In May 2012, the Tea Party shocked the Republican establishment by knocking off one of the GOP’s most venerable and influential Senators of recent decades. In the Indiana Republican primary, six-term Senator Richard Lugar lost by twenty points to a little-known, hard-right challenger, State Treasurer Richard Mourdock. Lugar outspent the upstart three to one, but the latter mobilized a fervent Tea Party grassroots effort while garnering the support of the National Rifle Association, the national Tea Party PAC, FreedomWorks, and the conservative Club for Growth PAC. Conservative activists charged that, after 36 years in Washington, Lugar (who did not own a residence in Indiana) was out of touch with the state. But his real crime was, as the Christian Science Monitor put it, that he was “too willing to compromise with Democrats on issues ranging from foreign policy and the DREAM Act to U.S. Supreme Court nominees and the auto bailouts.”
Lugar was one of several prominent Republican senatorial candidates (including a few incumbents) to be defeated during the primary elections of 2010 and 2012. All were perceived as being too moderate, flexible, and pragmatic, even though they were anything but liberal. Lugar’s voting record in the 112th Congress (his final two years) was 6 out of 100 on the Liberal Action Score and 63 on the Conservative Action Score.
In his final two years in the Senate before losing to Tea Party candidate Mike Lee (and finishing third) at Utah’s 2010 Republican Convention, Senator Robert Bennett had a liberal score of 10 and a conservative score of 68. The ultra-conservative Republicans who dominated Utah’s nominating convention got what they were looking for in Lee: In the 114th Congress, he has kept a perfect 100 percent conservative voting record (compared to an average of 66 percent for Senate Republicans), as rated by the Heritage Foundation.
The only other Republican Senator to score a perfect 100 on the Heritage scorecard so far in this Congress has been Ted Cruz, who defeated another Republican establishment figure, the moderate conservative Texas Lieutenant Governor David Dewhurst, in the 2012 Texas Republican primary. That race was uniquely instructive of a key feature driving political polarization in America’s electoral environment: the intense motivation of highly ideological voters, who are much more likely to dominate in low-turnout primary elections. Texas requires a run-off primary election if no candidate wins a majority in the first round. Dewhurst, who was endorsed by most Texas Republican office-holders, including Governor Rick Perry, and who spent huge sums of his personal fortune, won 45 percent in the first round to 34 percent for Cruz. But with nearly 300,000 fewer voters in the Republican run-off elections, Cruz crushed Dewhurst in the run-off with 57 percent of the vote.
Elsewhere as well, Tea Party Republican primary voters have been getting what they mobilized for. Nebraska Senator Deb Fischer, who with the help of the Tea Party upset two better-known and much better-funded Republicans in the 2012 primary, ranks in the top third of the Senate’s most conservative Republicans, with a 79 percent Heritage score for the current Congress. Two other Tea Party-endorsed insurgents who defeated less conservative Republican primary opponents in 2010, Rand Paul (Kentucky) and Marco Rubio (Florida), have among the top ten most conservative Republican Senate voting records this term (at 86 percent and 91 percent respectively).
It is generally believed that Tea Party victories in 2010 Republican primaries cost the party seats it could have won in Nevada and in Delaware, where nine-term Representative (and former Governor) Mike Castle, one of the few remaining Republican moderates in the House, was dealt a stunning defeat by conservative activist Christine O’Donnell. In the general election, she lost badly to a Democrat who seemed sure to lose to the popular Castle.
These Tea Party victories are only the latest chapter in a long-term and abundantly documented trend of increasing polarization in the U.S. Congress. Following an extended period of ideological polarization between the parties in the Senate and especially the House in the late 19th and early 20th centuries, party polarization in Congress attenuated for well over half a century, from roughly the late 1920s to the late 1980s. With moderate Republicans from the northeast and west and conservative southern Democrats, there was sufficient diversity within the parties to enable shifting bipartisan coalitions that could produce deals on budgets, taxation, civil rights, environmental protection, and other major legislation. But since the late 1980s, party moderates have become a dying species. Save for minority-majority districts in the House, the South has gone virtually solid conservative Republican (with moderate Democrat Bill Nelson from Florida the lone holdout in the Senate). Meanwhile, the coasts have been drifting more solidly leftward: All six West Coast Senate seats, and most in the Northeast as well, are now held by liberal Democrats.
As Thomas Mann and Norman Ornstein note in It’s Even Worse Than It Looks, the 111th Congress (during the first two years of Barack Obama’s presidency), marked the first time in modern history that there was no ideological overlap between Republicans and Democrats in the Senate. In 2014, National Journal reported that 2013 marked the fourth straight year when there was zero ideological overlap in the Senate, while the overlap in the House was the lowest ever (with only two Republicans more liberal than the most conservative House Democrat). It concludes, “The ideological sorting of the House and Senate by party, which has been going on for more than three decades, is virtually complete.”
Mann and Ornstein argue that Congressional polarization is “asymmetrical”, driven mainly by the right-wing lurch of the Republican Party. But as the remaining few Democratic centrists in the House gradually retire and moderate Southern Democrats have been defeated, Congressional Democrats have also become increasingly liberal. Meanwhile, the Democratic Party seems to be drifting leftward in the 2016 presidential primaries, with Vermont Senator Bernie Sanders (a self-described “democratic socialist”) mobilizing crowds and rising in the polls with a leftwing populist message. This is the third consecutive contested Democratic presidential primary campaign when the more left-of-center candidate (Howard Dean in 2004 and Barack Obama in 2008) is running significantly better than expected. Not surprisingly, the more moderate frontrunner, Hillary Clinton, is being pulled to the Left in response.
With the rising economic populism of the party’s left wing and the mobilization of progressive interest groups around Massachusetts Senator Elizabeth Warren and her anti-establishment agenda, there are growing signs of the symmetry Mann and Ornstein could not earlier discern in the polarization of party elites and activists. In the race to succeed Senator Barbara Mikulski in Maryland, firebrand four-term Congresswoman Donna Edwards is facing off against one of the most accomplished and effective members of the House Democratic caucus, Chris Van Hollen. Although the two are not that far apart in their voting records, Van Hollen has a history of attracting Republican co-sponsors to his bills, while Edwards is a party purist. Because of this, and especially because of his willingness to consider restructuring social security as part of a broad budget deal, Van Hollen has earned the enmity of a variety of “progressive” groups that have mobilized behind Donna Edwards. Because Maryland is one of the most heavily Democratic states in the country, the party seems little concerned about losing the seat if its Senate nominee is too extreme. In this sense, Edwards’s bid in Maryland may resemble Ted Cruz’s 2012 campaign in Texas, though the race could be complicated by the entry of several other strong Democratic candidates.
The growing ideological distance between the two parties has deep sociological roots and will not yield dramatically to institutional fixes. A wealth of research testifies to the impact of “residential sorting”, in which people with strong partisan and ideological views literally move toward one another. Thus, the South becomes more solidly conservative and Republican, and the West Coast increasingly becomes “the left coast.” Within states, conservatives are moving to the exurbs and voting solidly Republican, leaving safe Democratic seats in the cities (often dominated by African-American and/or Hispanic populations) and contested space in the suburbs. Eliminating partisan control of legislative redistricting is a good thing for democracy, because it will eliminate the ability of dominant parties at the state level to entrench and deepen their electoral dominance through gerrymandering. In the absence of gerrymandering, we are likely to see at least somewhat more competitive districts, which are more likely to elect more moderate candidates who must appeal to swing voters. But we should have no illusion that non-partisan redistricting will have more than a modest remedial effect on partisan polarization in the Congress.
A better clue to a remedy can be found in the recent campaigns of two moderates who fought back and won after losing their party primaries to more ideologically pure candidates. In 2006, the centrist Democratic incumbent, Joe Lieberman, lost the Connecticut Senate primary to left-of-center challenger Ned Lamont by 52 to 48 percent. While Lieberman was one of the most conservative and hawkish Senate Democrats, he was also a strong supporter of labor, social security, the environment, and gay rights, and he had an “F” grade from the National Rifle Association. In short, he was a moderate. But unlike other moderates who lost their party primaries and gave up the contest, Lieberman gathered the necessary signatures to contest the general election as an independent and thoroughly beat Lamont in November, 50 to 40 percent.
What enabled Lieberman to come back in the general election was a quirk in Connecticut election law—or rather, the absence of one that prevails most everywhere else. In 44 of the 47 American states that hold party primaries, defeated primary candidates are prevented from running as independents in the general election, either by the presence of “sore loser” rules (which explicitly forbid it) or by simultaneous registration dates for the primary and general elections. Eliminating the sore loser rule (explicit or de facto)—and thus enabling more moderates to come back in the general election if they are defeated by party militants in the primaries—should thus be an important target of electoral reform.
The most passionate advocate for eliminating this arcane rule is one of the country’s shrewdest and most experienced Congressional moderates, former Republican Representative Mickey Edwards of Oklahoma. Edwards notes that in the run-off election for the Texas Republican Senate Primary in 2012, Ted Cruz won only 630,000 votes in a state of 26 million people. If David Dewhurst had been able to come back in the general election as an independent, he likely would have beaten Cruz for the same reason that Lieberman beat Lamont: defectors from the opposition party choosing “the lesser of two evils.” The same would have been true for Robert Bennett in Utah. Lugar in Indiana and Castle in Delaware faced more complicated situations, since they were not running in solidly Republican states. Still, if the other reform I will propose had also been in place, they, too, might well have prevailed in the general election as independents.
The option of running on the ballot as an independent was not available to Alaska Senator Lisa Murkowski when she lost to Tea Party candidate Joe Miller in the August 2010 Republican Primary by a mere 2,000 votes. But Murkowski also did not go gentle into that good night. Instead, she mounted an implausible bid as a write-in candidate, and beat Miller in the general election by 10,000 votes, 39 to 35 percent. Thus, one of the most moderate Republican Senators (with a Heritage Foundation scorecard of only 33 percent in the current Congress) was re-elected, and perhaps emboldened to steer an even more moderate course.
What if moderate Senators and House members from both parties had confidence that moderation would pay—that instead of costing them a primary election, it might ensure them re-election in November, even if by a more difficult and unconventional path? What if centrist and pragmatic challengers—both partisans and independents—had a way to face down the most ideologically committed party members (who are also the most likely to vote) and win? What would this option look like?
There is a simple and feasible electoral reform that stands a good chance of emboldening moderates and facilitating their electoral success: Ranked Choice Voting (otherwise known as the Instant Run-off or the Alternative Vote). Under Ranked Choice Voting (RCV), voters rank all the candidates, or some number of them, in order of preference. If no candidate obtains a majority of first-place votes, the candidate with the lowest number of first-place votes is eliminated, and his or her second-place votes are redistributed to the other candidates. The process of elimination and redistribution of lower-preference votes then continues until a candidate gains a majority or wins a final two-candidate face-off.
For several reasons, RCV holds some promise of reducing polarization and enhancing democracy. Most of all, it would enable voters to opt for moderate and independent candidates in a general election without fearing that they would be “wasting” their vote on a “spoiler”, thereby electing the candidate they like the least. Increasingly over the past quarter century, American voters have identified not with one of the two parties but as independents. According to the Gallup Poll, this trend reached a new peak in 2013, when 42 percent said they identified as independents, compared to 31 percent with the Democratic Party and 25 percent with the Republicans. The bulk of these independents are more moderate than party identifiers. But most of them lean Democrat or Republican, and so they have a pretty clear idea whom they lean away from and hope will lose. This is why most three-way general election contests eventually see support for the independent melt down as the general election approaches. In the 2014 Maine gubernatorial election, an exit poll showed that independent Elliot Cutler (who came within two points of winning in 2010) would have beaten the Republican incumbent under RCV—or under Approval Voting, which also puts a premium on moderation and punishes the extremes. In reality, he won less than 10 percent of the vote due to voters’ fear of the spoiler problem.
The beauty of RCV is that it frees voters to vote sincerely rather than strategically. Left- or right-leaning independents who would prefer to vote for an independent centrist could do so knowing that if their first preference finished third, their vote would be transferred to the candidate or side of the partisan divide that they found least obnoxious.
RCV would enable the election of a growing number of candidates for governor or other statewide office, or for the Senate or House or a state legislature, who promise to be flexible and pragmatic in making budgets and forging compromises on polarizing issues. By neutralizing the spoiler problem it would likely increase the supply of moderate, unaffiliated candidates. In a legislative chamber, a sufficient number of independents could form a block that would gather in more flexible Democrats and Republicans and build winning coalitions from the center out. Some day they might even hold the balance of power in organizing the Senate or House, vowing to give partisan control to whichever party offered to be more flexible and responsible with respect to a core issue agenda—for example reducing the Federal budget deficit through a package of both spending cuts and tax increases, or addressing climate change by raising Federal fuel taxes on gasoline and diesel fuels while lowering other taxes to compensate for the added burden on lower-income consumers.
A moderate does not have to be a milquetoast, and centrism is not always a point in the bland middle between two extremes. There is lurking to break free from the current stultified confines of the two major parties a creative center ready to pursue innovative and even bold policy options that challenge some of the ideological orthodoxies and untouchable interests of both. In a context of more fluid bargaining, more moderate (or at least pragmatic) legislators in each party might be induced to pursue some grand bargains on policy issues that are now frozen by the deepening partisan and ideological divides.
Of course, it would not only be moderates seeking influence under the new electoral rules of RCV. Other parties would also contest, some of them to the Left of the Democrats (for example, the Green Party) or to the Right of the Republicans (such as the Libertarian party on taxation, and perhaps a new social conservative party). But the difference between RCV and another globally popular electoral reform, proportional representation, is that RCV uses single-member districts and requires that one candidate obtain a majority (or large plurality) of the vote. Thus it tends to isolate rather than reward the extremes. Nominees of the two parties would have to weigh whether they would pick up more second- and third-place votes than they would lose by trying to appeal to the ideological flanks rather than to the center. In most races, more extreme voters would have to give their lower-preference votes to the major-party candidate or risk electing someone even less to their liking. This could leave the major party nominees free to move toward the center to pick up the second-preference votes of a moderate independent. Or in districts or states where one party is pre-eminent, it might enable a moderate party rebel—a Richard Lugar or a Lisa Murkowski—to defeat a more ideologically pure competitor by rallying a general election coalition of moderates from their own party, independents, and partisans of the other party looking for the lesser evil.
One great advantage of the RCV is that it would not require any new national legislation, not to mention constitutional change. The same system of single-member districts would be used, and indeed the same districts could be used (but hopefully over time the elimination of gerrymandering would produce more competitive districts). Creating multi-member districts under PR would require new enabling legislation from the Congress. Very small three-member districts could attenuate the danger that the current two-party system would fragment into multiple parties (a common problem with PR, since the larger the district size the more proportional the system and hence the more parties in parliament). The advantage of three-member districts to elect members of the House is that they would be likely to elect one member from each party in a three-person district, since it would only take 25 percent plus one vote to elect at least one candidate. Thus Massachusetts might have two or three Republican Congressmen, rather than nine Democrats and no Republicans, while Alabama (6-1) might elect at least one more Democrat. The nearly vanished breed of moderate northern Republican and conservative southern Democrat might return to the House. But even three-member Congressional districts would increase the distance between constituents and their representatives, and proponents of the reform would probably be best advised to first test its efficacy at the level of state legislatures.
RCV is not only a simpler and less risky electoral reform, it is also more feasible, for two reasons. First, expanding choice is intuitively appealing to voters, who have embraced it when given the opportunity. This is why RCV is used in a growing number of American cities, including (in California) San Francisco, Oakland, Berkeley, and San Leandro, as well as Minneapolis and St. Paul, Minnesota, and Portland, Maine. According to the non-partisan election reform center, FairVote, more than two million Americans now use RCV to elect their local officials (up from just 100,000 in 2003).
Second, since RCV uses the existing district system, any state can adopt it at any time. Of course, state legislatures—dominated almost entirely by the two parties—are unlikely to adopt a reform that would challenge their electoral duopoly by making it easier for independents to win. But in 24 states voters can impose this reform at the ballot box through the initiative process. As a result of a successful signature drive this year, this is exactly what the voters of Maine will have the chance to do in November 2016. Having elected an independent (Angus King) to the governorship and then to the U.S. Senate, only to see his Republican successor as governor polarize the state after twice edging out Democratic and independent challengers in first-past-the-post elections, it is not surprising that Maine may become the first state to adopt RCV for all state and U.S. Congressional elections.
Another tantalizing potential windfall of RCV is a possible reduction in negative campaigning. Research commissioned by FairVote shows that voters in cities that used RCV to elect their officials were somewhat less likely to perceive the campaign as being very negative than voters from non-RCV cities.
Ranked choice voting is a modest reform in another respect. It is perfectly compatible with the preservation of the party primary as the means for choosing a party’s nominee in the general election. Indeed, a strong case could be made for using RCV within each party primary to choose the nominees (and certainly RCV should be used to avoid a primary run-off election). That might slightly increase the odds of more moderate figures winning party primaries.
The best way to break the trend toward polarizing outcomes in party primaries, however, would be to widen the voter base in those contests through at least partially open primaries (in which independents could vote in either party primary but not both), and by moving the primary election date as far back toward the general election as possible. In 2014, well over half the states held their party primaries before the end of June—more than four months before the general election. The further in advance of the general election, the lower voter interest (and thus turnout) is likely to be. And in addition, the longer the general election campaign, the more expensive it is. Several American states (including Florida) held their primaries in late August, and a few shortly after Labor Day (perhaps the ideal time, as it would still leave two months for the general election campaign).
Some states have eliminated party primaries altogether. In 2010, California adopted the Top Two system. As in Washington State, all candidates contest in a single, non-partisan “blanket” primary election. Candidates can indicate their party preference, but this does not imply that the party endorses them, and the top two vote-getters advance to the general election irrespective of their party and regardless of whether one of them won a majority of the vote. (In Louisiana, the first round is the November general election, and there is a “top two” runoff election only if a candidate fails to obtain a majority in the first round.)
The Top Two system has several potential weaknesses in comparison with RCV. First, it undermines parties by eliminating partisan primaries. The problem with democracy in American today is not political parties per se, which remain indispensable for recruiting candidates, structuring policy choices, and mobilizing participation, but rather the increasing ideological (and in a way cultural) distance between the parties. Rather than erasing or degrading party identity, the goal should be to facilitate the nomination of more pragmatic and flexible partisan candidates, to encourage moderation and compromise by legislators once they are elected, and to “grease the wheels” of compromise through the limited and occasional election of independents who can build creative coalitions from the center out.
A second problem with Top Two is that it can deprive a political party of a chance to compete in November even if its candidates win a majority of the vote in the first round. This could happen if one party splits its vote evenly between two candidates while the other party’s larger vote total splinters in many directions. Top Two thus puts a premium on parties muscling out primary competitors for fear of dividing their vote—not exactly a democratic virtue.
A separate but related problem may arise when the general election contains two people from the same party. This was in fact the dream of the architects of Top Two—that in some general election contests, competing candidates from the same party would have to moderate to win voters from outside their party. But if two Democratic candidates are competing in a heavily Democratic district, the temptation will be to replicate the ideological outbidding of a party primary, leaving voters from the other party to sit on their hands in November. As FairVote’s executive director Rob Richie has noted, “the single biggest decline in general election turnout among all 50 states from 2010 to 2014 took place in California, where the Top Two primary had been introduced in 2012.”
And that is another major defect of Top Two: It winnows the field brutally in a (low-turnout) primary election, often several months in advance, leaving many voters feeling they have no one to vote for in the general election, and under-representing the views of younger and non-white voters, who are less likely to vote in party primaries. While Top Two tends to disenfranchise some voters, RCV gives every voter a chance to influence the general election outcome through the process of vote transfers to lower-ranking preferences. Perhaps most damningly, Top Two appears to have done little to reduce polarization; research by Boris Shor and Nolan McCarty showed California to be off the charts in 2013 with the most polarized state legislature in the country (far worse even than in the U.S. Congress).
Because of the defects of Top Two, Richie has proposed using RCV in both primary and general elections, starting with a “Top Four” primary election. This creative system would increase the number of contested general elections in one-party-dominant districts or states. As Richie notes, well over 80 percent of House races have virtually pre-determined outcomes based on party registration. The same (and worse) is true in the states, where four in ten state legislative races lacked competing party candidates, “and most of the rest were contested only in name.” Top Four would also virtually guarantee that a generally successful moderate incumbent—a Lugar or Bennett or Lieberman—would make it into the general election race (thus obviating the need for changes to “sore loser” rules). And it would make it easier for minor parties and independents to get on the November ballot. Top Four could carry some risk of undermining political parties by eliminating partisan primaries, but FairVote’s proposed version would allow a party’s endorsement of a candidate to be noted on the ballot. And RCV could also be utilized in a multi-seat constituency. A more modest reform would be to preserve party primaries while making them open (or partially open) and later in the electoral cycle, and then eliminate the sore loser rule so that a defeated moderate incumbent could contest as an independent in the general election.
A different reform that is quietly gaining momentum would ensure that the winner of the national popular vote for President wins the office. The instrument for doing so is not a constitutional amendment but a bill (so far adopted in 11 jurisdictions with 165 electoral votes) mandating that the state’s (or the District of Columbia’s) electoral votes be cast for the winner of the national popular vote. If states with just 105 more electoral votes adopt the bill, the winner of the next national popular vote for President will be assured of an Electoral College victory.
It is difficult to predict with confidence how reforms will work out in practice. That is why it makes sense to start with modest reforms and see what happens. In many ways, federalism has been a curse for electoral administration in the United States, inhibiting the development of national standards for impartial, effective administration and fair, inclusive rules. But in at least one sense federalism is a blessing, enabling a number of different experiments in electoral rules at the local and state levels. The growing pace of interest in and adoption of ranked choice voting is testimony to these possibilities. As Top Two and RCV (and who knows what other reforms) are exposed to experimentation, we will learn what works under what conditions. Several intriguing alternatives are emerging. However, I suspect that RCV, along with elimination of the sore loser rule, will wind up being embraced as the reform of popular choice. It does the least violence to our current system, and it may hold the most realistic promise of making more elections interesting, inclusive, and meaningful again.