting turned out a much larger and younger electorate that is somewhat more moderate politically and marginally more Latino. Again, there was no evidence of any statistically significant effect on the partisanship or ideological leaning of voters. 

In sum, politically, election consolidation appears to be a boon to neither the left nor the right, and there’s no evidence that it skews the opposite directions in red and blue states. In California, it may help progressives slightly; other places, it may mildly advantage moderates. But the overwhelmingly dominant effects are a roughly doubled turnout and better representation of working-age voters. 

Subtler effects? 

Is consolidation truly neutral in its effects, then? Not likely. Statistical analyses reveal overall trends and averages, not specifics. Anything that doubles turnout will have some effect. Perhaps election consolidation boosts particular constituencies within the local electorate rather than entire parties or their ideologies. That’s one possibility another line of research supported. 

Let’s get into this evidence by returning to Clint Bolick, the conservative jurist whose quote kicked off this article. He said in the same testimony to the Arizona legislature: 

James Madison in the Federalist papers warned that the smaller the governmental entity the more likely it is to be subject to special interest manipulation. What he did not foresee is that local governments often would schedule elections when no one knows about them, which increases the potential for special interest manipulation. 

The “special interest manipulation” he had in mind was not progressives’ bogeyman of political spending by business PACs, though. It was the get-out-the-vote prowess of public-sector workers, such as teachers’ unions. 

Austin VanDerHeyden, municipal affairs liaison at the conservative Goldwater Institute, recently told me, “From our perspective, the lack of uniformity in candidate election dates leads to confusion among voters on Election Day. This gives a systematic edge to labor unions and special-interest groups in elections because they are the only groups with the infrastructure to turn out voters in the off-season.” 

The case of teachers’ unions 

A substantial body of evidence suggests that Mr. Bolick and Mr. VanDerHeyden are right: off-cycle elections amplify the power of public employee unions, such as teachers’ and firefighters’ unions, which are central members of the Democratic Party’s political coalition. Scholar Michael Hartney wrote for the Manhattan Institute, “Off-cycle local elections tend to give organized interests extra power in local politics. Among these interest groups, public-employee unions loom large, especially teachers’ unions.” By consolidating elections, conservatives may be able to dampen the power of a key Democratic interest group. 

Meanwhile, however, another substantial body of evidence pushes in the other direction. It suggests that on-cycle elections amplify the power of progressive voters on school funding and education policies. Research by Vladimir Kogan of Ohio State University and his coauthors shows that the increased influence of unions in off-cycle elections is (if real) completely swamped by those elections’ domination by older voters. In Montana, for example, some 53 percent of voters in off-cycle elections are older than 55. Off-cycle elections are practically gerontocratic. On many issues of taxation and school spending, older voters lean to the right. Shifting local elections to on-cycle would therefore help progressives by turning out younger voters on those issues. 

Which is it? Does election consolidation dilute Democratic unions’ influence or boost the power of working-age voters? Perhaps it does both. Logically, one would expect on-cycle elections to elevate the interests of working-age voters at the expense of others, but working-age voters are not politically monolithic. The net effects undoubtedly vary from place to place and time to time. But as with the big-data analyses discussed above, we’re left without any reason to believe election consolidation systematically skews either left or right. 

Which leaves us as unable as ever to explain bizarre-partisanship. 

The byproduct hypothesis 

So rather than looking to data to explain it, perhaps we should posit that the lack of data explains it. What if in each state where consolidation is considered, partisans do not have an astute and precise understanding of how it will play out? What if, instead, they are following their gut? 

In Western states, when I have asked partisans how they know consolidation will help their party, they tend not to cite actual analysis of local voting patterns. Instead, they invoke general rules (“Younger voters are more progressive,” or “In this state, high turnout helps Republicans.”). Or they point to specific, egregious, low-turnout off-cycle elections that hurt their team. On-cycle elections, they believe, will mean more wins for their side. Meanwhile, when I have asked opponents how they know consolidation will be bad for them, they have voiced similarly weak arguments, reciting long-since debunked concerns about long ballots, overwhelmed voters, or in one case, that street corners will literally lack space for all the yard signs if elections are synchronized. 

Such conversations have left me suspicious. I have started to think that legislators are mostly taking their cues from each other. One party proposes it, so the other opposes it. If you’re for it, then I’m against it. Both sides are accustomed to their adversaries cynically gaming everything for their own electoral advantage; they assume that election consolidation is just the same game. 

This hypothesis—that bizarre-partisanship is an accident, a byproduct of negative partisanship—may seem far-fetched. Surely elected officials must know exactly which electoral reforms will help or hurt their team, given the enormous stakes. Surely they must be talking with their political consultants, who must be calculating the net effect on their prospects for reelection, right? They must have better data than do a handful of academics, right? Right?! 

Maybe. But maybe not. 

Negative partisanship? 

Remember, election consolidation is usually a quiet, low-profile issue. It’s far from the klieg lights of the national, polarized political main stage. Elite opinion on both the right and the left supports it, as do voters of all parties, if only tepidly. Few legislators have studied the issue, the history of election scheduling, or whom the change would benefit. From a legislator’s perspective, election consolidation is almost always just one more bill or amendment among dozens. Few constituents are writing or calling or lobbying or testifying or yelling in town hall meetings about the scheduling of elections. And maybe legislators just vote against the other team. 

Majority parties run legislatures, so majority-party bills are the ones that get debated, scheduled, and voted on. In blue states, election consolidation proposals scheduled for hearings are almost always Democrats’ bills; in red states, they are almost always Republicans’. Perhaps negative partisanship explains the rest. Republicans and Democrats oppose each other’s bills, and what they state as their reasons are mere rationalizations. That would explain the script-swapping. 

Negative partisanship is probably why campaigners for election consolidation in Boulder, a lefty university enclave in Colorado, framed their central case for election consolidation as being in opposition to conservatism. Their campaign website warned, “The GOP don’t want more people voting. Democracy is under attack across the nation by MAGA Republicans and conservatives alike. Here in Boulder, we believe that democracy is stronger when we all vote.” 

Negative partisanship could also be why Tucson, a Boulder-blue university city in red Arizona, has been trying to fend off election consolidation in court for a decade. In Arizona, as noted, election consolidation has been a conservative cause. The Republican state legislature has forced it on cities, and Tucson leaders have rebelled. Not only that but Tucson citizens voted against consolidation, making themselves the sole exception to the rule that local ballot measures to consolidate elections always pass. Tucson shows that in this era of polarization, negative partisanship can trump even voters’ two-to-one preference for concurrent elections. 

Do all roads lead to consolidation? 

As dispiriting as that hypothesis might be, as an indicator of the health of democracy, it could also be a reason for optimism about the prospects for election consolidation. All roads may lead there. 

If the byproduct hypothesis is wrong and governing parties do know that election consolidation will help them, they will presumably continue to advance it. They will add to the three states and scores of cities that have consolidated elections in the past decade. Partisan proponents of synchronized elections in trifecta states are already pushing forward. Montana’s supermajority Republicans came close to passing a sweeping consolidation bill in 2023 and will try again in 2025. New York Democrats adopted consolidation for counties and towns in 2023 and are pushing to add cities in 2024. Meanwhile, Republicans in Kentucky and Tennessee, and Democrats in Washington, all passed consolidation bills through one but not both legislative chambers in 2024 and will be back again in 2025. At present, among states that ban on-cycle city elections are six Democratic-trifecta states, including Illinois and Massachusetts, and 12 Republican-trifecta states, including Georgia and Ohio; all are good candidates for the partisan road to consolidation. 

If, on the other hand, the byproduct hypothesis is right and negative partisanship is the main thing driving legislators to polarize in the current bizarre pattern—swapping scripts and reversing roles—then patient and diplomatic reformers might be able to build those rarest of things in 21st-century politics: bipartisan coalitions. Such coalitions would be needed to consolidate elections in the off-cycle states that lack trifectas, such as Pennsylvania and Wisconsin. They would also help speed and secure progress everywhere else, as bipartisan bills are more likely to pass and less likely to be reversed. 

Election consolidation by the bipartisan road came to Nevada in 2019. That year the legislature took up Assembly Bill 50. As usual, proponents recited the exact same arguments as in other states. Somehow, though, legislators did not polarize, perhaps because the bill had prominent supporters from each side of the aisle. The Republican secretary of state introduced the bill, and the registrar of voters in the Democratic stronghold of Clark County (home to Las Vegas) was a vocal supporter. Some cities had already moved to on-cycle elections, so everyone knew what to expect: higher turnout, better demographic representation, more accountability of government to voters’ values, and less public spending on elections. The bill passed with the votes of every single Democrat and a majority of Republicans in each chamber. 

This pattern—bizarre-partisanship-turned-bipartisanship—is worth trying to replicate. After all, legislators and advocates are already singing the same song. Why not do it together, in the same state, at the same time? 

Either way, whether by the partisan or the bipartisan path, election consolidation appears to have a bright future. The proposition of having fewer, better elections is overwhelmingly popular among the public and policy thinkers of all political stripes. And even in states where legislators are divided, supporters at least always seem to be the ones in power: Republicans in red states and Democrats in blue states. 

Thanks to Anna Campbell and Todd Newman for research support, and to reviewers Avi Green, Zoltan Hajnal, Neal Ubriani, and Ben Weinberg for comments on an earlier draft of this article.