Recent delays in the release of government labor market data have led researchers to seek alternative ways to measure labor market health. Thirteen out of the 24 variables typically used in the Kansas City Fed’s Labor Market Conditions Indicators (LMCI) come from the Bureau of Labor Statistics’ Employment Situation Report, which has faced delays. However, with 11 data series still available, a restricted version of the LMCI’s headline indicators—level of activity and momentum—remains usable.
According to researchers at the Federal Reserve Bank of Kansas City, both the official and restricted versions of the level of activity indicator have moved closely together over time. Before the pandemic, their correlation was very strong at 0.99. While there was some divergence between 2020 and 2021, this strong relationship resumed afterward. In recent months, both indicators have declined at similar rates and are now near their historical averages.
The momentum indicator built from only 11 input series also shows similar behavior to its unrestricted counterpart. The pre-pandemic correlation between these two measures was 0.95. The restricted momentum indicator excludes payroll growth and unemployment rate data, resulting in smaller swings during the pandemic but showing stronger responses to announced layoffs earlier this year. Both measures remain below zero but have recovered somewhat from earlier dips.
Researchers used this restricted LMCI to assess labor market conditions in September 2025. They found that “In September, our restricted level of activity declined from 0.06 to zero, which represents the indicator’s historical average.” They added: “Although the zero reading suggests that activity is no longer above its historical average, the pace of decline in the index is in line with recent month-to-month declines in both the official and restricted level of activity indicators.” Additionally, “In September, our restricted momentum indicator increased by 0.18, from −0.88 to −0.70.” This result is consistent with other recent changes observed in both official and restricted indicators.
Because these indicators are designed so that their historical average is zero and their standard deviation is one, current readings indicate that while labor market momentum remains below average, it is within typical ranges seen historically. Overall, researchers concluded: “these readings suggest that the labor market is continuing to cool at the slow but steady rate we have seen over the last two years, and activity remains near historical levels.”
The LMCI can also be used for real-time economic forecasts when official data sources are delayed or unavailable. Researchers noted that using a simple forecasting model based on previous work (Glover et al., 2021), “the forecast based on the restricted LMCI implies an unemployment rate for September 2025 of 4.4 percent,” only slightly higher than August’s official unemployment rate of 4.3 percent.
Similarly, drawing on another forecasting model (Lusompa & Mustre-del-Río, 2025), they found that “the forecast implies that the economy added an average of 60,000 jobs a month from July to September,” which marks a slight increase compared to an official three-month average of 29,000 jobs added from June to August.
Alternative estimates for payrolls for this period vary; for example, ADP’s latest release points to a three-month moving average of 23,000 jobs while Revelio Labs estimates it at 41,000 jobs. The differences among these estimates are not statistically significant due to large standard errors.
“In summary, when official data sources are delayed or unavailable, the LMCI can still provide a timely measure of the labor market’s health,” according to researchers at the Federal Reserve Bank of Kansas City. They added: “Additionally, this data-restricted LMCI can be used to forecast key labor market measures such as the unemployment rate or payroll growth.” The findings for September suggest little change in labor market conditions compared with August.
The authors note that while official statistics remain important for research purposes, “the LMCI’s long history and wide variety of inputs may make it a useful alternative for policymakers facing data delays.”
José Mustre-del-Río is an assistant vice president and economist at the bank; Johnson Oliyide is a research associate; Emily Pollard is an associate economist. The views expressed are those of the authors and do not necessarily reflect those of the Federal Reserve Bank of Kansas City or the Federal Reserve System.
