I’d like to exploit the panel dimension of the monthly files to calculate monthly wage growth for individuals. My naive quick attempt in stata was:
gen modate = ym(year, month)
format modate %tm
xtset cpsidp modate
gen wgegrwth=hourwage/L.hourwage-1
However, I don’t get any values generated for this (all data missing). No doubt this is too naive but any thoughts on best practice welcome.
The reason you are finding all missing data for this calculation is that the HOURWAGE variable comes from the Outgoing Rotation Group/Earner Study questions. These are a special group of questions that are asked to households who are in their 4th or 8th month in the CPS sample. Therefore when you lag by a month, there are no corresponding households. In general, it may be quite difficulty to calculate the change in wages over two consecutive months using CPS data. This is because most of the wage or income data come from either these Outgoing Rotation Group/Earner Study questions or the ASEC supplement which is only asked once per year.
I don’t think you can do this with CPS data. Hourly wages are only available for individuals who are working at the time of the survey and who are in their fourth or eighth interview (Month in sample 4 or 8).
Hope this helps
thanks both. In that case, is it possible to calculate the 4 month wage growth for the individuals in sample 8 since I should have their wages when they were in sample 4? If possible, any suggestions about how best to achieve this would be greatly appreciated.
*sorry that should read 12 month wage increase.
I don’t imediately see a reason why what you are describing wouldn’t work. If you are calculating a 12 month increase in wages or income, however, it may prove much better to use the ASEC supplement. The sample will be larger and the ASEC includes many more variables pertaining to income or wages than the Outgoing Rotation Group/Earner Study questions.
ah great so if a mis4 individual appears in the ASEC in year x they will appear in year x+1 ASEC with same unique CPSIDP number (now in mis8)?
Actually, IPUMS CPS has been working hard over the previous few years to develop tools for linking together ASEC samples to allow users to exploit the panel nature of the data. See this working paper for more information.
Many thanks Jeff - looks like incredibly useful work. Given ASEC has wage information (e.g. INCWAGE) - it seems I could calculate 12 year wage growth just using this, without linking to the March basic files. I guess I could do this using MARBASECID to link individuals between ASECs in consecutive years but this is only possible going back to 89. This is already useful, but do you think there is any reliable way of linking individuals across consecutive ASECs pre 89?
You can try to replicate the pre-1989 process yourself or wait until the IPUMS CPS Team releases the pre-1989 version of MARBASECID. There is lots of demand for this variable so we are all hoping it will be available soon.
ah that sounds great. the million dollar question: any idea when the revised data will be released, and how far back it will go?
Unfortunately, I’m not able to provide an exact timeline on the release yet. Hopefully, this summer or by the end of the year.
No problems I thought that would be the case. I might try and have a crack myself. Given I am trying to link individuals between ASECs, i.e. between ASEC in year x and year x+1, rather than between ASEC’s and monthly files, I’m hoping it should be a bit simpler than recreasting MARBASECID. But that may be optimistic. Any tips gratefully received, and thanks for all your help.