The purpose of an evaluation of fiscal sustainability is to assess the need for
policy reforms. Such an evaluation naturally takes its outset in current or un-
changed policies, that is, the purpose is not to predict future developments, but
to create an informed basis on which to discuss the need and nature of reforms.
Leaving aside the practical problems of defining what is meant by unchanged
policies, cf below, there is the fundamental problem that a fully specified in-
tertemporal general equilibrium model necessarily builds on the premise that
the intertemporal budget constraint of the public sector would have to be ful-
filled. That is, an instrument for adjustment needs to be specified to close the
model. This property should not be mixed up with Ricardian equivalence since
it follows from a consistent modelling of all interrelationships - the attractive
property and disciplining device of formulating an explicit intertemporal general
equilibrium model. To illustrate this point note that in a standard OLG model
Ricardian equivalence does not hold, however, this does not imply that there is
no intertemporal budget constraint for the public sector.
A pragmatic way of solving this problem is to close the budget by assuming
a lump sum transfer in the far future, cf below. This would allow debt to
accumulate and make it possible to assess the consequences of "passive" policies.
While a pragmatic solution, this approach is not without its problems. The
expectations of agents would differ between a case where the budget constraint
is fulfilled by an outside transfer and a case with a future lump sum tax increase.
However, this problem may have less importance for the path in the initial
periods
Pay-as-you go
A simple constraint to impose on the analysis is to require a balanced budget
on a period-by-period basis. Requiring that the total balance is zero for all future
periods, i.e.
b(xt,yt,zt) - rtdt =0 for all t (6)
This ensures an unchanged debt position, i.e.
dt = dt+1 for all t
Clearly this condition can still be fulfilled for an infinite of combinations of the
policy instruments, and it is not immediately clear from (6) what order of
magnitudes should be attained by policy changes aiming at ensuring that the
condition holds. Clarity can be gained by choosing one instrument (say θ1) and
solve for the value of the instrument ensuring that (6) holds leaving the policy
functions and all other policy parameters constant and thus time-invariant.
This approach yields a time dependent value of the policy parameter θp1ayg (t)
which ensures that (6) is fulfilled. The relation and path of θ1payg (t) relative
to the current value of θ1 (0) yields a perspective on the need and order of
magnitude of future policy changes as well as the underlying time profile. If
there is a systematic difference between θ1 (0) and θ1payg (t), current policies are
not sustainable if θp1ayg (t) is systematically above (below) θ1 (0) and θ1 is a tax
(expenditure) instrument.