Paul Krugman has covered this topic a lot recently in his New York Times column (e.g., New Perspectives on the Feel-Bad Economy, Nov 2023, and numerous others). The answer seems to be a mix of things, so here's my attempt to distill them into a list, ranked approximately by importance.

1. People don't follow economic indicators all that closely. Most people base their assessment of the economy on various heuristics, what Krugman calls "vibes" (e.g., Inflation, Disinflation and Vibeflation, Dec 2023). If the heuristics don't comport with economic indicators, then people's perception of the economy will follow likewise diverge from the story told by the indicators.

2. Perceptions of inflation are sticky. When you have an episode of inflation, once inflation is brought under control, prices do not go back down to their previous level. Rather, they resume their normal slow growth from the higher price level. When people compare the prices of things today to their idea of what those things ought to cost, they see that everything is more expensive than it "should" be. Often they interpret this as due to current inflation, rather than as the aftermath of a previous episode of inflation.

3. Some of the most visible indicators are lagging indicators. Most workers' rent and pay rate get adjusted at most once a year, so at any given time many of them are still working off of last year's number.

4. Bad news sells. The news media devote more column inches and air time to bad news than they do to good news. So, when economic indicators are bad you get a prominent and lengthy article about what it all means. When they are good it often gets relegated to a "news in brief" sidebar where it is easily missed.

5. Some of the strong economic indicators are not that relevant to individuals. GDP growth in Q3 of 2023 was 5.2% annualized, which is phenomenal, but what does that mean to individual citizens? In the long run it's better for growth to be high than for it to be low, but in the short run it's just a number. People are more concerned with their personal situation here and now than they are with the economy as an abstraction.

If you take all of that together, it's a recipe for public sentiment on the economy that responds quickly when things deteriorate and recovers slowly when things get better, which is pretty much what we are seeing in the polls.

Paul Krugman has covered this topic a lot recently in his New York Times column (e.g., New Perspectives on the Feel-Bad Economy, Nov 2023, and numerous others). The answer seems to be a mix of things, so here's my attempt to distill them into a list, ranked approximately by importance.

1. People don't follow economic indicators all that closely. Most people base their assessment of the economy on various heuristics, what Krugman calls "vibes" (e.g., Inflation, Disinflation and Vibeflation, Dec 2023). If the heuristics don't comport with economic indicators, then people's perception of the economy will follow likewise diverge from the story told by the indicators.

2. Perceptions of inflation are sticky. When you have an episode of inflation, once inflation is brought under control, prices do not go back down to their previous level. Rather, they resume their normal slow growth from the higher price level. When people compare the prices of things today to their idea of what those things ought to cost, they see that everything is more expensive than it "should" be. Often they interpret this as due to current inflation, rather than as the aftermath of a previous episode of inflation.

3. Perceptions of inflation are biased toward frequent purchases and away from significant purchases. I saw this one in an article today about consumer misperceptions of inflation. The author makes the point that price indexes are weighted by the fraction that they make up in an average consumer's budget. Therefore, big-ticket purchases like furniture and consumer electronics get a larger weight than day-to-day purchases because on average they make up a larger share of consumers' spending. However, the day to day purchases are more salient in consumers' memory because they are likely to have bought those things more recently.

4. Some of the most visible indicators are lagging indicators. Most workers' rent and pay rate get adjusted at most once a year, so at any given time many of them are still working off of last year's number.

5. Bad news sells. The news media devote more column inches and air time to bad news than they do to good news. So, when economic indicators are bad you get a prominent and lengthy article about what it all means. When they are good it often gets relegated to a "news in brief" sidebar where it is easily missed.

6. Some of the strong economic indicators are not that relevant to individuals. GDP growth in Q3 of 2023 was 5.2% annualized, which is phenomenal, but what does that mean to individual citizens? In the long run it's better for growth to be high than for it to be low, but in the short run it's just a number. People are more concerned with their personal situation here and now than they are with the economy as an abstraction.

If you take all of that together, it's a recipe for public sentiment on the economy that responds quickly when things deteriorate and recovers slowly when things get better, which is pretty much what we are seeing in the polls.