IFS
Uncertainty in the forecasts
Carl Emmerson
-4.0
-2.0
0.0
2.0
4.0
6.0
Actual
8.0
76
-
19 77
78
19 79
80
19 81
82
19 83
84
19 85
86
-8
19 7
88
19 89
90
19 91
92
19 93
94
19 95
96
-9
19 7
98
20 99
00
20 01
02
-0
3
10.0
19
Percentage of national income
Previous HMT forecasting errors
Financial year
© Institute for Fiscal Studies, 2004
Source: HM Treasury (2003)
-4.0
-2.0
0.0
2.0
4.0
Actual
Callaghan
Thatcher
Major
Blair
6.0
8.0
76
-
19 77
78
19 79
80
19 81
82
19 83
84
19 85
86
-8
19 7
88
19 89
90
19 91
92
19 93
94
19 95
96
-9
19 7
98
20 99
00
20 01
02
-0
3
10.0
19
Percentage of national income
Previous HMT forecasting errors
Financial year
© Institute for Fiscal Studies, 2004
Source: HM Treasury (2003)
Errors in borrowing forecasts
% of GDP
Average error
1 year
ahead
2 years
ahead
3 years
ahead
4 years
ahead
0.05
0.05
–0.58
–1.10
© Institute for Fiscal Studies, 2004
Source: HM Treasury (2003); Emmerson, Frayne and Love (2004)
Errors in borrowing forecasts
% of GDP
1 year
ahead
2 years
ahead
3 years
ahead
4 years
ahead
Average error
0.05
0.05
–0.58
–1.10
Average absolute error
1.03
1.65
2.22
2.95
Standard deviation of
errors
1.26
1.98
2.82
3.62
© Institute for Fiscal Studies, 2004
Source: HM Treasury (2003); Emmerson, Frayne and Love (2004)
Uncertainty over path of PSNB
• Forecast errors are normally distributed
• Average error is zero
– But forecasts are based on ‘cautious assumptions’
• Accuracy is the same as past
– This will understate true forecasting errors as
previous errors not adjusted for subsequent policy
changes
© Institute for Fiscal Studies, 2004
-4.0
-2.0
0.0
2.0
Central
4.0
6.0
-2
00
20 0
00
-0
20 1
01
-0
20 2
02
-0
20 3
03
-0
4
20
04
-0
20 5
05
-0
20 6
06
-0
20 7
07
-0
20 8
08
-0
9
8.0
19
99
Percentage of national income
Uncertainty over path of PSNB
Financial year
© Institute for Fiscal Studies, 2004
Source: IFS calculations
-4.0
-2.0
20%
0.0
2.0
Central
4.0
20%
6.0
-2
00
20 0
00
-0
20 1
01
-0
20 2
02
-0
20 3
03
-0
4
20
04
-0
20 5
05
-0
20 6
06
-0
20 7
07
-0
20 8
08
-0
9
8.0
19
99
Percentage of national income
Uncertainty over path of PSNB
Financial year
© Institute for Fiscal Studies, 2004
Source: IFS calculations
-4.0
-2.0
40%
0.0
20%
2.0
Central
4.0
20%
6.0
40%
-2
00
20 0
00
-0
20 1
01
-0
20 2
02
-0
20 3
03
-0
4
20
04
-0
20 5
05
-0
20 6
06
-0
20 7
07
-0
20 8
08
-0
9
8.0
19
99
Percentage of national income
Uncertainty over path of PSNB
Financial year
© Institute for Fiscal Studies, 2004
Source: IFS calculations
-4.0
60%
40%
20%
Central
20%
40%
60%
-2.0
0.0
2.0
4.0
6.0
-2
00
20 0
00
-0
20 1
01
-0
20 2
02
-0
20 3
03
-0
4
20
04
-0
20 5
05
-0
20 6
06
-0
20 7
07
-0
20 8
08
-0
9
8.0
19
99
Percentage of national income
Uncertainty over path of PSNB
Financial year
© Institute for Fiscal Studies, 2004
Source: IFS calculations
-4.0
80%
60%
40%
20%
Central
20%
40%
60%
80%
-2.0
0.0
2.0
4.0
6.0
-2
00
20 0
00
-0
20 1
01
-0
20 2
02
-0
20 3
03
-0
4
20
04
-0
20 5
05
-0
20 6
06
-0
20 7
07
-0
20 8
08
-0
9
8.0
19
99
Percentage of national income
Uncertainty over path of PSNB
Financial year
© Institute for Fiscal Studies, 2004
Source: IFS calculations
Uncertainty over path of CBS
• Forecast errors are normally distributed
• Average error is zero
– But forecasts are based on ‘cautious assumptions’
• Accuracy is the same as past
– This will understate true forecasting errors as previous errors
not adjusted for subsequent policy changes
• Assume similar errors in forecasting current budget
surplus and net borrowing
• Account for correlation in errors from all forecasts
made at the same time
© Institute for Fiscal Studies, 2004
Probability of meeting golden rule
Will the golden rule be met?
1.0
0.8
0.6
50/50 chance
1yr out
2yrs out
3yrs out
4yrs out
0.4
0.2
0.0
-4
-3
-2 -1
0
1
2
3
4
5
6
7
8
Expected cumulative CB surplus over cycle (% of GDP)
9
10
© Institute for Fiscal Studies, 2004
Source: Emmerson, Frayne and Love (2004)
Probability of meeting golden rule
Will the golden rule be met?
1.0
0.8
Budget 2002
0.6
50/50 chance
1yr out
2yrs out
3yrs out
4yrs out
0.4
0.2
0.0
-4
-3
-2 -1
0
1
2
3
4
5
6
7
8
Expected cumulative CB surplus over cycle (% of GDP)
9
10
© Institute for Fiscal Studies, 2004
Source: Emmerson, Frayne and Love (2004)
Probability of meeting golden rule
Will the golden rule be met?
1.0
0.8
Budget 2002
Budget 2003
0.6
50/50 chance
1yr out
2yrs out
3yrs out
4yrs out
0.4
0.2
0.0
-4
-3
-2 -1
0
1
2
3
4
5
6
7
8
Expected cumulative CB surplus over cycle (% of GDP)
9
10
© Institute for Fiscal Studies, 2004
Source: Emmerson, Frayne and Love (2004)
Probability of meeting golden rule
Will the golden rule be met?
1.0
0.8
Budget 2002
Budget 2003
0.6
Budget 2004
50/50 chance
1yr out
2yrs out
3yrs out
4yrs out
0.4
0.2
0.0
-4
-3
-2 -1
0
1
2
3
4
5
6
7
8
Expected cumulative CB surplus over cycle (% of GDP)
9
10
© Institute for Fiscal Studies, 2004
Source: Emmerson, Frayne and Love (2004)
Probability of meeting golden rule
Will the golden rule be met?
1.0
0.8
Budget 2002
Budget 2003
PBR 2004
0.6
Budget 2004
50/50 chance
1yr out
2yrs out
3yrs out
4yrs out
0.4
0.2
0.0
-4
-3
-2 -1
0
1
2
3
4
5
6
7
8
Expected cumulative CB surplus over cycle (% of GDP)
9
10
© Institute for Fiscal Studies, 2004
Source: Emmerson, Frayne and Love (2004)
Will the golden rule be met?
100%
82.2%
74.2%
Percentage
80%
62.0%
59.1%
60%
37.0%
40%
33.7%
20%
0%
Budget
02
Budget
03
Budget Budget
04
04, 200405 as
trend
PBR 04, PBR 04,
2004-05 but 2004correct
05 as
trend
© Institute for Fiscal Studies, 2004
Source: IFS Calculations
Conclusions
• Over the current economic cycle:
– Treasury forecast for Golden rule to be met
– At best a 62% chance of success
– Relies on stronger growth in receipts and slower
growth in spending
• Looking forwards
– Treasury forecast for Golden rule to be met
– Relies on strong growth in tax receipts
– Not expected to be met under ‘cautious case’
© Institute for Fiscal Studies, 2004