I want to use an IV strategy to estimate the effect of an electrification policy on female labour supply. I have data on the policy that indicates how many new households were connected to the grid in each year in each municipality for a total span of time of ten years (2000 to 2010).
I also have census data for two years (2000 and 2010) that contain information on whether a woman is employed or not.
My dependent variable is the change in the rate of employed women in each municipality.
My first question is: What is the correct way to code my dependent variable so that I can account for the change in female labour supply rates between 2000 and 2010?
My second question is: how should I code my electrification variable so that it takes into account the intensity of electrification each municipality experienced. (I cannot make the treatment variable a dummy, since almost all municipalities have received connections through this policy and limiting the treatment variable to a dummy reduces my sample size too much > the control group then would be too small).
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input long zipcode byte LPT2003 int(LPT2004 LPT2005 LPT2006 LPT2007 LPT2008 LPT2009 LPT2010 ano) float(LPT_yr LPTchange fem_employment10 fem_employment00) 3156700 0 0 44 427 1 1 12 172 2000 5 .022265904 0 0 2918001 0 0 107 433 445 271 509 325 2000 5 .05561173 0 0 2414605 0 1 49 63 116 80 41 1 2000 6 .13305534 0 0 3204807 0 28 85 51 24 106 12 16 2010 6 .1024825 0 0 2933257 0 0 0 0 1 2 82 14 2010 3 .05316864 0 0 2927408 0 0 0 0 0 0 1 0 2000 1 1.514603e-06 0 0 4211504 0 3 3 3 2 0 3 3 2000 6 .00604337 0 0 4128658 0 0 0 38 0 0 32 17 2000 4 .08430233 0 0 5300108 0 0 0 0 0 0 0 0 2000 . 0 0 0 2304236 0 22 96 28 21 44 21 72 2000 6 .0810883 0 0 end
Thanks in advance for help on where I could read up on these issues.
0 Response to How to code my independent and dependent variable in an IV strategy with an intensity-of-treatment effect
Post a Comment