Growth elasticity and inequality effect of poverty reduction

[Zuhumnan Dapel]

Dear All,
Kakwani (1993) derived growth elasticity of poverty and inequality effect of poverty reduction. Recent version of this can be found on page 8 of the document http://www.ipc-undp.org/pub/IPCWorkingPaper1.pdf

To me Kakwani's approach is a bit different from Datt and Ravallion's (1992) "Decomposition of Change in Poverty by Growth and Redistribution Components" (which I have no difficulty running using Stata) and Elasticity of FGT Indices to Per capita Consumption Expenditure using (I also have not difficulty computing using ADePT). If I'm right that there is some difference, how do I compute Kakwani's using Stata, dealing with the impact of growth (and the mitigating role of inequality) on poverty?


Thank you very much,
Dapel

[Michal Brzezinski]

I think that these methods are implemented in the Stata DASP package:
http://dasp.ecn.ulaval.ca/

I hope this helps,
Michal

[Zuhumnan Dapel]

Dear Michal, Thank you very much for the link. However, there seems to be no version for Mac OS.

[Nick Cox]

As at http://www.stata.com/statalist/archi.../msg00637.html, DASP is hidden behind a registration wall: the implication is that you should seek support from the authors.

[Zuhumnan Dapel]

Thank you Dr Cox. I'm now getting along, successful installed and now using it

[Zuhumnan Dapel]

I have been thinking that the poverty and growth and poverty elasticities are computed between two periods t-1 and t. But the

efgtineq

command seems to be doing so for one period. Any further help please?

[Nick Cox]

Where does that come from? Stata FAQ #12 applies as always. If it's from DASP, see #4 again.

[Zuhumnan Dapel]

One is trying to measure the index 'pro-poor' growth between two periods t1 and t2 using Kakwani's(1993) methodology. In the formula he derived, growth elasticity of poverty and growth elasticity of inequality are parts of. Which of the periods, (t1 or t2) should one used in computing the elasticities?

Thanks,
Dapel

[Pablo Bonilla]

For the Datt-Ravallion decomposition, you may check the user-written command drdecomp .

Best,