When the 43-day government shutdown finally ended on November 12, 2025 the BLS (Bureau of Labor Statistics)—an agency that typically works quietly behind the scenes—was suddenly back in the national spotlight. For weeks, analysts, policymakers, and reporters had been operating without any fresh labor-market data from the BLS: no payroll jobs, no unemployment figures, no wage or information on hours worked, and no indication of whether the economy was cooling or heating up. Now that the shutdown has lifted, a big concern is what the BLS will do about the missing data it did not collect or process during the long shutdown.
Two headline numbers drive that urgency: the job count and the unemployment rate. Although they are scheduled to be published simultaneously each month in The Employment Situation Report, they come from two different surveys—the Establishment Survey (CES), which measures payroll employment, and the Household Survey (CPS), which produces the unemployment rate. Understanding how these surveys function helps explain why a prolonged shutdown crippled one dataset more severely than the other.
What Are the CPS and CES?
The Household Survey, formally the Current Population Survey (CPS), interviews roughly 60,000 households each month to measure labor-force status among the civilian noninstitutional population age 16 and older. The survey is conducted by the Census Bureau using a 4-8-4 rotation system: households are interviewed for four consecutive months, leave the sample for eight months, and then return for another four for the following year. All questions focus on a strict reference period—the calendar week that includes the 12th day of the month. Interviewers ask individuals whether they worked during that week, how many hours they worked, whether they actively sought employment, and why they were unemployed if they did not work. Because each response is tied precisely to that reference week, the timing of data collection is fundamental to the survey’s integrity.
The Establishment Survey, known as the Current Employment Statistics (CES) program, collects data directly from employers. It surveys about 121,000 businesses and government agencies, covering roughly 631,000 worksites. Like CPS, CES also anchors its statistics to the pay period that contains the 12th of the month. Employers report payroll counts, total hours worked, and wage information. But unlike CPS, CES submissions arrive through automated payroll feeds, secure online systems, email, and in some cases telephone reporting, making it far less dependent on in-person field operations.
How the Government Shutdown Disrupted Data Collection
A prolonged federal shutdown severely disrupted the Household Survey because CPS relies on Census Bureau field staff to conduct interviews each month. When these staff are furloughed, interviewing simply stops. That means CPS did not collect responses tied to the reference week, the week that defines unemployment status for the entire month. Missing this narrow window lowers response rates, breaks the rotation structure, and increases reliance on imputation, all of which diminish the accuracy and consistency of unemployment estimates. During this shutdown, Census interviewers were unable to contact households, so the unemployment numbers for both September 2025 (reference week October 12th) and October 2025 (reference week November 12) were not collected.
By contrast, the Establishment Survey continued to receive much of its data because CES relies heavily on automated payroll systems and electronic reporting pipelines. Even during disruptions, these submissions continue flowing into BLS data systems. Once BLS staff return to work, they can in principle process the backlog and publish payroll employment figures. But unemployment statistics—derived solely from household responses—cannot easily or practically be produced retroactively if interviews never occurred.
The BLS has not yet announced what it will do with respect to missing data from the long shutdown period. Many data collection programs were halted and the impact of the shutdown will be widespread. Basically, both the CES - the establishment survey – and the CPS – the household survey – were compromised by the shutdown, but it is easier in many ways for the establishment survey to produce statistics retroactively for the missing months. Conducting household surveys retroactively is fraught with additional difficulties and will likely lead to a decision to simply not calculate the unemployment rate for the months not surveyed on schedule.