
JOLTS Experimental State Estimates Methodology
The JOLTS sample of 16,000 establishments does not directly support the production of sample based state estimates.
However, state estimates have been produced by combining the available sample with modelbased estimates, and
smoothed by taking a 3month moving average. These data are experimental. As such, they have not been subject to the
same level of review as the current official JOLTS national and regional estimates. BLS is inviting data users to comment
on both the methodology used to produce these estimates and on the usefulness of these data. The eventual goal is to
produce and provide JOLTS statelevel estimates on a monthly basis.
These estimates consist of four major estimating models; the Composite Regional model (an unpublished
intermediate model), the Synthetic model (an unpublished intermediate model), the Composite Synthetic
model (published historical series through the most current benchmark year), and the Extended Composite Synthetic
model (published currentyear monthly series). The Composite Regional model uses JOLTS microdata, JOLTS regional published estimates,
and Current Employment Statistics (CES) employment data. The Composite Synthetic model uses JOLTS
microdata and Synthetic model estimates derived from monthly employment changes in microdata from the Quarterly Census of Employment and Wages
(QCEW), and JOLTS published regional data. The Extended Composite Synthetic extends the Composite
Synthetic estimates by ratioadjusting the Composite Synthetic by the ratio of the current Composite Regional model estimate to the Composite
Regional model estimate from one year ago.
The Extended Composite Synthetic model (and its major component—the Composite Regional model) is used to extend the Composite
Synthetic estimates because all of the inputs required by this model are available at the time monthly estimate are produced.
In contrast, the Composite Synthetic model (and its major component—the Synthetic model) can only be produced when the latest
QCEW data are available. The plan is to use Extended Composite Synthetic model estimates to extend the Composite Synthetic model
estimates during the annual JOLTS retabulation process. The extension of the Composite Synthetic model using current databased
Composite Regional model estimates will ensure that the Composite Synthetic model estimates reflect current economic trends.
The following outlines each model in a nontechnical summary format. Each model is summarized separately, and answers the following:
 What is the approach attempting to do?
 What data inputs are used in the approach?
 How does the approach attempt to use that data?
 What data outputs are produced by the approach?
 What limitations does the approach have?
 What more needs to be done?
Composite Regional Model
What Approach?
The Composite Regional approach calculates statelevel JOLTS estimates from JOLTS microdata using sample weights, and the
adjustments for nonresponse (NRAF). The Composite Regional estimate is then benchmarked to CES statesupersector employment
to produce statesupersector estimates. The JOLTS sample, by itself, cannot ensure a reasonably sized sample for each
statesupersector cell. The small JOLTS sample results in quite a number of statesupersector cells that lack enough data to
produce a reasonable estimate. To overcome this issue, the statelevel estimates derived directly from the JOLTS sample
are augmented using JOLTS regional estimates when the number of respondents is low (that is, less than 30). This approach is
known as a composite estimate which leverages the small JOLTS sample to the greatest extent possible and supplements that with
a modelbased estimate. Previous research has found that regional industry estimates are a good proxy at finer levels of
geographical detail. That is, one can make a good prediction of JOLTS estimates at the regionallevel using only national
industrylevel JOLTS rates. The assumption in this approach is that one can make a good prediction of JOLTS estimates at the
statelevel using only regional industrylevel JOLTS rates.
In this approach, the JOLTS microdatabased estimate is used, without model augmentation, in all statesupersector cells that have
30 or more respondents. The JOLTS regional estimate will be used, without a samplebased component, in all statesupersector cells
that have fewer than five respondents. In all statesupersector cells with 5–30 respondents an estimate is calculated that is a
composition of a weighted estimate of the microdatabased estimate and a weighted estimate of the JOLTS regional estimate. The
weight assigned to the JOLTS data in those cells is proportional the number of JOLTS respondents in the cell (weight=n∕30,
where n is the number of respondents).
What data inputs?
 All JOLTS microdata records
 All weights from JOLTS estimation (final weights that account for sampling weight, NRAF, aggcodes, etc.)
 JOLTS published regional rates estimates (regional JO, H, Q, LD, and TS rates)
 CES statesupersector employment
How are data used?
 All JOLTS microdata are weighted using final weights. A weighted estimate is made for each JOLTS respondent.
 Counts are made for each statesupersector cell.
 Each JOLTS respondent is paired with its regional rate estimate for all variables.
 Based on the count of respondents in the statesupersector cell the JOLTS respondent belongs to, a Composite Model Weight (CMW) is calculated.
 If the count is>30, then the CMW for the respondent data=1. The CMW for the regional estimate=0.
 If the count<5, then the CMW for the respondent data=0. The CMW for the regional estimate=1.
 If the count is 5–30, then the CMW for the respondent data=n∕30, where n is the number of respondents. The CMW for the regional estimate=1n∕30.
 The statelevel rate estimate is therefore the final weighted respondentbased JOLTS rate times the CMW added to the regional rate times the CMW,
benchmarked to CES statelevel estimate:
 FINAL ESTIMATE=CES STATE EMP×((final weight JOLTS rate×CMW)+(regional rate×CMW))
 To stabilize the estimate, the sum of state Composite Regional estimates within each region is then benchmarked to the published JOLTS regional estimates.
How are outputs produced?
 This model produces statelevel estimates of JO, H, Q, LD, and TS. These estimates provide estimates for the most current month of
estimates and can be produced during monthly JOLTS estimation production.
What are the limitations?
 JOLTS data are somewhat volatile at the national and regional levels due to the small sample size which in turn results in volatile state estimates.
 The Composite Regional estimates can vary substantially from Composite Synthetic estimates for states that exhibit seasonal employment patterns that
differ substantially from the JOLTS region to which they belong. For example, Alaska has a pronounced seasonal employment pattern that differs from the
West region in which it resides. Consequently, the Composite Regional estimates derived using West region JOLTS rates substantially understate the JOLTS
rates in that state.
What more is needed?
These estimates are based upon a model. There is, as of yet, no methodology in place that can produce an estimate of error for the estimates
the model produces. Research on a methodology to produce an error estimate is currently underway.
The Composite Regional supersector estimates are summed across state industry supersectors to the nonfarm level.
Synthetic Model
What approach?
The Synthetic model differs fundamentally from the Composite Regional model. The Synthetic approach does not use JOLTS microdata but rather it uses
data from the QCEW that have been linked longitudinally (Longitudinal Database—LDB), the QCEWLDB. The Synthetic model
attempts to convert QCEWLDB monthly employment change microdata into JOLTS job openings, hires, quits, layoffs and discharges, and total separations data.
What data inputs?
 All monthly employment changes for each record on the QCEWLDB
 JOLTS published regional estimates (regional JO, H, Q, LD, and TS)
How are data used?
 Every record on the QCEWLDB is classified as expanding, contracting, or stable based on monthly employment change.
 For expanding records, the amount of employment growth is converted to JOLTS hires. They are given no separations.
 For contracting records, the amount of employment decline is converted to JOLTS separations. They are given no hires.
 For stable records, no attribution of JOLTS hires or separations is made.
 The entire QCEWLDB is summarized to the US Census regional level.
 The QCEWLDB regional summary is ratio adjusted to the JOLTS published regional estimate for hires and total separations.
 For each region, the ratio of QCEWLDB based regional hires and total separations to JOLTS published hires and total separations is calculated
(RatioH for hires and RatioTS for total separations).
 Each record on the QCEWLDB within each US Census region will have their converted JOLTS data multiplied by RatioH and RatioTS, by region.
 For expanding records, the amount of employment growth is then: (JOLTS hires×RatioH). They remain with no separations.
 For contracting records, the amount of employment decline is then: (JOLTS separations×RatioTS). They remain with no hires.
 For stable records, they remain with no JOLTS hires or separations.
 To produce statelevel estimates, sum the regional hires×RatioH by state to produce a statelevel JOLTS hires estimate and sum the TS×RatioTS by state
to produce a statelevel JOLTS total separations estimate.
How are the outputs produced?
 Statelevel JOLTS estimates for hires and total separations come directly from the model outlined above.
 Synthetic job openings are a function of the ratio of industryregional job openings and hires. This ratio of published job openings to hires is
applied to model hires estimates to derive model job opening estimates. Ratioadjusting the JOLTS model hires and separations to the regional published
JOLTS hires and separations estimates ensures that the JOLTS published churn rate is fully accounted for.
 Synthetic quits and layoffs and discharges are a function of the relative percentage of the individual components of total separations at the industryregional
level. The relative percentages of each component are applied to the model separations estimates to derive model quits and layoffs and discharges.
What are the limitations?
 This approach is NOT meant to model individual QCEWLDB data records. It would not be prudent to use this approach to model small populations
(30 or fewer establishments). The model works best at the statelevel, and while it is possible to model smaller populations, there potentially is a reduction
in the strength of the model proportionate to the reduction in the size of the population being modeled
 The model does generate statelevel job openings and separations breakouts. However, these estimates are based upon ratios that are common across the region
to which a state belongs. If there are significant differences in the ratio of job openings to hires or separations breakouts for any particular state (or set of
states) within a region, the model cannot detect that and estimates will not reflect those differences.
 Since the model is based on QCEWLDB data, the model cannot produce current statelevel estimate since QCEWLDB data lags current JOLTS estimation production by 6–9 months.
What more is needed?
These estimates are based upon a model. There is, as of yet, no methodology in place that can produce any estimate of error for the estimates the model produces.
Research on a methodology to produce an error estimate is currently underway. The Synthetic model may be augmented in the future with the Census Bureau’s QWI series
of hires and separations.
Composite Synthetic Model
What approach?
The Composite Synthetic model is nearly identical to the Composite Regional model. The primary difference is the use of the Synthetic model estimates (described in the
first section) rather than JOLTS published regional estimates when there is an insufficient amount of JOLTS microdata to produce a statesupersector estimate.
Just like the Composite Regional approach, the JOLTS microdatabased estimate is used in all statesupersector cells that have 30 or more respondents. However, in
contrast to the Composite Regional approach, the Composite Synthetic approach uses the Synthetic estimate in all statesupersector cells that have fewer than five
respondents. In all statesupersector cells with 5–30 respondents an estimate is calculated that is a composition of a weighted estimate of the microdatabased
estimate and a weighted estimate of the Synthetic estimate. The weight assigned to the JOLTS data in those cells is proportional the number of JOLTS respondents in
the cell (weight=n∕30, where n is the number of respondents).
The Composite Synthetic supersector estimates are summed across statesupersectors to the nonfarm level. Composite Synthetic estimates are averaged across 3 months,
creating a 3month moving average.
What data inputs?
 All JOLTS microdata records
 All weights from JOLTS estimation (final weights that account for sampling weight, NRAF, aggcodes, etc.)
 Synthetic estimates (regional JO, H, Q, LD, and TS rates)
 JOLTS regionallevel estimates (to benchmark the state estimates)
 CES statesupersector employment
How are data used?
 All JOLTS microdata are weighted using final weights. A weighted estimate is made for each JOLTS respondent.
 Counts are made for each statesupersector cell.
 Each JOLTS respondent is paired with its Synthetic rate estimate for all variables.
 Based on the count of respondents in the statesupersector cell the JOLTS respondent belongs to, a Composite Model Weighted (CMW) estimate is calculated.
 If the count is>30, then the CMW for the respondent data=1. The CMW for the Synthetic estimate=0.
 If the count<5, then the CMW for the respondent data=0. The CMW for the Synthetic estimate=1.
 If the count is 5–30, then the CMW for the respondent data=n∕30, where n is the number of respondents. The CMW for the Synthetic estimate=1−n∕30.
 The statelevel rate estimate is therefore the final weighted respondentbased JOLTS rate times the CMW added to the Synthetic rate times the CMW,
benchmarked to CES statelevel estimate:
 FINAL ESTIMATE=CES STATE EMP×((final weight JOLTS rate×CMW)+(synthetic rate×CMW))
 To stabilize the estimate, the sum of state Composite Synthetic estimates within each region is then benchmarked to the published JOLTS regional estimates.
How are outputs produced, and what are the limitations?
 This model produces statelevel estimates of JO, H, Q, LD, and TS. These estimates cannot be produced without lag.
What more is needed?
These estimates are based upon a model. There is, as of yet, no methodology in place that can produce any estimate of error for the estimates the model produces.
Research on a methodology to produce an error estimate is currently underway.
Extended Composite Synthetic Model
What Approach?
The Extended Composite Synthetic model is designed to project the Composite Synthetic forward until QCEWLDB data are available to produce Composite Synthetic estimates.
The Composite Synthetic estimates are extended using the ratio of the current Composite Regional state industry estimate to the Composite Regional state industry estimate from one year ago.
This approach ensures that the Extended Composite Synthetic state estimates reflect current JOLTS regional and industrylevel economic conditions. The Extended Composite
Synthetic estimates reflects current JOLTS state economic conditions to the extent that sufficient JOLTS microdata are available.
What data inputs?
 The historical series of Composite Synthetic model estimates at the stateindustrylevel
 The historical series of Composite Regional model estimates at the stateindustrylevel
How are data used?
The Composite Synthetic model estimates are produced at a lag since QCEWLDB data are only available at a 6–9 month lag relative to JOLTS data. The Composite Regional model
estimates, in contrast, are not produced at a lag and are available concurrent with JOLTS data. Therefore, Composite Synthetic estimates can be extended by ratioadjusting the Composite
Synthetic estimates by the ratio of current Composite Regional estimates to the Composite Regional estimates from one year ago at the stateindustrylevel as follows:
Where
 is the Extended Composite Synthetic state industry estimate for month t
 is the Composite Synthetic state industry estimate for month t12 (one year ago)
 is the Composite Regional state industry estimate for month t
 is the Composite Regional state industry estimate for month t12 (one year ago)
Statelevel estimates are produced by summing the Extended Composite Synthetic estimates over industry.
How are outputs produced, and what are the limitations?
 This model will produce statelevel estimates of JO, H, Q, LD, and TS. These estimates are produced without lag. The methodology allows the Extended Composite Synthetic data to
reflect current economic trends at the CESID IndustryRegion level. The projection reflects current state economic trends where sufficient JOLTS microdata are available.
What is the sample size allocation for the inputs used to produce the JOLTS state estimates?
The JOLTS experimental state estimates sample allocation table below provides a snapshot of the sample used in the state estimates. Sample are utilized in both components
of the model. The sample component incorporates JOLTS MSA respondent data. The model component incorporates JOLTS regionallevel respondent data, CES Metro Area respondent
data, and QCEW establishment counts.
SAMPLE ALLOCATION: For State Estimator Components FIPS Code  State  JOLTS State Respondents^{[1]}  JOLTS Regional Respondents^{[2]}  QCEW Establishments^{[3]}  CES State respondents^{[4]} 

2018  2019  2018  2019  2018  2019  2018  2019  1  Alabama  145  135  3,086  2,979  127,057  129,836  14,790  14,880 

2  Alaska  33  29  2,034  1,933  22,064  22,346  2,658  2,690 

4  Arizona  169  163  2,034  1,933  163,385  168,006  10,685  10,780 

5  Arkansas  85  70  3,086  2,979  90,170  91,348  6,657  7,390 

6  California  887  822  2,034  1,933  1,596,644  1,636,051  80,219  80,150 

8  Colorado  181  184  2,034  1,933  205,521  213,236  9,669  9,870 

9  Connecticut  140  119  1,893  1,862  120,353  122,420  8,326  8,330 

10  Delaware  29  27  3,086  2,979  33,266  34,339  2,388  2,390 

11  District of Columbia  44  41  3,086  2,979  40,421  41,459  1,680  1,860 

12  Florida  473  465  3,086  2,979  681,386  704,202  42,642  39,580 

13  Georgia  263  255  3,086  2,979  264,580  270,797  25,136  24,710 

15  Hawaii  31  37  2,034  1,933  41,573  42,600  2,960  3,120 

16  Idaho  53  54  2,034  1,933  62,622  66,589  4,884  4,880 

17  Illinois  405  381  2,153  2,096  367,357  370,118  23,844  24,200 

18  Indiana  219  179  2,153  2,096  166,974  167,895  13,207  13,790 

19  Iowa  113  107  2,153  2,096  102,019  103,255  9,976  10,060 

20  Kansas  119  105  2,153  2,096  88,549  88,903  7,844  7,660 

21  Kentucky  108  96  3,086  2,979  123,587  122,392  8,792  9,120 

22  Louisiana  143  123  3,086  2,979  131,654  133,076  10,094  10,630 

23  Maine  44  47  1,893  1,862  50,574  51,244  4,876  5,210 

24  Maryland  156  138  3,086  2,979  172,964  175,946  10,016  10,360 

25  Massachusetts  249  248  1,893  1,862  263,123  268,730  14,884  15,790 

26  Michigan  264  275  2,153  2,096  250,360  254,321  15,477  15,930 

27  Minnesota  195  175  2,153  2,096  178,398  180,442  10,448  10,530 

28  Mississippi  81  77  3,086  2,979  73,953  73,484  7,303  7,620 

29  Missouri  186  195  2,153  2,096  207,228  212,816  15,701  16,180 

30  Montana  50  54  2,034  1,933  50,002  50,062  3,960  3,950 

31  Nebraska  65  76  2,153  2,096  72,663  72,252  5,601  5,930 

32  Nevada  107  90  2,034  1,933  82,599  82,582  4,322  4,360 

33  New Hampshire  48  55  1,893  1,862  53,095  54,109  4,484  4,520 

34  New Jersey  274  270  1,893  1,862  261,193  265,349  17,243  17,020 

35  New Mexico  76  70  2,034  1,933  60,617  62,972  6,273  6,500 

36  New York  636  639  1,893  1,862  612,862  603,777  39,969  38,940 

37  North Carolina  271  279  3,086  2,979  280,304  286,747  24,763  24,950 

38  North Dakota  38  49  2,153  2,096  31,175  31,153  3,087  3,140 

39  Ohio  323  327  2,153  2,096  301,400  305,330  27,764  27,920 

40  Oklahoma  114  103  3,086  2,979  110,791  111,779  8,580  8,730 

41  Oregon  106  121  2,034  1,933  150,988  154,578  12,984  12,840 

42  Pennsylvania  451  425  1,893  1,862  365,599  368,424  25,550  27,540 

44  Rhode Island  33  38  1,893  1,862  37,458  39,239  2,099  2,290 

45  South Carolina  119  105  3,086  2,979  138,022  141,227  10,510  10,410 

46  South Dakota  32  39  2,153  2,096  33,419  33,943  2,842  2,980 

47  Tennessee  143  144  3,086  2,979  164,216  168,859  12,706  13,110 

48  Texas  667  664  3,086  2,979  692,217  713,990  48,805  50,380 

49  Utah  107  100  2,034  1,933  107,457  111,204  7,765  7,880 

50  Vermont  18  21  1,893  1,862  25,789  26,021  2,554  2,540 

51  Virginia  201  208  3,086  2,979  271,736  272,382  17,881  19,940 

53  Washington  194  178  2,034  1,933  240,720  243,936  14,193  12,990 

54  West Virginia  44  49  3,086  2,979  51,091  51,522  5,792  6,260 

55  Wisconsin  194  182  2,153  2,096  173,764  181,594  11,287  11,630 

56  Wyoming  40  31  2,034  1,933  26,322  26,948  3,074  2,990 

00  Total US  9,166  8,864  9,166  8,870  10,021,281  10,205,830  689,244  697,450 

What more is needed?
These estimates are based upon a model. There is, as of yet, no methodology in place that can produce any estimate of error for the estimates the model produces.
Research on a methodology to produce an error estimate is currently underway.
Last Modified Date: August 19, 2020
