The Great Sorting · Perception
Ask Americans whether the economy got better or worse last year and they will mostly tell you who sits in the White House. The gap between the parties' answers has roughly doubled — and it snaps shut and reopens at every inauguration.
In principle, last year's economy is a matter of fact. Prices rose by some amount; the unemployment rate was what it was. Yet when the American National Election Studies asks whether the national economy has gotten better or worse, the answer turns out to depend less on the economy than on the respondent — specifically, on whether the respondent's party holds the presidency.
The pattern has a name in political science: partisan-motivated reasoning about objective conditions. Larry Bartels showed three decades ago that partisans could not even agree on whether inflation had fallen under Reagan, when it plainly had.1 What the long ANES series adds is the trajectory: the partisan gap in economic perception has widened into a chasm, and it inverts cleanly the moment the White House changes hands.
Who sees a good economy depends on who is president
Weighted mean response to “is the economy better or worse than a year ago” (1 = worse, 3 = better), by party. Shaded bands mark presidential terms.
The mirror has a frame, though, and the frame is reality. The clearest evidence that conditions still matter is the one moment the partisan gap nearly vanished: in 1990, as a recession bit, the in-party's advantage in optimism fell to +0.04 — even the president's own partisans could not pretend the economy was fine. The 2008 financial crisis did the same, compressing the premium to +0.09. Catastrophe is bipartisan. Everything short of catastrophe is increasingly not.
Collapse the two lines into one number — how much rosier the president's co-partisans see the economy than the opposition does — and the long arc is unmistakable. The premium averaged +0.32 on the three-point scale in the 1980s; across 2012–2024 it averaged +0.73, and in 2024 it reached +0.82, its highest in the series.
The optimism gap has roughly doubled
In-party minus out-party mean economic sentiment, by year. Defined relative to the sitting president.
Catastrophe is bipartisan. Everything short of catastrophe is increasingly a matter of party.
None of this means voters are lying. Achen and Bartels argue that partisans genuinely perceive the world through a partisan lens rather than cynically reporting a party line.2 Gerber and Huber show the same logic in behavior, not just survey talk: consumption shifts when the party in power changes.3 The implication for democratic accountability is uncomfortable. If economic judgment — the classic mechanism by which voters reward and punish incumbents — is itself a product of partisanship, then the “referendum” on the economy is increasingly a referendum the electorate has already decided before the votes, and the economy, are counted.
Data. ANES Cumulative File, 1980–2024 (18 surveyed years), weighted. The item is the retrospective sociotropic question (“Would you say that over the past year the nation's economy has gotten better, stayed about the same, or gotten worse?”), recoded so 1 = worse, 2 = same, 3 = better; orientation verified against in-party means.
In-party / out-party. A respondent is “in-party” if their party identification (leaners folded into Democrats or Republicans) matches the sitting president's party in the survey year. The premium is the in-party weighted mean minus the out-party weighted mean.
Honest limits. 2020 and 2024 are mixed-mode (face-to-face plus web); the partisan gap is large in both, so the trend is not a mode artifact, but the exact 2024 level sits at the end of a changing instrument. The 1990 and 2008 troughs are reported as a feature, not hidden: they are the strongest evidence that real conditions still discipline perception at the extremes.