Skip to content | Skip to institutional links

Common menu bar links

Institutional links

Archived

This Web page has been archived on the Web.

- Table of Contents - Previous - Next -

Annex 2
Perspective: Long-Term Track Record of Fiscal Forecasting

Related Document:

Details on Differences by Fiscal Year

Introduction

The Government of Canada provides annual reconciliations between its fiscal forecasts and final outcomes in the Annual Financial Report of the Government of Canada and in the budget documents. However, in a review of the Government's forecasting accuracy,^[1] Dr. Tim O'Neill recommended that the Government provide a longer-term perspective on its fiscal-forecasting record. This annex provides a perspective on the Government's fiscal-forecasting record from 1994-95 to 2004-05.

Specifically, for this period it compares the fiscal estimate for the upcoming fiscal year, as well as the updated in-year estimate, to the final outcome as published in the Annual Financial Report of the Government of Canada for that year. The estimate for the upcoming fiscal year is traditionally provided in the budget tabled just before the beginning of that fiscal year, while the updated in-year estimate is provided in the budget tabled near the end of that fiscal year. As there were only four fall Economic and Fiscal Updates (in 1999, 2002, 2003 and 2004), these are excluded from the summary analysis in this annex. A detailed evolution of the fiscal forecasts over the period 1994-95 to 2004-05 by fiscal year is provided in Perspective: Long-Term Track Record of Fiscal Forecasting-Details on Differences by Fiscal Year, which is available on the Department of Finance Canada website at www.fin.gc.ca.

The Federal Fiscal-Forecasting Process

Since 1994, the economic forecasts used in the budget have been based on the average of private sector economic forecasts. About 20 forecasters are surveyed following the release of the quarterly National Income and Expenditure Accounts by Statistics Canada. The average of the forecasts for a number of key economic variables is used to derive the fiscal projections by major component on a National Accounts basis of accounting, consistent with the accounting framework used by Statistics Canada. These projections are then converted to a Public Accounts basis.

On a historical basis, Statistics Canada provides the Department of Finance Canada with reconciliations between its estimates and the fiscal results presented in the Public Accounts of Canada and the monthly Fiscal Monitor. For the current fiscal year and outer years, the conversion is done by the Department of Finance. In most cases, the adjustments are relatively straightforward, reflecting conceptual and classification differences. However, for the current fiscal year, adjustments are required to take into account more current information provided in the monthly Fiscal Monitor. A description of these adjustments can be found in the Economic and Fiscal Update for 2004 and 2005.

The forecast of the budgetary balance is developed on a bottom-up basis. Individual components of budgetary revenues are based on their applicable tax bases and tax rates. For example, personal income tax revenues are based on the forecast of taxable income, deductions and exemptions, and the applicable tax rates at different levels of income, based on the Income Tax Act's definition of taxable income. Budgetary revenues are sensitive to changes in the composition of income, lags in the receipt of data and cyclical factors. As a result, current economic assumptions may not be fully reflective of the fiscal outcome. In fact, it has been argued that current budgetary results, especially those for budgetary revenues, are the best contemporaneous and leading indicator of the economy rather than the other way around.

For program spending/expenses, some components are driven by economic assumptions (employment insurance benefits), demographics (elderly benefits) and legislation (the various federal transfers in support of health and other social programs and equalization). Direct program spending/expenses, which include the operating costs of the federal government, are based on the Main Estimates (tabled by the President of the Treasury Board), adjusted to include the impact of announced policy initiatives not included in the Estimates. Adjustments are also made to include provisions for liabilities that the Government may incur during the course of the year, such as the costs associated with employee contract settlements, other employee benefits, court decisions, natural disasters, and changes in environmental and Aboriginal liabilities, among others. As departments cannot overspend their appropriations, they usually commit less than what they were appropriated. An estimate of this "lapse" in appropriations is also included in the forecasts based primarily on historical experience. Although these adjustments are reviewed throughout the year, lack of current information requires professional judgment in determining the estimated amounts. Public debt charges are based on the outstanding stock of interest-bearing debt, the rollover of this debt and forecast interest rates. In deriving these individual estimates, consideration is given to risks based on historical experience and current developments.

The final outcome for the previous fiscal year (which normally becomes available in late September/early October of the current fiscal year) provides important information on the effective tax rates for the various revenue components, the lapse and adjustments to liabilities. In addition, with the release of the first quarter National Income and Expenditure Accounts in late May, Statistics Canada also publishes revised estimates of economic growth and its components for previous years. The final budgetary outcome for the previous fiscal year, coupled with the revised economic data, forms the new basis for the current-year estimates, often resulting in a reassessment of average effective tax rates, lapses and provisions for liabilities. This information, combined with current monthly financial results (usually the financial results to the end of August or September), often results in significant adjustments to the individual components of the budgetary balance and the bottom-line number itself.

The fiscal forecast for the current fiscal year is updated in the budget presented near the end of the fiscal year, primarily reflecting current monthly developments to the end of December. Although this update usually occurs in February-a month before the end of the fiscal year-there are still significant developments in the final quarter of the fiscal year, as well as accrual accounting adjustments after March 31, that could have significant impacts on the final outcome. For further details on these adjustments, see Annex 1.

Over the period under review, the sensitivity of the budgetary balances to economic developments has nearly doubled. In the 1994 budget, a 1-per-cent increase in nominal income was estimated to improve the budgetary balance by $1.3 billion in the first year. In the February 2005 budget, the same change to nominal income was estimated to improve the budgetary balance by $2.5 billion, reflecting the growth in the economy over this time period. In contrast, the sensitivity to interest rate changes has been dampened since the Government has lengthened the term to maturity of its market debt. In the 1994 budget, a 100-basis-point reduction in all interest rates improved the budgetary balance by $1.7 billion in the first year. In the 2005 budget, this was reduced to $1.0 billion.

The Track Record: 1994-95 to 2004-05-Comparison of the First Estimate to the Final Outcome

Table 1 compares the first estimate of the budgetary balance as contained in the budget immediately preceding the fiscal year to the final audited results as originally published.

The first estimate was underestimated by an average of $5.6 billion over the 1994-95 to 2004-05 period. In four years (1996-97, 1997-98, 1999-00, and 2000-01), the difference was substantially above the average. Excluding these years, the average difference was just $1.3 billion. In 2 of the 11 years under review, the final outcome was lower than the first estimate.

Table 1
Evolution of Fiscal Forecasts: First Estimate to Outcome

Fiscal year	First estimate	Outcome¹	Difference


	(billions of dollars)
1994-95	-39.7	-37.5	2.2
1995-96	-30.2	-28.6	1.7
1996-97	-21.8	-8.9	12.9
1997-98	-14.0	3.5	17.5
1998-99	3.0	2.9	-0.1
1999-00	3.0	12.3	9.3
2000-01	4.0	17.1	13.1
2001-02	8.3	8.9	0.6
2002-03²	5.0	7.0	2.0
2003-04	4.0	9.1	5.1
2004-05	4.0	1.6	-2.4

Average difference			5.6

¹ As originally published in the Annual Financial Report or the Government of Canada. ² For 2002-03, the first estimate was presented on a modified accrual basis of accounting while the final outcome was presented on a full accrual basis. The first estimate was adjusted to a full accrual basis, using the accrual estimates presented in the final budget estimate.

However, these estimates are not presented on a comparable basis. During the course of the fiscal year, initiatives were announced when it became clear that the budgetary balance would be better than estimated. This has been part of the prudent approach to budget planning: the Government has stated that it would only undertake new initiatives if it was certain it had the resources to do so. As such, funds were released during the course of the fiscal year when it became apparent that the outcome would be better than expected. These new initiatives were subject to the normal parliamentary approval process before any payment could be made.

Table 2 adjusts the first estimate and final outcome to adjust for the fiscal cost of policy initiatives made during the course of the fiscal year, as well as the impact of accounting and technical adjustments that were not anticipated at the time of the first budget estimate. On this basis, the first estimate was underestimated by an average of $10.2 billion over the 1994-95 to 2004-05 period. Just over half of this difference was attributable to budgetary revenues coming in higher than expected, with lower-than-expected program spending/expenses accounting for about one-third of the difference and the remainder due to lower-than-expected public debt charges.

Table 2
Evolution of Fiscal Forecasts: First Estimate to Outcome Adjusted for In-Year Policy Decisions/Technical Adjustments

Fiscal year	First estimate	Outcome¹	Difference	Revenues	Program spending/ expenses	Public debt charges

				Sources of difference²


	(billions of dollars)
1994-95	-39.7	-37.5	2.2	-0.6	3.8	-1.0
1995-96	-30.2	-28.6	1.7	-2.9	2.0	2.6
1996-97	-21.8	-10.6	11.1	2.6	5.6	2.8
1997-98	-14.0	6.8	20.8	14.9	3.3	2.6
1998-99	3.0	8.6	5.5	5.0	-1.5	2.1
1999-00	3.0	18.5	15.5	9.4	5.2	0.9
2000-01	4.0	22.4	18.4	18.2	0.3	-0.1
2001-02	8.3	12.7	4.4	-1.6	2.2	4.0
2002-03³	3.0	13.0	10.1	1.0	8.7	0.6
2003-04	4.0	13.8	9.8	1.5	6.4	1.8
2004-05	4.0	16.3	12.3	11.2	0.1	1.3

Average difference			10.2	5.3	3.3	1.6

Note: Totals may not add due to rounding. ¹ As originally published in the Annual Financial Report of the Government of Canada, adjusted for policy initiatives. 2 A positive number implies an improvement in the fiscal balance. A negative number implies a deterioration in the fiscal balance. ³ The first estimate was presented on a modified accrual basis of accounting while the final outcome was on a full accrual basis. The first estimate was adjusted to a full accrual basis using the accrual estimates presented in the 2003 budget.

Table 3 shows the differences between the first estimate and the final outcome for the various components of revenues and spending/expenses. In 4 of the 11 years (1997-98, 1999-2000, 2000-01 and 2004-05), budgetary revenues were underestimated by an average of $4.9 billion, accounting for over 90 per cent of the difference in revenues and nearly 50 per cent of the difference in the budgetary balance. The differences in these years can largely be explained by changes in the economic forecasts and better-than-expected fiscal outcomes in the previous year-which were not known at the time the first estimate was made. In the other years, differences among the revenue components were largely offsetting, with underestimates in four years and overestimates in three.

Table 3
Policy-Adjusted Budget Balance Forecast Difference by Component: First Estimate to Outcome


	1994-1995	1995-1996	1996-1997	1997-1998	1998-1999	1999-2000	2000-2001	2001-2002	2002-2003	2003-2004	2004-2005

	(billions of dollars)
Budgetary revenues
Personal income tax	-3.2	-0.2	-0.2	4.3	1.5	4.4	7.9	2.2	-1.4	-1.7	2.9
Corporate income tax	1.3	0.5	1.9	6.3	1.1	2.3	4.3	-4.2	-0.9	3.1	3.7
Other income tax	0.0	0.3	0.9	0.9	0.6	0.6	0.8	0.5	0.4	0.0	0.3
Goods and services tax	0.3	-1.0	0.2	2.0	-0.2	1.2	1.9	-0.8	1.0	-1.6	1.2
Excise taxes and duties	0.2	-0.4	-0.1	0.5	-0.3	-0.6	0.6	0.9	1.3	0.0	0.0
Employment insurance premiums	-0.4	-1.2	0.0	-0.3	1.1	0.6	0.6	-0.4	0.3	0.0	0.3
Other revenues	1.2	-0.7	-0.1	1.1	1.2	0.8	1.7	0.1	0.6	1.9	2.8

Total budgetary revenues	-0.6	-2.9	2.6	14.9	5.0	9.4	18.2	-1.6	1.0	1.5	11.2
Program spending/expenses
Elderly benefits	-0.1	-0.2	-0.3	-0.1	-0.1	-0.1	0.1	0.2	-0.1	0.1	-0.1
Employment insurance benefits	-3.5	-0.8	-1.4	-1.7	-0.7	-2.1	-0.6	1.5	-1.4	-0.7	-1.0
Federal support for health and other social programs	0.5	0.3	-0.2	-0.1	0.0	-0.1	0.0	0.0	0.0	-0.1	0.1
Fiscal arrangements	-0.4	-0.5	-0.1	0.6	2.4	0.6	1.2	-0.5	-2.1	-4.0	0.7
Alternative Payments for
Standing Programs			0.0	0.0	0.1	-0.1	-0.1	-0.2	0.2	0.0	-0.1
Direct program spending/expenses	-0.3	-0.8	-3.5	-2.0	-0.1	-3.5	-0.9	-3.1	-5.3	-1.7	0.5

Total program spending/expenses	-3.8	-2.0	-5.6	-3.3	1.5	-5.2	-0.3	-2.2	-8.7	-6.4	0.1
Public debt charges	1.0	-2.6	-2.8	-2.6	-2.1	-0.9	0.1	-4.0	-0.6	-1.8	-1.3
Net impact on budgetary balance	2.2	1.7	11.1	20.8	5.5	15.5	18.4	4.4	10.1	9.8	12.3

Note: Totals may not add due to rounding.

With the exception of 1998-99 and 2004-05, program spending/expenses were overestimated in each year. Lower-than-forecast employment insurance benefits and direct program spending/expenses explain most of the difference in program spending/expenses. The difference in employment insurance benefits primarily reflected an assumption with respect to the relationship between the forecast number of unemployed and the resulting number of beneficiaries. Assumptions with respect to the lapse in departmental appropriations and provisions for liabilities explain most of the difference in direct program spending/expenses. Large prior-year adjustments to equalization entitlements were a major factor in explaining the overall differences in program spending/expenses in 3 of the 11 years. The differences in public debt charges can largely be explained by changes in the interest rate forecasts.

The Track Record: 1994-95 to 2004-05-Comparison of the Final Budget Estimate to the Final Outcome

This section compares the estimate in the budget tabled just before the end of the fiscal year to the final outcome as published in the Annual Financial Report. The comparison is on an adjusted basis, excluding the impact of policy initiatives and accounting/technical changes that occurred after the budget was tabled. In 4 of the 11 years, policy initiatives or accounting/technical adjustments were made after the budget was tabled. In 1997-98, there was a change in accounting for commitments to international financial institutions; in 2000-01 and 2001-02, there were changes to the costs of previously announced budget measures; and in 2004-05, additional support was provided to farmers and there was a change in accounting for the recognition of the liability with respect to the Offshore Revenues Accords.

On an adjusted basis, the fiscal balance was underestimated by an average of $4.6 billion over the 1994-95 to 2004-05 period, with both higher-than-expected revenues and lower-than-expected program spending/expenses contributing about equally to the difference (Table 4). As a result, about half of the difference between the first estimate and the final outcome occurred after the final budget estimate. This underscores the difficulties highlighted in Annex 1, especially with respect to lags in information.

Table 4
Evolution of Fiscal Forecasts: Final Estimate to Outcome Adjusted for In-Year Policy Decisions/Technical Adjustments

Fiscal year	Final estimate	Outcome¹	Difference	Revenues	Program spending/ expenses	Public debt charges

				Sources of difference²


	(billions of dollars)
1994-95	-37.9	-37.5	0.5	-1.7	2.2	0.0
1995-96	-30.2	-28.6	1.7	-0.3	1.8	0.1
1996-97	-19.0	-8.9	10.0	4.7	4.8	0.5
1997-98	0.0	5.0	5.1	5.6	-1.1	0.6
1998-99	3.0	2.9	-0.1	-0.8	0.7	0.0
1999-00	3.0	12.3	9.3	5.7	3.7	-0.1
2000-01³	11.9	17.3	5.4	4.9	0.5	0.1
2001-02⁴	1.5	7.2	5.8	-0.2	4.5	1.5
2002-03	3.0	7.0	4.1	-0.9	5.0	-0.1
2003-04	1.9	9.1	7.2	5.1	2.1	0.0
2004-05	3.0	5.3	2.3	2.6	-0.9	0.6

Average difference			4.6	2.2	2.1	0.3

Note: Totals may not add due to rounding. ¹ As originally published in the Annual Financial Report of the Government of Canada, adjusted for policy initiatives. 2 A positive number implies an improvement in the fiscal balance. A negative number implies a deterioration in the fiscal balance. ³ Final estimate for 2000-01 is from the Economic Statement and Budget Update (October 18, 2000). ⁴ Final estimate for 2001-02 is from the December 2001 budget.

Table 5 shows the differences between the final budget estimate and the final outcome, on an adjusted basis, for the various components of revenues and spending/expenses. Budgetary revenues were overestimated in five years and underestimated in six years. However, the differences in those years in which budgetary revenues were underestimated were considerably larger than when overestimates occurred. The differences among the revenue components are largely random, with the possible exception of other revenues, which were underestimated in 9 of the 11 years.

Direct program spending/expenses account for nearly three-quarters of the overestimates in total program spending/expenses. In only two years was this component underestimated. The differences in direct program spending/expenses are largely due to difficulties in estimating the lapse in departmental appropriation and provisions for liabilities incurred during the year. In contrast, the differences in public debt charges were relatively small, with the exception of 2001-02.

Table 5
Policy-Adjusted Budget Balance Forecast Difference by Component: Final Estimate to Outcome


	1994-1995	1995-1996	1996-1997	1997-1998	1998-1999	1999-2000	2000-2001	2001-2002	2002-2003	2003-2004	2004-2005

	(billions of dollars)
Budgetary revenues
Personal income tax	-0.5	-0.3	0.0	2.4	-1.2	2.9	0.9	2.5	-2.5	1.4	0.2
Corporate income tax	-1.4	1.6	1.2	2.5	-0.4	0.7	-0.3	-1.3	0.3	1.5	1.5
Other income tax	0.0	0.2	0.7	0.6	0.0	0.2	0.8	0.1	0.4	-0.1	0.0
Goods and services tax	0.2	-0.8	1.2	-0.3	0.1	0.7	0.8	-0.8	-0.4	0.6	-0.5
Excise taxes and duties	-0.2	0.0	0.2	-0.1	0.0	0.1	0.9	0.2	0.2	0.2	0.0
Employment insurance premiums	0.0	0.0	0.2	0.1	0.2	-0.1	0.0	0.2	-0.4	0.4	0.2
Other revenues	0.2	-0.7	1.1	0.4	0.4	1.1	1.7	-1.2	1.6	1.3	1.2

Total budgetary revenues	-1.7	-0.3	4.7	5.6	-0.8	5.7	4.9	-0.2	-0.9	5.1	2.6
Program spending/expenses
Elderly benefits	-0.1	-0.2	0.0	0.0	0.0	0.1	0.0	0.1	-0.1	-0.1	-0.1
Employment insurance benefits	-0.5	0.0	-0.7	-0.2	-0.2	-0.4	-0.2	-0.4	-0.5	-0.4	-0.5
Federal support for health and
other social programs	0.2	0.1	-0.1	0.1	0.0	-0.1	0.0	0.0	0.0	0.0	0.0
Fiscal arrangements	-0.8	-0.6	0.1	0.5	0.0	-0.2	0.6	-0.2	-2.4	0.6	0.2
Alternative Payments for
Standing Programs			0.0	0.0	0.1	0.0	-0.1	-0.3	0.2	-0.3	0.0
Direct program spending/expenses	-1.0	-1.1	-4.1	0.7	-0.6	-3.1	-0.8	-3.7	-2.2	-1.9	1.3

Total program spending/expenses	-2.2	-1.8	-4.8	1.1	-0.7	-3.7	-0.5	-4.5	-5.0	-2.1	0.9
Public debt charges	0.0	-0.1	-0.5	-0.6	0.0	0.1	-0.1	-1.5	0.1	0.0	-0.6
Net impact on budgetary balance	0.5	1.7	10.0	5.1	-0.1	9.3	5.4	5.8	4.1	7.2	2.3

Note: Totals may not add due to rounding.

¹ Review of Canadian Federal Fiscal Forecasting: Processes and Systems, Dr. Tim O'Neill (June 2005). This annex is based on the analysis in Dr. O'Neill's report. Some adjustments were made to the data to present original forecasts and reported outcomes on a similar basis. However, the overall results are similar to those in Dr. O'Neill's report. [Return]