How confident are we that the votes we cast in this year’s presidential election will be counted accurately? This is a standard question researchers ask voters to assess the many different ways we cast and count ballots. We all don’t vote on the same day, at similar locations or on same type of voting equipment, which may be important sources of variation in the confidence voters have that their ballots are counted accurately. For example, national studies show that early voters, those who ballot in person or by mail before Election Day, are less confident than their Election Day counterparts that their ballot will be accurately counted. Mail-in ballots are subject to handling by the U.S. postal service and county election officials. Here an opportunity for ballots to be lost, damaged and even destroyed has a non-trivial probability. The ballots of in-person early voters have to be stored until they are counted on Election Day, again raising the possibility of lost or damaged ballots. In 2008 and again in 2012 researchers at Rice University asked Harris County voters if they “were confident that their ballots would be accurately counted.”
Table 1 reports respondents’ answers to this question in 2008 and 2012. What is most obvious and dramatic is the precipitous decline in voter confidence between 2008 and 2012. The decline in voter confidence, however, is not uniform across either race or partisan affiliation. Table 2 shows the decline in confidence is far greater among African-American and Hispanic voters than Anglos, though all racial and ethnic groups exhibited a significant decline in the confidence their vote will be counted accurately. The steep decline in voter confidence is observed among voters of all partisan persuasion (Table 3). However, the drop in voter confidence is significantly greater among Democrats and independents than Republicans.
Many factors account for this decline, including the rise of partisan polarization. In Texas and Harris County considerable partisan rhetoric has been spent on the state’s newly adopted voter ID law that requires voters to present a government issued photo ID in order to vote (though this is not required of mail-in voters). Supporters of the photo identification requirement claim this reform assures the integrity and accuracy of votes cast and election outcomes. Opponents claim there is little voter fraud and that a photo ID requirement suppresses participation among eligible voters who lack a driver’s license or other government photo ID, persons who Texas State officials recognized are disproportionately young, non-Anglo and mostly Democratic voters. Lawsuits challenging photo ID requirements have prevented implementation of these new voting requirements in several states, including Texas. The U.S. Supreme Court will decide these cases next spring. The recent flap over the “dead voter” purge in Harris County, where voters were dropped from the polls because their names also showed up as deceased on a national data base, may have also contributed to belief that our elections not are conducted in sound and professional manner.
The partisan rancor of the last four years, including the debate over party ID laws, may have contributed to the decline in voter confidence. Neither those who support voter ID laws (Anglos and Republicans) nor those who see these laws as either unnecessary and/or an obstacle to eligible voters casting their ballot (non-Anglos and Democrats) have been exempt from the diminution in voter confidence. The fact that those who perceive the voter ID law as objectionable have experienced a greater drop in the confidence that their vote will be accurately counted should be of some concern to legislators who adopted the voter ID law.
Table 1: How confident are you that your vote will be accurately counted? (Percent by column)
| 2008* | 2012** | |
| Very confident | 63.0 | 44.3 |
| Confident | 16.9 | 23.9 |
| Unsure | 12.6 | 4.1 |
| Somewhat confident | 3.3 | 19.8 |
| Not confident | 4.1 | 7.8 |
Table 2: How confident are you that your vote will be accurately counted? (Percent by column)
| Anglo | African-American | Hispanic | ||||
| 2008 | 2012 | 2008 | 2012 | 2008 | 2012 | |
| Very confident | 64.0 | 50.3 | 62.4 | 36.0 | 64.7 | 43.8 |
| Confident | 16.2 | 24.6 | 17.1 | 19.4 | 16.3 | 27.1 |
| Unsure | 13.0 | 4.3 | 12.4 | 3.8 | 10.9 | 3.1 |
| Somewhat confident | 3.0 | 15.7 | 4.0 | 24.8 | 2.7 | 19.8 |
| Not confident | 3.7 | 5.0 | 4.6 | 12.3 | 5.0 | 6.2 |
Table 3: How confident are you that your vote will be accurately counted? (Percent by column)
| Republican | Independent | Democrat | ||||
| 2008 | 2012 | 2008 | 2012 | 2008 | 2012 | |
| Very confident | 74.3 | 52.3 | 57.3 | 47.8 | 58.6 | 38.3 |
| Confident | 13.3 | 22.2 | 21.9 | 26.5 | 17.1 | 24.9 |
| Unsure | 8.7 | 4.3 | 15.2 | 2.9 | 14.0 | 4.3 |
| Somewhat confident | 1.7 | 16.5 | 2.6 | 13.2 | 4.9 | 23.2 |
| Not confident | 1.9 | 4.3 | 2.9 | 9.6 | 5.3 | 9.3 |
* In 2008 1,679 Harris County, Texas, voters leaving their Election Day or early voting place were asked on a scale of 1 to 5, where 1 means strongly disagree and 5 means strongly agree do you agree or disagree with the following statement: “I am confident that my vote will be accurately counted.”
** In 2012 802 Harris County, Texas, voters and likely voters were asked in a telephone survey (October 17-27) if they were “very confident, confident, somewhat confident, not confident or unsure that their vote would be accurately counted.”
Robert Stein, Ph.D., is the fellow in urban politics at the James A. Baker III Institute for Public Policy. He is also Rice University’s Lena Gohlman Fox Professor of Political Science and a nationally recognized political analyst and expert on elections.