Field Estimate← Methodology
Home / Sample Analysis
Sample AnalysisDomain: economicsResolved YES · September 18, 2024

Will the Federal Reserve begin its rate-cutting cycle before January 2025?

Analysis date: June 15, 2024 · This is a historical demonstration — the question resolved YES on September 18, 2024.

CROWD PROBABILITY
68%
Average of Metaculus, Manifold, Polymarket
CALIBRATED ESTIMATE
72%
Le (2026) slope b=1.2 (economics, 1mo+)
BRIER SCORE
0.077
Lower = better (0 = perfect, 0.25 = random)
01

Causal Factors

PCE inflation declining toward 2% targetYES · High

Core PCE at 2.6% in April 2024, down from 4.7% in early 2023. The pace of disinflation was consistent with Fed projections.

Labor market showing controlled softeningYES · Medium

Unemployment at 4.0% by May 2024, up from 3.4% trough. Wage growth slowing to ~3.9% YoY. Goldilocks conditions favored a cut.

Forward guidance from Fed officialsYES · High

Multiple FOMC members signaled openness to cuts 'later this year' in May–June 2024 speeches. The dot plot showed median expectation of one cut in 2024.

Services inflation remaining stickyNO · Medium

Shelter and services CPI components running above target. Risk that last-mile disinflation would stall, pushing the first cut to 2025.

Political sensitivity of timingNO · Low

Any September cut would come weeks before the November election. The Fed's historical reluctance to act near elections introduced some timing uncertainty.

02

Reference Classes

Fed rate cycle reversals (1980–2020)78%

In 8 of 9 tightening cycles, the Fed began cutting within 24 months of the final hike. The final hike in this cycle was July 2023.

Cases where inflation fell >2pp in 12 months80%

Historically, when core PCE dropped more than 2 percentage points in a 12-month window without recession, the Fed cut within 6 months in 4 of 5 cases.

FOMC dot plot accuracy (2015–2023)70%

When the median dot plot showed ≥1 cut by year-end, the Fed delivered at least one cut 70% of the time. Dot plot signaled 1 cut for 2024.

03

Inside-View Adjustment

+4pp above reference class averagefrom reference class average (~75%)

The convergence of disinflation data, forward guidance, and labor market softening is stronger than the typical pre-cut environment. The main risk — sticky services inflation — is real but the Fed has historically accepted 'good enough' conditions rather than waiting for perfection.

KEY ASSUMPTIONS

  • No significant reacceleration in core PCE above 3.5%
  • No sharp deterioration in financial stability requiring rate hikes
  • No major exogenous supply shock (energy, food) reversing disinflation
04

Community Consensus

71%
Metaculus
65%
Manifold
68%
Market consensus
68%
Average

Market-implied probabilities are sourced from Polymarket and Metaculus at the time of analysis. These represent crowd probability estimates — the aggregate view of many independent forecasters — and serve as a calibration anchor.

05

Structural Synthesis

The reference class base rate centers around 75% probability of a 2024 cut. The crowd estimate at 68% is modestly below this, likely reflecting residual uncertainty about services inflation and election-timing concerns. The Le (2026) calibration slope for economics contracts at the 1mo+ horizon is 1.20, indicating mild underconfidence — the crowd is slightly compressing probabilities toward 50%. Applying this recalibration shifts the estimate to 72%. The inside-view analysis supports a similar adjustment: conditions favoring a cut are notably stronger than the average pre-cut environment, while the risk factors are identifiable and manageable. The calibrated estimate of 72% is the final forecast.

FINAL CALIBRATED ESTIMATE
72%
Applied Le (2026) economics × 1mo+ slope (b = 1.2) to crowd estimate of 68%. Formula: p* = 0.68^1.2 / (0.68^1.2 + 0.32^1.2). Post-resolution Brier score: 0.077.
This is a historical demonstration analysis. All probability estimates shown are as of June 15, 2024 — before the question resolved. The calibration methodology and workflow steps are fully described on the methodology page. Source code is available on GitHub.