The Obama Effect: Voting Numbers

One of the narratives that has come front and center in the last two Presidential Elections is that of race in America, with an emphasis on its role in the  American political landscape .The 2008 election of Barack Obama was marked by increasing voter turnout rates for African Americans and younger voters. As these trends went against historical trends, many “students of voter behavior” were curious whether these trends would continue in the Congressional election of 2010 and the Presidential election of 2012.

William H. Frey, a demographer at the Brookings Institution and a friend of Data Driven Detroit, analyzed the 2012 elections for the Associated Press using census data on eligible voters and turnout, along with November’s exit polling. He estimated total votes for Obama and Romney under a scenario where 2012 turnout rates for all racial groups matched those in 2004. Overall, 2012 voter turnout was roughly 58 percent, down from 62 percent in 2008 and 60 percent in 2004.[1]

The results of his analysis suggested that America’s blacks voted at a higher rate than other minority groups in 2012, and by most measures surpassed the white turnout for the first time. This reflected a polarized presidential election in which blacks strongly supported Barack Obama while some whites stayed home. Had people voted last November at the same rates they did in 2004, when black turnout was below its current historic levels, Republican Mitt Romney would have won narrowly.

Confirmation of the Obama EffectVoter Turnout Text_3

Wednesday May 8th’s release by the Census Bureau confirms Frey’s research and other surveys estimating a rise in African American voting in the last 2 presidential elections. Results of the 2012 Current Population Survey show that African Americans turned out to vote in 2012 at rates higher than any other race/ethnic group for the first time since the Census Bureau began collecting voting data by citizenship in 1996.

The 2012 results, in Figure 1, show that African American voter turnout exceeded that of White, non-Hispanics by 2.1 percent! This compares with a rate that had run 5 to 7 percentage points behind whites from 1996 through 2004. The first election of Barack Obama in 2008 brought with it a 4.7 percentage point increase for African Americans and a 1.1 percentage point decrease for whites. While white turnout took another 2 percent drop in 2012, African Americans increased by another 1.5 percentage points.

Figure 1_final

“The 2012 turnout is a milestone for blacks and a huge potential turning point,” said Andra Gillespie, a political science professor at Emory University who has written extensively on black politicians. “What it suggests is that there is an ‘Obama effect’ where people were motivated to support Barack Obama. But it also means that black turnout may not always be higher, if future races aren’t as salient.”

Changing Demographics & Future Elections

Census data and exit polling show that whites and blacks will remain the two largest racial groups of eligible voters for the next decade. Overall, the findings represent a tipping point for blacks, who for much of America’s history were disenfranchised and then effectively barred from voting until passage of the Voting Rights Act in 1965. Last year’s heavy black turnout came despite concerns about the effect of new voter-identification laws on minority voting, outweighed by the desire to re-elect the first black president.

The numbers also offer a cautionary note to both Democrats and Republicans after Obama won in November with a historically low percentage of white supporters. While Latinos are now the biggest driver of U.S. population growth, they still trail whites and blacks in turnout and electoral share, because many of the Hispanics in the country are children or noncitizens.

The Asian and Latino populations continued to show turnout rates below 50 percent. This is not a factor of eligibility, however, because the Census Bureau has used a measure of citizenship to determine the eligible electorate.

It is clear that this new information will be critical inputs to the strategies pursued by both political parties. The increasing diversity of our electorate, marked by a slow growth African American population, a medium growth, immigration-based, Asian population, and a high growth, non-citizen Hispanic/Latino population, requires strategic investment on both sides of the political spectrum.

The turnout rates seen below bring forth a number of questions.

  • The white, non-Hispanic population is aging and decreasing in numbers. Was it the candidate choice that resulted in their decreased turnout or something else?
  • Will the increasing turnout of African Americans continue if there is no African American in the next Presidential race?
  • What are the strategies that need to be followed in order to get greater turnout for Asians and Latinos?

African American turnout, which trailed non-Hispanic whites by 7.7 percentage points in 1996, surpassed them by 2.1 percentage points in 2012. Both Asian and Latino turnout rates, while showing some variability over the years, have not moved in comparison to 1996 – still trailing non-Hispanic Whites by 17 and 16 percent, respectively.

Figure 2 provides one more way of looking at the 2012 results. Here we are comparing each group’s share of the electorate (eligible voters) with their share of actual voters. It is clear that both non-Hispanic Whites and African Americans are over-represented as voters, while Asians and Latinos are under-represented.

Figure 2_final

Gender’s Effect of Voter Turnout

There are subtexts to these trends and those involve the gender and age composition of voters within each of the major race/ethnic groups. Voting rates have historically varied according to gender. In every presidential election since 1996, women have voted at higher rates than men. The 2012 election produced a gap in the favor of women of some 4 percentage points. We can understand this gap in greater detail by looking at it across race and Hispanic origin.

Figure 3 provides a comparison of female voting rates to male voting rates. The differential for African Americans is by far the greatest and has remained high throughout the 16 year period. The 2012 differential was the highest over this period – 8.7 percentage points. The gap for non-Hispanic Whites has been consistent across the period and averaged less than half that for African Americans. While the gap for Latinos has decreased over the years, the Asian gap, which was the only to ever favor men, has increased to be close to that for non-Hispanic Whites.

Figure 3_final

Youth Today: What happened to the 18-24 year old voters?

We finish up this analysis with a quick look at age trends. In 2012, overall turnout rates decreased in comparison with both 2004 and 2008, a drop in voting characterized by large decreases in youth voting rates for all race groups and Hispanics. Statistically significant voting rate decreases were observed in the 18 to 24 years cohort across all three groups – non-Hispanic Whites, Blacks and Hispanics. The Census Bureau was not able to track age patterns within the Asian community.

What happened to the young voters who made all the noise in 2008? Have they lost their passion through voter-based civic engagement? How do candidates at all levels of government re-engage the young voter?

These data provide a great deal of food for thought. Data Driven Detroit is working diligently to provide web-based tools and resources for the elections of 2013. Detroit will be electing a new Mayor and a district-based City Council. D3 and the Michigan Nonprofit Association want to make sure that we do everything we can to get out the vote in November. Detroit has an opportunity to show that its citizens are engaged and ready to move the city forward.

Join us in that effort.


[1] His analysis also used population projections to estimate the shares of eligible voters by race group through 2030. The numbers are supplemented with material from the Pew Research Center and George Mason University associate professor Michael McDonald, a leader in the field of voter turnout who separately reviewed aggregate turnout levels across states, as well as AP interviews with the Census Bureau and other experts. The bureau is scheduled to release data on voter turnout in May (which I detail in a moment).