Getting Started Guide

User Manual:

Open the PDF directly: View PDF PDF.
Page Count: 15

DownloadGetting Started Guide
Open PDF In BrowserView PDF
Getting started with Consumer Expenditure Survey (CE)
Public-Use Microdata (PUMD)∗
Division of Consumer Expenditure Survey, Bureau of Labor Statistics†
September 9, 2014

∗ Version
† For

3.0
questions or comments, please contact the CE by phone, (202) 691-6900, or email, CEXINFO@bls.gov.

Contents
1 Introduction
1.1 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2 What is CE PUMD? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Do I really need CE PUMD for my project? . . . . . . . . . . . . . . . . . . . . . . .

1
1
1
1

2 PUMD file structure
2.1 Interview survey data . . . . . . .
2.1.1 Quarterly files . . . . . . .
2.1.2 “Annual” or “5 quarter” files
2.1.3 Paradata . . . . . . . . . .
2.2 Diary Survey Data . . . . . . . . .
2.2.1 Quarterly files . . . . . . .
2.3 File naming conventions . . . . . .
2.3.1 File names and directories .
2.3.2 What’s the “X” about? . . .

.
.
.
.
.
.
.
.
.

2
2
2
3
3
3
3
4
4
5

3 PUMD file content overview
3.1 Interview quarterly files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 FMLI-CU characteristics, income, and summary expenditure variables . . . .
3.1.2 MEMI-CU member characteristics and income . . . . . . . . . . . . . . . . .
3.1.3 MTBI-Monthly expenditure file classified by UCC . . . . . . . . . . . . . . .
3.1.4 ITBI-Consumer Unit monthly income classified by UCC . . . . . . . . . . . .
3.1.5 ITII-Consumer Unit monthly imputed income file classified by UCC. . . . . .
3.1.6 NTAXI-Federal and state tax information classified by tax unit . . . . . . . .
3.2 Interview annual files–EXPN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1 CLA-Clothing and sewing materials . . . . . . . . . . . . . . . . . . . . . . .
3.2.2 IHB-Hospitalization and health insurance detailed questions . . . . . . . . . .
3.2.3 VEQ-Vehicle maintenance and repair . . . . . . . . . . . . . . . . . . . . . . .
3.3 Diary files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.1 FMLD-CU characteristics, income, weights, and summary level expenditures
3.3.2 MEMD-CU member characteristics and income . . . . . . . . . . . . . . . . .
3.3.3 EXPD-Detailed weekly expenditure file categorized by UCC . . . . . . . . . .
3.3.4 DTBD-Detailed annual income file categorized by UCC . . . . . . . . . . . .
3.3.5 DTID-Consumer Unit imputed income file categorized by UCC. . . . . . . .

5
5
5
6
6
7
7
8
8
9
9
10
11
11
11
12
12
12

4 Documentation

13

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

.
.
.
.
.
.
.
.
.

1

Introduction

1.1

Purpose

The Getting Started Guide is intended for the novice user of CE Public-Use Microdata (CE PUMD).
The guide introduces the structure and content of the CE PUMD files and lists the available
documentation.

1.2

What is CE PUMD?

The CE PUMD are a series of data files that contain a wealth of detailed variables concerning each
individual response from the two CE surveys within a collection year. This level of detail is useful
for a broad spectrum of research, but to analyze the data requires researchers to have an advanced
knowledge of a statistical software package, such as SAS.

1.3

Do I really need CE PUMD for my project?

Many research projects do not require that much detail. For these projects, CE provides tabulated
data with pre-calculated population means, aggregates, and standard errors as well as selected
demographic characteristics.
The online CE tabular data, available at http://www.bls.gov/cex/tables.htm, are broken
out by the following characteristics:
ˆ Age of reference person
ˆ Composition of consumer unit
ˆ Highest education level of any member of the consumer unit
ˆ Higher income before taxes (> $70,000)
ˆ Hispanic or Latino origin of reference person
ˆ Housing tenure and type of area (Urban/Rural)
ˆ Income before taxes (< $70,000)
ˆ Number of earners in consumer unit
ˆ Population size of area of residence
ˆ Quintiles of income before taxes
ˆ Race of reference person
ˆ Region of residence
ˆ Size of consumer unit
ˆ Selected age of reference person

1

The tabular data are constructed from internal CE microdata, and are widely used by researchers, journalists, businessmen, policy makers, academics, non-profits, students, and local,
state, federal, and foreign governments.
If you are looking for more detailed expenditures than what are available in the online tables,
then you can request disaggregated tables (“prepubs”) by contacting our office at: http://www.
bls.gov/cex/csxcont.htm. The disaggregated tables will have more granular expenditures such
as “bacon” or “pork chops” instead of “pork.” You can see the level of detail available in the tables
by looking at the underlying items in the glossary of expenditures terms: http://www.bls.gov/
cex/csxgloss.htm#expn.
If your research needs cannot be met with the available tables, then the CE PUMD files may
be for you. With the CE PUMD files, you will be able to calculate means and other statistics by
demographic groups that you define, as well as conduct more complex analysis such as regression
analysis or analysis of variance.
However, the CE PUMD is a large, complex dataset that requires an extensive time commitment
by users to study the documentations to learn how to use the data appropriately. In addition to
the time needed to learn how to use the files, users must also have a good working knowledge of
an advanced statistical software package, such as SAS, SPSS, or STATA, to manipulate the files.
Because of the large file sizes, it may be difficult to use a package such as Microsoft Excel to conduct
analysis.
If you have any questions about whether the CE PUMD is right for you, our staff would be
happy to give you guidance. Contact us at: http://www.bls.gov/cex/csxcont.htm. We also host
an annual CE Microdata Users’ Workshop to help our users learn how to use the CE PUMD files
through hands-on practical training. Look for an announcement on the CE website for our next
workshop.

2

PUMD file structure

2.1
2.1.1

Interview survey data
Quarterly files

“Quarter” refers to the calendar quarter in which the interview occurred. For example, any Consumer Unit (CU) interviewed in April, May, or June would have their data stored in the quarter 2
(Q2) datasets. During an interview, the CU is asked to report expenditures for a reference period of
three months. So, for a CU interviewed in April, their expenditures in the Q2 file are for January,
February, and March. (This is important to remember when calculating calendar year estimates.)
Five quarters of interviews are included in each PUMD release, Q1 of (RELEASE YEAR) through
Q1 of (RELEASE YEAR + 1). There are six types of data files organized by quarter:
1. FMLI - a file with characteristics, income, weights, and summary level expenditures for the
CU.
2. MEMI - a file with characteristics and income for each member in the CU.
3. MTBI - a detailed monthly expenditure file categorized by Universal Classification Code
(UCC).
4. ITBI - a Consumer Unit monthly income file categorized by UCC.
2

5. ITII - a Consumer Unit monthly imputed income file categorized by UCC.1
6. NTAXI - a file with federal and state tax information for each tax unit in the CU.2
2.1.2

“Annual” or “5 quarter” files

The Interview EXPN files contain expenditure and non-expenditure information collected directly
from all major sections of the quarterly Interview survey. For each PUMD release, there are roughly
50 annual EXPN datasets, with each dataset corresponding to a specific section of the Interview
survey (http://www.bls.gov/cex/csxsurveyforms.htm#interview). Because these datasets correspond directly to the Interview survey sections, there are varying reference periods for the expenditures contained within each EXPN dataset, dependent upon the questions within each section.
Each annual EXPN dataset is made up of the five quarters of data within each release.
2.1.3

Paradata

Beginning in 2009, the CE began releasing paradata, which is data about the interview process.
Paradata is available in two datasets:
1. FPAR - a file with data about the survey, including timing for each section and whether the
respondent used records.
2. MCHI - a file with data about the contact history between the field representative and the
respondent, including reasons for interview refusal and time of contact.
Typically, each paradata file has nine quarters of data, starting with Q1 from the (RELEASE YEAR
- 1) to Q1 from the (RELEASE YEAR + 1). Nine quarters are included within each PARADATA
file so users can have paradata records across all interviews for a given CU. (A CU completing their
last interview during Q1 of (RELEASE YEAR) has the potential to have had their first (bounding)
interview during Q1 of (RELEASE YEAR - 1)).3

2.2
2.2.1

Diary Survey Data
Quarterly files

“Quarter” refers to the calendar quarter in which the Diary booklet was placed in the home of the
CU by the Census Field Representative. All Diary files are organized as quarterly files. Those
include:
1. FMLD - a file with characteristics, income, weights, and summary level expenditures for the
CU.
2. MEMD - a file with characteristics and income for each member in the CU.
3. EXPD - a detailed weekly expenditure file categorized by UCC.
1 Available

beginning in
beginning in
3 With the 2009 PUMD
quarters of data, while the
2 Available

2004
2013Q2
release, historical paradata for 2005-2007 were released. The dataset for 2005 contains 3
datasets for 2006 and 2007 contains 4 quarters of data each.

3

4. DTBD - a detailed annual income file categorized by UCC.
5. DTID - a Consumer Unit imputed income file categorized by UCC.4

2.3

File naming conventions

The naming conventions used for each PUMD release is common to both Interview and Diary data.
YY stands for the two digit year and Q stands for the calendar quarter of the Interview or calendar
quarter of the placement of the Diary booklet. Below, the example file directories and names are
for SAS datasets. The only difference in the naming convention for other dataset types (i.e., SPSS,
STATA, etc.) is the file extension.
2.3.1

File names and directories

Interview quarterly files
FMLI \INTRVWYY \FMLIYYQ.sas7bdat
MEMI \INTRVWYY \MEMIYYQ.sas7bdat
MTBI \INTRVWYY \MTBIYYQ.sas7bdat
ITBI \INTRVWYY \ITBIYYQ.sas7bdat
ITII \INTRVWYY \ITIIYYQ.sas7bdat
NTAXI \INTRVWYY \NTAXIYYQ.sas7bdat
Interview annual files
Note: Only selected EXPN files are listed here. All other EXPN files follow the same convention.
CLA \EXPNYY \CLAYY.sas7bdat
IHB \EXPNYY \IHBYY.sas7bdat
VEQ \EXPNYY \VEQYY.sas7bdat
Diary quarterly files
FMLD \DIARYYY \FMLDYYQ.sas7bdat
MEMD \DIARYYY \MEMDYYQ.sas7bdat
EXPD \DIARYYY \EXPDYYQ.sas7bdat
DTBD \DIARYYY \DTBDYYQ.sas7bdat
DTID \DIARYYY \DTIDYYQ.sas7bdat
4 Available

beginning in 2004.

4

2.3.2

What’s the “X” about?

You may have noticed there is an “X” in the name for some of the Interview quarterly files (eg.
FMLIYY 1x.sas7bdat or MEMIYY 1x.sas7bdat). You may have also noticed that this “X” is only
in the name of quarter 1 files for the current calendar year. This is because files for the first quarter
of any calendar year will appear in two CE PUMD releases:
1. As the “fifth” quarter file in the previous calendar year’s release,
2. And as the “first” quarter file in the current calendar year’s release.
The “X” signifies that the first quarter file of the current calendar year release is not identical to
the fifth quarter file of the previous calendar year release. One reason the files are not identical is
because PUMD production processes use the five quarters of data within the release to calculate
certain values, such as the critical and topcode values of income and expenditures. During years in
which there was a sample design change, there is a bigger difference in the “X” files. For example,
the CE changed its sample design in 2005Q1. When this happens, the CE draws an “overlapping
sample” for 2005Q1, i.e., some CUs are for the old design, some CUs are for the new design, and
some are for both designs.5 For the 2004 PUMD release, the CUs in the 2005Q1 files are those for
the old design. For the 2005 PUMD release, the CUs in the 2005Q1 files are those for the new
design. CUs that were in both the old and new designs are contained within the 2005Q1 file for
both the 2004 and 2005 PUMD releases.

3

PUMD file content overview

NEWID, the unique identifier for each CU, is included within every record on each data file.
It can be used to link records of a particular CU across files. For example, linking demographic
information in the Interview FMLI file to expenditure information in the MTBI file. Be aware that
the samples from the Interview and Diary surveys are independent, so there is no CU contained
within both the Interview and Diary files.

3.1
3.1.1

Interview quarterly files
FMLI-CU characteristics, income, and summary expenditure variables

ˆ There exists one record per CU.
ˆ Each CU is uniquely identified by NEWID.
ˆ Variables include demographics for the reference person and spouse of reference person, income
at the CU level, sample housing unit information, summary level expenditures, and weighting
variables.

Note: Not all the variables within the FMLI file are shown in the sample below.
5 CUs can be in both the old and new designs because the designs are based on areas of the United States called
Primary Sampling Units (PSU). So if a CU was in a PSU that was in both designs, then they are used in both designs.

5

NEWID AGE_REF FAM_SIZE FINCBTXM FDAWAYPQ FDAWAYCQ FINLWT21
1 2545235
82
2 11807.0
0
0 1980.829
2 2545245
69
3 33773.0
370
0 22347.181
3 2545255
22
1 11000.0
180
0 15272.973
4 2545275
57
1 163793.8
680
0 14774.904
5 2545305
45
2 116227.0
770
0 18557.247
6 2545355
37
1 120000.0
2000
0 16921.900
7 2545365
29
4 31400.0
625
0 15776.683
8 2545375
53
2
200.0
1530
0 15271.262
9 2545385
64
1 60000.0
455
0 16723.973
10 2545405
46
2 84608.2
1540
0 16474.727
3.1.2

MEMI-CU member characteristics and income

ˆ There exist multiple records per CU.
ˆ There exists one record per member.
ˆ Unique records are defined by the combination of NEWID and MEMBNO.
ˆ Variables include demographics about CU members, member level income, and member relationship status to the reference person.

Note: Not all the variables within the MEMI file are shown in the sample below.
NEWID MEMBNO AGE CU_CODE EARNER EDUCA SEX MEMBRACE SALARYXM
1 2667582
4
1
3
NA
NA
1
1
1
2 2667602
1 26
1
1
2
1
1
745
3 2667602
2 23
0
1
5
2
1
3282
4 2667602
3
2
3
NA
NA
1
1
1
5 2667612
1 22
1
1
5
1
4
871
6 2667612
2 24
6
1
5
1
4
1605
7 2667612
3 57
7
1
4
1
4
817
8 2667612
4 49
7
2
4
2
4
1
9 2667642
1 29
1
1
4
1
1
312
10 2667652
1 28
1
1
7
1
1
1
3.1.3

MTBI-Monthly expenditure file classified by UCC

ˆ There exist multiple records per CU.
ˆ There exists one record per expenditure per purchase month/year.
ˆ Prior to the 2010 PUMD release, unique records were not identified. From the 2010 release forward, unique records can be identified using a combination of NEWID, EXPNAME,
SEQNO, and ALCNO.
ˆ The monthly expenditure amounts are mapped from corresponding EXPN variables.

6

Note: Not all the variables within the MTBI file are shown below.

1
2
3
4
5
6
7
8
9
10

NEWID
2554935
2554935
2554935
2545765
2545765
2545765
2562785
2562785
2562785
2560385

3.1.4

UCC COST COST_ GIFT PUBFLAG REF_MO REF_YR
210110 232
1
2
2
3
2013
210110 232
1
2
2
2
2013
210110 232
1
2
2
1
2013
210110 600
1
2
2
3
2013
210110 600
1
2
2
2
2013
210110 600
1
2
2
1
2013
210110 500
1
2
2
4
2013
210110 500
1
2
2
3
2013
210110 500
1
2
2
2
2013
210110 660
1
2
2
4
2013

ITBI-Consumer Unit monthly income classified by UCC

ˆ There exist multiple records per CU.
ˆ There exists one record per income item per month/year.
ˆ Unique records are defined by NEWID, UCC, REFMO, and REFYR.
ˆ The monthly amounts are mapped from the corresponding CU-level imputed income mean
values.

Note: This ITBI sample shows ALL variables that are in the full ITBI file.6

1
2
3
4
5
6
7
8
9
10

NEWID REFMO REFYR UCC PUBFLAG
VALUE VALUE_
2545235
1 2013 5100
2 520.8333
NA
2545235
2 2013 5100
2 520.8333
NA
2545235
3 2013 5100
2 520.8333
NA
2545235
1 2013 5110
2 520.8333
NA
2545235
2 2013 5110
2 520.8333
NA
2545235
3 2013 5110
2 520.8333
NA
2545235
1 2013 5200
2 2500.0000
NA
2545235
2 2013 5200
2 2500.0000
NA
2545235
3 2013 5200
2 2500.0000
NA
2545235
1 2013 5210
2 2500.0000
NA

3.1.5

ITII-Consumer Unit monthly imputed income file classified by UCC.

ˆ There exist multiple records per CU.
ˆ There exists one record per income item per month/year.
ˆ Unique records are defined by NEWID, UCC, REFMO, REFYR, and IMPNUM.
6 The

dataset views in this document were generated using R, which represents missing values as “NA.” Other
statistical software packages may represent missing values as a period, blank space, or in another manner.

7

ˆ The monthly amounts are mapped from the corresponding CU-level imputed income imputes(5 imputes per income item).
ˆ This file is necessary to calculate variances and standard errors for imputed income.

Note: This ITII sample shows ALL variables that are in the full ITII file.

1
2
3
4
5
6
7
8
9
10

NEWID REFMO REFYR
UCC PUBFLAG
VALUE VALUE_ IMPNUM
2545235
1 2013 900030
2 982.8333
NA
1
2545235
1 2013 900030
2 982.8333
NA
2
2545235
1 2013 900030
2 982.8333
NA
3
2545235
1 2013 900030
2 982.8333
NA
4
2545235
1 2013 900030
2 982.8333
NA
5
2545235
2 2013 900030
2 982.8333
NA
1
2545235
2 2013 900030
2 982.8333
NA
2
2545235
2 2013 900030
2 982.8333
NA
3
2545235
2 2013 900030
2 982.8333
NA
4
2545235
2 2013 900030
2 982.8333
NA
5

3.1.6

NTAXI-Federal and state tax information classified by tax unit

ˆ There may exist multiple records per CU.
ˆ There exists one record per tax unit.
ˆ Unique records are defined by the combination of NEWID and TAX UNIT.
ˆ Variables include federal and state tax rates, tax liabilites, and various tax-related characteristics of the tax unit.

Note: Not all the variables within the NTAXI file are shown in the sample below.

1
2
3
4
5
6
7
8
9
10

NEWID TAX_UNIT DEPCNT FILESTAT SRATE_CY SRATE_PY
2545235
1
2
3
0.0000
0.0000
2545245
1
2
3
0.0000
0.0000
2545255
1
0
1
0.0000
0.0653
2545275
1
0
1
0.0575
0.0575
2545305
1
0
2
0.0475
0.0475
2545355
1
0
1
0.0758
0.0764
2545365
1
2
2
0.0930
0.0930
2545375
1
2
3
0.0000
0.0000
2545385
1
0
1
0.0610
0.0610
2545405
1
1
3
0.0400
0.0400

3.2

Interview annual files–EXPN

ˆ Each EXPN file is different, with some having multiple records per CU and others with one
record or no records per CU interviewed each quarter.

8

ˆ Unique records are defined differently in each file.
ˆ Common variables to all EXPN files: NEWID is the unique CU identifier. QYEAR defines
the year and interview quarter for a given record. SEQNO is assigned sequentially during
the interview as each expenditure record is recorded into the database. ALCNO is assigned
sequentially for each record that has been allocated from one expenditure during the process.
For example, a CU may report spending $50 on a pair of men’s pants and a shirt. The CE
will allocate out that record into separate records, one for men’s pants and shorts($30) and
one for men’s shirts($20). REC ORIG describes the origin of the record, i.e., whether the
expenditure data was collected during the current interview, expenditure data collected for
a current reference period in a previous interview, household inventory data collected during
a previous interview that was pulled forward to the current, or data that was created by the
CE processing edit system.
ˆ Variables include those directly collected from the Interview survey questionnaire on expenditures and screener questions. Variables that are recalculated by CE are also included.

Examples:
3.2.1

CLA-Clothing and sewing materials

ˆ Unique records in CLA are defined by the combination of NEWID, SEQNO, and ALCNO.
ˆ CLOTHYA is an item code variable for the type of clothing expenditure (suits, dresses,
undergarments, etc.)
ˆ CLOTHMOA is an indicator variable for the month in which the purchase occurred.
ˆ CLOTHXA is the amount of the clothing expense.

Note: Not all the variables within the CLA file are shown below.

1
2
3
4
5
6
7
8
9
10

QYEAR
20131
20131
20131
20131
20131
20131
20131
20131
20131
20131

3.2.2

NEWID SEQNO ALCNO CLOTHYA CLOTHMOA CLOTHXA
2511505
15
0
140
12
204
2511555
13
0
270
11
1
2511555
13
1
200
11
203
2511555
13
2
210
11
4
2511555
13
3
190
11
725
2511565
15
0
275
12
1
2511565
15
1
190
12
572
2511565
15
2
150
12
78
2511565
16
0
140
12
658
2511565
17
0
200
12
759

IHB-Hospitalization and health insurance detailed questions

ˆ Unique records in IHB defined by NEWID, SEQNO, ALCNO, and HHIPDLIB.

9

ˆ HHIPDLIB is the policy number for the health insurance plan. This number is derived by
the order in which the respondent lists their health insurance plans.
ˆ HHICOVQ is the number of people covered by health insurance plan.
ˆ QHI3MCX is the amount paid in health insurance premiums during the reference period
(derived by CE).
ˆ HHICODE is an indicator variable for the type of insurance plan (HMO, FFS, Medicare
supplement, or other special purpose plan).

Note: Not all the variables within the IHB file are shown below.

1
2
3
4
5
6
7
8
9
10

QYEAR
20131
20131
20131
20131
20131
20131
20131
20131
20131
20131

3.2.3

NEWID SEQNO ALCNO HHIPDLIB HHICOVQ QHI3MCX HHICODE
2511505
24
0
1
3
2
2
2511505
25
0
2
2
2
2
2511555
30
0
1
1
1075
1
2511555
31
0
2
1
1297
1
2511555
32
0
3
1
1297
1
2511565
27
0
1
1
1
1
2511565
29
0
3
1
1
4
2511575
33
0
1
4
1
1
2511585
35
0
1
2
1
2
2511595
37
0
2
2
329
1

VEQ-Vehicle maintenance and repair

ˆ Unique records in VEQ defined by NEWID, SEQNO, and ALCNO.
ˆ VOPSERVY is an indicator variable that decribes the type of maintenance or repair untaken.
ˆ VOPMOA is an indicator variable for the month in which the expense occurred.
ˆ VOPEXPX is the total cost of the maintenance or repair expense.

Note: Not all the variables within the VEQ file are shown below.

1
2
3
4
5
6
7
8
9
10

QYEAR
20131
20131
20131
20131
20131
20131
20131
20131
20131
20131

NEWID SEQNO ALCNO VOPSERVY VOPMOA VOPEXPX
2511505
63
0
100
12
275
2511575
61
0
100
11
103
2511595
76
0
370
12
201
2511625
39
0
115
11
819
2511625
39
1
100
11
1
2511625
39
2
110
11
1
2511675
54
0
100
11
715
2511695
126
0
100
12
808
2511695
127
0
365
10
417
2511695
128
0
365
11
382

10

3.3

Diary files

ˆ Most Diary files are analogous to Interview files.
ˆ There are slightly different variable names in the Diary files.
ˆ The summary level expenditures in the Diary FMLD file are not as extensive as those in the
Interview FMLI file.

3.3.1

FMLD-CU characteristics, income, weights, and summary level expenditures

ˆ Analogous to FMLI file in Interview.

Note: Not all the variables within the FMLD file are shown below.
NEWID STRTMNTH WEEKI CUTENURE AGE_REF AGE2 FOODHOME FINLWT21
1 1451201
4
1
1
70
50
52.43 34683.44
2 1451202
5
2
1
70
50
14.00 34683.44
3 1451211
4
1
4
33
1
0.00 31140.85
4 1451212
4
2
4
33
1
28.03 31140.85
5 1451221
4
1
3
88
65
0.00 40310.51
6 1451222
4
2
3
88
65
0.00 40310.51
7 1451241
5
1
1
43
19
152.65 35929.24
8 1451242
5
2
1
43
19
124.05 35929.24
9 1451261
4
1
2
31
1
108.78 33222.60
10 1451262
5
2
2
31
1
61.03 33222.60
3.3.2

MEMD-CU member characteristics and income

ˆ Analogous to MEMI file in Interview.

Note: Not all the variables within the MEMD file are shown below.

1
2
3
4
5
6
7
8
9
10

NEWID MEMBNO AGE EDUCA MARITAL EMPLTYPE WAGEXM
1451201
1 70
5
1
NA
1
1451201
2 68
5
1
NA
1
1451202
1 70
5
1
NA
1
1451202
2 68
5
1
NA
1
1451211
1 33
5
3
NA
1
1451211
2
9
NA
5
NA
1
1451211
3
8
NA
5
NA
1
1451211
4
8
NA
5
NA
1
1451212
1 33
5
3
NA
1
1451212
2
9
NA
5
NA
1

11

3.3.3

EXPD-Detailed weekly expenditure file categorized by UCC

ˆ Analogous to MTBI files in Interview.
ˆ There are multiple records per CU.
ˆ There is one record per expenditure per purchase month/year.
ˆ Unique records are not defined.
ˆ The weekly amounts are recorded directly from listings in the completed Diary.

Note: Not all the variables within the EXPD file are shown below.

1
2
3
4
5
6
7
8
9
10

NEWID ALLOC COST GIFT PUB_FLAG EXPNSQDY
UCC
1451201
0 6.63
2
2
7 10210
1451201
0 1.40
2
2
1 20210
1451201
0 1.60
2
2
1 110210
1451201
0 42.80
2
2
1 170310
1451201
0 10.00
2
2
6 190111
1451201
0 24.00
2
2
5 190112
1451201
0 10.00
2
2
4 190211
1451201
0 2.00
2
2
5 190321
1451201
0 8.00
2
2
1 190322
1451201
0 2.00
2
2
3 190322

3.3.4

DTBD-Detailed annual income file categorized by UCC

ˆ Analogous to ITBI file in Interview.

Note: This DTBD example shows ALL variables that are in the full DTBD file.
NEWID
1 1451201
2 1451201
3 1451201
4 1451201
5 1451201
6 1451201
7 1451201
8 1451201
9 1451201
10 1451201
3.3.5

UCC AMOUNT AMOUNT_ PUB_FLAG
900030 46918
NA
1
980000 46918
1
1
980010
2
NA
1
980020
70
NA
1
980040
2
NA
1
980060
2
NA
1
980070 46918
1
1
980090
100
NA
1
980210
100
NA
1
980230
100
NA
1

DTID-Consumer Unit imputed income file categorized by UCC.

ˆ Analogous to the ITII file.

Note: This DTID example shows ALL variables that are in the full DTID file.
12

NEWID
1 1451201
2 1451201
3 1451201
4 1451201
5 1451201
6 1451201
7 1451201
8 1451201
9 1451201
10 1451201

4

UCC PUB_FLAG AMOUNT AMOUNT_ IMPNUM
900030
1 46918
NA
1
900030
1 46918
NA
2
900030
1 46918
NA
3
900030
1 46918
NA
4
900030
1 46918
NA
5
980000
1 46918
1
1
980000
1 46918
1
2
980000
1 46918
1
3
980000
1 46918
1
4
980000
1 46918
1
5

Documentation

To aid in using the PUMD, the CE produces annual user guides and data dictionaries. Since the
2007 PUMD release, the user guides and data dictionaries, for both Interview and Diary data, have
been parsed out into separate documents. Prior to 2007, the user guides and data dictionaries were
combined in one document. The user documentations include:
ˆ A summary of changes from the previous release of CE PUMD.
ˆ More in depth information of the PUMD files.
ˆ Information on CE PUMD nondisclosure requirements.
ˆ Formulas and procedures for estimating aggregate statistics from the data.
ˆ Data reliability statements.
ˆ And other information, including information on sampling design and cooperation levels.

The data dictionaries include a listing of all the CE PUMD variables and descriptions. User guides
and data dictionaries dating back to the 1996 release, along with other miscellaneous documentations, can be accessed online at the CE website, http://www.bls.gov/cex/csxmicrodoc.htm.

13
Source Exif Data: 
File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.5
Linearized                      : No
Page Count                      : 15
Page Mode                       : UseOutlines
Author                          : 
Title                           : 
Subject                         : 
Creator                         : LaTeX with hyperref package
Producer                        : pdfTeX-1.40.14
Create Date                     : 2014:09:09 09:46:20-04:00
Modify Date                     : 2014:09:09 09:46:20-04:00
Trapped                         : False
PTEX Fullbanner                 : This is MiKTeX-pdfTeX 2.9.4902 (1.40.14)
EXIF Metadata provided by EXIF.tools

Navigation menu