The analysis did not directly count individuals’ FGM/C status. Instead, it combined population and prevalence data to estimate how many people were likely to have undergone the practice and how many remained potentially at risk. Population estimates came from the 2015–2019 American Community Survey Public Use Microdata Sample. Unlike earlier US studies that primarily selected people by place of birth, the study selected its population primarily by ancestry.

Age-disaggregated prevalence estimates were derived from Multiple Indicator Cluster Surveys, Demographic and Health Surveys and other nationally representative surveys using methods developed in earlier work by Callaghan. The included survey samples were described as nationally representative at the 95% confidence level, with reported FGM/C error margins between 0.2% and 1.78%.

Earlier US estimates used relatively simple calculations. Jones and colleagues multiplied the relevant population by an average prevalence measure, while Goldberg and colleagues used age-based prevalence assumptions. European studies frequently applied an additional adjustment for migration and acculturation, commonly estimating that native-born girls had half the risk of foreign-born girls. Callaghan’s model aimed to represent variation that could be concealed by combining these elements into a single adjustment and by assuming a common risk boundary for all communities.

The new model incorporated a coefficient for socio-economic selective migration, drawing on differences between national prevalence and prevalence among more educated, wealthier or urban groups. It also introduced an Ethno-religious Loading Factor for Nigerian and Egyptian populations. This factor was used in sensitivity analyses to illustrate how ethnic and religious selection might affect estimates, rather than to claim an exact measured effect, and results were presented both with and without it.

Three approaches were used to estimate the proportion of FGM/C risk occurring before migration, after migration and in the future. The first used a universal risk boundary of age 19, consistent with an earlier framework. The second replaced that boundary with an age above which FGM/C was considered rare in each relevant community. The third fitted a generalised gamma distribution to age-of-cutting data, providing a continuous probability curve for estimating risk at individual ages. All fitted quartiles were within the 90% confidence level of the original quartiles, and 95.2% were within the 95% confidence level.

The third approach assumed that the age distribution of FGM/C was unimodal, although this may be less suitable for Liberia, where survey data suggested a bimodal distribution. The model also did not include the possible effects of long-term residence in refugee camps before migration, transnational travel or acculturation. The Ethno-religious Loading Factor values for Nigeria and Egypt were explicitly described by the author as speculative and intended to demonstrate a mechanism’s potential impact rather than predict its exact scale.

All three migration-timing approaches produced the same total estimate of 587,386 women and girls in the affected population. However, the more detailed methods classified more people as already living with FGM/C and fewer as remaining at risk. Relative to the age-19 approach, the generalised gamma model reclassified 85,512 women and girls from the at-risk category to the living-with category, representing a 66.8% reduction in the estimated number considered at risk.

Under the hypothesised deterrence scenario, represented by a Prevention Impact Factor of 0.25 and incorporating the hypothetical effect of ethnic and religious selective migration, the model estimated that the Jones method had overstated the overall impacted population by 12.9%, or 79,817 people, while the Goldberg method had overstated it by 6.3%, or 36,388. In the generalised gamma results, 94.1% of the estimated affected population, or 508,953 people, were classified as likely already living with FGM/C; 5.9%, or 31,826, were classified as potentially at risk. These figures describe a model-based population classification rather than a direct prevalence survey. The findings change the composition of the modelled population rather than its total size, and should not be interpreted as proof that earlier studies directly misclassified individual women’s status.

Subject of Research: Modelling migration-related FGM/C prevalence and risk among US immigrant and diaspora populations

Article Title: Modelling the impact of migration on FGM/C prevalence and risk in immigrant and diaspora populations

Article References: Callaghan, S. (2026). Modelling the impact of migration on FGM/C prevalence and risk in immigrant and diaspora populations. PLOS One, 21(10), e0360038. https://doi.org/10.1371/journal.pone.0360038

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0360038

Keywords: female genital mutilation, FGM/C, migration, diaspora populations, United States, selective migration, age of cutting, prevalence modelling, acculturation, American Community Survey, Modelling, impact

The stakes of this question are considerable. Disability insurance programs in the United States provide a financial lifeline to millions of workers whose medical conditions prevent them from maintaining substantial employment. An allowance decision can determine whether an individual has access to monthly cash benefits, health insurance coverage through Medicare after a qualifying waiting period, and a measure of economic security during what is often the most vulnerable period of their working life. Because these decisions carry such weight, any systematic difference in how applicants from different racial and ethnic groups fare in the process would raise serious concerns about fairness and equal treatment under the law. That is precisely why researchers have repeatedly examined allowance rates by race and ethnicity, and why unadjusted statistics showing gaps between groups have drawn attention from policymakers and analysts alike.

The methodological approach taken in the study is central to its contribution. Rather than relying on aggregate approval statistics alone, the investigators combined data from the American Community Survey, one of the largest and most detailed ongoing household surveys in the United States, with Social Security administrative records. This linkage allowed the researchers to observe not just whether an applicant was allowed or denied, but a rich set of characteristics describing who that applicant was before the determination was made: their marital status, their health conditions, their earnings history prior to disability, and a range of other demographic and socioeconomic attributes. This kind of linked administrative-survey infrastructure is relatively rare and highly valued in social science research, because it permits analysts to move beyond simple comparisons of group averages and toward a more complete accounting of the factors that actually drive outcomes.

The central analytical question is one that economists describe as a problem of composition versus process. Raw, unadjusted data showed differences in allowance rates across racial and ethnic groups. But a raw gap of that kind can arise in two very different ways. It could indicate that the determination process treats similarly situated applicants differently, which would suggest a disparity within the system. Alternatively, it could reflect the fact that applicants from different groups arrive at the process with different characteristics on average, such as different distributions of impairments, different work histories, different ages, or different household circumstances, and that those differences in applicant composition produce different outcomes even if the process itself is applied uniformly. Distinguishing between these two explanations requires statistical adjustment, and that is exactly what the study provides.

When the researchers accounted for socioeconomic, health, and demographic factors, the picture changed substantially. Most of the observed differences in approval rates became much smaller, and many were no longer statistically distinguishable from zero. In other words, once the researchers compared applicants who were similar in terms of the characteristics that matter for disability determinations, the residual differences associated with race and ethnicity themselves largely disappeared. The authors summarized their finding directly, writing that the differences in allowance rates by race and ethnicity in the raw, unadjusted data are mainly due to compositional differences between groups. This conclusion points away from the determination process as the primary source of the observed gaps and toward the underlying circumstances that shape who applies and with what profile of conditions and resources.

Among the factors the study identifies as associated with allowance outcomes are marital status, health conditions, and pre-disability earnings. Each of these has a plausible connection to how disability claims unfold. Health conditions are, of course, at the heart of any disability determination, since the severity and nature of impairments are what adjudicators assess. Pre-disability earnings matter because they relate to insured status and to the earnings requirements that govern eligibility for disability benefits, as well as to the kinds of work an applicant can realistically be expected to perform in the residual functional capacity framework used in evaluations. Marital status can influence household resources and the support structures available to applicants as they navigate a lengthy and often demanding application process. Differences across racial and ethnic groups in the distributions of these characteristics can therefore translate into differences in approval rates even in the absence of differential treatment within the process itself.

The findings carry important implications for how the raw statistics should be interpreted. A gap in unadjusted allowance rates is often read, in public discussion, as evidence of bias in the system. The study does not support that reading as the primary explanation. At the same time, the results should not be taken to mean that the underlying disparities are unimportant or that no policy concern remains. If applicants from different groups arrive with systematically different health conditions, earnings histories, and household circumstances, those compositional differences themselves reflect broader inequalities in American society, including differences in access to healthcare, in occupational structures, and in economic resources. The study’s contribution is to locate the source of the allowance-rate gaps in these upstream factors rather than in the determination process, which redirects attention toward the social and economic conditions that produce different applicant pools in the first place.

For policymakers, the research offers a more precise map of where intervention might be most effective. If the determination process were the main driver of the gaps, reforms would focus on adjudication standards, training, and oversight of decision-makers. Because the study finds that compositional differences dominate, the more relevant levers may lie in addressing the health and economic disparities that shape disability incidence and application behavior across communities. The findings also illustrate the value of linked administrative data for evaluating social insurance programs. Without the ability to connect survey-based measures of applicant circumstances to actual determination outcomes, analysts would be left with unadjusted statistics that invite misinterpretation. The study demonstrates how such data infrastructure can sharpen the evidence base on which fair and effective policy depends.

The study, published in Contemporary Economic Policy, a journal that has since 1982 provided a forum for scholarly analysis of important policy issues accessible to both economists and policymakers, adds a rigorous and carefully controlled data point to a debate that is often conducted with incomplete evidence. Its central message is methodological as much as substantive: raw differences between groups are not the same as disparities produced by a process, and distinguishing between the two requires data that capture the full circumstances of the people moving through the system. In the case of US disability insurance, the researchers find that when those circumstances are measured and held constant, race and ethnicity themselves largely fall out of the story, leaving marital status, health conditions, and pre-disability earnings as the factors that travel with applicants into the determination process and shape the outcomes they receive.

Subject of Research: Racial and ethnic differences in US disability insurance allowance rates and their socioeconomic and health-related explanations

Article Title: Why are there differences in US disability insurance allowance rates by race and ethnicity?

Article References: Why are there differences in US disability insurance allowance rates by race and ethnicity?. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: disability insurance, Social Security, allowance rates, race and ethnicity, American Community Survey, administrative data, health conditions, pre-disability earnings, marital status, Contemporary Economic Policy, social insurance policy, statistical adjustment

A new study by economist Taylor Mackay of California State University, Fullerton, published in the Atlantic Economic Journal, offers the first systematic look at what these laws do for recipients of non-voucher categories of protected income, including people receiving Supplemental Security Income and welfare payments. The findings are striking in their restraint: across every outcome the study could measure, the estimated effects of source of income policies are uniformly close to zero. In nationally representative survey data spanning fifteen years, the laws do not appear to generate large, detectable aggregate changes in the housing circumstances of this population.

The technical challenge the study confronts is a familiar one in policy evaluation. Source of income laws were not adopted randomly; they passed in particular cities, counties, and states at particular times, often in places with stronger progressive political movements or more acute affordability concerns. A naive comparison between places with and without such laws could therefore confound the effect of the policies with pre-existing differences between those places. To address this, Mackay employs an event-study estimator drawn from the recent difference-in-differences literature, specifically building on methods developed by Clément de Chaisemartin and Xavier D’Haultfœuille, which are designed to handle policy adoptions that occur at different times across jurisdictions and to guard against biased comparisons when treatment effects vary over time or across groups.

The data underlying the analysis come from the American Community Survey, the Census Bureau’s large ongoing household survey, accessed through the IPUMS USA project. Mackay aggregates microdata from 2006 through 2021 to the level of Consistent Public Use Microdata Areas by year, creating geographic units whose boundaries remain stable even as official Census definitions change. This consistency matters: source of income laws are frequently adopted at the county or municipal level, and the ability to match those policies to stable geographic areas over time allows the study to capture far more policy variation than analyses relying on coarser geographic identifiers would permit.

Within those geographic units, the study identifies likely-affected renters — households reporting income from Supplemental Security Income or welfare programs, the non-voucher income categories most clearly protected by source of income statutes. The analysis then estimates intent-to-treat effects on four housing outcomes: residential mobility, gross rent, household size, and crowding. The intent-to-treat framing is deliberate. Because the surveys cannot observe which individual renters actually attempted to move or apply for housing after a law passed, the estimates capture the average effect of living in a jurisdiction covered by such a policy, an appropriately conservative measure of the laws’ real-world reach.

The results are notable for how tightly they cluster around zero. For residential mobility, the estimates are precise enough to rule out increases larger than 2 to 3 percentage points. Given a baseline annual mobility rate of roughly 22 percent among these renters, that means the study can exclude relative increases above about 10 to 13 percent. In other words, if source of income laws do help non-voucher protected-income recipients find or change housing, the aggregate effect is small enough to escape detection even in a study with substantial statistical power. Estimates for gross rent, household size, and crowding tell the same story: no large movements on any of the housing margins observable in the survey data.

Why might the laws matter so little for this group when prior research suggests they do help voucher holders? The study’s framing points to several plausible mechanisms rooted in how the rental market treats different income sources. Housing Choice Vouchers are a large, visible, federally administered subsidy that landlords can easily recognize and may distrust, so a legal mandate directly targets a well-documented form of discrimination. Field experiments and correspondence studies have long documented that landlords reject voucher holders at high rates, and earlier work by Lance Freeman and colleagues found that source of income protections increased voucher utilization and expanded access to less disadvantaged neighborhoods. Supplemental Security Income and welfare income, by contrast, may be less visible to landlords during the application process, or renters with such income may face binding constraints — affordability, credit history, screening criteria — that anti-discrimination law alone cannot overcome.

The study is careful about the limits of what its data can show. The American Community Survey does not record where a household lived before a move at the neighborhood level, so the analysis cannot test whether source of income policies changed the quality of destinations among renters who did move — the locational question that has motivated much of the voucher literature, including work by Ingrid Gould Ellen and colleagues on whether protections help recipients reach lower-poverty neighborhoods. Nor can the survey capture subtler outcomes such as application denials, search durations, or landlord behavior during leasing. It is possible that the laws improve the housing search experience for non-voucher recipients in ways that never register in annual mobility rates or rent payments. The estimates also concern the protected-income categories most clearly covered by the statutes; the study treats Social Security recipients separately in supplemental analyses, noting that households living mainly on retirement or other benefit income likely differ in composition and in their exposure to source-of-income discrimination.

For policymakers, the findings carry a sobering implication. Source of income laws are often promoted as a broad shield for all renters whose income comes from outside the labor market, and more than two decades of legislative activity reflects that ambition. The new evidence suggests that whatever benefits such laws deliver, they are unlikely to show up as large aggregate improvements in mobility, rent burdens, or crowding for SSI and welfare recipients. That does not necessarily mean the laws fail these renters — targeted benefits to a subset of applicants, reduced discrimination in principle, or improved search experiences could all be real while leaving population averages essentially unchanged. But it does mean that the strongest empirical case for the policies remains the voucher-focused literature, and that advocates hoping to demonstrate measurable gains for other protected groups will need better data, perhaps from field experiments or administrative records, to make it.

Methodologically, the study also demonstrates how far the policy-evaluation toolkit has come. By combining stable geographic panels built from Census microdata with modern event-study estimators that account for staggered policy adoption, Mackay is able to extract credible causal estimates from the messy patchwork of state and local lawmaking. The approach of measuring intent-to-treat effects on a nationally representative sample, and then reporting precisely how small the detectable effects are, offers a model for evaluating other anti-discrimination statutes whose benefits may be diffuse, targeted, or invisible in standard survey outcomes. As source of income laws continue to spread — and as housing affordability climbs the political agenda — this study sets an important baseline: for renters relying on Supplemental Security Income and welfare payments, the observable housing margins barely move, and any real effects of these laws must be sought somewhere the Census cannot yet see.

Subject of Research: The effects of source of income anti-discrimination laws on renters receiving non-voucher protected income such as Supplemental Security Income and welfare payments

Article Title: The Effect of Source of Income Policies on Recipients of Non-Voucher Categories of Protected Income

Article References: Mackay, T. (2026). The Effect of Source of Income Policies on Recipients of Non-Voucher Categories of Protected Income. Atlantic Economic Journal. https://doi.org/10.1007/s11293-026-09859-1

Image Credits: AI Generated

DOI: 10.1007/s11293-026-09859-1

Keywords: source of income laws, housing discrimination, rental housing, Supplemental Security Income, welfare payments, Housing Choice Vouchers, residential mobility, fair housing policy, American Community Survey, event-study estimator, housing economics, public assistance

The reassessment of neighborhood disadvantage comes at a moment when composite indices have quietly migrated from academic journals into the machinery of American health policy. What began as a descriptive tool for sociologists studying concentrated poverty now shapes decisions about resource allocation, payment adjustments, and clinical risk stratification. This elevation gives new urgency to questions that might once have seemed purely technical: which variables belong in an index, how those variables should be combined, and whether the resulting scores actually measure the construct researchers intend them to measure. When an index influences whether a community receives additional medical resources or how a hospital is reimbursed, methodological ambiguity ceases to be an academic inconvenience and becomes a matter of distributive justice.

One of the most consequential issues raised by this work is the conceptual entanglement of disadvantage and affluence. The theoretical literature on neighborhoods has long treated these as related but distinct phenomena. Disadvantage, in the classic formulation, refers to the co-occurrence of economic hardship, family instability, and housing precarity that erodes collective efficacy and isolates residents from mainstream opportunity structures. Affluence, by contrast, refers to the presence of highly educated, high-income residents whose resources sustain local institutions, schools, and civic organizations. Empirical studies have repeatedly found that affluence predicts health outcomes more strongly than the absence of disadvantage does, suggesting that the two ends of the socioeconomic continuum operate through different mechanisms. Yet widely used indices such as the Area Deprivation Index and the Child Opportunity Index fold indicators of advantage, such as white-collar employment or college attainment, directly into their scoring. This practice makes it impossible to determine whether an observed association with health reflects the harms of deprivation or the protections of affluence, a distinction with very different policy implications.

The sheer proliferation of indices compounds the problem. With more than thirty publicly available measures of neighborhood disadvantage in circulation, researchers face a bewildering choice, and there is little consensus about which instrument is appropriate for which purpose. The scoping review summarized in the source material found that while all fifteen national indices examined included poverty, other variables appeared far less consistently. Educational attainment appeared in twelve, unemployment in eleven, housing characteristics in twelve, public assistance receipt in only four, and single-parent family structure in eight. Variables such as race and ethnicity raise particularly thorny questions. The spatial concentration of Black residents in certain neighborhoods is a legacy of redlining, restrictive covenants, and exclusionary zoning rather than a manifestation of socioeconomic deprivation itself. Including such variables risks conflating the consequences of structural racism with poverty, while excluding them may obscure the very processes that produce concentrated disadvantage. There is no methodologically neutral answer, only choices that must be justified in relation to the research question at hand.

Comparability failures between indices are not hypothetical. The Area Deprivation Index and the Social Vulnerability Index, two of the most frequently used measures, correlate only modestly, with Spearman coefficients in the range of roughly 0.49 to 0.57 depending on geographic scale. More striking, only about 35 percent of census tracts that the ADI places in the most disadvantaged decile receive the same classification from the SVI. These disagreements propagate into health research. Studies of Medicare beneficiaries undergoing coronary artery bypass surgery found measurable differences in thirty-day readmission rates depending on which index classified a tract as most disadvantaged. Analyses of primary care patients showed that odds ratios for diabetes, hypertension, chronic kidney disease, and mortality varied systematically between the ADI and the Social Deprivation Index. In other words, the choice of index is itself an analytical decision that can alter substantive conclusions, yet it is rarely reported or justified with the same care as other modeling choices.

The temporal dimension of index construction deserves particular scrutiny. Most of the composite measures in widespread use today descend from indices developed in the late 1990s and early 2000s, drawing on census variables selected to characterize the American metropolis of that era. The intervening decades have transformed the socioeconomic landscape in ways the original architects could not have anticipated. The rise of the gig economy and precarious work complicates simple unemployment measures. Housing costs have escalated dramatically in many metropolitan areas, decoupling homeownership from economic security in some markets while remaining a meaningful asset elsewhere. Educational attainment has risen overall, shifting the meaning of thresholds such as lacking a high school diploma. Household composition has changed, with growth in multigenerational living arrangements and delayed family formation. An index calibrated to the census of 2000 may misclassify contemporary neighborhoods, and because indices are often updated by swapping in newer census data without revisiting the underlying variable selection, the construct being measured may have drifted silently over time.

Geographic scale introduces another layer of complexity that researchers and policymakers frequently underestimate. Indices differ in whether they are computed at the block group, census tract, or ZIP code tabulation area level, and these units are not interchangeable. Census tracts, the most common choice, typically contain a few thousand residents and were designed to be relatively homogeneous, but they can still mask sharp internal variation, particularly in gentrifying areas where new development sits alongside long-standing low-income housing. ZIP code tabulation areas, used by the Community Need Index and the Distressed Communities Index, are larger and more heterogeneous, diluting localized pockets of deprivation. Block groups, used exclusively by the ADI, offer finer resolution but suffer from a serious data problem: because of their small populations, the Census Bureau suppresses many block group estimates to protect confidentiality, particularly for income-related items. More than half of block groups had at least one missing item in the construction of the 2022 ADI, requiring imputation that introduces its own assumptions and uncertainty. This is the modifiable areal unit problem in miniature: the same underlying population can appear more or less disadvantaged depending entirely on the boundaries drawn around it.

The statistical machinery of index construction also varies in ways that are rarely transparent to end users. Factor analysis and principal components analysis are the standard data reduction techniques, but decisions about variable standardization, ranking, weighting, and the number of components retained differ across indices and are often poorly documented. The resulting scores may be expressed as continuous values, national percentiles, state-level ranks, or quintile classifications, each of which carries different implications for statistical power and interpretation. A rank-based measure, for instance, discards information about the magnitude of differences between neighborhoods and is sensitive to the distribution of the underlying population. Whether an index is standardized nationally or within states changes which neighborhoods appear extreme, a nontrivial concern for studies spanning multiple regions or for federal programs that allocate resources across state lines.

These methodological concerns intersect with a broader movement toward transparency and reproducibility in population health research. When indices disagree, the disagreement is not merely noise; it reflects genuine uncertainty about the structure of the underlying construct. A rigorous response would involve empirically re-evaluating that structure with contemporary data, testing whether the variables that loaded together two decades ago still form coherent dimensions, and whether affluence and disadvantage emerge as separable factors. Such an approach treats measurement as a hypothesis to be tested rather than a convention to be inherited. It also opens the possibility of developing indices tailored to specific outcomes, since the dimensions of neighborhood context most relevant to cardiovascular disease may differ from those most relevant to child development or mental health.

For clinicians and health systems, the stakes are increasingly concrete. Indices of neighborhood disadvantage are being incorporated into risk adjustment formulas, screening protocols, and value-based payment models. A primary care practice might flag patients from highly disadvantaged areas for enhanced outreach, while payers adjust reimbursement based on the socioeconomic profile of the populations served. Each of these applications inherits the measurement problems described above. If an index conflates affluence with the absence of disadvantage, a hospital serving an affluent area might appear to serve a disadvantaged one, or vice versa. If indices disagree about which neighborhoods are worst off, payment adjustments and resource flows will follow the idiosyncrasies of whichever instrument was adopted. The modest correlations documented between leading indices suggest that these are not edge cases but pervasive features of the current measurement landscape.

The path forward suggested by this reassessment is not to abandon composite measurement, which has demonstrated value in capturing multidimensional context that no single indicator can, but to rebuild it on firmer conceptual and empirical foundations. That means returning to the theoretical distinction between the scarcity of resources and the abundance of them, testing whether contemporary data support that two-dimensional structure, and being explicit about the purpose each index is meant to serve. It means documenting variable selection decisions, reporting sensitivity analyses across alternative indices and geographic scales, and acknowledging the uncertainty introduced by missing data and imputation. Above all, it means recognizing that neighborhood disadvantage is not a fixed quantity waiting to be read off a census table but a construct whose meaning evolves with the society it describes. Measurement, like the neighborhoods it seeks to characterize, requires periodic re-examination.

Subject of Research: The measurement structure of neighborhood disadvantage and affluence in the contemporary United States

Article Title: Are We Measuring Neighborhood Disadvantage Wrong? A Methodological Reassessment of its Structure in the United States

Article References: Clarke, P., Rollings, K., Melendez, R., Sinkewicz, M., Duchowny, K., Gypin, L., & Noppert, G. (2026). Are We Measuring Neighborhood Disadvantage Wrong? A Methodological Reassessment of its Structure in the United States. SSM – Population Health, Article 101966. https://doi.org/10.1016/j.ssmph.2026.101966

Image Credits: AI Generated

DOI: 10.1016/j.ssmph.2026.101966

Keywords: neighborhood disadvantage, Area Deprivation Index, factor analysis, neighborhood affluence, census tracts, health disparities, social determinants of health, American Community Survey, confirmatory factor analysis, area-based indices, health policy, measurement validity

Ernesto F. L. Amaral and Nereyda Y. Ortiz Osejo de Lemus of Texas A&M University analyzed American Community Survey data spanning 2005 to 2024, covering adults aged eighteen and older across the entire country. Their central question was deceptively simple: are the places where migration rates run high the same for everyone, and do the groups with the strongest regional concentrations also have the highest individual likelihood of moving? The answer, they found, is frequently no—a discrepancy with significant implications for how researchers and policymakers understand population redistribution in America.

The backdrop to the study is what demographers have called the Great American Migration Slowdown. Between 1968 and 2009, annual interstate migration rates fell from 3.6 percent to just 1.6 percent, and between 1982 and 2015, moves across state and county lines dropped by roughly half. The new analysis confirms this trajectory continued into recent years: overall migration rates bottomed out at 4.57 percent in 2012, recovered modestly through 2019, dipped during the COVID-19 pandemic, and settled back down to 4.58 percent by 2024. The decline is pronounced compared with other developed countries and is thought to reflect reduced responsiveness to regional economic differences, growing place attachment, and shifts in job mobility.

But the aggregate numbers conceal enormous variation. Hispanic, non-Hispanic African American, and non-Hispanic Native American populations consistently showed lower migration rates than non-Hispanic whites and non-Hispanic Asians throughout the two decades studied, with foreign-born Hispanics recording the lowest rates of any group. Foreign-born individuals of all backgrounds moved less often than their U.S.-born counterparts, reflecting differences in settlement histories, social networks, and access to information about destinations. When the researchers modeled individual migration odds statistically, controlling for age, education, income, marital status, children, homeownership, employment, and region of residence, these group differences persisted rather than disappearing.

The spatial half of the analysis used a technique called Local Indicators of Spatial Association, or LISA, which identifies statistically significant clusters of high or low values across neighboring geographic areas. Applied to Migration Public Use Microdata Areas—the finest geography available in the public census microdata—the method revealed striking regional signatures. In 2019, hotspots of high internal migration appeared across parts of the West, including northern California, Nevada, Utah, southern Idaho, and northwestern Wyoming, along with central Texas and northern Florida. By 2024, most of the western hotspots had faded, while Texas and Florida concentrations endured. A long corridor of low migration, stretching from Texas through the Midwest to the Northeast, remained stable across the entire period and even expanded into California and New Mexico.

Those patterns, however, look very different depending on which demographic group is mapped. Non-Hispanic African Americans showed pronounced high-migration clusters across southeastern regions, alongside widespread low-migration zones in the Midwest, Northeast, and West. Hispanic clustering concentrated in California, the Mountain West, Texas, and Florida. Non-Hispanic Native Americans displayed high-migration clusters in the Northwest, Southwest, and parts of the Great Plains, including Arizona, Utah, New Mexico, Colorado, Oklahoma, and Alaska—areas where Indigenous populations are geographically concentrated. Non-Hispanic Asians, by contrast, were characterized mainly by low-migration clusters in southern and Midwestern regions, with some expansion into the Northwest by 2024.

Nativity proved to be a crucial modifier of these spatial patterns. U.S.-born Hispanics showed broad high-migration clustering across western and southern areas, while foreign-born Hispanics displayed far fewer hotspots, concentrated primarily in Texas and Florida by 2024. The contrast was equally stark for African Americans: U.S.-born Black migrants exhibited a broad southeastern pattern, whereas foreign-born Black migrants showed tightly localized clusters in eastern metropolitan areas, including Florida, Atlanta, and Washington, DC—patterns likely reflecting distinct immigrant origins, settlement networks, and labor market incorporation. Foreign-born whites, meanwhile, were marked predominantly by low-migration clusters across central regions of the country.

Perhaps the study’s most striking finding is that regional concentration and individual mobility are genuinely distinct dimensions of migration. Non-Hispanic African Americans and Hispanics both showed lower individual odds of moving than non-Hispanic whites after adjustment, yet both groups displayed strong, persistent regional hotspots. Conversely, non-Hispanic Asians showed relatively high individual mobility despite maps dominated by low-migration clusters. A group can be regionally concentrated precisely because most of its members who do move go to the same places, or because past settlement histories anchor its geography, without its members being frequent movers overall. The authors are careful to note that their LISA maps describe spatial clustering at selected points in time and cannot disentangle the cumulative flows, network effects, and historical settlement processes that may have produced them.

The research also has limitations that the authors acknowledge candidly. The analysis is associational rather than causal, and the logistic regression models do not formally account for spatial autocorrelation, which could lead to underestimated standard errors. The aggregated geographic units limit the ability to study very local dynamics, and detailed subgroup-specific flow analysis at the county level would require restricted Census data. The 2020 ACS, disrupted by pandemic data-collection problems, was tested in robustness checks and found not to alter the main results. Still, the authors argue the two-pronged approach—descriptive spatial clustering paired with adjusted individual-level models—offers a descriptive framework that previous single-lens studies could not provide.

What emerges is a compelling guide for future research. The findings point directly to places and populations where localized studies could uncover the mechanisms behind the patterns: Hispanic migration in Texas and Florida, African American migration in the Southeast, foreign-born African American settlement in Atlanta and Washington, DC, and Native American migration across the Southwest, Great Plains, and Alaska. The authors suggest combining administrative county-to-county flow data with survey microdata, and applying spatially explicit methods such as geographically weighted regression, to test whether the relationships they identify vary across places. As the United States continues to grow more demographically diverse while its residents grow more rooted, understanding precisely where and how different communities redistribute themselves—and why geography and mobility do not always move in tandem—has become an essential question for demographers, economists, and anyone concerned with regional inequality.

Subject of Research: Spatial patterns and demographic heterogeneity of internal migration in the United States by nativity and race/ethnicity

Article Title: Internal migration, demographic heterogeneity, and spatial patterns in the United States

Article References: Amaral, E. F. L., & Ortiz Osejo de Lemus, N. Y. (2026). Internal migration, demographic heterogeneity, and spatial patterns in the United States. Journal of Population Research, 43(3), Article 27. https://doi.org/10.1007/s12546-026-09434-0

Image Credits: AI Generated

DOI: 10.1007/s12546-026-09434-0

Keywords: internal migration, demographic heterogeneity, spatial clustering, LISA, nativity, race and ethnicity, American Community Survey, population redistribution, spatial demography, United States, logistic regression, migration slowdown