With Wisconsin’s Gill v. Whitford case now before the Supreme Court, gerrymandering and its effects are now back in the public consciousness. One response to state legislature’s gerrymandering of Congressional districts has been to promote a redistricting formula based on the idea of wasted votes. In new research, Micah Altman and Michael P. McDonald find that there are limitations to such a formula based approach, especially given that here is no consensus on which one is a good measure of representation. Instead, they propose that formulas are used alongside open and transparent systems that support public participation in the redistricting process.
Given the extreme partisanship that we can see in the United States Congress, it is perhaps no surprise that gerrymandering is in the news. After all, gerrymandering is both a symptom of partisan politics and a likely contributor to it.
This fall, the United States Supreme Court heard oral arguments for Gill v. Whitford, a case that is viewed as having the potential to set a new standard for regulating gerrymandering. The plaintiffs in Gill proposed a simple formula, based on ‘wasted’ votes, that they claim measures whether district lines are fair. This formula adds to the long list of dozens of formula-based approaches proposed to prevent gerrymanders over the last fifty years.
Formulas to slay the gerrymander have been proposed for half a century — but this decade introduced something new to the picture. The technological innovation of online redistricting software (and especially open-source systems) provided ordinary people unprecedented access to the tools and data to create legal districts. During the 2010 redistricting cycle, thousands of engaged members of the public used software to create maps of their communities, draw single districts, and even draw hundreds of complete redistricting plans meeting complex and exacting legal requirements. These activities provided governments with a wider range of options to explore and improved transparency to a process previously accessible only to skilled technocrats. And this large collection of crowd-created maps provides new insights into redistricting criteria, the trade-offs among them, and the potential for representation by formula.
Gerrymandering in Ohio
US states are primarily responsible for redistricting. They use boundary delimitation institutions ranging from highly politicized state legislatures and political commissions to ostensibly non-partisan citizen commissions. Ohio has been a hotbed of recent activity to adopt a less political process as its state legislative and congressional redistricting has been politically charged.
During the post-2010 census redistricting in Ohio, Republicans held a 4-1 majority on the state’s Apportionment Board, responsible for state legislative redistricting, by virtue of holding all three statewide elected offices that serve on the Board. Following the Board’s adoption of a legislative redistricting plan, Democrats alleged in court that the Board placed partisan interests over state constitutional requirements. On a separate track, Ohio’s legislature enacted one of the most partisan congressional redistricting plans in the country.
Revealing the Possibility Frontier and Legislative Preference through Crowdsourcing
Ohio reformers hosted congressional redistricting competitions in 2009 and again in 2011 that awarded prizes to the redistricting plan that scored best on a formula. We supported Ohio advocates’ reform efforts by providing open-source web-accessible redistricting software as a platform to run their competition on. These efforts culminated in a 2012 ballot initiative, this time amending the process to create a “citizen” commission that would immediately redraw districts. This attempt also failed, however, prominent elected officials pledged to support reform. In 2015 the legislature passed a bipartisan referendum that amended the state constitution’s process and criteria for state legislative redistricting. This effort was successful — requiring not an exact formula, but rough proportionality of outcomes.
We collected post-2010 congressional plans submitted to the competition and all other publicly available plans that are objectively legal Ohio congressional redistricting plans. We evaluated Ohio congressional redistricting plans using two sets of criteria, those formulated for the competition and those that scholars have used to evaluate redistricting plans. These criteria sets are closely related. They encompass the same concepts of population equality, minority representation, respect for county boundaries, district compactness, the number of competitive districts, and the plan’s partisan fairness.
Figure 1 below shows the trade-offs between partisan advantage and good government criteria. We draw lines to represent the Pareto frontier (the set of optimal tradeoffs between sets of criteria), with the addition that we draw circles around single or (apparently) disconnected points on the frontier.
Figure 1 – Democratic Seat Advantage vs. Standard Criteria
Figure 1 provides insights as to why the legislature chose the adopted plan over potential alternative plans. The adopted plan is almost always on the left-most edge of the possibility space. This suggests that the adopted plan was drawn to maximize the expected Republican seat advantage, secondary to other criteria.
Our analysis shows some limits of a formula based approach. Interestingly, the trade-off between Democratic seats and the overall score that the reformers adopted for their competition, at the bottom of Figure 1, reveals that although the score was presented as non-partisan, in practice, it appeared to favor Ohio Democrats — in that higher scoring plans tended to have a greater number of Democratic-majority districts.
Creating political boundaries requires political choices — these should be democratic
This research, and from our parallel studies of Virginia, Florida, and other states, yields generalizable findings about participative redistricting. First, with appropriate data and technology members of the public can readily create legal redistricting plans that incorporate criteria which they value. Second, public plans generally demonstrate a wider range of possibilities than plans created by the political process. Third, public plans generally do better on explicitly measured good-government criteria than political plans. And, fourth — the specific criteria one uses, and the way they are defined, can substantially affect the tradeoffs among criteria.
How a society conducts elections through selecting political boundaries, qualifying voters, casting and counting votes is critical to a well-functioning democracy. And the choice of electoral processes inevitably embed some political choices.
By DrRandomFactor (Own work) [CC BY-SA 3.0], via Wikimedia Commons
When the US was founded, the choice to elect representatives using geographic districts reflected a political decision that representatives should have ties to specific communities of constituents, and an assumption that geography reflected the most salient of political distinctions. In modern history, the redistricting process has come to reflect other political decisions — to systematically award spoils for partisan victory, and (sometimes) to protect incumbents. In the future, we hope that the redistricting process can embed values of good government process and a rough equity in the outcomes of the electoral system.
Formula alone cannot accomplish this. There is currently no consensus on what formula adequately measures representation. The understanding of how to evaluate good-government criteria such as competitiveness, and partisan proportionality changes over time and with context. Moreover, formula may be subverted — through the choice of which criteria to measure, how to quantify them, their relative weights, the algorithms used to find plans based on the formulas, and the software used to measure and generate plans.
We argue that formula work best in combination with open and transparent systems that support public participation — and independent commissions that aid in selecting, adapting, and applying criteria. Commissions can be designed to ensure independence, participation, and transparency — and provide flexibility for change over time; and to support examination of alternative data and criteria. This enables what might otherwise be static quantification of representation to be embedded in a living, democratic, transparent, and participative process.
Some reformers want to take people out of redistricting altogether and “just let the formula do it”. Our work leads us to conclude that no single formula can, on its own, reliably create fair electoral maps — public input and societal choices should not be entirely removed from the mapping process. Our research suggests that thoughtful, responsible electoral mapping by independent commissions, aided by open information technology, is very possible.
- This article is based on the paper ‘Redistricting by Formula: An Ohio Reform Experiment’ in American Politics Research.
Please read our comments policy before commenting.
Note: This article gives the views of the author, and not the position of USAPP – American Politics and Policy, nor of the London School of Economics.
Shortened URL for this post: http://bit.ly/2yeNoKn
About the authors
Micah Altman – MIT
Dr. Micah Altman is Director of Research and Head/Scientist, Program on Information Science for the MIT Libraries, at the Massachusetts Institute of Technology. Previously He served as a Non-Resident Senior Fellow at The Brookings Institution, and at Harvard University as the Associate Director of the Harvard-MIT Data Center, Archival Director of the Henry A. Murray Archive, and Senior Research Scientist in the Institute for Quantitative Social Sciences. He conducts work primarily in the fields of social science, information privacy, information science and research methods, and statistical computation — focusing on the intersections of information, technology, privacy, and politics; and on the dissemination, preservation, reliability and governance of scientific knowledge.
Michael P. McDonald – University of Florida
Dr. Michael P. McDonald is Associate Professor of Political Science at University of Florida. He received his Ph.D. in Political Science from University of California, San Diego and B.S. in Economics from California Institute of Technology. He held a one-year post-doc fellowship at Harvard University and previously taught at Vanderbilt University; University of Illinois, Springfield; and George Mason University. His research interests are in the areas of elections and methodology. His voter turnout research shows that turnout is not declining, the ineligible population is rising.