I am working on a perfect foresight growth model. I would like to know if one needs to specify fiscal rules in a derterministic case? I have a government budget constraint composed of public spendings, social transfers and 4 taxes.
In a stochastic case, one would specify 7 fiscal rules by assuming AR(1) exogenous processes for taxes, social transfers and government spendings. Can I assume as well AR(1) (without the white noise) in a perfect foresight model?
Or should I just specify in an initval and endval block the values for the government’s controled variables ? I am trying to study taxes changes.
I don’t think the question is fully specified. If you include fiscal instruments, the question is how you determine the value of those instruments. In stochastic models you include random shocks to see the transmission of the model or to estimate those shocks. In a perfect foresight context, you can still specify the feedback component of a fiscal rule without the shocks if you think that this is a realistic depiction of reality. But it could also be that you have particular values for the tax instruments in mind that you want to impose. In that case, they would simply be exogenous variables.
So in the end it depends on what exactly you are trying to do.
I’ve checked one discussion where you wrote “That can’t work. All the tax revenue and the spending are exogenous. There is nothing that assures that debt is not explosive, i.e. the government’s intertemporal budget constraint is not satisfied. In Leeper’s terms, at least one policy must be active, but in your case, there is only passive policy. You would need e.g. some feedback of debt to taxes or debt.”.
Did you mean that at least one policy must be passive? In Leeper’s term “the passive authority’s decision rule necessarily depends on the current state of government debt” &" a passive tax rule in which the tax authority adjusts taxes in response to government debt sufficiently to stabilize debt ".
Then if I have values in mind for my taxes (then being exogenous and active policies since not related to debt), I should then make spendings and social transfers depend on debt in order to have passive policies?
That comment was mode in a stochastic model. In a perfect foresight context, you could model explosive debt dynamics, but that is typically not desired. That is why you may want to have at least one stabilizing, i.e. passive fiscal policy instrument.