Hi,

having to use STATA as an Economics Undergrad for Economics to write my dissertation and keep coming stuck on what regressions to run for my data from 2013 to 2019.

i am looking into a particular educational policy reform which changed the nature of care packages for children and young adults with disabilities. it basically caused a huge increase in demand and lack of funding for these care packages. the data i have covers the total amount of care packages, the total amount of funding and the funding per capitia (all per local authority.)

i have a supervisor for my dissertation who wants me to:
- Get a year on year trend of per capita funding; the total number of people enrolled; and the total funding (these would therefore be my DVs).

- In order to find this trend I would regress these DVs on year, yeardummy (showing first effect in 2016) and the interaction (year*yeardummy). **i am confused about the interaction term - he said 'You want to capture a divergence in trend in the per capita funding for high needs students therefore Think about how year and post legislation enactment need to be interacted to capture this'.

- do the same with the growth of these variables using logs.

This would allow me to see the variation throughout the years and therefore see whether there was a mismatch between the amount of funding and the amount of people enrolled with these care packages, hence showing a need for government intervention.


Does anyone know from this information what regressions i should be running? is it simple OLS, fixed effects or random effects?