Any idea how to combine income and smoking behavior for 2011 into a regression? -Surveys done in different months

I am attempting to do a regression showing the effects of income, price, and education on smoking behavior. The problem I am running into is that tobacco surveys are conducted in different months that income surveys are completed. For the year 2011, I am looking to combine the data, but am uncertain what the best path of action for this would be.

Since CPS respondents are in the sample for 4 months, leave for 8 months, and then re-enter for a final 4 months, approximately half of the January Tobacco Supplement respondents should also be present in the March ASEC. You can match respondents between these samples by using HRHHID and HRHHID2 to link households along with LINENO, SEX, and AGE to link respondents. Once linked, you can use the income variables available in the March ASEC.

For those respondents in the January sample that are in months 4 or 8 of the CPS, your best option is to refer to the Outgoing Rotation Group/Earner Study variable EARNWEEK.

Alternatively, there is a categorical family income variable (hufaminc) available in the basic monthly sample; however, it is not currently available via IPUMS-CPS. See this previous answer for information on linking your IPUMS-CPS dataset to the source NBER data. If you choose to use hufaminc, it will be available for the full Tobacco Supplement sample.

Hope this helps.