2025 Federal Government Shutdown Impact on Consumer Expenditure Surveys (CE) and Consumer Price Index (CPI)

During the 2025 lapse in federal appropriations, the Bureau of Labor Statistics (BLS) could not collect Consumer Price Index (CPI) data, and the Census Bureau (Census) could not collect Consumer Expenditure (CE) data. This page summarizes information related to the impact of the missing data on BLS data products.

BLS did not collect CPI data from October 1, 2025, through November 12, 2025. Census conducts CE data collection for BLS and did not collect CE data in October and November 2025. Census resumed collection of the CE on December 5, 2025.

Missing CPI data affected October and November 2025 indexes. BLS described the effects of the 2025 federal government shutdown on the CPI. Missing October 2025 data also affected April 2026 rent and owner’s equivalent rent indexes, as described in the Monthly Labor Review article, “Counterfactual imputation approaches for the housing component of the October 2025 CPI.”

Missing CE data in October and November 2025 affects BLS estimates in planned publications, including the 2025 annual CE release, final revisions of the 2025 Chained CPI-U indexes, and 2027 CPI-U and CPI-W indexes. To mitigate the impact on data quality of these estimates, BLS considered several approaches to address the missing data. On May 1, 2026, BLS shared a summary of approaches under review based on data simulations and operational considerations.

BLS engaged the National Association for Business Economics (NABE) and the Committee on National Statistics (CNSTAT) to conduct expert panels regarding the approaches. Both expert panel meetings took place in May 2026. They were open to the public and recorded. NABE and CNSTAT provided BLS summaries of the panel discussion and recommendations.

• NABE expert panel:
  • May 7, 2026
  • Recording and meeting summary notes are available through NABE

• CNSTAT expert panel:
  • May 8, 2026
  • Recording and meeting summary notes are available through CNSTAT

Based on BLS analysis and panel feedback, BLS selected the survey weight adjustment factor approach to prioritize the timely release of planned publications. In June 2026, BLS informed the public of these plans in the Monthly Labor Review article, “Addressing missing consumer expenditure data due to the 2025 lapse in appropriations.” All estimates using the 2025 CE data are expected to remain on schedule. BLS will summarize information for data users on the BLS website prior to each release of impacted CE and CPI data.

NABE Panel

NABE provided BLS with a summary of the discussion:

The BLS/NABE meeting focused on alternative approaches the BLS might employ to develop estimates for Consumer Expenditure (CE) Survey data that were not collected during the government shutdown. These data are necessary both for the publication of annual CE estimates and for the development of expenditure weights used in calculating the Consumer Price Index (CPI). The BLS presented several alternatives for each program, outlining the advantages and disadvantages of each approach. Following a question-and-answer session with the BLS, the NABE panel considered the options in a private meeting and subsequently presented the following conclusions to the BLS.

The panel unanimously agreed that, given the BLS’s limited resources, priority should be placed on future operations and data continuity. Accordingly, the selected approaches should avoid imposing substantial additional costs or operational burdens in pursuit of only marginal gains in accuracy, and they should not increase the risk of delays in data publication. The panel also emphasized that the selected methodologies should be communicated publicly on BLS.gov well in advance of data release and accompanied by full methodological transparency.

Consumer Expenditure Survey

The panel agreed with the BLS that weight adjustment represented the preferred approach for estimating missing observations in the CE Survey. One panelist suggested that the BLS explore the feasibility of utilizing interview data to supplement missing diary data for items collected through both interview and diary methods. In addition, the panel encouraged the BLS to continue its research into domain estimation; however, the panel agreed that this method should not be implemented at this time due to potential risks to timeliness and additional costs.

Consumer Price Index Programs

The BLS presented six alternatives for addressing the needs of the CPI program. After extensive discussion, the panel agreed that the final approach must prioritize both data timeliness and the production of reliable estimates. The panel also emphasized the importance of preserving seasonality whenever possible.

One panel member demonstrated that the strongest simulation results were achieved through a combination of three alternatives: weight adjustment, year-over-year inflation adjustment, and linear interpolation. However, given the potential implementation delays associated with linear interpolation, the panel recommended adopting a combination of the weight adjustment and year-over-year inflation adjustment for the CPI calculation, with linear interpolation incorporated only if doing so would not affect the production timeline.

The panel reiterated the importance of clear public communication regarding any methodological changes, emphasizing that government shutdowns can materially affect statistical reporting. The methods adopted to address missing data should therefore be transparent and communicated well in advance to allow data users sufficient time to incorporate the changes into their models and analyses.

Other Recommendations

• Communications posted to the BLS website should include both an executive summary and a detailed technical discussion to meet the needs of a broad range of data users.
• The BLS should return to its prior practice of releasing seasonal factors and weights two days before the publication of official data.
• Publication of the experimental housing data (New Tenant Rent Index) should resume within the next four to five months, after methodological improvements have been implemented and adequately explained to users.
• The BLS should continue to advocate for and pursue additional appropriations to support IT infrastructure investments and operational flexibility.
• The BLS should consider technology solutions that improve the timeliness and accessibility of public communications.

CNSTAT Panel

CNSTAT provides a summary of closed session discussion:

• CPI-U, CPI-E, CPI-W are different from Chained CPI.
  • C-CPI is not used in as many external products, so there is more leeway to do something more creative, uses monthly weights.
  • CPI-U, CPI-E, CPI-W annual weights.
  • CPI-U, CPI-E, CPI-W feed into BEA directly, and the Federal Reserve directly and indirectly.
  • C-CPI 12-month average (September to August) is used to set tax brackets.
• Some CPI indexes need to be done quickly (annual data and CPI-U) and cannot be delayed.
• Absence of diary data has more effect on C-CPI.
• The approach of weights adjustment seems to do well and is reasonable (but should think more about how to create the weights).
• What are the measures of uncertainty? What are the costs with respect to uncertainty of approaches?
• Use weight adjustment to get the level of expenditure right.
• If we want accuracy of C-CPI for highly seasonal goods (e.g., turkeys), then we need something more sophisticated like unit imputation or the Echo State Network model.
• Absence of price data in October and November helps frame the importance of how to treat the expenditure weights.
• Longer term investments:
  • BLS’ system limits the timely ability to implement other strategies, e.g., update/modify the diaries data.
  • Strongly suggest considering adjusting the systems to support this, make it possible to go back to fill in data.
• Thoughts for “insurance” for the future:
  • Survey collection design: If we distributed diaries early enough in advance, we could tell sampled households to fill out for a specific week in the future.
  • External data sources: credit card transactions and other data sources. These could be expensive but would be particularly helpful for diary survey.

Last Modified Date: June 30, 2026