I havereceived excellent help from Statalist before, so now I try again. My co-authorand I are seeking to estimate household consumption behaviour using microdataand we are particularly interested in the marginal propensity to consume out ofdifferent forms of income.
Thehouseholds in our dataset are sampled twice implying that we effectively have a panel with a time dimension equal to only two periods. Neither myco-author nor I have experience in working with such �shallow� panels and would appreciate feedback on these questions:
1) Arethere issues with very shallow panels (two periods) one should be aware of?
2 Are therepaper or other references dealing with the issue of panels with an extremelyshort time dimension?
Thank youvery much!
Karsten Staehr
