Tuesday Apr 07

The Hidden Numbers Behind Voter Suppression: Purging and Caging


Voter Suppression: Fair or Foul?

Voter suppression and voter turnout are as old as elections. All sides try to get all of their “yes” voters out without activating the opponents’ voters. Whether it’s a whisper in the back of the room in a student council election, a story in a workplace union certification election that not voting is a vote for the union, or super sophisticated micro-targeting and social media messages, some part of the spectrum of voter turnout versus voter suppression is a fundamental practice of politics.

We should never make the mistake of confusing politics with democracy. Often, they are not in competition, but in direct conflict. The self-interest of political parties place achieving power over maintaining democracy. Practicing self-delusion in their own councils and direct manipulation and deceit in dealing with the voters, they rationalize their politics in the sheep’s clothing of democratic principles. Doing so when in power and with their hands on the levers of government takes what’s fair in politics into the land of evil voter suppression that subverts any semblance of a fair democratic process.

The relatively benign claims of state election stewards handling the registries of voters about the need to “clean” the lists of deceased voters and those that have moved out of state fall apart when beset by either skullduggery or incompetence. We can all agree that dead people should not vote. We can all agree that voters should not be eligible to vote in two states simultaneously. Then what happens?

The devil is in the details, or from the experience of the Voter Purge Project, it is more accurate to say the devil is in the data.

Take Ohio for Example

This should have been a story with a happy ending and in some ways, it is, but everything it teaches is frightening for democracy.

Ohio’s Republican Secretary of State, Frank LaRose, perhaps as an expression of civic duty, but more likely, thumbing his nose at frequent Ohio electoral critics and of course the Democrats, did something very unusual. When he got ready to purge the voting list, rather than doing so under the dark cloak of bureaucratic confusion that accompanies most data dumps, he posted the whole list of voters he planned to purge, all 230,000 names, publicly on his website. We should wish that every Secretary of State was as transparent, through principle or pique.

There was a big problem though. 40,000 of the voters set to be purged were valid voters! We could establish that because one of us (Tingley-Hock) had built a database that had been tracking the Ohio voter file for the last five years. The bottom line: these 40,000 voters were saved. Most will never know that they might have gone to the polls in 2020 and been turned away, wrongly.

If they did know, they have Tingely-Hock to thank. A thirty-five-year veteran of database management, trained worldwide in the corporate hotbox where speed – and accuracy – of the data engineering mean the difference between making versus losing millions, Tingley-Hock was a story worthy of Tocqueville’s praise of voluntarism. Offended by what he read about voter purges, he had made it a personal mission to obtain the lists in Ohio and some other states, devise a database, and dedicate weekends for years to watch over the intricate details of millions of records in order to be able to step in the breach in just such a situation and to do it quickly enough and with enough authority to carry the day and save the votes of his fellow Ohioans.

Between the lines of that story are other stories that are more fraught and perilous: how did this 40,000 person “error” occur? What really happened? The biggest problem, involving 25,000 on the purging block, was found in the discrepancy between the state totals and the totals of two of the largest Ohio counties, Cuyahoga County, think Cleveland, and Franklin County, where the state capital of Columbus is located. The numbers and names didn’t match on the query to the database. The local county election boards showed these 25,000 to be valid voters, active and correctly addressed, while the Secretary of State’s data base had them missing in action, despite the fact that the origin of their data is in fact the counties themselves. Without argument, the state conceded these 25,000, and they were back in good stead.

Another 7800 that had been slated for purging were still recorded as “active” in the state’s voter file. The last 7200 were marked as “INACTIVE” even though the voter had voted in at least one of the general fall elections in 2014, 2016 and 2018 technically making them “ACTIVE” no matter what status had been placed on them.

Let’s give everyone involved the benefit of the doubt and assume good faith, no matter how difficult that climb might be. There was no hacking involved. No Russian conspiracy. These were not complex one-off errors, but commonplace mistakes. Not reconciling the state database. The state not affirming valid voters in their own database. These are simple mistakes with profound consequences.

Donald Trump beat Hilary Clinton badly in Ohio in 2016 in the presidential election. The margin was over 8% and the vote total was almost 450,000 votes. Disenfranchising 40,000 voters in that election and likely the coming election would not change the outcome. At the same time, as citizens concerned with the democratic process, we have to note Clinton won only eight counties in that shellacking, and that the 25,000 voters being purged were in the only two counties, Franklin and Cuyahoga, where her margin was over 60%.

Coincidence? Perhaps? But, might that explain why the Republican Secretary of State and his team wouldn’t bother to check whether they were dropping 25,000 voters accurately or not? At least it might explain their indifference, and the “ok, we give in, you got us,” response when shown the error.

As troubling is the question, what if Tingley-Hock had not been on his private mission to monitor the voter file in Ohio with his superhero, supersonic speed database algorithms? Would anyone anywhere have been able to make these lost voters found?

A National Issue Not a Local One

The issue of voter purges, deleting voters from the rolls, or caging, targeting voter for deletion through mail to their addresses and interpreting no response as failure to notify the election board of a change of residence, is not a new problem. Every election cycle for the last fifty years it seems to arise in some locality or state or another.

In Louisiana, mailings were ruled discriminatory when they only targeted African-American area in order to suppress minority voters. The Brennan Center, using federal election assistance commission data, has found that between 2016 and 2018, counties with a history of voter discrimination have continued purging people from the rolls at much higher rates than other counties. Furthermore, the Center found that “…the median purge rate over the 2016–2018 period in jurisdictions previously subject to preclearance was 40 percent higher than the purge rate in jurisdictions that were not covered by Section 5 of the Voting Rights Act. If purge rates in the counties that were covered by Section 5 were the same as the rates in non-Section 5 counties, as many as 1.1 million fewer individuals would have been removed from voter rolls between 2016 and 2018.” [https://www.brennancenter.org/our-work/analysis-opinion/voter-purge-rates-remain-highanalysis-finds]

This isn’t an isolated situation obviously. According to the Pew Charitable Trust,

• In Oklahoma this April, state officials deleted the names of more than 88,000 voters who hadn’t voted in two elections and failed to respond to a postcard confirming their addresses. Prior to that, Oklahoma had removed 275,000 inactive voters since 2015.

• And in California, state officials this year began the process of removing an estimated 5 million inactive registrations from its systems, complying with a settlement with the conservative group Judicial Watch.

• In Kentucky, the purging process has led to a lawsuit and years of political drama. After the State Board of Elections placed 175,000 people on an inactive voter list, the Kentucky Democratic Party sued … arguing the move could infringe on voting rights.

• In the run-up to the 2018 midterm elections, for example, voting rights activists criticized Georgia for purging 1.4 million voters over six years, along with other policies activists said targeted voters of color.

The Georgia story has become the poster-child of purging abuse. The person responsible for the purging was the Secretary of State. 500,000 were purged in the year before the election in which he stood for governor against a popular African-American state legislator, Stacey Abrams. His narrow victory over Abrams is believed by many to have been the direct result of the purge process and voter suppression. Worse, Georgia has announced plans to purge an additional 300,000 in coming months before the 2020 general election.

According to the U.S. Constitution, a citizen who is a registered voter cannot be disenfranchised for not voting. The U.S. Supreme Court in a 2018 decision triggered by Ohio’s practices of purging for failure to vote in the last federal election ruled that failure to vote when coupled with lack of response to a direct mailing, thereby leaving the voter’s address questionable, could validate purging. This aggressive one-chance-and-you-lose system has now been adopted by Georgia and is pending in other states like Arizona. The consolation prize is that voters purged in recent years in Ohio have a process of reinstatement until 2022. In Arizona, the state also allows some temporary relief. In most states, voters are unlikely to even know what hit them until they show up to vote.

The Barriers to Transparency

A footnote to the Ohio story and Secretary of State LaRose speaks to the value of transparency. Because he put the list forward ahead of time, Tingley-Hock and any others watching had the opportunity to see what might be right and wrong with it. Even after everything has been said and done, LaRose has been quoted publicly as committed to continued transparency, and that’s very good news indeed.

LaRose’s belief in transparency also makes him very rare among either Republicans or Democratic Secretaries of State. Ohio along with some other states provides access to the entire voter data file free of charge. Many states see denying access to the list as something along the lines of a voter suppression fundraiser.

Here are some of the worst offenders:

• Alabama: $33,000
• Arizona: $32,500
• Wisconsin: $12,500
• New Hampshire: $8,500
• West Virginia: $6,000
• Indiana: $5,000
• Louisiana: $5,000
• Montana: $5,000
• North Dakota: $5,000
• Virginia: $5,000

In order to pursue the successful database monitoring strategy advocated by the Voter Purge Project ideally the voter file would be collected every time it is undated either weekly or monthly. Calculating all of the states (except California where the rates are set by each individual county), to acquire the lists from every state on a once per month basis would cost about $1.7 million per year. The above list of the top ten most expensive all would run between $60,000 and $396,000, putting them out of reach for virtually any nonprofit and certainly for any citizen this side of a billionaire! That’s really the case for any of the states where the cost would be over $5000 per year to acquire the list on a regular monthly basis for analysis. That list would include: Maine ($2200/mo.), Rhode Island ($700/mo.), Pennsylvania ($500/mo.), Tennessee ($2500/mo.), Mississippi ($2100/mo.), Illinois ($500/mo.), South Dakota ($2500/mo.), Nebraska ($500/mo.), Oregon ($500/mo.), Hawaii ($500/mo.), and Utah ($1000/mo.).

Remember, all of this is computerized now. Many states, including very large states like New York, Ohio, and Florida offer the lists completely free. The back-of-the-envelope calculations made by the Voter Purge Project is that the maximum out-of-pocket cost for a state to deliver even a CD with the voter file, including labor, would be $28 per month. The rest has to do with either an effort to avoid transparency, an effort at suppression of voters, or even under the most charitable construction an antidemocratic misplaced revenue scheme.

What the Lists Reveal, and What Needs to Be Done

There are many questions about the stewardship of voter lists by the secretaries of state that need to be answered. Many of those answers lie in a regular and thorough analysis of voter files week to week and month to month, depending on what is available. ACORN International and the Labor Neighbor Research & Training Center partnered with the American Voter Project to form the Voter Purge Project believing that a systematic monitoring of voter files state by state coupled with an aggressive campaign might be able to create a bulwark against voter suppression.

The story from Ohio should be a sufficient argument for a comprehensive, national program when multiplied by fifty states and the District of Columbia. The Voter Purge Project believes if the same effort applied in Ohio had been undertaken in Georgia over the last several years, then Stacey Abrams would be governor of that state now.

As we look at the data from more and more states (currently we are processing seven states including GA, OH, NC, NY, OK, FL, WA and harvesting an additional thirteen so that we can process them as we build greater capacity), we can see troubling patterns and directions that our work needs to follow. These include the following:

• Why, if death is a major factor in purges, are there more purges between 17 and 35, than over 65? If it were a matter of not voting, how can 17-year-old registrants who have never voted, already be purged on a regular basis.

• In states formerly under pre-clearance by the Justice Department under the Voting Rights Act, information is available by race and ethnicity. Are there patterns reflected in that data that indicate that purges are more concentrated based on race and ethnicity? Even in the states that do not provide such data by merging with Census data one could potential evaluate whether there are discriminatory impacts and designs in purging.

• What are “drops” in the voter file as distinguished from questions of mortality and residence? What criteria are used to make these determinations? How can we make this number accountable and transparent?

• Evaluating the databases, new registrants are regularly less than voters dropped or put in confirmation or other categories that could lead to purges? Are voter registration efforts achieving the results reported once they go through the flesh-eating machine of data handling in the local and state election management?

• By evaluating the reports available nationally (or at least for all of the states we have processed or harvested) are there patterns in the purges that demand reform?

• Given the disparate costs for regularly monitoring the accuracy of voter files that now range from zero dollars per year to almost $400,000 per year (Alabama and almost Arizona), what do we need to do to force reforms in this area?

• Why aren’t all Secretaries of State pre-filing their potential purge list in the same way potential property tax and unclaimed property filings are published so that voters can act on any problems with their status ahead of the election?

• Given the access to phone, email, and mail addresses in many of these voter files, why are Secretaries of State not making an affirmative, proactive effort to robodial or email, both of which involve minimal cost, to voters to confirm their status, verify their correct addresses, or re-register?

We Could Go On

There’s a lot of work that needs to be done to prevent voter suppression in the name of list cleaning. We believe we have some of the keys to making that happen through the Voter Purge Project. These are the times that we need to protect the ballot and the ability of citizens to access the ballot and apply their voice to their vote. Without more sunlight on the purging process and data accountability, along the lines of what we are arguing, hundreds of thousands, if not millions, of votes could be lost when they are most needed in 2020.

Steve Tingley-Hock is a database engineer based near Columbus, Ohio who has been monitoring these issues with the American Voter Project.

Wade Rathke is with ACORN International and the Labor Neighbor Research & Training Center, both based in New Orleans. They have partnered to form the Voter Purge Project to campaign for voters’ rights to access the voting booths in 2020.

Subscriber Login

Latest Issue

Read The New Issue


Joomla! Debug Console


Profile Information

Memory Usage

Database Queries