with only the Netherlands having a real time estimate that is more certain than the full
sample estimate. Full sample uncertainty rises over the time period and approaches that of
the real time estimate. Only in Spain are both estimates of the uncertainty in the gap low
and stable.
|
Table 7.3 |
Standard Deviations of Output Gap Estimates in Selected Years - Full | ||||
|
1990__________ |
1995 |
2000 |
2005 | ||
|
Belgium |
Full-sample |
0.881 |
0.881 |
0.887 |
1.069 |
|
Recursive |
1.145 |
1.380 |
1.283 |
1.261 | |
|
France |
Full-sample |
'"^2.'169 |
2.184 |
2.192 |
2.686 |
|
Recursive |
3.956 |
2.996 |
3.057 |
2.822 | |
|
Germany |
Full-sample |
1.018 |
1.021 |
1.043 |
1.448 |
|
Recursive |
1.309 |
1.605 |
1.976 |
1.557 | |
|
Italy |
Full-sample |
1.067 |
1.071 |
1.089 |
1.472 |
|
Recursive |
1.645 |
1.800 |
1.691 |
1.680 | |
|
Netherlands |
Full-sample |
1.718 |
1.722 |
1.738 |
2.043 |
|
Recursive |
1.164 |
1.373 |
1.422 |
2.078 | |
|
Spain |
Full-sample |
0.430 |
0.430 |
0.431 |
0.522 |
|
Recursive |
0.799 |
0.503 |
0.649 |
0.625 | |
|
UK |
Full-sample |
1.393 |
1.431 |
1.470 |
2.226 |
|
Recursive |
2.526__ |
2.740 |
2.521 |
2.363 | |
7.3 Cyclically Adjusting the Budget Deficit
The patterns of evolving uncertainty detailed above should be reflected in our uncertainty
about the scale of the cyclically adjusted budget deficit. The relationship between the
deficit as a percent of GDP (BUD) and the cyclically adjusted deficit as a per cent of
GDP (CABUD) depends on the sensitivity of revenues and expenditure to the cycle. The
links between output and the budget deficit are discussed, for instance in Mélitz (2000),
Blanchard and Perotti (2002) and Wyplosz (2003). It is common to assume that an
increase in the output gap of 1.0 per cent of GDP (more use of capacity) reduces the
deficit by 0.5 per cent of GDP; we discuss this assumption further in section 7.4 below.
We may write the cyclically adjusted budget deficit (CABUD) for the UK, for example,
as dependent on the observed deficit (BUD) and the output gap (OG)
UKCABUD= UKBUD -0.5*UKOG (7.1)
Given the uncertainty in the output gap, and on the assumption that the budget deficit and
the cyclical coefficient are known, then we may write the standard deviation SD of the
CABUD as a function of the standard deviation of the output gap:
SD(UKCABUD) = 0.5*SD(UKOG) (7.2)
We apply this formula across seven European countries, applying the time varying
estimates of the standard deviation of the output gap to produce bounds of uncertainty
around budget deficits. The SD of the full sample estimate has generally settled down by
1990, but the real time estimates of the SD vary noticeably over time.
We focus here on estimating the real time cyclically adjusted deficits. These are of
concern to policy-makers, who have to make decisions in real-time. In Barrell et al.
192