Design, compare, and defend labor-market policy in one place.
Pick your audience and country, dial policy levers, and watch the country-calibrated Markov flows re-solve for u*, LFPR, and monthly churn in real time. Save named scenarios, compare up to three side-by-side, cite the evidence behind each lever, and export a printable brief for cabinet, board, or council.
Draft scenario
High UβE job-finding, high EβU separations, elevated NβE from a large hidden-worker pool. Beveridge-curve has shifted outward post-2020; prime-age men participation still below 2000s peak; long-term unemployment share cyclical.
Year 1: STC + wage insurance (cyclical). Year 2: childcare and DI reform (structural). Year 3: licensing reciprocity via interstate compacts.
- Expand short-time-compensation to raise EE / lower EU
- Portable benefits + occupational licensing reform to raise UE
- Childcare and DI reform to reactivate NβE
Policy levers for United States
Evidence & citations
Click any lever on the left to see its mechanism, implementation steps, lead agency, and academic citations. Every dial in this studio is defensible.
- Short-Time Compensation (STC / work-sharing) β Abraham-Houseman (2014); Cahuc-Kramarz-Nevoux (2021).
- Universal childcare tax credit + supply subsidy β Blau-Kahn (2013); Bick (2016); IMF WP 24/122.
- Portable benefits + national licensing reciprocity β Kleiner-Krueger (2013); Johnson-Kleiner (2020).
- SSDI vocational rehab expansion (Ticket-to-Work v2) β Autor-Duggan (2003, 2010); Maestas-Mullen-Strand (2013).
- Wage-insurance for displaced workers β Kletzer-Litan (2001); Hyman-Kovak-Leive-Naidu (2022).
Transition matrix β scenario vs baseline
| i \ j | E | U | N |
|---|---|---|---|
| E | 0.9632 | 0.0135 | 0.0233 |
| U | 0.3072 | 0.5114 | 0.1813 |
| N | 0.0621 | 0.0317 | 0.9063 |
| i \ j | E | U | N |
|---|---|---|---|
| E | 0.9632 +0.00pp | 0.0135 +0.00pp | 0.0233 +0.00pp |
| U | 0.3072 +0.00pp | 0.5114 +0.00pp | 0.1813 +0.00pp |
| N | 0.0621 +0.00pp | 0.0317 +0.00pp | 0.9063 +0.00pp |
Saved scenarios
Save the current scenario to start building a comparison set.
Three structural shocks reshaping the U.S. Markov chain
Predicted directional effects on the E/U/N monthly transition matrix and the induced steady-state stocks. Signs and magnitudes reflect the published empirical literature rather than free-form speculation. Every claim links back to a citation you can defend in a policy meeting.
Minimum wage rises (federal floor β $15 / state indexation)
Compresses the low-wage tail; small Uβ at the margin, larger NβE for second earners.
A binding federal minimum-wage increase or aggressive state indexation lifts the reservation wage in the bottom decile. In a search model, higher w raises UβE hazard for eligible workers but modestly cuts vacancy posting in the most exposed industries (food service, retail, home care, warehouse).
Modelled as a proportional log-odds bump scaled to 100% of the +$7.25β$15 central case.
Predictions recompute continuously as you drag: cell deltas are scaled by intensity, applied as log-odds bumps to the U.S. baseline P, then Ο = ΟPΜ is solved to steady-state. Sign and magnitude labels below still come from the published estimates β the slider only rescales their strength.
- UE: month 60, +8.73pp
- EU: month 25, +0.40pp
- NE: month 60, +2.38pp
- NU: month 60, -0.08pp
- u*: month 60, Ξ -0.16pp
- LFPR: month 60, Ξ +3.25pp
- ΞΈ: month 5, Ξ +1.077
Per-cell shape(t) = permanent Β· (1 β eβt/rise) + (1 β permanent) Β· eβ((t β peak)/halfLife)Β², rise = peak/2. Timing parameters reflect the literature: minimum-wage EU disemployment lags UE reactivation; the immigrant-influx UβE dip reverses within ~12β18 months (Foged-Peri); the crackdown EβN wave builds over 3β5 years as the retirement cohort clears without a replacement.
Each sweep re-solves the steady state Ο = ΟPΜ at 25 points across Β±30% of the slider range around the current setting, holding other sliders fixed. Faint horizontal lines mark the no-shock baseline; the vertical dashed line marks the current value. Slope is a rough local elasticity for briefing purposes β the response is mildly nonlinear because log-odds bumps are then re-normalised into row-stochastic P.
| Cell | Sign | Magnitude | Mechanism |
|---|---|---|---|
| pUE | + | moderate | Higher take-home wage raises the value of accepting an offer; reservation wage constraint less binding for LTU. Cengiz-Dube-Lindner-Zipperer (2019) find no significant employment loss but faster reemployment in bunching studies. |
| pNE | + | moderate | Second earners and prime-age reactivators cross the participation threshold. Especially strong for women 25β54 (Bailey-DiNardo-Stuart 2021). |
| pEU | + | small | Disemployment risk concentrates in tipped, sub-minimum, and teen segments (Neumark-Wascher). Aggregate EU flow moves by <5% relative to baseline in most credible estimates. |
| pUN | β | small | Discouraged-worker exits fall because search pays more; hazard from U into N declines. |
| pEE | β | small | Job-to-job churn in low-wage sectors slows slightly β the outside option compresses (Dube-Giuliano-Leonard 2019). |
- u*: u* β +0.05 to +0.15 pp (net small β faster job finding largely offsets marginal separations).
- LFPR: LFPR β +0.20 to +0.40 pp (dominant channel: NβE reactivation).
- ΞΈ (tightness): ΞΈ (vacancies/unemployed) β ~2β4% in exposed sectors, unchanged aggregate.
- Wages: Bottom-decile wages +8β12%; median +1β2% via spillovers up to 120% of new floor.
Winners: Low-wage incumbents, second earners, single parents, tipped workers in states banning sub-minimum.
Losers: Sub-minimum teens, small businesses in low-cost-of-living areas, gig/1099 workers below new floor.
Timeline: 6β12 months for UβE and NβE to lift; 12β24 months for the disemployment tail to show up in EU.
- Cengiz, Dube, Lindner, Zipperer (2019, QJE) β bunching design, no disemployment.
- Dube, Lester, Reich (2010, ReStat) β contiguous-county evidence.
- Neumark & Wascher (2007) β meta of adverse teen effects.
- Cadena (2014) β immigration + minimum wage interaction.
Effects concentrate at the bottom of the wage distribution; aggregate Markov flows move less than sector-specific ones. Monopsony intensity in local labor markets is the single largest source of dispersion in estimates.
Reading the three scenarios together: a minimum-wage rise + immigration crackdown + collapsing TFR compound into a regime where measured unemployment looks excellent () but the labor force itself is shrinking. The Markov chain is not broken β it is simply converging to a smaller, older, higher-wage but lower-output steady state. The right diagnostic is not ; it is in absolute terms and the dependency ratio . Policy makers reading only the headline number will systematically under-react to the fiscal and productivity squeeze arriving in the second half of the decade.
Predictions are directional and cite the peer-reviewed literature. Point-estimated magnitudes come from published elasticities and should be re-fit against country-specific gross-flow data before any statutory action. States E, U, N follow the E/U/N convention used elsewhere in this workbench.
Recommendations engine
Set target u* and LFPR for United States. The engine searches over lever directions and intensities and returns the smallest set whose combined effect on the stationary distribution best hits the target β with a rationale and a sensitivity note per lever.
- Baseline P is the SOC-weighted whole-economy monthly transition matrix, tilted by a small country offset (Kurzarbeit for DE, lifetime employment for JP, tight labor market for AU, etc.).
- Each lever applies a multiplicative log-odds bump of λ·intensity to its mapped cell (λ = 0.35 at intensity = 100%), then rows are renormalized.
- Steady-state u*, LFPR, and churn are the stationary distribution Ο = ΟPΜ (500 iterations). This is a comparative-statics tool, not a business-cycle forecast.
- Every lever cites peer-reviewed empirical work; open the evidence panel to see the source. Reject any scenario whose levers you cannot defend at the citation level.