Redistricting reform has been on the march in recent years, with about a dozen states embarking on systematic reforms of how district lines are drawn, especially out West (California, Arizona, Colorado, Washington, Idaho, Alaska) but also spreading to states back East like Ohio and New Jersey. And gerrymandering has become a front page issue nationwide, fueled by Barack Obama and other leading Democrats who have decried Republicans’ fiendish ingenuity in stacking the process to their advantage - although their own party has done likewise in many states where it has had a chance to call the shots. Republicans like former California Gov. Arnold Schwarzenegger have joined the chorus. And the constitutionality of partisan gerrymandering is before the U.S. Supreme Court, not for the first time.

Through all this, libertarians have mostly stayed on the sidelines. When I mention that I am active in efforts to curb gerrymandering, some people react with surprise: “Oh, is that that a libertarian issue?”

It should be, I think. Libertarians are in some ways especially well situated to spot the harms that can result when politicians get to select which constituents they would like to represent rather than vice versa. And the issue fits well into a long tradition of classical liberal thinking about electoral process and representation, among the goals of which is to restrain existing establishments from gathering too much power unto themselves.

But first some background.

 

Deciding Who Decides - And How

In any system based on geographical representation, someone must decide which voters should be grouped with which others into districts. As population grows and voters move around, someone must also redraw the lines periodically to reflect those shifts, at least under modern standards of fairness. Under our system, both state and federal legislative districts must be reapportioned at least every ten years to reflect new census results.

Here in the United States, the task of apportioning both Congressional and state legislative seats has historically fallen to state governments. Note, however, that the U.S. Constitution from its start has expressly granted Congress a role in overseeing how states hold elections for its own members. Article I, Section 4 of the document provides: “The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof; but the Congress may at any time by Law make or alter such Regulations.”

With the power to draw lines comes the power to punish or favor selected candidates. Even before we had the familiar name – derived from the dragon-shaped Massachusetts state senate district that Gov. Elbridge Gerry helped devise back in 1812 – gerrymandering was a well-known practice. Historians debate whether Patrick Henry used it to draw a deliberately unfavorable Virginia district for James Madison in hopes of keeping him out of the House in the 1789 election. (Madison won anyway.)

The first line of defense against the practice is to specify clear rules in advance governing how districts are drawn and make them legally binding upon the line-drawers. Compactness in districts, for example, is almost universally acknowledged to be one good principle, and success in achieving it need not be left to intuition: mathematical formulas are available by which to quantify the compactness of a map and compare it to alternatives. Likewise with another common redistricting principle, the congruence of district lines with smaller political subdivisions, such as counties and towns: a state can adopt a rule minimizing the number of county splits or providing that more populous counties should be split in preference to less. In safeguarding against manipulation, clear and specific marching orders are better than broad grants of vague authority for line-drawers to balance multiple factors or to discern so-called communities of interest. Likewise, states can enact more or less effective methods of judicial review to ensure that line-drawers follow the announced principles.

These background rules account for no small part of the difference in the gerrymandering landscape. States with firm rules tend to have less gerrymandering than those without. And that suggests one way in which state-by-state reform, without need for institutional or legal innovation, can improve matters: enact clear and objective marching orders backed by judicial review.

 

Why It Matters

In a world with many injustices and bad government policies, is this an issue worth caring about? Elections will always have unhappy losers, after all, and complaints of unfairness can seem amorphous, especially if no one’s individual rights have been violated.

Every imperfection in a representational system introduces a bias for or against some political group: large or small states, rural or urban constituents, committed or low-intensity voters. But gerrymandering introduces a bias specifically to the benefit of incumbents and those they favor.

In this respect one is reminded of the movement for term limits, in which libertarians have been prominent. Could America, or some parts of it, have developed a dominant political class that tends to perpetuate itself, shutting out rivals and newcomers? If so, then high among its preoccupations will be to keep its grip on office. One symbol of this entrenchment is the five-term Senator or fifteen-term Representative, too powerful and feared for anyone to dislodge. But another symbol is the dominant party that cannot be ousted from its control of a legislature unless the other party manages to score a sweeping win, with not 50 but more like 55 or 60 percent of the vote.

These days, the gerrymander is known above all as an instrument of party advantage: the governing party draws as many districts as it can in which its own adherents hold a comfortable though not overwhelming lead, while packing opposition voters as densely as possible into as few districts as it can.

Gerrymandering serves to entrench incumbents in other ways as well. Within a party, for example, it can safeguard incumbents from primary challenge. Unfriendly internal factions can be broken up or submerged in opposition districts. And since districts that sprawl around without logic are also more expensive to campaign in – the cost of advertising in multiple media markets is higher, for example – the advantage may go to those candidates, especially incumbents, who can raise money readily and cultivate allies among those interest groups that can turn out disciplined voters statewide.

The threat of drawing a hostile district can also be an important means of preventing dissident voices from being heard within a majority party: cross the leadership, and you just might get cut into a tough new district next time.

Beyond all this, rotation in office has a claim to standing itself among the safeguards of liberty. When there is no plausible prospect of being thrown out in favor of a rival party, there is less constraint on the thievery and high-handedness of the party in power, and less likelihood that the light of scrutiny will be cast upon it.

 

Beyond Partisanship

The fight over gerrymandering can sometimes come off as a battle between partisan insiders. But that can put libertarians – who often feel themselves standing outside the main political tribes – in a position to offer some distinctive perspectives.

Consider, for example, one of the most common fixes offered for redistricting problems, the bipartisan commission. What could be a more straightforward fix than to empanel a half-dozen loyal Republican lawmakers, a half-dozen of their Democratic colleagues, and perhaps a tie-breaking retired dignitary to supervise the drawing of a district map?

Libertarians will perhaps be quicker than many others to spot the weak points of this plan: to begin with, bipartisan does not mean nonpartisan. Maps drawn by such methods may avoid gross bias between parties by adopting a cozy “you can protect your members if we can protect ours” approach. But independent and third-party voters – not to mention insurgent political movements not yet represented in insider circles – will have no one to look out for their interests at all.

Or consider the question of “blinding,” that is, directing a panel not to acquire or consider data on such matters as current party registration or past voting records in assembling population blocks. An even more powerful technique is to blind a panel to the residence of any individual, such as any incumbent. Many bipartisan panels consisting of Republican and Democratic loyalists would reject such a proposal out of hand, while an outsider might be more intrigued by it as a way to help avoid catering to incumbent interests.

Then again, libertarians will often have in mind the constitutionalist maxim that official powers may be best divided among actors with some jealousy of each other, with ambition checking ambition and interest checking interest. When redistricting plans are adopted through the ordinary state legislative process, there is already some of that: a governor’s political goals and interests will often diverge from key legislators’, so that a veto threat can serve as a check on certain excesses.

But there might also be other or additional ways to divide power or counterpoise interest against interest. Under some reform plans, for example, multiple interested parties, or members of the public at large, are invited to submit proposed maps, and the redistricting panel then chooses the plan it considers to hew most closely to the stated redistricting principles. One big advantage of this approach can be to lay a more favorable groundwork for later judicial review, since the panel that rejects a facially better plan may find its decision coming under later scrutiny by a court.

 

In Search of a Process

A more fundamental and challenging question is: Is there some way that is both practicable and constitutionally sound to take the whole process out of the hands of those with a vested interest?

It’s not as easy as it may sound. For example, “hand the whole thing over to the courts” might look like an attractive option. But judges are incumbents too,  and if they are not federal judges they probably lack life tenure. In some states the bench is already too politicized for comfort, and one consequence of giving judges much more authority to draw district lines might be to incentivize other actors to politicize state judicial selection further.

These days a lot of the momentum backs the idea of placing the responsibility with independent redistricting commissions made up of citizen volunteers. The Arizona and California models, and others proposed since, each have their own details. They typically exclude persons who are considered too close to elected officials and involve a stage in which some neutral entity screens citizen-volunteers in search of those with enough civic aptitude to rise to the complexities involved. A randomization stage – think jury selection – may be used to reduce the likelihood that any powers that be can stack a panel with known and trusted friends.

How have these innovations worked in practice? Arizona’s has been in operation for two census cycles, California’s for one, and both have results that could be described as mixed. Arizona’s independent commission has come under pressure from litigation and political meddling, while in California, savvy interest groups managed to organize quietly to influence some of the proceedings, as a ProPublica investigation found. Still, many observers believe the quality of districting has improved in both states, and both systems are works in progress, capable of correcting design problems as they go and attracting new constituencies to monitor and counteract attempts to manipulate them from the outside. In short, it’s still too early to pronounce with confidence on how they will do.

We may dare to hope, however, for an enduring success. The great British classical liberals, like Bright and Macaulay, did not draw a sharp distinction between the substantive reforms for which they crusaded – say, reducing tariffs or removing legal hardships for minority religions – and the reform of representation, the franchise, and electoral procedure. Among the goals of the Great Reform Act of 1832 was to do away with the notorious “pocket” or “rotten” boroughs, settlements, sometimes tiny or half-abandoned, which had ancient rights to elect members to Parliament and which were in practice the possession of some great landowner who directed his tenants’ choice. Gerrymandering is neither as flagrant nor as readily abolished an evil as the pocket boroughs, else it would not have lasted so long. But in a sense it recalls a particular subgroup of the pocket boroughs, the “crown boroughs,” in which the landlord was the admiralty or some other department of state, and which could thus be voted by the government itself to advance its majority at the next election.

The crown boroughs were at length abolished, and no one misses them.

It is my pleasure to contribute to the dialogue on redistricting reform. I agree with Walter Olson that libertarians should care about redistricting reform. A little over a decade ago, when I held an appointment at the Brookings Institution, I collaborated with another Cato Institute scholar, John Samples, to conduct a project to promote electoral competition. In our co-edited book, The Marketplace of Democracy, we argue that like free markets democracy works best when candidates compete for votes through a free exchange of ideas. Gerrymandering violates this principle. When politicians are selecting voters, rather than voters selecting politicians, incumbents become insulated from the will of the people. Our democracy then resembles some of the world’s worst despotic regimes, where the people have little voice in who represents them.

I’ve been involved in redistricting since the late 1980s. I’ve been on the inside consulting for legislatures and commissions. I’ve been on the outside as an expert witness in litigation challenging adopted maps and working with public interest groups advocating reform. From my experience, I largely agree with Olson’s able synopsis of the available reform and legal options to curb gerrymandering. I’d like to elaborate on one reform pathway in particular that is compatible with the concept of a free market: requiring transparency and public participation.

 

Redistrict in Sunlight

Transparency in redistricting requires that a legislature or commission operate in sunlight. Transparency enables the public - and the courts – to determine if a district line is drawn to the benefit a political party or candidate.

Politicians prefer opaque redistricting to shield their actions from the public. Legislators do this by hiding behind legislative privilege, the deference courts give legislatures to keep their internal discussions private. Plaintiffs in litigation are hard-pressed to show the presence of illicit intent if they can only provide circumstantial evidence. All too often, defendants counter gerrymandering allegations by arguing that a suspect line was drawn to respect a “community of interest.” Without knowledge of the internal deliberations about the supposed community, there is no way for plaintiffs to verify defendants’ claims. Since the evidentiary burden lies with plaintiffs, they lose.

The good news with respect to transparency is that courts have become less willing to grant legislative privilege in redistricting cases. In recent Florida, Maryland, and Wisconsin litigation the courts have limited legislative privilege. The bad news is that politicians have dug deeper by hiring law firms that in turn hire consultants to draw redistricting plans. Politicians can then exert a different right: attorney-client privilege. Consultants may operate with weaker records-retention policies than a government and may thus destroy evidence before it can be divulged through court-imposed discovery.

Maps considered during the redistricting process are key pieces of evidence to establish the presence of gerrymandering. Redistricting is a very complex problem. Consultants typically draw many test maps before a final map is proposed. These test maps may embody different approaches to achieve the goals of those in charge of redistricting. Comparing maps drawn in earlier stages of the redistricting process to those drawn later can reveal how districts are fine-tuned to the advantage of a political party or an incumbent. A final map may have more districts that lean toward a political party than earlier versions, or an opponent’s home may be cut out of a final version of a district, whereas it was in the district in an earlier draft map. Transparency is a powerful tool to reveal such chicanery.

Public participation also assists in revealing illicit intent. When a member of the public draws a redistricting plan that conforms to federal and state requirements, they put a legislature or commission on notice that their final plan must at least meet these requirements to the same degree as the public’s map. Indeed, the Pennsylvania Supreme Court invalided the state legislative plan because a member of the public – a piano teacher – drew a plan that fared better on the state constitutional requirements of compactness and respect for local political boundaries.

 

Inside and Outside Strategies

There are two strategies to promote transparency and public participation in redistricting: from the inside and from the outside.

From the inside, legislatures and commissions can commit to the principles of transparency and public participation. They can hold all meetings in public, so that everyone witnesses the deliberations on how district lines are drawn. They can create web portals for the public to obtain access to the data and mapping tools needed for the public to create DIY maps and to submit these for consideration. When a legislature or commission adopts a final plan, they can state what considerations went into the drawing of each district so they cannot devise post-hoc explanations to later cover for their gerrymandering.

Some legislatures and commissions are bound by state constitutions or statutes to adhere to the principles of transparency and public participation. Where they are not, those in charge of redistricting can choose to bind themselves to these principles. It is in their best interest to do so, since courts may very well permit plaintiffs access to internal deliberations and test maps, which can prove very embarrassing and damning to defendants.

If legislatures and commissions will not impose transparency and public participation on themselves, we can do it for them. Prior to the 2010 redistricting, my colleagues – Micah Altman, at MIT, and Azavea, a Philadelphia-based GIS software company – and I created web-accessible redistricting software called DistrictBuilder that enables novices to draw legal redistricting plans. We developed the open-source DistrictBuilder software with support from the Sloan Foundation, the Joyce Foundation, the Judy Ford Wason Center for Public Policy at Christopher Newport University, and Amazon Corporation. We are not the only such effort; another notable application is Dave’s Redistricting App. Commercial vendors have also entered this sphere.

Advocacy groups have hosted redistricting competitions and other public mapping efforts across the country. An important but simple lesson from these events was that the public is capable of drawing legal redistricting plans. High school students, retirees, and everyone in between successfully created legal maps. Editorial boards used these alternative plans as yardsticks to measure plans adopted by legislatures and commissions to show that there were better ways to do redistricting. Legislators introduced some of these public plans as legislative bills, although no state legislature seriously considered them. However, when New York and Virginia courts invalidated districts, court-appointed special masters did draw remedial maps with ideas found in the public’s maps.

Our most notable success showed how the public can become full partners in the redistricting process through their participation. We deployed our software to draw Minneapolis city council districts. A newly formed citizen redistricting commission found their single copy of a commercial desktop mapping application inconvenient to use, so they turned to our web-accessible application to draw districts from their homes and offices. So, too, did a local Latino community group and a local Somali community group. These groups expressed their representational needs by drawing districts for their communities. The commissioners incorporated these districts into their final map, using DistrictBuilder’s sharing features. In the 2012 election the first Latino and Somali members were elected to the city council.

The Minneapolis experience also shows that there is a much cheaper way to do redistricting. Commercial vendors make millions of dollars by selling their software again and again to each state and local government that does redistricting, but only one application is really needed for an entire state. DistrictBuilder simultaneously supported in Minnesota congressional, state legislative, and local Minneapolis redistricting.

 

Conclusion

We are preparing for the next redistricting. We are collaborating with others to collect election data so we can evaluate the political nature of alternative redistricting plans. We are upgrading the DistrictBuilder software to take advantage of computing advancements. We hope to create versions of our software loaded with the necessary data in all fifty states plus the District of Columbia, so that anyone can perform DIY redistricting on the computing cloud. You can follow our efforts on www.publicmapping.org and can obtain the open-source DistrictBuilder code at www.districtbuilder.org.

 

I commend Walter Olson’s lucid framing of the gerrymandering dilemma. The libertarian perspective aptly focuses on improving electoral markets. He illustrates well how prevailing redistricting practices can stifle robust competition, as well as new voices and political accountability. Olson’s claims are rooted in a principled and reformist tradition of classical liberals who worked to banish pocket boroughs in Great Britain. Here I point up a few critiques of his position, noting that first principles and neutral procedures do not guarantee fairness, depending on how one defines it. I also argue that a bargaining framework might be better for pursuing fairness rather than trying to take politics out of the line-drawing process.

 

Assessing the Problem

Olson’s main concern is how gerrymandering entrenches incumbents. By incumbents he means individual officeholders as well as the parties in power. I endorse his broader point even though he overstates his case. Gerrymandering is not necessarily the most significant factor aiding incumbents. Regardless of line-drawing, officeholder advantages emerge from name recognition, cultivating a personal vote, and the ability to raise campaign money easily.

To be sure, incumbents try to safeguard their seat through redistricting. However, their self-serving behavior may bump against the collective interests of the party seeking to maximize seats in the legislatures. Incumbents prefer a bipartisan gerrymander that protects incumbents in both parties by increasing the size of their winning coalitions in their respective districts.  Party leaders, however, prefer a partisan gerrymander that entails spreading out the party’s voters across more districts while wasting the votes of the opposition party by packing them excessively into a district, or else dividing a cluster of opposition voters (“cracking”) so they cannot win majorities. The partisan gerrymander favors the party in power but it is a strategy that puts some incumbents at greater risk. Conceptually, Olson appears to conflate the incumbent and partisan gerrymander  In Gill v. Whitford, the case currently before the Supreme Court, the plaintiffs focus on the partisan gerrymander.

 

Competition Is Not the Only Goal

It is by no means clear that maximizing competition should come at the expense of other important values in a political system. Maps that maximize competition do not necessarily ensure fair representation of various interests, particularly for minorities. Nor is government likely to be stable or policymakers knowledgeable if politicians churn through the legislature from coin-flip elections. Additionally, legislators who live with intense fear of losing the next election may find it difficult to compromise on matters of interest to the broader polity. In short, there are trade-offs left unsaid by a singular focus on electoral competition.

To push the argument further, I want to take a stand for rent-seeking. Self-serving behavior in politics is not all bad if some structural accountability exists to constrain excess, and if previously marginalized voters stand to gain from it. Boss Tweed may have been corrupt, but a by-product of his corruption was that Tammany mobilized and empowered the much ostracized Irish-Americans. Less dramatically, gerrymandered rent-seeking gives incumbents the opportunity to build experience and influence that often benefits constituents. It also encourages an institutional loyalty among these incumbents and legislative power that bolsters the Constitution.

Let me be clear. I am not advocating rent-seeking but suggesting it might be tolerated to the degree it lubricates the capacity to govern in a diverse polity. The problem, of course, is when legislators lack any accountability because there is no latent threat of a challenge.

 

The Problem of Neutral Procedures

I want to address Olson’s focus on neutral procedures to guide the redistricting process. Using criteria such as compactness, congruence, and contiguity to guide line-drawing is invaluable. Such rules provide a foundation for consistency and legitimacy, while constraining the grossest of gerrymanders. But it must be acknowledged that any sort of principle could lead to unfair outcomes and is prone to manipulation.

Take the “blinding” approach. It has a powerful logic because it ignores political factors like partisan registration of voters or the home address of incumbents. But ignoring political factors – even if possible – may undermine Olson’s chief goal of increasing political competition. The natural spatial concentrations of partisan voters would make it harder to generate competitive districts without knowledge of voter loyalties. Moreover, given that racialized voting exists, blinding is less likely to produce districts that give blacks greater opportunities to choose preferred members of the legislature. Blinding would certainly make it difficult to meet the requirements of the Voting Rights Act. Finally, ignoring incumbency may undermine representation of voters who believe their current legislator knows and serves them well.

I agree with Olson that it is important to leave the courts out of these deliberations over matters of fairness and line-drawing as much as possible. Judges have their own biases. More critically, they may impose inflexible principles with long-term consequences that would be difficult to adjust down the road. Since there is no consensus in the polity about what constitutes “fair” we should avoid having the courts lock us into something before the democratic process can deliver on what might be an emerging consensus.  This concerns me about the upcoming decision in Gill v. Whitford where the court might rule on whether partisan redistricting is justiciable.

If not the courts, who should decide what is fair? Olson raises the possibility of citizen panels. I am not opposed to them, in principle, but I think he underplays their problematic nature. Such panels tend to follow the frameworks set out by elites who manage these bodies. And citizen panels, like juries, are just as prone to politics, particularly when stakes are high. To whom are they accountable? And how much will they be pressured by interest groups or rely on experts who manage the technical tasks of line-drawing requirements?

 

Better to Bargain

I have less faith in randomly selected individuals with weak ties to constituencies than politicians and interest groups who know how to bargain for themselves and those they represent. Rather than take politics out of it, better to channel it into a productive bargaining process. Political scientist Bruce Cain of Stanford University calls this a “pluralist” approach that emphasizes bargaining over neutral procedures. The aim is to ensure a place at the table for as many groups as possible, knowing the outcome will always be imperfect and that the logic in drawing maps may be inconsistent in any given cycle. The more important point is to balance interests in a domain that is highly complex and where notions of fairness are not settled. In these circumstances, muddling through makes sense.

The challenge of course is to create a legitimate bargaining process, and one that does not always end up in the courts. Cain points to the New Jersey Redistricting Commission, which uses a neutral mediator who facilitates the bargaining process. Another possibility is to have parties and interest groups commit to a binding arbitration process if they want to participate. Thinking outside the box, I could envision a political process in which the national political parties jointly draft a Memorandum of Understanding, with input from state parties, that lays out criteria. Both parties know what it is like to be on the losing side of a partisan gerrymander, which would temper their enthusiasm for publicly supporting the most egregious rent-seeking strategies. The MOU would not be legally binding but establishes a benchmark that the public and experts can hold them accountable for.

In the end, no matter how we slice it, any design introduces bias against a different set of voters. In cutting the cake, one’s notion of fairness depends on where we sit. By all means, let us look to criteria that improve the electoral market. But claims to take politics out of the process should be viewed with skepticism.

Thanks to Michael McDonald and Raymond J. La Raja for their kind words on my essay. I agree almost entirely with Michael McDonald’s points about the value of transparency, in particular open source data sharing methods that allow members of the public to employ free or low-cost software platforms to analyze underlying data and talk back. Public map submission can be especially powerful in providing a basis for judicial review, as his Pennsylvania anecdote (which is among my own favorites) shows. Public data transparency was among the key recommendations of Gov. Larry Hogan’s Maryland Redistricting Reform Commission, which I helped chair, when we issued our report two years ago.

I agree with much of what Raymond La Raja writes as well. Indeed, I may agree with him more than he realizes, since I have not been an advocate for elevating party competitiveness into a redistricting goal in itself by introducing some sort of artificial bias in favor of maps that yield more party-competitive districts. A better course, it seems to me on current evidence, is to pursue other neutral and impartial districting criteria and let the chips fall where they may on competitiveness. Given the geographic lumpiness of voter preference, the result is likely to be a mix of districts competitive between parties, districts competitive within a party via primary, and districts that count as safe for some incumbent or other well-situated person. Nothing wrong with those outcomes, so long as they have not resulted from the placing of a thumb on the scales.

Parenthetically, we can all agree that the federal Voting Rights Act introduces complexities by sometimes validating or requiring departures from otherwise neutral redistricting practice. This necessarily builds an additional stage into decisionmaking, since qualified experts or lawyers versed in VRA rules must be afforded a look at maps under consideration so as to flag any that (even by inadvertence) are VRA-noncompliant. But the VRA as I understand it is a law aimed at protecting minority voters, not particular incumbents or parties, so it’s hard for me to see why it should stand in the way of a project of blinding decisionmakers to the latter sorts of data.   

I can’t agree with, but as an intellectual matter will admit to being fascinated by, the contrarian case for deliberately allowing some measure of rent-seeking, incumbent perks, and insulation so as to grease the gears for cooperation within lawmaking bodies. It reminds me of the argument that it was a mistake to have curtailed congressional earmarks - they helped with logrolling, which led in turn to more consensus and better civility, or so the argument goes. And I know from overhearing elected officials chatting among themselves that some of them do yearn to loosen the tight grip that they believe the electoral process currently holds on them. We have to campaign constantly! We have to fundraise constantly! How can they expect us to act like Burke or Disraeli, Daniel Webster or Henry Clay? At least without a nice safe seat. That’s the ticket - more safe seats all round!

But if one conceives of the relationship between voters and representatives as one between principals and agents, I am not convinced that the answer is to give the agents more independence from the fear of being fired by their principals. Note also that one of the neutral redistricting criteria followed in many jurisdictions - minimize the shifts in district lines between one census and the next, so that not too many more voters are thrown into new districts than population shifts would require - already works to the benefit of political continuity and, I suspect in more cases than not, incumbent interest. 

Let me return to a note of agreement, though. All the available models of reform - the New Jersey “harness politics to tame politics” model, the Iowa civil service model, the citizen commissions, and so forth - have pluses and minuses. It’s way too early to call an end to experimentation and declare one the winner, especially when different methods may fit better into the political cultures and histories of different states. Prof. Bruce Cain has made a good case that the New Jersey model shows promise, even if it some of us worry that its structure might encourage bipartisan incumbent protection. Let a dozen other state-level experiments proceed. 

As the conversation unfolds I see areas of agreement with Walter Olson and Michael McDonald. First, we all view the current state of political competition as suboptimal for a healthy democracy. Second, we agree on data transparency and trying to engage the public in the process. I am less sanguine, however, about prospects for removing politics from the process (more on that below). Finally, Olson and I – and probably McDonald – agree that ongoing experimentation is good. There is no consensus about which model of redistricting is better than others, so variation and probing of outcomes is good.

Here are key remaining points of difference. First, regarding the goal of increased competition I believe we need to be realistic about what can be accomplished through redistricting. I think Olson recognizes this but I am not sure about McDonald. By all means, look for geographic areas where it is possible to draw competitive districts, but it is not clear to me how many we can produce, given structural constraints like geography, and “neutral” guidelines required by courts and statutes.

Perhaps a better way to introduce competition than redistricting is to loosen the regulations on third parties so they can be more competitive in lopsided partisan districts. I am guessing that Olson would agree with me heartily on this. We should also consider changing the winner-take-all rule and possibly use ranked-choice to stimulate competition. If we can reduce the bipolar standoff between the major parties, we might increase political competition and reduce the high stakes for redistricting.

I also want to emphasize my skepticism that a reformed redistricting process will dampen polarization very much, as McDonald suggests. We observe ideologues running even in competitive districts. The lack of competition and increased polarization has antecedent roots, which redistricting is unlikely to address that much. For example, new research by Andrew Hall and Danielle Thomsen suggests that moderates are disappearing from the ranks of potential candidates primarily because party activists are very unfriendly to them and governing is not fun when leadership holds all the power. In this way, voters are offered choices that are already polarized.

The discussion of how candidates emerge raises a related point. Perhaps we focus too narrowly on the general election in debates over redistricting and fair representation, particularly if one believes that nominations tend to produce candidates who are very far from the median voter. If independent commissions are to play a larger role in redistricting, they might evaluate the make-up of primary electorates when designing districts, particularly when the primary election is de facto the one that picks the officeholder. When a district cannot be made two-party competitive it might be constructed in a way that does not necessarily give winning pluralities to the most partisan and ideological voters. I am not holding my breath on this. But as McDonald asks rhetorically, where are the districts for moderates and independents? Point well-taken.

A second difference I have with my colleagues is about the process itself. I argue that we should not try to sidestep politics, but to channel it.  Unless we incorporate politics into the public bargaining process, its rivulets will submerge and resurface through interest groups with less accountability than parties and officeholders. By all means, continue developing criteria to attenuate the conflict of interest of politicians, but why not create an explicit Madisonian framework in which ambition is made to counteract ambition?

Highly polarized parties make it even more difficult to sidestep politics through independent commissions. The stakes are just too high. A small thumb on the scale can make a potentially balanced district tilt toward a conservative Republican or a liberal Democrat. More thumbs on the scale might ensure that majority power goes to a very conservative or very liberal party. In these circumstances, the parties will find surrogates or turn to the courts to wage their battles.

Instead, we should encourage a process by which each side – and the public – make iterative offers and counteroffers to move the process toward median positions that would lower the stakes. The commissions would then choose among the best plans that fit particular criteria rather than design the districts. As Bruce Cain argues, this would create incentives to build coalitions around plans rather than go to court in the service of parochial interests.

Many reformers have set up a false dichotomy based on what Christopher Achens and Larry Bartels call the folk theory of democracy. In this narrative, disinterested voters somehow make better choices than self-serving politicians. This is an unhelpful mythology. Either because of ignorance, manipulation, or their own self-interest, pure citizen commissions might bring as many problems to the table as officeholders or experts. Under poorly designed institutions, citizen commissions can produce outcomes that several groups think are unfair. And so the battle goes to the courts. The process repeats itself every decade and undermines its legitimacy. Any process should try to avoid the courts by lowering the stakes through bargaining.