This page addresses the effects of the 2025 federal government shutdown on the Current Population Survey (CPS), a monthly survey of U.S. households conducted by the U.S. Census Bureau for the Bureau of Labor Statistics. The CPS provides a wide range of information about employment and unemployment.
Federal government agencies were shut down or operating at reduced staffing levels during a lapse in appropriations from October 1, 2025, through November 12, 2025. All CPS operations, including data collection, were suspended during that time period.
We will update this page as more information becomes available. We are working to fully assess the situation and determine how to address the impact on CPS procedures and products.
You may be interested in these resources
1. Were September 2025 data affected?
No. Collection of CPS data for September 2025 had been completed in accordance with our normal schedule prior to the federal government shutdown. Publication of September data was delayed by more than 6 weeks (from October 3 to November 20) because of the shutdown. (See the September 2025 Employment Situation news release.)
2. Were October 2025 data affected?
Yes. There are no CPS estimates for October 2025. CPS data were not collected for October 2025 due to the lapse in appropriations and were not collected retroactively. All CPS operations were suspended during the federal government shutdown, which began before and continued beyond the scheduled collection period for October data.
BLS did not issue an October 2025 Employment Situation news release.
3. When was the November 2025 reference period?
The CPS reference week was November 9–15, 2025. The reference week usually is the 7-day calendar week (Sunday–Saturday) that includes the 12th of the month.
4. What was the impact on November data collection?
As announced, the November 2025 collection period was extended. Data collection typically begins on the Sunday immediately following the reference week and usually lasts 10 days.
The absence of October data increased the burden on collecting data in November. To allow adequate time for the Census Bureau staff to collect the November data and account for the limited availability of respondents around the Thanksgiving holiday, the collection period was extended through November 30.
Collection began on Monday, November 17, one day later than in a typical collection cycle, and continued through Sunday, November 30. No interviews were conducted on Thanksgiving Day (November 27).
5. What happened to households that were not contacted in October?
The CPS is a panel survey in which households are in the sample for 4 consecutive months, out of the sample for 8 months, back in the sample for 4 consecutive months, and then retired from the sample. A new panel of households is introduced into the sample each month, just as another panel is retired. Each panel represents about one-eighth of the monthly sample and is assigned a number that indicates where it is in the survey cycle. For example, households in the sample for the first month are referred to as MIS-1 (month-in-sample 1) and those returning after their 8-month break are referred to as MIS-5.
CPS continued sample rotation in November as originally scheduled, even though there were no households contacted in October. That is, the MIS-1 households in October 2025 became the MIS-2 households in November, the MIS-2 households in October became the MIS-3 households in November, and so forth. The MIS-4 households in October rotated out of the sample and will be contacted in 8 months. The MIS-8 households in October were retired out of the sample as planned and were not contacted in November.
Because of this, one-fourth of the November sample households were participating in the CPS survey for the first time (both the MIS-1 and MIS-2 panels), amounting to twice as many new households as in a typical month. Additionally, one-fourth of the November sample households were returning from a long break (MIS-5 and MIS-6 panels).
The larger number of first interviews and returning interviews in November 2025 added to the collection burden. Generally, households entering the sample for their first month are interviewed through a personal visit, and the interview is longer because demographic characteristics are collected for all household members. (Interviews other than the first one are usually collected by telephone.) Households in their fifth month, which are collected after the 8-month break, require some additional questions to update demographic and other details. Interviews for other months may be shorter in length.
6. What statistical weighting changes were made in November 2025?
CPS labor force estimates are derived using a composite estimator that takes advantage of the 75 percent overlap in the CPS sample from month-to-month. The use of composite estimation helps improve the accuracy of the labor force estimates by reducing the standard errors associated with month-to-month changes. In the CPS, composite estimation relies on data from the previous month as an input in developing statistical weights for the current month’s data.
Specifically, CPS uses the AK form of composite estimation which includes a difference term that accounts for the change in labor force estimates within overlapping panels from one month to the next. Statistical weights for November 2025 CPS data were generated following the usual AK composite estimation process, except that the difference term was modified to reflect the 2-month change in overlapping panel estimates from September to November because no October data were collected. November 2025 CPS labor force estimates will have slightly higher variances than usual because the difference term reflects that only 50 percent of the September and November samples overlap rather than the typical 75 percent overlap in the CPS sample from month-to-month.
This weighting change for the November 2025 estimates will not be needed for the December estimates, which will return to the usual composite weighting methodology. The impact on standard errors will get progressively smaller over the next several months before becoming negligible.
For more information about composite estimation see Chapter 2-3 of CPS Technical Paper 77. For more technical information about this November 2025 change in statistical weighting, see Current Population Survey Composite Estimation, November 2025.
7. What is the impact of the statistical weighting changes made in November 2025?
Compared with usual monthly labor force estimates from CPS, the November 2025 estimates have slightly higher variances due to a change in how the November statistical weights were produced. (See item 6 above.) For example, because of the weighting methodology change, the standard error for the national unemployment rate is larger than usual by a factor of 1.06. The slight increases in standard errors may persist for a few months after November 2025. For more technical information about this November 2025 change in statistical weighting, see Current Population Survey Composite Estimation, November 2025.
8. How is the missing October 2025 data represented in BLS products?
Missing data for October 2025 appears differently in different CPS products.
A dash is used to indicate missing October 2025 data in CPS tables, including the household survey “A” tables associated with the Employment Situation news release.
For the online Labor Force Statistics (LN) database, the appearance of missing data is determined by the format of the output.
For html displays, missing data for October 2025 appears as -(9) or - (9), where the dash indicates missing data and the text of footnote 9 reads “Data unavailable due to the 2025 lapse in appropriations.” In XLSX displays, the missing data are represented as blank cells with the footnote: “* Data unavailable due to the 2025 lapse in appropriations.” CSV and text files show dashes with no footnote.
In the BLS Public Data API, users will see a dash to represent the missing data value. Both net changes and percent are not visible. Footnote text “Data unavailable due to the 2025 lapse in appropriations” is included.
There is no public use microdata file for October 2025.
9. Was the seasonal adjustment methodology changed in November?
No. The seasonal adjustment methodology for the CPS did not change. The methodology can accommodate missing values in a time series and can compute seasonally adjusted estimates even in the absence of a month of data.
10. How did dependent interviewing work in November?
The CPS utilizes computer-assisted interviewing that allows for dependent interviewing techniques, that is, the use of information in the current interview that was obtained in a previous month's interview. For example, a person’s duration of unemployment is asked in the first month they are reported as unemployed, and this information is automatically updated by either 4 or 5 weeks if they remain unemployed in the subsequent month.
Because data were not collected in October, the information related to the duration of unemployment that would normally be carried forward in a dependent interview in November was not available, and questions needed to be re-asked. In November, all jobseekers and people who were on layoff were asked how long they had been looking for work or on layoff. Typically, only the relatively small number of people seeking a job or on layoff who did not respond to the survey the previous month, or who were newly unemployed, would have been asked about the duration of their unemployment.
11. What was the November response rate?
The CPS response rate was 64.0 percent in November 2025, a series low. The prior 12-month average was 68.4 percent. (The previous low was 64.9 percent in June 2020.)
12. How were variances and statistical significance in November 2025 labor force estimates affected?
Each month, CPS publishes tables that present the changes needed for labor force estimates to show statistically significant movements over selected time intervals. That is, they show the minimum changes needed in the labor force estimates, at a 90-percent confidence level, to indicate that the observed changes are not due to sampling error. (Learn more about the reliability of CPS estimates.)
For November 2025 data, the changes needed for statistical significance are slightly larger than usual. For example, the November unemployment rate required a 0.26 percentage point change to be statistically significant compared with a required change in September of 0.21 percentage point.
There are several reasons for this. Lower survey response (i.e., fewer households participating in the survey, see item 11) than usual and changes in statistical weighting (see items 6 and 7) both increased the variances associated with the November labor force estimates, resulting in higher thresholds for statistically significant changes.
Additionally, because October 2025 CPS data were not collected, the November 2025 statistical significance tables present the 2-month changes needed for statistically significant movements rather than the 1-month needed changes that are usually displayed. In general, comparisons over 2 months require larger changes to be statistically significant than comparisons over 1 month; this is not specific to November 2025.
Table: Multiplicative factors affecting the standard error for the unemployment rate in November 2025
| Lower response rate |
1.04 |
|
Changes to weighting
|
1.06 |
|
2-month change
|
1.12 |
|
Total impact in relation to September
|
1.23 |
The statistical significance tables for November 2025 (XLSX) (PDF) already incorporate these multiplicative factors in the calculation of the change needed to be statistically significant at a 90 percent confidence level. For example, the November unemployment rate required a 0.26 percentage point change to be statistically significant, about 1.23 times larger than the required change of 0.21 percentage point in September.
13. Will CPS produce annual estimates for 2025?
Yes. CPS will produce annual average estimates for 2025 by using the 11 months of available data, which excludes October. (CPS data were not collected in October due to the federal government shutdown.) As a result, 2025 annual estimates are not strictly comparable with annual averages for other years.
14. Will CPS produce quarterly estimates for the fourth quarter of 2025?
No. CPS data were not collected in October due to the federal government shutdown. Without one third of the data used for a quarterly estimate, CPS cannot produce reliable estimates for the fourth quarter of 2025.
15. How will the introduction of updated population controls in January be affected?
The annual population control adjustments that are usually incorporated with the release of January estimates in February will instead be introduced with the release of February 2026 estimates in March. Consequently, the initial January 2026 household survey estimates will continue to use short-term projections of monthly population estimates derived from population adjustments introduced in January 2025 (based on Vintage 2024 population estimates provided by the U.S. Census Bureau).
As soon as practicable, BLS plans to revise January 2026 estimates to incorporate the updated population controls. Additional information will be announced in the population control documentation.
16. How will December 2025 data be affected?
There was no impact on the collection of CPS data for December 2025. The Employment Situation news release will be published on January 9, 2026, as originally scheduled.
The CPS will return to the usual composite weighting methodology in December. (See item 6 above for changes in November.) The impact of the November weighting change (see item 7 above) on the standard error for the December unemployment rate will be negligible.
Due to complexities in the estimation process and the use of historical data, the production of December 2025 labor force status flows estimates will be delayed. Without the production of December flows estimates, the labor force status flows seasonally adjusted time series will not be subject to routine annual revisions. This standard revision will take place with the delayed production of the December labor force status flows.
Last modified date: January 8, 2026
United States Department of Labor