US layoff data from 1988 to present

What is WARN Database?

WARN Database standardizes labor market records that would otherwise sit scattered across dozens of state and federal filings. Its core product cleans up the mass layoff notices that large employers must file with state governments under the federal Worker Adjustment and Retraining Notification Act, and it also maintains a separate monthly employment series for US airlines sourced from the Department of Labor.

‍

What academic researchers should know about WARN Database labor market data

On Dewey, WARN Database publishes two datasets. The layoff notice dataset records individual mass layoff notices by state, with historical coverage from January 1, 1998 through the most recent month, for 49 states and territories. Coverage varies by state: Wyoming and Arkansas do not make WARN notices available, New Hampshire, Oklahoma, and New York have pending data awaiting FOIA responses, and states like Minnesota, New Jersey, and Pennsylvania only report at a monthly level rather than the exact notice date. The airline employment dataset reports monthly full and part time employment by carrier and by carrier size category, sourced from the Department of Labor's Bureau of Transportation Statistics, with coverage from January 1990 to the present.

‍

Why academic researchers choose WARN Database on Dewey

Building a dataset like this from scratch means requesting records from 50 different state agencies, each with its own format, its own level of detail, and its own idea of what counts as a mass layoff. WARN Database does that collection and standardization work already, through direct state pulls supplemented by FOIA requests, so researchers get a single cleaned dataset instead of fifty separate ones. Because notices are dated and geographically specific down to the state, researchers can study layoffs as they were announced rather than waiting for layoffs to show up months later in aggregate employment statistics. It pairs naturally with other Dewey partners for a fuller labor market picture: LinkUp for real time job postings at the same employers, BrightQuery for the private company financials and employment history behind a layoff, and People Data Labs for independent firmographic context.

‍

WARN Database academic research ideas and use cases

• Early signals of economic downturns. Because WARN notices are filed 60 days before layoffs take effect, state level notice counts can be tested as a leading indicator against official unemployment claims and payroll data.
• Private equity and corporate ownership changes. Layoff notices can be linked to firm level financial and ownership data to study whether layoffs cluster around leveraged buyouts, mergers, or changes in ownership structure.
• Industry and regional labor market shocks. State and date stamped layoff notices support research on how job losses in one industry or region spill over into local labor markets, housing demand, or small business formation.
• Airline industry employment cycles. Monthly employment by carrier and carrier size, going back to 1990, supports research on how airline employment responds to fuel price shocks, mergers such as the 2024 FedEx consolidation, or broader travel demand cycles.

‍