According to the most recent State Employment and Unemployment release from the Bureau of Labor Statistics, 13 states and the District of Columbia registered statistically significant unemployment rate increases over the year ending February 2026. Only two states — Indiana and Ohio — saw decreases. Thirty-five states had little change. The national rate sat at 4.4%, just 0.2 percentage points higher than a year earlier, but that aggregate masked sharp geographic divergence.
Here's what's actually happening: The labor market isn't deteriorating uniformly — it's deteriorating in specific places while staying frozen elsewhere. Delaware's unemployment rate climbed 1.1 percentage points year over year. Connecticut, Florida, and Minnesota each rose 1.0 point. Oklahoma rose 0.8, Maryland 0.7, and DC 0.6. On the employment side, year-over-year payroll losses were significant in three places: Maryland (-52,700 jobs), DC (-42,200, or -5.5%), and Iowa (-19,200). Only California (+120,500) and Nevada (+34,500) posted statistically significant gains.
Why it matters for you:
- Retention math depends on geography. In a deteriorating local market, your existing workers have fewer external options, which strengthens your retention position without requiring above-market compensation. The opposite holds in states where employment is growing. National compensation benchmarks treat the country as one labor market when it isn't. Workers in Delaware, Connecticut, or Minnesota face materially fewer outside opportunities than they did twelve months ago.
- Headcount risk concentrates regionally. If you operate in DC, Maryland, or Iowa, you're hiring into shrinking employment markets. The 5.5% decline in DC payrolls isn't a forecast — it's already happened. Hiring plans built on national assumptions about labor availability will miss in both directions: harder to fill roles in Texas and Nevada, easier in markets where unemployment is climbing. Workforce planning by national headline number gets the variance wrong.
- Multi-state comp policies are misallocating retention dollars. When a worker in Ohio (where unemployment fell 0.7 points) and a worker in Connecticut (where it rose 1.0 point) get identical raise budgets, you're overspending in one market and underdefending in the other. The fact that BLS classifies 35 states as statistically unchanged means most managers will assume their state matches the national story — even though the dispersion across the 50 states has widened materially over the past year.
Source: U.S. Bureau of Labor Statistics, State Employment and Unemployment (February 2026)
Watch this: With the national rate flat at 4.4% but the over-the-year state-level scoreboard running 14-to-2 negative, the next monthly release will reveal whether this is the leading edge of broader deterioration or contained regional weakness. The count of states tipping into "statistically significant increase" matters more than the headline number for anyone planning workforce decisions through the rest of 2026.
The contrarian play: While competitors negotiate retention using national unemployment data ("the market is tight at 4.4%"), pull your state-specific year-over-year change before any compensation conversation. In the 14 places where unemployment rose, your leverage is higher than the national average suggests. In Indiana and Ohio, the only two states where conditions improved, your leverage is lower. Treating geography as a primary input — not a footnote — separates managers who hold compensation costs from those who lose workers because the macro story didn't match the local one.