I am currently looking at INCWAGE (Row) SEX (Column) with OCC2010 (Control). I had a few questions regarding the data returned in IPUMS USA. Does INCWAGE account for part-time or full-time status? How reliable are the mean wages? I noticed that the standard deviation was quite high. Thank you!


INCWAGE does not account for full-time or part-time status. It is simply the reported amount of money received as an employee for the previous year. If you’d like to control for the amount of time spent working, you can do so using the UHRSWORK variable, which identifies the number of hours per week that the respondent usually worked. Also note that INCWAGE has values, such as 999999, that represent “N/A” or “Missing” observations. For valid calcuations, you’ll need to filter these values out.


Thank you for the prompt response. We will try to use UHRSWORK.

We filtered out the missing variables but still noticed a high standard deviation. Would this still be the most reliable way to obtain this data? Thank you.


It sounds like you are using the online data analysis system. This tool is useful for quick analysis of Census data. For more involved projects, I’d suggest extracting the microdata through the online data extract system.


I reviewed the data table from the online analyzer and did not find any missing values. Instead, I found 0 listed for a few occupations. Would it be best to remove these 0s? Would reviewing INCWAGE (Row) SEX (Column) with OCC2010 (Control) be the most reliable way to obtain mean salary data by sex by occupation? Any recommendations or suggestions are greatly appreciated. Thank you!


Technically speaking, there is no “bottom code” on the INCWAGE values. You could remove observations with values of zero, and then interpret your results as the mean wage among income earners.