Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XVI. APPENDIX 5). The microdata are available on CD-ROMs. These microdata files present detailed expenditure and income data from the Interview component of the CE for 2004 and the first quarter of 2005. The Interview survey collects data on up to 95 percent of total household expenditures. In addition to the FMLI, MEMI, MTBI, ITBI, and ITBI_IMPUTED1 files, the microdata include files created directly from the expenditure sections of the Interview survey (EXPN files). The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLI or MTBI files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Estimatesof average expenditures in 2004from the Interview Survey, integrated with data from the Diary Survey, will bepublished in the report Consumer Expenditures in 2004(due out in 2006). A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics,

Analysis unit

Consumer Units

Kind of data

Sample survey data [ssd]

Mode of data collection

Computer Assisted Personal Interview [capi]

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index.The CE program consists of two surveys, the quarterly Interview Survey and the Diary Survey (ICPSR 29883). The quarterly Interview survey is designed to collect data on major items of expense which respondents can be expected to recall for 3 months or longer. These include relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occur on a regular basis, such as rent or utilities. The Interview survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures.The microdata in this collection are available as SAS, STATA, SPSS data sets or ASCII text and comma-delimited files. The 2009 Interview release contains seven groups of Interview data files (FMLY, MEMB, MTAB, ITAB, ITAB_IMPUTE, FPAR, and MCHI), 50 EXPN files, and processing files.The FMLY, MEMB, MTAB, ITAB, and ITAB_IMPUTE files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly data sets for each of these files, running from the first quarter of 2009 through the first quarter of 2010. The FMLY file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMB file contains member characteristics and income data; the MTAB file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITAB file contains income data converted to a monthly time frame and assigned to UCCs; and the ITAB_IMPUTE file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs.The FPAR and MCHI datasets are grouped as 2-year datasets (2008 and 2009), plus the first quarter of the 2010. The FPAR file contains CU level data about the Interview survey, including paradata collected about the interview within the interview collection instrument (CAPI). This data includes information on the amount of time required to collect each interview and interview section, as well as other interviewer entered information about the resulting survey. The MCHI file contains data about each interview contact attempt, including reasons for refusal and times of contact. Both FPAR and MCHI files contain five quarters of data.Each of the 50 EXPN files contains five quarters of data. The EXPN files contain data directly derived from their respective questionnaire sections.The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. The processing files are: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), (2) a UCC file that contains UCCs and their abbreviated titles, identifying the expenditure, income, or demographic item represented by each UCC, (3) a vehicle make file (CAPIVEHI), and (4) files containing sample programs. The processing files are further explained in the Interview User Guide, Section III.F.6. PROCESSING FILES. There is also a second user guide, "User's Guide to Income Imputation in the CE", which includes information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships each CU member. Income information, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over was also collected.

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates (for consumer units or CUs) of average expenditures in news releases, reports, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (see Section XVI. Appendix 5). The microdata are available on CD-ROMs. These microdata files present detailed expenditure and income data for the Diary component of the CE for 2010. They include weekly expenditure (EXPN), annual income (DTAB) files, and imputed income files (DTID). The data in EXPN, DTAB, and DTID files are categorized by a Universal Classification Code (UCC). The advantage of the EXPN and DTAB files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLY and MEMB files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLY files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2010 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2010. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: “U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2010”.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2010 sample is composed of 91 areas. The design classifies the PSUs into four categories: • 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 38 "X" PSUs, are medium-sized MSAs. • 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. • 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2010 survey is generated from the 2000 Population Census file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XV. APPENDIX 4). The microdata are available online at http://www/bls.gov/cex/pumdhome.htm. These microdata files present detailed expenditure and income data for the Diary component of the CE for 2002. They include weekly expenditure (EXPD) and annual income (DTBD) files. The data in EXPD and DTBD files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2002 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2002. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2002".

STATE IDENTIFIER Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data. To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

INTERPRETING THE DATA

Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially. Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs. CUs with members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates (for consumer units or CUs) of average expenditures in news releases, reports, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (see Section XVI. Appendix 5). These microdata files present detailed expenditure and income data for the Diary component of the CE for 2005. They include weekly expenditure (EXPD), annual income (DTBD) files, and imputed income files (DTID). The data in EXPD, DTBD, and DTID files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2005 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2005. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: “U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2005”.

State Identifier Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data. To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

Interpreting the data Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family Members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially.
Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs.CUs with Members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XVI. APPENDIX 5). The microdata are available on CD-ROMs. These microdata files present detailed expenditure and income data from the Interview component of the CE for 2007 and the first quarter of 2008. The Interview survey collects data on up to 95 percent of total household expenditures. In addition to the FMLY, MEMB, MTAB, and ITAB_IMPUTE files, the microdata include files created directly from the expenditure sections of the Interview survey (EXPN files). The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLY or MTAB files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Estimates of average expenditures in 2007 from the Interview Survey, integrated with data from the Diary Survey, will be published in the report Consumer Expenditures in 2007 (due out in 2009). A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Interview Survey, 2007."

Analysis unit

Consumer Units

Kind of data

Sample survey data [ssd]

Sampling procedure

Samples for the CE are national probability samples of households designed to be representative of the total U.S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2007 and 2008 samples is composed of 91 areas. The design classifies the PSUs into four categories: • 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 38 "X" PSUs, are medium-sized MSA's. • 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. • 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2007 survey is generated from the 2000 Census of Population 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. Interviewers are then assigned to list these areas before a sample is drawn. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. The Interview Survey is a panel rotation survey. Each panel is interviewed for five consecutive quarters and then dropped from the survey. As one panel leaves the survey, a new panel is introduced. Approximately 20 percent of the addresses are new to the survey each month.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XVI. APPENDIX 5). The microdata are available on CD-ROMs. These microdata files present detailed expenditure and income data from the Interview component of the CE for 2006 and the first quarter of 2007. The Interview survey collects data on up to 95 percent of total household expenditures. In addition to the FMLY, MEMB, MTAB, and ITAB_IMPUTE files, the microdata include files created directly from the expenditure sections of the Interview survey (EXPN files). The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLY or MTAB files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Estimates of average expenditures in 2006 from the Interview Survey, integrated with data from the Diary Survey, will be published in the report Consumer Expenditures in 2006 (due out in 2008). A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Interview Survey, 2006."

Analysis unit

Consumer Units

Kind of data

Sample survey data [ssd]

Sampling procedure

The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2006 and 2007 samples is composed of 91 areas. The design classifies the PSUs into four categories: • 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 38 "X" PSUs, are medium-sized MSA's. • 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. • 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI. The sampling frame (that is, the list from which housing units were chosen) for the 2006 survey is generated from the 2000 Census of Population 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good Samples for the CE are national probability samples of households designed to be representative of the total U.S. civilian population. Eligible population includes all civilian noninstitutional persons. addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. Interviewers are then assigned to list these areas before a sample is drawn. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. The Interview Survey is a panel rotation survey. Each panel is interviewed for five consecutive quarters and then dropped from the survey. As one panel leaves the survey, a new panel is introduced. Approximately 20 percent of the addresses are new to the survey each month.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure (CE) program provides a continuous and comprehensive flow of the data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of the revision of the Consumer Price Index. To meet the needs of users, The Bureau of Labor Statistics produces population estimates (for consumer units or Cu's) of average expenditure in new releases, reports and articles in the Monthly Labour review. Tabulated CE data are also available on the internet and by facsimile transmission. These microdata files present detailed expenditure and income data for the Diary component of the CE for 2004. They include weekly expenditure (EXPD), annual income (DTBD) files, and imputed income files (DTBD_IMPUTED1). The data in EXPD, DTBD, and DTBD_IMPUTED files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU Members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2004 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2004 (Due in 2006).A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: “U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2004”.

State Identifier Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data. To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

Interpreting the data Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family Members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially.
Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs.CUs with Members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer unit

Kind of data

Sample survey data [ssd]

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XV. APPENDIX 4). The microdata are available online at http://www/bls.gov/cex/pumdhome.htm.

These microdata files present detailed expenditure and income data for the Diary component of the CE for 2002. They include weekly expenditure (EXPD) and annual income (DTBD) files. The data in EXPD and DTBD files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files.

Estimates of average expenditures in 2002 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2002. A list of recent publications containing data from the CE appears at the end of this documentation.

The microdata files are in the public domain and with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2002".

Analysis unit

Consumer Units

Kind of data

Sample survey data [ssd]

Sampling procedure

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2002 sample is composed of 105 areas. The design classifies the PSUs into four categories: • 31 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 46 "B" PSUs, are medium-sized MSA's. • 10 "C" PSUs are nonmetropolitan areas that are included in the CPI. • 18 "D" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2002 survey is generated from the 1990 Population Census 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (ED's) from the Census that fail to meet the criterion for good addresses for new construction, and all ED's in nonpermit-issuing areas are grouped into the area segment frame. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year. During the last 6 weeks of the year, however, the Diary Survey sample is supplemented to twice its normal size to increase the reporting of types of expenditures unique to the holidays.

STATE IDENTIFIER Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data. To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

INTERPRETING THE DATA Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially. Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs. CUs with members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract (en): The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index.The CE program consists of two surveys, the quarterly Interview Survey and the Diary Survey (ICPSR 29883). The quarterly Interview survey is designed to collect data on major items of expense which respondents can be expected to recall for 3 months or longer. These include relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occur on a regular basis, such as rent or utilities. The Interview survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures.The microdata in this collection are available as SAS, STATA, SPSS data sets or ASCII text and comma-delimited files. The 2009 Interview release contains seven groups of Interview data files (FMLY, MEMB, MTAB, ITAB, ITAB_IMPUTE, FPAR, and MCHI), 50 EXPN files, and processing files.The FMLY, MEMB, MTAB, ITAB, and ITAB_IMPUTE files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly data sets for each of these files, running from the first quarter of 2009 through the first quarter of 2010. The FMLY file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMB file contains member characteristics and income data; the MTAB file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITAB file contains income data converted to a monthly time frame and assigned to UCCs; and the ITAB_IMPUTE file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs.The FPAR and MCHI datasets are grouped as 2-year datasets (2008 and 2009), plus the first quarter of the 2010. The FPAR file contains CU level data about the Interview survey, including paradata collected about the interview within the interview collection instrument (CAPI). This data includes information on the amount of time required to collect each interview and interview section, as well as other interviewer entered information about the resulting survey. The MCHI file contains data about each interview contact attempt, including reasons for refusal and times of contact. Both FPAR and MCHI files contain five quarters of data.Each of the 50 EXPN files contains five quarters of data. The EXPN files contain data directly derived from their respective questionnaire sections.The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. The processing files are: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), (2) a UCC file that contains UCCs and their abbreviated titles, identifying the expenditure, income, or demographic item represented by each UCC, (3) a vehicle make file (CAPIVEHI), and (4) files containing sample programs. The processing files are further explained in the Interview User Guide, Section III.F.6. PROCESSING FILES. There is also a second user guide, "User's Guide to Income Imputation in the CE", which includes information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships each CU member. Income information, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over was also collected. Refer to the Interview User Guide documentation for a detailed explanation of the weight variables used. Eligible population includes all civilian noninstitutional persons. National probability sample of households designed to represent the total United States noninstitutionalized civilian population. 2011-01-03 A folder containing addendum data files (CES09_interview_addendum_files) has been added to the zipped package. The addendum files correct for a processing error that affected several records in the 2009Q3 Interview MEMB file. compu...

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XV. APPENDIX 4). The microdata are available online at http://www/bls.gov/cex/pumdhome.htm. These microdata files present detailed expenditure and income data from the Interview component of the CE for 2003 and the first quarter of 2004. The Interview survey collects data on up to 95 percent of total household expenditures. In addition to the FMLI, MEMI, MTBI, and ITBI files, the microdata include files created directly from the expenditure sections of the Interview survey (EXPN files). The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLI or MTBI files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Estimates of average expenditures in 2003 from the Interview Survey, integrated with data from the Diary Survey, will be published in the report Consumer Expenditures in 2003. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Interview Survey, 2003."

Analysis unit

Consumer Units

Kind of data

Sample survey data [ssd]

Sampling procedure

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian non-institutionalized persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2003 and 2004 samples is composed of 105 areas. The design classifies the PSUs into four categories: • 31 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 46 "B" PSUs, are medium-sized MSA's. • 10 "C" PSUs are nonmetropolitan areas that are included in the CPI. • 18 "D" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2003 and 2004 surveys is generated from the 1990 Census of Population 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in non-permit-issuing areas are grouped into the area segment frame. Interviewers are then assigned to list these areas before a sample is drawn. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. The Interview Survey is a panel rotation survey. Each panel is interviewed for five consecutive quarters and then dropped from the survey. As one panel leaves the survey, a new panel is introduced. Approximately 20 percent of the addresses are new to the survey each month.

WEIGHTING Each CU included in the CE represents a given number of CUs in the U.S. population, which is considered to be the universe. The translation of sample families into the universe of families is known as weighting. However, since the unit of analysis for the CE is a CU, the weighting is performed at the CU level. Several factors are involved in determining the weight for each CU for which an interview is obtained. There are four steps in the weighting procedure: 1) The basic weight is assigned to an address and is the inverse of the probability of selection of the housing unit. 2) A weight control factor is applied to each interview if subsampling is performed in the field. 3) A noninterview adjustment is made for units where data could not be collected from occupied housing units. The adjustment is performed as a function of region, housing tenure, family size and race. 4) A final adjustment is performed to adjust the sample estimates to national population controls derived from the Current Population Survey. The adjustments are made based on both the CU's Member composition and the CU as a whole. The weight for the CU is adjusted for individuals within the CU to meet the controls for 14 age/race categories, 4 regions, and 4 region/urban categories. The CU weight is also adjusted to meet the control for total number of CUs and total number of CUs who own their living quarters. The weighting procedure uses an iterative process to ensure that the sample estimates meet all the population controls. NOTE: The weight for a consumer unit (CU) can be different for each quarter in which the CU participates in the survey, as the CU may represent a different number of CUs with similar characteristics.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XV. APPENDIX 4). The microdata are available online at http://www/bls.gov/cex/pumdhome.htm.

These microdata files present detailed expenditure and income data for the Diary component of the CE for 2003. They include weekly expenditure (EXPD) and annual income (DTBD) files. The data in EXPD and DTBD files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (or income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files.

Estimates of average expenditures in 2003 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2003. A list of recent publications containing data from the CE appears at the end of this documentation.

The microdata files are in the public domain and with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2003".

STATE IDENTIFIER

Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data.

To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

INTERPRETING THE DATA

Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially.

Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs. CUs with members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

A. SURVEY SAMPLE DESIGN

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons.

The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2003 sample is composed of 105 areas. The design classifies the PSUs into four categories:

• 31 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 46 "B" PSUs, are medium-sized MSA's. • 10 "C" PSUs are nonmetropolitan areas that are included in the CPI. • 18 "D" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2003 survey is generated from the 1990 Population Census 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (ED's) from the Census that fail to meet the criterion for good addresses for new construction, and all ED's in nonpermit-issuing areas are grouped into the area segment frame.

To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance.

Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year. During the last 6 weeks of the year, however, the Diary Survey sample is supplemented to twice its normal size to increase the reporting of types of expenditures unique to the holidays.

B. COOPERATION LEVELS

The annual target sample size at the United States level for the Diary Survey is 7,800 participating sample units. To achieve this target the total estimated work load is 11,275 sample units. This allows for refusals, vacancies, or nonexistent sample unit addresses.

Each participating sample unit selected is asked to keep two 1-week diaries. Each diary is treated independently, so response rates are based on twice the number of housing units sampled.

Mode of data collection

Computer Assisted Personal Interview [capi]

Response rate

The response rate for the 2003 Diary Survey is 73.4%. This response rate refers to all diaries in the year.

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index.The CE program is comprised of two separate components (each with its own questionnaire and independent sample), the quarterly Interview Survey and the Diary Survey (ICPSR 32482). This data collection contains the quarterly Interview Survey data, which was designed to collect data on major items of expense which respondents could be expected to recall for 3 months or longer. These included relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occurred on a regular basis, such as rent or utilities. The Interview Survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures.The microdata in this collection are available as SAS, STATA, SPSS data sets or ASCII text and comma-delimited files. The 2010 Interview Survey release contains seven groups of Interview data files (FMLY, MEMB, MTAB, ITAB, ITAB_IMPUTE, FPAR, and MCHI), 50 EXPN files, and processing files.The FMLY, MEMB, MTAB, ITAB, and ITAB_IMPUTE files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly data sets for each of these files, running from the first quarter of 2010 through the first quarter of 2011. The FMLY file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMB file contains member characteristics and income data; the MTAB file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITAB file contains income data converted to a monthly time frame and assigned to UCCs; and the ITAB_IMPUTE file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs.The FPAR and MCHI datasets are grouped as 2-year datasets (2009 and 2010), plus the first quarter of the 2011. The FPAR file contains CU level data about the Interview survey, including paradata collected about the interview within the interview collection instrument (CAPI). This data includes information on the amount of time required to collect each interview and interview section, as well as other interviewer entered information about the resulting survey. The MCHI file contains data about each interview contact attempt, including reasons for refusal and times of contact. Both FPAR and MCHI files contain five quarters of data.Each of the 50 EXPN files contains five quarters of data. The EXPN files contain data directly derived from their respective questionnaire sections.The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. The processing files are: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), (2) a UCC file that contains UCCs and their abbreviated titles, identifying the expenditure, income, or demographic item represented by each UCC, (3) a vehicle make file (CAPIVEHI), and (4) files containing sample programs. The processing files are further explained in the Interview User Guide, Section III.F.6. PROCESSING FILES. There is also a second user guide, "User's Guide to Income Imputation in the CE", which includes information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships for each CU member. Income information, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over was also collected.

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates (for consumer units or CUs) of average expenditures in news releases, reports, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (see Section XV. Appendix 4). The microdata are available on the public BLS website for free download. These microdata files present detailed expenditure and income data for the Diary component of the CE. They include weekly expenditure (EXPN), annual income (DTBD), and imputed income (DTID) files. The data in EXPN, DTBD, and DTID files are categorized by a Universal Classification Code (UCC). The advantage of the EXPN and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLY and MEMB files contain data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLY files permits the data user to link consumer spending, by general expenditure category, to household characteristics and demographics on one set of files. Estimates of average expenditures from the Diary survey, integrated with data from the Interview survey, are published online in the CE annual reports.. A number of recent publications containing data from the CE are available on the public website as well. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: ?U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2011.?

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2011 sample is composed of 91 areas. The design classifies the PSUs into four categories: 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. 38 "X" PSUs, are medium-sized MSAs. 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2011 survey is generated from the 2000 Population Census file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year.

Mode of data collection

Computer Assisted Personal Interview [capi]

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index.The CE program is comprised of two separate components (each with its own questionnaire and independent sample), the quarterly Interview Survey and the Diary Survey (ICPSR 34442). This data collection contains the quarterly Interview Survey data, which was designed to collect data on major items of expense which respondents could be expected to recall for 3 months or longer. These included relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occurred on a regular basis, such as rent or utilities. The Interview Survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures.The microdata in this collection are available as SAS, SPSS, and STATA datasets or ASCII comma-delimited files. The 2011 Interview Survey release contains seven groups of Interview data files (FMLY, MEMB, MTBI, ITBI, ITII, FPAR, and MCHI), 50 EXPN files, and processing files.The FMLY, MEMB, MTBI, ITBI, and ITII files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly datasets for each of these files, running from the first quarter of 2011 through the first quarter of 2012. The FMLY file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMB file contains member characteristics and income data; the MTBI file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITBI file contains income data converted to a monthly time frame and assigned to UCCs; and the ITII file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs.The FPAR and MCHI datasets are grouped as 2-year datasets (2010 and 2011), plus the first quarter of the 2012 and contain paradata about the Interview survey. The FPAR file contains CU level data about the Interview survey, including timing and record use. The MCHI file contains data about each interview contact attempt, including reasons for refusal and times of contact. Both FPAR and MCHI files contain five quarters of data.The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLY or MTBI files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Each of the 50 EXPN files contains five quarters of data, directly derived from their respective questionnaire sections.The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. The processing files are: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), (2) a UCC file that contains UCCs and their abbreviated titles, identifying the expenditure, income, or demographic item represented by each UCC, (3) a vehicle make file (CAPIVEHI), and (4) files containing sample programs. The processing files are further explained in the Interview User Guide, Section III.G.8. "PROCESSING FILES." There is also a second user guide, User's Guide to Income Imputation in the CE, which includes information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships for each CU member. Income information, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over was also collected.

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. The CE program is comprised of two separate components (each with its own questionnaire and independent sample), the quarterly Interview Survey and the Diary Survey (ICPSR 4415). This data collection contains the quarterly Interview Survey data, which was designed to collect data on major items of expense which respondents could be expected to recall for 3 months or longer. These included relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occurred on a regular basis, such as rent or utilities. The Interview Survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures. The microdata in this collection are available as SAS, SPSS, and STATA datasets or ASCII comma-delimited files. The 2004 Interview Survey release contains five groups of Interview data files (FMLY, MEMB, MTAB, ITAB, and ITAB_IMPUTE), 50 EXPN files, and four processing files. The FMLY, MEMB, MTAB, ITAB, and ITAB_IMPUTE files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly datasets for each of these files, running from the first quarter of 2004 through the first quarter of 2005. The FMLY file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMB file contains member characteristics and income data; the MTAB file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITAB file contains income data converted to a monthly time frame and assigned to UCCs; and the ITAB_IMPUTE file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs. The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLY or MTAB files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Each of the 50 EXPN files contains five quarters of data, directly derived from their respective questionnaire sections. The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. The processing files are: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), (2) a UCC file that contains UCCs and their abbreviated titles, identifying the expenditure, income, or demographic item represented by each UCC, (3) two vehicle make and model files (VEHI and CAPIVEHI), and (4) files containing sample programs (See Section VII.A. SAMPLE PROGRAM). The processing files are further explained in the Interview User Guide, Section III.F.6. "PROCESSING FILES." There is also a second user guide, User's Guide to Income Imputation in the CE, which includes information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships for each CU member. Income information, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over was also collected.

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers, including data on their expenditures, income, and consumer unit (families and single consumers) characteristics. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. The CE program is comprised of two separate components, each with its own questionnaire and independent sample: (1) the quarterly Interview Survey, and (2) the Diary Survey. This data collection contains the quarterly Interview Survey data, which was designed to collect data on major items of expense which respondents could be expected to recall for 3 months or longer. Items include relatively large expenditures, such as those for property, automobiles, and major durable goods, and those that occurred on a regular basis, such as rent or utilities. The Interview Survey does not collect data on expenses for housekeeping supplies, personal care products, and nonprescription drugs, which contribute about 5 to 15 percent of total expenditures. The 2013 Interview Survey contains eight groups of Interview data files (FMLI, MEMI, MTBI, ITBI, ITII, NTAXI, FPAR, and MCHI), forty-three Detailed Expenditure (EXPN) files, and processing files. The FMLI, MEMI, MTBI, ITBI, ITII, and NTAXI files are organized by the calendar quarter of the year in which the data were collected. There are five quarterly datasets for each of these files, running from the first quarter of 2013 through the first quarter of 2014 (with NTAXI files starting the second quarter of 2013). The FMLI file contains consumer unit (CU) characteristics, income, and summary level expenditures; the MEMI file contains member characteristics and income data; the MTBI file contains expenditures organized on a monthly basis at the Universal Classification Code (UCC) level; the ITBI file contains income data converted to a monthly time frame and assigned to UCCs; and the ITII file contains the five imputation variants of the income data converted to a monthly time frame and assigned to UCCs. The NTAXI file contains federal and state tax information for each tax unit within the CU. The FPAR and MCHI datasets are grouped as 2-year datasets (2012 and 2013), plus the first quarter of 2014, and contain paradata about the Interview survey. The FPAR file contains CU level data about the Interview survey, including timing and record use. The MCHI file contains data about each interview contact attempt, including reasons for refusal and times of contact. Both FPAR and MCHI files contain five quarters of data. The EXPN files contain expenditure data and ancillary descriptive information, often not available on the FMLI or MTBI files, in a format similar to the Interview questionnaire. In addition to the extra information available on the EXPN files, users can identify distinct spending categories easily and reduce processing time due to the organization of the files by type of expenditure. Each of the 43 EXPN files contains five quarters of data, directly derived from their respective questionnaire sections. The processing files enhance computer processing and tabulation of data, and provide descriptive information on item codes. There are two types of processing files: (1) aggregation scheme files used in the published consumer expenditure survey interview tables and integrated tables (ISTUB and INTSTUB), and (2) a vehicle make file (CAPIVEHI). The processing files are further explained in the Interview Survey Users' Guide, Section III.H.9. "Processing Files." In addition to the primary users' guide, the Users' Guide to Income Imputation provides information on how to appropriately use the imputed income data. Demographic and family characteristics data include age, sex, race, marital status, and CU relationships for each CU member. Income information was also collected, such as wage, salary, unemployment compensation, child support, and alimony, as well as information on the employment of each CU member age 14 and over. The unpublished integrated CE data tables produced by the BLS are available to download through NADAC (click on "Other" in the Dataset(s) section). The tables show average and percentile expenditures for detailed items, as well as the standard error and coefficient of variation (CV) for each spending

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates (for consumer units or CUs) of average expenditures in news releases, reports, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (see Section XVI. Appendix 5). The microdata are available on CD-ROM as SAS data sets or ASCII text files. These microdata files present detailed expenditure and income data for the Diary component of the CE for 2007. They include weekly expenditure (EXPN), annual income (DTAB) files, and imputed income files (DTID). The data in EXPN, DTAB, and DTID files are categorized by a Universal Classification Code (UCC). The advantage of the EXPN and DTAB files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLY and MEMB files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLY files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2007 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2007. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: “U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2007”.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

A. SURVEY SAMPLE DESIGN Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2007 sample is composed of 91 areas. The design classifies the PSUs into four categories: • 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 38 "X" PSUs, are medium-sized MSAs. • 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. • 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI. The sampling frame (that is, the list from which housing units were chosen) for the 2007 survey is generated from the 2000 Population Census file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year.

Mode of data collection

Computer Assisted Personal Interview [capi]

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates for consumer units (CUs) of average expenditures in news releases, reports, issues, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (See Section XV. APPENDIX 4). The microdata are available online at http://www/bls.gov/cex/pumdhome.htm.

These microdata files present detailed expenditure and income data for the Diary component of the CE for 2002. They include weekly expenditure (EXPD) and annual income (DTBD) files. The data in EXPD and DTBD files are categorized by a Universal Classification Code (UCC). The advantage of the EXPD and DTBD files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLD and MEMD files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLD files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files.

Estimates of average expenditures in 2002 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2002. A list of recent publications containing data from the CE appears at the end of this documentation.

The microdata files are in the public domain and with appropriate credit, may be reproduced without permission. A suggested citation is: "U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2002".

STATE IDENTIFIER Since the CE is not designed to produce state-level estimates, summing the consumer unit weights by state will not yield state population totals. A CU's basic weight reflects its probability of selection among a group of primary sampling units of similar characteristics. For example, sample units in an urban nonmetropolitan area in California may represent similar areas in Wyoming and Nevada. Among other adjustments, CUs are post-stratified nationally by sex-age-race. For example, the weights of consumer units containing a black male, age 16-24 in Alabama, Colorado, or New York, are all adjusted equivalently. Therefore, weighted population state totals will not match population totals calculated from other surveys that are designed to represent state data. To summarize, the CE sample was not designed to produce precise estimates for individual states. Although state-level estimates that are unbiased in a repeated sampling sense can be calculated for various statistical measures, such as means and aggregates, their estimates will generally be subject to large variances. Additionally, a particular state-population estimate from the CE sample may be far from the true state-population estimate.

INTERPRETING THE DATA

Several factors should be considered when interpreting the expenditure data. The average expenditure for an item may be considerably lower than the expenditure by those CUs 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 of those purchasing. (See Section V.B. for ESTIMATION OF TOTAL AND MEAN EXPENDITURES). Also, an individual CU may spend more or less than the average, depending on its particular characteristics. Factors such as income, age of family members, geographic location, taste and personal preference also influence expenditures. Furthermore, even within groups with similar characteristics, the distribution of expenditures varies substantially.

Expenditures reported are the direct out-of-pocket expenditures. Indirect expenditures, which may be significant, may be reflected elsewhere. For example, rental contracts often include utilities. Renters with such contracts would record no direct expense for utilities, and therefore, appear to have no utility expenses. Employers or insurance companies frequently pay other costs. CUs with members whose employers pay for all or part of their health insurance or life insurance would have lower direct expenses for these items than those who pay the entire amount themselves. These points should be considered when relating reported averages to individual circumstances.

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

A. SURVEY SAMPLE DESIGN

Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons.

The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2002 sample is composed of 105 areas. The design classifies the PSUs into four categories:

• 31 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. • 46 "B" PSUs, are medium-sized MSA's. • 10 "C" PSUs are nonmetropolitan areas that are included in the CPI. • 18 "D" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI.

The sampling frame (that is, the list from which housing units were chosen) for the 2002 survey is generated from the 1990 Population Census 100-percent-detail file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (ED's) from the Census that fail to meet the criterion for good addresses for new construction, and all ED's in nonpermit-issuing areas are grouped into the area segment frame.

To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance.

Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year. During the last 6 weeks of the year, however, the Diary Survey sample is supplemented to twice its normal size to increase the reporting of types of expenditures unique to the holidays.

B. COOPERATION LEVELS

The annual target sample size at the United States level for the Diary Survey is 7,800 participating sample units. To achieve this target the total estimated work load is 11,275 sample units. This allows for refusals, vacancies, or nonexistent sample unit addresses.

Each participating sample unit selected is asked to keep two 1-week diaries. Each diary is treated independently, so response rates are based on twice the number of housing units sampled.

Mode of data collection

Computer Assisted Personal Interview [capi]

Response rate

The response rate for the 2002 Diary Survey is 74.2%. This response rate refers to all diaries in the year.

Abstract

The Consumer Expenditure Survey (CE) program provides a continuous and comprehensive flow of data on the buying habits of American consumers. These data are used widely in economic research and analysis, and in support of revisions of the Consumer Price Index. To meet the needs of users, the Bureau of Labor Statistics (BLS) produces population estimates (for consumer units or CUs) of average expenditures in news releases, reports, and articles in the Monthly Labor Review. Tabulated CE data are also available on the Internet and by facsimile transmission (see Section XVI. Appendix 5). The microdata are available on CD-ROM as SAS data sets or ASCII text files. These microdata files present detailed expenditure and income data for the Diary component of the CE for 2006. They include weekly expenditure (EXPN), annual income (DTAB) files, and imputed income files (DTAB_IMPUTE). The data in EXPN, DTAB, and DTAB_IMPUTE files are categorized by a Universal Classification Code (UCC). The advantage of the EXPN and DTAB files is that with the data classified in a standardized format, the user may perform comparative expenditure (income) analysis with relative ease. The FMLY and MEMB files present data on the characteristics and demographics of CUs and CU members. The summary level expenditure and income information on the FMLY files permits the data user to link consumer spending, by general expenditure category, and household characteristics and demographics on one set of files. Estimates of average expenditures in 2006 from the Diary survey, integrated with data from the Interview survey, are published in Consumer Expenditures in 2006. A list of recent publications containing data from the CE appears at the end of this documentation. The microdata files are in the public domain and, with appropriate credit, may be reproduced without permission. A suggested citation is: “U.S. Department of Labor, Bureau of Labor Statistics, Consumer Expenditure Survey, Diary Survey, 2006”.

The Diary survey PUMD are organized into five major data files for each quarter: 1. FMLD - a file with characteristics, income, and summary level expenditures for the household 2. MEMD - a file with characteristics and income for each member in the household
3. EXPD - a detailed weekly expenditure file categorized by UCC 4. DTBD - a detailed annual income file categorized by UCC
5. DTID - a household imputed income file categorized by UCC

Analysis unit

Consumer Unit

Kind of data

Sample survey data [ssd]

Sampling procedure

A. SURVEY SAMPLE DESIGN Samples for the CE are national probability samples of households designed to be representative of the total U. S. civilian population. Eligible population includes all civilian noninstitutional persons. The first step in sampling is the selection of primary sampling units (PSUs), which consist of counties (or parts thereof) or groups of counties. The set of sample PSUs used for the 2006 sample is composed of 91 areas. The design classifies the PSUs into four categories: 21 "A" certainty PSUs are Metropolitan Statistical Areas (MSA's) with a population greater than 1.5 million. 38 "X" PSUs, are medium-sized MSAs. 16 "Y" PSUs are nonmetropolitan areas that are included in the CPI. 16 "Z" PSUs are nonmetropolitan areas where only the urban population data will be included in the CPI. The sampling frame (that is, the list from which housing units were chosen) for the 2006 survey is generated from the 2000 Population Census file. The sampling frame is augmented by new construction permits and by techniques used to eliminate recognized deficiencies in census coverage. All Enumeration Districts (EDs) from the Census that fail to meet the criterion for good addresses for new construction, and all EDs in nonpermit-issuing areas are grouped into the area segment frame. To the extent possible, an unclustered sample of units is selected within each PSU. This lack of clustering is desirable because the sample size of the Diary Survey is small relative to other surveys, while the intraclass correlations for expenditure characteristics are relatively large. This suggests that any clustering of the sample units could result in an unacceptable increase in the within-PSU variance and, as a result, the total variance. Each selected sample unit is requested to keep two 1-week diaries of expenditures over consecutive weeks. The earliest possible day for placing a diary with a household is predesignated with each day of the week having an equal chance to be the first of the reference week. The diaries are evenly spaced throughout the year. B. COOPERATION LEVELS The annual target sample size at the United States level for the Diary Survey is 7,200 participating sample units. To achieve this target the total estimated work load is 12,200 sample units. This allows for refusals, vacancies, or nonexistent sample unit addresses. Each participating sample unit selected is asked to keep two 1-week diaries. Each diary is treated independently, so response rates are based on twice the number of housing units sampled. The response rate for the 2006 Diary Survey is 74.2% as shown below. This response rate refers to all diaries in the year. Number of Eligible housing unit interviews
diaries designated for the survey Type B or C ineligible cases Number of potential diaries Type A nonresponse Total respondent interviews
24,320 4,844 19,476 5,021 14,455

Mode of data collection

Computer Assisted Personal Interview [capi]