An official website of the United States government

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Bureau of Labor Statistics > Consumer Expenditure Survey > Methods

Summary of Research into 2017 Consumer Expenditure Survey Changes

Brett McBride

Division of Consumer Expenditure Surveys

This document provides the high level summaries of research into changes in the 2017 Consumer Expenditure Diary Survey and Interview Survey. Findings are summarized in terms of the changes’ impact on data, perceived or measured burden, cost (where applicable), and finally whether the overall goal of the change was met.

Change and intent	Data	Burden	Cost	Goal Met?
Full diary double-placement to reduce cost, simplify data collection procedure	There was no limit to the drop-off in entries across weeks in initial 2 months vs. prior year, but there were more diary entries in both weeks.	NA	Impact on costs was inconclusive.	Simplified procedures
Diary forms change to simplify details that previously made the diary task seem imposing for respondents	Minimal – although there was more reporting for section with more diary space, number of items reported were consistent.	NA	NA	Simplified form
EPD/LPD (earliest/latest diary placement date) restrictions removed to increase response rates and simplify the collection process	Provisional analysis showed no across-the-board decline in the share of seasonal items that might have been affected by change.	There was a 65% decline in diaries ‘placed too late’ in the first half of 2017 versus 2016, and response rates appeared to increase.	Not analyzed	Increased response
Interview construction section changes to reduce the burden on respondents	Analysis indicated changes in different directions in quarterly amounts reported for construction, but similar expenditure shares.	There was no overall change in section length (averaging 1.5 minutes) observed for those with positive durations in this section.	NA	Timing unchanged
Mortgage question changes to improve data quality by better capturing current mortgage information from respondents	New question version indicated about 40% of respondents asked new question in initial quarter had a refinanced mortgage, and reported mortgage origination dates were more recent.	NA	NA	Undetermined (many respondents had inventoried data instead of getting new question), but initial findings promising
Vehicle fuel question changes to only ask fuel expenditure follow-up questions for consumers with alternative vehicles	Similar (small) share of consumers reporting expenditures for these fuels.	There was an improvement by no longer asking fuel type questions for respondents without alternative fuel vehicles.	NA	Fuel questions asked of appropriate subset of respondents
Earned Income Tax Credit question removal to improve data collected on this source of income	Producing amounts via tax estimation process limited underreporting found with collected estimate.	Eliminated a question previously asked of respondents.	NA	Improved estimate
Transit questions revised to simplify the questions	Increased median amounts spent for transit were noted, signaling improved data quality.	Those reporting mass transit had section lengths that were, on average, 8 seconds longer in the 3 months following changes.	NA	Simplified question set
Long-term care insurance now classified with other health insurance questions	There was a slight year-over-year decline in proportion reporting LTC in 2nd quarter 2017 and lower average premiums were reported.	Question now more logically classified amid other types of health insurance.	NA	Included with health section
Health insurance changes intended to reduce measurement error	The incidence of misreported Medicare in private insurance decreased; however, there was a decrease in reporting of Medicare and Medicaid.	There was a slight increase in the proportion reporting ‘Don’t know’ or ‘Refused’ compared to the same months in 2016.	NA	Improvements in correct reporting of Medicare as public insurance
Entertainment question changes to capture more detail in reporting of movie and park/museum expenses	There were increases in the proportion reporting these expenses and amounts reported, year-over-year in quarter 2.	Respondents encountered one additional response option for which to report expenses.	NA	More data reported on entertainment expenses
Reference period changes to increase survey consistency among sections that ask total amount questions	Examination of equivalent months found decreased reporting for non-health insurance types, but the question itself was not significant when controlling for other factors. Amounts reported for mortgage pre-payments increased.	Results showed minor increases in the proportion reporting ‘Don’t Know’ and ‘Refusal’ responses for many types of insurance. For mortgage prepayments, there were lower missing data rates with the new questions.	NA	Greater consistency in how questions asked
Interview burden measurement to examine changes in time and responses to burden questions	NA	Mixed findings; the length of the interview did increase in 2017, particularly due to an increase in the health care section, but perceived burden and the proportion saying survey was ‘too long’ decreased when compared with questions asked in 2013.	NA	Respondents reporting feeling less burdened despite increase in average interview time

Last Modified Date: March 12, 2019