← Back to scans

How high will unemployment get before 2027?

KXU3MAX-27-5 · Economics · 2026-04-04
37%
Multi-Step
37%
Market Price
-0.4%
Edge
50%
Confidence
Final Rationale
Starting from 4.3% unemployment in March 2026, reaching 5.0% requires a 0.7pp rise in ~9 months. My corrected estimate of ~35% was adjusted upward from 25% after accepting critiques about mechanical dilution from sq3 weighting, underweighting the $119/barrel oil shock (which historically preceded recessions producing exactly this magnitude of unemployment increase), and the weak two-month payroll average (~22.5K/month). The market at 40.1% with high liquidity deserves deference, but I remain slightly below it because the most recent high-frequency data — initial claims at 202K (lowest in months) and March payrolls at +178K — are genuinely strong leading indicators that the labor market hasn't yet cracked. My 3pp deviation from market reflects this specific recent data strength.
What Would Change My Mind
April or May payrolls coming in below +100K or negative, initial claims rising above 250K sustained for 2+ weeks, a further escalation in the Iran conflict pushing oil above $140/barrel, or announcement of broad tariff implementation with immediate effect would push me above 50%. Conversely, a ceasefire with Iran driving oil below $90, or two consecutive months of +200K payrolls, would push me below 25%.
1. Planning Agent Sonnet
binary economics kalshi_data kalshi_orderbook fred_data article_search web_search code_execution wikipedia_lookup
Sub-questions
IDQuestionWeight
sq1 Will the U-3 unemployment rate rise above 5% at any point in 2026? 50%
sq2 Is the U.S. economy currently on a trajectory toward a recession or significant labor market deterioration in 2026? 30%
sq3 Has the U-3 unemployment rate already exceeded 5% in early 2026 (January–March)? 20%
Reasoning Approach
This question asks whether U-3 unemployment will exceed 5% at any point in 2026. The key distinction is that this is a MAX question — it resolves YES if unemployment breaches 5% even once before 2027, not just at year-end. The current unemployment rate (early 2026) is the critical starting point. As of late 2025, unemployment was hovering around 4.1-4.2%. To breach 5% by December 2026, unemployment would need to rise ~0.8-0.9 percentage points over ~9 months — roughly 0.09pp/month. Historical context matters: in non-pandemic recessions, unemployment typically rises 0.1-0.2pp/month during deterioration phases. Key causal factors include: (1) tariff impacts and trade policy uncertainty under Trump administration, (2) federal workforce reductions (DOGE), (3) tightening financial conditions, (4) leading indicators like jobless claims and JOLTS. The prediction market price (Kalshi) will be highly informative as it aggregates sophisticated forecasters. The reference class for 'unemployment rising 0.8pp in 9 months from ~4.2%' should be examined — this requires meaningful recession conditions but not a catastrophic shock. The question resolves based on any BLS monthly release showing U-3 >= 5.0%, so I need to assess recession probability and the pace of any deterioration.
2. Research Execution No LLM
ToolStatusTime
kalshi_data OK 2.97s
kalshi_orderbook OK 0.17s
fred_data OK 0.86s
article_search OK 0.62s
web_search OK 47.61s
code_execution FAILED 0.01s
wikipedia_lookup OK 1.54s
Errors: code_execution: __import__ not found
3. Evidence Synthesis Sonnet · 20 items
#StrengthDirectionSourceClaimPriced In
1 STRONG ↓ DOWN web_search BLS reported U-3 unemployment at 4.3% in March 2026 (released April 3, 2026), unchanged from February, confirming unemployment has NOT exceeded 5% in Q1 2026. Yes
2 STRONG ↓ DOWN fred_data FRED data confirms UNRATE at 4.3% as of March 2026, with a YoY increase of only +0.1 percentage points, well below the 5% threshold. Yes
3 STRONG ↓ DOWN web_search March 2026 payrolls rose 178,000 — a strong reversal from February's 133,000 decline — beating the Dow Jones consensus estimate of 59,000, suggesting labor market resilience rather than collapse. Yes
4 MODERATE ↑ UP web_search Labor force participation rate fell to 61.9% (lowest since fall 2021), suggesting the stable unemployment rate is partly driven by labor force shrinkage (~400,000 drop in March) rather than pure job gains — a hidden weakness. No
5 STRONG ↑ UP web_search February 2026 saw a 133,000 decline in nonfarm payrolls (revised down from initial report), indicating significant volatility and at least one month of notable labor market weakness in early 2026. Yes
6 MODERATE ↑ UP web_search U-6 (broader unemployment including discouraged workers and involuntary part-time) rose to 8% in March 2026, suggesting underlying labor market slack beyond what U-3 shows. No
7 STRONG ↓ DOWN fred_data Initial jobless claims (ICSA) were 202,000 as of late March 2026, up only 1,000 YoY — still historically low, not signaling a surge in layoffs. Yes
8 STRONG ↓ DOWN fred_data JOLTS job openings at 6,882,000 in February 2026, down 70,000 YoY — modest decline suggesting slight cooling in labor demand but not a sharp deterioration. Yes
9 STRONG ↑ UP fred_data Total nonfarm payrolls grew by only 152,000 YoY as of March 2026, indicating very slow annual job creation — a potential sign of structural weakening. Yes
10 STRONG ↑ UP article_search A war with Iran began around February 28, 2026, pushing oil prices to $119.48/barrel by March 9, raising mortgage rates and creating significant economic uncertainty — a new negative shock to the U.S. economy. Yes
11 MODERATE ↑ UP article_search The ECB warned the Iran war has made the economic outlook 'significantly more uncertain,' citing upside inflation risks and downside growth risks — global spillover effects onto U.S. economy are likely. Yes
12 MODERATE ↑ UP article_search The Federal Reserve paused rate cuts in both January and March 2026 meetings, with inflation stalling and unemployment rising — policy is not providing stimulus to buffer labor market deterioration. Yes
13 MODERATE ↑ UP article_search Experts warn the U.S. unemployment insurance system is unprepared for a recession, with state benefits falling far short of wages — if a recession hits, its labor market impact could be amplified. No
14 MODERATE ↑ UP article_search Federal government employment continued to decline in March 2026 per BLS data, consistent with ongoing DOGE-related federal workforce reductions adding to labor market headwinds. Yes
15 MODERATE ↑ UP article_search Canada lost 84,000 jobs in February 2026 with unemployment at 6.7% — a deteriorating labor market in the U.S.'s largest trading partner signals regional economic stress and potential spillover. Yes
16 STRONG ↑ UP web_search Wage growth slowed to 3.5% YoY in March 2026 — the lowest since May 2021 — suggesting cooling labor demand and reduced wage pressure, consistent with labor market softening. Yes
17 STRONG ↑ UP web_search Long-term unemployment (27+ weeks) rose by 322,000 over the year to 1.8 million by March 2026 — a significant increase suggesting structural labor market deterioration underway. No
18 STRONG — NEUTRAL kalshi_data Kalshi prediction market prices the 'above 5%' unemployment scenario at 40.1% as of early April 2026, with high liquidity (avg 2,871 contracts/day) and a 7-day decline of 6.2 percentage points following the strong March jobs report. Yes
19 STRONG ↓ DOWN kalshi_orderbook Orderbook shows NO bid at $0.51 vs YES bid at $0.04 with depth of 11,215 — very high liquidity on the NO side, suggesting sophisticated market participants lean toward unemployment staying below 5%. Yes
20 MODERATE ↑ UP kalshi_data The related 'above 9% before 2030' market is priced at 52% — implying a meaningful probability of a severe recession eventually, but not necessarily in the 2026 timeframe. Yes
Information Gaps
  • No data on April 2026 or later unemployment readings — the most critical remaining months for whether 5% is breached
  • No quantitative estimates of Iran war's direct impact on U.S. GDP or employment from economic models or forecasters
  • No data on magnitude of DOGE-related federal workforce reductions and their pass-through to U-3 unemployment
  • No recession probability forecasts from major banks (Goldman Sachs, JPMorgan, etc.) for 2026
  • No data on tariff implementation timeline or scope under the Trump administration in 2026
  • No data on consumer confidence or spending trends in Q1 2026 that would signal demand-side recession risk
  • No information on whether the February payroll decline of -133,000 was revised due to data anomaly or genuine contraction
  • Absence of leading indicator composite data (e.g., Conference Board LEI) for early 2026
Key Uncertainties
  • Duration and escalation of the Iran war — a prolonged conflict with higher oil prices and potential ground troops could significantly worsen the economic outlook
  • Whether the February 2026 payroll decline (-133,000) represents an isolated shock or the start of a trend of monthly job losses
  • Fed policy response: will the Fed cut rates if unemployment accelerates, and how quickly could such cuts buffer the labor market?
  • Pace of federal workforce reductions (DOGE) — if layoffs accelerate in Q2-Q3 2026, this could push unemployment meaningfully higher
  • Whether the labor force participation decline (shrinking denominator keeping U-3 stable) masks more severe underlying labor market weakness
  • Tariff policy uncertainty under Trump and its impact on manufacturing and trade-sensitive employment sectors
  • Whether the strong March payroll number (+178,000) is sustained in subsequent months or reverses again
Market Context
Price: 0.401 · 7d: -0.062 · 30d: +0.041
Volume: HIGH — 2,871 avg daily contracts is substantial for a prediction market · Depth: HIGH — depth of 11,215 with heavy NO-side liquidity at $0.51
Reliability: HIGH — high volume, deep orderbook, 89 days of data; recent price drop likely reflects strong March jobs report; market has ranged 28-63% over its life indicating genuine uncertainty
4. Independent Forecaster Opus · No market price Did NOT see market price
Independent estimate: 25% (medium confidence)
sq1: Will the U-3 unemployment rate rise above 5% at any point in 2026?
25%
Base rate: Starting from 4.3% in March 2026, unemployment needs to rise 0.7+ percentage points in ~9 remaining months. Historical base rate for such a move: In non-recession years, unemployment rarely rises this much. In recession years (2001, 2008), unemployment rose ~0.5-1.0pp over 9-month windows. The probability of a recession in 2026 is elevated but not dominant. Key evidence: (1) March payrolls at +178K beating estimates significantly pushes DOWN ~5% — suggests labor market still generating jobs. (2) Initial claims at 202K are historically very low, pushing DOWN ~3% — no surge in layoffs visible. (3) JOLTS at 6.88M, only modest decline, pushes DOWN ~2%. Against these: (1) Iran war and oil at $119/barrel pushes UP ~5% — a significant negative supply shock. (2) February payroll decline of -133K shows volatility and potential for sharp monthly drops, pushes UP ~3%. (3) Long-term unemployment rising 322K YoY pushes UP ~3% — structural deterioration. (4) Labor force participation at 61.9% (lowest since 2021) pushes UP ~2% — hidden weakness. (5) DOGE federal workforce reductions ongoing, pushes UP ~2%. (6) Wage growth slowing to 3.5% pushes UP ~1%. Net assessment: From a starting point of ~15% base probability (historical frequency of 0.7pp rise in 9 months from ~4.3%), the Iran war shock and various headwinds push this up meaningfully, but the strong March payroll report and low claims push back down. I estimate ~25%.
sq2: Is the U.S. economy currently on a trajectory toward a recession or significant labor market deterioration in 2026?
40%
Base rate: Unconditional recession probability in any given year is roughly 15-20%. Current conditions are notably worse than average due to: (1) Active war with Iran pushing oil to $119/barrel — STRONG UP ~10%. Oil shocks preceded recessions in 1973, 1979, 1990, 2008. (2) Fed pausing rate cuts with inflation stalling — monetary policy not accommodative, pushes UP ~3%. (3) Long-term unemployment surge of 322K — historically a leading/coincident indicator of recessions, pushes UP ~5%. (4) Canada losing 84K jobs — regional spillover, pushes UP ~2%. (5) DOGE federal workforce reductions — fiscal drag, pushes UP ~2%. (6) Wage growth at lowest since May 2021 — cooling demand, pushes UP ~2%. Against this: (1) March payrolls strong at +178K — economy still creating jobs, pushes DOWN ~5%. (2) Claims at 202K very low — pushes DOWN ~3%. Starting from ~18% base, net adjustments push to approximately 40%.
sq3: Has the U-3 unemployment rate already exceeded 5% in early 2026 (January–March)?
1%
BLS has already reported U-3 at 4.3% for March 2026, and January-February were similar. This is essentially resolved NO with near certainty. The only uncertainty would be a future benchmark revision, but BLS monthly releases are the resolution source. P(YES) = 0.01.
Show full reasoning trace
base rate: Starting from U-3 at 4.3% in March 2026, need 0.7+pp rise. Historical frequency of such rises over 9 months: roughly 15-20% in any given period, higher (~35-45%) conditional on recession. Unconditional recession probability for 2026 elevated to ~35-40% given Iran war, DOGE, tariffs.
evidence updates: Strong March payrolls (+178K) and very low initial claims (202K) are the strongest DOWN signals. Iran war with oil at $119/barrel is the strongest UP signal. Long-term unemployment surge and falling labor force participation provide moderate UP pressure. Net: elevated but not dominant probability.
combination method: sq1 directly answers the question with 50% weight. sq2 informs recession trajectory (30% weight). sq3 confirms no breach yet (20% weight). Weighted average gives 0.247, which aligns with my direct sq1 estimate of 0.25.
final: The U.S. labor market as of early April 2026 shows mixed signals — resilient job creation but underlying structural weakening, compounded by a major geopolitical shock (Iran war). The 0.7pp gap from current 4.3% to 5.0% threshold is substantial but not insurmountable if conditions deteriorate through the remaining 9 months. The strong March report provides a meaningful buffer. I estimate a 25% probability that U-3 will breach 5% before 2027.
5. Devil's Advocate Sonnet Market price revealed
Math Error Asymmetric Evidence Reasoning Flaw Missing Info
Challenges
  1. The combination method is flawed. sq3 (whether unemployment already exceeded 5% in Jan-March) is essentially a known NO and contributes nothing informative — giving it 20% weight mechanically dilutes the estimate. A proper combination would weight sq1 (the direct question) much more heavily, or simply use sq1 as the primary estimate with sq2 as a cross-check.
  2. The base rate calculation appears inconsistent. The forecaster states 'roughly 15-20% in any given period' for a 0.7pp rise over 9 months, but then says it's '35-45% conditional on recession' and estimates recession probability at ~35-40%. This would imply: P(>5%) = P(>5%|recession)*P(recession) + P(>5%|no recession)*P(no recession) ≈ 0.40*0.375 + 0.05*0.625 = 0.15 + 0.03 = 0.18, which is below the stated 25%. The evidence updates should push this higher, not the base rate itself.
  3. The forecaster notes February payrolls were revised to -133K but treats this as both a strong UP signal AND partially offset by March's +178K. However, the net of these two months (roughly +45K total over two months, or ~22.5K/month average) is actually quite weak job growth. The March 'beat' narrative may be overstated given the two-month average is below sustainable levels.
  4. The Iran war shock is being somewhat underweighted relative to historical precedent. Oil at $119/barrel represents roughly a 50-60% increase from pre-conflict levels. The 1990 oil shock (Iraq-Kuwait) saw unemployment rise from 5.4% to 6.8% over the following year. The 2008 oil spike to $147 coincided with unemployment rising from 5.0% to 7.3%. These are exactly the magnitude of moves that could push from 4.3% to 5%+, and the forecaster assigns only a net ~5% uplift from this factor.
  5. The forecaster lists 13 UP signals vs 6 DOWN signals but arrives at only 25% — this seems directionally inconsistent with the evidence balance. The DOWN signals are strong (March payrolls, low claims) but the sheer number and variety of UP signals (war, oil shock, DOGE cuts, falling participation, rising long-term unemployment, slowing wages, weak YoY job growth, Canada spillover) suggest more systemic risk than a 25% estimate captures.
  6. The information gaps are substantial and systematically bias toward underestimation. Missing consumer confidence data, recession probability estimates from major banks, tariff implementation details, and leading indicator composites would likely reveal additional downside risks given the current environment (war + potential tariffs + DOGE). Uncertainty should widen the estimate, and given the asymmetric nature of the question (unemployment tends to rise faster than it falls), this should push the probability up.
  7. The forecaster notes the 'above 9% before 2030' market is priced at 52% but doesn't adequately incorporate this. If there's a 52% chance unemployment exceeds 9% before 2030, the implied probability of exceeding 5% at some point in 2026-2029 is very high (likely 70%+), which means the probability of exceeding 5% in 2026 alone should be meaningfully above 25%, especially given the active geopolitical shock.
Suggested adjustment: +10pp
The combination method mechanically dilutes the estimate by weighting a near-certain NO sub-question at 20%. The evidence balance (13 UP vs 6 DOWN) is inconsistent with 25%. The oil shock historical precedent is underweighted — $119/barrel oil has historically been associated with recessions that easily produce 0.7pp unemployment increases. The two-month payroll average (~22.5K/month) is much weaker than the March headline suggests. Correcting these reasoning flaws pushes the estimate to approximately 35%.
Market comparison: Forecaster: 25%, Market: 40.1%. The 15pp divergence is partially justified by the forecaster's emphasis on strong March data and low claims, but the forecaster's reasoning has several identifiable flaws (mechanical dilution from sq3, underweighting oil shock precedent, inconsistent evidence weighting). After correcting for reasoning errors, the adjusted estimate of ~35% narrows the gap with the market. The remaining ~5pp divergence could be justified by the forecaster's emphasis on the most recent labor market data being genuinely strong.
6. Final Forecast Opus
37% (medium confidence)
Starting from 4.3% unemployment in March 2026, reaching 5.0% requires a 0.7pp rise in ~9 months. My corrected estimate of ~35% was adjusted upward from 25% after accepting critiques about mechanical dilution from sq3 weighting, underweighting the $119/barrel oil shock (which historically preceded recessions producing exactly this magnitude of unemployment increase), and the weak two-month payroll average (~22.5K/month). The market at 40.1% with high liquidity deserves deference, but I remain slightly below it because the most recent high-frequency data — initial claims at 202K (lowest in months) and March payrolls at +178K — are genuinely strong leading indicators that the labor market hasn't yet cracked. My 3pp deviation from market reflects this specific recent data strength.
Pipeline Timing
StepTime
Planning 26.14s
Research 47.77s
Synthesis 49.49s
Forecaster 31.13s
Critic 27.31s
Final 22.32s