Skip Navigation
Report

America’s Voting Machines at Risk

Published: September 15, 2014

Tech­no­logy has changed dramat­ic­ally in the last decade, but Amer­ica’s voting machines are rapidly aging out. In 2016, for example, 43 states will use elec­tronic voting machines that are at least 10 years old, peril­ously close to the end of most systems’ expec­ted lifespan. Old voting equip­ment increases the risk of fail­ures and crashes — which can lead to long lines and lost votes on Elec­tion Day — and prob­lems only get worse the longer we wait.

Based on 10 months of inde­pend­ent research, and inter­views with more than 100 elec­tion offi­cials and special­ists in all 50 states, this compre­hens­ive study looks at the chal­lenges asso­ci­ated with outdated equip­ment and the new tech­no­lo­gies that can help solve the impend­ing crisis.

Read the Exec­ut­ive Summary

View on Scribd


AnchorExec­ut­ive Summary

In Janu­ary 2014, the bipar­tisan Pres­id­en­tial Commis­sion on Elec­tion Admin­is­tra­tion (PCEA) issued a stern warn­ing that should be of grave concern to all Amer­ic­ans: There is an “impend­ing crisis … from the wide­spread wear­ing out of voting machines purchased a decade ago. … Juris­dic­tions do not have the money to purchase new machines, and legal and market constraints prevent the devel­op­ment of machines they would want even if they had funds.”

This report, nearly two years later, docu­ments in detail the extent of the prob­lem and the steps we must take in the coming years to address it. Over the past 10 months, the Bren­nan Center surveyed more than 100 special­ists famil­iar with voting tech­no­logy, includ­ing voting machine vendors, inde­pend­ent tech­no­logy experts, and elec­tion offi­cials in all 50 states. In addi­tion, we reviewed scores of public docu­ments to quantify in greater detail the extent of the crisis. We explore the current chal­lenge in three parts: (1) the danger, look­ing at the age of machines around the coun­try relat­ive to their expec­ted lifespans and the prob­lems that we can expect; (2) the new tech­no­lo­gies that can help solve the prob­lem going forward; and (3) recom­men­ded solu­tions to the impend­ing crisis.

Among our key find­ings:

  • Unlike voting machines used in past eras, today’s systems were not designed to last for decades. In part this is due to the pace of tech­no­lo­gical change. No one expects a laptop to last for 10 years. And although today’s machines debuted at the begin­ning of this century, many were designed and engin­eered in the 1990s.
     
  • While it is impossible to say how long any partic­u­lar machine will last, experts agree that for those purchased since 2000, the expec­ted lifespan for the core compon­ents of elec­tronic voting machines is between 10 and 20 years, and for most systems it is prob­ably closer to 10 than 20.
    • The major­ity of machines in use today are either peril­ously close to or exceed these estim­ates. Forty-three states are using some machines that will be at least 10 years old in 2016. In most of these states, the major­ity of elec­tion districts are using machines that are at least 10 years old.
       
    • In 14 states, machines will be 15 or more years old.
       
    • Nearly every state is using some machines that are no longer manu­fac­tured and many elec­tion offi­cials struggle to find replace­ment parts.
  • The longer we delay purchas­ing new equip­ment, the more prob­lems we risk.
    • The biggest risk is increased fail­ures and crashes, which can lead to long lines and lost votes.
       
    • Older machines can also have seri­ous secur­ity and reli­ab­il­ity flaws that are unac­cept­able today. For example, Virginia recently decer­ti­fied a voting system used in 24 percent of precincts after find­ing that an external party could access the machine’s wire­less features to “record voting data or inject mali­cious data.
       
    • Smal­ler prob­lems can also shake public confid­ence. Several elec­tion offi­cials mentioned “flipped votes” on touch screen machines, where a voter touches the name of one candid­ate, but the machine registers it as a selec­tion for another.
       
  • Elec­tion offi­cials who believe they need to buy new machines do not have suffi­cient resources.
    • Elec­tion juris­dic­tions in at least 31 states want to purchase new voting machines in the next five years. Offi­cials from 22 of these states said they did not know where they would get the money to pay for them.
       
    • Based upon recent contracts and assess­ments provided by elec­tion offi­cials, the Bren­nan Center estim­ates the initial national cost of repla­cing equip­ment over the next few years could exceed $1 billion, though that could be partially offset by lower oper­at­ing costs and better contracts than are currently used in many juris­dic­tions.
       
    • As elec­tion juris­dic­tions diverge in how they respond to the crisis, we see an increas­ing divide among, and even within, states in the abil­ity to ensure elec­tions can be conduc­ted without system fail­ures and disrup­tion.
       
    • A prelim­in­ary analysis by the Bren­nan Center lends support to the concern expressed by some offi­cials that without federal or state fund­ing, wealth­ier counties will replace aging machines, while poorer counties will be forced to use them far longer than they should.

These are troub­ling find­ings, but our study also provides hope for the future. Tech­no­logy has changed dramat­ic­ally in the last decade, offer­ing the possib­il­ity of machines that are more reli­able, more usable, and less expens­ive. Several recent innov­a­tions — often driven by elec­tion offi­cials who have worked with vendors, academ­ics, and voters — could point the way to more afford­able and flex­ible 21st century machines. While such advances may help us in future years, they will not resolve today’s crisis. There is no escap­ing the imme­di­ate need to plan and set aside suffi­cient funds to buy new machines.