Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nick Cox <njcoxstata@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: lorenz ordinates in stata |
Date | Mon, 5 Nov 2012 18:38:27 +0000 |
Neglecting (ignoring) within-county variation is what you seem to be proposing to do, so I don't know why you want me to explain it. You are neglecting or ignoring it necessarily because, on the information you give us, you don't have data to do otherwise. If all you have is one value -- presumably a mean, but you haven't explained "per capita", which just means "per person", and doesn't imply mean, median or anything else necessarily -- for each county in each year, then you can proceed only by assuming that the distribution of county means (or whatever) approximates the distribution of income across persons. I imagine that you are an economist, and I know I am not one, but that seems to me to be likely to be a dubious assumption, to put it mildly. I am making the same very simple point: do you really have the data to do what you want? Because if not, which programs could do it is immaterial. I am asking a question, not giving advice on what you should do. Nick On Mon, Nov 5, 2012 at 6:26 PM, Ajita Atreya <ajita63@hotmail.com> wrote: > Nick, > > Could you please illustrate more on what kind of distribution can I get by neglecting within-county variation and how should I proceed. > > Thanks, > Ajita > ---------------------------------------- >> Date: Mon, 5 Nov 2012 18:21:40 +0000 >> Subject: Re: st: lorenz ordinates in stata >> From: njcoxstata@gmail.com >> To: statalist@hsphsun2.harvard.edu >> >> Thanks for the further details. On this information, you can only >> proceed to a distribution by neglecting within-county variation. >> >> Nick >> >> On Mon, Nov 5, 2012 at 6:11 PM, Ajita Atreya <ajita63@hotmail.com> wrote: >> > Hi Nick, >> > >> > Thank you for your reply. >> > >> > I might have misrepresented my data here. >> > >> > I have data on per capita flood premium paid and percapita income for each county (144 counties) and year (1978-2010). I am trying to determine progressivity by examining the departure from proportionality using lorenz and concentration curve. I believe there must be some program in stata for this kind of analysis. >> > >> > Thank you for your help >> > Ajita >> > >> > ---------------------------------------- >> >> Date: Mon, 5 Nov 2012 18:00:46 +0000 >> >> Subject: Re: st: lorenz ordinates in stata >> >> From: njcoxstata@gmail.com >> >> To: statalist@hsphsun2.harvard.edu >> >> >> >> It seems that you have one observation for each county and year, at a >> >> wild guess some sort of county mean. Where is the information on >> >> variability that would enable you to construct a Lorenz curve? At most >> >> you can pool the counties to get a distribution for each year, but >> >> then you still have no information on variability within county. >> >> >> >> Typing -findit- with appropriate keywords will reveal several >> >> user-written programs, but they seem beyond what you can use. >> >> >> >> Nick >> >> >> >> On Mon, Nov 5, 2012 at 4:56 PM, Ajita Atreya <ajita63@hotmail.com> wrote: >> >> > Dear statalist, >> >> > >> >> > Can anyone help me on the best way to get lorenz ordinates for county income in deciles. I have the data in the following form >> >> > >> >> > county Year Per capitaIncome >> >> > A 1978 1000 >> >> > A 1979 2000 >> >> > B 1978 500 >> >> > B 1979 600 >> >> > B 1980 700 >> >> > c >> >> > . * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/