Skip to main content
Menu
Republicans And Democrats Should Be Worried About 2020

Welcome to Pollapalooza, our weekly polling roundup.

Poll of the week

The electorate in 2020 is likely to be similar to 2016’s. This should worry people in both parties — for different reasons, of course.

Our poll of the week is not technically one poll, but an in-depth study that looks at the changing demographics of the U.S. electorate and how those shifts are likely to affect future presidential elections. The study was a joint project of the Bipartisan Policy Center, the Brookings Institution, the Center for American Progress and the Public Religion Research Institute, all think tanks based in Washington, D.C.

Using U.S. Census Bureau and other data, the authors of the study project that the eligible voting population in 2020 will break down like this:

  • 44 percent white people without college degrees (down from 46 percent in 20161
  • 23 percent white people with college degrees (compared to 22 percent in 2016)
  • 13 percent black people (12 percent in 2016)
  • 13 percent Latinos (12 percent in 2016)
  • 8 percent people who are either Asian or another race or ethnicity that is not black, white or Latino (7 percent in 2016)

Those numbers are good for Democrats and bad for Republicans in the sense that the parts of the electorate that are expected to increase by 2020 (non-whites and white people with college degrees) are generally much more supportive of Democrats. In 2016, according to exit polls, Donald Trump won among white people without degrees by 37 percentage points and among white people with degrees by just 3 points. He lost among non-white people by 53 points.

At first glance, the projected growth of white people with college degrees and non-white people in the electorate seems fairly small. But it’s building on a longer-term trend that is helping Democrats. Over the past 20 years, the percentage of college graduates and non-white people in the U.S. electorate has increased while the share of whites without college degrees has decreased. And in that period (1997 through 2017), the Republican candidate has lost the popular vote in four out of five presidential elections2 (in 2000 and 2016, Republicans won the Electoral College despite losing the popular vote).

Indeed, the report’s authors argue that this slight shift in the projected makeup of eligible voters by 2020 would have tipped the 2016 election toward Hillary Clinton. Basically, if you rerun the 2016 election — if the various voting blocs backed Democrats and Republicans at the same percentages as in 2016 and turnout stayed the same by group — but you use the projected 2020 population, these small increases in the number of minorities and college-educated voters and the decline of white-working class voters would turn Michigan, Pennsylvania and Wisconsin from very narrow GOP wins to similarly narrow Democratic wins, according to the authors. Clinton would have won the Electoral College if she had carried those three states.

But, of course, demographics are not destiny — they never have been. There are problems in these projections for Democrats, too:

  1. The population of eligible voters is less diverse than the U.S. population overall (the latter is about 39 percent non-white, while the pool of potential 2020 voters is projected to be about 33 percent non-white).
  2. In 2020, almost half the electorate (44 percent of voters) is projected to belong to a demographic (white and working class) that now overwhelmingly backs Republicans. These voters were about equally split between the two parties as recently as the 1990s.
  3. Non-white people are a huge swath of the electorate in some states, but not in many others. In 2020, white people are projected to make up more than 80 percent of the electorate in 23 states, according to this study. Non-white people are expected to be more than 40 percent of the electorate in just six states (California, Georgia, Hawaii, Maryland, New Mexico and Texas) and Washington, D.C., by that point.

And the dynamic of a country gradually becoming less white while some states stay very white is likely to endure beyond 2020. Even in 2036, when the researchers think that about 41 percent of the electorate nationally will be non-white, they project that 11 states will have an eligible voter population that is more than 80 percent white.

Again, demographics are not destiny, in the short or long term. Democrats could reverse their slide among whites without college degrees. Republican candidates could become more popular with Asian, black and Latino voters. Maybe a strong Democratic presidential candidate in 2020, particularly with Trump’s soft approval ratings, can improve Democrats’ performance among non-white voters, white people with college degrees and white people without college degrees all at the same time. It’s difficult to see Trump altering these demographic trends, but perhaps a future GOP presidential candidate could. Moreover, focusing only on the demographic makeup of the electorate ignores turnout, and Republican-leaning groups tend to turn out to vote at higher rates than Democratic-leaning groups.3

But this study underlines two broad truths in today’s U.S. politics. The Democrats need to do a better job wooing white working-class voters and getting more blacks to the polls (black populations are larger than Asian or Latino ones in states like Michigan and Ohio that have a lot of Electoral College votes). If they don’t, they’ll have a problem winning states in the middle of the country and therefore the Electoral College. At the same time, Republicans have a huge problem with non-white voters that imperils their ability to win national elections and should not be ignored because of Trump’s victory in 2016.

The authors of this study, in looking at the parties’ demographic coalitions, wrote that “quite a few future scenarios could mimic the result of the 2016 election — a Democratic win in the popular vote with a Republican win in the Electoral College.” That is really bad news for Democrats, but hardly a great place for the GOP to be in either: trying to lead a country where a plurality of voters voted for the other party.

Other polling nuggets

  • A new Marist poll found that Republicans are beginning to view Robert Mueller’s investigation less favorably. The share of Republicans who have an unfavorable view of Mueller increased from 30 percent (in a March poll) to 49 percent. The share of Republicans who think the investigation is fair decreased from 30 percent to 22 percent. That said, the percentage of Republicans who believe that the special counsel should be fired remained steady at about 25 percent.
  • Texas Sen. Ted Cruz is facing a strong challenge from Democratic opponent Beto O’Rourke according to a new Quinnipiac poll, which shows Cruz leading 47 percent to 44 percent.
  • On Tuesday, voters in Arizona’s 8th Congressional District will cast ballots on Tuesday to replace Republican Rep. Trent Franks, who resigned after he was accused of offering an aide $5 million to impregnate her. While one April poll found the Republican candidate, Debbie Lesko, ahead of Democrat Hiral Tipirneni by 10 points, two other polls — one independent and one conducted for the Tipirneni campaign — found a tied race.
  • A Morning Consult poll found that 66 percent of registered voters in the U.S. supported the airstrikes that the U.S., UK and France launched on Syria, but 57 percent said they weren’t confident that the attack would prevent the Syrian government from using chemical weapons again.
  • A YouGov poll conducted after the strike on Syria found that 27 percent of adults believe the U.S. has a responsibility to do something about the fighting in that country; many more — 49 percent — said they think the U.S. should do something about Syria’s “alleged use of chemical weapons.”
  • 68 percent of Republicans and 74 percent of Democrats support tougher sanctions against Russia, according to an ABC News/Washington Post poll.
  • A SurveyMonkey poll found that 78 percent of Republicans support President Trump’s tariffs on steel and aluminum imports, but half still think that free trade agreements help the U.S. economy.
  • 75 percent of white evangelical Protestants have a favorable view of Trump, according to a new PRRI survey. That’s the best number for Trump with that group since PRRI started asking the question in 2015.
  • A SurveyMonkey poll found that 49 percent of Mississippi residents consider their state’s economy either “very bad” or “fairly bad,” compared with only 28 percent in Southern states more broadly.4 Sixty-two percent of Mississippians said they were willing to pay more taxes to fund better infrastructure.
  • 81 percent of Canadians think the U.S. decision to remove net neutrality was a bad one, and 72 percent are concerned that it could increase costs of web-based services for Canadians, according to a poll by the Angus Reid Institute.
  • Despite being jailed on corruption-related charges, former Brazilian president Lula da Silva remains the front-runner in polls to win another term in the election scheduled for October.

Trump’s approval rating

Trump’s job approval rating is 40.2 percent; his disapproval rating is 54.2 percent. Last week, his approval rating was 40.6 percent, compared with a disapproval rating of 53.3 percent. Broadly, the president’s approval rating has been fairly stable. About a year ago (on April 20, 2017), Trump’s approval rating was 42.0 percent, and his disapproval rating was 52.0 percent.

The generic ballot

The Democrats hold a 46.4 percent to 39.8 percent advantage on the generic congressional ballot this week. Last week, Democrats were up 46.2 percent to 39.6 percent.

Footnotes

  1. Two of the three authors of this week’s report released a previous study with estimates of the 2016 electorate. We are using those figures here. Those estimates differ from the exit polls conducted by the major news outlets, as the authors argue that the exit polls underestimated the number of white, non-college-educated people in the electorate.

  2. 2004 was the exception

  3. If college-educated voters continue to trend Democratic-leaning, they would be an exception to this pattern.

  4. Alabama, Arkansas, Florida, Georgia, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Texas and Virginia

Perry Bacon Jr. is a senior writer for FiveThirtyEight.

Dhrumil Mehta is a database journalist at FiveThirtyEight focusing on politics.

Comments