The USC Dornsife/Los Angeles Times "Daybreak" poll tracks about 3,000 eligible voters until election day, asking on a regular basis about their support for Hillary Clinton, Donald Trump or other candidates as well as their likelihood of actually casting a ballot.
We update the data each day based on the weighted average of poll responses over the previous week. That means results have less volatility than some other polls, but also means the poll lags somewhat in responding to major events in the campaign. More about the poll and why it differs from others.
We ask voters what the chance is that they will vote for Trump, Clinton or someone else, using a 0-100 scale. The overall level of support for each candidate reflects the weighted average of those responses.
* Results in the shaded areas fall within the poll’s margin of error, meaning the difference between the candidates could be the result of chance.
Voters 35-64 have so far shown the least volatility, splitting about evenly between Clinton and Trump. Trump holds the advantage among those 65 or older. The two candidates have exchanged leads among younger voters.
Trump holds an advantage among voters without a college degree. White voters who have not graduated from college are a core source of support for Trump. By contrast, Clinton has done better among voters with college degrees than previous Democrats.
Clinton holds a distinct edge among lower-income voters, reflecting her strong support among blacks and Latinos. Trump has a lead among middle-income voters.
Disaffected white middle-class voters have been the backbone of Donald Trump's presidential run, as the poll shows. Black and Latino voters lean heavily toward Clinton. Trump's statements critical of Mexicans in the U.S. illegaly have harmed him politically among Latinos.
Trump saw a significant bounce in his support from women after the Republican convention, but Clinton rebounded quickly after her convention. Trump's support among men has remained fairly steady.
We ask voters who they expect to see win, regardless of which candidate they support. Over the years, asking voters their expectation about which candidate will win often has proved to predict elections more reliably than asking how they plan to vote. That’s particularly true when the election is still many weeks away.
* Results in the shaded areas fall within the poll’s margin of error, meaning the difference between the candidates could be the result of chance.
People of all ages think Clinton will win the election, but immediately after the Republican convention, a majority of voters 65 and older briefly said they thought Trump would win.
As education levels increase, so do voters' perceptions that Clinton will win.
Voters with annual incomes of more than $75,000 are closely divided on which candidate they support, but by a significant margin, they expect Clinton will win.
No racial or ethnic group currently has a clear majority that thinks Trump will win the election.
A majority of men support Trump, but a majority of both genders believe Clinton will win.
Turnout is a key factor in any election, but may be particularly central in this one. If one candidate’s supporters are less committed to turning out than the other’s, that could point to an important weakness.
* Results in the shaded areas fall within the poll’s margin of error, meaning the difference between the candidates could be the result of chance.
The USC Dornsife/L.A. Times Presidential Election "Daybreak" Poll asks more than 400 people each day about their voting intentions. The poll is part of the Understanding America Study (UAS) at the University of Southern California's Dornsife Center for Economic and Social Research.
Each day's poll respondents are a subset of the UAS election panel, roughly 3000 U.S. citizens who were randomly recruited from among all households in the United States. Respondents are asked three predictive questions: What is the percent chance that (1) you will vote in the presidential election? (2) you will vote for Clinton, Trump, or someone else? and (3) Clinton, Trump or someone else will win?
Results are weighted to match demographic characteristics, such as race and gender, from the U.S. Census Current Population Survey, and are aligned to the 2012 presidential election outcome using self-reported votes in that election.
These charts are updated daily (just after midnight) with an average of all of the prior week's responses. The Daybreak Poll began on July 4, 2016, and will run through the November election. It is being conducted in partnership with the Jesse M. Unruh Institute of Politics and the Los Angeles Times. For more information about the survey's methods and to download the source data please visit http://election.usc.edu/.
Updated Sept. 19: The display of the poll results originally showed too large an area of uncertainty. It was corrected on Sept. 19 for all results back to the start of the poll. The error did not affect the poll results, only the way the uncertainty was visualized.
Additional credits: Lily Mihalik. Photos: Irfan Khan and Carolyn Cole / Los Angeles Times.