Technical Notes
The principal objective of the Consumer Expenditure Surveys (CE) is to collect information on the buying habits of U.S. consumers. Consumer expenditure data are used in a variety of research activities by government, business, labor, and academic analysts. In addition, the data are required for the regular revision of the Consumer Price Index (CPI) market basket.
First collected in 1888, the CE was conducted approximately every 10 years through the 1972–73 survey; annual collection began in late 1979. The CE is conducted by the U.S. Census Bureau for the Bureau of Labor Statistics (BLS). The CE consists of two surveys, each with independent samples that were first implemented in the 1972–73 series. There is a Diary Survey completed by participating consumer units for two consecutive 1-week periods, and an Interview Survey in which expenditures of consumer units are obtained in four interviews over the course of 10 months, conducted at 3-month intervals. The results in this report are based on integrated data from both surveys. Survey participants record dollar amounts for goods and services purchased during the reporting period, regardless of whether payment is made at the time of purchase. Expenditure amounts include all sales and excise taxes for items purchased by the consumer unit. All business-related expenditures are excluded from both surveys, as are most items for which the consumer unit is reimbursed. However, the Interview Survey collects information on insurance reimbursements for medical care costs.
For the Diary Survey, the CE Program draws a sample of 12,000 addresses per year, with approximately 10,000 of those addresses found to be occupied housing units. The other 2,000 addresses are generally nonexistent, nonresidential, or vacant. Approximately 6,000 of those occupied housing units respond to the survey, and each of those households completes two weekly diaries for a total of 12,000 weekly diaries per year. Likewise, for the Interview survey, the CE Program visits 12,000 addresses per quarter, with 10,000 of those addresses having occupied housing units. Then approximately 6,000 of those occupied housing units respond to the survey for a total of 6,000 interviews per quarter. Each consumer unit is interviewed once per quarter, for four consecutive quarters. With the rotating panel, some consumer units rotate out of the survey each quarter, while others rotate in. Data from both the Diary and Interview Surveys are collected on an ongoing basis in 91 geographic areas of the United States.
The Diary Survey is designed to capture expenditures on small, frequently purchased items that are normally more difficult for respondents to recall buying over an extended period. Respondents keep detailed entries of expenses for food and beverages—both for food at home (e.g., purchased from grocery stores, convenience stores, farmers markets) and food away from home (e.g., purchased from restaurants, carry outs, employer and school cafeterias)—and for tobacco, housekeeping supplies, nonprescription drugs, personal care products and services, most apparel, and any other goods or services purchased within the period. Although this survey was designed to collect information on expenditures that could not be recalled easily over an extended period, respondents are asked to report all expenses, except those spent while traveling overnight or longer, that the consumer unit incurs during the survey week.
The Interview Survey is designed to capture expenditure data that respondents can reasonably recall for a period of 3 months or longer. In general, these expenditures are relatively large, such as expenditures for real property, automobiles, and major appliances, or they occur on a regular basis, such as rent, utility payments, and insurance premiums. The Interview Survey also collects data on expenditures incurred on leisure trips. Including broad estimates of food spending, it is estimated that about 95 percent of expenditures are covered in the Interview Survey. Nonprescription drugs, household supplies, and personal care items are excluded.
Integrated data from the Diary and Interview Surveys provide a complete accounting of consumer expenditures and income that neither survey is designed to do on its own. One reason is that data on some expenditure items are collected in only one of the surveys. For example, the Diary Survey does not collect information on reimbursements, whereas the Interview Survey does. Examples of expenditures for which reimbursements are included in the Interview Survey are medical care; automobile repairs; and construction, repairs, alterations, and maintenance of property. For items that are unique to one survey or the other, the choice of which survey to use as the source of data is obvious. However, there is considerable overlap in coverage between the surveys, and integrating the data involves a determination of the appropriate survey from which to select specific expenditure items. When data are available from both surveys, the CE program uses statistical methods to determine which survey is more reliable. In this way, some items are selected from the Interview Survey, and others are selected from the Diary Survey. The CE program reviews the survey source every 2 years for this purpose. See “CE Source Selection for Publication Tables” in the Consumer Expenditure Survey Anthology, 2011 (BLS Report, no. 1030), for source selection details.
As noted above, the CPI market basket is revised in accordance with results from the CE. However, the population coverage of the CE differs from the CPI. The CE data cover the total population including rural areas, whereas the CPI covers only the population in urban and metropolitan areas. With the latest CPI geographic revisions, efforts are being made to align the urban areas sampled by the CPI and the CE.
Definitions of components also differ between the CE and CPI. For example, homeownership is treated differently in the two surveys: actual expenditures of homeownership are reported in the CE, whereas the CPI uses a rental equivalence approach that estimates the change in the cost of obtaining, in the rental marketplace, services equivalent to those provided by owner-occupied homes. (For an overview of the CPI methodology, see the Handbook of Methods section on the Consumer Price Index.)
Interpreting the data
Expenditures are averages for consumer units with specified characteristics, regardless of whether a particular unit incurred an expense for a specific item during the recordkeeping period. Therefore, the average expenditure shown for an item may be considerably lower than the average expenditure by the consumer units that purchased the item. The less frequently an item is purchased, the greater the difference between the average for all consumer units and the average for those purchasing the item. Similarly, an individual consumer unit may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family members, and geographic location influence expenditures. In addition, even within groups with similar characteristics, the distribution of expenditures varies substantially, as personal preference, prices, and other factors not collected in the CE (e.g., health status of consumer unit members) also influence expenditures. These points should be considered when comparing reported averages with the individual circumstances of any particular consumer unit of interest. Users of these survey data should also keep in mind the data reflect conditions at the time they were collected, which may be different than current circumstances. For example, prices may have changed. All prices, as measured by the CPI-U, increased 1.8 percent from 2018 to 2019 (annual average index); undoubtedly, some prices rose more than the average (1.8 percent), some rose less, some stayed the same, and some have even declined. In addition, sample surveys are subject to two types of error: sampling and nonsampling. Sampling errors occur because the data are collected from a sample representing the population, rather than from the entire population. Therefore, the mean of the sample may differ from the mean that would be obtained if data from the entire population were available. Nonsampling errors result from the inability or unwillingness of respondents to provide correct information, differences in interviewers’ abilities, mistakes in recording or coding, and other processing errors. For additional information on these types of errors see sampling and nonsampling errors.
Standard tables, 1960–61, 1972–73, 1984–2019
Tables in this report include integrated data from both the Diary and Interview components of the CE, enabling data users to associate the full range of expenditures with consumers’ demographic characteristics. Users can access the CE tables beginning with 1960–61, which have more detail than is given in this report. Also available are tables showing average annual data over a 2-year period for
-
income before taxes, cross-tabulated by age, consumer unit size, or region;
-
single consumers by gender, cross-tabulated by either income or age; and
-
selected metropolitan statistical areas (MSAs).
Detailed tables, 1984–2019
In addition, the CE also produces detailed prepublication tables, which contain additional sub-categories of spending by demographic characteristics, but also have larger variances. They are available by request; email: CEXInfo@bls.gov. Prepublication tables at the all-consumer unit level are available on the CE research products web page.
Public-use microdata
The 2019 CE public-use microdata, including Interview Survey data, Diary Survey data, and paradata (data about the survey process), are available on the CE website for electronic download. The Interview Survey files contain expenditure data in two different formats: MTBI files that present monthly values by coding items based on the CPI pricing scheme, and expenditure files that organize expenditures by the section of the Interview Survey questionnaire in which they are collected. Values on the expenditure files cover different periods depending on the specific questions asked, and the files also contain relevant non-expenditure information not found on the MTBI files.
Beginning with the 2013 public-use microdata, new estimates of local, state, and federal tax liabilities are included. The CE introduced these estimates to improve the quality of surveyed tax liabilities, which suffered from low item response. The respondent-provided tax data were still available in the public-use microdata for 2013, after which they were no longer collected. For more information on the improvements, see Improving data quality in the Consumer Expenditure Survey with TAXSIM, and for new applications of the tax estimates, see New estimates of personal taxes in the Consumer Expenditure Survey. Free public-use microdata are available at PUMD data files from the years 1980 to 2019.
Other survey information
Other survey information available on the website includes answers to frequently asked questions and a glossary of terms. Beginning with the 2000 data, estimates of standard errors for integrated Diary and Interview Survey data are also available.
The CE program also publishes expenditure data within the context of various topics of interest in Beyond the Numbers. As of December 2020, recent articles are: How have healthcare expenditures changed? Evidence from the Consumer Expenditure Surveys and How does consumer spending differ among households in California, Texas, and New York? A new BLS data product can tell us. Additional CE articles are presented in the Monthly Labor Review (MLR). As of December 2020, the recent analytical MLR articles using CE data are entitled, “Comparing characteristics and selected expenditures of dual- and single-income households with children,” and “Not fun for young and old alike: how the youngest and oldest consumers have fared in recession and recovery.” In addition, the CE research library includes general articles and research papers using CE data, including documents.
Information on the methodology used to calculate and collect CE data, including the CE data quality profile, can be found in the CE Handbook of Methods.
Upcoming events
Each year, the CE program sponsors a symposium and a microdata users’ workshop. Both events are free, although registration is required. The CE Symposium focuses on survey methodology, and typically features presenters from the CE program, other BLS programs, and researchers who are not affiliated with BLS. The symposium typically meets for 1 day. The CE Microdata Users’ Workshop is a 3-day event that combines practical “hands-on” training in use of the data (Interview and Diary), with presentations from users who are not affiliated with BLS. The practical training builds from an introduction to the data files and structure on the first day to expert topics on the final day. Presentations from researchers are selected from those who answer the call for presenters. As with registration, there is no fee for submitting a proposal to present in answer to the call. Both events are held in the same week of July. The next CE Symposium will be held on July 20, 2021. The next CE Microdata Users’ Workshop will be held on July 21-23, 2021. More information about these events is available on the CE website (CE Survey Methods Symposium and Microdata Users’ Workshop). Reports on these events are also published in the Monthly Labor Review (MLR). The most recent report available at the time of publication of this news release describes the 2019 events (Consumer Expenditure Surveys Methods Symposium and Microdata Users’ Workshop, July 16–19, 2019). Reports on earlier events are available on the CE MLR publications webpage (Monthly Labor Review articles by CE staff and others using CE data). Reports on subsequent events (2020 onward) will be posted at a later date.
Contact information
For more detailed information on the availability of current and earlier data, contact the Division of Consumer Expenditure Surveys, Office of Prices and Living Conditions, Bureau of Labor Statistics, 2 Massachusetts Avenue, NE, Washington, DC 20212-0001; call (202) 691-6900; email: CEXInfo@bls.gov; or visit the Consumer Expenditure Surveys page. Information in this publication is in the public domain and, with appropriate credit, may be reproduced without permission. Upon request, the information in this report will be made available to individuals who are sensory-impaired. Voice phone: (202) 691-5200. Federal Relay Service: 1-800-877-8339.
United States Department of Labor
