Exhibit 3.
Probability of waiting > 20 min | |||||
---|---|---|---|---|---|
No Controls | Controlsa + Census Region Fixed Effects | Controls + Diagnosisb + Census Region Fixed Effects | Controls + Practice Fixed Effects | Controls + Medical Practioner Fixed Effects | |
(1) | (2) | (3) | (4) | (5) | |
Medicaid | 0.0170 | 0.0338 | 0.0318 | 0.0153 | 0.0090 |
(0.0068) | (0.0071) | (0.0075) | (0.0071) | (0.0017) | |
R2 | 0.0003 | 0.0406 | 0.0679 | 0.1242 | 0.1620 |
SOURCE Author’s analysis of data from athenahealth. NOTES Table is based on the baseline sample, restricted to patients with either Medicaid or private insurance (N=15.8 million outpatient visits). The mean of the dependent variable for privately insured is 0.1626.
Controls include six bins for patient age (0-4 yrs, 5-14 yrs, 15-24 yrs, 25-44 yrs, 45-64 yrs, 65+ years), patient gender, five time of day bins (before 10am, 10am-12pm, 12pm-2pm, 2pm-4pm, 4pm and after), seven days of the week, four arrival binary variables (arrived at least twenty minutes early, arrived at least ten minutes early, arrived on time, and arrived less than ten minutes late). We also control for twelve physician specialty groups and total work relative value units associated with the visit which reflects the procedures preformed.
Controls for diagnosis consist of indicator variables for each observed combination of (up to eight) ICD-9 diagnosis codes for the visit.