Skip to main content
. Author manuscript; available in PMC: 2016 Nov 1.
Published in final edited form as: J Econ Behav Organ. 2015 Nov 1;119:72–83. doi: 10.1016/j.jebo.2015.07.009

Table A.6.

Additional regression results.

County fixed effects, ages 65-plus Condition categories,
ages 66-plus
County fixed effects,
ages 65 and 67



HMO PFFS HMO PFFS HMO PFFS
Age 66–69 −1.364*** −0.939***
0.023 0.0193
Age 70–74 −1.688*** −1.033*** −0.243*** −0.0415***
0.039 0.0241 0.0218 0.0133
Age 75–79 −1.934*** −1.132*** −0.432*** −0.0991***
0.046 0.0313 0.0258 0.0168
Age 80–84 −2.150*** −1.261*** −0.569*** −0.209***
0.0503 0.0414 0.0342 0.031
Age 85-plus −2.462*** −1.469*** −0.840*** −0.394***
0.0442 0.0412 0.0232 0.027
Age 67 −1.395*** −0.941***
0.0225 0.0111
Plan count 0.135*** 0.0787** 0.254*** 0.106*** 0.155*** 0.0602***
0.0495 0.0309 0.0514 0.02 0.022 0.0158
Plan count above 16 −0.146*** −0.0586* −0.208*** −0.113*** −0.157*** −0.0498***
0.0494 0.0307 0.0518 0.0205 0.0221 0.0154
Plan count above 45 0.00205 0.0219*** −0.0307*** 0.0234*** 0.0011 0.00771***
0.00426 0.00609 0.00436 0.00531 0.00224 0.00242
Premium 0.000608 −0.00975*** 0.00518*** −0.00638*** −0.000596 −0.0105***
0.00153 0.00136 0.00167 0.00125 0.00111 0.0009
Non-premium OOPC 0.00496* −0.00557** 0.00652** −0.00648** 0.00292 −0.0112***
0.00284 0.00283 0.00303 0.00289 0.0024 0.00182
Year 2008 0.207* −0.713*** −0.882*** −0.287*** 0.192*** −0.413***
0.109 0.0954 0.0636 0.0763 0.0741 0.0511
Year 2009 0.311*** −0.889*** −0.699*** −0.626*** 0.400*** −0.142***
0.0905 0.078 0.0558 0.0584 0.0607 0.0437
Year 2010 0.540*** −1.249*** 0.107* −1.392*** 0.691*** −0.456***
0.0686 0.0703 0.0553 0.0528 0.048 0.0355
Black 0.268*** 0.604*** 0.219*** 0.577*** 0.0217 0.426***
0.0445 0.0324 0.046 0.0352 0.0445 0.0251
Male −0.0821*** −0.0736*** 0.0441*** 0.00613 −0.235*** −0.205***
0.00969 0.0109 0.00826 0.00891 0.00722 0.00576
Constant −3.676*** −4.122*** −7.740*** −4.836*** −4.009*** −4.149***
0.737 0.459 0.756 0.301 0.322 0.235
Observations 2,461,442 9,932,390 7,155,814

Note: This table presents results from three multinomial logit regressions. All three regressions model the choice between traditional Medicare (TM, the omitted category), Medicare Advantage (MA) HMO/PPO plans, and MA private fee-for-service (PFFS) plans. The first regression was run on a 5% sample of beneficiaries aged 65 and above, and included county-level fixed effects. The second model was run on a 20% sample beneficiaries aged 66 and above, and included condition category indicator variables. The third model was run on beneficiaries aged 65 and 67, and included county-level fixed effects. The coefficient estimates are suppressed for the county fixed effects and condition category terms. Standard errors are clustered at the county level.

*

p < 0.1.

**

p < 0.05.

***

p < 0.01.