Table II.
Heterogeneous Effects, Utilization
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | |
|---|---|---|---|---|---|---|---|---|
|
|
||||||||
| Dependent Variable: Number Outpatient Visits | ||||||||
|
|
||||||||
| By Income Level | By Educational Status | By Prevalence Black Doctors | By Marital Status | |||||
|
|
||||||||
|
income > black male median |
income <= black male median |
educ > black male median |
educ <= black male median |
black MD > median |
black MD <= median |
married | unmarried | |
| Pj*postt*blackr*maleg | −0.546 (0.548) | −1.725** (0.705) | −0.061 (0.409) | −2.801*** (0.839) | −1.359*** (0.373) | −2.052 (1.460) | −1.398*** (0.326) | −1.665 (1.061) |
| Pj*postt*maleg | 0.060 (0.084) | −0.049 (0.184) | −0.013 (0.092) | −0.130 (0.294) | 0.042 (0.082) | −0.110 (0.238) | 0.029 (0.106) | −0.188 (0.240) |
| Pj*postt*blackr | −0.150 (0.221) | −0.211* (0.108) | −0.465** (0.215) | 0.468 (0.617) | −0.019 (0.115) | −0.566 (0.726) | 0.419** (0.204) | −0.511** (0.191) |
| Fixed Effects | State-Year, Race-Gender-Year, Race-Gender-State | |||||||
| Observations | 143,554 | 77,400 | 178,756 | 42,198 | 176,032 | 44,922 | 160,335 | 60,619 |
| No. Clusters | 49 | 49 | 49 | 49 | 25 | 24 | 49 | 49 |
| Adj R-squared | 0.013 | 0.030 | 0.014 | 0.036 | 0.016 | 0.017 | 0.014 | 0.030 |
Notes: OLS estimates of equation (2) assessing heterogeneous effects by income level, education level, black doctor prevalence, and marital status. Specifically, in the first two columns we divide the sample by median black male income. In the following two columns we divide the sample by median black male education. In the next two columns, we calculate black physicians as a percentage of all physicians using occupational data from the 1970 Census and bifurcate states as above or below median on this dimension. In the last two columns, we divide the sample by whether the survey respondent was married. Utilization data are from the harmonized version of the National Health Interview Survey (NHIS) available from IPUMS and merged with restricted identifiers for use in the Restricted Data Center (RDC) and cover the period 1969–1977. Post is an indicator variable equal to 1 in the years following 1972. The unit of observation is the individual, and the sample includes non-veteran black and white men and women ages 45–74. The outcome variable across all panels and columns is the number of physician interactions in the last 12 months. In addition to the listed fixed effects, individual-level controls in every specification for utilization include indicator variables for educational status, income, age, marital status, telephone ownership, and rural/urban status. Regressions are weighted using provided survey weights. Standard errors are clustered at the state level.
p<0.01,
p<0.05 and
p<0.10, respectively.