CE Experimental Research Products
This page provides an introduction to special tabulation tables and other experimental products being developed by the Consumer Expenditure Survey Division. New research products will be added and updated periodically.
Information on this page includes the All Consumer Unit Prepublication Table (All CU Prepublication MVP), new crosstabulated tables with higher incomes, the 2014 Higher Income Table which expanded the top income range to $200,000 and over, a Generational Table that sorts expenditures by generation/cohorts, information on how to get the most from the CE published and prepublication tables, and the experimental New Jersey State weight publicuse microdata files.
Note that the 2014 through mid2016 Generational tables were research work, and had not been produced using BLS production methods and standards. The 2016 annual Generational table is now included in the standard tables.
How to get the most from CE tables
This document provides information to help one delve deeper into the published as well as the very detailed nonpublished (prepublication) tables. This short description describes how one can use consumer unit characteristics to uncover useful underlying information about the data.
State Weights for New Jersey
Guidance to the user: Care should be taken when analyzing PublicUse Microdata using the state weights, as the small number of households for some expenditures can cause the mean dollar estimate to be imprecise. The more aggregated summary variables will produce more precise estimates. Additionally, it should be noted that these weights are only for the state of New Jersey and cannot be used to make inferences about any other geographic areas. Similarly, the national weight is still unable to provide state level estimates. The provided data must be used in conjunction with the PublicUse Microdata to obtain state level estimates.
The State Weights initiative by CE is an effort to produce experimental microdata products that can allow users to explore consumer expenditure data at the state level, a feature previously unavailable in the data. We intend to explore the viability of the CE sample to support weight creation for as many states as possible. The first available state is New Jersey. As others become available, they will be added to this page for download and use.
The NEW State Weights for New Jersey are available with their documentation.
State Weights for New Jersey
(XLSX) (STATA) (SAS)
An Overview of the StateLevel Weighting Procedure
Using the State Weights on the PUMD
Sample code: (STATA)
All CU Prepublication Means, Variances, and Percent reporting (MVP) Tables
All consumer units: Annual detailed expenditure means, standard errors, coefficients of variation, and weekly (D) or quarterly (I) percents reporting, Consumer Expenditure Survey, 201316
Guidance to the user: Care should be taken when analyzing detailed expenditure subcategories in the prepublication tables, as the small number of households reporting an expenditure can cause the mean dollar estimate to be imprecise. Users should consider the coefficient of variation (CV). Expenditures with CVs of 25 percent and over are generally considered unreliable. For further information on standard error and variance in the CE, see Standard Errors in the 2016 Consumer Expenditure Survey.
This table was developed to show the large amount of additional detailed expenditure data calculated by the CE. The table contains Means, Variances, and Percent reporting (MVP). The CE standard published tables are an aggregate of the integrated data, and contain only higher level summary rows. This Prepublication table contains data pulled from three different tables. The three underlying tables summarized in this table can be obtained by contacting us.
Available Prepublication tables in this disaggregated format are based on the same set of fifteen published annual tables on our home page. Displaying the disaggregated rows at the lowest level allows for the addition of either weekly or quarterly percent reporting statistics. Furthermore, one can calculate the weekly or quarterly means expenditure for only those who pay for particular items. The three types of tables are available for the same demographics as our published annual calendar tables and contain the following information.
Integrated tables which contain selected data from both the Interview and Diary surveys. These integrated tables present the weekly and quarterly reported expenditures in an annual mean format that matches published tables, but in a detailed disaggregated form. The tables also include standard errors, and coefficients of variation. These contain the most detailed set of expenditures calculated by the CE.
Interview Survey tables containing only Interview data on quarterly expenditures. The table includes annual expenditure means, quarterly percent reporting and variances. For example, one can use this table to find out what percentage of consumer units reported paying Mortgage interest in a quarter and how much they spent on average.
Diary Survey tables contain weekly expenditure mean and percent reporting. As an example, one can use this table to find out what percentage of consumer units bought Bacon or Eggs in a week and how much they spent on average.
The MVP tables displaying these data can be viewed in pdf or xlsx formats.
ALL CU MVP table 2016 (PDF)
(XLSX)
ALL CU MVP table 2015 (PDF)
(XLSX)
ALL CU MVP table 2014 (PDF)
(XLSX)
ALL CU MVP table 2013 (PDF)
(XLSX)
CrossTabulated Tables
The four new 2year experimental tables with rebalanced income categories and higher income ranges are produced on a research basis using 20152016 data. This work was done to increase the usefulness of CE data. Rebalancing was undertaken to consolidate and balance income ranges. As income rose when the economy recovered, the number of consumer units in the highest income range had become unbalanced and top heavy, while the lowest income ranges contained fewer and fewer CUs. For example, the current 2year income for single men table has seven income ranges, but the $40,000 and over top income range has 36.4 percent of all CUs in it, and contains 35 percent more CUs than the lowest three income ranges combined.
 Age of reference person by income before taxes — (XLSX)
 Region of residence by income before taxes — (XLSX)
 Size of consumer unit by income before taxes — (XLSX)
 Singles by income before taxes — (XLSX)
Generational Tables, 2014 through mid2016
The midyear generational table for July 2015 through June 2016 with an All consumer units column and five additional columns using the birth year of the reference person grouped into generations has been added to this section. The source for the years used in the generational groupings comes from the following report.
The five generational columns are:
Birth year of 1981 or later  sometimes referred to as the Millennial Generation
Birth year from 1965 to 1980  also known as Generation X
Birth year from 1946 to 1964  Baby Boomers
Birth year from 1929 to 1945  Silent Generation
Birth year of 1928 or earlier  Greatest Generation (G.I. Generation)
The birth year of the reference person was approximated based on the collected age of the reference person variable in the CE data.
 2016 Midyear Generational Table — (XLSX)
 2015 Generational Table — (XLSX)
 2015 Midyear Generational Table — (XLSX)
 2014 Generational Table — (XLSX)
Higher Income Table, 2014
CE evaluated the income ranges on the published standard table, and found that over time the number of consumer units in the lowest income columns had decreased. This research table divided the existing $150,000 and over income column into two new columns: $150,000 to $199,999, and $200,000 and over. In 2015, these columns were integrated into the annual “Income before taxes” table.
 2014 Higher Income table — (XLSX)
The generational and higher income tables are now part of our standard table series.
Last Modified Date: December 7, 2017
