While working with ACS data I need some state controls like: unemployment rate, income per capita, racial and age composition, percentage of state population with positive welfare income - and for this I need the number of population in each states across years ( 2005-2015).
Could you kindly suggest how I should do it ? Wolfers & Stevenson (2006) used these controls , and I need that too for my research.
Thanks in advance !
The ACS sample design includes sample weights that enable researchers to estimate populations that are nationally representative. To obtain representative statistics of a population (estimate total population in relation to the statistic you are calculating, in your case income), you can apply sample weights to your analysis. Read more on sample weights in this blogpost and suggested procedures to incorporate sample weights into your analysis, here. For your analysis estimating per capita income by state, the appropriate person-level weight to use with INCTOT is PERWT (preselected in your extract). Please note that you will need to exclude NIU (not in universe) cases. Here is an example of how to incorporate sample weights into your analysis using Stata:
mean inctot [pw=perwt], over(statefip)
We provide exercises to help users practice working with IPUMS data in various statistical packages, including code examples, that you might find useful, specifically Exercise 1 and Exercise 2 on working with IPUMS USA data in Stata.