Browse Adzuna Blog »

Was there any connection between the job market and the US election?

For most people, this US Presidential election came at a really interesting point as the country continued to deal with the covid-19 pandemic, and the economic cost that this had. At the time of writing, there are still some disputes about certain votes, but most of the legal challenges have been defeated, and most publications have called all the states. So we wanted to have a look at it through our favored prism of the job market.

In the exit polls posted on election night, the economy came out as a bigger factor for voters in this election than the pandemic-handling. Most of the economic indicators including the employment numbers were improving throughout the first 3 years of Donald Trump’s presidency before being radically upended by the virus. 

We’ll take a look at the states that were in the best shape leading up to the election, those that had suffered the greatest losses in job volumes, and how the votes ended up. 

The best & worst performing states 

If we accept the theory that a well-performing economy will benefit the incumbent more than the challenger, then the states that were doing well in the lead up to the election would likely favor Trump over Biden. Of the 10 states that had the highest job volumes compared to their pre-pandemic levels, nine were in pretty safe states for the eventual winner – so we weren’t going to see big changes there. 

However the one state in that top 10 that was competitive was arguably the most important state this time around, and the one still under legal disputes: Pennsylvania. That state appears to have swung from Red to Blue, which goes against that theory. The only state that was considered a battleground – Ohio – stuck with Trump, which does support the theory a bit.

At the other end in the bottom 10 states, there are more competitive states – Texas, Florida and Nevada, but none of them switched parties from 2016 to 2020, so we can’t tell too much there. However, we should look a bit deeper.

Assessing the net gain

Let’s be clear, there are many reasons why people would switch their votes from one party to another. We also have to accept that Joe Biden and Hillary Clinton are different candidates and will have provoked a different response among voters. So we naturally should judge any movements in vote with that in mind.

However, if we look at the net gain in vote share from 2016 to 2020, and how the gap between the two major parties changed, we can see some of the movements below the surface even if it wasn’t enough to flip an individual state.

Returning to the states with the highest job volumes compared to pre-pandemic levels, the top 10, Biden got an average net gain of 2.77% of the votes. This is a little below the average across states (2.87%), so not strong evidence that a good job market favors the incumbent.

At the other end, if states that had recovered less well reacted by voting against the current president we should see a greater net gain for Biden than the average. What we saw was the opposite. There was only a 2.47% net gain for Biden in the bottom 10 states, below the national average.

Based on this, there seems to be no obvious connection between job market performance and the way people voted.

Swing states

Of the 10 or so closest states, there were only two that appeared to go against the grain and swing towards Trump: Nevada and Florida. Florida is an interesting case for a number of reasons not relevant to this blog. But in terms of the job market, it’s one of the few states that has fewer jobs than pre-pandemic times. It currently has only 88% of normal volumes.

Nevada appears to have moved by a tiny margin (0.02%) so there aren’t huge conclusions we can draw from this, but it was also the state with the worst job market recovery of this group. Of the remaining 8 states, all moved towards Biden. Of these, Texas and Minnesota had not recovered their job numbers, and saw large net gains for Biden and the Democrats.

Looking at the lowest ebb

As the election has rumbled on over a year, it might be more telling to look at states which suffered the greatest loss in job vacancies and see how the population reacted. Week 18 of the year was the point where the US vacancy levels were at their lowest, so what do the top and bottom 10 look like there.

The top 10 states – those that suffered the least on paper – gave Biden a net gain of 4.15% on average, well above the average. This goes against our instincts again.

The bottom 10 – those that lost the most vacancies – gave Biden a net gain of 2.72% on average. Which matches the average swing across states – which doesn’t tell us much!

Again – no obvious connection, except to say that states that seemed to have the best job market performance ended up coming out more for the challenger (Biden) than the incumbent (Trump).

Taken in aggregate

Rather than just looking at the top and bottom ten states, what happens when we look at things in aggregate?

Well, the picture is not much clearer! In fact, there is almost no correlation (R² = 0.025) between the net gain for Biden and the proportion of jobs lost at the peak of the crisis.

If that weren’t bad enough, there is even less of a correlation (R² = 0.001) between the net gain for Biden and the recovery in job numbers by state.


Well – it appears that the answer is a fairly resounding “No”. You can slice and dice the data and if you squint you might see a connection, but not in the way you’d expect. But if we look at the data in aggregate, we would need to see a really strong correlation in order to begin to hint at a connection. Instead we have data that appears to be totally random.

The fact that so many jobs had recovered may have been a factor in making the election seem so close. A previously strong economy recovering from an external shock that few reasonably predicted does not paint the current administration in the worst light. But it appears that was not enough to shift the needle the other way.

But as we said throughout – there are many factors at play here, of which economic confidence is one, and the number of jobs available are only part of that. As we’ve seen elsewhere, while the total number of jobs is back to normal, the sectors and industries that make up the current job market are very different.