In a new poll of 1006 likely voters nationwide, former Vice-President Joe Biden holds on to a two-point lead over President Donald Trump, 49% to 47% with 4% not sure. The random sample online poll was taken on Friday beginning after 7 PM and has an overall margin-of-sampling error of +/-3.2 percentage points.

Our July 8th poll showed Biden leading Trump 49% to 42% and the previous poll released on August 29th had Biden leading by six points (48% to 42%).

The former VP is leading among Democrats 91%-8%, voters 18-29 years of age (60%-35%) and those 30-49 (50%-45%), as well as among women (56%-41%), Progressives (85%-15%), Liberals (85%-14%) and Moderates (58%-36%). He also leads among Hispanics is 61%-34% and Blacks (87%-11%).

The President, who was hospitalized before the poll was launched, leads among fellow Republicans (94%-6%), voters 50-64 (53%-45%) and those over 65 (50%-47%). As of now, Mr. Trump is capturing a slight majority of Catholic voters (51%-46%), Evangelicals (68%-32%), Conservatives (78%-18%), and Very Conservative voters (92%-8%). He also is ahead among Whites (56%-40%) and parents of children under 17 living at home (54%-40%).

The two candidates are tied among Independents at 44% each with 12% undecided.

In a 4-way test including two prominent third-party candidates, Mr. Biden leads 47%-45%, with 2% each for Dr. Jo Jorgensen of the Libertarian Party and Green nominee Howie Hawkins.

In other results:

  • By a margin of 43%-40%, voters still think that Mr. Trump will win the election
  • The President’s job approval has climbed to 50% while 47% disapprove
  • 31% say the country is heading in the right direction, 60% feel it is moving in the wrong direction
  • Overall, the top issues in rank order are Covid-19 (38%), Jobs/Economy (34%), Health Care (33%), Crime/Public Safety (19%), Taxes/Budget (15%), Immigration (13%), and Climate Change/Environment (12%). Democrats are far more likely to select Covid-19 than Republicans (47% to 28%), while Republicans select Jobs/Economy 40% to 29% over the Democrats, as well as Crime/Public Safety (24%-12%).
  • For 18-29 year-olds Climate Change/Environment is the second highest issue (22%)
  • Among voters who have voted already, or plan to vote by mail, Biden leads 61% to 36%. On the flip side, Among those who intend to vote in person, Trump leads 58% to 39%.

Pollster John Zogby: Contrary to my own observations, it looks like the President has not been hurt by his debate performance nor his hospitalization. His 47% performance is actually one point higher than his vote percentage in 2016. For now, he appears to have consolidated his base of Whites, parents, conservatives, men, and his own party’s voters. Joe Biden looks as if he is on his way to doing the same with his base. His numbers among Hispanics are respectable but not quite at the 66%-67% he really needs. The same with Blacks. His 86% is better than our last poll but he needs to hit 90%, especially in those key battleground states of Pennsylvania, Michigan, Wisconsin, North Carolina, and Georgia. His 60%-35% lead among young voters is about where he needs to be. He is going to have to get his position on climate change out better through all the noise. Our last poll had Mr. Biden leading strongly among independents but now the two candidates are tied with 12% still undecided. It is important to note that since 2000 the Presidential races have had a see-saw quality and this may continue to play out for the remaining month.  This poll shows that the potential for a “red to blue shift” could be very real.  With Trump winning the count on Election day, and Biden’s mail-ins coming in after the election.


Click HERE for crosstabs


John Zogby Strategies

Nearly 40 years of excellence providing the most accurate polling and market research – servicing clients in over 80 countries, ranging from global NGOs, Fortune 500 companies, candidates from all parties, business startups, and local municipalities trying to redefine their strategy amidst dynamic changes.  JZS asks revealing questions that get to the core and point to the future, offer meaningful analysis, and guiding strategies.

EMI Research Solutions

EMI Research Solutions is a leading online sample provider. Founded in 1999, leverages more than 150 partners in over 95 countries to provide high quality, most cost-effective sample for any project. Combining superior expertise, white glove service, and unmatched reach makes EMI your one-stop sample provider – getting it done without compromise!



  1. You say it is a random online poll. How does that work? How do you randomize an online poll? How do you control for a self-selection bias?

    • Great questions. Here is our methodology statement.

      Using an in-house list from a longtime vendor colleague, JZS sent email invitations to a random sample of the total pool of addresses in the U.S. and were stratified to be representative by region. The total list of emails is over 15 million nationwide and includes datapoints that cover demographics and behaviors of all potential participants. The overall sample is already representative of the population nationwide. Efforts were made to target groups that are normally under-represented in typical samples. Each email invitation, randomly selected, was password coded and secured so the IP addresses were not tracked as well as to prevent each respondent from taking the survey more than once. Subsets of the data have a larger margin of error than the whole data set. As a rule, we do not rely on the validity of very small subsets, especially smaller than 70 respondents. At that subset, we can make estimations based on the data, but in such cases, the data is more qualitative than quantitative.

      Here are some of thoughts to consider:
      Random digit dialing in the year 2020 – reduces already dismal response rates down to the infinitesimal.

      We have done and seen studies that show that the distribution of our online samples (i.e. the representation by race, age, etc.) are better than the distribution in landline/cell phone sampling.

      Final question to consider. How many have landlines today? How many with cell phones actually pick up to participate in the survey. Compare this with a nation that has 94% connection to online.

      We still do live call methodology for clients that request it.

  2. Mr. Zogby, you seem to be Republic-favoring even more than Rasmussen. “2%”? Really?

    Is it a poor methodology or personal preference? Just saying )

    • See our methodology statement under the “in the news” tab title “Just How Big Is Joe Biden’s Lead? A Guide To The Difference in Polls”

      There you will see why our polls are different.

  3. John, how confident (For the lack of a better term on my mind right now) is the data regarding minorities? 11% or 15% of the black vote might not sound much but could very well be the difference for who’s winning come November, (Likewise with the Hispanic vote).
    Also, do you plan to do detailed polls for Battleground states? I think they’d give another angle on the race.
    I have to say, your polls to me based on anecdotal evidence seem more likely than the ones mentioned most by the media, which to me seem to over sample Democrats by a lot (Suggesting an even more one sided turnout than 2008) and your methodology seems very smart too. (I don’t think people have the patience for phone calls these days).
    I look forward towards more polls from you as the race continues over the next month!

    • Ori, thank you for your kind and thoughtful note.

      For the most reliable data on the Black vote, see the poll we did for Newsmax in which we surveyed 600 black likely voters nationwide.

      Our typical nationwide poll usually has about n=120 which is representative of the population as that’s what they are nationally in terms of percent of population. When reporting on a number like n=120 the margin of error is about +/- 9 percentage points. We still report on it but we always mention that subgroups have a higher margin of error than the overall sample.

      Which is why I encourage you to look at our poll of black voters that I have linked. It was early September so over a month ago, but it is consistent with what we have been finding.

      Regarding battleground state polling. We’d love to. We have been pitching clients but so far our clients have been mostly interested in Nationwide polls of subgroups.

  4. Thanks for sharing this results Mr. Zogby. I have searched your site on more information about your methodolgy and details on the stratified method, who you polled, etc. Do you provide more detailed reports on the sampling methods of how you collected your data as well as a breakdown of who are these 1006 people surveyed, for example, that might be similar to other polling organizations like those made available by Gallup or Pew?

    • did you read the methodology report linked on the site? Please send us an example of what you are referencing. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *

Post comment