Developing an innovative data collection platform for qualitative and mixed-methods research within the U.S. and globally

The U.S. has well-developed research and data tools for monitoring the physical and mental health of the country, but not for monitoring its social health. We need a new monitoring platform to track and uncover emerging challenges in this arena, such as experiences related to racial and gender conflict and discrimination; social isolation and loneliness; misinformation and toxic and conspiratorial beliefs; meaninglessness and distrust; and social deprivation in its many forms.

CPI’s Voices Lab addresses this need through the American Voices Project (AVP), an innovative new research infrastructure for collecting and analyzing mixed-methods data across diverse research topics. The AVP is the country’s first large-sample, nationally representative, mixed methods data platform that makes it possible to discover what people are thinking, feeling, doing, and experiencing. The core of the AVP’s innovative methodology is long-form immersive conversations with individual respondents that promote trust and authentic sharing about one’s family, work, friends, politics, religion, and everyday experiences. These conversations are then followed up with more conventional survey questions and linked to administrative data (when respondents consent to linkage).

This approach is a useful supplement to data collected through conventional surveys because it uncovers feelings and experiences that survey-designers hadn’t thought to elicit and that survey-takers may not be aware of or willing to disclose on a one-off questionnaire. The AVP also complements conventional qualitative research because it is based on large representative samples, allows for secondary analysis, and thus energizes a new cumulative form of qualitative analysis.

The first round of AVP interviewing, funded by a large consortium of top foundations, was completed in 2022. We are now piloting a new public data analysis and dissemination platform in which qualified researchers apply for access to the AVP data, analyze the deidentified data on a secure server, and then write and release research reports (after disclosure avoidance review to ensure that confidentiality is fully protected). The Russell Sage Foundation has released two special issues of the Russell Sage Foundation Journal of the Social Sciences (Part I and Part II) devoted to this initial round of analyses – and see also this series of AVP-based Crisis Monitoring Reports, focused on Americans' experiences as the COVID-19 pandemic unfolded.

CPI’s Voices Lab is currently building a new and improved version of the AVP to serve as a permanent public-use research platform. We are also supporting an international initiative to build a cross-national immersive interviewing platform. This work is supported by a broad interdisciplinary team of scholars.

Funders of CPI’s work in this area have included the Annie E. Casey Foundation; the Bill & Melinda Gates Foundation; the Center for Research on Child Wellbeing at Princeton University; the Chan Zuckerberg Initiative; the David and Lucile Packard Foundation; the Federal Reserve Banks of Atlanta, Boston, Cleveland, Dallas, New York, Philadelphia, Richmond, and San Francisco; the Ford Foundation; The James Irvine Foundation; the JPB Foundation; the National Science Foundation; the Pritzker Family Foundation; and the Russell Sage Foundation.

Using big data to understand interactions within social networks and neighborhoods and to develop new interventions to reduce segregation

The networks of social connections in which all people are embedded have profound effects on social and economic mobility, employment, physical and mental health, access to information, participation in sociopolitical and religious life, and many other everyday experiences. Indeed, the country’s most important social problems often take the form of “network problems,” with examples including loneliness, the spread of disease, and unequal access to advantageous social connections. Because healthy networks are critical for social health, it’s important to build research tools that can be used to examine which neighborhoods, social groups, or schools and universities have healthy networks. CPI’s Social Networks Lab is advancing research in this area by developing a nationally-representative “big data” network dataset, drawing on anonymized cellphone data, that includes a rich set of economic, demographic, and other variables.

This dataset can be used to examine the sources of loneliness, the causal effects of neighborhood policy on social interactions, racial and ethnic differences in social interaction, and much more. To date, our team has completed both an early policy study that identified restaurants as a primary source of COVID-19 infections, and a basic science study showing that, contrary to conventional expectations, large cities promote segregation by supporting restaurants, shopping centers, and other venues that are targeted to very narrow socioeconomic groups (such as fancy restaurants for the one percent).

Examining the reach and impact of refundable tax credits for low-income households and behavioral responses to tax rate increases (such as “millionaire migration”)

Tax policy is an increasingly important tool for taking on poverty and inequality, particularly through federal and state refundable tax credits that deliver resources to families and individuals with low incomes. CPI is studying the reach and impact of existing and proposed tax credits targeted to low-income tax filers, such as EITCs, child tax credits, and renter’s tax credits. This research builds on CPI’s work on the California Poverty Measure.

Tax policy also determines how much revenue will be available to support public spending priorities, such as poverty- and inequality-reducing initiatives. Understanding behavioral responses to these tax policies is important for effective policy design. For example, if increased taxes cause many individuals to move away from higher-tax areas, these policies will be less effective in raising revenues. CPI’s influential research on “millionaire migration,” based on analysis of administrative tax return data, has helped inform this policy debate.

Evaluating and identifying promising evidence-based early childhood interventions

The first few years of a child’s life are a sensitive developmental period during which poverty and other stressors can cause long-lasting harm. This critical period also represents a window of opportunity for carefully targeted policy interventions that can have long-lasting positive impact. CPI has an ongoing stream of research on interventions targeted to young children, such as baby bonds (see also the Economic Inclusion Lab), early childhood training (see also the California Lab), and tax credits (see also the Tax Policy Lab).

Leader: Luigi Pistaferri

The study of poverty and inequality has long privileged measures of income over those of consumption. Does this lead us awry? The purpose of the Consumption RG, which the CPI has founded only recently, is to monitor trends in consumption-based poverty and inequality and to understand the sources of those trends. The analyses within this RG will focus on the individual components of consumption because they are not always moving in lockstep. For example, inequality in the ownership of major durables has been declining, whereas inequality in nondurables, services, and food consumption has been increasing. The trends in food consumption are further complicated, however, because coloric intake is not growing more unequal, partly as a result of assistance provided by SNAP. The main rationale for setting up this RG is that patterns of consumption, especially among the poor, appear to be changing in complicated ways that require careful study and documenting. Although in principle the well-being of households may be better assessed with consumption than income, a consumption-based approach is thus complicated by the very different trends across the components of consumption and by difficulties in knowing the value assigned to the quality of goods that are consumed.

Leaders: Linda Burton, Kathryn Edin, David Grusky

The Supplemental Poverty Measure (SPM) reveals substantial post-1970 reductions in poverty under a constant (i.e., “anchored”) threshold, but this trend masks worrisome developments at the very bottom of the distribution. Although the overall SPM has trended downward since 1970, the SPM for households with less than half of the anchored threshold level (i.e., “deep poverty”) has remained stable since 1968. Even more worrying, the most extreme forms of poverty, such as living on less than $2 per day (per person), have in fact increased over the last two decades. The main tasks of our Poverty and Deep Poverty RG are to describe trends in poverty and deep poverty, to assess the effectiveness of current anti-poverty programs, and to examine the likely payoff to introducing new anti-poverty programs. We present a sampling of relevant projects below.

Frequent Reporting Project: Why are unemployment statistics reported monthly whereas poverty statistics are reported only once a year (and with such a long lag)? The CPI is hard at work solving this problem.

California Poverty Project: The CPI, in collaboration with the Public Policy Institute of California, issues the California Poverty Measure (CPM) annually. There are plans afoot to make it an even more powerful policy instrument. 

Ending Poverty in California: Is it possible to substantially reduce poverty in California by relying entirely on evidence-based programs? It indeed is.

The National Poverty Study: The country’s one-size-fits-all poverty policy ignores the seemingly profound differences between suburban poverty, immigrant poverty, reservation poverty, rural white poverty, deindustrializing poverty, and the many other ways in which massive deprivation plays out in the U.S. The National Poverty Study, which will be the country’s first qualitative census of poverty, takes on the problem.

Income supports and deep poverty: The U.S. does not rely heavily on unconditional cash transfers in its poverty programming. Is this a mistake? The CPI is assisting Y Combinator in providing the first U.S. evidence on unconditional income support since the negative income tax experiments of the 1970s.

Disability and deep poverty: The country’s disability programs are an important anti-poverty weapon. In evaluating their effectiveness, it is important to determine whether the low employment rates among program recipients reflects an underlying (low) capacity for employment, as opposed to the labor-supply effects of the programs themselves. Although it’s long been difficult to assess such labor-supply effects, now there’s a way forward.

Evictions and deep and extreme poverty: Are evictions an important cause of deep and extreme poverty? This line of research examines the extent to which deep and extreme poverty can be reduced with a “housing first” policy that ramps up federal housing programs.

Deep poverty and TANF add-ons: The country is implicitly running hundreds of experiments on how best to structure TANF programs, but it hasn’t had the capacity to evaluate them. Are administrative data the answer?

Leaders: David Harding, Stephen Raphael, Joan Petersilia

Since the mid-1970s, the United States has experienced a precipitous rise in incarceration, with about 2.3 million U.S. adults now incarcerated in state and federal prisons. In recent years, there has been increasing pressure to wind down this commitment to mass imprisonment, and it’s accordingly important to study ways to reintegrate successfully. The Incarceration RG is tasked with monitoring and evaluating the relationship between poverty and sentencing, parole reform, probation, reintegration, and recidivism.

Poverty and the decline in prison population: Will the ongoing decline in California’s prison population bring about an increase in homelessness, mental health service use, and other poverty-relevant outcomes? This line of research will reveal whether ongoing declines in incarceration should be coordinated with increased funding for programs that may substitute for incarceration.

Arrests, race, and poverty: Is reducing arrests the only way to reduce criminal bookings (and the employment-reducing effects of such bookings)? There may be another way.

Leaders: Sanjay Basu, Mark Cullen, Jeremy Freese, David Rehkopf 

The affiliates within the Health Disparities RG are using new computer modeling and statistical techniques to examine how poverty affects the health of children and adults and how some anti-poverty programs are reducing those effects. Here’s a sampling of our projects. 

Income, geography, and life expectancy: Using deidentified tax data and Social Security Administration death records, Raj Chetty and his coauthors have shown that the richest 1 percent live 14.6 years longer, on average, than the poorest 1 percent. Although poor people typically have much shorter lives, Chetty also shows that the extent of this disadvantage depends on the place of residence, thus suggesting that there may be opportunities for policy to reduce the gap in life expectancy.

Infant health and poverty: Which poor neighborhoods are associated with very low birth weights? By identifying neighborhoods that are yielding very low birth weights, we can start to target home visiting and related programs.

Biological mechanisms of disadvantage: We all know that poverty “gets under the skin” and creates lasting disadvantage. Is this because children exposed to poverty-induced stress experience epigenetic changes? We’re going to know very soon.

Income and the developing brain: The prevailing view is that poverty is especially likely to shape children’s early development because of the high plasticity and rapid growth of the brain during the first three years of life. It’s high time for a rigorous study of how income affects the brain function and development of infants and toddlers. 

Leaders: Nicholas Bloom, Raj Chetty, Emmanuel Saez

The CPI is home to some of the country’s most influential analyses of the income and wealth distribution. The purpose of the Income and Wealth RG is to monitor the ongoing takeoff in income inequality, to better understand its sources, and to analyze its implications for labor market performance, educational attainment, mobility, and more. The following is a sampling of the CPI’s research projects within this area.

Trends in income and wealth inequality: What are the key trends in U.S. income and wealth inequality? The U.S. increasingly looks to Emmanuel Saez and his research team for the latest data on U.S. economic inequality.

Distributional National Accounts: In an ambitious infrastructural project, Emmanuel Saez and his team are building a “Distributional National Accounts” based on tax returns, a data set that will eliminate the current gap between (a) national accounts data based on economic aggregates and (b) inequality analysis that uses micro-level tax data to examine the distribution of income but is not consistent with national aggregates. This new data set will in turn make it possible to evaluate the extent to which economic growth, which has long been represented as a preferred poverty-reduction approach, is indeed delivering on that objective.

The rise of between-firm inequality: How much of the rise in earnings inequality can be attributed to increasing between-firm dispersion in the average wages they pay? This question can be addressed by constructing a matched employer-employee data set for the United States using administrative records.

Rent and inequality: It is increasingly fashionable to argue that “rent” accounts for much of the takeoff in income inequality. The Current Population Survey can be used to assess whether this claim is on the mark. 

Leaders: Gregory Acs, David Card, Michael Hout, Jesse Rothstein

The labor market was of course hit very heavily by the Great Recession, as evidenced by (a) the slow recovery of the unemployment rate, (b) and the even slower recovery of the long-term unemployment rate and the prime-age employment ratio (defined as the ratio of employed 25-54 year-olds to the population of that same age). This “jobs problem,” which is especially prominent among low-skill workers, has led to a sharp rise in the number of poor households without any working adults. It also underlies, in part, the sharp increase in the number of disability insurance claims and awards, which in turn has further reduced the supply of labor among low-skilled individuals.

If the first type of “jobs problem” is that there still are not enough of them, the second is that the jobs that are available do not always provide the requisite hours, wages, or security that are needed for a sure pathway out of poverty. As a result, low-skill individuals are not just working less but, even when they are working, there is no guarantee that their jobs will lift them and their families out of poverty. The Labor Markets RG is tasked with conducting research on these and related problems and exploiting administrative and other data to assess possible policy responses to them. We list below a few examples of the work being carried out in this group.

Long-run effects of work incentives: As nonworking poverty increases, the U.S. might well want to turn to new types of work incentive programs. Have these programs worked elsewhere?

Minimum wages and poverty: Throughout the west coast, there are a host of minimum wage “experiments” underway, experiments that have the potential to reset the low-wage labor market in quite fundamental ways. How are these experiments playing out?

Leaders: Linda Burton, Tomás Jiménez, Hazel Markus, Douglas Massey, C. Matthew Snipp

The CPI has an extensive research program on race, ethnicity, immigration, and poverty. The National Poverty Study, for example, is designed to rigorously compare differences across rural black, deindustrialized, reservation, and other “racialized” poverty forms. The CPI also runs a comprehensive program on Hispanic poverty that explores such topics as the “chilling effect” of anti-immigrant laws on program use, the reasons why, contrary to much speculation, the Hispanic poverty rate has not taken off, and the causes of the so-called Hispanic Health Paradox (see, for example, our Pathways Magazine special report on poverty, inequality, and mobility among Hispanics). And one of the CPI’s most distinguished affiliates, Jennifer Eberhardt (who is on the CPI directorate), is carrying out a groundbreaking big-data analysis of policing and race. We list below a sampling of other CPI projects on race, ethnicity, immigration, and poverty.

Poverty among refugees: The U.S. refugee population faces very high rates of poverty, yet we know very little about the effects of different resettlement programs and approaches. There are efforts afoot to exploit available administrative data and begin to find out what works and what doesn’t.

Arrests, race, and poverty: Why are some arrests resolved informally while others are converted into a criminal record that then has a life-long scarring effect? The process of converting an arrest into a criminal booking may play an important role in generating downstream racial disparities.

Reducing the race gap in test scores: The new Stanford Education Data Archive (SEDA) is a rich resource that is providing the most systematic evidence to date on the capacity of school-district policies to reduce the racial gap in test scores.

Poverty and schooling on reservations: Why are test scores and educational outcomes on Native reservations so low (relative to the national average)? In a new project by the noted ethnographer Martin Sánchez-Jankowski, we’ll be learning more about how traditional and formal education are viewed and the ways in which they might be better integrated.