In 1994, the budgetary planning process was significantly changed.  An average of private sector economic forecasts for a number of selected economic variables was used for budget planning purposes rather than the Department of Finance’s economic forecast.

The budget planning period was restricted to a short-term focus of two years rather than a medium-term term of five years. The budget included a Contingency Reserve and economic prudence to order to ensure that the stated fiscal targets would be met.  Four private sector economic forecasting firms were contracted to produce fiscal forecasts for the fall update. 

A number of changes to the process have been made since 2006.  This note assesses the various elements of the budgetary process and argues that further changes be made.

The Finance Department Should Stop Using the Average Private Sector Economic Forecasts

Each quarter, the Department of Finance surveys those private sector economists who undertake quarterly economic forecasts.  The average of their forecasts for a number of key economic and financial variables is then used for budget planning. The impetus for this approach were the  recommendations made by Ernst & Young (E&Y) in their 1994 Report, “Review of the Forecasting Accuracy and Methods of the Department of Finance[1]”.

 The government commissioned this Report to address two key questions:

• How accurate have the department’s economic and fiscal forecasts been over the past decade; and,
• How could the department’s forecasting accuracy be improved?

E&Y concluded that the Department of Finance’s economic forecasts were more accurate than any individual private sector forecast, on average, over the 1982-83 to 1992-93 period.  However, E&Y made 29 recommendations that they felt would improve the  forecasts.  One of these recommendations (#27) stated that:

“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.” 

E&Y felt that this could be accomplished by having the House of Commons Standing Committee on Finance engage a panel of independent reviewers to critique the government’s economic and fiscal forecasts or alternatively establish an independent agency to provide the economic and fiscal forecast. Neither of the recommendations was adopted. In fact, they were ignored for obvious political reasons.

Instead, the government at the time decided to use the average of the private sector economic forecasts, rather than those of the Department of Finance. This practice has been in existence to date.  The prime reason for adopting this approach was to focus the debate on the actions required to bring the deficit under control at that time rather than on the economic and fiscal forecasts themselves.

There are a number of issues with respect to this approach.  Currently, the Department of Finance surveys about 15 private sector forecasters, most of which are financial institutions.  In most cases, they focus primarily on the short-term – one to two years.  In Budget 2010, there were only three forecasters (Conference Board of Canada, Global Insight and the University of Toronto) that had large-scale econometric models capable of providing medium-term economic forecasts.  This raises issues of comparability between the short and medium term forecasts.

Most forecasters update their forecasts on a quarterly basis, following the release of the Canadian economic accounts by Statistics Canada (first quarter results published in May, second quarterly results in August, third quarter results in November and fourth quarter and preliminary estimate for year as a whole in February of the following year).  For a February or early March budget, private sector economists would not have time to incorporate the fourth quarter results of the previous year in their forecasts.  This could impact on the fiscal projections, if the fourth quarter results differed from their original forecast, which is very often the case. 

There have been wide variations among private sector forecasters about future economic developments, both in the short and medium term.  At times, the government has taken this into consideration and adjusted the average private sector economic forecasts, based on the Department of Finance’s assessment of future economic developments.  In these cases, it is the Department of Finance’s economic forecast that is used, rather than the average of the private sector economic forecasts.  

The private sector economists are surveyed for only a selective number of aggregate economic and financial indicators: real gross domestic product (GDP) growth; GDP inflation, nominal GDP; the 3-month treasury bill rate; the 10-year government bond rate; the unemployment rate; the  consumer price index; the  exchange rate (US cents/Cdn$); and finally, and U.S. real GDP growth.  The Department of Finance decomposes the forecast values for real and nominal GDP into their  expenditure and income components, respectively.  How this decomposition is made can have a significant on the fiscal projections as different components of income have different effective tax rates.  For example, corporate profits have a much higher average effective tax rate than that personal income. 

 E&Y also stated that the implementation of recommendation #27 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.  However, this is usually not the case as governments have often argued that missing budget targets was due to the private sector economic forecasts being off track.  Unfortunately, the private sector forecasts are discussed in camera with the economists surveyed so that there is no way of knowing who is responsible for what.

As E&Y noted in their report, the Department of Finance’s track record was better than those in the private sector. This should not be surprising. Finance has a branch of about 100 professionals solely dedicated to monitoring and analyzing current economic developments and assessing their impacts on future developments.  The branch is staffed by highly qualified and professional economists, who are recruited directly by the Department of Finance.  They maintain one of the country’s most comprehensive econometric models of the Canadian economy.  They prepare detailed quarterly forecasts of the Canadian economy as well as for Canada’s major trading partners.  They can update their economics quickly if new information becomes available. These forecasts are used internally to brief the Minister of Finance and Cabinet and to assess the private sector economic forecasts.

The Office of the Parliamentary Budget Officer (PBO) also uses the average of the private sector economic forecasts for their fiscal updates.  There is no reason to have two agencies use the same economic forecasts to produce fiscal forecasts.  The Bank of Canada uses its own internal economic forecasts for monetary policy purposes.

The Government should, once again, take full responsibility for the economic and fiscal forecasts presented in its budgets and economic and fiscal updates.  The Department of Finance’s detailed economic forecasts should be used to prepare the fiscal forecasts, rather than using the average of private sector forecasts for a selected number of major aggregate.  The Department’s economic forecasts should be compared with the average private sector economic forecasts, with differences fully explained and justified.  Forecast details on both real and nominal GDP should be publicly provided so that assessments can be made on the composition of GDP.  This should be provided on a consistent basis.  The PBO also argues for this. Like the Bank of Canada, Department of Finance officials should appear on a more regular basis before committees on Parliament to explain the impact of current economic developments.  

Changing the process will open the debate on the applicable economic assumptions to use for budget planning.  It will once again ensure that the government is fully accountable and more transparent for both the economic and fiscal projections.  It will once again engage the private sector economic forecasters in a public debate.

The Focus of the Buget should be on the Medium term

From 1979 to 1993, five-year budget forecasts were provided.  This was changed in the 1994 Budget to two years.  The rationale given for the change was that government would be required to take immediate corrective actions to ensure that the short-term fiscal targets would be met if it appeared that they could be at risk. This had not been the case previously where there was a tendency always to forecast the meeting of fiscal targets at the end of the medium term and to postpone real policy actions.  Beginning in 1994, five-year economic and fiscal forecasts were provided in the fall update.

In the 2005 Budget, five-year fiscal projections were re-introduced.  The rationale given was that the majority of initiatives proposed in that budget extended beyond the traditional two-year horizon.  The 2006 Budget, however, returned to a two-year time horizon, arguing that uncertainties are fewer over the short term and that the Government can be reasonably held to account for its short-term fiscal targets.  However, this didn’t last long, as five-year budget forecasts were once again provided in the 2008 budget and have been since, although the Government still argues that its focus is the short term.

A short-term budget focus can be appropriate in a period of unsustainable debt levels and the need to aggressively deal with high budget deficits.  Previously, short-term fiscal targets were regularly missed and corrective fiscal action delayed. However, both the OECD and IMF recommend the presentation of medium-term budget plans rather than just focusing on the short term (one or two years).  They argue that a medium-term planning horizon ensures that fiscal policy is more appropriate to the business cycle.  Short-term targets often result in governments taking pro-cyclical measures rather than letting automatic stabilizers work.  A medium-term focus ensures that measures taken in the short term are sustainable over a longer period.  A medium-term focus also identifies emerging structural pressures, which often are better addressed early in the process rather than later on.

Currently, the government argues that a two-year budget horizon is appropriate because it can be more reasonably held to account for its short-term forecasts as the uncertainties are greater over the medium term.  However, in both the 2009 and 2010 budgets, it provided an update of the five-year fiscal projections in the fall updates, while stating it was still focusing on the short term.  This was done to demonstrate that the deficit was on a downward track and that a balanced budget was achievable by the end of the five-year period.

With the exception of most of the measures announced in its Economic Action Plan, the measures proposed in most budgets extend beyond the two-year budget time horizon.  Tax changes are usually ongoing and it is important to assess their impacts over a longer time horizon.  The same is applicable for most spending initiatives.  They too are ongoing, although the government may wish to re-announce them on a regular basis.

It is more important now that the budget plan be for at least five years and that the fiscal measures proposed in the budget be put on the same time frame. Potential economic growth is slowing for demographic reasons and there are major upward pressures on government spending as the population ages. This means rising deficits.  Strategies need to be developed now to deal with these emerging structural issues. The Department of Finance has provided some impact of the consequences of the aging of the population in previous working papers.  This analysis should be updated and presented on a regular basis.

The Budget needs to be more Prudent

Budgets from 1994 to 2005 contained a Contingency Reserve as well as some form of economic prudence – either indirectly through adjustments to key economic variables or explicitly.  This prudence recognized the uncertainties in economic forecasting and the difficulties inherent in translating the economic forecasts into fiscal projections.  In the mid-1990s, prudence was important to enhance the credibility of the annual fiscal targets.  With the emergence of balance budgets and increasing surpluses, the need for prudence was less evident.  However, prudence was still required to ensure that the annual fiscal targets would be met without having to take in-year action, which could be extremely disruptive to departmental planning.

In the 2006 Budget, the Contingency Reserve and explicit economic prudence were eliminated.  Instead, the government committed to reducing debt by $3 billion a year, primarily counting on better-than-expected fiscal results at year end to honour that commitment. 

The financial meltdown underscores the uncertainties inherent in economic and fiscal forecasting.  Other unexpected developments experienced during the course of the fiscal year (floods, SARS, Avian flu, etc) exacerbate the impact of economic uncertainties.  The impact of structural imbalances affecting the fiscal framework further compounds these uncertainties.

No one could have forecast the fiscal impact of the latest recession.  Nor should one attempt to offset the impact of automatic stabilizers on the fiscal balance.  However some degree of prudence is required to manage non-cyclical factors, especially errors in translating economic forecasts into fiscal forecasts.  This would re-enforce the credibility of the announced fiscal targets.   A prudence reserve of 2% of total revenues would not be unreasonable

Eliminate Use of Private Sector Forecasters to Produce Fiscal Forecasts

As part of the changes to the budgetary process in 1994, four private sector forecasting organizations[2] develop detailed fiscal projections on a National Accounts basis, based on the average of the private sector economic forecasts and the tax and spending policies in place at the time of the last budget for the next five years.  The Department of Finance would convert these projections to a Public Accounts basis.  The average of these four projections was published in the fall updates and formed the bases for the pre-budget consultations.

 In the November 2006 Economic and Fiscal Update, the fiscal projections were prepared by the Department of Finance.  However, the four private sector forecasting organizations were asked to develop fiscal projections on a National Accounts basis, based on their own economic forecasts.  These projections were converted to a Public Accounts basis by the Department of Finance, subject to review and approval by the private sector organizations.  The fiscal forecasts by each of the private sector organizations were then compared to the fiscal projections prepared by the Department of Finance.  The use of the four private sector forecasting organizations to prepare fiscal projections for the fall update was discontinued with the September 2009 Update of Economic and Fiscal Projections.  No reason was given.

There were considerable differences among the projections prepared by the four private organizations and the Department of Finance.  This is not surprising.  First, the each of the private sector organizations used their own forecasts of economic growth rather than the average of the private sector economic forecasts.  Second, they also used their own distribution of the various component of nominal income to forecast the applicable components of budgetary revenues.  Third, the specifications of their econometric models differ, thereby resulting in different results, even if they used the same economic forecasts and distributions. 

This process was not entirely independent. A number of the key fiscal variables were provided by the Department of Finance on a National Accounts basis and not derived independently by the private organizations. Converting the National Accounts fiscal projections into Public Accounts projections requires in-depth knowledge of the difference between the two accounting bases.  In also requires information on liabilities that is internal to the government. The potential fiscal outcome for the current fiscal year was often based on the monthly financial results to date, as determined by the Department of Finance. This affected not only the current year’s fiscal estimates but the projections for the outer years as well, given that they are projected on the current year’s base.

The decision to discontinue this practice was the correct one. The use of the four private sector forecasting agencies to produce fiscal forecasts was first proposed when there was no PBO.  Now PBO presents its fiscal forecasts on a regular basis.  There is no need for the four private sector forecasting agencies to produce fiscal forecasts, given the significant involvement of the Department of Finance in this process.  It can not be considered independent. Following the release of the PBO forecast, the House of Commons Standing Committee of Finance should undertake a review of the two fiscal forecasts.

[1] Review of the Forecasting Accuracy and Methods of the Department of Finance: Ernst & Young September 1994.

[2] Global Insight, the University of Toronto, the Conference Board of Canada and the Centre for Spatial Economics