The past couple of months have been nothing short of tumultuous for the United States and the rest of the world. While the global outbreak and spread of COVID-19 sent everyone indoors to safety, the abrupt awakening of many Americans to the racial injustices exposed by the murder of George Floyd threw even more volatility into the mix and led many out to the streets.
From the protests in all 50 States, and the numerous violent videos of police officers abusing peaceful protestors, to the national guard spraying tear gas to clear out protestors for President Trump’s infamous photo-op, American voters have been given a lot to critically think about.
Amidst significant changes to life as we know it, it can be difficult to interpret the meaning and implication of current events. As discussed in a previous piece, COVID-19 created a very difficult situation for polling and data analysis.
Conventionally, in order to develop an understanding of current events, we rely on past data and trends to explain or provide insights into current data. As the pandemic has been irreversibly disruptive and transformative to modern society, past data and trends are rendered unrelatable. Now, between the protests around the country and the president’s handling of the situation, we’re even further into uncharted territory.
This isn’t an excuse, but more a long way of saying: if you want to know what’s going to be happening in the numbers a couple weeks from now, let alone November, the honest answer is we have no idea.
With that being said, not surprisingly, polling has shown large movements over recent weeks and that movement has been towards Democrats. If we look at the Real Clear Politics National Average (with the caveat being we do not elect our President in a nationwide vote) we see that Biden had an average level of support of just over 50 percent on March 13th (right when companies in the United States started to close) and Trump was at 44 percent. For the next month both candidate’s support dropped and on April 14th Biden was as 47 while Trump was at 42.
A month later Biden moved up to 48 percent, while Trump moved up a point as well to 43 percent. As you can see in the chart below, there was some tightening in the race in early May, but then both candidates started to grow their support.
Floyd’s murder occurred on May 25th and since that point in time Biden’s support went up to 50 percent while Trump has dropped to his lowest level of 2020 at 41 percent.
The pattern is very clear, but to further drive the point home, since the middle of May there have been twenty national polls reported on Real Clear Politics. Of those surveys, all twenty show Biden winning. In thirteen of them, Biden is leading by more than 5 points and in just four of them Biden is leading by 3 points or less. It’s also worth noting that in early April Fox News reported a tie in their published survey and then six weeks later reported Biden had an 8 point lead on Trump.
As you may have heard, the president took issue with a poll released by CNN indicating President Trump stands 14 points behind Biden. The 41 percent who indicated their support for President Trump is the lowest in CNN’s tracking on this topic since April 2019, and Biden’s 55 percent support is the highest mark we have seen to date. Could this be an outlier? Sure, but there are four other polls in the past month that show Biden with a double-digit lead.
Should we trust polling?
Polling certainly took a few black eyes after the unexpected outcome of the 2016 election, and more recently the Trump campaign took issue with the CNN survey mentioned above. The campaign even went as far as asking CNN for an apology and retraction of the poll.
In this unorthodox move, the Trump campaign based their demand on the basis of poor methodology and inaccurate sampling. In their letter, the Trump camp argued that the CNN poll is “designed to mislead American voters through a biased questionnaire and skewed sampling.”
But similar to any professionally conducted public opinion research, CNN published the methodology and sampling breakdown they used. Regardless, as much as the president’s campaign takes issue with the numbers drawn and conclusions made, the fact remains that across the board polling data indicates Biden has a strong national lead.
Further, this movement isn’t simply a national trend nor exclusively relevant to the presidential race. While there are less state-specific polls to interpret, we certainly have seen a steady shift in states like Arizona where in early March Democrat Mark Kelly had a five-point advantage and now surveys show him with a double-digit lead. Also, we’ve seen Sen. Gary Peters (D) in Michigan move from a five-point to a 15-point advantage over a similar time period.
All this is to say is that the pattern is very clear of what’s happening right now.
What does this mean for November?
Overall public opinion data is a reliable source of information about what’s currently happening. That, however, shouldn’t be viewed as having the ability to accurately forecast an outcome for the November election four-plus months away.
As we’ve seen in past presidential elections, opinions change fast and can turn the tide of an election within weeks.
In 2008, Barack Obama only took a decisive lead over John McCain as late as mid-September that year. For the majority of the race, margins between the candidates never reached double digits — and often flipped.
Based on polling data from Real Clear Politics, as of Sept. 15, 2008, Obama was trailing by a margin of 1.6 points, only to take the lead days later and continually increase his leading margin until Election Day and end the race with a 7.6 point lead. Only until after Labor Day, as little as eight weeks out from the election was polling data indicative of the outcome of that election.
If the poll data we’re seeing now were conducted this coming fall, it may have a different meaning and implication for the outcome of the election. As for right now, polls being conducted in June simply tell a story of how Americans feel about the actions of the Trump Administration in response to current events. They’re not necessarily predictive. The November election is still over four months away — a lot can happen within that time frame.
In fact, given the year we’re having, we would almost be willing to bet that a lot will.
Stefan Hankin is founder and president of Lincoln Park Strategies, a Washington D.C.-based public opinion firm. Follow him on Twitter at @LPStrategies