«Contesting the streets Volume 18, number 1 • 2016 U.S. Department of Housing and Urban Development | Office of Policy Development and Research ...»
PSID began with a national sample of about 5,000 households with approximately 18,000 individuals (Hill, 1992). The study has followed these individuals and their descendants at each wave, leading to sample growth over time. PSID’s 2015 wave includes about 10,000 households containing 25,000 individuals. Respondents have been interviewed by telephone since 1973, with interviews conducted annually from 1968 to 1997 and biennially thereafter. Wave-to-wave core reinterview response rates typically range between 96 and 98 percent. PSID data are available free of charge to the public and have been used for approximately 4,000 peer-reviewed publications, including more than 700 dissertations. The study’s design has been replicated in many countries around the world. PSID is regularly used for policy analysis by U.S. federal government agencies.
On the National Science Foundation’s (NSF’s) 60th anniversary, it named PSID as 1 of the 60 most significant scientific advances ever funded by NSF.
PSID’s unique features include its national representativeness, the long duration of the panel, its genealogical design, and its broad and deep content. PSID includes adult respondents of all ages and follows individuals across the entire lifecourse. Adult children are interviewed in their own family units after they achieve economic independence from their parents’ households. This unique self-replacing design means that, for many families, PSID includes self-reported information on three (and occasionally four, or even five) generations of the same family at various points in their lifecourse. PSID is the only survey ever collected on lifecourse and multigenerational economic conditions in a long-term panel representative of the full U.S. population (see McGonagle et al.,
2012). With sample weights, PSID data are nationally representative of U.S. families. Results based on analyses of PSID data can therefore be used to make statements about the entire U.S. population and also major demographic subgroups defined by age, gender, income, and race/ethnicity.
In addition to collecting rich information on housing and neighborhood characteristics, PSID collects data on a wide array of economic, social, demographic, geospatial, health, and psychological factors, supporting multidisciplinary research. In 2015, the 76-minute interview collected data on employment; earnings; income from all sources; education; expenditures; transfers; health;
emotional well-being; mortality and cause of death; marriage and fertility; housing; residential location; participation in government programs; financial distress, including problems paying debt such as mortgages and foreclosure; vehicle ownership; wealth and pensions; and philanthropy.1 Many of these areas have been included in the PSID instrument since 1968 and measured consistently over time. Hundreds of additional variables in other domains have been collected in various waves throughout the history of PSID. Most of the data are publicly available on PSID’s online Data Center (http://www.psidonline.org/), with certain sensitive or disclosive variables available under contractual arrangements.
Substantial data on home learning environments, neighborhood characteristics, and housingrelated decisionmaking are collected in the PSID Child Development Supplement (CDS) and the PSID Transition into Adulthood Supplement (TAS), major ongoing studies of children and young adults in PSID families. CDS began in 1997, with the goal of providing researchers with a comprehensive, nationally representative, prospective database of young children and their families for studying how family, neighborhood, and school characteristics influence cognitive and behavioral development and health. Children and caregivers were reinterviewed 5 years and 10 years after the original interview. Between 2005 and 2015, the same children were followed into young adulthood once they turned 18 years of age in the six-wave TAS. TAS bridges the period between childhood, when data were collected as part of CDS, and economic independence in adulthood, when sample members become eligible to be interviewed as household heads in PSID. Together, the resulting CDS-TAS archive of this original cohort of CDS children provides up to 18 years of prospective information on a cohort of 3,500 children. A new round of CDS was launched in 2014 (CDSand will collect information on all children in PSID households every 5 years. Children from CDS-2014 will continue to be followed into adulthood in future waves of TAS and PSID.
Information on Housing and Neighborhoods Considerable information about housing and neighborhood characteristics has been collected in every wave of PSID (see exhibit 1). Topics include dwelling characteristics, housing utilities, residential mobility and migration, housing-related financial information and consumption expenditures, mortgage distress, and neighborhood characteristics based on geospatial identifiers and administrative data. Information on home and neighborhood characteristics and the emergence of financial independence and housing-related decisionmaking has also been collected in CDS and TAS.
Since the start of PSID, data have been collected on dwellings characteristics, including dwelling type and number of rooms. Information is also collected about characteristics of retirement and senior The 2015 Core PSID questionnaire is available at ftp://ftp.isr.umich.edu/pub/src/psid/questionnaires/q2015.pdf.
Housing and Neighborhood-Related Questionnaire Content in PSID-CDS-TAS PSID Housing-Related Content Dwelling characteristics Housing type: house, duplex, apartment, condominium, townhouse, mobile home.
Number of rooms.
Number of individuals sharing living space.
Whether retirement or senior community and services offered.
Housing utilities: type, cost, and frequency of payments Home heat, water, sewer, electricity, telephone service, air-conditioning, cable television, Internet connection.
Use of government programs for utility costs.
Housing finances Whether owns or rents.
Current market value of dwelling.
Rental agreement detail.
Property tax amounts.
Home insurance amounts.
Use of government programs for housing.
Housing consumption expenditures Annual expenditures for home repairs and maintenance and for household furnishings and equipment.
Mortgage distress Falling behind in housing payments.
Expectations about housing payment difficulties in coming year.
Residential mobility, reasons for moving, moving intentions Residential change timeline, including timing and address of all residential moves occurring during past 2 calendar years.
Reason for moving, including changes in employment, school attendance, or marital status; for an improved living situation; to save money; or because of a financial shock, such as bankruptcy, foreclosure, or eviction.
Likelihood of moving in near future and reason.
CDS and TAS Housing and Neighborhood-Related Content Neighborhood characteristics and home environment (CDS) Ratings by primary caregiver on neighborhood stability, social cohesion, safety, and satisfaction.
Ratings by interviewer on HOME Inventory, including availability of reading material, technology, musical instruments, and toys; features of play areas; lighting adequacy; clutter; cleanliness; space; noise;
and condition of nearby homes and buildings.
Location and economic independence (TAS) Where living during different parts of year, including parent’s home, college dormitory, apartment or rented home, military base, or other institution.
Whether moved for an employment opportunity.
Help received from parents and relatives for housing payments and amounts received.
Restricted Use Data Assisted housing administrative linkages (PSID) Receipt of government housing subsidies (waves 1968, 1970 through 2009).
Type and class of subsidy, including public housing, low-income housing tax credit, Farmers Home Administration, other federal, other state, other project-based housing, other tenant-based housing.
Geospatial data (PSID-CDS-TAS) For all waves: FIPS county and place; 5-digit ZIP Code; MSA and CBSA; census tract, block, and block group; match-quality indicators.
CBSA = Core Based Statistical Area. CDS = Child Development Supplement. FIPS = Federal Information Processing Standard. HOME = Home Observation Measurement of the Environment. MSA =metropolitan statistical area. PSID = Panel Study of Income Dynamics. TAS = Transition into Adulthood Supplement.
housing communities. Since the earliest waves, information has been collected on the type and cost of utilities, including source of home heat, air-conditioning, and the use of government subsidies for utility costs. Questions were added more recently about cable television and Internet connections.
Data on residential mobility, moving intentions, and reasons for moving have been collected throughout the study. For each wave, information is obtained on all recent residential moves and their timing.
Specific reasons for each move are also collected. Respondents provide estimates of the likelihood of moving in the near future and describe life events that may trigger potential moves.
Detailed housing finance information has been obtained since the earliest waves of PSID, including current market value of the dwelling, details of rental agreements and mortgages, and the use of government subsidies. Starting in 2005, data have been collected on housing-related consumption expenditures, including annual costs of home repairs and maintenance and of household furnishings and equipment as part of a complete series on consumption expenditures.
At the onset of the 2009 housing crisis, PSID began collecting extensive information about mortgage distress, including falling behind in payments, mortgage modifications, foreclosure activity, and expectations about housing payment difficulties in the coming year. This information has been used extensively to describe and analyze families’ difficulties with home mortgages during the Great Recession (December 2007 to June 2009), including by the Federal Reserve Board (for example, Sherlund, 2010) and by others (for example, Lin, Liu, and Xie, 2016).
Although nearly all the data are freely available in the public domain, certain information about housing and geography is available only through a restricted data use contract to maintain the confidentiality of PSID respondents, including geospatial identifiers below the level of state and administrative linkages to external databases. These geospatial identifiers and administrative data have been widely used as a means of characterizing the neighborhoods in which respondents live. Three levels of geospatial data are available: census tract, block group, and block. Residential addresses have been geocoded for all waves of the study using four different versions of census geography: addresses from 1968 through 1985 were geocoded using both the 1970 and 1980 census geography;
those from 1968 through 1999, using the 1990 census geography; those from 2001 through 2009, using the 2000 census geography; and those since 2011, using the 2010 census geography.
Linkages of PSID data to administrative records, including the receipt of government housing subsidies, are also available. These linkages are generated through a process that matches addresses of PSID families in each wave with those in the Assisted Housing Database collected by the U.S.
Department of Housing and Urban Development (HUD). Information is available regarding whether a PSID address in a given year corresponds to an assisted housing address, and, if so, the type of assisted housing, including whether subsidized by HUD, by the former Farmers Home Administration, by tax credits administered by the U.S. Department of the Treasury, or through state-level housing subsidy programs.
Other administrative data include identifiers for primary and secondary schools attended by children in CDS and TAS. These school identifiers link PSID children to detailed information about their schools from the Common Core of Data and Private School Universe Survey prepared by the U.S. Department of Education’s National Center for Education Statistics (NCES). PSID and TAS
sample members who have attended college, university, or technical and vocational postsecondary institutions have identifiers that can be linked to data from the NCES Integrated Postsecondary Education Data System.
With a central goal of providing information about how child health and well-being are influenced by home and neighborhood environments, CDS has collected particularly rich information on these latter topics. All waves include detailed information collected from the child’s primary caregiver on neighborhood stability, social cohesion, safety, and satisfaction. The Home Observation Measurement of the Environment Inventory (Caldwell and Bradley, 2003), designed to measure the quality and quantity of stimulation and support available to a child in the home environment, has been included in all waves of CDS.
Finally, TAS collects housing-related content that reflects the high rates of mobility and emerging financial independence characteristic during young adulthood. For instance, information is obtained about where young adults live “most of the time,” including in parent’s home, a college dormitory, an apartment or rented home, a military base, or other institution. Because young adults move frequently, this information is collected for different parts of the year (October through April and May through August); information regarding whether a move occurs for an employment opportunity is also captured. TAS also assesses young adults’ economic independence by collecting information about help received from parents and relatives in paying rent or a mortgage. Additional data are collected across many other domains, such as self-perceptions; future expectations for schooling, careers, and employment; and information regarding health, wealth, and income that can support rich models of housing decisions and their effects on social and economic outcomes during young adulthood.