A daily composite score measuring how much a president can still get done. Four weighted components, 70 years of back-data, zero opinion. The duck rises as the presidency floats out.
The Index is a composite score, not an opinion. Four components, all based on publicly available polling and economic data — the same sources available for presidents back to Truman. Higher = lamer.
30-day trailing average of Trump's approval minus disapproval, from the NYT presidential tracker (aggregates Gallup, YouGov, AP-NORC, Reuters/Ipsos, Quinnipiac).
Democratic minus Republican advantage on the national congressional ballot, smoothed over 30 days from our own poll database (515 polls, Dec 2024–present).
VoteScope's own projected probability of Democrats controlling the House and Senate, from our Bayesian midterms model updated daily.
University of Michigan Consumer Sentiment Index (UMCSENT via FRED). Available monthly since 1952 — enabling true historical comparison across all post-war presidencies.
Hover any component for data source details. All four components use real public data fetched automatically from NYT, FRED, and VoteScope's own model.
Drag the handle to scrub through the full series from January 20, 2025. Yellow = Lame-Duck Index. Blue = net approval (NYT). All four components are plotted from real data.
Retroactive LDI trajectories computed from the same four components — approval (Gallup archive), generic ballot (ANES), congressional control, and UMCSENT (FRED). Current Trump trajectory in yellow.
VoteScope's Bayesian model, updated daily. These probabilities feed directly into the LDI congressional control component. A high LDI historically predicts larger midterm losses.
Common questions about the index, its methodology, and how to read it.
No. The Index measures structural constraints on any presidency — it is scale-invariant to party. The same four components, applied to the Obama or Clinton second terms, produce the readings shown in the Historical Comparison section. A Democratic president with low approval, a bad generic ballot, and weak consumer sentiment would score just as high.
We picked components that (1) are available as public data going back to Truman, (2) can be fetched automatically, and (3) have documented historical predictive power for midterm results. Cabinet churn and legislative friction are interesting but harder to backfill reliably across 70 years. We may add components in future versions.
The approval component refreshes with the NYT tracker (quasi-daily). The generic ballot updates each time we run our polling pipeline (typically several times per week). The congressional control component updates with every model run. Economic sentiment is monthly (FRED releases UMCSENT mid-month).
It means Trump currently sits in the "Quacking" zone (50–75), the second-highest severity level. This is comparable to Obama in mid-2014 or Bush in mid-2006 — both of whom went on to suffer significant midterm losses. The score does not predict the future; it describes the current structural situation.
Because charts alone do not stick. The duck rises, sinks, quacks — and you remember where the number was last month. Also: "lame duck" is the actual political science term we are operationalizing.