LinkedIn’s jobs-per-jobseeker index outperformed a standard labor market measure in estimating wage and price inflation over nearly seven years, Federal Reserve Bank of New York researchers said on Oct. 30.
The index produced smaller errors for both current inflation estimates and forecasts than the ratio of job openings to unemployed people, the authors found. Their Liberty Street Economics article, “Evaluating New LinkedIn Labor Market Tightness Indicators,” was published Oct. 30.
The authors examined LinkedIn and Bureau of Labor Statistics data from July 2019 through April 2026, comparing three LinkedIn indicators with established measures. They assessed how well each estimated inflation before official releases and predicted the next period’s inflation.
The researchers were Elizabeth B. Li, a research analyst, Giorgio Topa, an economic research advisor, and Jacob P. Weber, a research economist, all at the New York Fed.
They said their findings do not necessarily reflect the central bank’s position.
The authors found that LinkedIn’s Jobs per Jobseeker Index closely followed the official ratio of vacancies to unemployment. Their calculated correlation was 0.73, with higher positive values indicating that the measures moved more closely together.
The LinkedIn Hiring Rate also tracked the government’s hiring measure closely, the researchers found. Their correlation reached 0.95 when the comparison began in early 2021, excluding the sharp movements at the start of the pandemic.
The jobs-per-jobseeker index divides qualifying openings by the number of people applying for jobs on LinkedIn, the authors said. The hiring rate measures members adding a new employer to their profiles, while a third indicator divides qualifying openings by applications.
The authors assessed price inflation using the Bureau of Labor Statistics’ monthly Consumer Price Index and wage inflation using its quarterly Employment Cost Index. They compared prediction errors with those from the vacancies-to-unemployment ratio, which measures how many openings are available relative to unemployed people.
The hiring rate beat that benchmark for current price inflation estimates but performed worse for current wage inflation estimates, the authors found. It beat the benchmark for forecasts of both, although the jobs-per-jobseeker index performed better.
The applications-based index and the government hiring rate generally performed worse than the benchmark, the authors found. One exception was the applications-based index, which beat it when forecasting wage inflation.
LinkedIn’s monthly indicators could arrive in the following month’s first week, compared with about five weeks for the Bureau of Labor Statistics’ Job Openings and Labor Turnover Survey, the authors said. That timing could help policymakers estimate inflation before official figures become available.
The researchers said LinkedIn’s coverage may not represent the entire U.S. economy because it captures only jobs posted and applied for on the platform. Limited historical availability of two LinkedIn indicators also restricted the study’s starting date.
