It’s (still) the economy (and politics) — stupid

Stay informed with free updates

Simply sign up to the Global Economy myFT Digest — delivered directly to your inbox.

Partisan loyalties clearly influence Americans’ views of the economy — perhaps even more so than in the past. But the narrative that Joe Biden is not getting credit for America’s solid economy because of it, feels a little too basic?

The argument often starts by looking at consumer sentiment. The chart below shows that since mid-2022 Democrats’ view of the US economy has soared higher than that of Republicans’ as the post-pandemic recovery picked up steam. (Americans would also put more faith in Donald Trump to handle the economy).

This is used as basis to suggest that Biden’s low approval ratings — currently around 39 per cent nationally — would be higher if partisan attitudes were not holding him back. That may be true to an extent, but there are a few flaws:

I) Consumer sentiment scores are not always the best gauge of households’ day-to-day experience of the economy. They often clump a number of backward and forward-looking questions awkwardly into an index.

Numerous surveys suggest that voters regard the cost of living as the most important factor in evaluating how the economy is doing. Indeed, if we splice out the closest proxy to this element in the University of Michigan’s Indicator of Consumer Sentiment — households’ view of their finances, relative to the year before — it hardly paints a picture of optimism:

II) Relatedly, America’s falling inflation rate, near record-low unemployment and resilient growth perplexes many who would expect Biden’s approval rating to be higher. But, these are static indicators. Inflation may be 3.2 per cent, but households are currently battling price levels that are around a fifth higher than when Biden’s term began.

III) It may be true that America’s headline economic data looks healthy — but not every household experiences that. Inflation and unemployment is, of course, not the same everywhere. Might approval ratings reflect local-level economic conditions?

We can address these flaws together.

The Misery Index — the sum of the inflation and unemployment rate — is a more explicit indicator of economic sentiment. We can replace the inflation rate with the change in price-level between January 2021 and January 2024, to target more closely the change in cost of living that households have faced, and split it by state. Current voting intentions can then be mapped on from various pollsters.

The chart below shows all states above the US Augmented Misery Index (AMI) for January 2024 — which comes in at 21.6 — three years since Biden’s inauguration. Notice anything?

Of the 15 states above the US’ AMI score, 11 are currently Republican-leaning. Two swing states are also represented in the top two slots: Arizona and Nevada. And, of the top 25 US states by AMI score, 18 are Republican-leaning.

Looking across all states there is a weak negative correlation — a higher AMI score is generally associated with a lower approval rating. (If we remove North Dakota, which is a bit of an outlier — it has a low misery score but a very low approval rating — the correlation looks even stronger)

Other measures of distress, such as transitions into debt delinquency, map on to this picture too. Indeed, in Florida, Nevada and Texas — where Biden has a below national-level approval rating — the rate of delinquency is also above the US-wide level.

What insights can we glean from this?

I) A number of states with lower than average approval ratings for Biden have indeed experienced relatively more “miserable” economic conditions, offering some non-partisan justification for their sentiment

II) Some states — Colorado, New Mexico and Delaware, for instance — have high approval of Biden, despite their relatively more miserable economic experience

Together, this should put some doubt on the narrative that Biden has a low approval rating simply due to partisanship. Party affiliation certainly plays its part — in both directions — but its salience will differ depending on the household and region. Counties, rural areas, and cities will undoubtedly reveal further differences.

With a country the size of the US — which also has vast inequalities — national-level data can be a bit blinding. The election, after all, can be won and lost at a local-level.