Table 2.
Backward myopic cost share | Forward myopic cost share | |||||
---|---|---|---|---|---|---|
|
||||||
Model: log(1/12 + Y) | Coefficient | Elasticity | S.E | Coefficient | Elasticity | S.E |
All Spending | −1.26 | −0.44 | 0.044 | −1.26 | −0.41 | 0.041 |
Outpatient | −0.86 | −0.29 | 0.034 | −0.91 | −0.28 | 0.033 |
Inpatient | −2.18 | −0.30 | 0.045 | −17.37 | −2.35 | 0.357 |
Pharmaceutical | −1.00 | −0.44 | 0.040 | −1.12 | −0.51 | 0.045 |
Maternity | −0.27 | −0.09 | 0.042 | −0.39 | −0.12 | 0.060 |
MH/SA | −0.62 | −0.26 | 0.046 | −0.93 | −0.39 | 0.069 |
ER | −0.12 | −0.04 | 0.024 | −0.51 | −0.17 | 0.102 |
Room and Board | −1.44 | −0.20 | 0.069 | −8.18 | −1.14 | 0.389 |
Non specialty visits | −0.73 | −0.25 | 0.041 | −0.90 | −0.28 | 0.046 |
Home visits | −0.06 | −0.01 | 0.071 | −0.25 | −0.05 | 0.326 |
Prevention | −0.34 | −0.02 | 0.009 | −4.86 | −0.29 | 0.123 |
Surgical procedures | 1.33 | 0.07 | 0.098 | 28.85 | 1.52 | 2.139 |
Surgical supplies/devices | −3.13 | −0.22 | 0.090 | −80.88 | −5.76 | 2.318 |
Specialty visits | −0.80 | −0.32 | 0.033 | −1.10 | −0.42 | 0.043 |
Dialysis | 0.17 | 0.02 | 0.156 | 0.31 | 0.04 | 0.290 |
PT, OT, speech therapy | −0.33 | −0.15 | 0.032 | −0.60 | −0.27 | 0.057 |
Chiropractic | −0.36 | −0.23 | 0.065 | −0.47 | −0.30 | 0.083 |
CAT scans | −0.49 | −0.15 | 0.023 | −2.64 | −0.80 | 0.123 |
Mammograms | −1.02 | −0.11 | 0.027 | −19.92 | −1.97 | 0.499 |
MRIs | −0.95 | −0.29 | 0.022 | −7.30 | −2.21 | 0.171 |
PET scans | −1.13 | −0.17 | 0.081 | −7.35 | −1.13 | 0.528 |
Radiology - diagnostic | −0.42 | −0.15 | 0.018 | −1.09 | −0.38 | 0.044 |
Ultrasounds | −0.38 | −0.14 | 0.015 | −1.54 | −0.56 | 0.060 |
Diagnostic services | −0.53 | −0.15 | 0.021 | −1.54 | −0.42 | 0.059 |
DME | −0.39 | −0.18 | 0.038 | −1.00 | −0.47 | 0.097 |
Ambulances | −0.31 | −0.07 | 0.052 | −4.75 | −1.00 | 0.806 |
Note: N = 170,963,286 individual months. Each row is a different regression using the log(spending plus 1/12) as the dependent variable. Each model includes individual*year fixed effects, 12 monthly time dummies and uses employer’s average cost share for that year*plan*family*month, without the own individual, as an instrument. Cluster corrected standard errors use the employer*plan for clusters. Elasticities are evaluated at the mean cost share for that particular type of service, and S.E obtained using the delta method.