Skip to main content
ABC News
Introducing Our New-And-Improved Swing-O-Matic

Will Donald Trump be able to capitalize on American voters’ desire for change, or will the nation’s changing demographics doom him and propel Hillary Clinton to victory in the Electoral College?

Last December, FiveThirtyEight unveiled the Swing-O-Matic, an interactive tool that shows how demographic groups could affect the 2016 election. Today, we’re relaunching it with several refinements and improvements to help you assess Clinton’s and Trump’s Electoral College fortunes.

You can use the Swing-O-Matic to see how national shifts in turnout and partisanship within seven demographic groups would affect the outcome of the 2016 election in each state. Here’s what’s new:

1. White breakdown by gender — The Clinton-Trump matchup has featured a historic gender gap, particularly between white women and men. By popular demand, the Swing-O-Matic now has an option to break down white voters by gender, in addition to educational attainment. Use the toggle button below the maps to choose the breakdown you prefer.

2. More comprehensive census data — Since the Swing-O-Matic’s December launch, several analyses have questioned discrepancies between demographic figures reported in 2012 exit polls, the Census Bureau’s post-election voter turnout report (which features a much larger sample size), and actual voter files. For example, the exit poll found college graduates made up 47 percent of the electorate, while the Census Bureau reported their share at 37 percent. We have overhauled our interactive to blend data from exit polls, actual election results, the Census Bureau’s Current Population Survey and its annual population estimates to build what we feel is a more accurate profile of each state’s demographic landscape and electorate.1

3. Third-party slider — The 2016 election is on track to feature the highest share of non-major-party voters in two decades. Use the third-party slider below the map to simulate how more — or fewer — votes for Libertarian Party nominee Gary Johnson, Green Party candidate Jill Stein and others could affect the outcome in each state. Third-party support is capped at 20 percent, well above their current level in the polls, and the default is set at 2 percent — the 2012 share.

4. Nebraska and Maine electoral vote splits — Unlike the other 48 states, which are winner-take-all in the Electoral College, Nebraska and Maine award two votes to the statewide winner and one vote to the winner in each congressional district. In the 2016 campaign, Clinton has targeted Nebraska’s 2nd District, and Trump has targeted Maine’s 2nd District. The Swing-O-Matic now calculates shifts in both states’ districts and reflects their individual results in the Electoral College tally.

So that’s what’s new. The rest of the analysis is the same as before. To build a baseline model of the 2016 presidential election, we started with the results of the 2012 election, looking at support for Mitt Romney and President Obama by race, gender and educational attainment. We then adjusted the size of those demographic groups based on four years of population change.

From there, you can choose your own adventure: When you adjust each group’s national turnout and party breakdown, the Swing-O-Matic automatically recalculates each state’s election results, along with the outcome of the Electoral College and national popular vote.

FiveThirtyEight: Re-introducing our Swing-O-Matic interactive


  1. We started with the census’s turnout data by demographic group in each state and then adjusted it proportionally to reported votes cast. We then used exit polls to help estimate each demographic subgroup’s vote preferences. Finally, we adjusted each group’s voting size using census population estimates released since 2012.

David Wasserman is the U.S. House editor for the Cook Political Report.