Search

Consumption is a well-being measure that is determined by a combination of resources (e.g., income, in-kind benefits, assets, debt, time) available to households, their circumstances, and their preferences. In this study, we derive consumption poverty statistics using a consumption measure that includes the flow of services from owner-occupied housing and vehicles and in-kind transfers. The base data are from the US Consumer Expenditure Survey Interview from 2015 through 2020. The consumption poverty rate (using an absolute threshold anchored to the 2015 relative consumption poverty rate) declines from 16.8 percent in 2015 to 11.5 percent in 2020.

When developing anti-poverty policies, policymakers need accurate data on the prevalence of poverty. In this paper, we focus on subjective poverty, a concept which has been largely neglected in the literature, though it remains a conceptually appealing way to define poverty. The primary goal of this study is to re-examine the concept of subjective poverty measurement and to estimate trends in subjective income poverty rates in the European Union. Our estimations are based on a Minimum Income Question using data from a representative survey, EU-SILC. We find robust empirical evidence of decreasing trends in subjective poverty in 16 of 28 EU countries. We conjecture that trends in subjective poverty may reflect changes in societies which are not captured by official poverty indicators, and our results thus enrich the existing data on general poverty trends in the EU. Lors de l’élaboration des politiques de lutte contre la pauvreté, les décideurs ont besoin de données précises sur la prévalence de la pauvreté. Dans cet article, nous nous concentrons sur la pauvreté subjective, un concept qui a été largement négligé dans la littérature, bien qu’intéressant pour définir la pauvreté. L'objectif principal de cette étude est de réexaminer le concept de mesure subjective de la pauvreté et d'estimer les tendances des taux de pauvreté monétaire subjective dans l'Union européenne. Nos estimations sont basées sur une question relative au revenu minimum en utilisant les données d'une enquête représentative, EU-SILC. Nous trouvons des preuves empiriques solides que la pauvreté subjective a tendance à baisser dans 16 des 28 pays de l'UE. Nous supposons que les tendances de la pauvreté subjective peuvent refléter des changements sociétaux qui ne sont pas pris en compte par les indicateurs officiels sur la pauvreté. Ainsi, nos résultats viennent enrichir les données existantes sur les tendances générales de la pauvreté dans l'UE.

Studies into the relation between subjective perceptions of individuals and objective economic conditions have usually resulted in ambiguous empirical findings. Whilst most studies perceive subjective welfare as being operationalized by indicators of happiness or life satisfaction, we narrow the approach to an economic domain of subjective well-being—perceptions of poverty. We argue that our approach better reflects the economic dimension, as the former may include numerous non-economic domains. We use a case study of two countries—Czechia and Slovakia—which underwent early economic transition as a common state in 1989–1992, then became independent states in 1993, after the dissolution of Czechoslovakia. We base our findings on three historical data sets covering a period from around the end of the communist era to the early years after the split, and recent data from EU Statistics on Income and Living Conditions (2005–2016). Despite initially small differences in subjective poverty levels in socialist Czechoslovakia, a considerably larger drop in economic performance during the transition period in Slovakia than in Czechia resulted in a sharp widening of the subjective poverty gap. The recent data suggests that, despite a high degree of actual economic convergence of Slovakia and Czechia, the gap in subjective perceptions of poverty is declining at a remarkably slower pace. We argue that relatively fast economic growth is not necessarily associated with a commensurate decline in subjective poverty perceptions. Our results thus support the Easterlin Paradox, although we substitute happiness by an economic dimension of subjective well-being.

The onset of the coronavirus disease 2019 (COVID-19) pandemic led to considerable changes in consumer spending behavior in the United States Using data from the Household Pulse Survey, this article examines the extent of pandemic-related behavioral changes reported in August 2020 The article also shows how these changes differed across generations and geography

When comparing economic well-being using income or expenditures, an equivalence scale is often used to adjust for differences in characteristics that affect needs. For example, a family of two is a...

When comparing economic well-being using income or expenditures, an equivalence scale is often used to adjust for differences in characteristics that affect needs. For example, a family of two is assumed to need more income than a single person, but not twice as much due to the economies of scale in consumption. In this study, we ask whether it is appropriate to use a common equivalence scale when comparing economic well-being across countries and/or time if consumption expenditure patterns differ?Based on an Engel methodology, we estimate equivalence scales for a diverse set of countries (Canada, France, Israel, Poland, South Africa, Switzerland, Taiwan, United States) in different time periods (1999-2012). We find considerable differences in economies of scale across countries, as well as increases over time. Notably, we find that economies of scale are larger than those implied by the widely accepted 'square root of household size' equivalence scale.Our results indicate that using a common equivalence scale to compare economic well-being across countries and/or time is misleading. Specifically, if economies of scale are understated (as is the case when using the 'square root of household size'), the relative poverty experienced by larger versus smaller families is being overstated.

This chapter describes the research and historical events that led to the production of the Supplemental Poverty Measure for the United States. This chapter includes an overview of the official measure, a description of numerous attempts to replace the official measure with an alternative poverty measure, and finally the recommendations for the production of a measure that supplements rather than replaces the official measure. These recommendations were based on an enormous body of research designed to improve poverty measurement in the United States undertaken by economists and statisticians at the Bureau of Labor Statistics; Census Bureau; other federal, state, and local agencies; universities; and think tanks around the country. The ultimate result of this work is a poverty measure that is designed to better reflect current living standards in the United States and the goals of today's society as reflected in tax and transfer policies.

The poverty measure presented compares spending needs to resources available to meet those needs. The analysis is for the U.S.; however, lessons from other countries regarding desirable properties of a poverty measure are considered. A primary focus is internal consistency between thresholds and resources. This study is among the first for the U.S. to describe an internally consistent poverty measure, drawing from recommendations of the U.S. National Academy of Sciences (NAS). Thresholds reflect spending needs as “outflows.” Resources measure “inflows” available to meet spending needs. The U.S. Consumer Expenditure Survey is used for thresholds, and the Current Population Survey is the basis for resources. Trends are reported with comparisons to the official and a relative measure. An important finding is that increases in expenditures for shelter, captured in the NAS thresholds, suggest a greater increase in the number of families not able to meet basic needs than is reflected by official poverty statistics over this time period.

Research linking macro and micro statistics of dwelling services is in its infancy in the U.S. including work by the Bureau of Economic Analysis, Bureau of Labor Statistics, and the Census Bureau. Comparisons of aggregated estimates generated from micro-level data to estimates at the macro-level can inform both levels on the accuracy and precision of methods and data sources. In this study, the treatments of housing in the macro statistics of the National Accounts and in the micro statistics of household expenditure and income surveys are examined. Three approaches to value dwelling services using household survey data are compared: capitalization rate, hedonic, and rental equivalence. Estimates are produced using data from the U.S. Consumer Expenditure Survey and American Housing Survey. Estimated aggregates of implicit net rental income from owner-occupied housing are compared to the aggregate value in the National Accounts. Possible sources of differences in the macro- and micro-based aggregates are discussed. The effects of adding net implicit rental income on income distributions are examined, particularly on inferences about the relative well-being by the age of householder. Overall, only marginal reductions in income equality result when net rental incomes are added to before tax money income; this only occurs when reported rental equivalence and return to home equity are used as methods of rent estimation.

Aggregate under-reporting of household spending in the Consumer Expenditure Survey (CE) can result from two fundamental types of measurement errors: higher-income households (who presumably spend more than average) are under-represented in the CE estimation sample, or there is systematic under-reporting of spending by at least some CE survey respondents. Using a new data set linking CE units to zip-code level average Adjusted Gross Income (AGI), we show that the very highest-income households are less likely to respond to the survey when they are sampled, but unit non-response rates are not associated with income over most of the income distribution. Although increasing representation at the high end of the income distribution could in principle significantly raise aggregate CE spending, the low reported average propensity to spend for higher-income respondent households could account for at least as much of the aggregate shortfall in total spending.

The purpose of this paper is to further the understanding of subjective measures used to assess poverty and to add to the literature on poverty measurement methodology. In particular, the paper focuses on the minimum income question (MIQ) first proposed by Goedhart and colleagues (1977). Data from the U.S. Consumer Expenditure Survey and from a Dutch newspaper survey are used. The primary contribution of the paper is the inclusion of household expenditures as additional explanatory variables of minimally necessary income. Significant differences between the coefficients of several categories of expenditures, particularly for leisure, appear to reveal differences in the interpretation of the minimum income question by respondents. Thus, we question the underlying assumption of the MIQ that everyone adheres the same welfare meaning to the phrase “minimally necessary income,” and conclude that the resulting thresholds should not be used as to measure poverty before further research has been carried out to explore what respondents are thinking when they answer questions such as the MIQ.

In March 2010 an Interagency Technical Working Group (ITWG) released guidelines on thresholds and resources for a Supplemental Poverty Measure (SPM). The ITWG recommended that thresholds include in-kind benefits that are accounted for in resources; however, only limited in-kind benefit information is available in the Consumer Expenditure Survey (CE) Interview component, the data source upon which the thresholds are based. For example, the CE collects information on food expenditures that implicitly include the cash value of benefits from the Supplemental Nutrition Assistance Program (SNAP) but no information on other food programs. This study introduces a new method, the CPS Program Participation Method, of imputing benefits for the National School Lunch Program (NSLP) and the Women, Infants, and Children program (WIC). In this study, data from the Current Population Survey (CPS), the data source upon which the SPM resource measure is based, are used to model the participation of CE households in the NSLP and WIC using the CPS Program Participation Method. These CPSbased participation rates for NSLP and WIC are then used along with U.S. Department of Agriculture information to assign benefit levels to the CE households. Thresholds based on the CPS Program Participation Method are produced for 2009 and compared to thresholds based on a method based on program eligibility guidelines, the CE Eligibility Method. SPM thresholds are produced by housing types as well as overall. No poverty rates using these thresholds are produced. Results reveal that the CE Eligibility Method overall threshold is higher than the CPS Program Participation Method overall threshold. This is not surprising since the CE threshold is based on eligibility while the CPS threshold is based on program participation. The paired CE and CPS based thresholds are also statistically significantly different from each other for owners with mortgages and for owners without mortgages. When housing tenure thresholds are compared to each other within each method group, statistically significant differences arise for two of the three pairs of thresholds. In particular, the thresholds for owners without a mortgage were found to be different from the thresholds of both owners with a mortgage and renters, while the thresholds for owners with a mortgage and renters did not differ from each other at the significance levels used for testing.

Inequality in U.S. consumption expenditures is examined using the Lerman and Yitzhaki covariance method for decomposing the Gini coefficient by factors. From the decomposition, nonparametric estimates of elasticities with respect to total expenditures are derived. Using data from the 1987 U.S. Consumer Expenditure Interview Survey, the overall Gini based on total annual consumption expenditures is 0.33. Consistent with parametric expenditure elasticities, the Gini expenditure elasticities for household operations, apparel and services, and entertainment are relatively high. An implication of this result is that taxing such goods and services will increase the progressivity or reduce the regressivity of the tax system. Copyright 1993 by MIT Press.

In this note I provide results from ongoing research at the Bureau of Labor Statistics on experimental poverty thresholds.

This research compares median-based thresholds with ones based on the 33rd percentile using Consumer Expenditure Interview Survey (CE) data from 2004 quarter one through 2009 quarter one. Thresholds for reference families are produced using two different approaches: (1) calculating the thresholds based on the expenditure records of reference families composed of two adults with two children; and (2) calculating the thresholds using expenditures from all consumer units participating in the CE, but first converting their expenditures into adult equivalent value using the three-parameter equivalence scale before identifying the median and percentile values. Different updating mechanisms examined include: (1) an annual recalculation of the 33rd percentile of the reference family’s outlays on FCSU, and (2) changes in the reference family’s median spending or consumption based on needs. Thresholds, based on spending and consumption concepts to value needs for food, clothing, shelter, and utilities (FCSU), are produced.

Previous research (Verbrugge, 2008a) demonstrated that housing rents and ex ante user costs diverge markedly for extended periods of time, a finding with profound implications for income and inflation measurement. But the primary data sources in that study were various indexes, based upon largely disjoint data sources, constructed using different aggregation techniques, and each subject to various criticisms. This raised doubts about the quality of the comparison. The relationship between user costs and rents might well be much tighter at the micro level; after all, house prices and rents (and their growth rates) can vary dramatically within cities, and rents are notoriously sticky. Furthermore, the use of indexes precludes both cross-sectional and dollar cost comparisons. In this study, we use Consumer Expenditure Interview Survey (CE) data to examine the relationship between user costs and rents at the individual unit level, in dollars, using unit-level information on house value, rent, taxes, and the like. This allows us to accurately estimate unit-specific user costs and to control for unobservables like structure and neighborhood quality. We also make the point that in theory, after-tax user costs should equal net rent, i.e., expected rental income, rather than gross rent. Our findings are striking. In keeping with most previous research, we find tremendous divergence between conventional measures of user costs and net rents, thus ruling out index construction errors as a possible explanation. This divergence does not result from a faulty rent measure: we find that reported rents are sensible, in that they move similarly to official rent indexes, and are not simply out-of-pocket expenses. Instead, and most perplexing, we find a surprisingly close correspondence between net rents and a particular estimate of user costs, one implicitly assuming zero transactions costs and constructed using an appreciation measure that is both theoretically suspect and empirically a poor predictor of actual appreciation.

The purpose of this research was to explore economic dimensions of a consumer gift-giving model. Two dimensions of extrahousehold gift expenditures were modeled: the probability of giving and the expected value of the corresponding expenditures. Data were from 4,139 households in the Quarterly Interview component of the 1984-85 U.S. Continuing Consumer Expenditure Survey. The results demonstrated that both the probability of giving and the value of annual expenditures for gifts given outside the consumer unit are related to total expenditures (a proxy for income), family size, life-cycle stage, and education. In addition, the probability of gift giving is related to the number of female adults, ethnicity, and urbanization, and the value of gift expenditures is related to region. Extrahousehold gift expenditures appear to be a luxury--as income increases, gift expenditures increase more rapidly. Copyright 1991 by the University of Chicago.

This paper presents an experimental poverty measure and compares it to the current official measure, now more than 40 years old. The experimental measure is based on an approach, drawn from work by a National Academy of Sciences (NAS) expert Panel, to consistently define basic needs and family resources. The experimental thresholds are based on out-of-pocket spending by families on basic goods and services and are based on an “outflows” concept. The resource measure is based on an “inflows” concept and reflects money coming into the household that is available to meet one’s basic needs. The U.S. Consumer Expenditure Survey serves as the basis for the experimental thresholds and the Current Population Survey Annual Social and Economic Supplement serves as the basis for the resource measure. Results for 1996 to 2005 are reported with trends examined. An important finding is that increases in expenditures for shelter and utilities, captured in the new thresholds, suggest a greater increase in the number of families not able to meet basic needs than is reflected by the official poverty statistics.

This paper constructs, for the five largest cities in the United States, user costs and rents for the same structure, in levels (i.e., measured in dollars). The levels formulation is a major advantage over indexes since one can answer questions like "Is it cheaper to rent or to own?" or "Are houses overvalued?" because such questions are essentially about the levels of rents and house prices and their fundamentals. These new measures are constructed using Consumer Expenditure Survey (CE) Interview data from 1982 to 2002, along with house price appreciation forecasts from Verbrugge (2007a). Characteristics, current market value, and rental equivalence of owner-occupied housing are used in a regression framework to predict the rent associated with a structure with median characteristics in each city. The property value of this median house is used to construct a user cost estimate for this structure. We find that, for the median structure in each city, estimated user costs and rents diverge to a surprising degree, in keeping with the previously noted findings of Verbrugge (2007a). It is not always cheaper to own: user costs sometimes lie well above rents. Finally, the dynamics of the estimated price-to-rent ratio are generally similar to those found in conventional estimates based upon indexes, suggesting that the present study might be useful for scaling or normalizing other estimates.

Subjective minimum income (MIQ) and minimum spending (MSQ) are the study focus. Basic Needs Module (1995) data from the U.S. Survey of Income and Program Participation are analyzed. A regression intersection approach is used to estimate household thresholds. MIQ thresholds are higher than MSQ thresholds. Both are higher than U.S. official poverty thresholds, and thresholds based on a National Academy of Sciences (NAS) methodology. Subjective threshold based equivalence scales imply greater economies of scale than those in the other two measures but are similar to behavioral scales. This finding suggests that families make trade-offs to meet their minimum needs.

This paper reviews a procedure that is being followed in the United States of America (USA) to experimentally test and evaluate recommendations made for redefining poverty measurement in that country. The recommendations were made in 1995 by the US National Academy of Sciences (NAS) Panel on poverty measurement. In this paper these recommendations are reviewed and the impact of implementing the recommendations on measures of inequality and poverty are examined. In conclusion, a discussion concerning possible lessons for India is provided. The recommended poverty measure (based on new measures of thresholds and resources) is examined in terms of its impact on inequality statistics, as well as poverty statistics, and results are compared to similar statistics based on the official measure. The standard Gini index, and three generalized entropy inequality measures are used to examine inequality. For the poverty analysis simple head count ratios, poverty gaps, and Foster-Greer-Thorbecke poverty measures are computed. Data from the 1991 U.S. Consumer Expenditure Survey (CE) Interview are used to produce the thresholds, and data from the 1992 through 1997 Current Population Survey (CPS), and in some analyzes, the 1991 panel of the Survey of Income and Program Participation (SIPP), are used to define resources. The proposed measure produces a distribution of resources that is, in general, more equal than is the distribution of official income. The poverty analysis reveals that changes in the poverty rates based on the official and the experimental measures are similar over time. However, poverty as measured by the NAS measure is greater than official poverty. The experimental poverty measure yields a poverty population that looks slightly more like the total U.S. population in terms of various demographic and socioeconomic characteristics than does the current official measure. Geographically adjusting the thresholds results in greater equality and lower poverty rates than when non-adjusted thresholds are used. With regard to India, poverty measurement is likely not to be based on income and expenditures primarily. Alternative measures based on other needs and resources are reviewed. However, regardless of the measure used, systematic evaluations of the measure are necessary and the USA model may be one to consider in this evaluation process.

Responses to minimum income and minimum spending questions are used to produce economic well-being thresholds. Thresholds are estimated using a regression framework. Regression coefficients are based on U.S. Survey of Income and Program Participation (SIPP) data and then applied to U.S. Consumer Expenditure Survey (CE) data. Three different resource measures are compared to the estimated thresholds. The first resource measure is total before-tax money income, and the other two are expenditure based. The first of these two refers to expenditure outlays and the second to outlays adjusted for the value of the service flow of owner-occupied housing (rental equivalence). The income comparison is based on SIPP data while the outlays comparisons are based on CE data. Results using official poverty thresholds are shown for comparison. This is among the earliest work in the U.S. in which expenditure outlays have been used for economic well-being determinations in combination with personal assessments, and the first time rental equivalence has been used in such an exercise. Comparisons of expenditures for various bundles of commodities are compared to the CE derived thresholds to provide insight concerning what might be considered minimum or basic. Results reveal that CE and SIPP MIQ thresholds are higher than MSQ thresholds, and resulting poverty rates are also higher with the MIQ. CE-based MSQ thresholds are not statistically different from average expenditure outlays for food, apparel, and shelter and utilities for primary residences. When reported rental equivalences for primary residences that are owner occupied are substituted for out-of-pocket shelter expenditures, single elderly are less likely to be as badly off as they would be with a strict outlays approach in defining resources.

The purpose of this paper is to review recent data made available through the Luxembourg Income Study (LIS) that include expenditures and asset valuations. The LIS data are augmented with comparable data from the U.S. Consumer Expenditure Survey. The surveys with expenditure data are reviewed in terms of collection units and variable definitions. Inequality statistics are produced and compared using income, expenditures, and market value of owned home. Rankings of countries by income and expenditure inequality are similar but not the same across the countries studied. Suggestions are made for the LIS to improve the expenditure data available following the COICOPS framework.

In many countries we observe a gap between macroeconomic and microeconomic statistics. In order to explain the underreporting observed in microeconomic statistics, the present paper tests the misreporting hypothesis through the double hurdle model. The misreporting hypothesis is based on some key assumptions: It is assumed that there are two categories of 'zero expenditure households' in the Survey - one category involves households that did not purchase consumer durables during the survey period and reported zero expenditure correctly on the Survey while the other category of 'zero expenditure households' involves households that, although they purchased consumer durables during the survey period, reported zero expenditure on the Survey. This is the source of misreporting. We also assume that there are 'positive expenditure households' for consumer durables in the Survey. These positive expenditure households for consumer durables are assumed, in the misreporting hypothesis, to have reported correctly their expenditure on consumer durables in the Survey. This model enables us to correct for over-reporting of zero expenditure households in the micro data. The data used for estimation involves thirteen clusters of consumer durables from the Consumer Expenditure Survey complied by the Bureau of Labor Statistics. The empirical results are satisfactory in supporting the conclusion that misreporting plays an important role in the underreporting in microeconomic statistics compiled by the Bureau of Labor Statistics compared with macroeconomic statistics compiled by the Bureau of Economic Analysis, the US Department of Commerce

The purpose of this paper is to review recent data made available through the Luxembourg Income Study (LIS) that include expenditures and asset valuations. The LIS data are augmented with comparable data from the U.S. Consumer Expenditure Survey. The surveys with expenditure data are reviewed in terms of collection units and variable definitions. Inequality statistics are produced and compared using income, expenditures, and market value of owned home. Rankings of countries by income and expenditure inequality are similar but not the same across the countries studied. Suggestions are made for the LIS to improve the expenditure data available following the COICOPS framework.

Equivalence sales are used to adjust income by family size to obtain income distribution measures. Recently, the concept of equivalence scale elasticity has been introduced to characterize the effect that scales have on distribution measures. We produce utility-based equivalence scales that have the property of constant elasticity. By assuming a particular functional form for the scales and that the scales are independent of the base level of utility, we obtain unique equivalence scales. In contrast to previous estimates of utility-based scales, we do not restrict our sample to particular family types. We determine price-dependant scales by estimating a characteristic-dependent almost ideal demand system using quarterly expenditure data from the US Consumer Expenditure Survey and price indices from the US Consumer Price Index. We use our scales and those implicit in the US official poverty thresholds to adjust expenditures and show that these scales have similar effects on inequality measures.

In recent years there has been a renewed interest among social science researchers and policymakers concerning the definition, measurement, and incidence of poverty. With government agencies producing statistics for quite some time concerning the incidence of poverty and the characteristics of poor individuals and households, there have also been parallel activities to improve the identification and measurement of poverty. This has been particularly true within the United States (U.S.), countries in Europe, Canada and Australia.

Household expenditure survey data can be used to provide information concerning the economic status and material well-being of families within and across countries. The purpose of this paper is two-fold: first, to describe the survey which is conducted in the United States (U.S.) to collect these data, and second, to present results from recent studies in which these data have been used to examine the distribution of household expenditures. We hope this paper will stimulate greater interest among researchers internationally to examine household expenditure data when addressing distributional issues. The paper is divided into four sections. Section one provides a description of the U.S. Consumer Expenditure Survey (CEX) and procedures. Section two includes results from analyses of the distribution of expenditures across households and the inequality in these expenditures. Section three discusses limitations of the CEX data. Section four is the summary and conclusions.

Although child poverty rates continue to surpass those of others, there is growing consensus that current official poverty measure has become outdated and flawed. Using data from the Current Population Survey and the Survey of Income and Program Participation, we implement an experimental poverty measure based on recommendations by a National Academy of Sciences panel. We find that while child poverty rates continue to surpass those of others, the gap between child and adult poverty rates is smaller under the experimental measure. Results highlight the impact of noncash government benefits and the Earned Income Tax Credit in reducing child poverty.

The cost-effectiveness of various end-stage renal disease (ESRD) treatments was compared using two different cost measures. The first measure, gross social costs, excluded output gains due to treatment, whereas the second measure, net social costs, included output gains from both market and nonmarket activities. The cost-effectiveness criterion was the cost-per-life year gained or the implicit value of a year of life. The lower the cost-per-life year gained, the more cost-effective the treatment was. Four ERSD treatments were evaluated over 20 years. Home dialysis and transplantation were more cost-effective than in-center dialysis, regardless of whether gross or net social costs were used. However, lower values were obtained in the case of net social costs reflecting a provision for output gains due to treatment. The use of net social costs also resulted in greater variations in costs-per-life year gained by age. Changes in survival probabilities affected the results for transplant patients and dialysis patients differently.

International audience; We consider the link between poverty and subjective well-being, and focus in particular on the role of time. We use panel data on 49,000 individuals living in Germany from 1992 to 2012 to uncover three empirical relationships. First, life satisfaction falls with both the incidence and intensity of contemporaneous poverty. Second, poverty scars: those who have been poor in the past report lower life satisfaction today, even when out of poverty. Last, the order of poverty spells matters: for a given number of years in poverty, satisfaction is lower when the years are linked together. As such, poverty persistence reduces well-being. These effects differ by population subgroups.