New York Fed Researchers Find LinkedIn Hiring Data Can Improve Inflation Forecasts

New York Fed researchers said a new set of labor-market indicators built from LinkedIn activity could improve how economists track inflation in real time, with one measure standing out as especially useful for forecasting both consumer prices and wages.

In a Liberty Street Economics post published Friday, Elizabeth B. Li, Giorgio Topa and Jacob P. Weber evaluated three LinkedIn-based indicators of labor market tightness against more established gauges. Their main result was that LinkedIn’s Jobs per Jobseeker Index, or JPJI, outperformed the traditional vacancies-to-unemployment ratio in both nowcasting and forecasting price and wage inflation over the sample they studied.

Researchers tested whether faster hiring data add signal

The basic question behind the research is straightforward. If inflation pressure partly depends on how hard it is for employers to find workers, then a timely measure of labor market tightness can help policymakers and investors judge where wages and prices may head next. Standard gauges exist, but they often arrive with a lag. The researchers argued that private-sector platform data may offer a faster read on conditions between official releases.

The New York Fed post examined three LinkedIn measures: the Jobs per Application Index, the Jobs per Jobseeker Index, and the LinkedIn Hiring Rate. The first two are based on the number of what LinkedIn classifies as attractive job openings relative to either applications or active jobseekers. The hiring-rate measure tracks how many LinkedIn members add a new employer to their profile, scaled against the member base.

Those data have a timing advantage. The authors said LinkedIn measures for a given month can be available in the first week of the next month, well ahead of official labor-market datasets. By contrast, the job-openings figures used in the researchers’ benchmark vacancies-to-unemployment ratio come from the Bureau of Labor Statistics’ Job Openings and Labor Turnover Survey, which is released with a lag of roughly five weeks.

The research focused on whether the LinkedIn indicators merely move with familiar labor-market series or actually improve inflation tracking. To answer that, the authors compared them with common benchmarks, including the vacancies-to-unemployment ratio and the JOLTS hires rate. They then tested how well each indicator helped predict monthly Consumer Price Index inflation and quarterly Employment Cost Index wage inflation using rolling real-time regressions.

Jobs per Jobseeker Index stood out against standard benchmarks

Among the three LinkedIn indicators, the Jobs per Jobseeker Index delivered the strongest results. The researchers said it tracked the vacancies-to-unemployment ratio reasonably well, with a correlation of 0.733 over the period they studied, which ran from July 2019 through April 2026. That matters because it suggests the private dataset is capturing a similar underlying concept of labor-market tightness even though it is drawn from activity on a single hiring platform rather than the full labor market.

The LinkedIn Hiring Rate also compared favorably with an established official measure. Starting the comparison in early 2021, the authors found a 0.95 correlation between the normalized LinkedIn Hiring Rate and the JOLTS hires rate. Even so, the hiring-rate measure did not beat the JPJI in the inflation exercises. The researchers said the JPJI produced the lowest nowcasting error among the indicators they studied, outperforming the vacancies-to-unemployment ratio for both CPI and ECI inflation.

The same pattern largely held in the forecasting exercise. The researchers found that the JPJI again had a lower forecasting error than the vacancies-to-unemployment benchmark for both consumer-price and wage inflation. The LinkedIn Hiring Rate also performed better than the traditional benchmark in forecasting both inflation measures, but it still ranked behind the JPJI. By comparison, the Jobs per Application Index and the JOLTS hires rate produced weaker results across most of the tests.

That finding is notable because it points to a measure based on jobseekers, not just posted openings, as a better read on inflation pressure. In tight labor markets, the number of available workers relative to attractive openings may tell economists more than raw vacancy counts alone. When the pool of active seekers is small relative to openings, wage pressure can build more quickly as employers compete for workers, which can later feed into broader inflation.

Useful signal comes with limits on representativeness

The paper does not argue that LinkedIn data should replace official statistics. In fact, the authors explicitly cautioned that the measures are not necessarily representative of the U.S. economy as a whole because they capture only jobs posted, searched and accepted through LinkedIn’s platform. That qualification matters. LinkedIn skews toward certain industries, occupations and white-collar workers, so a platform-based indicator may miss parts of the labor market where hiring conditions differ materially.

Still, timeliness can be valuable even when coverage is imperfect. For policymakers, a faster indicator can help fill the gap between official releases, especially when inflation and labor-market dynamics are changing quickly. For markets, earlier clues about wage pressure can shape expectations for monetary policy, bond yields and rate-sensitive sectors. The authors’ conclusion was measured rather than sweeping: the JPJI and the LinkedIn Hiring Rate track existing labor-market-tightness gauges reasonably well, and in several cases they improve inflation nowcasts and forecasts relative to more familiar measures.

The work also fits a broader trend in economic analysis. Central banks and researchers increasingly test private and high-frequency datasets to supplement traditional government statistics, not because official data are dispensable, but because they are often slower and more heavily revised. What makes the New York Fed paper more useful than a simple data novelty story is that it benchmarks the LinkedIn indicators directly against standard measures and evaluates them with a clear error metric, root mean squared error, rather than relying on anecdotal correlations alone.

Readers should also keep the institutional context in mind. The analysis appeared on Liberty Street Economics, the New York Fed’s research blog, and the post notes that the views expressed are those of the authors and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Even with that caveat, the findings are likely to attract attention because they speak directly to one of the hardest problems in macroeconomic analysis: how to get a timely read on inflation pressure before the official data arrive.

For now, the strongest takeaway is narrow but practical. The researchers did not claim that every LinkedIn hiring metric improves inflation forecasting. They found that one measure, the Jobs per Jobseeker Index, was the most consistently helpful, while the LinkedIn Hiring Rate also added value in several cases. Their sample runs through April 2026, and the appeal of the indicators rests largely on the fact that they can be observed earlier in the month than comparable official labor-market measures.