Welcome to the CPS Data Application

The Current Population Survey (CPS) is a monthly survey of about 60,000 U.S. households. The survey collects labor force characteristics and a variety demographic information from respondents. Designed as a nationally representative survey, the CPS data is the source for many key labor market indicators and the official source of the unemployment rate. The CPS is also used to calculate other labor market statistics including: labor force participation, hours worked, class of worker, industry, occupation, and reasons for unemployment.

Datasets

As Published (1994 - Present)

Last updated Invalid Date

Data DictionaryDownload Dataset (CSV)

Data taken directly from the CPS results. (From 1994 - present)

Last updated Invalid Date

Data DictionaryDownload Dataset (CSV)

Layouts

The monthly data and the original documentation files from this time period contain multiple layouts (adult, children, and armed forces). Some of the earlier CPS data contains those and other layouts. To identify which rows in a monthly file correspond to which layout, a combination of variables are used in the original documentation. To simplify this, the warehouse version contains a single variable (v9000) that can be used to identify which rows are associated with which layout. The primary layout, and the one likely of most interest to users, is the adult record layout. From a user perspective, the different layouts are not likely a major concern. But they needed to be addressed for the purpose of putting the data into a table.

v9000Layout TypeFirst Included
1Adult
2ChildJanuary 1982
3Non-interview Type AJanuary 1982
4Non-interview Type B-CJanuary 1982
5Armed ForcesJanuary 1984


Variable Name Modification

The original documentation provides variable name prefixes, composed of a letter or two (H, HG, L, or A in the adult record layout; C in the children record layout, and M in the armed forces layout) followed by a symbol ($, -, or %) (for example, H$MONTH). The different symbols are used to distinguish between unedited, edited, and allocated data. In the warehouse version of the data, the symbols are replaced by numbers, as described in the table below. Additionally, in some cases the letters used in the prefixes were also replaced. To line up variables from the adult, children, and armed forces record layouts the letters A, C, and M.

OriginalUsedIndicates
$1Unedited Data
-2Edited Data
%3Allocated Data


Full Set of Prefixes
OriginalModified
H$H1
H-H2
H%H3
HG-HG2
L-L2
L%L33
A$C$M$v1
A-C-M-v2
A%C%M%v3


Last updated Invalid Date

Data DictionaryDownload Dataset (CSV)

Layouts

The monthly data and the original documentation files from this time period contain multiple layouts (adult, children, and armed forces). Some of the earlier CPS data contains those and other layouts. To identify which rows in a monthly file correspond to which layout, a combination of variables are used in the original documentation. To simplify this, the warehouse version contains a single variable (v9000) that can be used to identify which rows are associated with which layout. The primary layout, and the one likely of most interest to users, is the adult record layout. From a user perspective, the different layouts are not likely a major concern. But they needed to be addressed for the purpose of putting the data into a table.

v9000Layout TypeFirst Included
1Adult
2ChildJanuary 1982
3Non-interview Type AJanuary 1982
4Non-interview Type B-CJanuary 1982
5Armed ForcesJanuary 1984


Last updated Invalid Date

Data DictionaryDownload Dataset (CSV)

The CPS Linked Variable data are designed to facilitate the study of the CPS in a consistent manner over time. This dataset "links" variables in the 1976, 1989, and 1994 versions of the CPS where possible, allowing researchers to trace major survey concepts over time. Linked variables in this dataset can be linked either between the 1976 and the 1989 versions of the CPS, the 1989 and the 1994 versions of the CPS, or across all three versions.

Last updated Invalid Date

Data DictionaryDownload Dataset (CSV)

The CPS Linked Variable data are designed to facilitate the study of the CPS in a consistent manner over time. This dataset "links" variables in the 1976, 1989, and 1994 versions of the CPS where possible, allowing researchers to trace major survey concepts over time. Linked variables in this dataset can be linked either between the 1976 and the 1989 versions of the CPS, the 1989 and the 1994 versions of the CPS, or across all three versions.

The Atlanta Fed’s Harmonized Variable and Longitudinally Matched dataset is a grab and go version of the monthly basic CPS. It is intended for researchers doing labor market analysis, thus it includes data for 16+ year olds only. The data are designed to be used right out of the box. Several Atlanta Fed products are created using this dataset, including the Wage Growth Tracker and Labor Force Participation Dynamics. The data contain harmonized variables that are coded consistently through time as well as identifiers for longitudinally matching individuals and households over time.

Each variable’s start year is indicated at the end of the variable name. For example, mlr76 is coded with consistent values (1= employed, 2= unemployed and 3=not in labor force) from 1976 until today. The stata (.dta) file includes these labels already attached. For users wishing to perform time series analysis, lags of many variables are provided on the dataset already (for example, mlr76_tm12 is the individual’s labor force status from 12 months ago). For a visual of each variable over time and summary statistics see “Charts”. For information on how each variable is created see “Codebook”.