Creating variables from stored FGT poverty measures

[Zuhumnan Dapel]

Dear All,
I just constructed the following poverty measures

            
group(zon e1 year)  
        a=0           a=1       a=2
            
1    0.45051    0.15565    0.07424
2    0.57981    0.22978    0.12236
3    0.39341    0.15794    0.08509
4    0.79021    0.42030    0.26352
5    0.45631    0.17127    0.08639
6    0.55371    0.21982    0.11454
7    0.52296    0.19020    0.09092
8    0.55920    0.22276    0.11705
9    0.33838    0.12851    0.06735
10    0.61907    0.25688    0.13994
11    0.32306    0.10053    0.04533
12    0.57190    0.22638    0.11925
13    0.29774    0.09968    0.04957
14    0.51338    0.18552    0.09115
15    0.42791    0.16115    0.08363
16    0.67895    0.30197    0.17228
17    0.31564    0.11616    0.05968
18    0.50939    0.17982    0.08679
19    0.31267    0.13610    0.08576
20    0.58398    0.23640    0.12305
21    0.45392    0.18429    0.10000
22    0.65345    0.31652    0.19166
23    0.61066    0.28144    0.16747
24    0.67679    0.28936    0.15706
25    0.35864    0.13116    0.06538
26    0.57635    0.22272    0.11029
27    0.50073    0.20173    0.10729
28    0.77152    0.39423    0.24309
29    0.63412    0.25309    0.13067
30    0.77116    0.35862    0.20526
31    0.54763    0.21473    0.11026
32    0.65842    0.25118    0.12816
33    0.42415    0.14907    0.07161
34    0.77879    0.36587    0.21069
35    0.67673    0.27356    0.14352
36    0.75806    0.35490    0.20447

obtained from six regions over six-year period.
I wan to create three variables: p0,p1 and p2. First, is important I clarify the structure of the results. 0.45051 (italised) is the headcount index of poverty in region i, first region (where i= 1, 2, ...., 6) at date t, first year (where t=1,2, ..., 6); 0.52296 (bolded under group 7) is the index in first region and in the second year. Note, there are 1,378 individual units observations such that the 45% is the poverty rate in the region at the given time. The values for p0 that one is intending to create should be 0.45051 for all the 1,378 individual units from the region at the time.

Help is much appreciated.

Thank you,
Dapel

[Stephen Jenkins]

I suggest that you (1) sort this output by region and year, and then (2) list region year p0 p1 p2, sepby(region). The structure of your data would then be much clearer than above. If you have region and year variables in your individual-level data set, then you can use merge to attach these estimates shown above (saved to a data set) to each observation as appropriate using region and year as the match keys.

[Zuhumnan Dapel]

Thank you very much Prof Jenkins. The results are already as required in (1) of your post. Now I have performed (2), we have

. list zone year p0 p1 p2, sepby(zone)

+----------------------------------------+
zone   year       p0       p1       p2 
----------------------------------------
1.     1   1980   .00051   .00041   .00036 
2.     1   1985    .0006   .00051   .00045 
3.     1   1992   .00026    .0002   .00018 
4.     1   1996   .00018   .00019   .00022 
5.     1   2004   .00062   .00054   .00049 
6.     1   2010     .001   .00092   .00086 
----------------------------------------
7.     2   1980   .01111   .01036   .01002 
8.     2   1985   .01685   .01579   .01507 
9.     2   1992   .01625   .01382   .01233 
10.     2   1996     .017   .01619   .01531 
11.     2   2004   .02626   .02388   .02148 
12.     2   2010   .03952   .03547   .03288 
----------------------------------------
13.     3   1980   .01796   .01696    .0166 
14.     3   1985    .0138   .01233   .01174 
15.     3   1992    .0258   .02286   .02155 
16.     3   1996   .01941   .01854   .01828 
17.     3   2004     .028   .02654   .02564 
18.     3   2010   .01736   .01435   .01252 
----------------------------------------
19.     4   1980    .0612   .07659   .08723 
20.     4   1985   .02999   .02928   .02898 
21.     4   1992   .04813   .05036   .05219 
22.     4   1996   .02052   .02338   .02572 
23.     4   2004   .02708   .03255   .03646 
24.     4   2010    .0657   .07262   .07596 
----------------------------------------
25.     5   1980   .02446   .02161    .0198 
26.     5   1985   .01602   .01173    .0096 
27.     5   1992   .01771   .01534   .01431 
28.     5   1996   .04117   .04464   .04826 
29.     5   2004   .03977   .03734   .03502 
30.     5   2010   .05081   .04833   .04606 
----------------------------------------
31.     6   1980   .03905   .03647   .03452 
32.     6   1985   .03109   .02895    .0277 
33.     6   1992   .04749   .03944   .03458 
34.     6   1996   .04705   .04733   .04667 
35.     6   2004   .04926    .0539   .05604 
36.     6   2010   .09102   .10025   .10492 
+----------------------------------------+

.
Problem now solved with good results

  Result                           # of obs.
    -----------------------------------------
    not matched                             0
    matched                            97,646  (_merge==3)
    -----------------------------------------

.

Once again, thanks a million.
Dapel

[achintya widhi]

Dear all, i want to calculate p0, p1, and p2. My data is cross section year 2014, households and individuals survey data. my variabels are percapita expenditure, province, age, agesquare, education, gender, landownership. how can i calculate P0, P1, and P2? because those index will be used as outcomes variabel to running propensity score matching.