Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Re: Weighting by State Population: Use iweight?


From   Gordon Hughes <G.A.Hughes@ed.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   st: Re: Weighting by State Population: Use iweight?
Date   Fri, 27 Jul 2012 10:56:37 +0100

Use aweights - i.e. [aw=state_pop]. If you were to use iweights, the implied sample size and the standard errors would depend upon the arbitrary scaling of state_pop.

In this context aweights are different from the weights used by the BLS, etc to construct state-level statistics. What aweights do is to give a greater weight to rates (crime, unemployment, etc) for states with large populations than to those with small populations.

Note one potential trap, which concerns the basis for normalising the data. For some purposes you can treat each state & year combination as a separate observation, so that AZ may have a larger weight in 2005 than in 1980. This is ok if you use OLS and then calculate standard errors adjusted for clustering by state. On the other hand, if you wish to use panel data procedures, these will normally insist that the weight is the same for all years for a particular state (panel unit). In that case, you have to use (for example) population in 2000 as the weight variable or the average state population over the data period, which can be calculated easily using -egen-.

Gordon Hughes
g.a.hughes@ed.ac.uk

=====================================

I am working on running a panel data regression that compares various
crime rates (dependent) to unemployment and income inequality
(independent). The panel consists of 50 states from 1976 to 2009. Most
of the previous literature when providing summary statistics and OLS
regression results simply state that the statistics and regressions
are "weighted by state population". I am very confused on how to
weight by state population. I do not think I need to use pweight or
aweight as the data is already aggregated by the US Census and Bureau
of Labor Statistics.Would I, therefore, use  the iweight command?
Could create a new variable that is the percent of the total US
population that a state posses and use this as the weight?

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index