News in Charts: The economic outlook in the key swing states

Economics is rarely the sole factor in an election campaign, but it can and often does play a role in deciding outcomes. As we noted three months ago, US economic fundamentals were slightly supportive of the Democratic Party’s attempts to retain the White House for a second successive term. In that note, we highlighted six key states — Arizona, Georgia, Michigan, Nevada, Pennsylvania and Wisconsin — which could swing the vote either way; and we noted that polling data suggested that Donald Trump was outperforming Joe Biden in each of them. Now that Kamala Harris has taken over as the Democrats’ nominee, however, that picture has changed. The polls now suggest that the Democrats have the lead in Michigan and Wisconsin, and that with the possible exception of Nevada the gap to the Republicans has narrowed in each of the other swing states.

Refresh this chart in your browser | Edit the chart in Datastream

So what does economics have to say about the incumbent party’s odds of winning the vote in these six key states? Looking first at state-level GDP data, strong growth should offer a boost to the party of government. These data unfortunately only cover the past four presidential elections, but current four-quarter growth rates in each of six states flagged above are generally better than they were at the time of three of the past four elections, and comparable to the rates seen in 2016. That said, while GDP growth was a fairly helpful predictor of changes in the vote shares for the incumbent president’s party in the 2008 and 2012 elections (alone it can explain more than 50% of the variation in this), it yielded no useful insight into voting patterns in the 2016 and 2020 elections.

Refresh this chart in your browser | Edit the chart in Datastream

Want more charts and analysis? Access a pre-built library of charts built by Fathom Consulting via Datastream Chartbook in LSEG Workspace.

There is generally evidence that changes in the unemployment rate are also a sound predictor of electoral outcomes. If that proves the case, then Kamala Harris might have some cause for optimism — unemployment rates in the six key states are generally close to all-time lows and have seen large declines over the course of the Biden presidency (in part reflecting a return to pre-COVID norms). Indeed, there are only three instances of comparable labour market improvements during the course of a single presidential term — during Ronald Reagan’s and Barack Obama’s second terms in office (1984–88 and 2012–16 respectively), and Bill Clinton’s first (1992–96). However, somewhat worryingly for VP Harris, only President Clinton managed to translate these job market improvements into a higher vote share in these six states, with the incumbent party’s vote share falling in the other occurrences.

Refresh this chart in your browser | Edit the chart in Datastream

Another economic factor that is likely to be exercising the minds of US voters is the cost of energy — petrol pump prices are currently higher than they have been at any previous election since the inception of the EIA’s measure in 1990. While Ms Harris can point to external factors as a key driver of this (and she may even attempt to deflect some of the criticism onto President Biden), it is still likely to weigh on her election prospects and may be a decisive factor for many.

Refresh this chart in your browser | Edit the chart in Datastream

Overall, recent experience suggests that traditional economic indicators have been less effective in determining electoral outcomes in the key swing states for this year’s presidential contest. Clearly, there are often other factors at play but, to the extent that we can infer anything from economic data, it seems as though if two out of three key macro indicators (GDP and unemployment) are probably pulling in favour of Kamala Harris. Of course — absent significant adverse and unforeseen economic shocks — as the date of the election nears it is polling data itself that will be increasingly more informative.

The views expressed in this article are the views of the author, not necessarily those of LSEG.

____________________________________________________________________________________