Caila Pinkleton is an economist and Frances Tirado is a statistician in the Division of Current Employment Statistics – National, Office of Employment and Unemployment Statistics, Bureau of Labor Statistics. Telephone: (202) 691‑6555; email: Contact CES
Introduction
With the release of January 2017 data on February 3, 2017, the Bureau of Labor Statistics (BLS) introduced its annual revision of national estimates of employment, hours, and earnings from the Current Employment Statistics (CES) monthly survey of nonfarm establishments. Each year, the CES survey realigns its sample-based estimates to incorporate universe counts of employment—a process known as benchmarking. Comprehensive counts of employment, or benchmarks, are derived primarily from unemployment insurance (UI) tax reports that nearly all employers are required to file with State Workforce Agencies.
The March 2016 benchmark level for total nonfarm employment is 142,814,000.
Compared with the sample-based, published estimate for March 2016, the March 2016 benchmark level is 81,000 lower. This is a downward adjustment of −0.1 percent.
Table 1 below shows the recent history of total nonfarm percent and level benchmark revisions. Over the prior ten years, the annual benchmark revision at the total nonfarm level has averaged 0.3 percent (in absolute terms), with an absolute range of 0.0 percent to 0.7 percent.
Table 1. Percent and level differences between nonfarm employment benchmarks and estimates by industry supersector, March 2006 – 2016 (in thousands) (1)
Footnotes (1) The differences listed in this table reflect the error due to normal benchmarking procedures. Typically this error is equal to the March benchmarked level minus the published March estimated level. However in some years, other factors beyond normal benchmarking procedures influence the difference between the benchmarked and published March estimate levels. Those years are footnoted. (2) A review of industries for the possible presence of noncovered employment in benchmark 2011 yielded 13 additional industries. As a result of including these industries, employment in the amount of 95,000 was added to the total nonfarm benchmark level. The difference between the benchmarked and published March 2011 estimate level was 162,000. For this table, the 95,000 amount was added to the original published total nonfarm and total private March 2011 estimates before calculating the percent and level differences. Portions of the 95,000 amount were also added as appropriate to the original published March 2011 estimates of supersectors financial activities and education and health services before calculating the percent and level differences. (3) The percent and level differences in this column reflect reconstructions to series within CES supersectors financial activities and education and healthcare services. Each first quarter, the Quarterly Census of Employment and Wages (QCEW) program, whose data account for approximately 97 percent of the CES universe scope (see www.bls.gov/web/empsit/cestn.htm#section1), incorporates updated industry assignments. In 2013, these updates included two substantial groups of nonrandom, noneconomic code changes, one to funds, trusts, and other financial vehicles (NAICS 525), and the other, a reclassification of approximately 466,000 in employment from private households (NAICS 814), which is out of scope for CES, to services for the elderly and persons with disabilities (NAICS 62412), which is in scope. These changes also had an impact, beyond what would be considered typical for a given benchmark year, on corresponding CES series. For more information about the changes to these industries, see the QCEW First Quarter 2013 News Release available at www.bls.gov/news.release/archives/cewqtr_09262013.htm or the special notice regarding reconstructed data section of the 2013 CES Benchmark Article at www.bls.gov/ces/cesbmart13.pdf. (4) With the 2015 benchmark, CES reconstructed the national employment series for CES series 65-624120 services for the elderly and persons with disabilities back to January 2000. CES previously reconstructed this series with the 2013 benchmark; however, between the 2013 and 2015 benchmark, a better source of information for the employment within NAICS 62412 for the state of California was found. The inclusion of the reconstructed series resulted in total nonfarm and total private employment that was 27,000 less than the originally published March 2015 estimate level. The difference between the benchmarked and originally published March 2015 estimate level is −199,000 or −0.1 percent. This table displays March 2015 data after accounting for the decrease of 27,000 from the reconstructed series. Similarly, for the education and health services supersector, this table displays March 2015 data after incorporating the reconstructed series.
Tables 2, 5, and 6 on this page contain clickable headers that sort the table by the clicked column.
Table 2 shows the nonfarm employment benchmarks for March 2016, not seasonally adjusted, by industry. The revision to the reconstructed total nonfarm employment is −81,000.
Seven supersectors saw negative revisions. The largest downward revision occurred in professional and business services, −125,000, or −0.6 percent. Within this supersector, the largest revision was in professional and technical services (−51,800, or −0.6 percent). The next largest downward revision occurred in trade, transportation, and utilities, −110,000, or −0.4 percent. This downward revision was driven by a large downward revision in retail trade (−118,200, or −0.8 percent). Education and health services was revised downward by −83,000, or −0.4 percent. This downward revision was driven by a large downward revision in health care (−120,600, or −0.8 percent). Mining and logging experienced the largest downward revision in percentage terms among supersectors with a revision of −22,000, or −3.2 percent, driven by a large downward revision in support activities for mining (−26,600, or −9.9 percent). Other services also experienced a downward revision of −12,000, or −0.2 percent. The two smallest downward revisions occurred in financial activities (−4,000, or 0.0 percent) and information (−2,000, or −0.1 percent).
The remaining four supersectors had upward revisions. The largest upward revision occurred in leisure and hospitality by an amount of 102,000, or 0.7 percent. Within this supersector, the largest revision was in accommodation and food service (101,600, or 0.8 percent). The next largest upward revision occurred in government, which experienced an upward revision of 70,000, or 0.3 percent. Manufacturing was revised upward 58,000, or 0.5 percent. Within this supersector, the largest revision was in food manufacturing (22,400, or 1.5 percent). Construction had an upward revision of 47,000, or 0.7 percent.
Securities, commodity contracts, investments, and funds and trusts
919.1
918.3
0.8
0.1
0.8
0.1
55-524000
Insurance carriers and related activities
2,565.6
2,585.5
-19.9
-0.8
19.9
0.8
55-530000
Real estate and rental and leasing
2,085.0
2,078.7
6.3
0.3
6.3
0.3
55-531000
Real estate
1,523.7
1,527.9
-4.2
-0.3
4.2
0.3
55-532000
Rental and leasing services
538.0
527.6
10.4
1.9
10.4
1.9
55-533000
Lessors of nonfinancial intangible assets
23.3
23.2
0.1
0.4
0.1
0.4
60-000000
Professional and business services
19,704
19,829
-125.0
-0.6
125.0
0.6
60-540000
Professional and technical services
8,815.2
8,867.0
-51.8
-0.6
51.8
0.6
60-541100
Legal services
1,115.6
1,120.1
-4.5
-0.4
4.5
0.4
60-541200
Accounting and bookkeeping services
1,083.2
1,126.0
-42.8
-4.0
42.8
4.0
60-541300
Architectural and engineering services
1,383.3
1,414.4
-31.1
-2.2
31.1
2.2
60-541400
Specialized design services
134.5
137.4
-2.9
-2.2
2.9
2.2
60-541500
Computer systems design and related services
1,943.3
1,952.8
-9.5
-0.5
9.5
0.5
60-541600
Management and technical consulting services
1,329.5
1,299.3
30.2
2.3
30.2
2.3
60-541700
Scientific research and development services
670.3
660.9
9.4
1.4
9.4
1.4
60-541800
Advertising and related services
483.0
488.7
-5.7
-1.2
5.7
1.2
60-541900
Other professional and technical services
672.5
667.4
5.1
0.8
5.1
0.8
60-550000
Management of companies and enterprises
2,221.4
2,252.2
-30.8
-1.4
30.8
1.4
60-560000
Administrative and waste services
8,666.9
8,710.2
-43.3
-0.5
43.3
0.5
60-561000
Administrative and support services
8,273.5
8,313.5
-40.0
-0.5
40.0
0.5
60-561100
Office administrative services
485.7
483.9
1.8
0.4
1.8
0.4
60-561200
Facilities support services
144.0
146.6
-2.6
-1.8
2.6
1.8
60-561300
Employment services
3,419.3
3,461.5
-42.2
-1.2
42.2
1.2
60-561320
Temporary help services
2,787.5
2,802.6
-15.1
-0.5
15.1
0.5
60-561400
Business support services
899.0
903.0
-4.0
-0.4
4.0
0.4
60-561500
Travel arrangement and reservation services
214.6
202.0
12.6
5.9
12.6
5.9
60-561600
Investigation and security services
883.7
881.2
2.5
0.3
2.5
0.3
60-561700
Services to buildings and dwellings
1,919.4
1,917.5
1.9
0.1
1.9
0.1
60-561900
Other support services
307.8
317.8
-10.0
-3.2
10.0
3.2
60-562000
Waste management and remediation services
393.4
396.7
-3.3
-0.8
3.3
0.8
65-000000
Education and health services
22,590
22,673
-83.0
-0.4
83.0
0.4
65-610000
Educational services
3,697.1
3,676.0
21.1
0.6
21.1
0.6
65-620000
Health care and social assistance
18,892.8
18,997.0
-104.2
-0.6
104.2
0.6
65-620001
Health care
15,263.5
15,384.1
-120.6
-0.8
120.6
0.8
65-621000
Ambulatory health care services
6,988.7
7,041.1
-52.4
-0.7
52.4
0.7
65-621100
Offices of physicians
2,502.5
2,531.1
-28.6
-1.1
28.6
1.1
65-621200
Offices of dentists
912.2
924.5
-12.3
-1.3
12.3
1.3
65-621300
Offices of other health practitioners
837.2
839.6
-2.4
-0.3
2.4
0.3
65-621400
Outpatient care centers
840.6
820.3
20.3
2.4
20.3
2.4
65-621500
Medical and diagnostic laboratories
263.3
259.0
4.3
1.6
4.3
1.6
65-621600
Home health care services
1,345.2
1,374.1
-28.9
-2.1
28.9
2.1
65-621900
Other ambulatory health care services
287.7
292.5
-4.8
-1.7
4.8
1.7
65-622000
Hospitals
4,979.9
5,036.0
-56.1
-1.1
56.1
1.1
65-623000
Nursing and residential care facilities
3,294.9
3,307.0
-12.1
-0.4
12.1
0.4
65-623100
Nursing care facilities
1,632.4
1,645.9
-13.5
-0.8
13.5
0.8
65-623200
Residential mental health facilities
613.2
612.5
0.7
0.1
0.7
0.1
65-623300
Community care facilities for the elderly
886.9
886.7
0.2
0.0
0.2
0.0
65-623900
Other residential care facilities
162.4
161.9
0.5
0.3
0.5
0.3
65-624000
Social assistance
3,629.3
3,612.9
16.4
0.5
16.4
0.5
65-624100
Individual and family services
2,208.0
2,204.1
3.9
0.2
3.9
0.2
65-624200
Emergency and other relief services
162.4
158.1
4.3
2.6
4.3
2.6
65-624300
Vocational rehabilitation services
336.5
333.0
3.5
1.0
3.5
1.0
65-624400
Child day care services
922.4
917.7
4.7
0.5
4.7
0.5
70-000000
Leisure and hospitality
15,143
15,041
102.0
0.7
102.0
0.7
70-710000
Arts, entertainment, and recreation
2,067.8
2,066.7
1.1
0.1
1.1
0.1
70-711000
Performing arts and spectator sports
438.9
454.4
-15.5
-3.5
15.5
3.5
70-712000
Museums, historical sites, and similar institutions
148.6
145.8
2.8
1.9
2.8
1.9
70-713000
Amusements, gambling, and recreation
1,480.3
1,466.5
13.8
0.9
13.8
0.9
70-720000
Accommodation and food services
13,075.4
12,973.8
101.6
0.8
101.6
0.8
70-721000
Accommodation
1,878.6
1,861.5
17.1
0.9
17.1
0.9
70-722000
Food services and drinking places
11,196.8
11,112.3
84.5
0.8
84.5
0.8
80-000000
Other services
5,642
5,654
-12.0
-0.2
12.0
0.2
80-811000
Repair and maintenance
1,291.8
1,289.7
2.1
0.2
2.1
0.2
80-812000
Personal and laundry services
1,424.2
1,415.1
9.1
0.6
9.1
0.6
80-813000
Membership associations and organizations
2,925.8
2,948.8
-23.0
-0.8
23.0
0.8
90-000000
Government
22,552
22,482
70.0
0.3
70.0
0.3
90-910000
Federal
2,767.0
2,758.0
9.0
0.3
9.0
0.3
90-911000
Federal, except U.S. Postal Service
2,162.7
2,153.9
8.8
0.4
8.8
0.4
90-919120
U.S. Postal Service
604.6
604.5
0.1
0.0
0.1
0.0
90-920000
State government
5,247.0
5,234.0
13.0
0.2
13.0
0.2
90-921611
State government education
2,573.0
2,565.7
7.3
0.3
7.3
0.3
90-922000
State government, excluding education
2,673.6
2,667.9
5.7
0.2
5.7
0.2
90-930000
Local government
14,538.0
14,490.0
48.0
0.3
48.0
0.3
90-931611
Local government education
8,243.6
8,208.1
35.5
0.4
35.5
0.4
90-932000
Local government, excluding education
6,294.6
6,281.5
13.1
0.2
13.1
0.2
Footnotes
(1) During processing for the 2016 benchmark it was found that several establishments originally coded in CES industry 65-621111 (offices of physicians, except mental health) should have been coded in CES industry 65-621491 (HMO medical centers) back to January 1990. Additionally with the 2016 benchmark, CES reconstructed the CES national employment series for 90-921611 (state government education) and 90-931611 (local government education) back to April 2014 to address an ownership reclassification of some colleges and universities. For more information, see the Reconstruction section in the 2016 Benchmark Article.
From April 2016 to December 2016, also known as the postbenchmark period, estimates were recalculated for each month based on new benchmark levels and new net birth/death factors. Net birth/death factors were revised to incorporate information from the most recent year of universe employment counts. Table 3 shows the net birth/death model values for the supersectors over the postbenchmark period. From April 2016 to December 2016, the net birth/death model cumulatively added 1,021,000, compared with 881,000 in the previously published April 2016 to December 2016 employment estimates.
Table 3. Net birth/death estimates by industry supersector, April to December 2016 (in thousands)
Table 4 presents revised total nonfarm employment data on a seasonally adjusted basis for January 2016 through December 2016. The revised data for April 2016 forward incorporate the effect of applying the rate of change measured by the sample to the new benchmark level, as well as updated net birth/death model adjustments and new seasonal adjustment factors. Revisions to November and December also reflect incorporation of additional sample receipts.
Table 4. Differences in seasonally adjusted levels and over-the-month changes, total nonfarm employment, January – December 2016 (in thousands)
With the release of the January 2017 first preliminary estimates on February 3, 2017, CES incorporated series changes related to annual sample adequacy and disclosure review.
Series changes
All CES series are evaluated annually for sample size, coverage, and response rates. The following changes resulted from a re-evaluation of the sample and
universe coverage for CES industries, which are based on North American Industry Classification System (NAICS) industries. Some small industries no longer have sufficient sample to be
estimated and published separately and were combined with other similar industries for estimation and publication purposes. The following information is also available at
www.bls.gov/web/empsit/cesnewseries.htm.
Some series have new CES industry codes or titles as a result of the series changes (Exhibit 1).
These CES industry code or title changes were applied to all data types published for the designated series.
Historical data for these series with new CES industry codes or CES industry titles are impacted as noted on the remainder of this page;
historical data are available under the new CES industry codes or CES industry titles.
Exhibit 1. Series with CES industry code or title Changes
NAICS Code
Previous
New
CES Industry Code
CES Industry Title
CES Industry Code
CES Industry Title
335313,4
31-335314
Relays and industrial controls
31-335314
Switchgears, switchboard apparatus, relays, and industrial controls
3122,60
32-329200
Tobacco and tobacco products
32-329900
Other miscellaneous nondurable goods manufacturing
Exhibit 2 through Exhibit 7 show the new CES industry codes and titles, not the previous CES industry codes and titles, as noted in Exhibit 1.
Only directly estimated data types1 are included in the exhibits:
All Employees (AE)
AE Average Weekly Hours (AE AWH)
AE Average Hourly Earnings (AE AHE)
AE Average Weekly Overtime Hours (AE AWOH)
Women Employees (WE)
Production Employees (PE)
PE Average Weekly Hours (PE AWH)
PE Average Hourly Earnings (PE AHE)
PE Average Weekly Overtime Hours (PE AWOH)
The directly estimated data types listed except for AE are collectively called non-AE data types. In order to more easily identify affected series, since AE series are published at a more detailed industry
level than non-AE series, series changes are shown for AE and non-AE data types. The non-AE tables cover all directly estimated non-AE data types.
The first three exhibits reference the AE data type and the second three exhibits reference all non-AE data types. The three exhibits in each group display the discontinued, collapsed, and new series.
Discontinued series exhibits (Exhibit 2 and Exhibit 5) display series for which the data types noted are no longer published. Collapsed series exhibits
(Exhibit 3 and Exhibit 6) display series for which the data types noted are no longer published because the industry no longer has sufficient sample to be estimated
and published separately. Affected industries have been combined with other similar industries for estimation and publication purposes. Historical data for these series were reconstructed to provide consistent
time series. New series exhibits (Exhibit 4 and Exhibit 7) display series for which the data types noted are now published.
Several series were reconstructed as part of the 2016 CES benchmark. Exhibit 8 outlines the CES industries, data types, and timespans that were impacted by the reconstructions.
Exhibit 8. CES detailed series involved in data reconstructions
HMO medical centers and offices of physicians, except mental health reconstructed from January 1990 forward
During processing for the 2016 benchmark, a change to NAICS codes for reports in health care services affecting two industries was discovered. This change was determined to be non-economic and affected the entire series going back to 1990.
Several establishments originally coded in CES industry 65-621111 (offices of physicians, except mental health) should have been coded in CES industry 65-621491 (HMO medical centers). A ratio of the employment in the reclassified establishments to the employment originally in 65-621111 was calculated for March 2016. This March 2016 ratio was applied back to January 1990 to each month’s employment level for 65-621111. The resulting employment amounts were taken out of 65-621111 and put in industry 65-621491. This change affected all employees (AE) for 65-621491 and AE, production and nonsupervisory employment (PE), and women’s employment (WE) for 65-621111. At higher levels of aggregation, average weekly hours and average hourly earnings for both AE and PE (AE AWH, AE AHE, PE AWH, and PE AHE) were affected for 65-621111. Aggregate levels of AE, WE, and PE were unaffected at the 3-digit level. All industries impacted by this reconstruction are included in Exhibit 9 below.
This series was benchmarked according to the usual methodology after the reconstruction. The reconstruction and subsequent benchmark resulted in revisions to seasonally and not seasonally adjusted data from January 1990 forward. Due to the time span of the reconstruction, the entire series was re-seasonally adjusted back to 1990. More information about CES benchmarking methods is available in the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section7b. More information about seasonal adjustment of CES series is available in the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section6e.
Exhibit 9. Industries impacted by CES reconstruction of HMO medical centers and offices of physicians, except mental health, January 1990 forward
State and local government education reconstructed from April 2014 forward
A reclassification of some colleges and universities from state to local ownership caused a shift in employment from CES series 90-921611 (state government education) into 90-931611 (local government education). Prior to benchmarking, sample-based links2 for AE were constructed from original matched sample data after accounting for the ownership change. These sample-based links were applied to AE from July 2014 to the March 2015 benchmark, after which the series were wedged back to April of 2014. The months following March 2015 were re-projected forward using the newly calculated sample links for AE. The previous level of employment was maintained for government by calculating the differences in employment between the summed state, federal, and local government employment and the previous aggregate government employment. Differences were raked proportionally into 90-921611 and 90-931611.
New ratios of WE to AE employment were calculated from the reclassified sample data. The new WE ratios were applied to the reconstructed AE series from April 2014 forward to create a reconstructed WE series.
The reconstruction resulted in state government education AE that was 35,700 less than the originally published March 2016 estimate level and local government education AE that was 34,800 more than the originally published March 2016 estimate level. All industries impacted by this reconstruction are included in Exhibit 10 below.
The series were benchmarked according to the usual methodology after the reconstruction. The reconstruction and subsequent benchmark resulted in revisions to the not seasonally adjusted data from April 2014 forward. Seasonally adjusted series are subject to revisions back 5 years. More information about CES benchmarking methods is available in the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section7b. More information about seasonal adjustment of CES series is available in the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section6e.
Exhibit 10. Industries impacted by CES reconstruction of state and local government education, April 2014 forward
A benchmark revision is the difference between the benchmark employment level for a given March and its corresponding sample-based estimate. The overall accuracy of the establishment survey is usually gauged by the size of this difference. The benchmark revision often is regarded as a proxy for total survey error, but this does not take into account error in the universe data or infrequent events such as historical reconstructions. The employment counts obtained from quarterly UI tax forms are administrative data that reflect employer record-keeping practices and differing state laws and procedures. The benchmark revision can be more precisely interpreted as the difference between two independently derived employment counts, each subject to its own error sources.
Like all sample surveys, the establishment survey is susceptible to two sources of error: sampling error and nonsampling error. Sampling error is present any time a sample is used to make inferences about a population. The magnitude of the sampling error, or variance, relates directly to sample size and the percentage of the universe covered by that sample. The CES monthly survey captures slightly under one-third of the universe, exceptionally high by usual sampling standards. This coverage ensures a small sampling error at the total nonfarm employment level. More information on the reliability of CES estimates is available in the Reliability section of the CES Technical Notes available at www.bls.gov/web/empsit/cestn.htm#section1c.
Both the universe counts and the establishment survey estimates are subject to nonsampling errors common to all surveys – measurement, response, and processing errors. The error structures for both the CES monthly survey and the UI universe are complex. Still, the two programs generally produce consistent total employment figures, each validating the other.
The routine benchmarking process results in revisions to the series for employment of production and nonsupervisory employees (PE) and employment of women employees (WE). There are no benchmark employment levels for these series; they are revised by preserving ratios of employment for the particular data type to employment of all employees (AE) prior to benchmarking, and then applying these ratios to the revised employment of all employees. These figures are calculated at the basic cell level and then aggregated to produce the summary estimates.
Average weekly hours, average hourly earnings, and in manufacturing industries, average weekly overtime hours are not benchmarked; they are estimated solely from reports supplied by survey respondents at the basic estimating cell level.
The aggregate industry levels of the hours and earnings series are derived as a weighted average. The employment of all employees estimates or the employment of production and nonsupervisory employees estimates for the basic cells essentially act as weights for their respective hours and earnings estimates for broader industry groupings. Adjustments of the employment of all employees estimates to new benchmarks may alter the weights used for both AE and PE hours and earnings, which, in turn, may change the estimates for both AE and PE hours and earnings at higher levels of aggregation.
Generally, new employment benchmarks have little effect on hours and earnings estimates for major industry groupings. To influence the hours and earnings estimates of a broader industry group, employment revisions have to be relatively large and must affect industries that have hours or earnings averages that are substantially different from those of other industries in their broader group. Table 5 and Table 6 provide information on the levels of specific hours and earnings series resulting from the March 2016 benchmark. At the total private level, there was no change in average weekly hours estimates for both AE and PE from the previously published level. Total private average hourly earnings increased by one cent for AE from the previously published level and did not change for PE.
Table 5. Effect of March 2016 benchmark revisions to AE AWH and AE AHE estimates, selected industries
(1)
During processing for the 2016 benchmark, several establishments originally coded in CES industry
65-621111 (offices of physicians, except mental health) were corrected to be coded in CES industry
65-621491 (HMO medical centers). Estimates for these industries were adjusted back to 1990 to account for this change.
All employees (AE) for 65-621491 and AE; women employees (WE); production employees (PE); and average weekly hours,
average hourly earnings, and average weekly earnings for both AE and PE (AE AWH, PE AWH, AE AHE, PE AHE, AE AWE, PE AWE)
were affected for 65-621111. At aggregate levels AE, PE, and WE were not affected, although there were changes to
hours and earnings estimates for both AE and PE. For more information, see the Reconstruction
section in the 2016 Benchmark Article.
(1)
During processing for the 2016 benchmark, several establishments originally coded in CES industry
65-621111 (offices of physicians, except mental health) were corrected to be coded in CES industry
65-621491 (HMO medical centers). Estimates for these industries were adjusted back to 1990 to account for this change.
All employees (AE) for 65-621491 and AE; women employees (WE); production employees (PE); and average weekly hours,
average hourly earnings, and average weekly earnings for both AE and PE (AE AWH, PE AWH, AE AHE, PE AHE, AE AWE, PE AWE)
were affected for 65-621111. At aggregate levels AE, PE, and WE were not affected, although there were changes to
hours and earnings estimates for both AE and PE. For more information, see the Reconstruction
section in the 2016 Benchmark Article.
Establishment survey benchmarking is done on an annual basis to a population derived primarily from the administrative file of employees covered by UI. All employers covered by UI laws are required to report employment and wage information to the appropriate Labor Market Information (LMI) agency four times per year. Approximately 97 percent of total nonfarm employment within the scope of the establishment survey is covered by UI. A benchmark for the remaining 3 percent is constructed from alternate sources, primarily records from the Railroad Retirement Board (RRB), County Business Patterns (CBP), and Annual Survey of Public Employment and Payroll (ASPEP). This 3 percent is collectively referred to as noncovered employment. Noncovered employment data from these sources are available only on a lagged basis. More information on calculating noncovered employment in the CES program is available in the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#NCE.
The time required to complete the benchmark revision process—from the full collection of the UI population data to publication of the revised industry estimates—is about ten months. The benchmark adjustment procedure replaces the March sample-based employment estimates with UI-based population counts for March. The benchmark therefore determines the final employment levels, while sample movements capture month-to-month trends.
On a not seasonally adjusted basis, the sample-based estimates for the year preceding and the nine months following the benchmark month (March) are also subject to revision; seasonally adjusted data are subject to revision back five years. Employment estimates for the months between the most recent March benchmark and the previous year's benchmark are adjusted using a linear "wedge-back" procedure, which assumes that the total estimation error accumulated at a steady rate since the last benchmark. For the nine months following the March benchmark (also referred to as the postbenchmark period), CES applies previously derived over-the-month sample changes to the revised March level to get the revised estimates (see Revisions in the post benchmark period). New net birth/death model estimates also are calculated and applied during postbenchmark estimation. More information on benchmarks in the CES program is available in the Benchmarks section of the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section7b.
The CES sample alone is not sufficient for estimating the total nonfarm employment level because each month new firms generate employment that cannot be captured through the sample. There is an unavoidable lag between a firm opening for business and its appearance on the CES sample frame. The sample frame is built from Unemployment Insurance (UI) quarterly tax records. These records cover virtually all U.S. employers and include business births, but they only become available for updating the CES sampling frame 7 to 9 months after the reference month. After the births appear on the frame, there is also time required for sampling, contacting, and soliciting cooperation from the firm, and verifying the initial data provided. In practice, BLS cannot sample and begin to collect data from new firms until they are at least a year old.
BLS has researched both sample-based and model-based approaches to measuring birth units that have not yet appeared on the UI universe frame. The research demonstrated that sampling for births was not feasible in the very short CES production timeframes, so the Bureau utilizes a model-based approach for this component. This model incorporates two components; the first component is an indirect imputation for business deaths and the second component is an Autoregressive Integrated Moving Average (ARIMA) time series model designed to estimate the net birth/death employment not accounted for by the imputation from the first component. More information on the CES birth/death model is available in the Birth/Death Model section of the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm#section6c.
Only error from the second component is directly measurable. Error from this component is measured by comparing the actual net of births and deaths with the model-based adjustment that was used in the CES sample-based estimates. Most recently, the data from April 2015 to March 2016 can be measured. As Table 7 shows, the actual net birth/death for April 2015 to March 2016 was approximately 99,000 above the forecasted amount used in the CES monthly estimates for the same time period.
Table 7. Differences between forecasted and actual net birth/death, total private employment, April 2015 – March 2016 (in thousands)
Typically, five years’ worth of seasonally adjusted data are revised with each CES annual benchmark. As with the not seasonally adjusted data, if additional seasonally adjusted revisions to the seasonal adjustment data are made with a benchmark, those revisions will be noted in previous sections of this article.
For technical information on how seasonal adjustment is performed in the CES program, refer to the Seasonal Adjustment section of the CES Technical Notes, available at www.bls.gov/web/empsit/cestn.htm#section6e.
For more information on seasonal adjustment model specifications and special model adjustments, please see the Seasonal Adjustment Model Specification List section of the CES Seasonal Adjustment Files and Documentation page, available at www.bls.gov/web/empsit/cesseasadj.htm#samodel.
LABSTAT, the BLS public database, contains all historical employment, hours, and earnings data revised as a result of this benchmark, including both not seasonally adjusted and seasonally adjusted data. The data can be accessed at www.bls.gov/ces, the CES homepage.
Previously published data are available on both a not seasonally adjusted and seasonally adjusted basis for all CES industries down to the 3-digit level from the CES Vintage Data page, available at www.bls.gov/web/empsit/cesvininfo.htm.
1CES estimates data in two ways: directly and indirectly. Directly estimated data types refer to data types for which estimates are calculated directly from the responding sample. Indirectly estimated data types refer to data types for which estimates are calculated from other directly estimated data types. Average weekly earnings of all employees and indexes of aggregate weekly hours of all employees are examples of indirectly estimated data types. For more information on indirectly estimated data types, see the CES Technical Notes available at www.bls.gov/web/empsit/cestn.htm.
2A matched sample link is defined to be all sample members that have reported data for the reference month and the month prior. Excluded from the matched sample link is any sample unit that reports that it is out-of-business and has zero employees. For more information, please refer to the CES Technical Notes at www.bls.gov/web/empsit/cestn.htm.
