Aggregating across states

CPS
1

My coauthor and I seek to aggregate across states using the basic monthly data. So far we have tried calculating the labor force participation rate but we are not able to closely match the monthly unadjusted state-level data published by the BLS.

Because of the ease of using IPUMS published datasets relative to those published by NBER over time, we would like any assistance you may be able to provide.

For example we are collapsing the individual-level data to monthly aggregates and using ‘WTFINL’ to weight the observations. Is this the recommended approach to move form individual-level data to state-level data, or is there another approach or weight that should be used?

2

It sounds to me that you are using the correct sampling weight for your analysis. WTFINL is the final person-level weight that should be used in analyses of basic monthly data. When analyzing ASEC data, researchers should use the person weight ASECWT. For analyses including the outgoing rotation group variables, such as, EARNWEEK, HOURWAGE, PAIDHOUR, and UNION, researchers should use the EARNWT variable. If you continue to see a persistent discrepancy in your calculations, feel free to email ipums@umn.edu with a link to the statistics you are aiming to replicate and a detailed description of how you are constructing your estimates.

1 Like