CE Overview Jay T. Ryan Chief, Division of Consumer Expenditure Survey December 8, 2010.

CE Overview

Jay T. RyanChief, Division of

Consumer Expenditure Survey

December 8, 2010

2

Overview of the Consumer Expenditure Survey (CE)

Goals of the CE Scope / Coverage Sampling Stakeholders Weighting Data Collection Classification System Estimation Publication

3

Goals of the CE

To meet the need for timely and detailed information about the spending patterns of different types of households.

To provide the basis for revising the expenditure weights for the Consumer Price Index (CPI) every two years.

4

Scope / Coverage The CE is a nationwide survey designed

to represent the U.S. civilian noninstitutionalized population.

The U.S. Census Bureau collects CE data from consumer units (CUs) - people living at one address who share living expenses. In most cases, CUs are the same as households.

Survey participants report dollar amounts for all non-investment purchases. Business expenses and reimbursements are excluded.

5

Sampling The CE is a designed to represent all urban

and rural areas in the U.S.

Sample selection begins with the definition and selection of geographic areas called primary sampling units (PSUs).

Within the (currently) 91 PSUs, the sampling frame (list of addresses from which the sample is drawn) is generated from the 2000 Census 100-Percent Detail File.

The sampling frame is augmented by a sample of addresses drawn from new construction permits.

6

Stakeholders CPI

Bureau of Economic Analysis (BEA)

IRS

Defense

Census

Other Federal Agencies

State governments

Businesses

Researchers

General Public

7

Quick facts Program budget: FY10, $43 million CE staff at BLS: 60 economists,

statisticians, survey methodologists, systems analysts

Years as a continuous survey: 30 Response rates: about 75% for both

surveys Interview addresses in sample: 60,000

annually (and about 35,000 good interviews)

Diary addresses in sample: 13,000 annually (and about 14,000 good weekly diary interviews)

Interview surveys conducted by telephone: 33%

8

Weighting

Estimating average expenditures is done by weighting.

Each CU is assigned a weight, which is the number of similar consumer units in the U.S. civilian noninstitutionalized population represented by the sampled CU.

Population weights are updated quarterly based on U.S. population figures from the U.S. Census Bureau’s Current Population Survey (CPS).

9

Data Collection

The U.S. Census Bureau collects CE data for the BLS using two different surveys with separate samples:- the quarterly Interview Survey - the weekly Diary Survey

Respondents in both surveys are asked to describe all of their spending.

10

Data Collection

The Interview Survey is used to measure amounts spent on large purchases, such as major appliances or vehicles, and those purchases that occur regularly like monthly payments for rent and utilities.

11

Data Collection The Diary Survey

collects information about frequent purchases like food and personal care items that are difficult to remember over longer periods of time, like the three-month recall used in the Interview Survey.

12

Classification System

Detailed expenditures are grouped into narrowly related subcategories designed to closely match CPI categories. For example, fresh pineapples are grouped in “Other fresh fruits.”

13

Classification System

The subcategories are combined into major groups for publication. “Other fresh fruits” are combined with apples, bananas, oranges, and other citrus fruits into the published “Fresh Fruits” category.

14

Estimation

Average annual expenditures are estimated using weighted means for each CU.

Source data from the Interview and Diary Surveys are integrated by income and other demographic variables to create total expenditures.

15

Estimation

Source selection from either the Diary or the Interview is determined by the number and average amounts of reported purchases from each survey. The higher mean is selected if statistically significant.

Sources are reevaluated every two years.

Published dollar expenditures are the means for all consumer units, both purchasers and nonpurchasers.

16

Estimation

Data Quality Adjustment Methods used to improve estimates:

Allocation

Imputation

17

Estimation

Allocation is used when a CU reports a single combined expenditure amount for multiple purchased items.

For example, the CU reports spending $100 for shirts, pants, and socks. The $100 is divided into the shirts, pants, and socks item categories.

18

Estimation

Imputation corrects for missing or invalid entries. All fields except assets and income taxes are subject to imputation.

Beginning with the 2004 published data, all missing income fields are imputed to have a complete record for all households.

19

Major categories of published annual average

expenditures include: Food

Housing

Apparel and services

Transportation

Healthcare

Entertainment

Reading

Personal care products and services

Education

Cash contributions

Personal insurance and pensions

20

Publication & Microdata Published tables are classified by 13

standard characteristics, such as Quintiles of Income, Size of CU, Age, Education, Race, Owner/Renter, etc.

Microdata data CDs available for purchase

CE website:www.bls.gov/cex

Gemini project:http://www.bls.gov/cex/geminiproject.htm

Contact Information

Jay T. RyanDivision Chief

Consumer Expenditure Surveywww.bls.gov/cex

[email protected]