When the dust settled after the most recent Texas party primaries, newspaper editorial writers across the state were able to dust off and wheel out their evergreen essays about the deplorable lack of voter turnout.
As Professor Rick Hasan has noted, a partisan split does show up in arguments over boosting voter turnout. To grossly simplify the philosophical positions, conservatives tend to argue that low turnout isn’t really a problem, and might actually be a good thing (because boosting turnout might encourage the wrong element to vote for the wrong kinds of candidates and policies, thereby derailing civilization). Liberals, on the other hand, tend to argue that low turnout is a problem (because the vast majority of the public have no say in the way that they are governed, thereby derailing civilization), and that boosting turnout might finally bring about some simple decency and policy balance to our dangerously unresponsive plutocratic government institutions.
In reality, the philosophical split is probably less partisan than it appears, because both those who favor voting by a small, professional elite, and those who favor voting by a large, representative cross section of society would probably agree that independent of turnout, elections should end with the right person or policy being chosen.
Right now, I’m taking a fascinating online Coursera course (A Beginner’s Guide to Irrational Behavior) being taught by Duke University’s Dan Ariely, a leading figure in the trendy world of behavioral economics and choice architecture design. These are fields of social science inquiry that basically didn’t exist when I was an undergraduate, but that arose to challenge long-held assumptions about rationality and choice. Professor Ariely’s research interests are broad and difficult to summarize in a brief way, but at the risk of being too flip and general, I would say that the founding hypotheses of behavioral economics are:
(1) The thousands of discretionary choices we make every day are not guided by some sort of objective optimality; none of our behaviors are rational in the sense of being objectively correct. We act in reaction to a complex mixture of emotional and sensory biases, environmental cues, neurological and endocrine short-cuts, make-works, and work-arounds, and so on.
As one could say with respect to lots of complex biological systems (our cobbled-together bipedalism, our barely adequate accommodations for giving birth to big-headed kids, our badly-engineered mammalian ears, our badly-engineered vertebrae, our “work-in-progress” bodies in general), the evolution of human behavior is the triumph of the low bidding contractor – our brains are only just as good as they minimally have to be in order to keep us alive and reproducing.
(2) The same neurological limitations that make us blind to the actual causes of our behavioral choices also make us very bad at recognizing our lack of rational agency.
(3) On the contrary, we are really good at overestimating our rationality, modeling rationality into our memories and interpretations of our own actions, and selectively editing our sensory evidence of reality for the sake of preserving our illusory capacity for reasoning.
So let me posit the following with respect to polities, nation-states, and the general conduct of human government: that we cannot presume that low or high voter turnout in any election is in any way correlated with the objective quality of the election outcome. Further, that there is no reason to believe that any particular voter or group of voters will vote in such a way as to promote their own or anyone else’s self-interest, even when they think that they are doing so.
The notion of having elections is a cultural artifact, and is useful to us to the extent that we accept the premise that having elections somehow works out for us over the long run.
You might assume from all this that I am a defeatist, or that my message is that we are irrational and therefore shouldn’t try to make the right decisions. This is not my view.
Yes, our mental constructions that lead us to one outcome or another are not built on logic or facts. But there really are better and worse outcomes, and our decisions (however we may arrive at them) are critically important. Therefore, we must engineer our elections so as to mitigate or subvert our biases, with the objective of forcing us to make good decisions whether we know it or not.
We already engage in this kind of engineering (or choice architecture, in the current social science parlance) in an ad hoc and non-systematic way.
Consider the ballot drawing. We tend to vote more for the first candidate on the ballot. It’s a crazy visual bias that accounts for a good 5% of the vote in any election (more or less). So as a sort of rough justice, we randomize the order of the candidate names in non-partisan elections. Tex. Elec. Code Section 52.094. Whoever is lucky enough to come out on top will get that magic 5% of the vote.
It’s an odd approach, given that the optimal solution would be to get rid of the magic 5% before the election ever begins. But it’s cheaper and easier than trying to figure out how to reverse the way our brains are wired.
And in another (and major) way, we take advantage of the visual bias when the election is a partisan one, because we gift the dominant political party (based on the most recent gubernatorial election) with the magic 5% in the November general election by listing that party’s candidates first in each race. Tex. Elec. Code Section 52.092.
We therefore more-or-less deliberately favor the more successful party in the general election, but not in local or primary elections. The mechanism for straight-party voting reinforces this bias by offering the voter a path of least resistance – to vote for all the candidates in the first listed party without having to make any other choices on a lengthy ballot. Tex. Elec. Code Section 52.071.
Ballots are fascinating cultural artifacts, because independent of the actual votes cast on them, they reflect irrational biases and both deliberate and unconscious manipulations of those biases. For example, we list the races in a descending geographic hierarchy that reflects our irrational tendency to view”big” elections as being more important that “little” elections.
Which is odd, when you think about it, because the more local the office, the more likely that the discretionary policies of that office will affect us immediately and directly. The President is important, but when was the last time the Office of the President of the United States stuck a parking ticket on your windshield?
Here’s a radical proposal for ballot reform. Imagine a ballot that looked like this:
General and Special Election – Saturday, November 9, 2080 [following the landmark reform in the mid-21st century shifting November general elections to the second Saturday of that month, to boost turnout].
BALLOT INSTRUCTIONS: Choose the offices and propositions that you would like to vote on from the issues and positions menu. After making your selections and composing your ballot, choose the candidates you would prefer for each office, and your position on each proposition.
For each choice you make, you will be asked why you made that choice. Your reasons for choosing will become part of the public record of this election and will be subject to peer review, which will determine the weight to be given to your vote. In no case will your vote ever count as less than 1 vote for each of your choices.
Depending on the peer review of the reasons for each of your choices, your vote for a candidate or proposition may be counted from 1 to 3 times. The adjusted vote total of all votes cast by all voters following peer review will determine the outcome of each race.
Resurrecting Submissions (But Not Preclearance) Using Title VI of the Civil Rights Act – Tracking Election Procedural Changes After Shelby County
What’s Up With Shelby County v. Holder?
Most readers of this blog already know the answer to that question, but to recap — on June 25, 2013, the United States Supreme Court issued an opinion in a voting rights lawsuit involving Shelby County, Alabama, and the Attorney General of the United States. As court watchers had expected, the Supreme Court found a key section of the Voting Rights Act (Section 5) to be unconstitutional. One consequence of the decision was that a process called “preclearance” ended, meaning that as of June 25, 2013, the U.S. Department of Justice no longer reviews and approves changes in voting procedures that have been adopted in certain jurisdictions.
I think the U.S. Department of Justice (DOJ) has a strong and defensible statutory reason for tracking election procedures across all jurisdictions in the United States, independent of the specific requirement in Section 5 of the Voting Rights Act (VRA) that covered jurisdictions had to submit changes in their voting procedures for preclearance. (For those of you asking, “What the heck is Section 5?” the Voting Section of the Civil Rights Division at DOJ has an excellent short summary of this key element of the Voting Rights Act of 1965, along with historical background.)
Mandatory Notification of Changes in Election Procedures Under Title VI of the Civil Rights Act
I propose that DOJ should promulgate some new rules that would more or less recapitulate the guidelines for preclearance submissions under Section 5 of the Voting Rights Act, but in the context of compliance reporting under Title VI of the Civil Rights Act of 1964 (hereafter “the Civil Rights Act” or “the Act.” With statutes such as the Civil Rights Act of 1964, one has to specify the date, because there are a lot of laws with the name “Civil Rights Act”).
Some explanation is in order for people who may not be familiar with how the Act is organized. The Civil Rights Act of 1964 (along with later amendments) is a breathtakingly broad and powerful response to a century’s worth of endemic racial discrimination; the Act has lots of parts that accomplish different goals, all with the objective of shutting down the often slippery and duplicitous efforts of Southern segregationists to avoid meaningful enforcement against institutional racism.
Because the Act is so large, it is broken down into smaller sections (Titles) that focus on different ways that institutional racism manifests. Title I is the “voting part” of the Civil Rights Act, but it is pretty weak — unlike the Voting Rights Act of 1965, Title I of the Civil Rights Act doesn’t get rid of tests or qualifications to vote, but just requires that such tests be applied fairly. Title II of the Civil Rights Act allows the use of court-issued injunctions against private businesses that cater to the general public but refuse service on the basis of race (examples would include restaurants or hotels that might prohibit African-Americans, or that might offer inferior or different service based on race). Title III of the Civil Rights Act accomplishes the actual desegregation of public accommodations. Title IV desegregates public education, and Title V creates the Commission on Civil Rights.
Title VI (the part of the Civil Rights Act that I’m talking about) prohibits acts of discrimination by agencies or organizations that receive federal financial assistance. I’m focusing on Title VI because that part of the law includes very broad authorization of rulemaking authority by federal agencies such as the DOJ. In terms of rulemaking, Title VI of the Civil Rights Act is very different from Section 5 of the Voting Rights Act, which contains no explicit rulemaking authority.
In fact, one of the limitations of the Voting Rights Act is the incapacity of federal agencies to make use of that law to justify formal administrative enforcement. The Voting Rights Act enables access to the Federal courts in order to protect against discrimination in voting, but it doesn’t explicitly enact much in the way of non-judicial bureaucracy to enforce it.
Some of the other titles in the Civil Rights Act are famous (such as Title VII, which creates equal employment opportunities and guards against discrimination in employment), while others are less well-known (Title VIII authorizes the Commerce Department to collect statistical data on voter registration, and Title IX authorizes removal of civil rights cases from the courtrooms of segregationist Southern judges). There are other titles as well, but as I mentioned, I’m mostly interested in looking at Title VI.
Title VI of the Civil Rights Act and Part 42 of Title 28, Code of Federal Regulations
The rules that Federal agencies adopt are collected into a series of volumes entitled the “Code of Federal Regulations,” which is organized thematically into titles and sections. Title 28 of the Code of Federal Regulations contains rules promulgated by DOJ, and Part 42 of Title 28 contains DOJ rules for enforcing Title VI of the Civil Rights Act of 1964.
This is where my proposed rules would logically be placed, but this is not where you would look to find the “administrative guidance” adopted by DOJ to guide entities in making submissions under Section 5 of the Voting Rights Act. That “guidance” (because, remember, the Voting Rights Act didn’t authorize the creation of agency rules) is found in Part 51 of Title 28.
I mention Part 51 because I freely cut and pasted huge swaths of that preclearance submission guidance into my proposed rules. That’s because I want to enrich and expand the submission process, preserving the administrative apparatus of Section 5 submission as much as possible. I don’t want to reinvent the whole thing because it’s better public policy to build off the established framework.
Thus, 28 C.F.R. Part 42 (the administrative rules promulgated by the DOJ to administer the Civil Rights Act) should be amended to include a new Subpart J (labeled beginning with Sections 42.801 through 42.826) that would recapitulate the structure and contents of the submission guidelines in 28 C.F.R. Part 51, but exclude the enforcement provisions or selective categories of covered jurisdictions (that had previously been addressed in Section 5 of the Voting Rights Act).
Replacing “Submissions” with “Reports”
Notice the stylistic changes I made in the proposed rules. No mention is made of “submissions,” and there’s no mechanism for reviewing the “reports” that federally funded agencies must, under these proposed rules, submit to DOJ. That removal of review is intentional — no longer are jurisdictions (e.g., states, or any part of states, or any other jurisdiction) being singled out for preclearance of their voting changes. Instead, the proposed rules would compel jurisdictions to report voting changes to continue to qualify for federal funding.
Applying the “Report” Requirements to All Political Subdivisions
You’ll also notice something pretty major when you read the proposed rules. As drafted, the rules apply to all political subdivisions, not just in the South or in “trouble spots,” but everywhere. This drafting constitutes a dramatic expansion of reporting requirements, far beyond the old submission process, and it’s not accidental.
First, the expansion would enable DOJ and Congress to acquire the hard data they need to reconfigure and revive the coverage formula in Section 4 of the Voting Rights Act of 1965. Second, the expansion addresses the implicit or explicit complaints of conservative critics of the Voting Rights Act, which was (in a nutshell) that the Voting Rights Act of 1965, as applied, was unfairly burdening the South in comparison to other jurisdictions, without regard to whether those other jurisdictions exhibited patterns of racist behavior.
These Rules Don’t Substitute For a Robust Voting Rights Act
I applaud efforts to fix or rehabilitate the Voting Rights Act of 1965 (and there are all sorts of good proposals, from piggybacking some new “non-binding guidance” onto Section 2 of the Act, to getting Congressional reauthorization of Section 4 with beefed-up sociological data showing that the South is still a hotbed of racial discrimination to a greater degree than are other jurisdictions. See, for example, Christopher Elmendorf and Douglas Spencer, The Geography of Discrimination in Voting: MRP Meets the VRA., UC Davis Legal Studies Research Paper Series No. 339 (May 2013). at 47.)
My proposed “reporting” rules aren’t intended to diminish or reroute efforts to restore Section 4(b) VRA covered jurisdiction formulas. If anything, the proposed rules are intended to complement, but not replace, the submission process. We still need federal preemptive blocking of potentially discriminatory changes in voting procedures (e.g., in addition to voter ID and voter registration laws, the shortening of early voting windows and ending of Sunday voting, distribution of polling places based on population, physical layout of polling places, and so on).
Saving the Voting Rights Act of 1965 is a big, intimidating goal, with complicated political risks. Arguably the House of Representatives has been hijacked by extremists who have no interest in governing the country, the conservative-dominated Supreme Court has already succinctly demonstrated its bias with respect to civil rights laws, and DOJ faces having to litigate, court by court and state by state, to mitigate discriminatory election procedures.
Consider the proposed “reporting” rules as having far more modest goals. The intention here is to recreate the mechanical, or “easy,” part of the submission process while working toward other bigger and harder goals.
Why We Need Something Analogous to Section 5 of the Voting Rights Act Submissions
Section 5 of the Voting Rights Act (that is, preclearance) enabled a federal agency (the Department of Justice) to preemptively shut down state and local legislation that weakened minority voting rights. But preclearance was also a powerful statistical and data-gathering tool for policymakers and planners.
The identification of minority voting interests should be a “best practice” of legislative drafting, but such identification often isn’t done unless there is some external requirement compelling involvement and comments from protected classes of language and racial minority groups. Preclearance submissions enforced greater care in the legislative drafting process across all covered jurisdictions.
Preclearance submissions also encouraged greater awareness by lawmakers of the geographic distribution and demographics of the voting-age population in a covered jurisdiction, and not just for purposes of planning redistricting, but also for placing polling sites and scheduling early voting. Preclearance encouraged accountability to local interests and cut down on accidental errors in voting site placement, protected against the underestimation of the level of language assistance needed in a territory, and encouraged the review and remediation of architectural barriers affecting voters with disabilities.
The administrative framework of at least the reporting part of the preclearance framework has to be rebuilt. As a nation, we can either do it now (when there’s still enough institutional memory and structure to make the process slightly less painful) or we can do it later (when it will be much more expensive and painful) after the successor to Section 4 of the Voting Rights Act of 1965 (in whatever form that future law may exist) finally gets passed.
It’s also time for “non-covered jurisdictions” — that is, the jurisdictions that weren’t identified by the formula for coverage in Section 4(b) of the Voting Rights Act of 1965, and that therefore weren’t subject to the Section 5 preclearance process — to start getting used to the idea of tracking and reporting changes in voting procedures to the DOJ. The stated reasons for historical exemption of most non-Southern jurisdictions merely undercut the goals promoted by the Voting Rights Act, and contributed to the gutting of the law when the Supreme Court ruled that Section 4 of the Voting Rights Act was unconstitutional.
Cost Considerations and Administrative Burdens
It would be misleading to suggest that “reporting” detailed changes in voting procedures wouldn’t take any money or effort. Preclearance submissions were tedious, expensive, and time-consuming. When the Shelby County decision was handed down, we election attorneys were in some sense like kids who had been told that we had a snow day.
We were giddy with excitement that we wouldn’t have to go through the laborious process of tracking and submitting every change in voting procedures to the Voting Section of the Civil Rights Division of the Department of Justice. No more having to read numbing 300-page omnibus bills looking for the one section that might affect election procedures. Suddenly, rooms full of boxed manila folders were free to be used for other things. Shelby County meant no more maintenance of detailed lists of legislative materials dating back to the mid-1970s, and no more endless rounds of phone tag with legislative aides, law firms, and federal preclearance analysts. It would be fair to say we were delighted.
That was what the atmosphere felt like in the Elections Division at the Texas Secretary of State’s office, even among attorneys who were philosophically predisposed to support the Voting Rights Act. I suspect it was the mood elsewhere, given that the immediate practical effect of the verdict in Shelby County v. Holder was the dismantling of a decades-long structure of administrative guidance and office practices to support regular written submissions of detailed dossiers on changes in election procedures. Money that would otherwise have gone toward the Section 5 submission industry was saved for other uses.
I don’t doubt that Shelby County cut down on the billable hours at law firms representing political subdivisions subject to the Section 5 preclearance requirements, shifted the workload of preclearance analysts at DOJ, and cleared out file rooms across the Deep South.
For all I know, the civil servants at DOJ also treated the end of preclearance as a holiday — the administration of preclearance certainly cost money and time for DOJ as well as for the states. Analysts at DOJ would describe the Texas submissions file room (a few years ago, a DOJ analyst told me that Texas alone accounted for half of the submissions received by DOJ each year) as a cramped warren of buckling, teetering towers of cardboard boxes stacked floor to ceiling. Not surprisingly, the staff at the Voting Section of the Civil Rights Division seemed to dread the labor associated with retrieving a specific file from the unruly mess.
But just as every snow day must eventually be made up, our vacation from the mechanical tedium of tracking and reporting changes in voting procedures needs to end. It’s time to get back to work.
Despite the practical unpleasantness and cost involved, I think it is still the case that the reestablishment of some archive of procedural changes in election administration is cheaper if it’s done now, rather than later when years of legislative efforts will have to be retroactively reconstructed.
Reporting Changes in Election Procedures After June 25, 2013
As I mentioned , the Voting Rights Act of 1965 did not provide DOJ with any explicit binding rulemaking authority, and DOJ did not claim rulemaking authority in issuing various guidelines for compliance with the Voting Rights Act. In fact, the administrative procedures for Section 5 preclearance submissions were alternatives to judicial review and therefore not binding on any of the covered jurisdictions.
At least one scholar has speculated that the DOJ’s lack of formal rulemaking authority may have further undermined potential Supreme Court deference to DOJ as an executive agency, and in fact created an atmosphere of “anti-deference.” See Arpit K. Garg, A Deference Theory of Section Five (Draft as of April 1, 2012), online at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2209636; Jennifer Nou, Sub-Regulating Elections (Public Law and Legal Theory No. 462, 2014) (Chicago Unbound, University of Chicago Public Law and Legal Theory Working Paper), at p. 27.
The reason why the non-binding Section 5 submission “guideline” process was so popular (accounting for well over 99% of all Section 5 preclearances) is that it was so fast and cheap compared to judicial preclearance. With Section 5 reduced to a hollow shell, DOJ now responds to any requests for preclearance with the following boilerplate language:
On June 25, 2013, the United States Supreme Court held that the coverage formula in Section 4(b) of the Voting Rights Act, 42 U.S.C. 1973b(b), as reauthorized by the Voting Rights Act Reauthorization and Amendments Act of 2006, is unconstitutional and can no longer be used as a basis for subjecting jurisdictions to preclearance under Section 5 of the Voting Rights Act, 42 U.S.C. 1973c. Shelby County v. Holder, 570 U.S. ___, 2013 WL 3184629 (U.S. June 25, 2013) (No. 12-96). Accordingly, no determination will be made under Section 5 by the Attorney General on the specified change. Procedures for the Administration of Section 5 of the Voting Rights Act, 28 C.F.R. 51.35. We further note that this is not a determination on the merits and, therefore, should not be construed as a finding regarding whether the specified change complies with any federal voting rights law.