Review of Canadian Federal Fiscal Forecasting, Processes and Systems: 3
- Table of Contents - Previous -

## Archived - Tables and Charts

Archived information

Archived information is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Table 1
Comparison of Estimates of the Error in Forecasts of the Fiscal Balance
($Billions)  Balance forecast in the budget before the fiscal year began Actual balance (outcome)1 Difference (forecast less actual balance)2 In-year policy initiatives3 Adjusted balance (outcome had there been no policy initiatives) Adjusted difference (forecast less adjusted balance)2 A. Policy and Economic Analysis Program - including policy initiatives (PEAP - CIRANO 1) 1995-1996 -30.3 -28.6 -1.7 1996-1997 -21.8 -8.9 -12.9 1997-1998 -13.3 3.5 -16.8 1998-1999 3.0 2.9 0.1 1999-2000 3.0 12.3 -9.3 2000-2001 4.0 17.1 -13.1 2001-2002 8.2 8.9 -0.6 2002-2003 2.0 4.0 -2.0 2003-2004 4.0 9.1 -5.1 B. Policy and Economic Analysis Program - excluding policy initiatives (PEAP- CIRANO 2) 1995-1996 -30.3 -28.6 -1.7 0.0 -28.6 -1.7 1996-1997 -21.8 -8.9 -12.9 -0.7 -9.6 -12.2 1997-1998 -13.3 3.5 -16.8 5.9 9.4 -22.7 1998-1999 3.0 2.9 0.1 5.6 8.5 -5.5 1999-2000 3.0 12.3 -9.3 6.2 18.5 -15.5 2000-2001 4.0 17.1 -13.1 5.4 22.5 -18.5 2001-2002 8.2 8.9 -0.6 3.6 12.5 -4.3 2002-2003 2.0 4.0 -2.0 7.4 11.4 -9.4 2003-2004 4.0 9.1 -5.1 4.7 13.8 -9.8 C. Dale Orr, Global Insight 1995-1996 -30.3 -28.6 -1.7 1996-1997 -21.8 -8.9 -12.9 1997-1998 -14.0 3.5 -17.5 1998-1999 3.0 2.9 0.1 1999-2000 3.0 12.3 -9.3 2000-2001 4.0 17.1 -13.1 2001-2002 8.3 8.9 -0.7 2002-2003 2.0 7.0 -5.0 2003-2004 4.0 9.1 -5.1 D. Jim Stanford, Canadian Auto Workers 1995-1996 -32.7 -28.6 -4.1 1996-1997 -24.3 -8.9 -15.4 1997-1998 -17.0 3.8 -20.8 1998-1999 3.0 3.1 -0.1 1999-2000 3.0 12.7 -9.7 2000-2001 3.0 18.1 -15.1 2001-2002 1.5 8.9 -7.4 2002-2003 3.0 7.0 -4.0 2003-2004 3.0 9.1 -6.1 Note: Some totals may not add due to rounding. 1 With two exceptions, actual balances are those reported at the end of the fiscal year in question. Many have since been revised to a full accrual accounting basis. The first exception is in the actual outcomes used by Stanford, who used Finance Canada's Fiscal Reference Tables published in 2002 for all years up to 2001-02. The second exception is 2002-03, where PEAP – CIRANO adjusted the actual outcome to a partial accrual accounting basis so that comparisons could be made on an equal footing.2 In the difference columns, a negative value indicates that the balance was better than forecast, i.e., the balance was under-forecast. When there were deficits, they were smaller than forecast; when there were surpluses, they were larger than forecast. 3 A positive value indicates the balance would have improved (a smaller deficit or a bigger surplus) by this amount had the initiatives not been taken. I.e., since 1997-98, the initiatives reduced the potential surplus. Source: PEAP – CIRANO, Orr, Stanford. Table 2 Budget Balance – Forecasts, Outcome and Differences ($ Billions)

  Balance forecast in the budget before the fiscal year began  Actual balance (outcome)1    Difference (forecast less actual balance)2  In-year policy initiatives3 Adjusted balance (outcome had there been no policy initiatives)  Adjusted difference (forecast less adjusted balance)2 1994-1995 -39.7 -37.5 -2.2 0.0 -37.5 -2.2 1995-1996 -30.3 -28.6 -1.7 0.0 -28.6 -1.7 1996-1997 -21.8 -8.9 -12.9 -0.7 -9.6 -12.2 1997-1998 -13.3 3.5 -16.8 5.9 9.4 -22.7 1998-1999 3.0 2.9 0.1 5.7 8.5 -5.5 1999-2000 3.0 12.3 -9.3 6.2 18.5 -15.5 2000-2001 4.0 17.1 -13.1 5.3 22.5 -18.4 2001-2002 8.2 8.9 -0.7 3.6 12.5 -4.3 2002-2003 2.0 4.0 -2.0 7.3 11.4 -9.4 2003-2004 4.0 9.1 -5.0 4.8 13.8 -9.8 Some totals may not add due to rounding. 1 With one exception, actual balances are those reported at the end of the fiscal year in question. Many have since been revised to a full accrual accounting basis. The exception is 2002-03, where PEAP adjusted the actual outcome to a partial accrual accounting basis so that comparisons could be made on an equal footing.2 In the difference columns, a negative value indicates that the balance was better than forecast, i.e. the balance was under-forecast. When there were deficits, they were smaller than forecast; when there were surpluses, they were larger than forecast. 3 A positive value indicates the balance would have improved (a smaller deficit or a bigger surplus) by this amount had the initiatives not been taken. I.e., since 1997-98, the initiatives reduced the potential surplus.Source: PEAP – CIRANO analysis

Table 3a
Budget Balance and Components, Including Policy Initiatives
($Billions) Budget Balance Total Revenues Program Expenditures Public Debt Charges Balance forecast in the budget before the fiscal year began Actual balance (outcome)1 Difference (forecast less Revenue forecast in the budget before the fiscal year Program expenditures forecast in the budget before the fiscal year began Actual program expenditures Difference (forecast less actual program expenditures)2 actual balance)2 began Actual revenue (outcome)1 Difference (forecast less actual (outcome)1 revenue)2 1994-1995 -39.7 -37.5 -2.2 123.9 123.3 0.6 122.6 118.7 3.9 41.0 42.0 -1.0 1995-1996 -30.3 -28.6 -1.7 133.2 130.3 2.9 114.0 112.0 2.0 49.5 46.9 2.6 1996-1997 -21.8 -8.9 -12.9 135.0 140.9 -5.9 109.0 104.8 4.2 47.8 45.0 2.8 1997-1998 -13.3 3.5 -16.8 137.8 153.2 -15.4 107.8 108.8 -0.9 43.3 40.9 2.4 1998-1999 3.0 2.9 0.1 151.0 155.7 -4.6 104.5 111.4 -6.9 43.5 41.4 2.1 1999-2000 3.0 12.3 -9.3 156.7 165.7 -9.0 111.2 111.8 -0.6 42.5 41.6 0.9 2000-2001 4.0 17.1 -13.1 162.0 178.6 -16.6 116.0 119.3 -3.3 42.0 42.1 -0.1 2001-2002 8.2 8.9 -0.7 172.8 173.3 -0.5 123.9 126.7 -2.8 40.8 37.7 3.0 2002-2003 2.0 4.0 -2.0 174.2 174.1 0.1 136.0 135.0 1.0 36.3 35.0 1.3 2003-2004 4.0 9.1 -5.0 184.7 186.2 -1.5 143.0 141.4 1.7 37.6 35.8 1.9 Source: PEAP – CIRANO and another’s calculations Table 3b Policy Adjusted Budget Balance and Components ($ Billions)

Budget Balance Total Revenues Program Expenditures Public Debt Charges  Balance forecast in the budget before the fiscal year began  Adjusted balance (outcome had there been no policy initiatives)  Adjusted difference (forecast less adjusted  Revenue forecast in the budget before the fiscal year began  Adjusted revenue (outcome had there been no policy initiatives)  Adjusted difference (forecast less adjusted revenue)2 Program expenditures Adjusted program expenditures Adjusted difference (forecast less adjusted program expenditures)2 balance)2 forecast in the budget before the fiscal year began (outcome had there been no policy initiatives) 1994-1995 -39.7 -37.5 -2.2 123.9 123.3 0.6 122.6 118.7 3.9 41.0 42.0 -1.0 1995-1996 -30.3 -28.6 -1.7 133.2 130.3 2.9 114.0 112.0 2.0 49.5 46.9 2.6 1996-1997 -21.8 -9.6 -12.2 135.0 137.7 -2.7 109.0 102.3 6.7 47.8 45.0 2.8 1997-1998 -13.3 9.4 -22.7 137.8 152.7 -14.9 107.8 102.4 5.4 43.3 40.9 2.4 1998-1999 3.0 8.5 -5.5 151.0 156.0 -4.9 104.5 106.0 -1.5 43.5 41.4 2.1 1999-2000 3.0 18.5 -15.5 156.7 166.0 -9.3 111.2 105.9 5.3 42.5 41.6 0.9 2000-2001 4.0 22.5 -18.4 162.0 180.3 -18.2 116.0 115.7 0.3 42.0 42.1 -0.1 2001-2002 8.2 12.5 -4.3 172.8 173.4 -0.5 123.9 123.1 0.7 40.8 37.7 3.0 2002-2003 2.0 11.4 -9.4 174.2 174.1 0.1 136.0 127.7 8.3 36.3 35.0 1.3 2003-2004 4.0 13.8 -9.8 184.7 186.3 -1.6 143.0 136.7 6.3 37.6 35.8 1.9 Note: Some columns may not add due to rounding. 1 With one exception, actual balances are those reported at the end of the fiscal year in question. Many have since been revised to a full accrual accounting basis. The exception is 2002-03, where PEAP adjusted the actual outcome to a partial accrual accounting basis so that comparisons could be made on an equal footing.2 In the difference columns, it is important to remember that negative values can have different meanings. In the case of the balance, a negative value indicates an under-forecast, i.e. an actual deficit was smaller than predicted and an actual surplus was bigger than predicted. In the case of revenues, a negative value also indicates that revenue was under-forecast and actual revenues were bigger than forecast; this would contribute to the under-forecast of the balance. However, under-forecasts of spending have the opposite effect on the balance. In most years, spending was over-forecast and the actual sums were lower than predicted. These positive differences also contributed to the under-forecast of the balance.

Table 4
Shares of Actual Revenues and Expenditures

 1994-95 1995-96 1996-97 1997-98 1998-99 1999-00 2000-01 2001-02 2002-03 2003-04 Total Revenues ($Billions) 123.3 130.3 140.9 153.2 155.7 165.7 178.6 173.3 174.1 186.2 % of Total Revenues Personal Income Tax 45.7 46.2 44.9 46.2 46.6 47.9 46.1 48.3 46.2 45.6 Corporate Income Tax 9.4 12.2 12.1 14.7 13.9 14.0 15.8 13.9 12.2 14.7 EI/UI Premiums 15.3 14.2 14.1 12.3 12.4 11.2 10.5 10.4 10.5 9.4 GST Revenues 13.6 12.6 12.8 12.7 13.3 13.8 14.0 14.4 15.8 15.2 Customs Duties 2.9 2.3 1.9 1.8 1.5 1.3 1.6 1.7 1.8 1.6 Excise, Other Energy Taxes 5.5 5.6 5.9 5.6 5.4 4.8 4.7 5.0 5.6 5.5 Miscellaneous Tax Revenues 1.4 1.6 2.0 1.9 1.9 2.1 2.4 1.8 2.5 1.7 Non-Tax Revenues 6.2 5.3 6.3 4.7 5.1 5.0 5.0 4.5 5.4 6.4 Total Expenditures ($ Billions) 160.8 158.9 149.8 149.7 152.8 153.4 161.4 164.4 170.1 177.1 % of Total Expenditures Public Debt Charges 26.2 29.5 30.0 27.3 27.1 27.1 26.1 23.0 20.6 20.2 Total Program Spending 73.8 70.5 70.0 72.7 72.9 72.9 73.9 77.0 79.4 79.8 Total Program Spending ($Billions) 118.7 112.0 104.8 108.8 111.4 111.8 119.3 126.7 135.0 141.4 % of Total Program Spending Direct Program Spending 47.7 45.6 46.1 49.8 46.0 48.1 48.1 48.1 47.1 49.5 EI/UI Benefits 12.5 12.0 11.8 10.9 10.7 10.1 9.6 10.9 10.7 10.7 OAS Benefits 17.3 18.8 20.6 20.4 20.5 20.9 20.3 20.0 19.5 19.0 Transfers to OLG – CHST 14.7 14.9 14.1 11.6 14.4 13.4 12.1 13.7 16.7 16.1 Transfers to OLG – Other 7.8 8.8 7.4 7.3 8.5 7.4 8.6 7.4 6.0 4.7 Note: In the 2000-01 fiscal year, transfer to persons for heating relief totalling$1.5 billion or 1.2% of program spending is included in the total but not separated out into a component. Source: PEAP – CIRANO

Table 5
Policy Adjusted Budget Balance Forecast Difference and Components
($Billions)  1994-95 1995-96 1996-97 1997-98 1998-99 1999-00 2000-01 2001-02 2002-03 2003-04 Budget Balance -2.2 -1.7 -12.2 -22.7 -5.5 -15.5 -18.4 -4.3 -9.4 -9.8 Total Revenues 0.6 2.9 -2.7 -14.9 -4.9 -9.3 -18.2 -0.5 0.1 -1.6 Personal Income Tax 3.2 0.2 0.2 -4.3 -1.5 -4.4 -8.0 -3.3 -0.4 1.7 Corporate Income Tax -1.3 -0.5 -1.9 -6.3 -1.1 -2.3 -4.4 4.3 4.3 -3.1 EI/UI Premiums 0.4 1.2 0.0 0.3 -1.1 -0.6 -0.6 0.4 -0.6 0.0 GST Revenues -0.3 1.0 -0.2 -2.0 0.2 -1.2 -1.9 0.8 -0.6 1.6 Customs Duties 0.2 0.3 0.1 -0.7 0.2 0.4 -0.6 -0.9 -0.3 0.4 Excise, Other Energy Taxes -0.4 0.1 0.0 0.2 0.3 0.2 -0.1 -0.1 -0.5 -0.4 Miscellaneous Tax Revenues 0.0 -0.3 -0.9 -0.9 -0.6 -0.6 -0.9 -0.6 -0.3 0.0 Non-Tax Revenues -1.2 0.7 0.1 -1.1 -1.2 -0.8 -1.8 -1.2 -1.6 -1.9 Total Expenditures 2.8 4.6 9.5 7.8 0.6 6.2 0.2 3.8 9.6 8.2 Public Debt Charges -1.0 2.6 2.8 2.4 2.1 0.9 -0.1 3.0 1.3 1.9 Total Program Spending 3.9 2.0 6.7 5.4 -1.5 5.3 0.3 0.7 8.3 6.3 Direct Program Spending 0.7 1.1 4.5 4.2 0.2 3.5 0.9 1.7 4.9 1.8 EI/UI Benefits 3.5 0.8 1.4 1.7 0.7 2.1 0.5 -1.5 1.4 0.7 OAS Benefits 0.1 0.2 0.3 0.1 0.1 0.1 -0.1 -0.2 0.1 -0.1 Transfers to OLG – CHST -0.4 -0.3 0.2 0.1 0.0 0.1 0.0 0.0 0.0 0.0 Transfers to OLG – Other 0.0 0.1 0.1 -0.6 -2.5 -0.4 -1.1 0.7 1.8 4.1 Note: In the 2000-01 fiscal year, transfer to persons for heating relief totalling$1.5 billion or 1.2% of program spending is included in the total but not separated out into a component.

Table 6
Measures of Forecasting Difficulty

Standard Deviation Standard Deviation/Mean Lagged Correlation 1984-1993 1994-2003 1984-1993 1994-2003 Real Growth 2.53 1.59 0.96 0.45 0.90 0.88 Unemployment Rate 1.48 1.17 0.15 0.14 0.97 0.93 Employment Growth 1.75 0.86 1.12 0.42 0.96 0.80 CPI Inflation 1.32 1.00 0.33 0.54 0.86 0.76 Nominal GDP Growth 3.23 2.55 0.55 0.48 0.94 0.85 3-Month T-Bill Rate 2.44 1.45 0.27 0.33 0.89 0.89 10-Year Benchmark Bond Yield 1.55 1.28 0.16 0.21 0.87 0.95 Current Account Balance 8.87 18.31 -0.47 2.99 0.86 0.87 Exchange Rate 5.29 3.87 0.07 0.06 0.98 0.88 Labour Income Growth 2.79 1.90 0.46 0.42 0.96 0.87 Corporate Profit Growth 20.44 21.25 6.38 1.35 0.84 0.84 Total Business Investment Growth 7.50 5.82 2.97 1.09 0.89 0.88 US Real GDP Growth 1.91 1.33 0.57 0.41 0.85 0.88 Source: PEAP – CIRANO

Table 7
Correlation Coefficients – Economic Differences and % Fiscal Differences
Year-Ahead Budget Forecasts

Real GDP Growth Nominal GDP Growth GDP Inflation Unemployment Rate Employment Growth CPI Inflation Short Rate Long Rate Real GDP Growth Total Revenues 0.36 0.34 0.15 -0.01 0.06 0.03 0.10 -0.15 Personal Income Tax 0.18 0.19 0.12 0.21 -0.08 0.33 -0.23 -0.44 Corporate Income Tax 0.17 0.18 0.12 -0.23 -0.21 -0.34 0.20 -0.01 EI Premiums 0.71 0.59 0.14 -0.35 0.41 0.02 0.41 0.21 GST Revenues 0.38 0.35 0.15 -0.15 0.58 0.04 0.21 0.08 Customs Duties -0.13 -0.19 -0.22 0.52 -0.10 0.15 -0.17 -0.07 Excise, Other Energy Taxes -0.07 -0.04 0.01 0.28 0.15 0.02 0.11 0.40 Miscellaneous Tax Revenues -0.15 -0.25 -0.30 0.27 -0.32 0.03 -0.59 -0.53 Non-Tax Revenues 0.29 0.25 0.08 0.02 0.56 -0.18 0.44 0.47 Total Spending 0.25 0.13 -0.14 -0.15 0.17 -0.22 0.23 -0.08 Total Program Spending 0.01 -0.14 -0.32 0.07 0.00 -0.15 -0.10 -0.38 Direct Program Spending -0.17 -0.34 -0.47 0.20 -0.20 -0.33 -0.14 -0.24 EI Benefits -0.70 -0.58 -0.14 0.84 -0.45 0.28 -0.62 -0.51 Old Age Security Benefits -0.18 -0.04 0.19 0.45 -0.07 0.17 0.06 0.07 Transfers to OLG – CHST 0.47 0.39 0.09 -0.44 0.63 0.09 0.36 -0.08 Transfers to OLG – Other 0.25 -0.01 -0.45 -0.39 -0.10 -0.45 -0.05 -0.05 Public Debt Charges 0.71 0.73 0.45 -0.69 0.45 -0.24 0.89 0.74 Note: Economic Differences are based on Statistics Canada 1st revised estimate. Source: PEAP – CIRANO

Table 8
Impact of Economic Forecast Differences on Year-Ahead Fiscal Forecasts Based on Department of Finance "Fiscal Sensitivities"

Total Revenues ($Billions) Budget Difference Excl Policy Budget Difference Economic Impact Economic Impact - Levels Adjusted$ Level % of Bud % of Bud $Level % of Bud ex Pol ex Pol 1994-1995 0.6 0.6 -2.1 * * -2.3 * * 1995-1996 2.9 2.9 2.4 81.7 81.7 2.2 76.8 76.8 1996-1997 -2.7 -5.9 1.0 * * -1.7 63.6 28.7 1997-1998 -14.9 -15.4 0.7 * * -4.2 28.5 27.6 1998-1999 -4.9 -4.6 2.3 * * -0.1 2.5 2.6 1999-2000 -9.3 -9.0 -5.1 55.0 56.9 -7.5 80.7 83.6 2000-2001 -18.2 -16.6 -4.5 24.5 27.0 -7.4 40.5 44.6 2001-2002 -0.5 -0.5 5.1 * * 3.0 * * 2002-2003 0.1 0.1 -4.7 * * -8.5 * * 2003-2004 -1.6 -1.5 0.5 * * -2.5 158.3 168.8 Total Program Expenditures ($ Billions)

Budget Difference Excl Policy Budget Difference Economic Impact $Level % of Bud ex Pol % of Bud 1994-1995 3.9 3.9 0.9 23.8 23.8 1995-1996 2.0 2.0 -0.7 * * 1996-1997 6.7 4.2 -0.2 * * 1997-1998 5.4 -0.9 0.7 13.7 * 1998-1999 -1.5 -6.9 0.8 * * 1999-2000 5.3 -0.6 0.9 16.9 * 2000-2001 0.3 -3.3 -0.1 * 3.6 2001-2002 0.7 -2.8 -0.5 * 18.1 2002-2003 8.3 1.0 0.8 10.0 82.7 2003-2004 6.3 1.7 -1.2 * * Public Debt Charges ($ Billions)

Budget Difference Excl Policy Budget Difference Economic Impact $Level % of Bud ex Pol % of Bud 1994-1995 -1.0 -1.0 -3.2 308.0 308.0 1995-1996 2.6 2.6 3.4 130.6 130.6 1996-1997 2.8 2.8 1.7 60.0 60.0 1997-1998 2.4 2.4 1.3 54.6 54.6 1998-1999 2.1 2.1 1.3 59.9 59.9 1999-2000 0.9 0.9 0.3 30.1 30.1 2000-2001 -0.1 -0.1 0.0 * * 2001-2002 3.0 3.0 1.5 48.0 48.0 2002-2003 1.3 1.3 -0.1 * * 2003-2004 1.9 1.9 0.6 34.3 34.3 * Denotes fiscal difference increases with adjustment for actual economic growth and interest rates Source: PEAP – CIRANO Table 9 Decomposition of Economic Forecast Impact −Total Revenues Percentage of Difference Explained By:$ Billion Level Difference Economic Difference Revenue Rate Difference Residual Difference 1994-95 0.6 -319.5 413.3 6.1 1995-96 2.9 63.2 37.3 -0.5 1996-97 -2.7 81.5 18.2 0.3 1997-98 -14.9 34.4 63.2 2.4 1998-99 -4.9 11.8 87.8 0.3 1999-00 -9.3 88.9 10.5 0.6 2000-01 -18.2 38.8 58.6 2.6 2001-02 -0.5 -492.5 601.8 -9.3 2002-03 0.1 -14,716.7 14,158.7 658.0 2003-04 -1.6 158.2 -57.4 -0.8 Source: PEAP-CIRANO

Table 10
Fiscal Forecasting and Budget Processes in Selected OECD Countries

Summary Table

Australia

 Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process Economic and Fiscal Snapshot 2004 Financial balance: 0.7% of GDP Net financial liabilities: 1.9% of GDP Structural balance: 0.5% of GDP Real GDP growth: 3.6% Charter of Budget Honesty Act (1998): - Debt at prudent levels. - Stability and predictability in the level of the tax burden. - Intergenerational report every five years. Fiscal Strategy Statement:(in every budget) - Balanced budget over the cycle. - No increase in overall tax burden from 1996-97. - Improving net worth position. Federal government has complied with its fiscal rules and targets since their introduction. Framework for federal government’s fiscal policy set by the Charter of Budget Honesty Act (1998). Requires the government to use external accounting and reporting standards and report any deviation from these standards. Economic and fiscal forecasts produced by Australian Treasury. 1 The fiscal snapshot is for the total government sector, which in the case of Canada includes federal, provincial-territorial and local governments as well as the Canada Pension Plan and Quebec Pension Plan. Sources: OECD Economic Outlook No. 76 (December 2004); national authorities’ publications; individual country’s OECD Economic Surveys; OECD/World Bank Budget Practices and Procedures Database.

Canada

 Compliance with Fiscal Economic and Fiscal Snapshot Fiscal Rules and Targets Rules and Targets Budget Process 2004 Financial balance: 1.2% of GDP Net financial liabilities: 31.1% of GDP Structural balance: 1.1% of GDP Real GDP growth: 2.8% No formal short-term budget balance target. First announced in Budget 2004, government's objective of lowering federal debt/GDP ratio to 25% by 2014-15 confirmed in Budget 2005. Debt Repayment Plan to ensure federal debt/GDP ratio remains on a permanent downward track. Federal government has complied with or surpassed its targets since their introduction. Each fall and prior to the Budget, the Department of Finance conducts extensive consultations with an economic advisory group. Mid-yearEconomic and Fiscal Update also includes private sector fiscal forecasts on a National Accounts basis, which are translated into Public Accounts projections in consultation with the private sector economic forecasting firms. At the time of the Budget, the status quo fiscal projections are updated based on the most recent economic and fiscal information.

New Zealand

 Budget Process Economic and Fiscal Snapshot Fiscal Rules and Targets Compliance with Fiscal Rules and Targets 2004 Financial balance: 2.9% of GDP Net financial liabilities: 9.9% of GDP Structural balance: 1.8% of GDP Real GDP growth: 4.8% Fiscal Responsibility Act (1994) - Surplus over reasonable time frame. - Gross debt and net worth at prudent levels. - Publication of fiscal policy intentions twice a year. - Surplus to pre-fund the New Zealand Superannuation Fund (NZSF). - Gross debt on a declining path with a target ratio of 20% of GDP before 2015. (Prior to 2004, debt target was 30% of GDP over the economic cycle.) Central government has complied with its fiscal rules and targets since their introduction. In 2004, modified its debt target to adapt to a better situation. NZ Treasury produces preliminary economic forecasts. Final version is adjusted to include new data, views from an external panel of experts and discussions with private sector representatives. Alternative scenarios are calculated to complement the central forecast. Fiscal forecasts are produced exclusively by the NZ Treasury for the current year and the following three fiscal years. Inland Revenue Department also produces alternative revenue forecasts, which are published in the budget documents.

Sweden

 Economic and Fiscal Snapshot Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process 2004 Financial  balance: 0.5% of GDP Net financial liabilities: 3.8% of GDP Structural  balance: 0.6% of GDP Real GDP growth: 3.3% Fiscal Budget Act 1996 - Nominal ceilings on 27 spending categories set out three years in advance. - Average surplus of 2% of GDP over the cycle. - Balanced budget requirement for local governments since 2000. Central government has complied with its expenditure ceilings each year since their introduction. However, unlikely to meet its surplus target over the 2000-2007 period (an average surplus of 1.3 per cent of GDP is expected for the period). Economic assumptions are prepared by the Ministry of Finance, which also looks at private sector forecasts. Fiscal forecasts, which are based on the economic forecasts and current tax rules, are prepared by the Ministry of Finance. Top-down budgetary process, multiyear expenditure ceilings and surplus target. Expenditure ceilings are the cornerstone of the budget framework.

United Kingdom

 Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process Economic and Fiscal Snapshot 2004 Financial balance: -3.2% of GDP Net financial liabilities: 36.3% of GDP Structural  balance: -3.4% of GDP Real GDP growth: 3.2% Code for Fiscal Stability, 1998: - Golden rule: borrow (run deficit) only for investment over cycle. - Sustainable investment rule: 40% net debt to GDP over cycle. Reporting requirements such as annual Pre-Budget Report and Long-Term Public Finance Report. In its 2005 budget, the government stated that it will meet its two fiscal rules over the current cycle (defined as starting in 1999-2000 and ending in 2005-06). Forecasts are based on "cautious assumptions" audited on a three-year rolling basis. Public finance forecasts based on trend economic growth rate of  0.25 percentage points lower than the government’s view (central case). An alternative scenario (cautious case) where trend growth is  1 percentage point lower than the central case is published to illustrate the risks. Fiscal rules are assessed in both the central and cautious cases. Illustrative 50-year projections included as an annex to the budget.

United States

 Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process Economic and Fiscal Snapshot 2004 Financial balance: -4.4% of GDP Net financial liabilities1: 37.0% of GDP Structural balance: -4.2% of GDP Real GDP growth: 4.4% Budget Enforcement Act (expired in 2001) contained discretionary spending caps and "pay-as-you-go" (PAYGO) rules for mandatory spending and revenue legislation. Any discretionary spending legislated above the cap or any mandatory expenditure increase or revenue reduction creating a net fiscal cost would trigger across-the-board cuts (sequestration) for non-exempt programs. Legislated debt level is increased as needed (automatically in the case of the House) and does not create difficulty for increasing the debt. President submits budget proposal to legislators (Congress). Houses of Congress reach joint resolution on budget aggregates. Congress passes appropriations bills and any revenue measures and debt legislation. President signs bills into law. The Office of Management and Budget (OMB) assists the President with fiscal and economic policy, preparation of the budget, and overseeing its administration of agencies. The Congressional Budget Office (CBO) assists Congress in evaluating the President’s budget proposals and in developing the budget resolution. It issues non-partisan economic and fiscal forecasts and reports and does not make recommendations. 1 Adjusted to exclude certain government employee pension liabilities to improve comparability with other countries' debt measures.

European Union

Economic and Fiscal Snapshot Fiscal Rules and Targets Compliance with Fiscal Budget Process Rules and Targets See individual country information. The 1992 Maastricht Treaty requires: - Deficit ceiling at 3 per cent of GDP. - Debt limit at 60 per cent of GDP. The Stability and Growth Pact, adopted in 1997, came into effect in  1999 to strengthen the Maastricht Treaty provisions. - Requirement to achieve a close-to-balance budgetary position in the medium term. In 2004, five EU-15 member countries are estimated to have had deficits equal or higher than 3 per cent of GDP, including France, Germany and Italy. See individual country information.

France

Economic and Fiscal Snapshot Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process 2004 Financial balance: -3.7% of GDP Net financial liabilities: 46.1% of GDP Structural balance: -3.1% of GDP Real GDP growth: 2.1% No official fiscal rule or target, except that the country is subject to the Maastricht Treaty provisions. France has been in breach of the Stability and Growth Pact 3-per-cent deficit ceiling since 2002. Ministry of Finance plays a central role in designing, implementing and controlling the budget. Economic assumptions are prepared by the Ministry of Finance and are revised twice a year. Fiscal forecasts are prepared by the Ministry of Finance.

Germany

Economic and Fiscal Snapshot Fiscal Rules and Targets Compliance with Fiscal Budget Process Rules and Targets 2004 Financial balance: -3.9% of GDP Net financial liabilities: 54.7% of GDP Structural balance: -2.6% of GDP Real GDP growth: 1.2% Constitutional rule requires a balanced budget at the federal level but allows borrowing for investment expenditures (golden rule). The Constitution specifies exceptions during economic downturns or war and that restrictive fiscal policy should not destabilize the economy or restrict growth and prosperity. For 2003 and 2004, all levels of governments agreed to a domestic stability pact including spending limits, although it is not enforced. In breach of the Stability and Growth Pact’s deficit ceiling since 2002. Golden rule breached over the last few years. A financial-planning council chaired by the federal Finance Ministry makes recommendations for coordinating federal, state and municipal fiscal policies. Twice a year, a panel of experts publishes estimates of tax revenues, economic growth, etc., which are then used as inputs by the Finance Ministry. Economic forecasts are prepared by the Ministry of Economy and are reviewed by an independent panel of experts.

Italy

 Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process Economic and Fiscal Snapshot 2004 Financial balance: -2.9% of GDP Net financial liabilities: 96.2% of GDP Structural balance: -2.7% of GDP Real GDP growth: 1.3%. No fiscal rule or target, except that the country is subject to the Maastricht Treaty provisions. Italy has met the Stability and Growth Pact (SGP) deficit ceiling since 1998. However, Eurostat refused to validate Italy’s deficit figures for  2003 and 2004, which could be revised upwards, marking a breach of the SGP deficit ceiling for these years. Italy’s gross debt burden is significantly higher than the SGP debt limit of 60 per cent of GDP. Economic assumptions for the budget are prepared internally at the Ministry of the Treasury. Fiscal forecasts are prepared by the Ministry of the Treasury for a three-year period. Economic growth assumptions used in the preparation of the three-year fiscal forecasts are up to 0.5 percentage points lower than the actual Treasury forecast. The annual budget includes a central reserve fund (1.1% of total expenditures) to meet unforeseen expenditures.

Netherlands

  Fiscal Rules and Targets Compliance with Fiscal Rules and Targets Budget Process Economic and Fiscal Snapshot 2004 Financial balance: -2.9% of GDP Net financial liabilities: 39.0% of GDP Structural balance: -0.6% of GDP Real GDP growth: 1.2% After an election, the government produces a "Coalition Agreement," which establishes overall budget policy and the fiscal targets for the duration of the term. Coalition Agreements presented in 1994, 1998 and 2003 included: - Cautious GDP growth assumptions. - Real fixed net expenditure ceilings, set for 4 years, on 3 sectors: central government, social security funds and health care. - Pre-determined allocation of revenue windfalls and shortfalls between budget balance and tax changes. Has generally met expenditure ceilings over the last decade. A severe cyclical downturn and recent new measures that reduced the budgetary margins led the country to a breach of the SGP 3-per-cent deficit ceiling in 2003. The government implemented substantial consolidation measures, which helped bring the deficit below 3 per cent of GDP in 2004. The country has a long history of coalition governments. Culture of fiscal responsibility highlighted by the use of the Central Planning Bureau (CPB), a trusted independent government institution that produces non-partisan economic and fiscal estimates. Prior to an election, the CPB produces an economic forecast on which all parties base their platform estimates. Parties then send platforms to CPB, which then costs the programs and estimates any economic impacts. The Minister of Finance is responsible for fiscal policy and the preparation of the annual budget. The Ministry of Finance bases its budget calculations on the independent economic assumptions prepared by the CPB. Deviations from the CPB economic forecasts do occur, albeit very infrequently.

Chart 1
Private Sector Forecasts

Average Forecast [ ■ ] Actual [ ● ]

Source: PEAP – CIRANO

Chart 2
Private-Sector Forecasts

Average Forecast [ ■ ] Actual [ ● ]

Source: PEAP – CIRANO

Chart 3
Private-Sector Forecasts

Average Forecast [ ■ ] Actual [ ● ]

Source: PEAP – CIRANO

Chart 4
Private Sector Forecasts

Average Forecast [ ■ ] Actual [ ● ]

Source: PEAP – CIRANO

## Appendix 1 Individuals Consulted

Canada

Elly Alboim
Principal
Earnscliffe Strategy Group

Scott Clark*
Board Director (Canada and Morocco)
European Bank for Reconstruction and Development

David Dodge*
Governor
Bank of Canada

Don Drummond
Senior Vice-President and Chief Economist
TD Bank Financial Group

Peter Dungan
Director
Policy and Economic Analysis Program
Institute for Policy Analysis
University of Toronto

James G. Frank
Chief Economist
Office of the Leader of the Opposition

Clément Gignac
Senior Vice President
Chief Economist and Strategist
National Bank Financial

Jamey Heath
Research and Communications Director
Federal New Democratic Party Caucus

Ron Kneebone
Department of Economics and
Institute for Advanced Policy Research
University of Calgary

Kevin Lynch*
Executive Director (Canada and Ireland)
International Monetary Fund

John Manley
Senior Counsel
McCarthy Tétrault LLP

Jack Mintz
President and CEO
C.D. Howe Institute
and
Deloitte & Touche Professor of Taxation
J. L. Rotman School of Management
University of Toronto

Steve Murphy
Research Economist
Policy and Economic Analysis Program
Institute for Policy Analysis
University of Toronto

Peter Nicholson
Deputy Chief of Staff, Policy
Prime Minister's Office

Dale Orr
Managing Director
Canadian Macroeconomic Services
Global Insight Inc.

Ellen Russell
Senior Research Economist
Canadian Centre for Policy Alternatives

Donald Savoie
Canada Research Chair in
Public Administration and Governance
Université de Moncton

Adrienne Warren
Senior Economist and Manager
Bank of Nova Scotia

Mary Webb
Senior Economist and Manager
Bank of Nova Scotia

Tom Wilson
Senior Adviser
Institute for Policy Analysis
and
Professor Emeritus of Economics
University of Toronto

United States

Douglas Holtz-Eakin
Director
Congressional Budget Office

Elizabeth Robinson
Deputy Director
Congressional Budget Office

Europe/UK

Tamin Bayoumi
Division Chief
North American Division
Western Hemisphere Department
International Monetary Fund

Simon Brooks
Director of Fiscal and Macroeconomic Policy
UK Treasury

Daniele Franco
Director – Public Finance Directorate
Research Department
Banca D’Italia

Val Koromzay
Director – Country Studies Branch
Economics Department
Organisation for Economic Co-operation and Development

Sandro Momigliano
Deputy Director
Research Department
Banca D’Italia

Martin Muhliesen
Deputy Division Chief
North American Division
Western Hemisphere Department
International Monetary Fund

Robert Price
Head – Monetary and Fiscal Policy Division
Economics Department
Organisation for Economic Co-operation and Development

Christopher Towes
Senior Advisor
Western Hemisphere Department
International Monetary Fund

* Former Deputy Minister, Finance Canada.

## Appendix 2-A Recommendations of the Ernst & Young Report

In 1994, the federal government engaged consulting firm, Ernst & Young to review the procedures for and assess the accuracy of its fiscal forecasts. The E & Y report, Review of the Forecasting Accuracy and Methods of the Department of Finance, made 29 recommendations designed to help the Finance Department improve its forecasting accuracy. The following are the recommendations made in the report under five separate categories.

Forecasting Methodologies

1. The Department of Finance should enhance the personal and corporate income tax revenue forecasts by explicitly incorporating micro-simulation models into the fiscal forecasting methodology.

2. The Department’s economic forecasters should specifically estimate corporate profits in a separate model, and use these results in the forecasting process.

3. The Department of Finance should build supplementary fiscal models to analyze and quantify the impact of structural shifts in the economy and specific aspects of the tax system that could have a major influence on tax revenues. The study team suggests prudent capital investment in the development of specific supplementary models to deal with these issues.

4. A detailed external review of the Department’s macroeconomic model should be undertaken, and a policy of periodic external reviews should be implemented.

5. The Department of Finance should publish the national accounts version of its fiscal forecasts, as well as the reconciliation between its national accounts forecasts and its public accounts forecasts.

Data Inputs

6. Statistics Canada should dedicate additional resources to more accurately capturing data for key components of gross domestic product, especially personal income and corporate profits.

7. The federal government should make changes to the accounting for tax revenues to mitigate year-end adjustment issues.

8. The entries for expenditures in the government’s spending accounts and the adjusting entries for changes in the valuation of government assets should be recorded at regular intervals during the year, perhaps quarterly, to minimize year-end surprises and to provide better information for tracking purposes during the year.

9. Revenue Canada should gather additional information on GST collections through modification of the GST return. This recommendation does not intend the collection of data not already calculated by vendors when preparing their GST remittance.

10. Revenue Canada should complete its database on industrial classification of vendors subject to the GST. This complete file would assist with an understanding of GST flows at the industry level, which should improve both forecasting and subsequent tracking.

11. Statistics Canada should coordinate corporation income tax data collection with Revenue Canada and the Department of Finance:

• to avoid unnecessary duplication;
• to ensure consistency among various data sources;
• to improve the timeliness and quality of the corporate income tax data; and
• to provide additional detail on companies in profit and loss positions, and their stocks of unutilized deductions and losses for tax purposes.

Forecast and Budget Process

12. The emphasis for forecasting and budget-making in the future should more evenly balance short-term and medium-term objectives.

13. The economic forecast included in the budget should reflect the "most-likely" short-term and medium-term prospects for the economy.

14. A medium-term fiscal projection based on the assumption of no policy changes should be included in the budget.

15. The fiscal forecast should be prepared against a back-drop of clearly enunciated budget-making rules.

16. The short-term and medium-term fiscal forecasts should intentionally tend to a cautious assessment of the likelihood of improved fiscal performance (the "prudent forecaster rule").

17. A common reporting scheme for the Department’s economic and fiscal forecasts should be adopted in all budgets.

Tracking Against Forecasts

18. The Department of Finance should prepare and publicly release regular mid-year updates on performance versus the budgetary economic and fiscal forecasts for the current year, and updated forecasts for the medium-term outlook.

19. When the economic and fiscal situation has developed significantly at variance from the budget forecasts, the Minister should report this to Parliament and the public as soon as possible. This report should outline the reasons for the deviations from the budget forecast, and present the government’s intentions for reacting to the situation.

20. The budget forecasts for the various kinds of tax revenues should be disaggregated by key dimensions (for example, industry sectors and geographic regions) to assist with effective tracking.

21. An analytical group should be created at Revenue Canada with the responsibility for tracking the collections of the various revenue sources against budget forecasts, reporting observed variances from forecasts on a regular basis, and analyzing the reasons for these variances. As well, this analytical group would assist in the revenue forecasting efforts of the Department of Finance.

22-
25. The analytical group:

• would collaborate with the Department of Finance in analyzing major trends affecting revenue collections;
• should take over the monitoring of receivables. The management and administration functions associated with receivables would remain in regional and local Revenue Canada offices;
• should have responsibility for quantifying the revenue impact of any ruling requests or potential change in administrative policy; and
• should cooperate with other data collection agencies to ensure efficiency in data collection, and to avoid duplication in data collection activity.

26. When significant and complex changes to the tax structure are made, such as the implementation of the GST, special monitoring efforts should occur.

Institutional Considerations

27. The Department of Finance should establish a mechanism to increase the "distance" between the economic and fiscal forecasts presented in the budget and the political process. However, the implementation of this recommendation should in no way obscure the responsibility of the Minister of Finance and the government for delivering on the fiscal plans set out in the budget.

28. To achieve greater distance between its economic and fiscal forecasts presented in the budget and the political process, the House of Commons Finance Committee should hold public hearings on the government’s annual mid-year forecast update. As part of this process, the Committee should engage an independent review panel to provide a thorough objective critique of the forecasts, which would be released publicly.

29. A Steering Group should be created, with officials from the Economic Analysis and Fiscal Forecasting and Tax Policy and Legislation Divisions of the Department of Finance, and the Analytical Group from Revenue Canada.

## Appendix 2-B Evolution of Budget Forecast Procedures

In September 1994, Ernst & Young released their final report, Review of the Forecasting Accuracy and Methods of the Department of Finance, which contained recommendations for how the federal government could improve its forecasting performance. Since that report, budget forecasting procedures have evolved over time, incorporating some of these recommendations as well as others made through repeated consultation with private sector economists (specifically, consultations at the time of the 1999 Economic and Fiscal Update) and the House of Commons Standing Committee on Finance.

1994 Budget

The following changes to the budget forecasting process were implemented in the 1994 Budget, prior to the release of the Ernst & Young report.

• The government began using economic assumptions below the middle of the range of private sector forecasts in developing the budget plan.
• The government began using a two-year horizon for budget decisions rather than a five-year time horizon.
• A fiscal target equivalent to a 3% deficit/GDP ratio by 1996-97 was set.
• A contingency reserve for the one-year-ahead and two-year-ahead forecasts was included.

1995 Budget

The 1995 Budget represented the first budget following the Ernst & Young report. It is stated in Annex 1 of that budget that half of the 29 recommendations made by Ernst & Young were included in (or incorporated by) the 1995 Budget. In particular:

• The government released its first public Economic and Fiscal Update, as recommended by the Ernst & Young report, in October of 1994 as a means for updating the public on the economic state of the country and the resulting fiscal situation and outlook.
• Prudence was built into the economic assumptions (via interest rates). It is explicitly stated that the economic assumptions underlying the fiscal projections "…are deliberately biased towards prudence." The government’s assumptions for the key economic variables affecting the deficit were more purposely cautious than the average of the 18 private sector firms.
• A contingency reserve was explicitly included for the one- and two-year-ahead forecasts of $2.5 billion and$3 billion respectively.
• The government’s fiscal forecast was prepared using a fiscal target equivalent to a 3% deficit/GDP ratio by 1996-97.

1996 Budget

• The 1996 Budget explicitly stated that prudence factors of 50 basis points in 1996 and 80 basis points in 1997 were applied to the average of private sector forecasters’ projections for interest
rates. [66] That is, interest rate assumptions were set 50 and 80 basis points higher than forecast by the economists surveyed. These assumptions translated into a more prudent outlook for GDP growth and inflation as well.
• No changes were made to the contingency reserve.
• In the 1995 Economic and Fiscal Update, the government outlined a more general fiscal anchor of a two-year rolling deficit target. The target set in the 1996 Budget was equivalent to a 2% deficit/GDP ratio for 1997-98.

1997 Budget

• The 1997 Budget explicitly stated that prudence factors of 80 basis points for short-term rates and 50 basis points for long-term rates would be added to the average of private sector forecasters’ projections for interest rates in  1997 and 1998.[67] These assumptions translated into a more prudent outlook for GDP growth and inflation as well.
• The contingency reserve for the one-year-ahead forecast was increased to $3 billion. • The fiscal target set in the 1997 Budget was equivalent to a 1% deficit/GDP ratio for 1998-99. 1998 Budget • The 1998 Budget explicitly stated that prudence factors of 80 basis points for short-term rates and 50 basis points for long-term rates would be added to the average of private sector forecasters’ projections for interest rates in 1998. Prudence factors for 1999 were set 100 basis points higher than the average of private sector forecasters’ projections.[68] These assumptions translated into a more prudent outlook for GDP growth and inflation as well. • No changes were made to the contingency reserve. • The fiscal anchor was modified to target balanced budgets within the rolling two-year horizon. 1999 Budget • Prudence factors of 70 basis points for short-term rates and 50 basis points for long-term rates were added to the average of private sector forecasters’ projections for interest rates in 1999 and 2000.[69] • No changes were made to the contingency reserve. • The fiscal anchor was revised to target balanced budgets or better within the rolling two-year horizon. 1999 Economic and Fiscal Update In the spring and fall of 1999, the Department of Finance conducted an unprecedented consultation process with the chief economists of Canada’s major chartered banks and four leading forecasting firms. The purpose was to agree upon a set of economic assumptions for budget planning purposes and engage the four forecasting firms to prepare five-year fiscal projections based on the agreed upon economic assumptions (assuming no change to tax and spending policies). As a result of these consultations, changes were made to the government’s forecasting methods and assumptions. Specifically, that: • Five-year time horizons be used for the purposes of public debate, although budget decisions still be made on a rolling two-year horizon. • The average of private sector economic forecasts continue to be used for budget planning purposes, but that any prudence, in addition to the contingency reserve – which until 1999 had been embedded into the economic assumptions – be explicitly shown. The 1999 Economic and Fiscal Update represented the first Update to: • Include five-year status quo fiscal projections for public debate. • Use the fiscal projections from the four leading forecasting firms. • Explicitly show economic prudence and extend it to the medium term. 2000 Budget • Fiscal forecasts continued to be based on the average of private-sector forecasters’ economic projections, but prudence factors were not built in. • Economic prudence of$1 billion in the one-year-ahead forecast and $2 billion in the two-year-ahead forecast were explicitly shown. • The contingency reserve remained unchanged. • The fiscal anchor (balanced budgets or better) remained unchanged. 2001 Budget • Fiscal forecasts continued to be based on the average of private sector forecasters’ economic projections.[70] Average medium-term economic projections were also included, although the number of private sector respondents was considerably lower.[71] • Five-year status quo fiscal projections were published and did not include the impact of policy initiatives, the contingency reserve and economic prudence. These projections were based on the private sector survey in October 2001. • Budget decisions were again made using a rolling two-year approach to budgeting and a fiscal target of balanced budgets over that period. In contrast to the five-year projections, the economic assumptions used were adjusted from the October 2001 survey based on consultations with private sector economists in December 2001. • The size of the contingency reserve and economic prudence was smaller than in previous budgets. 2003 Budget • Fiscal forecasts continued to be based on the average of private sector forecasters’ economic projections. • The size of the contingency reserve and economic prudence was restored to pre-Budget 2001 levels. • The fiscal anchor remained unchanged. • As recommended by the Auditor General, the government adopted full accrual accounting as its accounting standard. The fiscal projections in the 2003 Budget, as well as fiscal results back to 1993-94, were presented on a full accrual basis of accounting. 2004 Budget • Fiscal forecasts continued to be based on the average of private sector forecasters’ economic projections. • The size of the contingency reserve and economic prudence remained unchanged. • The fiscal anchor remained unchanged. ## Appendix 2-C Prudence Included in Fiscal Forecasts Since 1994 Summary Two forms of prudence have been incorporated into the fiscal framework since 1994. The contingency reserve has been the predominant and most visible measure of prudence included in the fiscal plan. Every Budget and Economic and Fiscal Update (EFU) since 1994 has included a contingency reserve, although the exact amount has varied. While economic prudence has also been an ongoing feature of the fiscal framework since 1994, the form of the economic prudence has evolved over time. Further, although the amount of prudence has varied from year to year, it has tended to increase with the length of the forecasting horizon, reflecting the greater degree of uncertainty in the later years of the forecast. Budgets Table 1 presents the prudence used in each Budget since 1994. In the table, Year 1 corresponds to the fiscal year that is ending as the Budget is tabled. For example, in the 2004 Budget, Year 1 represents 2003-04, while Years 2 and 3 represent 2004-05 and 2005-06. The table shows that prior to the 1999 Budget, Year 1 would generally not include a contingency reserve. Since then, however, each Budget has included some amount for the contingency reserve for Year 1, consistent with the government’s Debt Repayment Plan. That is, if the reserve was not used, it would automatically be used to pay down nominal debt. Although contingency reserves of less than$3 billion were common up to 1996, the $3 billion contingency reserve became the norm from 1997 onward, except in years where exceptional circumstances forced a reduction of the reserve. From the 1994 Budget to the 1999 Budget, economic prudence was embedded in the revenue and expenditure forecasts through the use of conservative economic assumptions. Interest rates were assumed to be slightly higher than the private sector average and GDP growth was assumed to be slightly lower (see Appendix 2-B for details). The fiscal impact of modifying these assumptions was usually not published. However, each document typically published a sensitivity analysis table, which provided "rules of thumb" regarding the fiscal impact of a 100-basis-point increase in interest rates or a 1-percentage-point reduction in economic growth relative to the budget assumptions. This allowed the reader to make a rough estimate of the fiscal impact of the "prudent" assumptions. Starting with the 1999 fall Update, the amount of economic prudence was explicitly shown as a separate item in the fiscal framework, and revenue and expenditure forecasts were based on the true average of private sector economic forecasts, without any prudence adjustments. This decision was made pursuant to a recommendation by the private sector economists, with whom the Minister began consulting on a regular basis in 1999. In the December 2001 Budget, in the aftermath of the September 11 terrorist attacks, the government did not set aside any economic prudence and used a portion of the contingency reserve in order to fund new security measures. However, the government indicated its intention to rebuild the normal contingency reserve and economic prudence as soon as possible, which was done in the 2002 fall Update. Table 1: Prudence Included in Budgets from 1994 to 2004 Amount of Prudence Document Type of Prudence Year 1 Year 2 ($ billions) Budget 1994 Contingency Reserve 0.0 2.4 3.0 February 1994 Economic Prudence 0.0 * * Budget 1995 Contingency Reserve 0.0 2.5 3.0 February 1995 Economic Prudence1 0.0 * * Budget 1996 Contingency Reserve 2.5 2.5 3.0 March 1996 Economic Prudence 0.0 * * Budget 1997 Contingency Reserve 0.0 3.0 3.0 February 1997 Economic Prudence 0.0 * * Budget 1998 Contingency Reserve 0.0 3.0 3.0 February 1998 Economic Prudence 0.0 * * Budget 1999 Contingency Reserve 3.0 3.0 3.0 February 1999 Economic Prudence 0.0 * * Budget 2000 Contingency Reserve 3.0 3.0 3.0 February 2000 Economic Prudence 0.0 1.0 2.0 Budget 2001 Contingency Reserve 1.5 2.0 2.5 December 2001 Economic Prudence 0.0 0.0 0.0 Budget 2003 Contingency Reserve 3.0 3.0 3.0 February 2003 Economic Prudence 0.0 1.0 2.0 Budget 2004 Contingency Reserve 1.9 3.0 3.0 March 2004 Economic Prudence 0.0 1.0 1.0 * Up to the 1999 Fall Update, the economic prudence was built into the fiscal framework through the use of more conservative economic assumptions than the private sector average. The fiscal impact of using these "prudent" assumptions was usually not published. 1The 1995 Budget explicitly showed that the fiscal impact of the "prudent" economic assumptions amounted to $1.3 billion in 1995-96 (Year 2) and$2.5 billion in 1996-97 (Year 3).

Fall Updates

Table 2 presents the prudence used in each EFU since 1999. Economic prudence has tended to increase with the length of the forecasting horizon, reflecting the greater degree of uncertainty in the later years of the forecast. The profile of economic prudence embedded in the Updates since 1999 has been consistent with recommendations of the Minister’s advisory group of private sector economists.

Table 2: Prudence Included in Fall Updates from 1999 to 2003

Amount of Prudence  Document  Type of Prudence Year 1 Year 2 Year 3 Year 4 Year 5 ($billions) Fall Update 1999 Contingency Reserve 3.0 3.0 3.0 3.0 3.0 3.0 November 1999 Economic Prudence 0.0 2.0 3.0 3.0 3.5 4.0 Economic Statement and Budget Update 2000 Contingency Reserve 0.0 3.0 3.0 3.0 3.0 3.0 October 2000 Economic Prudence 0.0 1.0 2.0 3.0 3.5 4.0 Economic Update 2001 Contingency Reserve 3.0 3.0 May 2001 Economic Prudence 1.0 2.0 Fall Update 2002 Contingency Reserve 3.0 3.0 3.0 3.0 3.0 3.0 October 2002 Economic Prudence 0.0 1.0 2.0 3.0 3.5 4.0 Fall Update 2003 Contingency Reserve 2.3 3.0 3.0 3.0 3.0 3.0 November 2003 Economic Prudence 0.0 0.0 0.0 1.0 3.0 4.0 ## Bibliography Auerbach, Alan (1999). "On the Performance and Use of Government Revenue Forecasts." National Tax Journal, vol. 52, no. 4, p. 765-782. Balassone, Fabrizio and Franco, Daniele. "EMU Fiscal Rules: A New Answer to an Old Question?" Banca d’Italia Fiscal Rules, Roma (2001). Boothe, Paul and Reid, Brad (1998). "Fiscal Prudence and Budgeting in the Medium-Term." Fiscal Policy and Economic Growth, Kingston: John Deutsch Institute and Institute for Policy Analysis. Crippen, Dan (2003). "Countering Uncertainty in Budget Forecasts." OECD Journal in Budgeting, vol. 3, no. 2, p. 139-151. Ernst & Young (1994). "Review of the Forecasting Accuracy and Methods of the Department of Finance." Hermanutz, Derek and Matier, Chris (2000). "Modelling Federal Finances Under Uncertainty." Department of Finance (Canada) Working Paper 2000-02. HM Treasury (2002). "Reforming Britain’s Economic and Financial Policy: Towards Greater Stability. Hampshire: Palgrave. International Monetary Fund (2002). Canada: "Report on the Observance of Standards and Codes – Fiscal Transparency Module." IMF Country Report 02/51. Washington, DC: IMF. International Monetary Fund (2005). "How do Canadian Budget Forecasts Compare With Those of Other Industrial Countries?" IMF Country Report. Washington, DC: IMF. Organisation for Economic Co-operation and Development (2001). "OECD Best Practices for Budget Transparency." Organisation for Economic Co-operation and Development (2004). "Enhancing the Effectiveness of Public Spending: Experience in OECD Countries." Economics Department Working Paper, no. 380. Orr, Dale (2005). "Does the Government Usually Under-Forecast the Fiscal Surplus? Why? Update." Global Insight. Penner, Rudy (2001). "The Uncertainty of Budget Estimates." Business Economics. (July). Policy and Economic Analysis Program (PEAP) and Center for Interuniversity Research and Analysis on Organizations (CIRANO) (2005). "Forecasting Process and Performance of the Department of Finance." Robson, William B.P. (2002). "Saving for Health: Prefunding Health Care for an Older Canada." C.D. Howe Institute Commentary 170. Toronto: C.D. Howe Institute. October. Rudin, Jeremy (2003). Unpublished notes for presentation. Scarth, William and Jackson, Harriet (1998). "The Target Debt-to-GDP Ratio: How Big Should It Be? And How Quickly Should We Approach It?" Fiscal Policy and Economic Growth, Kingston: John Deutsch Institute and Institute for Policy Analysis. Stanford, Jim (2005). "Submission to Tim O’Neill, Special Advisor to Finance Minister Ralph Goodale. "Review of Federal Fiscal Forecasts and Budget Planning Procedures." February. Strauch, Rolf, Hallerberg, Mark and von Hagan, Jürgen (2004). "Budgetary Forecasts in Europe: The Track Record of Stability & Convergence Programmes." European Central Bank Working Paper Series no. 307. NOTES: 1 This is a problem a number of other OECD countries would prefer to have, rather than the deficit situation in which they find themselves. [Return] 2 Although the accounting rules are not changeable, there are ways to accommodate initiatives other than debt reduction (even after the year-end) that are consistent with the rules. This is discussed in Section 4.3 below. [Return] 3 Although, arguably, the energy rebate provided in the Economic Statement and Budget Update in October 2000, was the result of there being unanticipated surpluses available. [Return] 4 Only the forecasting firms (currently PEAP, Conference Board of Canada, Global Insight and Centre for Spatial Economics) produce detailed fiscal forecasts and then only for the Economic and Fiscal Update in the fall. [Return] 5 In 2003-04, on a status quo basis, direct program spending was$1.8 billion lower than estimated while fiscal transfers were $4 billion lower. [Return] 6 The arithmetic of this argument is straightforward: since the federal debt is now about 40% of GDP, then any deficit smaller than 40% of the annual change in GDP would still permit the debt ratio to fall. In 2003-04, for example, the debt ratio fell by 3 percentage points to 41.1% from 44.1%. The numerator (the debt) fell to about$502 billion from $511 billion a year earlier. At the same time, the denominator (nominal GDP) increased to$1,219 billion in 2003 from $1,158 billion a year earlier, an increase of$61 billion. Even if the government had run a deficit of $9.1 billion in 2003-04 rather than a surplus of the same magnitude, the debt ratio would have fallen by 1.5 percentage points to 42.6%. [Return] Current survey invitees: Bank of Montreal, Laurentian Bank of Canada, UBS Warburg, CDP Capital, Centre for Spatial Economics, CIBC World Markets, Conference Board of Canada, Mouvement des caisses Desjardins, Global Insight, Economap Strategic Economic Advisors, J.P. Morgan, Financière Banque Nationale, Merrill Lynch, Nesbitt Burns, Royal Bank of Canada, Scotiabank, TD Canada Trust, University of Toronto (PEAP). [Return] Items marked with an asterisk (*) have been collected since 1994; the full set listed has only been collected since September 1999. [Return] See, for example, Rudy Penner (2001) and Ernst & Young (1994). [Return] 10 Strauch, Rolf, Hallerberg, Mark and von Hagan, Jürgen (2004). [Return] 11 This result appears to contradict results from the aggregate analysis, which found that, overall, forecast errors were symmetrically distributed around zero. The aggregate result is perhaps misleading, as there are differences in the behaviour of forecast errors across countries. [Return] 12 Auerbach considers the average revision at each forecast horizon, as well as the sum of these average revisions over all horizons. [Return] 13 Only the OMB revisions for the "pre-Clinton" period show statistically significant bias. [Return] 14 Crippen (2003), p.51. [Return] 15 The historical period used for the 2005-2014 budget projections was 1981 to 2003. [Return] 16 Boothe, Paul and Reid, Brad (1998). [Return] 17 The four levels of prudence considered are: zero,$3 billion, $6 billion, and$9 billion. The three budget rules considered are: a balanced budget in every year, a balanced budget over a two-year period, and a balanced budget over a four-year period. [Return]

18 Hermanutz, Derek and Matier, Chris (2000). [Return]

19 E.g. balancing the budget over a two-year rolling horizon, non-permanent shocks to GDP and partial debt rollover each year. [Return]

20 This will be discussed more fully below. [Return]

21 For one-year-ahead forecasts, the time difference is 18-20 months. [Return]

22 We have adopted for the most part the terminology used in the PEAP – CIRANO study for the difference between projected and actual values. They prefer "differences" to "errors," which is the more commonly used term. However, the two terms can be used interchangeably and the latter will also be used, albeit sparingly. [Return]

23 The results shown for PEAP – CIRANO 1, Orr and Stanford are inclusive of policy initiatives.[Return]

24 In those years, as mentioned in Section 2, Finance took the average of the private sector forecasts and then adjusted them (down for growth, up for interest rates) to make the fiscal projections more conservative. [Return]

25 To avoid confusion about terminology, an under-forecast, whether of revenues, expenditures or budget balance, will be entered as a negative value in the tables. However, the term under-forecast can have different meanings depending on the variable. An under-forecast for the budget balance implies that an actual deficit was smaller or a surplus larger (hence better) than predicted. An under-forecast of revenues – revenues ended up higher than projected – would contribute to the budget balance being better than expected. However, an under-forecast of expenditures – spending was greater than anticipated – would contribute to the budget balance being worse than projected. [Return]

26 Inclusion of other countries’ in-year policy initiatives conceivably could change Canada’s relative positioning if one or more had larger in-year initiatives than Canada. [Return]

27 However, the IMF does note, more generally, that "the evidence that forecasts tend to be more conservative in the presence of macroeconomic and fiscal volatility is relatively strong" and suggests the need for further research on this issue. [Return]

28 Over the years Statistics Canada has re-based and redefined elements of the National Accounts in ways that forecasters could not have anticipated. Using the most recent Statistics Canada data therefore leads to invalid estimates of economic forecast errors. [Return]

29 This is not meant as an implied criticism of the IMF as they were using the official budget document figures on economic assumptions for all the countries examined. Their treatment of Canada was consistent with that. [Return]

30 Op. cit. [Return]

31 The revenue/GDP ratio for individual revenue components and for total revenue is functionally equivalent to the revenue sensitivities used by Finance. They both indicate the impact on revenue of a $1 change in nominal GDP. [Return] 32 A correlation coefficient of 1.00 means that the two variables move perfectly together while 0.00 implies no link at all. A value over 0.25 is noteworthy. [Return] 33 There have also been significant revisions over time to the nominal levels of the expenditure and income components of GDP as well as to the overall level, and these could potentially have had their own independent impacts on the fiscal forecast differences. Note, however, that income and expenditure components are forecasted by the Department of Finance itself and not by the private sector forecasters. The components are made to total to the aggregate GDP obtained from the private sector. [Return] 34 Op. cit., p. 26. [Return] 35 Op. cit. p. 5. [Return] 36 Op. cit. p. 6. [Return] 37 Finance officials argue that they are not able to reach this total in their calculations from the forecasts made over that period. [Return] 38 Coming at the end of a period of sluggish labour market behaviour, the over-forecast is understandable. [Return] 39 Op. cit. p.31. [Return] 40 Op. cit. p.?. [Return] 41 The other three are stability through constrained [policy] discretion, credibility through sound, long-term policies and credibility through pre-commitment. [Return] 42 Defined as differences in access to data and information, acquired knowledge and expertise. [Return] 43 The option of choosing to balance the budget over the economic cycle is discussed in section 4.3 below. [Return] 44 For example, see Auerbach. [Return] 45 In the consultations, the observation was made that the revisions may simply be the result of the economy’s going through its expansion phase, in which upside economic surprises are typical. If true, the reverse should happen in downturns, i.e. Statistics Canada would be revising GDP down. That did not occur in the 2001-02 period and it is noteworthy that the economy continued to outperform expectations. [Return] 46 This is generally referred to as the "golden rule" and is applicable in several other European countries including the UK. [Return] 47 However, it is interesting to note that in public option polls, routinely over 40% of Canadians still think that the federal government has a deficit. [Return] 48 Boothe and Reid (1998) estimate that$6 billion to $9 billion in prudence would be required to virtually ensure no deficit at any time in the business cycle. [Return] 49Automatic stabilizers tend to moderate swings in the economy by increasing household spending capacity during downturns (tax withdrawals fall and employment insurance benefits rise) and reducing it in upturns. Because the PIT and EI adjustments are built in, they work automatically. [Return] 50 For a concise history of the debate among economists over the wisdom of balanced budgets, see Balassone and Franco (2001). [Return] 51 In unpublished remarks, Rudin (2003) makes the point that the Finance Department has estimated the cyclically adjusted balance since the 1980s but that the official estimate for some years has changed significantly over time as data revisions occur. This renders the initial estimate of dubious value. [Return] 52 Obviously, if the surplus target were set high enough (say$10 billion) it would almost certainly prevent deficits. [Return]

53 See, for example, the comments in the summary of consultations (Section 2) as well as Auerbach (p. 779). [Return]

54 The British government currently is, in its fiscal projections, using fiscal year 2006-07 as the end of this economic cycle in the UK and the start of the next one. [Return]

55 Of course, in the long term, it opens up more room for those same spending and tax initiatives. Therefore, this should be more correctly viewed as the trade-off between current and future consumption of government services and/or current and future tax reductions. [Return]

56 In the case of the budget surplus option, the debt reduction could either be a down payment on ensuring target achievement or it could be an adjustment to that target. [Return]

57 The no-deficit rule currently has an explicit debt reduction component (the unused portion of the contingency reserve) but that is not an inherent provision of such a rule. Explicitly targeting a surplus, on average, over the cycle implies allocation to debt reduction (or a special fund to be drawn down in the future under specific conditions). [Return]

58 See Scarth and Jackson (1998), for example. [Return]

59 Trend (or capacity) growth refers to a pace of economic expansion that matches the growth of economy’s capacity, which comprises (approximately) the rate of labour force growth (1%) and long-term productivity growth (estimated at 2%). That yields a real rate of 3% to which is added expected inflation (estimated 2%) to yield 5% nominal growth. [Return]

60 The surplus target in the positive balance rule or the contingency reserve of $3 billion in the current no-deficit rule. [Return] 61 If nominal GDP grows at 5% annually (trend rate) the nominal debt has to grow at the same pace. The size of the increase in debt will be 25% of the size of the change in GDP or 1.25% for a 5% GDP growth rate. [Return] 62 Robson (2002) estimates that the impact of population aging on health care costs will be to "raise their growth rate…by about one percent annually above the path that inflation and increased utilization would otherwise produce." (p. 1) [Return] 63By about$320,000 per year based on the provision that has been made for hiring the outside economists. [Return]

64 While made before the 2005 Budget, these commitments had a significant impact on other commitments introduced in the Budget. [Return]

65 The IMF (2005) study takes particular note of the unique role of outside forecasters in the Canadian process. In some countries, private sector analysts are sought for their expertise in the budget process but their input is not formalized. [Return]

66 The private sector average is based on 20 respondents for 1996 and 18 respondents for 1997.[Return]

67 The private sector average is based on 21 respondents for 1997 and 18 respondents for 1998. [Return]

68 The private sector average is based on 18 respondents for 1998 and 1999. [Return]

69 The private sector average is based on 19 respondents for 1999 and 2000. [Return]

70 The private sector average was based on 19 respondents for 2001 and 2002. [Return]

71 The number of respondents for 2006 and 2007 were 9 and 2 respectively. [Return]