The map of the presidential election has taken on familiar shades over the last few decades. Red states tend to stay red. Blue states tend to stay blue. There are just a few states, the swing states—states such as Nevada, Colorado, and Florida—where the outcome isn’t obvious. Those are the true battleground states, and so they get the most attention from presidential candidates.
The vote tally in those states might be close, but the American electoral system is such that winner takes all. The small margin of voters that push the state towards the winning candidate are incredibly powerful voters. These tipping point voters have more influence on the election than voters in predictable states.
“I think most voters sort have this naive notion. One person, one vote, each vote counts the same. Of course in an electoral college system that’s not the case,” says Doug McAdam, a professor of sociology at Stanford University who studies American politics. “That’s why campaigns overwhelmingly focus their resources in battleground states.”
We decided to drill down into the demographics of voters in America to better understand which voters are likely to have the most power to determine who our next president will be. We used turnout data from the U.S. Census and election forecasts based on polling data from Fivethirtyeight.com to build a model that gives a “power score” to voters. The score changes based on their age, gender, ethnicity, and where they live. It reflects the competitiveness of the state they live in and voter turnout rates for their demographic group; ultimately, it measures the statistical likelihood that a vote from that group will decide the election:
It’s important to note that while we’re looking at voter demographics by race or ethnicity, it is not our intent to imply that voters of the same race vote as a bloc, with one voice and identical interests. In fact, we know they don’t. The goal is to better understand how our electoral system impacts the voting power of different demographic groups. Also keep in mind that the model is based on turnout rates in 2012. It is likely that turnout rates in 2016 may be different. (To learn about how we calculated the power of your vote, read: How much does your vote count: Behind the model)
Here’s what we learned: In what are currently the most competitive states, Nevada, Colorado, and Florida, where the campaigns focus their organizing efforts, one group emerges as among the most powerful voters: black women ages 65-74 living in Nevada. They live in a pivotal swing state, and they turn out to vote at higher than average rates. Based on our model, their vote is likely to be 398% more powerful than that of the average voter nationally.
That doesn’t surprise David Damore a Professor of Political Science at the University of Nevada, Las Vegas and senior analyst at Latino Decisions, a firm that has been hired to do polling for the Clinton campaign. “In 2008 and 2012 African Americans voted above their population share. They are reliable, and they are overwhelmingly Democratic.”
According to Damore, women of color, many of whom work in the hospitality industry in Las Vegas and surrounding Clark County make up a key part of the Democratic base in Nevada. Clinton will need their votes to offset strong participation from voters in more rural parts of the state who lean Republican.
Our model suggests that some of the least powerful voters are likely to be male Hispanic and Asian voters ages 18-24 living in California. These voters are more than 99% less likely to influence the outcome of the election than the average voter in November. This is because California, a predictably Democratic-leaning state ranks very low in competitiveness and because these voters have much lower turnout rates than other groups.
So, what happens when we zoom out? How do these disparities between states affect the overall voting power of voters of all ages from different racial groups nationwide? Averaging these state numbers for each voter group gives us a more complete picture of the relative power of different racial groups to decide the election.
Voter Power Scores (national average) by Race and Hispanic Origin, 2012 - 2016
Data source: U.S. Census Bureau, The Diversifying Electorate—Voting Rates by Race and Hispanic Origin in 2012, May 2013. Please note: While the Census now allows respondents to report more than one race, figures here are derived from the Census Bureau’s report and reflect data for people who reported they were a single race alone only. Use of the single-race populations does not imply that it is the preferred method of presenting or analyzing data. Voting power estimates are derived from our model and are based on 2012 historic turnout rates and Fivethirtyeight.com’s Voter Power Index, part of their election forecast, as of September 15, 2016.
Once again, we see that voters of color have much less influence than they should. The differences come down to where people live. As a group, white voters have the highest overall voting power. Their voting power in this election is estimated to be 5% higher than other voters. This is because, even though they have slightly lower turnout rates than black voters on average, they are over represented in key swing states such as Pennsylvania, with large numbers of electoral votes up for grabs.
Asian voters as a group are likely to have the least influence on the outcome in November. Again, this isn’t because their vote counts less, but because relatively few Asian voters live in swing states like Ohio or Pennsylvania where the race is competitive.
As an individual voter, this ‘swing state effect’ is easy to see: If you live in Wisconsin, you’ve been lavished with campaign attention, but if you live in California, the only way you’ll get to see the candidate is at a high-dollar fundraiser.
For example, although its voter registration programs are national, the Clinton campaign told us it is focusing its organizing efforts among Latino voters in Florida, Colorado, Nevada and North Carolina, even though they are not the states with the largest Latino populations. Nearly half of America’s Hispanic citizens, of which Latinos represent the largest share, live in California and Texas, but those states are not likely to be competitive.
As a result, the Clinton campaign’s voter outreach programs to young Latino voters — program such as “ Mi Sueño, Tu Voto/ My Dream, Your Vote,” which works to empower DREAMers to ask U.S. citizens to vote in support of them, to name just one—touch many fewer voters than they might if the map of the race was less concentrated.
We reached out to the Trump campaign to learn more about their voter outreach efforts to voters of color, but did not receive a response. In July, the campaign hired ex-Apprentice star Omarosa Manigault as the director of African-American outreach. Based on her Twitter feed, this has largely meant touring with Women for Trump in places like Michigan, North Carolina and Ohio. (Why is the campaign so interested in reaching voters of color in North Carolina? It may be because it is a ‘swingy’ state and a substantial population of black voters with exceptionally high turnout rates: 80% voted in 2012.)
Young voters of color who don’t happen to live in swing states sense this lack of attention from the campaigns. For some voters it results in a vague sense of detachment. “You know I am just not interested in politics. It’s just something I don’t pay attention to that much, even though I should,” said Wendy Gramajo, 25, a Latino college student in California who has never voted. “Ok, I’ll watch, but voting personally is not my thing. I am kind of a bystander,” said Jose Garcia, 21, who also doesn’t plan to vote in November.
Others are more astutely aware of the issue: “With the electoral college I don’t think it matters. Really, California is a typically blue state. Same thing in Texas. It is typically a red state,” said Randy Nu, a Korean American college student in California. He doesn’t vote in presidential elections. “I focus more on state level and county races.”
The danger, of course, is that the effect becomes self-reinforcing. Voters who do not have contact with the campaigns, feel like their votes don’t make a difference and so they don’t vote. Low turnout rates, lead to even less attention from the campaigns in the next election. It’s a potentially vicious circle, making the true impact of the swing state effect hard to quantify.
Outside of election reform (changing the electoral college system would require a constitutional amendment), the solution, says McAdam, is more participation, not less. Hispanic voters represent 11% of the electorate, that’s one of every 10 eligible voters, and growing. Some 500,000 Hispanic voters are added to the voting rolls each year. And, 30% of Hispanic registered voters are not affiliated with a party. They are independent, forcing candidates to court their vote. The problem: Hispanic voters have the lowest turnout rates of any group.
“If this growing Hispanic electorate were to turnout at levels even close to white voters, you would be amazed at what states would be in play. The power of their vote would be magnified,” said McAdam. But shifting the electoral map would require a new generation of activist voters. “The potential power of the Hispanic voting group will only be realized if there is record turnout.”
Until then, a very small number of Americans are deciding the outcome of presidential elections. And, these voters don’t necessarily reflect the diversity of America as a whole.
Learn about how we calculated the power of your vote: How much does your vote count: Behind the model
Data modeling and analysis: Daniel McLaughlin
Design: Kent Hernandez
Editing: Kashmir Hill and Rachel Schallom
Daniel McLaughlin is a creative technologist exploring the 2016 presidential election. Before joining Fusion, Daniel worked at the Boston Globe and graduated from MIT with a BS in urban studies and planning.
Kate Stohr is a data journalist and community builder based in San Francisco, CA.