[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: panel data and impact estimation
I have 5 household surveys, from 1999-2004, all beneficiaries of a rural
development program in Mexico, each year with different people. The samples
are representatives at the state level (32 states), so I could make the
agrupation by state and obtain a panel of 32X5 observations. I also have a
variable for Rural Development Districts (DDR), that are territorialy
smaller entities than states, so I could try to build a bigger panel, but
I'm not sure if the samples by DDR are representaives. My intention is to
use instrumental varaibles for estiamting income.
My questions are:
1) If it is possible to estimate the impact of the program just having
beneficiaries into the panel?
2) How many are enough observations? (32X5?)
* For searches and help try: