Survey of Income and Program Participation
The Survey of Income and Program Participation (SIPP) is a panel survey operated by the Census Bureau.
Design
Sampling
The SIPP utilizes multi-level stratified sampling to select a panel of housing units from the Census Bureau's Master Address File (MAF). For the 2022 panel, the size was 55,000. All household members are contacted.
The first sampling stage is of primary sampling units (PSU): counties with a population of 7,500+ and collections of adjacent counties otherwise. A county that forms a PSU alone is labeled self-representing (SR), while the counties that are aggregated for a PSU are labeled non-self-representing (NST). For the 2022 panel, SR PSUs generally have a population of 100,000+.
SR PSUs are sampled with certainty. NSR PSUs are sampled with stratified probability proportional to size (PPS). For the 2022 panel, 252 SR PSUs and 434 NSR PSUs were sampled.
The second sampling stage stratifies PSUs by income. Addresses are randomly selected with inflated probability in the low income stratum. For the 2022 panel, this inflation was to a scale of 1.6. This is expected to result in 31% more responses in and near poverty.
Panels may be concurrent. Regardless of panel, data is published by year of collection; all data collected in 2020 is "2020 SIPP".
Frequency
Panels are tracked for several years.
Within a panel, there are waves of contact. Prior to 2014, a panel involved 8 to 16 waves of contact every 4 months. (For a total timeline of 2 to 4 years.) From 2014, a panel involves 4 waves of contact annually. (For a total timeline of 4 years.)
Generally, the SIPP questionnaire asks questions concerning the preceding calendar year (i.e., since January 1). There are exceptions that ask about the prior month or about the date of interview. If a household participates in the panel for all waves, data covering 4 calendar years is collected.
Interviewers attempt to follow movers.
History
The SIPP began in 1983. Panels were tracked for 32 months. Members were contacted every 4 months and asked about the prior 4 months.
The 1996 redesign of SIPP increased the panel duration to 4 years. Computer-assisted interviewing (CAI) was also introduced. Lastly, the policy of oversampling housing units from low income strata was introduced.
The 2014 redesign of SIPP aimed to reduce costs of survey administration. Panel members were contacted annually and the survey instrument was rewritten for builtin consistency checks.