zero wages


I am trying to study the annual wage, hourly wage and weekly wage. I realize there are too many zero wages even after dropping those who are unemployed, self-employed, in the IPUMS data, how should I handle those wages?

Another thing to consider, that I don’t see you meantioning, is to exclude people who are not in the labor force (see LABFORCE). These people are techniclly in the universe for wage questions, but will have zero wages.

Ultimately, it is really up to you how you handle these observations. The “correct” method likely depends on your research question and method. Personally, without additional information, I’d suggest you perform your analysis both with these observations in the sample and again without. Then you’ll be able to see if your results are meaningfully dependent on including these people or not.

Hi Jeff,

Thanks for answering my questions. I need to aggregate wage to the commuting zone level. Another question related is do I need to weight the wage using the perwt for conducting a commuting zone level analysis?


Yes, using the person-level samping weight variable will correct for the complex sampling design used by the Census Bureau.