Skip to main content

Estimated Costs of Covid-19 Election Resiliency Measures

Proper planning can ensure that the pandemic does not prevent a free and fair election. To be effective, funding is urgently needed.

https://www.brennancenter.org/our-work/research-reports/estimated-costs-covid-19-election-resiliency-measures

There is no question that the Covid-19 pandemic presents a difficult and, in many ways, unprecedented challenge to America’s elections. The Brennan Center has offered a detailed plan to ensure that the pandemic does not prevent a free and fair election. Implementing that plan must begin now. Below, we provide a preliminary cost estimate to implement all aspects of our plan, which could cost up to $2 billion nationwide. 1 Of course, the Brennan Center plan is not an exhaustive list, and states will have additional needs to ensure all of their citizens can vote with confidence during this pandemic.

Ensuring vote-by-mail option is available to all voters

Total estimated cost: $982 million–$1.4 billion

The following costs should be considered when increasing the option of mail voting to all voters across the country:

  • Ballot printing. Increasing the number of voters using vote by mail will require printing a larger number of ballots, absentee envelopes, and other materials. Jurisdictions should print enough ballots and ballot envelopes for 120 percent of registered voters to ensure sufficient ballots for all voters even if there are surges in voter registration close to the election and voters who change their minds and decide to vote in person instead of casting their ballot by mail. Estimated cost: $54 million–$89 million
    • Based on cost estimates provided by three ballot printing vendors, we estimate that the cost to print a ballot ranges from 21.4 cents per ballot to 35 cents per ballot. We multiplied these costs by 254 million registered voters, 120 percent of the registered voters in the United States, to obtain our estimate.
  • Postage costs. The costs of both sending and receiving ballots should be covered by the U.S. Postal Service (USPS). Estimated cost: $413 million–$593 million
    • We estimate the cost of mailing voters their ballots (including additional materials, such as return envelopes, instructions, and other informational materials) will cost $1.15–$2.00 per registered voter, or $243,455,000–$423,400,000 in total. This estimate is derived from interviews with election officials and ballot printing vendors (estimates varied widely, from $0.65 in Virginia to over $2.00 in California). In addition, voters will need to return their ballots. The cost per ballot will be less because additional materials will not be included in the return. Using an average of 80 cents per ballot for voters to return ballots, we estimate an additional $170 million to provide voters with prepaid postage for voters to return their ballots.
  • Drop boxes for absentee ballots and appropriate security. Jurisdictions should offer secure drop boxes in accessible locations for voters to drop off ballots directly. Drop boxes must be equipped with adequate security measures, such as cameras. Estimated cost: $82 million–$117 million for purchase and installation (excluding current infrastructure in vote-by-mail states) and $35 million–$47 million for operation and maintenance (excluding current infrastructure)
    • We know that at least four states — California, Colorado, Oregon, and Washington — already have drop boxes in place statewide. Washington State requires at least one ballot box per 15,000 registered voters. In Pierce County, Washington, ballot boxes provided by the company Laserfab cost between $7,000 and $10,000 to purchase and install. Snohomish County, Washington, which uses the same ballot boxes, estimates an annual ongoing operating and maintenance cost of approximately $3,000 per ballot box in a typical nonpresidential election year and $4,000 per ballot box in a presidential election year. Accounting for the four states that already have ballot boxes in place statewide, we estimate that 11,666 ballot boxes would be needed nationwide (~175 million registered voters/15,000 registered voters). To arrive at our cost estimate, we multiplied these various ballot box costs by 11,666 ballot boxes.
  • Secure electronic absentee ballot request technology. Voters must be allowed to request absentee ballots in person or through the mail, and states should offer additional methods to request ballots online or by phone. These costs must also include an increased use of online ballot delivery for uniformed and overseas citizens absentee (UOCAVA) voters. Estimated cost: $16.7 million (excluding current infrastructure)
    • Costs of obtaining or developing a secure electronic absentee ballot application tool vary widely, but we estimate an average of $325,000 per state, if the state currently has online voter registration (39 states and DC have OVR). For the purpose of estimating an online absentee ballot application tool cost, we assume that all states have OVR, since we account for the cost of implementing OVR in a different section of this document. We know that at least two states, Virginia and Pennsylvania, already have this tool and that in three states, Colorado, Oregon and Washington, voters do not need to apply to receive an absentee ballot. Therefore, we multiplied $325,000 by 46 (45 states and DC) to obtain a total cost estimate of $7 million to implement secure online absentee ballot tools nationwide.
    • We estimate a cost of $100,000 per state per year to provide a secure, online blank ballot delivery service, which allows voters to mark their absentee ballot on a computer before printing it. This assures accessibility for voters with disabilities. We estimate that at least 25 percent of states already offer a service like this. We multiplied $100,000 by 37 states to obtain a cost estimate of $3,700,000 for this service.
    • We estimate the total cost for secure electronic absentee ballot request technology/tool + annual cost for electronic vote-by-mail technology to be $2,300,000 + $3,700,000, or $6 million total.
  • Ballot tracking. Ballot tracking software should be used to provide confidence that ballots are reaching the appropriate destination in a timely manner. Jurisdictions should also set up a texting service for ballot tracking information, which will provide voters with reminders, confirmations of receipt, and confirmations of acceptance. Estimated cost: $4.2 million (excluding current infrastructure)
    • We estimate that at least 25 percent of states already have basic ballot tracking software. We estimate that this software will cost $50,000 per state. (38 states x $50,000 = $1,900,000). We are providing a separate estimate for the text delivery service, which only a handful of states currently utilize: $50,000 per state. This estimate includes setting up the platform plus costs of messages. (45 states x $50,000 = $2,250,000)
  • Improvements to absentee ballot processing. To manage the increase in absentee ballots, some jurisdictions will need to purchase resources that include signature verification technology, high-volume mail processing and sorting equipment, and high-speed ballot scanners. Estimated cost: $120 million–$240 million
    • Approximately 15 percent of local jurisdictions in the country have more than 25,000 voters (15 percent of 8,000 jurisdictions is 1,200 jurisdictions). High-speed scanners for tabulating absentee ballots cost in the range of $50,000 to $100,000 per unit. This gives a range of $60,000,000 to $120,000,000 for high-speed tabulators nationwide. The cost for high-speed automated mail sorting equipment is assumed to be in a similar range and also would only be needed in jurisdictions with more than 25,000 voters. This gives a range of $60,000,000 to $120,000,000 for high speed mail processing equipment nationwide.
  • Additional facilities. Jurisdictions will require substantially more space for ballot processing and storage.Estimated cost: $92 million
    • A surge in absentee ballots will require jurisdictions to set up an additional location for ballot processing. Most local election offices are not large enough to handle these needs and will likely need to obtain commercial space. For this estimate, we assume lease of a commercial space for 60 days to cover pre- and postelection processing work. For 85 percent of locals that have fewer than 25,000 voters (6,800 locals), we estimate rental costs of $5,000 per month for a total of $10,000. For the 15 percent of jurisdictions that are larger (1,200 locals), we estimate $10,000 per month for a total of $20,000. This gives us an estimated cost of $92,000,000.
  • Additional staffing to support absentee ballot processing. Staff will be needed for processing ballots and duplicating ballots onto the stock required for tabulation. Estimated cost: $164.6 million 
    • Assumptions include that additional seasonal staff will be needed to process absentee ballots before, during, and after Election Day for a total of 14 days. Hourly rate is assumed to be at least $15 per hour for eight hours of work per day. This would be $1,680 per additional worker. For jurisdictions under 25,000 voters, we assume 10 additional staff for a total of 68,000 seasonal workers. For jurisdictions larger than 25,000 voters, we assume 25 additional staff for a total of 30,000 seasonal workers. This would require $164,640,000 in additional staffing support nationwide.

Maintaining in-person voting

Total estimated cost: $271.4 million

Providing everyone with the option to vote by mail will not replace all in person voting by November. The handful of states that have all-mail elections took many years to get there. As we saw in the Iowa caucus, putting too much strain on an entirely new system is sure to result in breakdowns and failures. Furthermore, there are millions of Americans who will not be able to cast a private and independent vote by mail: people without Internet and mail access, those who need language assistance to vote, and people with disabilities who rely on voting machines to cast their ballots among them. There is evidence that the absence of in-person voting options could disproportionately and negatively impact Black, Latino, and young voters. We must maintain the safety-valve of in person voting, but in a way that reduces density and ensures health. To do so, the following costs must be incurred:

  • Polling facilities that meet public health standards. Poll workers will need additional resources to clean and sanitize all facilities, machines, and resources. Polling places that use hand-marked paper ballots may wish to give voters single-use pens. Jurisdictions may also incur costs due to the need to change polling locations close to Election Day if public health requires, or to acquire access to backup polling locations. Estimated cost: $29.2 million (funding for all states, even though some states may already be paying for some of this cost)
    • Cleaning supplies would cost an estimated $20 per precinct. A sample of three states with no-excuse absentee voting (Illinois, North Carolina, and Ohio) had an average of one precinct for every 1,454 registered voters. Cleaning supplies would therefore cost $0.013 per registered voter. Providing a single-use ballot-marking pen to every voter would cost about $0.50 per registered voter, if every registered voter voted in person. This will be a much lower cost if vote by mail increases. Estimate is based off of pens for 25 percent of registered voters. While this still may be high considering the number of voters using absentee ballots and voting machines, the estimate will help to cover additional facility costs.
  • Increased poll worker support. Jurisdictions must hire poll workers beyond the normal amount to overcome day-of absences. Poll worker pay may need to increase to provide an incentive for serving in-person voting. Estimated cost: $140 million (funding for pay raises for current level of poll workers in each state, and full payment for additional poll workers in each state)
    • A sample of three states with no excuse absentee voting (Illinois, North Carolina, and Ohio) had an average of one poll worker for every 208 registered voters, or about 1 million poll workers nationally. Increasing poll worker hiring by 20 percent as well as providing a raise, bringing pay from about $100 to $200 a day, would cost $100 million in raises for current levels of staffing and $40 million for the additional 20 million workers. 
  • Professional interpreters. Jurisdictions will need to offer language assistance by phone in case bilingual poll workers are absent or unavailable. Estimated cost: $43 million (funding for interpretive services for all counties covered under Section 203)
    • This estimate would cover interpreter services at a cost of $700 per day for each precinct located in a county covered under Section 203 of the Voting Rights Act. Notably, this estimate only covers interpreter services on Election Day, not during early voting periods.
  • Increased provisional materials. Jurisdictions should prepare for a surge in provisional voting due to delays in the processing of voter registration applications. Estimated cost: $21 million (funding for all provisional envelope printing, even though states and locals are already covering some of this cost)
    • Supplying enough provisional envelopes for 25 percent of registered voters at a cost of $0.40 per envelope would cost $21 million nationally.
  • Voter wait time tools. States and counties that use vote centers for in-person voting should develop online voter wait time tools to reduce lines and crowding. Estimated cost: $1.2 million (funding for all states that allow vote centers)
    • A mobile app that tracks wait times for one Texas county took 50 hours to develop in 2014. Our total estimate assumes average rates of mobile app development at $16 per hour and assumes that the time of development increases with the size of the jurisdiction.
  • Expanded early voting. Jurisdictions should expand early voting options to reduce lines and administrative stress on Election Day. This will increase all of the costs of in-person voting considered above. Estimated cost: $37 million (funding for states that don’t already have early in-person voting)
    • In 2010, Maryland counties spent $2.6 million to conduct early voting for a one-week period prior to the election, according to a legislative fiscal analysis. This represented $0.74 per registered voter. Adjusted for inflation, this would be $3.1 million in 2020, or $0.89 per registered voter. For a two-week period of early voting, this would then be $1.77 per registered voter. Excluding the all-mail states, there are 20.7 million voters in states that do not have early in-person voting. Expanding early voting to these voters would therefore cost an estimated $36.6 million. More money may be needed to expand early voting periods in states that offer in-person early voting for less than two weeks.

Developing and bolstering online registration

Total estimated cost: $85.9 million

In the months and weeks before every presidential election, millions of Americans update their voter registration information or register to vote for the first time. Covid-19 could severely disrupt this process, making it difficult for Americans to submit timely registration applications elections officials to process those applications. The outbreak will certainly reduce access to government offices that provide voter registration services.

States should adopt and bolster online voter registration systems (and they should consider implementing same-day registration, the costs of which will likely not be significant). Bolstering online registration will include the following costs:

  • Implementation of online registration for states where not used already. Thirty-nine states and DC have either fully implemented online voter registration or are in the process of doing so. The other states should do so before November. Estimated cost: $3.7 million
    • A 2014 survey of states by the Pew Charitable Trusts found that 11 of 13 states that had implemented online voter registration spent an average of $240,000 in initial startup costs. Two outliers reported $0 (Kansas) and $1.8 million (California). Since one of the remaining jurisdictions to implement online voter registration is a very high population state (Texas), an increased estimate for costs in Texas of $1 million is appropriate. $3.4 million was then adjusted for inflation to $3.7 million. 

      Note: some states may not be able or willing to move to online registration systems in time for the November election. These states will need to invest in public campaigns, voter outreach, education, and mailings to ensure voter registration is fully up to date. We do not believe the cost of these measures will be significantly less than our estimates for adoption of online registration.

  • Capacity and vulnerability testing. Online voter registration systems should be tested and their capacity bolstered to ensure that they can handle surges in web traffic. Estimated cost: $82.2 million 
    • A 2017 U.S. Election Assistance Commission (EAC) survey found that 15 states have either “bottom-up” or “hybrid” voter registration databases. For these states, added testing will be required, as individual counties that maintain their own online voter registration systems will need to conduct capacity and vulnerability testing of those systems. We estimate that capacity testing will cost approximately $25,000–$60,000 per jurisdiction and vulnerability testing will cost approximately $80,000–$100,000 per jurisdiction. Six states with bottom-up systems have 421 counties total for a total of 421 county and 6 state systems. County level systems are on the high end ($100,000) for vulnerability testing but midrange ($40,000) for load testing. Nine states have hybrid systems. In Texas, 39 counties operate their own system. Using this as a predictor of the average number of individual systems, we estimate 109 county and 9 state systems across those nine states, which also are on the high end ($100,000) for vulnerability testing but midrange ($40,000) for load testing. Thirty-four states operate top-down systems (North Dakota does not have registration) and DC is added for 35, each of which is on the high end for load testing ($60,000) and vulnerability testing ($100,000), adding up to $82.2 million

Public education

Total estimated cost: $252.1 million 

If you like this article, please sign up for Snapshot, Portside's daily summary.

(One summary e-mail a day, you can change anytime, and Portside is always free.)

Fear and confusion around a pandemic create a fertile environment for fear, disinformation, and efforts to manipulate the electoral process for improper purposes and partisan gain. State officials, advocates, and citizens should take steps to reassure citizens that voting will be safe and to guard against the use of Covid-19 to suppress voters or otherwise manipulate the election. The following costs should be considered:

  • Public education campaigns. Jurisdictions must inform voters of all changes to voting rules and all options available to register and vote. This must include advertising in non-English languages. Estimated cost: $250 million 
    • Only five states have essentially moved to an all or primarily vote-by-mail system. The rest, plus DC, will need to launch public education campaigns that include mailers, television, radio, social, and other media, all in multiple languages. The 2020 Census similarly involves significant changes that the public must learn about, such as an online option and multilanguage advertising needs. For the 2020 Census, California is spending about $2.52 per person who was counted in the 2010 Census, while New York City is spending about $0.50 per person. Houston and Harris County in Texas are jointly spending $4 million dollars, or about $0.88 per person. Similar levels of spending per voting-age member of the population — about 77 percent of the total population — would result in costs of between $129 million and $643 million. Our estimate for voter education about options during the Covid-19 pandemic is on the lower end of this range, even though these levels are over and above spending undertaken by the Census Bureau and independent organizations to ensure an accurate count.
  • Strengthened voter resources. Jurisdictions must provide accessible and easily used tools for voters to look up polling locations and registration status in order to proactively counter misinformation or malicious attacks to government systems. Estimated cost: $2.1 million
    • Capacity testing on these websites should cost approximately $40,000 per state plus DC and Puerto Rico.

1 Our estimates are conservative because they do not include cost estimates for Puerto Rico. We did not include Puerto Rico in our estimates because we relied on data from the most recent Election Administration and Voting Survey, which Puerto Rico did not participate in, as it did not conduct a federal election in 2018. Congress should of course provide funding for Puerto Rico to implement Covid-19 plans.


Lawrence Norden is the director of the Election Reform Program, where he leads the Brennan Center’s work in a variety of areas, including its effort to bring balance to campaign funding and break down barriers that keep Americans from participating in politics, ensure that U.S. election infrastructure is secure and accessible to every voter, and protect elections from foreign interference. His work has been featured in media outlets across the country, including the New York Times, the Wall Street Journal, Fox News, CNN, MSNBC, and National Public Radio. He has testified before Congress and several state legislatures on numerous occasions.

In 2009, Norden served as chair of the Ohio secretary of state's bipartisan Election Summit and Conference, authoring a report to the State of Ohio on improving that state’s election laws. The report was endorsed by the bipartisan Ohio Association of Election Officials and the Columbus Dispatch, which praised the report for “following an independent path.”

Norden was the keynote speaker at the Sixth Annual Votobit International Conference on Electronic Voting (Buenos Aires, 2008) and the 2009 Electronic Voting Technology Workshop/Workshop on Trustworthy Elections (Montreal, 2009). In June 2009, he received the Usability Professional Association's Usability In Civic Life Award for his “pioneering work to improve elections.”

Norden is the lead author of the book The Machinery of Democracy: Protecting Elections in an Electronic World (Academy Chicago Press, 2006) and a contributor to the Encyclopedia of American Civil Liberties (Routledge, 2006). He is a member of the Election Assistance Commission’s Board of Advisors, where he currently serves as vice chair of the Election Security Committee. He is a graduate of the University of Chicago and NYU School of Law.

Edgardo Cortés is an advisor to the Election Security team at the Brennan Center, where he consults on the development of regulation, legislation and litigation. He has more than 15 years of experience in all facets of the electoral process including campaigns, nonpartisan voter registration, federal and state election policy, and local and state election administration.

As Virginia’s first commissioner of elections, Cortés spearheaded voter registration and election administration modernization efforts in the commonwealth. Those efforts included: establishing paperless voter registration at DMV locations; fully integrating online DMV transactions with the online voter registration system; establishing an online, paperless absentee ballot request system; implementing an easier to use voter registration form; creating an online assessment of election administration at the local level, making election data more accessible to the public.

Cortés served as the chairman of the Board for the Electronic Registration Information Center (ERIC) and chairman of the U.S. Election Assistance Commission Standards Board. He was a charter member of the Election Infrastructure Government Coordinating Council established by the U.S. Department of Homeland Security. Cortés previously served as general registrar in Fairfax County, Va., and deputy director for policy and grants director at the U.S. Election Assistance Commission. He has also directed congressional field campaigns and a national nonpartisan voter registration program, and led efforts to implement automatic restoration of voting rights for individuals with prior felony convictions in Virginia.

Cortés holds a bachelor’s degree from Cornell University and a master’s degree in political management from the George Washington University. He lives in Springfield, Va., with his wife and son.

Liz Howard serves as counsel for the Brennan Center’s Democracy Program. Her work focuses on election security. Howard regularly comments for television, radio, and print media on issues relating to election security and election administration and has testified before U.S. House Committee on Homeland Security and in a variety of state legislatures. She has also co-authored multiple Brennan Center reports and white papers:  Better Safe Than Sorry (2018),  Defending Elections: Federal Funding Needs for State Election Security  (2019), Trump-Russia Investigations: A Guide Preparing for Cyberattacks and Technical Failures: A Guide for Election Officials  (2019).

Prior to joining the Brennan Center, Howard served as deputy commissioner for the Virginia Department of Elections. During her tenure, she coordinated many election administration modernization projects, including the decertification of all paperless voting systems, implementation of the e-Motor Voter program, and adoption of online, paperless absentee ballot applications, for which the department received a 2017 Innovations in American Government Bright Ideas Award from the Ash Center for Democratic Governance and Innovation at the Harvard Kennedy School.

She previously worked as general counsel at Rock the Vote, a nonprofit organization dedicated to engaging young people in politics and as a senior associate at Sandler Reiff in Washington, DC, where she specialized in election law with a focus on voting rights, campaign finance, and postelection disputes. Howard earned her JD from the William & Mary Law School and received the Alumnus of the Year award from the William & Mary Election Law Society.

Derek Tisler is a fellow with the Brennan Center for Justice’s Democracy Program, where he advocates for policies to protect the integrity of American elections against the threat of foreign interference.

Tisler is a recent graduate of the University of Chicago Law School, where he served as editor-in-chief of the University of Chicago Legal Forum and participated in the Jenner & Block Supreme Court and Appellate Clinic. He previously interned with the Brennan Center, the Voting Rights Project at the Chicago Lawyers’ Committee for Civil Rights, and the U.S. Department of Housing and Urban Development. Prior to law school, he worked in state legislative advocacy with a focus on urban policy. Derek holds a BA in economics from Michigan State University.

Gowri Ramachandran comes to the Brennan Center from Southwestern Law School, in Los Angeles, California, where she is on leave from her position as professor of law. At Southwestern, she taught courses in constitutional law, employment discrimination, critical race theory, and the Ninth Circuit Appellate Litigation Clinic, which received the Ninth Circuit’s 2018 Distinguished Pro Bono Service Award.

She received her undergraduate degree in mathematics from Yale College and a master’s degree in statistics from Harvard University. While in law school, she served as editor-in-chief of the Yale Law Journal. After graduating from law school in 2003, Ramachandran served as law clerk to Judge Sidney R. Thomas of the U.S. Court of Appeals for the Ninth Circuit in Billings, Montana. After a fellowship at Georgetown University Law Center, she joined the Southwestern faculty in 2006.

The Brennan Center fights to make elections fair, end mass incarceration, and preserve our liberties — in Congress, the states, the courts, and the court of public opinion. Join us in building an America that is democratic, just, and free.