Last week, Election Law @ Moritz was fortunate to host a conference on "Independent Election Administration: Who Draws the Lines, and Who Counts the Votes?" Organized by my colleagues, Dan Tokaji and Steve Huefner, this two-day event assembled from around the country a distinguished group of scholars in our field. (For a complete list of participants, see here.) While the conference, not surprisingly, did not yield any definitive conclusions on how best to achieve impartiality in the drawing of district lines or the administration of voting procedures, it did crystallize - at least for me - some key points concerning the quest for increased fairness in our nation's electoral systems.
Although the conference was initially structured to address both redistricting and vote tabulation, the proceedings developed a sharp impression that the two topics need to be addressed separately. To be sure, partisan bias may inappropriately infect both kinds of decisions, but the disease does not manifest itself in exactly the same way in both contexts, and the different conditions call for different cures.
Redistricting is the more complicated situation: ideally, a fair redistricting process would protect against undue incumbency entrenchment, as well as preventing one political party from capturing an excessive advantage against another. Nor is there any clear consensus on how a perfectly impartial "redistricting czar" should go about drawing the lines in any given case, in light of all the legitimate competing values at stake whenever lines are drawn (balancing, for example, both geographic and demographic considerations in an effort to make the legislature representative of the citizenry). Consequently, there was a sense at the conference that redistricting should be done by some kind of multi-member body that itself is representative of the citizenry, yet in a way that is not beholden to the self-interests of current legislators or political parties.
Heather Gerken offered the intriguing suggestion that randomly selected "citizens assemblies" develop redistricting plans to be submitted to voters for approval in a referendum. There were the inevitable questions about how such "citizens assemblies" would operate: who would lead their deliberations? who would provide information to them concerning redistricting options? for how long would they meet and how would they be paid? would service on them be voluntary or mandatory (as is the case with juries)? how large would they be and would pure randomness be preserved as the sole selection criterion, even if the result was a body that seemed distinctly unrepresentative of the citizenry on some important dimension (race, sex, age, occupation, level of educational attainment, level of personal income, and so forth)? Despite these important questions, the idea of redistricting by "citizen assemblies" remains worthy of pursuit.
The difficulty of structuring such "citizens assemblies" so that they represent the citizenry fairly reminds one of the difficulty of structuring a fair constitutional convention. The similarity is not surprising: the task of drawing district lines is part of setting up the basic framework for the operation of the legislature, which is the primary function of a constitution. Redistricting must occur more frequently that rewriting the state's constitution, but there is a strong consensus that it should not occur too frequently: no more often than once every ten years. One can think of redistricting as a process of episodically updating the basic constitutional architecture based on population shifts, and therefore the body that engages in redistricting can be conceived as a "limited-purpose constitutional convention" that confines itself to the fundamental assignment of drawing legislative lines without revisiting the rest of the constitutional design. In this light, it is appropriate to select members of the redistricting body in the same way as one would select members of a constitutional convention, and although random selection has not often been used as a method of appointing delegates to a constitutional convention, perhaps it should be - especially since any process by which delegates to the convention are elected from districts simply replicates the redistricting problem in the structure of the constitutional convention itself. In any event, the need for better structured redistricting bodies continues, and the idea of randomly selected "citizens assemblies" to conduct this task, subject to the approval of the citizenry as a whole, is a promising one.
By contrast, minimizing partisanship in the vote tabulation process, or what is often called "election administration" (addressing the range of voting procedures from registration through recounts), requires an institutional structure very different from a citizens assembly. Most states currently repose the ultimate administrative authority over these matters in a single chief elections officer, usually the Secretary of State. The problem with the current arrangement is that the Secretary of State is a partisan official, not that these administrative responsibilities are given to a single government official. Many of the administrative decisions that a state's chief elections authority must make are interpretations of existing statutes and regulations in the midst of fast-moving developments immediately before or after the election occurs. Given the need for speed, it is preferable that these decisions be made by a single individual rather than a multi-member body.
Moreover, because it is generally recognized that it is far preferable to have as much of the law of election administration codified in advance of the election (either by statute or by administration regulations promulgated through a notice-and-comment rulemaking proceeding), a well-designed system would minimize the degree of discretion available to the state's chief elections officer when making interpretative decisions in the midst of a fiercely contested election. Such discretion could never be eliminated entirely, to be sure - which is why this office must be insulated from partisanship insofar as is feasible. But the narrowly circumscribed discretion of this office, limited to the enforcement of specific statutory and regulatory directives, indicates that there is no need for a state's elections authority to be structured as a multi-member body representative of the citizenry in all its demographic and geographic diversity. This office truly functions as an elections umpire, and thus it should be structured in such a way to best achieve this umpire role.
The single most critical attribute of an umpire is that this individual not be a member of either team that is participating in the contest. Therefore, in our elections system, the umpire must be neither Democrat nor Republican - or at least equally acceptable to both sides. One potential way to achieve this neutrality is to adopt Rick Hasen's suggestion that a state's chief elections officer, before appointment by a governor, receive approval from three-fourths of the state's legislature. This suggestion assumes that state legislatures are not so dominated by one party that it controls three-fourths of the seats. Reasonable as that assumption may be, a more direct way to assure the equal assent of both Democrats and Republicans to the identity of this umpire would be to explicitly require the consent of the two parties' leaders. For example, the chief elections officer could be appointed upon submission to the governor of a name jointly approved by the leaders of both the majority and minority parties in the state legislature.
In any event, it seems particularly important that this office itself not be an elected position. Even if the election were nominally nonpartisan, candidates for this statewide position would seek and receive support from one or the other of the two major political parties during the campaign, and thus their ability to be an impartial umpire would be compromised. Being unelected does not mean this umpire would be unaccountable: for example, its term of office could be structured so that, after each two-year election cycle, the umpire would need reapproval from the leadership of the majority and minority parties in the legislature in order to be appointed to another term.
As an alternative to moving the functions of a state's chief elections officer from an elected Secretary of State to an appointed position of the kind just described, it has been suggested that certain limits be imposed on the partisan activities of elected Secretaries of States. While some reduction of partisanship is certainly better than none, the idea of impartial umpire remains compromised if the state's chief elections officer retains any partisan affiliation. In baseball would it suffice if the umpire could be on the payroll of one of the two competing teams as long as the umpire did not serve on the team's coaching staff? The only reason to promote such modest reform measures is that they are the only ones feasible in the short run.
But another significant lesson learned at the conference is the importance of achieving what's feasible. Perfection is often the enemy of the good, it was said several times. Or "politics is the art of the possible," to quote the old cliché. As scholars continue to explore ways to reduce inappropriate partisanship in either redistricting or election administration, increased attention needs to be given to the feasibility of scholarly proposals.