As a starting point, Figure 4.11 provides a two-dimensional decomposition of the budget
balance. Vertically, the budget balance is decomposed into factors which are economy-
induced and those which are policy-induced. Horizontally, a distinction is made factors
with a temporary impact on the budget as opposed to those which have a permanent
impact.
Figure 4.11 |
Determinants of the Actual Budget Balance Temporary Permanent | ||
Economy |
A |
C | |
Policy |
B |
D |
Square A would thus includes factors related to the temporary state of the economy at
given tax and expenditure rules. An example would be unemployment benefit
expenditures and temporary higher/lower tax revenues due to tax bases being temporarily
higher/lower than their trend values or tax elasticities being temporarily higher/lower than
normal. Square B, includes, for example, temporary higher expenditures due to
discretionary stabilisation policy measures. Square C would capture, for example, how
potential GDP and demographic trends affect the budget balance permanently, while
square D includes the impact on the budget from permanent changes to tax and
expenditure rules.65
On this basis a comparison of the coverage of the authors decomposition can be made to
the coverage of the CAB in relation to the different budget concepts. This is done in
Table 4.2.
Table 4.2 What Do the Different Budget Indicators Capture?
Budget Concept |
Determinants of budget |
Corresponding best to |
Use of CAB |
Structural budget balance |
C+D |
Core component |
CAB (B+C+D) |
Fiscal stance (change) |
B+C+D |
Discretionary component (B) |
Change CAB (B+C+D) |
Discretionary policy |
B+D |
Discretionary component |
Change CAB (B+C+D) |
Fiscal impulse (change) |
A+B+C+D |
(n.a. change nominal ________balance)________ |
(n.a. change nominal |
Let me go through the concepts in turn. First, if the aim is to capture the permanent
budget trend all temporary influences should be taken out. This would correspond to the
concept of a “structural budget balance” (C+D). The core component in the authors’
model would then seem most suitable to capture this. When the CAB on the other hand is
See also Braconier and Forsfalt (2004).
116