Skip to main content
. Author manuscript; available in PMC: 2014 Mar 1.
Published in final edited form as: J Ment Health Policy Econ. 2013 Mar;16(1):3–12.

Table 3.

Quantile regression coefficients from model of mental health care expenditure testing differential response by race, insurance, education, income, and language

95th percentile 97.5 percentile 99th percentile
Black-White Difference
 Insurance Private −2.896 ** −1.603*** −1.304***
Public −1.009 ***,††† 0.112††† −0.112†††
Uninsured −0.823 **,††† −1.798***,††† −1.752***,††
 Education < HS Grad −2.896 ** −1.603*** −1.304***
HS Grad −2.757 ***,††† −1.007***,††† −0.531†††
Any College −3.618 *** −1.544***,†† −0.948**,††
College Grad −2.833 ***,††† −0.843**,††† −0.459†††
 Income <FPL −2.896 ** −1.603*** −1.304***
100–125% FPL −3.023 ***,††† −0.769††† −0.879†††
125–200% FPL −3.199 *** −1.880***,††† −0.782
200–400% FPL −2.779 *** −1.922*** −1.439***
400%+ FPL −3.367 *** −1.706***,†† −1.305***
Latino-White Difference
 Insurance Private −1.888 *** −1.352** −0.819**
Public −2.100 ***,††† −1.569***,††† −1.482***,††
Uninsured −0.6361 ** −2.632***,††† −1.648***,††
 Education < HS Grad −1.888 *** −1.352** −0.819**
HS Grad −1.547 **,††† −0.489††† 0.078†††
Any College −2.421 *** −1.026***,††† −0.181†††
College Grad −1.724 ***,††† −0.715 −0.094†††
 Income <FPL −1.888 *** −1.352*** −0.819**
100–125% FPL −2.225 ***,††† −1.393***,††† −1.141**,†††
125–200% FPL −2.054 *** −1.399*** −0.973**
200–400% FPL −1.733 *** −1.394*** −1.193***
400%+ FPL −2.307 *** −1.580*** −0.694††

Source: 2004–2008 MEPS (n=83,878)

On log scale - proportional effects (no risk ratios)

In addition to race/ethnicity indicator variables, quantile regression models adjust for age, gender, health and mental health status variables (self-reported mental health and health status, mental and physical components of the SF-12 scale, K6 scale of severe psychological distress, PHQ-2 depression scale, list of chronic conditions, BMI), socioeconomic status variables (poverty status, education, and marital status), and interactions between race/ethnicity and education and income.

Indicates referent group used in the expenditure model. Coefficients presented for these groups are estimated using linear combinations.

**,***

Significant disparity at α=0.05 and α=0.01 levels

††, †††

Significantly different from prior row’s group (same quintile) at α=0.05 and α=0.01 levels