JOLTS Experimental State Estimates
The Job Openings and Labor Turnover Survey (JOLTS) data are based on a national sample of about 16,000 establishments.
These data are used by policymakers to better understand the current state of the U.S. economy, and by academics,
industry experts, economists, and others to understand the dynamic activity of businesses in the economy that lead to
aggregate employment changes. While the sample size is designed to support estimates for major industries at the national
and regional levels, the Bureau of Labor Statistics has been researching the possibility of leveraging the sample to
produce model assisted estimates at the state total nonfarm industry level. These estimates are experimental. We
encourage data users to review these estimates and provide input on both the technical aspects of the models and on the
usability of the resulting data.
The experimental data are developed using two composite models. These models are designated as:
- Composite Regional model: this model is used to produce the initial monthly state estimates
- Composite Synthetic model: this model is used to refine the state estimates as part of annual processing—by
incorporating data from the Quarterly Census of Employment and Wages (QCEW) that weren’t available when the initial
estimates were produced.
JOLTS Experimental State Data tables are provided for both models as a 3-month
moving average. The data will normally be presented as a combination of Composite Regional and Composite Synthetic data.
Table A includes data from the Composite Synthetic model from February 2001 to June 2018 with data from the Composite Regional
model from July 2018 to December 2018. Table B includes data only from the Composite Regional model from February 2001 to
December 2018. Comparing Table A and Table B allows the data user to examine the differences between final data and initial data.
BLS plans to update these data on a quarterly basis while assessing data user input on the models and on the utility of these
data. We encourage data users to provide input on these data at firstname.lastname@example.org.
Both models take the following form:
- J is a JOLTS data element: job openings; hires; total separations; quits; layoffs and discharges
- w1 is a weight between 0 and 1 and w2=1−w1
- D is a data-based estimate, as long as there are 5 or more sample units
- M is a modeled estimate, either using a regional approach or a synthetic approach (both are described below)
- Estimates of J are developed at the North American Industry Classification System (NAICS) supersector level.
- Once the NAICS supersector estimates are developed for the region, using either model, the sum of the estimated
state levels for the region are ratio adjusted to the regional level of the JOLTS data element.
- Within each state, the ratio adjusted values of J are then aggregated across NAICS supersectors to the total nonfarm level.
Composite Regional Model
- MR=regional_rate×state_employment. That is, the regional rate is multiplied by the state employment to get the state
level of the JOLTS data item.
Composite Synthetic Model
- MS=a synthetic model that distributes hires to all businesses that increase employment, and separations to all businesses
that lose employment, such that the hires and separations equal regional totals. All other JOLTS data elements are then modeled
relative to these data elements:
- Synthetic job openings are a function of the ratio of industry-regional job openings and hires. This ratio of published job openings
to hires is applied to model hires estimates to derive model job opening estimates.
- Synthetic quits and layoffs and discharges are a function of the relative percentage of the individual components of total separations
at the industry-regional level. The relative percentages of each component is applied to the model separations estimates to derive model
quits and layoffs and discharges.
A more complete description of both models can be found on the JOLTS Experimental State Estimates Methodology page.
Some common questions and answers can be found on our Frequently Asked Questions page.
Last Modified Date: May 24, 2019