Abstract
There has been considerable interest in using decomposition methods in epidemiology (mediation analysis) and economics (Oaxaca-Blinder decomposition) to understand how health disparities arise and how they might change upon intervention. It has not been clear when estimates from the Oaxaca-Blinder decomposition can be interpreted causally because its implementation does not explicitly address potential confounding of target variables. While mediation analysis does explicitly adjust for confounders of target variables, it typically does so in a way that effectively entails equalizing confounders across racial groups, which may not reflect the intended intervention. Revisiting prior analyses in the National Longitudinal Survey of Youth on disparities in wages, unemployment, incarceration, and overall health with test scores, taken as a proxy for educational attainment, as a target intervention, we propose and demonstrate a novel decomposition that controls for confounders of test scores (e.g., measures of childhood socioeconomic status [SES]) while leaving their association with race intact. We compare this decomposition with others that use standardization (to equalize childhood SES [the confounders] alone), mediation analysis (to equalize test scores within levels of childhood SES), and one that equalizes both childhood SES and test scores. We also show how these decompositions, including our novel proposals, are equivalent to implementations of the Oaxaca-Blinder decomposition but provide a more formal causal interpretation for these decompositions.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 825-835 |
| Number of pages | 11 |
| Journal | Epidemiology (Cambridge, Mass.) |
| Volume | 29 |
| Issue number | 6 |
| DOIs | |
| State | Published - Nov 1 2018 |
Fingerprint
ASJC Scopus subject areas
- Epidemiology
Cite this
Decomposition Analysis to Identify Intervention Targets for Reducing Disparities. / Jackson, John; VanderWeele, Tyler J.
In: Epidemiology (Cambridge, Mass.), Vol. 29, No. 6, 01.11.2018, p. 825-835.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Decomposition Analysis to Identify Intervention Targets for Reducing Disparities
AU - Jackson, John
AU - VanderWeele, Tyler J.
PY - 2018/11/1
Y1 - 2018/11/1
N2 - There has been considerable interest in using decomposition methods in epidemiology (mediation analysis) and economics (Oaxaca-Blinder decomposition) to understand how health disparities arise and how they might change upon intervention. It has not been clear when estimates from the Oaxaca-Blinder decomposition can be interpreted causally because its implementation does not explicitly address potential confounding of target variables. While mediation analysis does explicitly adjust for confounders of target variables, it typically does so in a way that effectively entails equalizing confounders across racial groups, which may not reflect the intended intervention. Revisiting prior analyses in the National Longitudinal Survey of Youth on disparities in wages, unemployment, incarceration, and overall health with test scores, taken as a proxy for educational attainment, as a target intervention, we propose and demonstrate a novel decomposition that controls for confounders of test scores (e.g., measures of childhood socioeconomic status [SES]) while leaving their association with race intact. We compare this decomposition with others that use standardization (to equalize childhood SES [the confounders] alone), mediation analysis (to equalize test scores within levels of childhood SES), and one that equalizes both childhood SES and test scores. We also show how these decompositions, including our novel proposals, are equivalent to implementations of the Oaxaca-Blinder decomposition but provide a more formal causal interpretation for these decompositions.
AB - There has been considerable interest in using decomposition methods in epidemiology (mediation analysis) and economics (Oaxaca-Blinder decomposition) to understand how health disparities arise and how they might change upon intervention. It has not been clear when estimates from the Oaxaca-Blinder decomposition can be interpreted causally because its implementation does not explicitly address potential confounding of target variables. While mediation analysis does explicitly adjust for confounders of target variables, it typically does so in a way that effectively entails equalizing confounders across racial groups, which may not reflect the intended intervention. Revisiting prior analyses in the National Longitudinal Survey of Youth on disparities in wages, unemployment, incarceration, and overall health with test scores, taken as a proxy for educational attainment, as a target intervention, we propose and demonstrate a novel decomposition that controls for confounders of test scores (e.g., measures of childhood socioeconomic status [SES]) while leaving their association with race intact. We compare this decomposition with others that use standardization (to equalize childhood SES [the confounders] alone), mediation analysis (to equalize test scores within levels of childhood SES), and one that equalizes both childhood SES and test scores. We also show how these decompositions, including our novel proposals, are equivalent to implementations of the Oaxaca-Blinder decomposition but provide a more formal causal interpretation for these decompositions.
UR - http://www.scopus.com/inward/record.url?scp=85056286316&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85056286316&partnerID=8YFLogxK
U2 - 10.1097/EDE.0000000000000901
DO - 10.1097/EDE.0000000000000901
M3 - Article
C2 - 30063540
AN - SCOPUS:85056286316
VL - 29
SP - 825
EP - 835
JO - Epidemiology
JF - Epidemiology
SN - 1044-3983
IS - 6
ER -