Diversified REITs · U.S. Bureau of Labor Statistics
BLS Public Data API - NAICS 56 Employment Series
Datadory delivers bls public data api naics 56 employment series ceu6056000001 data: the monthly all-employees count for administrative and support and waste management and remediation services, unadjusted, from the Current Employment Statistics national program - roughly 1,000 monthly observations back to 1939, nine documented fields per record, delivered as files or straight into your warehouse.
API, files, or your warehouse. Daily, weekly, or hourly.
- Where it covers
- United States, national total
- How far back
- Monthly observations back to 1939; roughly 1,000 readings in the full run; year ranges selectable at delivery
- How fine
- Monthly, one industry line - NAICS 56, administrative and support and waste management and remediation services, unadjusted
What is the BLS Public Data API - NAICS 56 Employment Series?
One line, one month, the entire support-services workforce of the United States. CEU6056000001 is a Current Employment Statistics national series from the U.S. Bureau of Labor Statistics: all employees, in thousands, not seasonally adjusted, for administrative and support and waste management and remediation services - NAICS code 56, filed under the professional and business services supersector. The identifier decodes on sight: CEU marks CES National, U marks the unadjusted variant, 60 the supersector, 60560000 the industry code mapping to NAICS 56, and 01 the all-employees measure.
The headline is the staffing, security, janitorial and waste-remediation headcount, counted from payroll records. At extraction in August 2026 the newest reading was July 2026 at 9,124.4 thousand employees, flagged preliminary; June 2026 read 9,161.1 thousand, also preliminary. Set those against a January 2020 level of 8,974.3 thousand and an April 2020 pandemic low of 7,555.9 thousand and one series carries the pre-shock peak, the deepest contraction and the recovery in a single comparable line.
What do sample rows look like?
Each record is one month of one measure. Four real observations, exactly as the series reads:
seriesID : CEU6056000001 # CES national, unadjusted
year : 2026
period : M07 # July 2026
value : 9124.4 # all employees, thousands
footnotes: P # preliminary
seriesID : CEU6056000001
year : 2026
period : M06 # June 2026
value : 9161.1 # all employees, thousands
footnotes: P # preliminary
seriesID : CEU6056000001
year : 2020
period : M01 # January 2020, pre-shock level
value : 8974.3 # all employees, thousands
seriesID : CEU6056000001
year : 2020
period : M04 # April 2020, pandemic trough
value : 7555.9 # all employees, thousands
footnotes: -Read the block once and the story tells itself: 9.16 million support-services jobs in June 2026 against 7.56 million in April 2020 - a swing of roughly 1.6 million positions inside one sector line - while January 2020 fixes the pre-pandemic baseline at 8.97 million. Every month since 1939 is keyed the identical way, so a chart of the whole run and a chart of one quarter come off the same schema.
Why does a diversified REITs shelf track a support-services payroll?
Diversified REITs hold mixed portfolios, and the tenant base underneath them is staffed, cleaned, secured and serviced by NAICS 56. When that payroll expands, office demand, service contracting and facilities spend expand with it; when it contracts, the pain lands quarters before it reaches a rent roll.
The series quantifies the exposure. Peak-to-trough between January 2020 and April 2020 the line fell from 8,974.3 to 7,555.9 thousand - roughly a 16 percent contraction in a single quarter, a depth few tenant industries matched. A monthly frequency catches turns that quarterly fundamentals report late, and the national scope matches how a diversified portfolio actually spreads its exposure.
How should an unadjusted series be modeled?
The U in CEU6056000001 means not seasonally adjusted: the values are raw payroll counts, with every seasonal wrinkle left in. Staffing and administrative work swings around holidays, fiscal year-ends and project cycles, so adjacent months can move for reasons that have nothing to do with the cycle.
Two habits keep the series honest. Compare the same calendar month across years rather than neighboring months when reading trend. And for smoothing work, pair it with the seasonally adjusted sibling in the same CES family, which carries the same industry definition on adjusted values. Datadory delivers either cut on request, with the field dictionary traveling alongside.
What comes back with every record?
Nine documented fields ride on every delivery. Five carry the economics: seriesID, year, period (M01 through M12), value in thousands of employees, and footnotes - where the P marking an observation preliminary lives. Four are delivery metadata: the Results.series container, a status marker, the processing responseTime in milliseconds, and an informational message.
That is the whole schema, and it never changes month to month, which is why the series loads cleanly into a warehouse table and diffs cleanly against yesterday's copy.
Field dictionary
Every field below is documented against real records. The full dictionary ships with the sample.
| Field | Type | Definition | Example |
|---|---|---|---|
seriesID | string | BLS series identifier. | CEU6056000001 |
year | string | Observation year. | 2026 |
period | string | Observation period code, M01 through M12 for the twelve months of the year. | M07 |
value | number | All employees in thousands of persons, not seasonally adjusted. | 9124.4 |
footnotes | string | Footnote codes riding beside the observation, such as P for preliminary. | P |
Results.series | string | Container object holding the returned series, each with its identifier and its block of observations. | - |
status | string | Status marker stating whether a delivery succeeded. | RESPONSE_SUCCESS |
responseTime | integer | Processing time in milliseconds, recorded for each delivery. | - |
message | string | Informational text accompanying a returned series. | - |
BLS Public Data API - NAICS 56 Employment Series - product specification
| Attribute | Value |
|---|---|
| Industry | Diversified REITs |
| Records | One national series of roughly 1,000 monthly observations back to 1939 |
| Fields | Nine documented fields per record |
| Geographic coverage | United States (national) |
| Temporal coverage | Monthly observations back to 1939; year ranges selectable at delivery |
| Granularity | Monthly by NAICS 56 industry, all employees, not seasonally adjusted |
| Source | U.S. Bureau of Labor Statistics |
| Quality score | 8/10 on the Datadory rubric |
What teams do with it
- Tenant-industry demand monitoring The monthly support-services headcount is the fastest readable proxy for whether the service layer beneath office and mixed-use properties is expanding or quietly thinning.
- Cycle timing and shock benchmarking January 2020 to April 2020 compressed a 16 percent sector contraction into three prints - the series doubles as a calibration curve for how sharply this tenant base moves in stress.
- Macro features for valuation models Eight decades of monthly observations give cap-rate, NOI and occupancy models a long, consistently defined macro regressor that survives backtests to the 1950s.
- Facilities and staffing spend planning Operators pricing cleaning, security and contingent labor contracts read the sector headcount as the demand side of their own book.
- Editorial and research grounding Writers and analysts cite one official line instead of assembling secondary estimates, with the footnote conventions documented beside every value.
Questions buyers ask
What does series CEU6056000001 measure?
All employees, in thousands, not seasonally adjusted, for administrative and support and waste management and remediation services - NAICS 56 - from the Current Employment Statistics national program of the U.S. Bureau of Labor Statistics. One observation per month, one consistent definition, roughly 1,000 monthly readings back to 1939.
How far back does bls public data api naics 56 employment series ceu6056000001 data go?
To 1939 for the national CES series, giving roughly eight decades of monthly history in a single uninterrupted line. Year ranges are selectable at delivery, so you can take the full run or bound it to the window your model actually needs.
Why is the series published unadjusted?
The CEU prefix marks the unadjusted variant of the CES national series: raw payroll counts with seasonal patterns left visible. Read the same calendar month across years to see trend, and reach for the seasonally adjusted sibling in the same family when a smoothed level is what the analysis calls for.
How current is the newest observation?
At our August 2026 extraction the newest reading was July 2026 at 9,124.4 thousand employees, and June 2026 read 9,161.1 thousand - both carrying the P preliminary footnote. Treat trailing months as provisional and anchor conclusions on settled vintages; earlier months arrive already revised.
Was April 2020 the largest move in the series?
It is the deepest contraction in the modern record: 8,974.3 thousand employees in January 2020 fell to 7,555.9 thousand by April 2020, a loss of roughly 1.42 million jobs - about 16 percent of the sector - inside one quarter. Nothing else in the postwar span approaches that slope.
Can I get a sample cut to my own window?
Yes. Name the years and months you want - say, January 2019 onward, or 2008 alone for a stress-period study - and the sample arrives bounded to that range with the field dictionary and the footnote conventions documented beside it.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.