Case Study 2: The Voter File Wars --- Who Owns Democracy's Data?

Case Study 2: The Voter File Wars --- Who Owns Democracy's Data?

The Dispute

In 2015, a small civic technology startup in a Western state attempted to purchase the state's voter file to build a nonpartisan voter information tool. The startup's founders---a former software engineer and a former community organizer---envisioned an app that would let citizens look up their registration, see their voting history, explore their district's demographics, and find information about upcoming candidates and ballot measures.

The state's secretary of state's office quoted them a price of $12,500 for the file, payable annually. The file could be used only for "political purposes" as defined by state law, which the secretary's office interpreted to include campaign activities and nonpartisan voter registration drives but not commercial applications. The startup would need to sign an agreement prohibiting them from reselling the data, publishing individual voter records, or using the file for non-political purposes.

The startup's founders were taken aback. The voter file, they argued, was a public record, created and maintained with taxpayer money. Why should accessing it require paying thousands of dollars and signing restrictive agreements? And who decided that building civic technology tools was an acceptable "political purpose" while other potential uses were not?

Their frustration was shared by researchers, journalists, and civic organizations across the country who had encountered similar barriers. The rules governing voter file access varied dramatically from state to state, creating a patchwork of policies that shaped who could participate in data-driven politics and who could not.

The Patchwork

A survey of voter file access policies across all fifty states reveals extraordinary variation:

Cost. Some states provide their voter files for free or for a nominal fee (under $100). Others charge thousands of dollars. A few charge on a per-record basis, making the full file prohibitively expensive for anyone without a significant budget. In one notable case, a Southern state charged more than $25,000 for its complete voter file---a price that effectively limited access to well-funded campaigns and national data vendors.

Eligibility. Most states make the voter file available to political parties, candidates, and elected officials as a matter of course. But access for other users---researchers, journalists, civic organizations, commercial entities, individual citizens---varies widely. Some states allow anyone to purchase the file; others restrict access to registered voters, political organizations, or entities that demonstrate a "legitimate political purpose."

Use restrictions. Many states impose conditions on how the voter file can be used. Common restrictions include prohibitions on commercial use, resale, and the publication of individual voter records. Some states require users to sign agreements pledging to comply with these restrictions; others rely on statutory provisions with enforcement mechanisms.

Format and quality. Even when the voter file is available, its format and quality vary. Some states provide clean, well-documented files in standard formats (CSV, fixed-width text). Others provide files in proprietary formats, with inconsistent field names, missing documentation, and coding schemes that vary from county to county within the same state.

Content. The information included in the voter file differs by state. Some states include race, ethnicity, and telephone numbers; others do not. Some include full voting history going back decades; others include only recent elections. Some include information on absentee ballot requests and early voting; others do not.

The Stakeholders

The voter file access debate involves multiple stakeholders with competing interests:

Campaigns and political parties want broad, affordable access to voter files because the data is essential for voter contact, targeting, and mobilization. Both major parties maintain national data operations that depend on regular access to state voter files. They generally oppose restrictions that would limit their access but are less concerned about restrictions that affect other users.

Data vendors (L2, TargetSmart, Aristotle) want access to voter files because they are the raw material for their enriched data products. These companies purchase voter files from every state, standardize and merge them, add consumer data, and sell the resulting products to campaigns and organizations. They are among the most frequent and highest-volume purchasers of voter file data.

Researchers and academics want access to voter files for studies of voter behavior, turnout, registration patterns, and electoral administration. Academic research on these topics is directly limited by the cost and availability of voter file data. Some states provide free or discounted access to researchers; others treat them identically to commercial users.

Journalists want access for investigations into voter registration problems, election administration, and voter suppression. Reporters have used voter file data to identify irregularities in voter rolls, investigate allegations of voter fraud, and assess the impact of voter ID laws and registration restrictions.

Civic technology organizations like ODA want access to build tools that serve the public. Their goals are typically nonpartisan and non-commercial, but they do not always fit neatly into the categories that state laws define as eligible users.

Privacy advocates worry about the potential for misuse of voter data. The voter file contains personally identifiable information---names, addresses, dates of birth---that could be used for stalking, identity theft, or harassment. Some advocates argue that voter files should be more restricted, not less, to protect individual privacy.

State election officials are responsible for maintaining voter files and administering access policies. They face competing pressures: legislative mandates to provide public access, administrative burdens of processing requests, concerns about data security, and political pressures from parties and campaigns that want favorable terms.

The Deeper Question

Beneath the practical disputes about cost, format, and eligibility lies a deeper question: who owns democracy's data?

One view holds that voter registration data is a public record, created by government agencies using public funds, and should be freely and universally accessible. This view emphasizes the democratic value of transparency and the importance of enabling broad participation in data-driven civic life. Adaeze Nwosu and ODA largely align with this position.

A contrasting view holds that voter data contains sensitive personal information that should be carefully controlled. This view emphasizes the privacy interests of individual voters and the potential for data misuse. Advocates of this position point to cases where voter file data has been used for harassment (targeting voters based on their party registration), commercial spam (using voter contact information for marketing), and data breaches (insecure storage of voter files by third parties).

A third view, often held by election administrators, emphasizes the practical costs of data access: the staff time required to process requests, the technical infrastructure needed to distribute data securely, and the administrative burden of monitoring compliance with use restrictions. From this perspective, fees are not barriers to access but reasonable cost-recovery mechanisms.

Implications for the Garza-Whitfield Race

The voter file access landscape has concrete implications for the Garza-Whitfield race. Both campaigns have access to enriched voter files through their party-aligned data vendors, giving them detailed information on every registered voter in the state. This information enables the sophisticated targeting operations that Nadia Osei and Jake Rourke use to direct their campaigns' resources.

But OpenDemocracy Analytics, which wants to build public-facing tools using voter file data, faces a different landscape. The state charges $8,000 for its voter file, a significant expense for a small nonprofit. The use agreement prohibits commercial applications, which is not a problem for ODA, but it also requires annual renewal and imposes data security requirements that strain ODA's limited technical infrastructure.

Sam Harding, ODA's lead data journalist, has argued publicly that voter file data should be freely available to all civic organizations. "If campaigns can access this data to target voters with persuasion messages," they write, "then civic organizations should be able to access the same data to empower voters with information."

The counterargument, from the state's election director, is pragmatic: "We process more than 200 data requests per year. We cannot do that for free. The fees cover our costs, and the use agreements protect voters' privacy."

This tension---between the democratic ideal of open data and the practical realities of cost, privacy, and administration---is not unique to voter files. It runs through the entire political data ecosystem, and it will accompany you throughout this book.

Discussion Questions

Should voter registration data be freely available to anyone, or should access be restricted? If restricted, who should have access and under what conditions? Develop a specific policy proposal and defend it.
The cost of voter file access ranges from free to over $25,000 across states. What are the consequences of this variation for civic organizations, academic researchers, and small campaigns? Should there be a national standard for voter file access?
Privacy advocates worry that open voter file access could enable harassment and data misuse. How would you balance the democratic value of data transparency against the privacy interests of individual voters? Are there technical solutions (anonymization, differential privacy) that could help?
Data vendors like L2 and TargetSmart purchase public voter files, enrich them with commercial data, and sell the resulting products for profit. Is this an appropriate use of public data? Should data vendors be required to share their enriched products with the public, since the foundation is public data?
Sam Harding argues that civic organizations should have the same data access as campaigns. The election director argues that fees are necessary for cost recovery. Who is right? Is there a middle ground that serves both interests?