UCLA CCPR Population Working Papers

Studying the social determinants of COVID-19 in a data vacuum

Kate Choi — 2020-05-13

The Canadian government has no plans to release data on the race or socioeconomic status of COVID-19 patients. Therefore, whether COVID-19 is disproportionately affecting certain socio-demographic groups in Canada is unknown. We fill this data void by merging publicly available COVID-19 data with tabular census data to identify risk factors rendering certain geographic areas more vulnerable to COVID-19 infections and deaths. We combine insights obtained from this analysis with information on the socio-demographic profiles of smaller geographic units to predict and display the incidence of COVID-19 infections and deaths in these locales. Like in the U.S., COVID-19 has disproportionately affected black and immigrant communities in Canada. COVID-19 death tolls are also higher in Canadian communities with higher shares of older adults.

Disparities in Vulnerability to Severe Complications from COVID-19 in the United States

Emily Wiemers — 2020-05-28

This paper provides the first nationally representative estimates of vulnerability to severe complications from COVID-19 overall and across race-ethnicity and socioeconomic status. We use the Panel Study of Income Dynamics (PSID) to examine the prevalence of specific health conditions associated with complications from COVID-19 and to calculate, for each individual, an index of the risk of severe complications from respiratory infections developed by DeCaprio et al. (2020). We show large disparities across race-ethnicity and socioeconomic status in the prevalence of conditions which are associated with the risk of severe complications from COVID-19. Moreover, we show that these disparities emerge early in life, prior to age 65, leading to higher vulnerability to such complications. While vulnerability is highest among older adults regardless of their race-ethnicity or socioeconomic status, our results suggest particular attention should also be given to the risk of adverse outcomes in midlife for non-Hispanic Blacks, adults with a high school degree or less, and low-income Americans.

New Approaches to Estimating Immigrant Documentation Status in Survey Data

Heeju Sohn — 2020-06-26

Approximately a quarter of the 43 million immigrants living in the United States are thought to be undocumented. Yet, the lack of accurate population-level information about undocumented immigrants provides fertile ground for public misconceptions, political and media hype, and false claims. The goal is to determine how well descriptions of the undocumented population are likely to mirror the reality of undocumented immigrants’ lives in the US. We compare (1) the distribution of the population by documentation status and (2) distributions of the characteristics of undocumented and documented immigrants produced by two methods. The first method (the “decomposition method”) is a commonly used strategy used in previous work and the second method is an alternative, independent method developed in this article. We used the Survey of Income and Program Participation (SIPP) and the Los Angeles Family and Neighborhood Survey (LAFANS). The existing decomposition method works reasonably well if the data contains information on whether respondents are naturalized citizens or and lawful permanent residents. However, when these variables are missing or problematic, the decomposition method produces biased results. The actual undocumented population in the US may be even more socioeconomically disadvantaged than studies based on existing decomposition methods indicate. This article evaluates methods to conduct reasonably accurate nationally representative, policy relevant research on the lives of undocumented immigrants without potentially jeopardizing members of this vulnerable population.

Explaining the Decline of Child Mortality in 44 Developing Countries: A Bayesian Extension of Oaxaca Decomposition for Probit Random Effects Models

Antonio Pedro Ramos — 2020-09-01

We develop a novel extension of Oaxaca decomposition methods for non-linear random effects models to investigate the decline of infant mortality in 42 low and middle income countries. We analyze micro data from 84 Demographic and Health Surveys where surveys from two time periods were available. We predict mortality at the birth level with a Bayesian hierarchical probit regression models. We use the predictions from these models as input for our new Oaxaca method. Our novel approach accounts for uncertainty in the decompostion results, and allows for point estimates, stan- dard deviations, and posterior distributions of the Oaxaca conclusions. Further, our approach does not depend on assumptions such as matched samples between two surveys and and marginalizes ran- dom effects for variables that are not comparable between surveys, such as location effects. For most countries, declines in infant mortality are due to changes in the regression coefficients, not on covari- ate distributions. However, our decomposition results show that there is considerable heterogeneity between countries and uncertainty on which variable matter the most within countries.

Leave No Child Behind: Using Data from 1.7 Million Children from 67 Developing Countries to Measure Inequality Within and Between Groups of Births and to Identify Left Behind Populations

Antonio Pedro Ramos — 2020-09-01

Background: Goal 3.2 from the Sustainable Development Goals (SDG) calls for reductions in national averages of Under-5 Mortality. However, it is well known that within countries these reductions can coexist with left behind populations that have mortality rates higher than national averages. To measure inequality in under-5 mortality and to identify left behind populations, mortality rates are often disaggregated by socioeconomic status within countries. While socioeconomic disparities are important, this approach does not quantify within group variability since births from the same socioeconomic group may have different mortality risks. This is the case because mortality risk depends on several risk factors and their interactions and births from the same socioeconomic group may have different risk factor combinations. Therefore mortality risk can be highly variable within socioeconomic groups. We develop a comprehensive approach using information from multiple risk factors simultaneously to measure inequality in mortality and to identify left behind populations.
Methods: We use Demographic and Health Surveys (DHS) data on 1,691,039 births from 182 different surveys from 67 low and middle income countries, 51 of which had at least two surveys. We estimate mortality risk for each child in the data using a Bayesian hierarchical logistic regression model. We include commonly used risk factors for monitoring inequality in early life mortality for the SDG as well as their interactions. We quantify variability in mortality risk within and between socioeconomic groups and describe the highest risk sub-populations.
Findings: For all countries there is more variability in mortality within socioeconomic groups than between them. Within countries, socioeconomic membership usually explains less than 20% of the total variation in mortality risk. In contrast, 2 country of birth explains 19% of the total variance in mortality risk. Targeting the 20% highest risk children based on our model better identi es under-5 deaths than targeting the 20% poorest. For all surveys, we report effciency gains from 26% in Mali to 578% in Guyana. High risk births tend to be births from mothers who are in the lowest socioeconomic group, live in rural areas and/or have already experienced a prior death of a child.
Interpretation: While important, di erences in under-5 mortality across socioeconomic groups do not explain most of overall inequality in mortality risk because births from the same socioeconomic groups have di erent mortality risks. Similarly, policy makers can reach the highest risk children by targeting births based on several risk factors (socioeconomic status, residing in rural areas, having a previous death of a child and more) instead of using a single risk factor such as socioeconomic status. We suggest that researchers and policy makers monitor inequality in under-5 mortality using multiple risk factors simultaneously, quantifying inequality as a function of several risk factors to identify left behind populations in need of policy interventions and to help monitor progress toward the SDG.

Measuring Within and Between Group Inequality in Early-Life Mortality Over Time: A Bayesian Approach with Application to India

Antonio Pedro Ramos — 2020-09-02

Most studies on inequality in early-life mortality (ELM) compare average mortality rates between large groups of births, for example, between births from different countries, income groups, ethnicities, or different times. These studies do not measure within-group disparities. The few studies that have measured within-group variability in ELM have used tools from the income inequality literature. We show that measures from the income inequality literature, such as Gini indices, are inappropriate for ELM. Instead we develop novel tools that are appropriate for analyzing ELM inequality. We illustrate our methodology using a large data set from India, where we estimate ELM risk for over 400,000 births using a Bayesian hierarchical model. We show that most of the variance in mortality risk exists within groups of births, not between them, and thus that within-group mortality needs to be taken into account when assessing inequality in ELM. Our approach has broad applicability to many health indicators.

Where has democracy helped the poor? Democratic transitions and early-life mortality at the country level.

Antonio Pedro Ramos — 2020-09-10

The effects of democracy on living conditions among the poor are disputed. Previous studies have addressed this question by estimating the average effect of democracy on early-life mortality across all countries. We revisit this debate using a research design that distinguishes between the aggregated effects of democracy across all countries and their individual effects within countries. Using Interrupted Time Series methodology
and estimating model parameters in a Bayesian framework, we find the average effect of democracy on early-life mortality to be close to zero, but with considerable variation at the country-level. Democratization was followed by fewer child deaths in 21 countries, an increase in deaths in eight, and no measurable changes in the remaining 32 cases. Transitions were usually bene cial in Europe, neutral or bene cial in Africa and Asia, and neutral or harmful in Latin America. The distribution of country-level effects is not consistent with common arguments about the conditional effects of democratic transitions. Our results open a new line of research into the sources of theses heterogeneous effects.

Baby Bonus, Fertility, and Missing Women

Wookun Kim — 2020-10-13

This paper presents novel causal evidence on the effects of pro-natalist financial incentives on
babies. I exploit rich spatial and temporal variation in the generosity of cash transfers provided
to families with newborn babies and the universe of birth, death, and migrant registry records
in South Korea. I find that the total fertility rate in 2015 would have been 3% lower without
the cash transfers. These cash transfers were particularly effective among working mothers
and encouraged them to have second and third children. This selection of working mothers into
childbearing led to a decrease in gestational age, which in turn led to an overall reduction in birth
weight, but no change in early mortality. The cash transfers had an unintended consequence of
correcting the unnaturally male-skewed sex ratio closer to its natural level.

Effects of Randomized Corruption Audits on Early-Life Mortality in Brazil.

Antonio Pedro Ramos — 2020-10-13

Background: Various studies suggest that corruption affects public health systemsacross the world. However, the extant literature lacks causal evidence about whether anti-corruption interventions can improve health outcomes. We examine the impact of randomized anti-corruption audits on early-life mortality in Brazil.

Methods: The Brazilian government conducted audits in 1,949 randomly selected municipalities between 2003 and 2015. To identify the causal effect of anti-corruption audits on early-life mortality, we analyse data on health outcomes from individuallevel vital statistics (DATASUS) collected by Brazil’s government before and after the random audits. Data on the audit intervention are from the Controladoria-Geral da Uni˜ao, the government agency responsible for the anti-corruption audits. Outcomes are neonatal mortality, infant mortality, child mortality, preterm births, and prenatal visits. Analyses examine aggregate effects for each outcome, as well as effects by race, cause of death, and years since the intervention.

Results: Anti-corruption audits significantly decreased early-life mortality in Brazil. Expressed in relative terms, audits reduced neonatal mortality by 6.7% (95% CI -8.3%, -5.0%), reduced infant mortality by 7.3% (-8.6%, -5.9%), and reduced child mortality by 7.3% (-8.5%, -6.0%). This reduction in early mortality was higher for nonwhite Brazilians, who face significant health disparities. Effects are greater when we look at deaths from preventable causes, and show temporal persistence with large effects even a decade after audits. In addition, analyses show that the intervention led to a 12.1% (-13.4%, -10.6%) reduction in women receiving no prenatal care, as well as a 7.4% (-9.4%, -5.5%) reduction in preterm births; these effects are likewise higher for nonwhites and are persistent over time. All effects are robust to various alternative specifications.

Interpretation: Governments have the potential to improve health outcomes through anti-corruption interventions. Such interventions can reduce early-life mortality and mitigate health disparities. The impact of anti-corruption audits should be investigated in other countries, and further research should further explore the mechanisms by which combating corruption affects the health sector.

A Practical Revealed Preference Model for Separating Preferences and Availability Effects in Marriage Formation

Shuchi Goyal — 2020-11-12

Many problems in demography require models for partnership formation that separate latent preferences for partners from the availability of partners. We consider a model for matchings within a bipartite population where individuals have utility for people based on known and unknown characteristics. People can form a partnership or remain unpartnered. The model represents both the availability of potential partners of different types and preferences of individuals for such people. We develop Menzel’s (2015) framework to estimate preference parameters based on sample survey data on partnerships and population composition. We conduct simulation studies based on new marriages observed in the Survey for Income and Program Participation (SIPP) to show that, for realistic population sizes, the model recovers preference parameters that are invariant under different population availabilities. We also develop confidence intervals that have correct coverage. This model can be applied in family demography to understand individual preferences given different availabilities.