----------------------------------------------------------------------------------- log: /netscr/khrapov/stata/rec10/rec10.log log type: text opened on: 3 Apr 2008, 08:15:08 . use spending . su Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- obs | 1302 651.5 375.9993 1 1302 airtravel | 1302 189.0568 502.8492 0 5133 carloans | 1302 1340.324 4745.544 0 45412 cars | 1302 1.006912 .8910425 0 6 highered | 1302 203.5361 1146.541 0 20490 -------------+-------------------------------------------------------- income | 1302 41798 40071.62 12 432834.4 members | 1302 2.648233 1.577827 1 12 midwest | 1302 .2142857 .4104836 0 1 northeast | 1302 .1612903 .3679398 0 1 south | 1302 .3218126 .4673511 0 1 -------------+-------------------------------------------------------- urban | 1302 .9132104 .2816346 0 1 west | 1302 .3026114 .4595649 0 1 cardum | 1302 .109831 .3127992 0 1 income_sq | 1302 3.35e+09 9.13e+09 144 1.87e+11 . ta carloans CARLOANS | Freq. Percent Cum. ------------+----------------------------------- 0 | 1,159 89.02 89.02 293 | 1 0.08 89.09 491 | 1 0.08 89.17 500 | 1 0.08 89.25 528 | 1 0.08 89.32 942 | 1 0.08 89.40 1090 | 1 0.08 89.48 1200 | 1 0.08 89.55 1527 | 1 0.08 89.63 1700 | 2 0.15 89.78 1739 | 1 0.08 89.86 2000 | 1 0.08 89.94 2086 | 1 0.08 90.02 2279 | 1 0.08 90.09 2295 | 1 0.08 90.17 2524 | 1 0.08 90.25 2786 | 1 0.08 90.32 2884 | 1 0.08 90.40 3000 | 1 0.08 90.48 3200 | 1 0.08 90.55 3250 | 1 0.08 90.63 3500 | 1 0.08 90.71 3519 | 1 0.08 90.78 3544 | 1 0.08 90.86 3700 | 1 0.08 90.94 3800 | 1 0.08 91.01 3843 | 1 0.08 91.09 4428 | 1 0.08 91.17 --Break-- r(1); . gen carloans_dum=carloans>0 . ta carloans_dum carloans_du | m | Freq. Percent Cum. ------------+----------------------------------- 0 | 1,159 89.02 89.02 1 | 143 10.98 100.00 ------------+----------------------------------- Total | 1,302 100.00 . kdensity carloans . kdensity carloans if carloans>0 . tobit carloans income income_sq urban members highered, ll(0) Tobit regression Number of obs = 1302 LR chi2(5) = 37.91 Prob > chi2 = 0.0000 Log likelihood = -1913.9852 Pseudo R2 = 0.0098 ------------------------------------------------------------------------------ carloans | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- income | .1994888 .0606466 3.29 0.001 .0805126 .3184649 income_sq | -6.89e-07 3.12e-07 -2.21 0.028 -1.30e-06 -7.62e-08 urban | -3915.967 3495.656 -1.12 0.263 -10773.73 2941.792 members | 1694.701 659.0258 2.57 0.010 401.8278 2987.574 highered | 2.109867 .6424819 3.28 0.001 .8494495 3.370285 _cons | -35981.97 4919.823 -7.31 0.000 -45633.65 -26330.29 -------------+---------------------------------------------------------------- /sigma | 22747.85 1653.26 19504.49 25991.2 ------------------------------------------------------------------------------ Obs. summary: 1159 left-censored observations at carloans<=0 143 uncensored observations 0 right-censored observations . mfx Marginal effects after tobit y = Fitted values (predict) = -28611.076 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- income | .1994888 .06065 3.29 0.001 .080624 .318354 41798 income~q | -6.89e-07 .00000 -2.21 0.027 -1.3e-06 -7.7e-08 3.4e+09 urban*| -3915.967 3495.7 -1.12 0.263 -10767.3 2935.39 .91321 members | 1694.701 659.03 2.57 0.010 403.034 2986.37 2.64823 highered | 2.109867 .64248 3.28 0.001 .850626 3.36911 203.536 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(pr(0,.)) Marginal effects after tobit y = Pr(carloans>0) (predict, pr(0,.)) = .1042413 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- income | 1.59e-06 .00000 3.39 0.001 6.7e-07 2.5e-06 41798 income~q | -5.48e-12 .00000 -2.25 0.025 -1.0e-11 -7.0e-13 3.4e+09 urban*| -.0339837 .03284 -1.03 0.301 -.098341 .030374 .91321 members | .0134757 .00515 2.62 0.009 .003388 .023564 2.64823 highered | .0000168 .00001 3.27 0.001 6.7e-06 .000027 203.536 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(e(0,.)) Marginal effects after tobit y = E(carloans|carloans>0) (predict, e(0,.)) = 10861.758 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- income | .0342027 .01024 3.34 0.001 .014135 .05427 41798 income~q | -1.18e-07 .00000 -2.22 0.026 -2.2e-07 -1.4e-08 3.4e+09 urban*| -699.7423 649.89 -1.08 0.282 -1973.51 574.026 .91321 members | 290.5592 112.07 2.59 0.010 70.9059 510.213 2.64823 highered | .3617401 .10969 3.30 0.001 .146754 .576726 203.536 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(ystar(.,.)) Marginal effects after tobit y = E(carloans) (predict, ystar(.,.)) = -28611.076 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- income | .1994888 .06065 3.29 0.001 .080624 .318354 41798 income~q | -6.89e-07 .00000 -2.21 0.027 -1.3e-06 -7.7e-08 3.4e+09 urban*| -3915.967 3495.7 -1.12 0.263 -10767.3 2935.39 .91321 members | 1694.701 659.03 2.57 0.010 403.034 2986.37 2.64823 highered | 2.109867 .64248 3.28 0.001 .850626 3.36911 203.536 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(xb) Marginal effects after tobit y = Linear prediction (predict, xb) = -28611.076 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- income | .1994888 .06065 3.29 0.001 .080624 .318354 41798 income~q | -6.89e-07 .00000 -2.21 0.027 -1.3e-06 -7.7e-08 3.4e+09 urban*| -3915.967 3495.7 -1.12 0.263 -10767.3 2935.39 .91321 members | 1694.701 659.03 2.57 0.010 403.034 2986.37 2.64823 highered | 2.109867 .64248 3.28 0.001 .850626 3.36911 203.536 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . clear . exit