Federal Reserve Bank of San Francisco introduces new labor market stress indicator

Researchers at the Federal Reserve Bank of San Francisco have constructed a new Labor Market Stress Indicator (LMSI). This indicator, based on state-level unemployment insurance claims, provides an early warning for economic slowdowns and recessions by tracking the geographical spread and depth of labor market stress.

Mapping Labor Market Stress Across States

The Labor Market Stress Indicator (LMSI) is primarily based on state-level unemployment insurance claims data, observable at weekly frequency.

By examining both the geographical spread and the depth of labor market stress buildup, the LMSI provides an early indicator designed to alert policymakers of potential economic slowdowns.

The indicator extends the logic of the Sahm-rule to state-level unemployment rates, classifying a state as experiencing 'accelerating unemployment' when its three-month moving average unemployment rate rises at least 0.5 percentage point above its 12-month lowest value.

Historically, whenever 30 or more states first experience accelerating unemployment, the national economy has always been in an NBER recession.

This dual perspective on breadth (number of states) and depth (labor-force share) helps distinguish genuine national downturns from more localized imbalances.

Complementing Traditional Recession Signals

The paper evaluates the LMSI as a recession indicator, comparing its performance with established tools like the national Sahm-rule, the Michaillat-Saez minimum indicator, and the Treasury yield curve.

In real time, the LMSI achieves an AUROC (Area Under the Receiver Operating Characteristic curve) of around 0.87, similar to the Sahm-rule and the Michaillat-Saez indicator.

While these indicators outperform the term spread in real time, the term spread shows stronger predictive power at longer horizons (one to two years).

The study suggests substantial gains from combining information across all indicators, with a logit model yielding an AUROC of about 0.96, highlighting their complementary strengths.

Sharpening the Early Warning System

This research significantly advances real-time economic monitoring by integrating granular, state-level data into recession forecasting.

The emphasis on geographical dispersion offers a crucial refinement over aggregate national indicators, providing a more nuanced understanding of labor market health.

While the LMSI's individual predictive power is strong, its true value lies in its ability to complement existing tools, forming a more robust and comprehensive early warning system for policymakers.