Bezon, I do not know why are you are frightened with running regression. It is very simple if you have data and correct question. Data about personal remittances you can find here: http://data.worldbank.org/indicator/BX.TRF.PWKR.CD.DT . There is also statistics about poverty level in different countries: http://data.worldbank.org/topic/poverty Food insecurity is here: http://www.fao.org/hunger/en/

You may study empirically how the growth in remittances influences reduction of poverty level. In some countries with high fraction of people working abroad the share of remittances in GDP is pretty high - about 30%. This should definetily make poverty lower.

Can you say a bit more about the structure of the data? Is remittance categorical or continuous? More importantly, why is regression not an option or desired here?

I would test the following given data. Suppose at time t in country A remittances formed x% of GDP and poverty (below $2 per day or other threshold - what you find) was y%. Then in t+1 remittances grow to z%, while poverty decline to w%. Form such data for several countries and run regression to see how 1% growth of remittances share in GDP influences poverty decline. Sure you need to extract data on GDP and other variables to form your data for regression, but this is available for many countries. Next you can control for other factors to make good empirical paper.

About your question: data should be numerical (neither categorical, nor continuous). In empirical study one always works with discrete set of data.

Dear Bezon Kumar,

Sorry for replying late. 

I suggest you to read a paper on the effect of remittances on poverty in Tajikista, without using any regression method.

Please have a look at:

https://www.researchgate.net/publication/235809009_The_impact_of_remittances_and_equivalence_scales_on_poverty_in_Tajikistan

Hope it can help.

Best regards,

Gianni

Article The impact of remittances and equivalence scales on poverty ...

Thank you a lot for your answer, Betti.