Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Michigan.do. 2. * construct new variables;. gen mi=state==26;. * michigan dummy;. gen hike=month>=33;. * treatment period dummy;. gen treatment=hike*mi;

Similar presentations


Presentation on theme: "1 Michigan.do. 2. * construct new variables;. gen mi=state==26;. * michigan dummy;. gen hike=month>=33;. * treatment period dummy;. gen treatment=hike*mi;"— Presentation transcript:

1 1 Michigan.do

2 2. * construct new variables;. gen mi=state==26;. * michigan dummy;. gen hike=month>=33;. * treatment period dummy;. gen treatment=hike*mi;

3 3. * get means of smoking rates for the 2x2 box;. sort mi hike;. by mi hike: sum smoked;

4 4 ------------------------------------------------------------------------------- -> mi = 0, hike = 0 Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- smoked | 26441.1950002.3962083 0 1 ------------------------------------------------------------------------------- -> mi = 0, hike = 1 Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- smoked | 18852.1827923.3865064 0 1 ------------------------------------------------------------------------------- -> mi = 1, hike = 0 Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- smoked | 17790.1957279.3967712 0 1 ------------------------------------------------------------------------------- -> mi = 1, hike = 1 Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- smoked | 12943.1783203.382797 0 1

5 5 Smoking Rates Time Period BeforeAfterDiff. States MI0.19572790.1783203-.01741 All others 0.19500020.1827923-.01221 -.0051997

6 6. * now run the regression;. reg smoked mi hike treatment; Source | SS df MS Number of obs = 76026 -------------+------------------------------ F( 3, 76022) = 8.59 Model | 3.95288903 3 1.31762968 Prob > F = 0.0000 Residual | 11663.5888 76022.153423862 R-squared = 0.0003 -------------+------------------------------ Adj R-squared = 0.0003 Total | 11667.5417 76025.153469803 Root MSE =.39169 ------------------------------------------------------------------------------ smoked | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- mi |.0007277.0037982 0.19 0.848 -.0067168.0081723 hike | -.0122079.0037337 -3.27 0.001 -.019526 -.0048898 treatment | -.0051997.0058668 -0.89 0.375 -.0166985.0062991 _cons |.1950002.0024088 80.95 0.000.1902789.1997215 ------------------------------------------------------------------------------ Notice the estimate for treatment is arithmetically Identical to the number from the 2x2 box

7 7. * now control for observed characteristics;. * generate dummy variables to control for race, parity, education and age;. xi i.age i.mrace3 i.meduc6 i.ageg;. * add these variables to the model;. reg smoked mi hike treatment _I*;

8 8 Source | SS df MS Number of obs = 76026 -------------+------------------------------ F( 15, 76010) = 474.06 Model | 998.136229 15 66.5424153 Prob > F = 0.0000 Residual | 10669.4055 76010.140368445 R-squared = 0.0855 -------------+------------------------------ Adj R-squared = 0.0854 Total | 11667.5417 76025.153469803 Root MSE =.37466 ------------------------------------------------------------------------------ smoked | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- mi | -.0039841.0036501 -1.09 0.275 -.0111382.00317 hike | -.0030636.0035745 -0.86 0.391 -.0100696.0039424 treatment | -.0051552.0056122 -0.92 0.358 -.0161551.0058447 _Iageg_2 |.105199.0052283 20.12 0.000.0949516.1154464 Delete some results _Imeduc6_6 |.0156193.0145937 1.07 0.284 -.0129844.044223 _cons |.1476714.0092008 16.05 0.000.1296378.165705 ------------------------------------------------------------------------------

9 9. * now control for all time periods and all states;. * need to define time and state dummies;. * in the regression, drop the mi and hike variables;. xi i.age i.mrace3 i.meduc6 i.ageg i.state i.month;. reg smoked treatment _I*;

10 10. reg smoked treatment _I*; Source | SS df MS Number of obs = 76026 -------------+------------------------------ F( 70, 75955) = 102.48 Model | 1006.80931 70 14.3829901 Prob > F = 0.0000 Residual | 10660.7324 75955.140355901 R-squared = 0.0863 -------------+------------------------------ Adj R-squared = 0.0854 Total | 11667.5417 76025.153469803 Root MSE =.37464 ------------------------------------------------------------------------------ smoked | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- treatment | -.0053194.0056134 -0.95 0.343 -.0163216.0056828 _Iageg_2 |.1051191.0052311 20.10 0.000.0948663.115372 Delete some results _Istate_26 | -.0046835.0051651 -0.91 0.365 -.0148071.0054402 _Istate_42 | -.0009752.0045039 -0.22 0.829 -.0098028.0078525 _Imonth_2 |.0273311.0136851 2.00 0.046.0005083.0541538 Delete some results _Imonth_3 |.0305209.014018 2.18 0.029.0030457.0579961 _Imonth_56 |.009216.0148571 0.62 0.535 -.0199038.0383358 _cons |.1292655.0138104 9.36 0.000.1021972.1563339 ------------------------------------------------------------------------------


Download ppt "1 Michigan.do. 2. * construct new variables;. gen mi=state==26;. * michigan dummy;. gen hike=month>=33;. * treatment period dummy;. gen treatment=hike*mi;"

Similar presentations


Ads by Google