Don’t Despair—Mobilize for Institutional Change

Even smart scholars think nothing can be done to solve the problem of polarization, but they’re unduly pessimistic: there are institutional reforms that would work if we muster the will to adopt them

Tom Edsall’s column for the N.Y. Times last week was full of despair. The theme was that partisan polarization in the United States has gotten so bad that the nation is at a point of no return and that there is nothing that we can do about.

I share the sense that the divide between Democrats and Republicans has gotten much deeper and wider than at any previous time since World War Two. But I emphatically don’t share the belief that nothing can be done about it. On the contrary, the main reason why I launched Common Ground Democracy is to publicize the point that there are institutional remedies for the problem of polarization.

It doesn’t have to be this way. Even accepting the preferences of voters as they are, which increasingly reflects the pernicious tribalism of political professionalism, there are institutional reforms that would cause government to reflect the views of the electorate’s center, rather producing politicians who increasingly cater to the extremes of their respective party’s “base” voters.

But before pointing to potential solutions, let’s review some of the despairing comments in Edsall’s essay.

Its very first sentence describes the current “racial and cultural conflict” as “insoluble.”

It quotes Richard Haass, the former head of the Council on Foreign Relations, saying “one of the comparative advantages of this democracy has been its ability to reform itself and correct mistakes, and our ability to do so now is much less certain.” Haass adds:

“I am no longer confident there is the necessary desire and ability to make this country succeed. As a result, I cannot rule out continued paralysis and dysfunction at best and widespread political violence or even dissolution at worst.”

Bruce Cain, a Stanford political scientist I know and admire greatly, told Edsall that we are stuck in a kind of electoral Catch-22: “We can’t have effective government until we have sufficient consensus, and we can’t have consensus unless the people in government aim for effective policy rather than notoriety and a media career.” As a result, Cain predicts: “we will wallow in a polarized, divided government for another term or two.”

Cain blames “the design of the Madisonian system” for preventing effective government when the nation is polarized as it currently is. (The “Madisonian system” refers to the fact that James Madison was the principal intellect behind the architecture of the U.S. Constitution, with its separation of powers and division of authority between the federal and state governments.)

Cain is not the only one. Isabel Sawhill, a policy analyst at Brookings, observed that the authors of the Constitution “thought they had designed a set of institutions that could weather the storms” caused by “human frailties and passions.” But Sawhill sees these Madisonian institutions as incapable of handling the forces that produced the partisan tribalism existing today; she calls this “the great misalignment between the institutions we have and those we need.”

Sawhill, however, can identify no better institutions that would remediate the misalignment she perceives. In fact, she tells Edsall that she and her colleagues “feel quite helpless — unable to find ways to improve matters.”

Pippa Norris, a comparativist at Harvard, echoes the critique of America’s existing electoral institutions. She points specifically to “plurality elections” and the “two-party system” they engender. (“Plurality elections” are those in which the winner needs to win only a plurality of votes—more than any other candidate—rather than a majority of votes, which is more than fifty percent.) As polarization between the two parties increases in a plurality-winner system, it “reinforces us-them intolerance among winners and especially losers, who increasingly come to reject the legitimacy of the rules of the game where they feel that the deck is consistently stacked against them.” But she offers no alternative institutional arrangement that would counteract this trend.

The truth is that there are alternatives to the plurality elections that, as part of the overall Madisonian system, are so problematic under conditions of high partisan polarization as currently exists. In future Common Ground Democracy essays, I will describe some institutional innovations that I’m exploring in my research. Here, I will summarize some reforms I’ve analyzed in prior work. (For those interested in the publications on which this essay is based, here are four papers: Tournament Elections with Round-Robin Primaries, Requiring Majority Winners for Congressional Elections, The Constitution and Condorcet, and Total Vote Runoff: A Majority-Maximizing Form of Ranked Choice Voting.)

One feature of plurality elections that is especially problematic under conditions of high partisan polarization is the use of party primaries to determine which candidates are on the ballot in the plurality-winner general election. (See Nick Troiano’s new book on this topic.) When the overall electorate is highly polarized, Team Blue’s primary will tend to produce a nominee who is considerably left of center, while Team Red’s primary will tend to produce a nominee on the far right. In the general election, the choice will be between two relatively extreme candidates with no option in the middle.

One reform that aims to eliminate this polarizing effect of partisan primaries is to adopt the kind of nonpartisan primary that California uses. In this nonpartisan primary, all candidates regardless of their party affiliation compete against each other. In California’s version of this system, the two candidate who receive the most votes in the nonpartisan primary advance to the general election. The two candidates might be two Democrats, or two Republicans, or an independent along with one candidate from either of the major parties.

In principle, this “top-2” nonpartisan primary gives moderate (less polarizing) candidates more of a chance to compete than a system of partisan primaries. Moderates, who can’t win their own party’s primary when the primary voters are themselves considerably more polarized—deep blue or deep red, clustered at each end of the partisan spectrum—than voters in the entire electorate, might be able to achieve one of the top two spots in a nonpartisan primary.

To illustrate this point, consider the idea of using California’s top-2 system in the context of this year’s presidential election. As I write this on the eve of Iowa’s Republican caucus, it seems very unlikely that Nikki Haley will be able to beat Donald Trump in the Republican primaries in order to win the party’s nomination, even though polls consistently show that Americans as a whole would prefer her instead of Trump. (She consistently outperforms Trump against Biden, demonstrating more general election voters would prefer her rather than Trump. See, for example, this latest CBS poll.)  But in a top-2 system, where all voters—independents as well as Democrats and Republicans—choose among all candidates, it is possible that Nikki Haley could emerge as one of the top two candidates. (In the general election, if the polls are an accurate prediction, she would beat Biden.)

Alaska uses a version of California’s nonpartisan primary that gives moderate candidates even more of a shot. Alaska’s system sends the top four candidates in the nonpartisan primary, rather than just two, to the general election. Then, for the general election, Alaska uses Ranked Choice Voting to pick the winner among these top four candidates.

Imagine Alaska’s system being used for this year’s presidential election. We can assume that Nikki Haley would be one of the top four to advance to the general election, along with Biden, Trump, and Ron DeSantis. Then, given the way Alaska’s version of Ranked Choice Voting works, Haley doesn’t even need to finish second among the four in order to be able to win. Suppose she comes in third behind Biden and Trump, but DeSantis comes in fourth. If enough DeSantis voters pick Haley rather than Trump second, then Haley could leapfrog over Trump because of the RCV tabulation procedure. Then, assuming Trump’s voters prefer Haley to Biden, once Trump is eliminated along with DeSantis, Haley could emerge victorious even though she started out in third place. (RCV eliminates the candidate with the fewest first-choice votes and then redistributes the ballots that ranked the eliminated candidate first to whichever candidate is ranked second on those ballots; RCV repeats this elimination-and-reallocation procedure until a candidate secures a majority of votes.)

As powerful as Alaska’s system is in counteracting partisan polarization, there is an even more powerful variation of Alaska’s system. Alaska’s particular method of tabulating the results from its RCV is not the most anti-polarization tabulation method available. Instead, a modest tweak in the tabulation procedure would assure that the least polarizing candidate always prevails.

In this essay, we don’t need to consider all the details of this alternative tabulation procedure. It suffices to say here that this alternative RCV method will always elect a candidate whom a majority of voters prefer to each other candidate on the ballot. So, consider the possible use of this alternative RCV method in the context of this year’s presidential election. Suppose, again, that the top four candidates in the nonpartisan primary would be Biden, Trump, DeSantis, and Haley. Regardless of the relative position of these four candidates in the first-choice rankings on the RCV ballots, and thus regardless of the result that Alaska’s elimination-and-redistribution procedure would reach, Haley would win the election if a majority of voters would prefer her to each of the three other candidates head-to-head (as is the case if current polls are correct).  

The point of this analysis is not to say that Haley should win the presidential election. On the contrary, if the kind of system I’m describing—with the alternative method for tabulating RCV results—is used, it likely would induce even more moderate candidates than Haley to enter the race. Larry Hogan, Joe Manchin, and other potential contenders for the No Labels centrist ticket would almost certainly jump into the race if this kind of electoral system were in place. (And they wouldn’t be a spoiler, as they would by entering the race with the existing electoral system.) Then, as long as a centrist was among the top four candidates in the nonpartisan primary—perhaps eking out DeSantis for fourth place—then the centrist (and not Haley) would win the general election if a majority of voters preferred this centrist to each of the other alternatives head-to-head. And, indeed, we can easily imagine a majority preferring Hogan or Manchin to either Biden, or Trump, or Haley, when considered one-on-one against each. (Republican voters, along with moderate independents would prefer Hogan or Manchin to Biden, and Democratic voters along with moderate independents would prefer Hogan or Manchin to either Trump or Haley.)

Thus, if this kind of alternative RCV system were used for presidential and congressional elections, American government would be entirely different from what it is today. The presidency and congressional seats would tend to be occupied by more centrist candidates, who can achieve policy compromises on behalf of the American people, instead of the polarizing politicians whom the existing system puts into office.  There wouldn’t be the electoral Catch-22 and resulting legislative paralysis that Bruce Cain and others lament.

Thus, the proper response to the pathology of hyper-polarization afflicting American politics is not despair and helplessness. Rather, it is to adopt the kind of electoral reform most likely to produce anti-polarizing winners. Getting this adopting won’t be easy. But it’s not impossible. Roughly half the states have ballot initiative procedures that could be used to enact this reform—just as ballot initiatives were used in California and Alaska to adopt those versions of nonpartisan primaries.  

The acute danger of the despair expressed in Edsall’s column is that it becomes a self-fulfilling—and self-defeating—prophecy. Instead, we should roll up our sleeves and get to work as fast as we can putting in place the kind of reforms that are capable of reversing this accelerating hyper-polarization.