Attachment T - 2021 Response Rates for the Interview and Diary Surveys

U .S. Department of Labor Bureau of Labor Statistics

2 Massachusetts Ave. N.E.

Washington, D.C. 20212

August 31, 2022

MEMORANDUM FOR: DAVID C. SWANSON, Supervisory Mathematical Statistician

Branch of Consumer Expenditure Surveys

Division of Price Statistical Methods

Office of Prices and Living Conditions

FROM: SHARON L. KRIEGER, Mathematical Statistician

Branch of Consumer Expenditure Surveys

Division of Price Statistical Methods

Office of Prices and Living Conditions

SUBJECT: 2021 Response Rates for the Interview and Diary Surveys

In 2021 the response rate for the Interview survey was 45.6%. That means 45.6% of “eligible” CUs participated in the survey. The table below shows this 2021 response rate along with those for 2017 thru 2020 as a comparison.

Interview Survey

Collection Year	CUs Designated for the Survey	Type B or C Nonresponses	Total Eligible CUs	Type A Nonresponses	Interviews	Response Rate for Eligible CUs
2017	47,949	7,756	40,193	15,714	24,479	60.9%
2018	48,018	7,652	40,366	17,207	23,159	57.4%
2019	47,799	7,410	40,389	18,688	21,701	53.7%
2020	50,939	8,027	42,912	22,667	20,245	47.2%
2021	51,944	7,306	44,638	24,304	20,334	45.6%

In 2021 the response rate for the Diary survey was 40.0%. That means 40.0% of “eligible” CUs participated in the survey. The table below shows this 2021 response rate along with those for 2017 thru 2020 as a comparison.

Diary Survey

Collection Year	CUs Designated for the Survey	Type B or C Nonresponses	Total Eligible CUs	Type A Nonresponses	Interviews	Response Rate for Eligible CUs
2017	24,696	4,586	20,110	8,452	11,658	58.0%
2018	24,522	4,389	20,133	9,054	11,079	55.0%
2019	24,566	4,322	20,244	9,562	10,682	52.8%
2020	36,132	5,693	30,439	19,824	10,615	34.9%
2021	36,106	5,902	30,204	18,137	12,067	40.0%

Type B or C nonresponses are housing units that are vacant, nonexistent, or ineligible for interview. Type A nonresponses are housing units which the interviewers were unable to contact or the respondents refused to participate in the survey. The response rates stated above are based only on the eligible housing units (i.e., the designated sample less Type B and C nonresponses).

COVID-19’s Impact on Response Rates

The COVID-19 pandemic contributed to a decrease in the response rates from 2019 to 2020, from 53.7% to 47.2% in the Interview survey, and from 52.8% to 34.9% in the Diary survey. In large part, the decrease was caused by changes to data collection protocols at the Census Bureau.

From mid-March until the end of June 2020, the Census Bureau stopped personally visiting CUs and instead started contacting them exclusively by telephone. The Census Bureau was able to find telephone numbers for most addresses in the survey, but not all of them. That meant some of the CUs had an unknown eligibility status. CUs with unknown eligibility were treated as Type A nonresponses, which caused a noticeable drop in the response rates.

After June 2020, the Census Bureau gradually resumed its usual data collection protocols. However, the pandemic’s impact on response rates lingered throughout the rest of the year anyway. This impact was bigger in the Diary survey than in the Interview survey because it was harder for the Census Bureau to get interviews in the Diary survey than in the Interview survey. In the Diary survey most CUs were being contacted for the first time and many of their telephone numbers were unknown. In contrast, in the Interview survey most CUs had already been contacted in a previous quarter and their telephone numbers were known.

In 2021, the pandemic’s impact on response rates diminished as the Census Bureau continued to return to its usual data collection protocols. This resulted in a relatively small decrease in the response rates from 2020 to 2021 for the Interview survey, from 47.2% to 45.6%, and in a relatively large increase in the response rates from 2020 to 2021 for the Diary survey, from 34.9% to 40.0%.

The Sample Expansion’s Impact on Sample Sizes

The “sample expansion” in 2020 contributed to an increase in the number of “CUs Designated for the Survey,” from 47,799 to 50,939 in the Interview survey, and from 24,566 to 36,132 in the Diary survey. The sample expansion went into effect in April 2020 in the Interview survey and in January 2020 in the Diary survey. The reason for the expansion is that in 2020 the CE survey added some outlet-related questions to its questionnaires to replace CPI’s outlet information formerly collected by the TPOPS survey, and this meant the sample size in urban areas needed to be increased to ensure outlet sampling frame sufficiency in those areas.

cc:	CESMD	DCES_SUPER	Others
	S. Ash	J. Brattland	A. Cobet
	S. Krieger	B. Creech	T. Garner
	B. Nix	S. Curtin
	B. Steinberg	M. Edwards
	D. Swanson	L. Erhard
	L. Vermeer	B. Rigg
		A. Safir

4