In a nutshell, I'm using DID to estimate the impact of a cash-transfer program on fertility in a specific country. The program offers families an untaxed, unconditional cash transfer for every second and subsequent child.
The dependent variable is TFR (total fertility rate).
I'm estimating this simple DID model: numerical example
I don't have individual level or microdata, only aggregate level fertility data.
I also only have 1 year of post data (2017) (program was implemented in early 2016), so at the moment I'm looking at 2016 (pre) and 2017 (post). I assume that any behavior changes in response to the program wouldn't show up in the data until 2017.
My treatment group consist of anyone eligible to receive the benefit (people with 2 or more children) and the control group consists of people not eligible (people with less than 2 children).
I've restricted the sample to childbearing aged women (15-45) and calculated the TFR for the treatment (2+ birth parity) & control (less than 2 children) groups pre (2016) and post (2017) program implementation.
The DID value I get is: -.06
- What does a negative value mean?
- Did I follow a correct procedure in estimating the simple DID?
- How would I do a t-test between these groups?
- How much of a limitation is the amount of post data I have? (1 year)
- Would it work to include an average of a larger time frame in the pre period (for example, 2010-2016) and just 2017 for post?
- What data would I need to estimate a simple DID regression?