Scan: How high will unemployment get before 2027?

← Back to scans

How high will unemployment get before 2027? — Above 5%

KXU3MAX-27-5 · Economics · 2026-05-04

29%

Agent

33%

Market Price

-4.4%

Edge

50%

Confidence

Volume: 287,490

Spread: 1.8c

Days to resolution: 249

Markets in event: 10

Final Rationale

My independent estimate of 24.7% rested on strong leading indicators (initial claims at 189K, continuing claims at 2-year low) and CBO/FOMC forecasts of 4.4-4.6% for 2026, well below the 5% threshold. I PARTIALLY ACCEPT the critique that the question likely resolves on peak 2026 unemployment, and that November's reported 4.5% is only 0.5pp from threshold — this widens the path to YES since any single month above 5% triggers resolution. I ACCEPT that my three sub-questions were highly correlated and gave false precision, and that information gaps (April BLS, post-Iran-war forecasts) warrant wider uncertainty. I REJECT full convergence to the 33% market price because the depth of NO-side conviction at $0.44 and record-low claims data provide genuine evidence against a near-term spike, but I move to 29% to reflect the legitimate reasoning corrections.

Price vs. prediction

How the market moved after the agent's call. Solid blue line is the actual market price; dashed line is the agent's static forecast at $t{=}0$.

$n = 58$ follow-ups

Re-scan Context

This market has been scanned before. Previous predictions:

Date	Predicted	Market Price	Confidence
2026-04-04	37%	37%	50%

1. Planner Opus · Fermi decomposition + tool plan

Tools chosen: kalshi_data kalshi_orderbook fred_data article_search web_search code_execution wikipedia_lookup

Sub-questions (Fermi decomposition)

Is the U-3 unemployment rate currently on an upward trend that would bring it above 5% before end of 2026? (w=0.35)
Will a recession or significant economic downturn occur in 2026 severe enough to push unemployment above 5%? (w=0.35)
Based on historical base rates and current trajectory, will unemployment rise by at least 1+ percentage points from current levels to exceed 5% within 8 months? (w=0.30)

Planner reasoning

This question asks whether the maximum U-3 unemployment rate reading in 2026 will exceed 5% before year-end (resolves Jan 8, 2027, covering all of 2026). Key analytical steps: 1. **Current baseline**: As of early May 2026, unemployment is approximately 4.2-4.4%. The threshold requires a rise of ~0.6-0.8 percentage points from current levels. 2. **Temporal dynamics**: With 8 months remaining (May-Dec 2026), the question is whether a sufficient shock occurs AND propagates fast enough. Historical recessions show unemployment can rise quickly (COVID: 11pp in 2 months; GFC: 5.5pp over 24 months). A moderate recession starting now could easily push unemployment past 5% within 8 months. 3. **Reference classes**: The Sahm Rule and other leading indicators matter here. If a recession is already underway or imminent, the probability rises sharply. If no recession, staying below 5% is highly likely given the Fed's dual mandate. 4. **Causal mechanisms**: Tariff-induced slowdowns, consumer confidence declines, financial market stress, and Fed policy tightening are potential triggers. The 2025 tariff escalation is a key risk factor that may have already been affecting the labor market. 5. **Key uncertainty**: Whether current economic headwinds (from 2025 tariff shocks) have already triggered a recessionary labor market dynamic, versus whether the Fed has room to cut rates and stabilize unemployment below 5%. 6. **Market prices** on Kalshi will serve as an important anchor, reflecting aggregate forecaster views on this specific threshold.

2. Research Parallel tool calls · legacy v2

Tool	Status	Time	Items
kalshi_data	OK	1.9s	-
kalshi_orderbook	OK	0.3s	-
fred_data	OK	0.9s	-
article_search	OK	0.7s	-
web_search	OK	35.8s	-
code_execution	OK	0.0s	-
wikipedia_lookup	OK	1.6s	-

3. Evidence Synthesis Sonnet · 21 items · legacy v2

#	Strength	Credibility	Direction	Source	Claim	Priced In
1	STRONG	95	↓ DOWN	fred_data	U-3 unemployment rate was 4.3% as of March 2026, with only a +0.1pp YoY change, suggesting a very gradual upward drift rather than a sharp trend.	Yes
2	MODERATE	88	NEUTRAL	web_search	February 2026 unemployment rose to 4.4% (above 4.3% January), near November's 4-year high of 4.5%; the trend is modestly upward but not accelerating.	Yes
3	STRONG	95	↓ DOWN	fred_data	Initial jobless claims for the week ending April 25, 2026 fell to 189,000 — the lowest since 1969 — suggesting the labor market remains robust with no imminent acceleration in layoffs.	No
4	MODERATE	87	NEUTRAL	web_search	State-level data as of April 22, 2026 shows national unemployment at 4.4%, up 0.2pp from a year earlier; 13 states and DC had YoY increases, consistent with a mild upward drift.	Yes
5	MODERATE	90	NEUTRAL	fred_data	Nonfarm payrolls increased by only 152,000 YoY as of March 2026, a slowdown suggesting some labor market softening but not a collapse.	Yes
6	STRONG	88	↓ DOWN	web_search	CBO's February 2026 projections forecast unemployment at 4.6% for 2026 — below 5% — suggesting mainstream forecasts do not anticipate exceeding the threshold.	Yes
7	STRONG	90	↓ DOWN	web_search	FOMC March 2026 median projection for unemployment is approximately 4.4%, well below the 5% threshold, reinforcing that Fed policymakers do not foresee a threshold breach.	Yes
8	STRONG	82	↑ UP	article_search	A war involving Iran began on February 28, 2026, causing oil prices to spike to $119/barrel and raising mortgage rates; this geopolitical shock creates downside risk to U.S. growth via higher energy costs.	Yes
9	MODERATE	75	↓ DOWN	article_search	The Iran-U.S. war ended in a ceasefire after Iran closed the Strait of Hormuz; the conflict lasted roughly 6 weeks, limiting but not eliminating its recessionary impact on the U.S.	Yes
10	MODERATE	78	↑ UP	article_search	European economic and consumer confidence plummeted in March 2026 due to the Iran war, with employment expectations under pressure — a leading indicator of potential global demand slowdown.	Yes
11	MODERATE	85	↑ UP	article_search	The Fed has paused rate cuts three times in 2026, keeping rates at 3.50%-3.75%; this removes a key policy cushion and could amplify labor market weakness if a downturn occurs.	Yes
12	WEAK	55	↓ DOWN	article_search	A former Trump White House economist noted U.S. economic resilience and the difficulty of predicting recession timing, suggesting no consensus that recession is imminent.	Yes
13	WEAK	60	NEUTRAL	article_search	Unemployment insurance systems in many U.S. states are described as inadequate for a recession, which could amplify a downturn's economic impact but does not itself signal recession is occurring.	Yes
14	MODERATE	72	↓ DOWN	article_search	China secured a trade war truce with the U.S. by weaponizing rare earth minerals in 2025-2026; this trade conflict resolution reduces one major recessionary risk vector.	Yes
15	STRONG	80	NEUTRAL	code_execution	To exceed 5% by end of 2026, unemployment needs to rise ~0.8pp in 8 months (0.1pp/month); historical recession pace averages ~1.5pp/month, making it mathematically feasible if recession occurs.	Yes
16	MODERATE	78	↓ DOWN	code_execution	Historical non-COVID recessions rose at 0.05-0.23pp/month; exceeding 5% without a full recession would require the upper end of non-recessionary drift sustained for 8 months.	Yes
17	STRONG	93	↓ DOWN	fred_data	Continuing jobless claims are at the lowest in two years (1,785,000) as of late April 2026, indicating little current recessionary momentum in the labor market.	No
18	STRONG	82	↓ DOWN	kalshi_data	Kalshi market prices the 'Above 5%' outcome at 33.4%, with a slight downward trend over 7 and 30 days, reflecting aggregate forecaster skepticism that unemployment will breach 5%.	Yes
19	MODERATE	75	↓ DOWN	kalshi_orderbook	The orderbook shows high liquidity (depth=16,380) with the no_bid at $0.44, indicating strong market conviction toward the 'No' side (unemployment stays below 5%).	Yes
20	WEAK	60	NEUTRAL	kalshi_data	Kalshi's 'Above 9% before 2030' market is priced at 44%, suggesting significant tail risk over a 4-year window but not necessarily concentrated in 2026.	Yes
21	MODERATE	88	↓ DOWN	fred_data	The Fed Funds rate is at 3.64% (YoY -0.69pp), giving the Fed meaningful room to cut rates aggressively if unemployment starts rising sharply, providing a policy buffer.	Yes

Information Gaps

April 2026 BLS Employment Situation report (due May 8, 2026) not yet available — this is the most critical near-term data point for trend assessment
No direct evidence on Sahm Rule indicator current reading or whether it has been triggered
No data on Q1 2026 GDP growth rate, which would confirm or deny recessionary conditions
No direct evidence on the impact of 2025 tariff escalation on U.S. labor market (layoffs, hiring freezes in affected sectors)
No information on current consumer confidence indices (e.g., Conference Board, University of Michigan)
No data on corporate earnings guidance or announced layoff plans in major sectors for Q2-Q3 2026
Unclear duration and economic impact of the Iran war on U.S. energy costs and supply chains beyond oil price spike
No recession probability forecasts from major banks (Goldman Sachs, JPMorgan) with post-Iran-war updates

Key Uncertainties

Whether the Iran war's oil price shock (up to $119/barrel) has already passed through to consumer prices and hiring decisions in ways not yet visible in labor data
Whether the Fed's rate pause (3 consecutive holds) reflects genuine inflation concerns that constrain their ability to cut rates if unemployment rises
Timing and magnitude of any tariff-related layoffs in manufacturing and trade-exposed sectors
Whether current leading indicators (low initial claims, low continuing claims) will reverse sharply in May-July 2026
Probability and severity of a U.S. recession in 2026 — the critical binary outcome that determines whether 5% is breached
Whether the China trade war truce is durable or could re-escalate, adding additional downside risk

Market Context

Price: 0.334 · 7d: -0.008 · 30d: -0.016
Volume: HIGH — 2,416 avg daily contracts is substantial for a prediction market · Depth: HIGH — depth of 16,380 with no_bid at $0.44 indicates strong NO-side conviction
Reliability: HIGH — large volume, high depth, 119 days of data; market has wide historical range (28%-63%) suggesting significant uncertainty has existed over time, but current pricing reflects cooling recession fears

4. Independent Forecaster Opus · No market price · legacy v2 Did NOT see market price

Independent estimate: 25% (58% confidence)

base rate

23%

60% conf

evidence driven

24%

60% conf

contrarian

27%

55% conf

Spread

4pp

Low disagreement

sq1: Is U-3 currently on an upward trend that would bring it above 5% before end of 2026?

22%

Current trend shows ~4.3-4.4% with only +0.1-0.2pp YoY drift. Initial claims at 189K (lowest since 1969) and continuing claims at 2-year low strongly contradict an accelerating trend. CBO forecasts 4.6% and FOMC median 4.4% for 2026. Base rate for unemployment rising 0.7+pp in 8 months absent recession is low (~15%). Modest upward drift exists but at current pace (~0.02pp/month) would only reach ~4.6% by year-end. Push down for strong claims data. Estimate: 22%.

sq2: Will a recession or downturn occur severe enough to push unemployment above 5%?

25%

Base rate of recession in any given year is ~15-20%. Iran war oil shock to $119/bbl is a meaningful negative shock, but ceasefire after 6 weeks limits damage. Historically, oil price spikes >50% have preceded recessions ~50% of the time, but this one was relatively brief. Fed has paused but has 350bp+ of room to cut. China trade truce reduces tail risk. Even if mild recession occurs, it might not push unemployment >5% by Dec 2026 given lag. Probability of recession severe enough AND fast enough: ~25%.

sq3: Will unemployment rise 0.7+pp from current levels to exceed 5% within 8 months?

20%

Mechanically requires ~0.09pp/month sustained — feasible only with recession or significant labor market deterioration. Strong leading indicators (low claims, low continuing claims) argue against imminent acceleration. Fed has policy room. Historical non-recession periods rarely see 0.7pp rises in 8 months. Conditional on recession (~25% chance), probability of >5% by Dec is ~60%. Conditional on no recession, ~5-8%. Weighted: 0.25*0.6 + 0.75*0.07 ≈ 20%.

Show full reasoning trace

base rate: Historical base rate: unemployment rising 0.7+pp in 8 months occurs roughly 15-20% of years, almost always during recessions. Recession probability in any given year ~15-20%. CBO forecast 4.6% and FOMC 4.4% for 2026 both well below threshold.

evidence updates: DOWN: Initial claims at 189K (lowest since 1969) and continuing claims at 2-year low strongly indicate no labor market deterioration in progress (-10pp from naive base). DOWN: Fed has 350bp room to cut (-3pp). UP: Iran war oil shock to $119/bbl creates real recession risk (+5pp). UP: Fed paused 3x suggesting inflation constraint (+2pp). NEUTRAL: Trend is mildly upward but decelerating.

combination method: Weighted average across three correlated sub-questions, all reflecting the same underlying question of whether labor market deteriorates sharply. No independent diversification benefit.

final: Starting from 4.3-4.4% with strong leading indicators (record-low claims) and only 8 months remaining, breaching 5% requires either recession or unprecedented non-recessionary acceleration. Despite Iran war shock, current data does not show recessionary momentum. Estimate: 23%.

ensemble: {'ensemble_count': 3, 'perspectives': {'base_rate': {'p_yes': 0.23, 'confidence': 0.6}, 'evidence_driven': {'p_yes': 0.24, 'confidence': 0.6}, 'contrarian': {'p_yes': 0.27, 'confidence': 0.55}}, 'spread': 0.04, 'individual_p_yes': {'base_rate': 0.23, 'evidence_driven': 0.24, 'contrarian': 0.27}}

5. Devil's Advocate Sonnet · legacy v2 Market price revealed

Missing Info Reasoning Flaw

Challenges

The question asks 'how high will unemployment GET before 2027' which likely includes ALL of 2026, not just the remaining 8 months. The forecaster should check whether unemployment already touched 5% earlier in 2026 (November is noted as 'near 4-year high of 4.5%' but the forecaster doesn't fully verify the YTD peak). If the question resolves on peak unemployment at any point in 2026, prior months matter.
The forecaster cites February 2026 unemployment at 4.4% with 'November's 4-year high of 4.5%' but treats current level as 4.3-4.4%. There's ambiguity about whether the recent peak is closer to threshold than acknowledged — only 0.5-0.6pp away, not 0.7pp.
Sub-question combination is described as 'weighted average across three correlated sub-questions' but sq1, sq2, and sq3 are essentially asking the same question three different ways. Averaging correlated estimates that all point to ~22-25% gives false precision; the actual uncertainty should arguably be wider given missing April BLS data and unclear Iran war economic impact.
The forecaster acknowledges significant information gaps (April BLS report due May 8, no Sahm Rule reading, no Q1 GDP, no post-Iran-war recession forecasts) but doesn't widen uncertainty to reflect these gaps. Confidence of 0.583 seems high given these acknowledged unknowns.
Historical base rate calculation conflates 'recession in any given year' (~15-20%) with 'recession severe enough to push U-3 above 5% within 8 months from 4.3-4.4%.' Given U-3 is already elevated and trending up modestly, the conditional probability given even a mild downturn is higher than base rate suggests.

Suggested adjustment: +4pp

Modest upward adjustment warranted because: (1) the question may include peak unemployment over all of 2026, and November's reported 4.5% reading is closer to threshold than the forecaster's framing suggests; (2) acknowledged information gaps (April BLS, Sahm Rule, post-war forecasts) should widen uncertainty rather than leaving point estimate unchanged; (3) the three sub-questions are highly correlated, so the convergence around 22-25% is not independent confirmation. This is not adjustment toward market — it's correction for under-weighted uncertainty.

Market comparison: Forecaster: 24.7%, Market: 33.4%. Market is liquid (high volume, high depth) and has had wide historical range. The 8.7pp divergence may be partly justified by the forecaster's strong recent claims data, but the forecaster's dismissal of uncertainty around missing data and ambiguity about peak-vs-current-level resolution is a genuine reasoning gap. A modest move toward 28-29% is justified by reasoning flaws, not market conformity.

6. Aggregator Opus · final calibrated forecast

29% (50% confidence)