How do term-average approval ratings correlate with re-election outcomes and historical legacy?
Executive summary
Term-average presidential approval ratings are a strong but imperfect predictor of re-election chances: presidents who sustain higher approval numbers across a term have historically enjoyed better odds at winning a second term, but many exceptions and contextual factors mean approval is neither necessary nor sufficient for victory [1] [2]. Approval’s relationship to historical legacy is even looser — some presidents with middling or low in-term approval are later rated highly by historians, and vice versa, so term-average polls capture contemporaneous popularity more than long-term judgment [3] [4].
1. How approval maps onto re-election outcomes
Empirical records assembled by Gallup and archival projects show a clear statistical link: incumbents who enter an election with higher job-approval averages tend to win reelection more often, and data series used by poll aggregators and academic archives are the basis for that conventional wisdom [1] [5]. Analysts and university compilations note that sustained approval above the 50% mark across much of a term greatly improves an incumbent’s electoral prospects, while plunges into the 30s or lower have historically presaged trouble at the ballot box [2] [6]. That correlation holds repeatedly across post‑war presidents, but it is not deterministic — electoral coalitions, opponents, and specific crises can override raw approval numbers [2].
2. The “first-term advantage” and the second-term drag
One persistent pattern is the better approval performance in first terms than second terms: Washington and Lee and other historical summaries find that presidents who serve more than four years usually had higher averages in the run-up to a reelection win than during the subsequent second term, indicating a structural decline in approval over time that complicates any simple forecasting rule [2]. The Roper Center and the American Presidency Project maintain the time-series that display these downtrends and the familiar “honeymoon” peaks followed by erosion, meaning an incumbent’s term-average is shaped by whether early surges or late slumps dominate the weighted mean [6] [1].
3. Approval versus historical legacy: short-term mood vs. long-run judgment
Retrospective polling and historian rankings diverge from in-term approval: Gallup’s retrospective measures show presidents like John F. Kennedy scoring extremely high in later public memory despite having middling lows while in office, whereas presidents with stronger contemporaneous approval can be judged harshly by historians later — demonstrating that legacy depends on outcomes, institutional changes, and post-presidential narratives beyond term averages [3] [4]. Academic work on presidential reputation emphasizes that policy durability, crises handled, and documentary record shape legacy in ways polls cannot fully capture [7].
4. Causal confounders that weaken the correlation
Approval numbers both reflect and shape political reality, but they are driven by many confounders: the macroeconomy, wars and crises, partisan polarization, and media cycles all push approval independently of policy competence, and incumbents sometimes exploit short-term “rallies” around foreign policy to temporarily boost numbers without long-term electoral payoff [2] [7]. Models and trackers that combine approval with economic indicators or state-level patterns produce better electoral forecasts than approval alone, underscoring how approval is one input among many [8] [5].
5. Measurement traps, institutional agendas, and polling variance
Approval data are generated and aggregated by institutions with differing methods — Gallup, Roper, university archives, and modern aggregators use different sampling frames and weighting schemes — which creates variation and debate about exactly how high or low an average is at any moment; averages like Nate Silver’s or Ballotpedia’s try to smooth this but carry methodological assumptions that shape the takeaway [1] [5] [9]. Media and political actors also selectively cite approval figures that support narratives (re-electability, crisis, mandate), producing implicit agendas in public discussion even while the underlying datasets remain valuable for trend analysis [10] [11].
6. Bottom line: useful signal, noisy predictor, weak historian
Term-average approval ratings are a useful, empirically grounded signal about an incumbent’s immediate political standing and a significant predictor of re-election probability when combined with economic and electoral-context indicators, but they are noisy and subject to confounders; and they are a poor single-variable predictor of long-term historical legacy, which depends on outcomes, policy durability, and later reinterpretation rather than contemporaneous popularity alone [1] [3] [7].