FINA Committee Meeting

Notices of Meeting include information about the subject matter to be examined by the committee and date, time and place of the meeting, as well as a list of any witnesses scheduled to appear. The Evidence is the edited and revised transcript of what is said before a committee. The Minutes of Proceedings are the official record of the business conducted by the committee at a sitting.

For an advanced search, use Publication Search tool.

If you have any questions or comments regarding the accessibility of this publication, please contact us at accessible@parl.gc.ca.

Previous day publication Next day publication

PDF

38th PARLIAMENT, 1st SESSION

Standing Committee on Finance

EVIDENCE

CONTENTS

Thursday, November 18, 2004

1535

The Chair (Mr. Massimo Pacetti (Saint-Léonard—Saint-Michel, Lib.))

Mr. Tim O'Neill (Executive Vice-President and Chief Economist, Bank of Montreal, As Individual)

1540

1545

The Chair

Mr. Tim O'Neill

1550

The Chair

Mr. Tim O'Neill

The Chair

Mr. Tim O'Neill

The Chair

Mr. Charlie Penson (Peace River, CPC)

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

1555

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

1600

Mr. Charlie Penson

The Chair

Mr. Tim O'Neill

The Chair

Mr. Yvan Loubier (Saint-Hyacinthe—Bagot, BQ)

1605

Mr. Tim O'Neill

1610

Mr. Yvan Loubier

Mr. Tim O'Neill

The Chair

Hon. John McKay (Scarborough—Guildwood, Lib.)

1615

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

1620

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

Hon. John McKay

Mr. Tim O'Neill

1625

Hon. John McKay

Mr. Tim O'Neill

The Chair

Ms. Judy Wasylycia-Leis (Winnipeg North, NDP)

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

1630

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

1635

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

Mr. Tim O'Neill

Ms. Judy Wasylycia-Leis

The Chair

Mr. Tim O'Neill

The Chair

Mr. Don Bell (North Vancouver, Lib.)

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

1640

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

Mr. Tim O'Neill

1645

Mr. Don Bell

Mr. Tim O'Neill

Mr. Don Bell

The Chair

Ms. Rona Ambrose (Edmonton—Spruce Grove, CPC)

Mr. Tim O'Neill

1650

Ms. Rona Ambrose

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

1655

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

The Chair

Mr. Guy Côté (Portneuf—Jacques-Cartier, BQ)

Mr. Tim O'Neill

Mr. Guy Côté

Mr. Tim O'Neill

1700

Mr. Guy Côté

Mr. Tim O'Neill

Mr. Guy Côté

The Chair

Mr. Yvan Loubier

1705

Mr. Tim O'Neill

Mr. Yvan Loubier

M. Tim O'Neill

The Chair

Mr. Yvan Loubier

The Chair

L'hon. Maria Minna (Beaches—East York, Lib.)

Mr. Tim O'Neill

1710

Hon. Maria Minna

Mr. Tim O'Neill

Hon. John McKay

The Chair

Mr. Yvan Loubier

Hon. John McKay

The Chair

Hon. Maria Minna

Mr. Tim O'Neill

Hon. Maria Minna

Mr. Tim O'Neill

Hon. Maria Minna

Mr. Tim O'Neill

Hon. Maria Minna

Mr. Tim O'Neill

Mr. Yvan Loubier

Mr. Tim O'Neill

Hon. Maria Minna

1715

The Chair

Mr. Tim O'Neill

The Chair

Mr. Yvan Loubier

The Chair

Mr. Charlie Penson

Mr. Tim O'Neill

Mr. Charlie Penson

Mr. Tim O'Neill

The Chair

Hon. John McKay

The Chair

Hon. John McKay

Mr. Charlie Penson

1720

The Chair

Ms. Rona Ambrose

Mr. Tim O'Neill

Ms. Rona Ambrose

Mr. Tim O'Neill

Ms. Rona Ambrose

The Chair

CANADA

Standing Committee on Finance

NUMBER 018

1st SESSION

38th PARLIAMENT

EVIDENCE

Thursday, November 18, 2004

[Recorded by Electronic Apparatus]

* * *

¹ (1535)

[English]

The Chair (Mr. Massimo Pacetti (Saint-Léonard—Saint-Michel, Lib.)): Good afternoon.

Mr. O'Neill, good afternoon; I want to thank you for being here with us today. I suppose you have an opening remark or opening statement for us, so I'll let you have the floor and then I'm sure the members will have questions for you.

Mr. Tim O'Neill (Executive Vice-President and Chief Economist, Bank of Montreal, As Individual): Thank you very much.

Thank you for the invitation. It's an opportunity for me to give you a bit of detail, but hopefully not overmuch detail, about what I plan to do and allow you time to ask questions, but it's also an opportunity for me to hear comments and suggestions from you about what paths and what issues may need to be addressed.

As you know, the Minister of Finance announced recently that I would be reviewing the government's economic and fiscal forecasting, and as you heard him reiterate this week, the work is primarily focused on the accuracy of both economic and fiscal forecasting. As well, the IMF is going to be involved in looking at inter-country comparisons of forecasting activity, and I'll come back to that.

What I see in the work, and ultimately in the report, are four main areas of coverage. One, and probably the largest and most important one, is doing some actual quantitative analysis or empirical analysis of the accuracy of the economic and fiscal forecasts that we have seen. With respect to the economic forecasts, it means doing an evaluation of how accurate the private sector economists' forecasts have been since the government chose to use the private sector economists' forecasts as a fundamental part of its fiscal projections.

Part of that also is an assessment of the kinds of volatility or changes in volatility in key macroeconomic variables that we may have observed. That becomes relevant because obviously I've spent most of my life as a forecaster, but these are things like changes in the degree of variability that we're used to seeing in growth or inflation, interest rates, even exchange rates. We'll look at how much these may have changed.

With respect to fiscal forecasts, the key focus is on the in-year and one-year-out forecasts, but also the two-year-ahead and the five-year-ahead forecasts as well. Obviously, from my point of view at least, the farther out in the future you project, the wider the range of possible variance from projections the actual results are likely to be, and that's an experience all countries have had to deal with.

In that context, I think the critical work is on assessing, evaluating, if you will, which factors are most critical in explaining the discrepancies between the projections and the actual outcomes. And part of that also will be trying to analyze the extent to which we have actually seen functional changes in the relationship between the key economic variables, or economic performance in more general terms, and the key revenue and spending variables.

For example, to give you an illustration, if there has been a significant change in the composition of the sources of income for households--wage and salaries, dividends, capital gains, interest-- because there are different tax rates associated with them and because there are greater degrees of volatility in year-to-year changes, that can have an influence. With the same economic performance a change in the composition can make a difference to the kinds of fiscal numbers and revenue numbers that we actually observe. So that's the first piece.

The second piece I've also made reference to is the work by the IMF to do a comparative analysis of Canadian federal fiscal forecasting and that done by central governments in other countries. That comparative analysis is already under way, and among the countries the IMF is looking at are the U.S., the U.K., Australia, New Zealand, the Netherlands, and several European countries. They're going to look at, among other things, what are the key characteristics of how the forecasting is done, what institutions are involved, what processes are used, what budget rules may be part of the process. And then they will assess the relative, as I put it here, quality of those countries' forecasts.

¹ (1540)

One of the key focuses is what kind of fiscal environment are these countries operating within, and that comes back to a point I made earlier: to what degree are countries different with respect to how volatile their revenue sources are. For example, suppose Canada relies to a greater extent on corporate income tax than do other countries. Because that tends to be a more volatile revenue source, that is, more variable from year to year, the nature of that particular composition of revenue sources would tend to, other things being equal, bias the Canadian numbers to be more volatile than other countries' numbers. So you want to look at that, and they will be looking at that very carefully.

The third part of this is really more just making sure that everybody, including me, is clear on what are the changes we've seen in the forecasting process and the budget-making process that are relevant to this particular study. I'm not going to examine all budget-making issues--that's far beyond the mandate I've been given--but I'll focus with specific respect to changes that have been made in the way we do budgets. For example, we have formally now included in the budget-making process a contingency reserve and a prudence factor in the projections. That has been part and parcel of the budgets we have seen come from the government over almost the last decade.

So an interesting question there is to what extent does that change the nature of the projections? Does it set a lower boundary on the projections one is going to work with, and how does that then flow through to the kind of flexibility a government would provide for itself in the budget-making process?

We've also seen in recent years another still modest but not insignificant change, and that is a shift to, in some instances, longer-term commitments. For example, most recently striking examples would be the health care agreement with the provinces and the equalization agreement. Those have long-term commitments entered into by the federal government, in this particular case, with the provincial governments. But if you go back even four or five years, we had a five-year program of tax reduction. Obviously, in any of these instances the government has a right to change that program, but the fact is we are making these longer-term commitments.

To what extent does that then affect the degree of flexibility one has in adjusting to changes that may have been unanticipated? I want to take a look at that carefully.

In the final part of the background work, I want to engage in some consultations with individuals external to the government who will have both some technical expertise, be able to talk intelligently and with insight to the whole forecasting process and the institutions associated with it, but also would have reasonably informed views on the broader issue of transparency and flexibility in the budget process. And here I'm only talking about process.

The public interest is obviously significant, as you are well aware, or more aware than I. There is a public interest in the accuracy of fiscal projections and the transparency of the process by which they are generated or produced. In particular, policy-makers are certainly better enabled to make appropriate decisions about what changes and what capacity for change there is in, for example, program spending or tax changes, and of course the public is better enabled to form opinions about the quality of those decisions and their implications.

¹ (1545)

What I'm looking at are consultations with fundamentally two groups of people: first, the economists who have been involved in this process, because they will clearly have a view on whether and to what extent it ought to be changed and how effective it has been; and secondly, other individuals who I think it's reasonable to suppose could provide me and the report with some insightful analysis of the issues I just described.

The last part of this talks about what will be in the final report, and I think it is fairly straightforward and obvious. One is to report the results of the analysis itself; the second is to talk about what the implications are, both with respect to improving forecast accuracy and with respect to any improvements in the transparency of the process in which we'll be engaged; and then, in both of those areas, to make recommendations as warranted by that analysis to the minister about changes to institutions, procedures, techniques, and so on, and perhaps recommendations about the appropriate role, the fiscal targets, the role of parliamentary committees, public consultations, and so on, in the budget-making process.

I may have taken a bit longer than you had planned on, Mr. Chairman, but I thought it was useful to lay out in some reasonable comprehensiveness what I'm planning to do.

The Chair: For some of us who are not aware of your background, how long have you been doing this as an economist? Basically, tell us what your background is.

Mr. Tim O'Neill: I have been with the Bank of Montreal, first as deputy and then as chief economist, since 1993. Before that I was with the Atlantic Provinces Economic Council as president. In both of those roles I was involved in consultations with the finance department and with officials and the minister on the budget process.

Specifically with respect to forecasting, that has been one of my main roles over the last 11 years, and with respect to this particular project, the work actually began about a month and a half ago.

¹ (1550)

The Chair: Are you still working at the bank?

Mr. Tim O'Neill: I am.

The Chair: So this is just a side project.

Mr. Tim O'Neill: This is my part-time job.

To be less facetious about it, the bank has been very generous in allowing me to use time when I would have otherwise been involved in bank-related activities to do this work. My plan is that I will be completing it by the early part of February. That is when I'm intending to report to the minister and of course to this committee.

The Chair: Colleagues, we have about 90 minutes and eight members. I'm going to give you all 10 minutes, and if there's some time left we'll give Judy some extra time, like she always requests.

We'll try to go back and forth. Mr. Penson is first, please.

Mr. Charlie Penson (Peace River, CPC): Thank you, Mr. Chair.

I'd like to welcome Mr. O'Neill here today. I think you're a very good choice for the project that's under way, and I'd like to welcome you to the kickoff of our standing committee's hearings on a special budgetary review process.

As you know, Mr. O'Neill, this matter has been referred to our committee for study, and we certainly hope to have a number of witnesses come to talk to us as well, so I think it is important that you're here today to give us the timeframes and the parameters of what you're looking. And I think you've already said that you would be open to committee members' suggestions..

Mr. O'Neill, just to set the tone, we do all have to appreciate how important it is to bring some credibility back to this process. I know that's why you're involved in it. You've been involved as well in private sector forecasts that the Minister of Finance has used in the past to develop budgets. You have our support. We think this is a very worthwhile project, and if it ends up with your recommendation that we go outside to a special budgetary review office, the Library of Parliament or something like that, it would help to bring some credibility back.

You've already addressed one thing in terms of timing, but I think it is very important, Mr. O'Neill, that the Minister of Finance has the benefit of your report and we have the benefit of some kind of expertise going into the next budget, because we've had seven budgets, I think, where the surplus has been larger than projected. The reason we're all here today is that we want to find out why that has been off so much.

I have two questions, just to kick it off. I understand that you'll have your report ready for the Minister of Finance sometime in early February.

Mr. Tim O'Neill: That's correct.

Mr. Charlie Penson: Are you planning to share that report with our committee, or will you be reporting just to him?

Mr. Tim O'Neill: My understanding, in particular from the comments made by the minister earlier this week, is that I would be reporting to him but that I would also be providing this committee with the report. I'm assuming—but I don't know this for sure—there will be an opportunity to discuss it with you if that in fact is your desire.

Mr. Charlie Penson: I think that would be helpful.

Mr. O'Neill, the fact that the whole process has been off so much in one particular direction in the last several years has raised a lot of concern. We're doing pre-budget consultations right now and we are getting an earful about that. A lot of people feel Canadians have been denied a chance to debate what should happen, and there's a lot of spending going on outside the budgetary process. The industry itself, the people who are giving advice to government, has been tainted a little bit by this process.

Now, it's my understanding that when they have asked private sector forecasters to forecast, in the past they have not been looking for forecasts on direct program spending, not reviewing that sector of it, which this year alone is some $70 billion. I see in your outline for us you're talking about the need to evaluate the accuracy since 1995, I think, and make an assessment for changes in volatility in the last 10 to 20 years. Will you be looking at the fact that the private sector has not been forecasting, has not been giving advice to government on the direct program spending component of it as part of that assessment?

¹ (1555)

Mr. Tim O'Neill: Let me just correct that. Initially the process involved simply the private sector forecasters providing the formal economic outlook. It was then averaged and that was used as the foundation for fiscal projections. But in more recent years there have been forecasting firms that have at least taken a first stab at fiscal projections on a national accounts basis, so there has been some involvement in the actual fiscal projections themselves.

But I think it's fair and accurate to say that ultimately the fiscal projections that appear in the budgets are the responsibility of the department and have been the result of work done within the Department of Finance.

Mr. Charlie Penson: In that regard, you can only forecast based on what information has been given to you by the Department of Finance. Is that correct?

Mr. Tim O'Neill: In terms of the fiscal forecast, of course that's correct. All of us would have views we would express on, for example, the risks associated with an economic forecast that might have a bearing on, for example, what the revenue streams are likely to be and therefore what the risks are to the actual fiscal projections themselves. We were perfectly free to provide that kind of advice.

But in the formal projections themselves, obviously there's a fair degree of detailed information most of us would not have the capacity to utilize within our own shops. Therefore, we wouldn't normally be inclined to try to do a detailed fiscal projection.

Mr. Charlie Penson: I'll just follow up on that, Mr. O'Neill. My understanding is that there are two or three components of how these surpluses have come in larger than the forecasts said. They have largely come from underestimating the amount of revenue that would be coming in, but there's another very important component and that's overestimating the size of expenditures, including debt service expenditures. Such estimates should really not be that difficult to do.

The reason we're all here today is that there's a problem. In your initial look at this, what have you found that is causing the difference on the expenditure side and has been putting this thing out so far every year for the last seven years?

Mr. Tim O'Neill: Let me be fair to myself by pointing out that since I am at the beginning of this, I don't have a lot of final, firm, definitive conclusions. There's a lot of work left to be done. But I think at least two or three general points are worth making here. The extent to which economic forecasts may differ from the actual outcomes can have an effect both on revenue projections and on spending projections. It's not just the level of growth in the economy that can make a difference, but it's also the composition of where that growth comes from.

For example, a couple of years ago there was a significant difference between what had been projected in the form of equalization payments and the actual payments that had to be made under the requirements of that particular program. A key reason for that was, in the weaker economy we experienced several years ago, one of the areas that was most adversely affected was the manufacturing base—i.e., Ontario and Quebec. So if you think about how equalization is derived, you'll see that the fact it was Ontario that was weak made a difference. That had an influence, therefore, on the spending numbers, a significant one.

Secondly, the economic performance data can make a significant difference to revenues, and I am sure you're well aware of that. Just a preliminary casual examination of the economic forecasts over the last seven years indicates that in every year since 1997 the economic forecasts have under-predicted nominal GDP growth. That makes a difference, clearly, to the government's revenue projections.

As far as debt-servicing costs are concerned, a key issue is the interest rate forecast, because that will affect what the rate of those debt service costs is, and the economic forecasts have also erred to some extent in the interest rate forecasts.

So as an explanation for why even something as seemingly straightforward as debt service costs might be misforecast, if I can put it that way, that would be it. But I think, generally speaking, that this component of the variance between the projection and the actual was relatively small.

º (1600)

Mr. Charlie Penson: But, Mr. O'Neill, I'm sure you know the Centre for Policy Alternatives did some work in February last year on projections, basically just using percentages of GDP, and they seemed to be pretty close to the mark. You must use this yourself, and private sector forecasters must use it.

All things being equal, if there were errors in judgment because it was difficult to forecast back and forth several ways, we should have had deficits rather than surpluses the odd year, but we have a trend that seems to be systemic. People who are using these other types of forecasts have plugged in the numbers and said, well, here's where I think it is, and they seem to be more accurate. Can't the Department of Finance do that too, or shouldn't they do it?

The Chair: I am sorry, but we have run out of time, so if you could, answer that very quickly.

Mr. Tim O'Neill: First of all, I made a commitment to Jim Stanford that I'll talk to about why he was more accurate. Second, I think it is fair to say that we've tried to use the normal rules of thumb that have served us well in the past for this growth rate in nominal GDP, this growth rate in revenue. Under-forecasting growth has been one of the reasons we've been off. Maybe the centre was able to be more accurate in their revenue and growth forecasts.

Also, and this is one of the reasons I mentioned in my outline, I think we need to look carefully at whether there have been some structural changes in the relations between the economic performance—economic growth—and the revenue numbers, whether there has been some sort of functional change in that relationship over time, because it certainly appears that this is the case in certain areas of revenue.

The Chair: Mr. Loubier.

[Translation]

Mr. Yvan Loubier (Saint-Hyacinthe—Bagot, BQ): Mr. O'Neill, thank you for coming. I am very pleased to see you again. You have an excellent reputation as an economist and I think that you will help us find our way through the forecasting issue.

I was listening to your outline, earlier, when you mentioned the first thing you were going to do. You will probably have some difficulties. You will analyze what the economists said in the past as well as their surplus forecasts. You will find it difficult to distinguish political projections from purely economic projections.

Let me explain. When consultations started, in 1994, I was a member of this Committee and Mr. Martin was boasting about the fact that he consulted private sector economists including economists from major Canadian banks. However, he did it behind closed doors. He said that after consulting these economists, he came to the conclusion that the surplus would not exceed x billion dollars per year. This week, in the economic update of Mr. Goodale, we got for the first time the exact projections of some forecasting firms concerning surpluses. Until now, we never had access to that kind of data.

Secondly, among the economists who intervened publicly, to my knowledge two have already worked with Mr. Martin as Deputy Ministers of Finance. I do not doubt their competence, but here again, their numbers were so off the mark, so far from reality, based on what we could already measure with the indicators, six months before final results, that they had to be somewhat tainted by a political agenda.

Members of the Bloc do not have a crystal ball, far from it. However, since 1997, we managed to forecast surpluses with reasonable accuracy. We did not say that we had forecasted these results after the fact. I took part in televised press conferences—you can check—in which we gave our projections. How did we do it? It was a challenge, but we based our numbers on the forecasts of the five major banks, the Caisse de dépôt et placement and the Mouvement Desjardins concerning the nominal GDP growth. We made an analysis of each of their forecasts and made some adjustments a few weeks before announcing our own projections and we gave a nominal GDP growth rate that made sense.

We had also noticed, particularly before 2002, before the full indexation of personal income tax brackets, that the elasticity between nominal GDP growth and personal income tax revenues was such that there was a spread of about three percentage points. So, if a 4% nominal GDP was forecasted, fiscal revenues from personal income tax were about 7%. I think it was 7.4% at a given time in 1998 or 1999.

We also learned, from the Department of Finance's fiscal review, to check periodically revenues and expenditures and identify periods when they were higher than at other times of the year. We were carefully watching all government decisions concerning tax cuts, fiscal agreements, health agreements, for example on CHST. Using all that information and our judgment, we succeeded in making projections 3 or 4% off actual results, except in 1997 when there had been a difference of about 10% between our forecasts and reality.

Firstly, in your analysis, will you be able to distinguish between what is political and what is not?

Secondly, should we not come back to simpler but basic methods, I mean by that the empirical observation of payments and receipts, government decisions which impact on receipts and maybe nominal GDP, even if we have to revise six months before the economic updates the forecasting made six months earlier? We could do that and update our numbers periodically. This was the first part of my comments.

º (1605)

Mr. Penson raised earlier the question I want to ask. It is about an independent forecasting office. It is working in the US, so why not here? It would be an institute at arms' length from the government that would give us that type of projections.

Secondly, that independent forecasting office should have access to more refined data from the Department of Finance on monthly payments and receipts. With that information, it could have a better picture of the surplus size at year-end.

I would like you to comment on what I said and to answer my question.

[English]

Mr. Tim O'Neill: With respect to an outside agency, the U.S. clearly is an example of a country where with the Congressional Budget Office they provide fiscal and economic projections that are non-partisan, if I can put it that way. The Netherlands also has an agency that is arm's length, but still a government agency, which also does economic and fiscal projections. They go one step further. Those projections become the baseline for all of the parties in the government or in Parliament—all of the parties' economic platforms start with that common base.

I've indicated in other comments I've made in the past regarding this work that I'm intending to look at those models. In fact, in addition to the work being done by the IMF, which will help give me some sense of, from an empirical point of view, whether it has made any difference, I'll also be spending time in both Washington and the Netherlands to talk to the individuals who are involved in the forecasting.

I can't say at this point whether I'm going to conclude that this is a good idea for Canada and whether or not it's a recommendation I would make. Because clearly, the issue will be to what end, to what benefit, to what purpose would we create a separate agency?

One might be to make more accurate projections, but obviously the issue of whether the countries involved have actually ended up with better or equal or worse fiscal projections than Canada will be of some importance in that analysis. I can say just from a casual examination of data on fiscal projections in both countries that at first glance at least it does appear as if their performance on fiscal projections is no better or may even be worse.

The second purpose might be—and this is of a different order—to improve the transparency of the process. That's one of the reasons I want to talk to the people in at least those two countries that are using that outside agency model to see what kind of impact it's had on the whole budget-making process.

At this point I'm afraid I can't give you a definitive answer.

º (1610)

[Translation]

Mr. Yvan Loubier: Mr. O'Neill, you said that at first sight, forecasting doesn't seem better in the US and other countries. Do they have a 50% variance between their surplus projections and real results? This is what happened here last time. I've never heard criticisms saying that they made such obvious mistakes in the US or Europe for political reasons. I have also one important remark to make concerning an independent forecasting office. It is important that everybody be as well informed as possible and that the process and numbers that are given to us be credible.

Secondly, we are here to serve democracy. When the real situation of public finances is hidden from us for a whole year citizens are not able to measure the performance of their government, its capacity to meet their expectations and their priorities. It could be interesting the two first years to get unexpected surpluses as they were presented by Minister Martin and later by Minister Manley. The two first years, it was cute to have an unexpected surplus and people were happy. In 1994, we had a $43 billion deficit. So it came as a good surprise. Later, Canadians have been faced with inescapable priorities such as health, education and the fight against poverty while we got a false picture of public finances. This is not a good way to serve democracy.

I think that you have a very important mandate. If we do not find a way more just and more credible to forecast the result of government operations, we will totally miss our goal and we will not serve the interest of the people. In that sense, your mandate is crucial.

Here is my last question. I was asked why the IMF is intervening. The last time, it was in 1994, when the IMF enjoined Canada to improve its poor budget results. Why is the IMF intervening?

[English]

Mr. Tim O'Neill: To answer the last question, there are really two reasons. One, we are operating with a much shorter timeframe than we did the last time this work was done. Secondly, they have a large network of analysts and have access to… I mean, we all have access to the data from other countries, but they're able to marshal it much more quickly, to pull it together far more quickly than I could or any of my staff at the Bank of Montreal or even staff at the Department of Finance could.

I guess a third reason would be precisely because they're an international agency, completely outside of Canada, outside of a national agency. I think we can rely very extensively on the responsibility that they will take and apply to the research they're doing. But it's really the first two reasons. Logistically it makes a lot of sense, and secondly, they can gather data far more quickly because they have those connections with all of the governments, especially in the OECD countries. So it will make the work a lot faster, for me at least. It will be a wonderful support for the work I'm doing.

The Chair: Thank you, Mr. O'Neill.

Mr. McKay.

Hon. John McKay (Scarborough—Guildwood, Lib.): Thank you, Mr. Chair, and thank you, Dr. O'Neill, for your assistance.

I just want to go over the current process. I assume you've been involved in the last number of years in this process. The first stage of the process, as I understand it, is that the 15 or 20 economists gather in a room and analyze the data. The first question I have, therefore, is whose data are you analyzing? Are you analyzing the department's data, or do you bring independent data to the table?

º (1615)

Mr. Tim O'Neill: The data we're working with, if we're talking about the economic forecasts, are of course the same data every forecaster is working with—current data on the economy. But all of us do economic forecasts, at least all of the bank economics departments do, and there are several forecasting firms as well that are involved in this process. We simply bring to the table the economic forecast that we have for the coming year or two years.

Hon. John McKay: So is this the same stuff you would say to your own employer?

Mr. Tim O'Neill: Absolutely. If you went to the Bank of Montreal economics department website, the numbers that you see there are precisely the same numbers that we would provide in this process to the Department of Finance.

Hon. John McKay: So you go into the room, and what are the parameters? Do you just give GDP, or do you give GDP plus interest? Do you give interest on inflation? What are the factors you put into that forecast?

Mr. Tim O'Neill: The key numbers would be GDP growth, inflation, short-term interest rates, and long-term interest rates.

Hon. John McKay: Okay, so those would be your main four—

Mr. Tim O'Neill: The four main ones, that's right.

Hon. John McKay: — and there would be probably a few others.

And what massaging, if any, do the economists do with those numbers? Presumably you all don't agree on what those four or five components might be. How does that work?

Mr. Tim O'Neill: The formal process is that we sit around in a group like this and talk about how we came to the conclusions we did about the forecast, what risks there may be, whether they're negative or positive risks. But the formal, final projection that's used in the budget comes simply from taking an average of the forecasts of each of the variables that have been provided by the forecasters. So whether it's seven or eight or ten or five, you simply average those numbers, average the growth rate, average the interest rate numbers, average the inflation rate, and that's—

Hon. John McKay: So it's nothing more sophisticated than that.

Mr. Tim O'Neill: It's not any more sophisticated than that.

Hon. John McKay: So does the department try to influence you in any one way or another at that stage, or does anybody at a political level try to influence you one way or another at that stage?

Mr. Tim O'Neill: No, and even if they did, it wouldn't make any difference. All of us have a public role that we play within our own institutions, so what I say in public at a presentation is bound to be the same thing I say internally within the bank when I'm advising my own colleagues, and it's going to be exactly the same thing I say to the Department of Finance or to the minister or to members of this committee or anybody else.

Hon. John McKay: Is there any gap that you've observed in terms of either the quality or the quantity or the character of the economists who are in the room? Is there something that's obvious to you that's missing there?

Mr. Tim O'Neill: No, I don't think so. I think we all do in our day-to-day lives as honest and as competent a job as we can, understanding that forecasting the future is very difficult and therefore there are likely to be errors. In fact, what I often say in public presentations is that the one thing I know about my forecast is it will be precisely wrong—that is to say, that in the precise point estimates it will never be perfectly accurate.

But there tend to be points in the economy when it's more difficult than normal to forecast, and that's when you're going through a transition, if the economy is slowing down or speeding up, because the timing can be very different and the components of that adjustment process can vary from one turning point to another.

Hon. John McKay: So the bunch of you in the room talk about these numbers, arrive at an average number; the department provides you with a bunch of cheap sandwiches and a table, and that's what it boils down to. Then, out of that, as I understand it, you hand those numbers off to modellers, econometric modellers. Is that the next stage of the process?

º (1620)

Mr. Tim O'Neill: That's right.

Hon. John McKay: Do you or anyone else in the room have anything to do with those econometric modellers?

Mr. Tim O'Neill: I do not because my particular department doesn't have its own economic model. We essentially subscribe to one that we then use and apply our own assumptions to. But there are firms that do provide modelling services that we subscribe to, and those folks are currently used in taking the economic forecasts and then using them as a foundation for fiscal forecasts.

Hon. John McKay: I understand there's what, three or four—

Mr. Tim O'Neill: That's correct.

Hon. John McKay: — modellers that are used. They run it through whatever their algorithms are and punch out.

What happens after that? Do they talk among themselves? And are you part of that conversation, or is it just the modellers that talk among themselves?

Mr. Tim O'Neill: The results are shared with us, but there's no particular need for us to be involved in the minor adjustments or tweaking of the projections. But certainly we are informed of what the results are. Usually you think of this as a two-stage process, where we meet initially with the finance officials, go through the initial economic numbers, eventually the final fiscal projections, and then later we meet with the minister to talk about those numbers and what the risks are, where we're worried, and so on.

Hon. John McKay: In effect, present it to the client.

Mr. Tim O'Neill: That's correct.

Hon. John McKay: Do you follow a similar type of process for the Bank of Montreal? When you're presenting to the chairman of the bank, is it a parallel process?

Mr. Tim O'Neill: It's reasonably parallel to that, because the sorts of questions that the chairman, the CEO, the risk management committee of the bank would be interested in may be somewhat different from what the minister is interested in or the department is interested in or you're interested in, because obviously it has more to do with what does it mean for our particular institution. But the process is parallel.

Hon. John McKay: Is there a correlation, or an inverse correlation, for that matter, between the accuracy of the numbers, the time that you're projecting out? Is that modelled as well? In other words, a three-month projection you can almost go to the bank on, so to speak—six-month, nine-month, that sort of thing. Is there a formula that's actually applied to the length of time to which the forecast will be used?

Mr. Tim O'Neill: There's no formula, as such, but I think it's fair to say, and I've mentioned in passing already, that the farther out the period is that you're forecasting, the greater is the potential degree of variance between what you anticipate and what you actually see happening.

Hon. John McKay: Your colleagues just participated in this process with the minister, what, two weeks ago?

Mr. Tim O'Neill: That's correct.

Hon. John McKay: When will they know whether or not they're finally right?

Mr. Tim O'Neill: In the case of the economic forecasts, the forecast for this year, 2004, we really won't have final numbers, or reasonably final numbers, until about the second quarter of next year, because obviously we rely on StatsCanada to do the actual estimate of growth and so on.

The interest rates, of course, we'll know because we can observe them. That's not a problem. But it's the growth numbers that we won't have a final take on.

As far as the fiscal numbers, well, of course you're well aware of the lags that are involved in getting the final numbers on the fiscal side. You know that we just recently closed the books on fiscal year 2003-04. Our lag isn't quite that long, but there is some lag.

Hon. John McKay: You'd be closing the books on the numbers you just gave the minister in what, August 2006?

Mr. Tim O'Neill: Sorry, I don't want to confuse you here.

Hon. John McKay: Okay, clarify it for me then, please.

Mr. Tim O'Neill: For the fiscal year 2003-04, we only now have the final numbers, and those were the ones that were significantly different from what had even been projected in the budget document in February. That's what I meant—in other words, that there is a lag in actually getting the information on the past. The lag is a bit longer in the case of the fiscal numbers than in the case of the economic numbers.

On the other hand, it's also fair to note that in almost every country, you can have backward revisions to the numbers. So, for example, two or three years after the fact you could have a revision in the growth numbers for 2002.

º (1625)

Hon. John McKay: Here is a final quick snapper. Did any of the folks who were in the room predict a 32% rise in the dollar vis-à-vis the American dollar, or its 32% decline? Did anybody?

Mr. Tim O'Neill: Nobody that I'm aware of. Nobody stepped forward to claim credit for it, so I'm assuming that they, like I, did not forecast it accurately.

The Chair: Thank you.

Ms. Wasylycia-Leis, and then Ms. Ambrose.

Ms. Judy Wasylycia-Leis (Winnipeg North, NDP): Thank you, Mr. Chairperson.

First of all, I'd like to say hello to you, Mr. O'Neill, and welcome you here and congratulate you on your position. I'm not going to say yet, as my colleague from the Conservatives did, that I'm glad you're here, because I'm not sure. I'm not sure yet what difference your appointment is going to make to what I consider to be a significant issue around forecasting.

My first question for you is, do you recognize a problem in terms of the forecasting for federal budgets currently?

Mr. Tim O'Neill: I'm a very careful economist, and before I would… There is certainly an issue about the degree of accuracy that we have observed in the fiscal projections. My job, as I see it, is to determine what the primary causes of those discrepancies have been and to suggest any changes that might be made in the forecasting processes and procedures that could ameliorate those inaccuracies, could reduce those inaccuracies.

It's important to look not only at our own performance in this regard relative to the past and to some perhaps objective benchmark that might be set, but it's also important to look at what we're seeing in other countries, if only to get a sense of whether there is something unique, something unusual, about the Canadian case. All I can say at this point is that I don't know the answer to that. I do know economic forecasters around the world have difficulty making accurate economic forecasts, and I do know many governments have had, at various points in time, significant difficulties, or if I can put it this way, variances in what they're projecting and the actual outcomes.

I've seen estimates of the range of error for CBO forecasts in the U.S. that five years out they can be out as much as 100%. The point is that nobody has a lock on or has a magic capacity to ensure absolute accuracy in any set of numbers, and we're no different in that regard.

Ms. Judy Wasylycia-Leis: Sure. I appreciate that. I've a little difficulty trying to describe, using your words, the situation in terms of “degree of accuracy” and “discrepancies” when we're talking about such big discrepancies in the forecasting of the surplus over an extended period of time on a repeated basis.

Would you at least agree that if we're looking, from the 1994 budget to the present, at a total low-balling or unforecasted surplus of $86 billion, that isn't a problem that has to be addressed by you or the government or whatever work we do as a committee?

Mr. Tim O'Neill: Well, it's a question of what one would normally have expected, and secondly, what actually would have caused those variations.

Again, a quick glance at the numbers would suggest that, on average, the amounts by which the budget surplus projections have been out over the last seven years is around $4 billion to $4.5 billion. That amounts to about 1.5% of—

º (1630)

Ms. Judy Wasylycia-Leis: I'm sorry, are you saying we've been out $4.5 billion since...

Mr. Tim O'Neill: Over the last seven years, on average.

Ms. Judy Wasylycia-Leis: That's for each year?

Mr. Tim O'Neill: That's an average, yes. I'm talking about the original projection and then the base line before you make adjustments for actual policy changes, because of course that's the foundation for making the projection. That's a quick, off-the-top—

Ms. Judy Wasylycia-Leis: We're getting close to our numbers, but I think you're still low in terms of my understanding of, in the last seven years, being out at least $61 billion. I'm looking now at the stats comparing the official budget target with the alternative federal budget estimate and the actual.

For example, in 1999-2000, we're talking about $3 billion in terms of the official budget target, $16 billion estimated in the AFB, and the actual was $12.7 billion. We can go down the line in each case. It was $3 billion in 2000-01, $15 billion by the AFB, $18 billion actual. At the end of this period, 2003-04, the total official budget target was $12.4 billion, the AFB was $55 billion, and the actual was $55.8 billion. So in fact the alternative federal budget folks were very close to the actual.

It begs the question, as others have done: how could the 19 economists—of whom I think you were one—be so out when in fact these folks, the economists with the alternative federal budget, were so accurate using federal government assumptions? We're not just talking about one person. It's not just Jim Stanford. He's clearly been involved in the last few years, but many progressive economists have worked very hard at this and have been very accurate, and you haven't—or the group of 19 with the federal ministers haven't been. They've been way out.

I'm just trying to get a sense of whether or not you recognize this as a big problem. The minister doesn't seem to want to recognize it as a big problem. The parliamentary secretary doesn't want to recognize it as a big problem. I think Canadians are very disturbed. I think they don't feel very comfortable when they see the numbers being so far out, especially this year, looking at the $1.9 billion versus the $9.1 billion.

I would like to know: you were part of the 19, weren't you, who have given the government advice over the years on these budgets?

Mr. Tim O'Neill: Well, there has been a little bit of inflation of the numbers of people. In actual fact, the formal forecasting group numbers about eight or nine, but you're right, it's a relatively large number of people.

One of the things I want to do is take a look at the precise nature of those forecasts. There is a simple difference, for example, between comparing the budget forecast with the actual outcome after adjustments are made for policy changes and doing it beforehand. Obviously, you want to do the comparison with the status quo.

Ms. Judy Wasylycia-Leis: The AFB always did their forecasting beforehand, as well.

Mr. Tim O'Neill: With the status quo?

Ms. Judy Wasylycia-Leis: Yes.

Mr. Tim O'Neill: Then it sounds to me, and I'll have to check it, as if their numbers are a little bit different from the ones I've been looking at. But fair enough: there is a difference.

Ms. Judy Wasylycia-Leis: The fact is, they were right. This means a lot to Canadians, because in fact we're looking at $86 billion that went automatically against the debt because we couldn't have the accurate figures before Parliament for debate and the Canadian people were left out of the whole participatory democracy agenda. One could criticize this from a number of points of view, but I think it has serious ramifications all the way around.

I would like to know if you would give your blessing to our attempt to establish an independent parliamentary budget office.

Mr. Tim O'Neill: I think I've answered the question already, in the sense that first of all I want to look at how such an independent body might function, and second—and we have at least reasonably comparable examples of agencies that operate in other countries—do it in the context of what it is we're trying to achieve. If it is simply greater accuracy in projections, it may turn out that it's not an achievable outcome. If it's in terms of improving the process and its transparency and so on—

Ms. Judy Wasylycia-Leis: How could it not be possible to achieve greater accuracy if in fact other economists have been accurate and it's the government that has been out? Surely it is possible. Now the goal is, yes, to get the expertise to give the government the accurate numbers.

º (1635)

Mr. Tim O'Neill: It sounds as though what you're suggesting is we ought to hire the Centre for Policy Alternatives.

Ms. Judy Wasylycia-Leis: I think we could have saved a lot of money if in fact the government had actually gone the route of hiring the AFB—

Mr. Tim O'Neill: I think it's worth examining precisely how they came to the results they did.

Ms. Judy Wasylycia-Leis: As far as I understand it, the problem of accurate forecasts goes back a good decade. It's been recognized, I think, by Liberal finance ministers to be a problem, whether in deficit years or surplus years.

Have there not been, to the best of your knowledge, recommendations in the past for an independent forecasting capacity with Parliament or the government?

Mr. Tim O'Neill: There were, I think, in the analysis of this that was done in 1994, suggestions for the use of outside agents, and in particular what resulted from that was the use of the private sector forecasters as agents for generating the economic forecast. So in a sense we've done that.

Ms. Judy Wasylycia-Leis: In 1994, when the Prime Minister was the finance minister, there was a problem that was recognized and a recommendation was made for some sort of independent advice in terms of getting better, accurate forecasts.

Surely, Mr. Chairperson, we have what we need to make the case for pursuing an independent parliamentary budget office, and for talking to Mr. O'Neill today about how we can shape that process, whether or not he agrees with our approved terms of reference, and ascertaining what kind of advice he may offer us as we go down that path.

The Chair: Mr. O'Neill, would you like to answer that in 10 seconds?

Mr. Tim O'Neill: I think I've answered it in the sense that I will certainly look at that among other changes that might be made in the process and will give you my best advice and recommendations.

The Chair: Thank you.

Mr. Bell.

Mr. Don Bell (North Vancouver, Lib.): Mr. Chair, in local government—my experience has been in budgeting—I know the kinds of discussions we've had. When you get three economists or three lawyers in a room, you get three different opinions.

Mr. Tim O'Neill: Often four.

Mr. Don Bell: Yes, and that's the challenge.

As I understand, what we're talking about here is looking forward, and repeated throughout the terms of reference is the terminology “to improve the process”, and I've heard you indicate that as well. That's what I think we all want.

My question to you is, do you believe the terms of reference are adequate? Do they give you the freedom necessary to do the job you would feel comfortable putting your professional signature on? I'm referring to the introduction. Were the pages we've been given here the terms and detail you were given, or are these terms you developed out of a smaller set of less distinct directions?

Mr. Tim O'Neill: Are you talking about the outline you had in advance that I just went through?

Mr. Don Bell: Yes.

Mr. Tim O'Neill: That's my own setting out of what I see as the mandate for me to follow. The terms of reference were much briefer than that.

Mr. Don Bell: That's the three on the top?

Mr. Tim O'Neill: That's right. In answer to your question, if I had felt any qualms about feeling constrained in any way by the terms of reference, I wouldn't have taken the mandate on.

Mr. Don Bell: Okay. I have a few other questions.

I know we've been skirting this issue or discussing it. How accurate generally are private economic forecasts—among the banks, for example, which I know we look to quite often—for the coming years? I guess I'm going to the question of budgeting, but I'm asking, do they tend to err on the conservative side, on the side of caution?

Mr. Tim O'Neill: No, I don't think there's any one rule of thumb. I grew up in Atlantic Canada, and probably because of that I tend to be more an optimist than anything else. I think you have to be to survive.

I think it varies across the group of economists. Where you tend to see bigger differences in the forecasts are when you're at turning points—when an economy is going into or coming out of a recession. When you're in the kind of environment we're in now, which is more or less steady as she goes, economic forecast numbers tend to be closer together.

If you look back over the last seven years of economic forecasts, as I indicated earlier, we have tended as a group to under-forecast the actual growth that we've seen. That has been with respect to nominal GDP growth, which of course is what's important for forecasting revenue. Every single year we were under. The average for the seven years was under by about one and a half percentage points on growth. That necessarily makes a difference in the revenue projections and some of the spending projections that one would get from that process.

I think the best way to describe it is to say that in the late 1990s we were surprised at how continually strong the economy was relative to expectations and particularly in 2001 how much stronger it was in the face of a recession in the U.S. So I don't know that you'd regard it as a systemic issue or problem, but if you take those three years, the last two of the 1990s and 2001, those were the years in which you had the biggest difference between what we were projecting and what the economy actually generated in the way of economic performance. Now I think the numbers have come closer to what we've actually seen in the economy.

º (1640)

Mr. Don Bell: In terms of this last year, would you be looking at the impact on the projections on the economy of SARS, BSE, and those other factors of the day? Are those the kinds of things you'd be looking at?

Mr. Tim O'Neill: Yes.

I guess it's important to distinguish two kinds of differences or factors that can make a difference in an economic forecast. One is you're simply surprised, based on all the evidence and all the data, at how strong the numbers actually turn out to be. I'll give you an example. Recently almost every single economist I know... I can't speak for the Centre for Policy Alternatives, but everybody else has been very surprised at how well the economy has adjusted to a higher Canadian dollar. None of us would have predicted that exports would have grown so strongly in the first half of this year. That's a pleasant surprise, but what it means is that growth was somewhat stronger than anticipated, at least in that part of the economy.

What you have, I think, are instances when under the normal kinds of behaviours that we expect to see you actually get better or worse performance than you thought you would. Other cases are instances when you had shocks that were completely unpredictable, such as the combination of SARS, power blackouts, and so on in 2003, which cumulatively had an impact on growth you could never have anticipated, or 9/11 and its impact on growth in late 2001. So there are two different ways in which your forecast can be out: one, you're just simply surprised; two, you have shocks that you could never have anticipated.

Mr. Don Bell: Finally, in response to the question from Ms. Wasylycia-Leis, the issue is, I gather, three things that you're going to look at. One is the idea of government having its own internal forecasters and relying totally on that. The other is some kind of a hybrid where they use the private advice tempered with the internal. The other is the independent or external, as is the case in the Netherlands, as you said, which is one variation, and the other is the United States.

Mr. Tim O'Neill: That's correct.

Mr. Don Bell: I guess each of those would be unique to some degree to the culture they're located in.

Mr. Tim O'Neill: First, I'm not sure how I would define the culture, and second, I'm not sure the cultures of the U.S. and the Netherlands, for example, would be regarded as comparable enough to say that's the reason why they've chosen the model. In the U.S., of course, as you know, they basically have two parties, and in the Netherlands they have a multiplicity of parties and most often have coalition governments. Even the different political structures would not give you reason to think that they would choose a similar model.

I'm not sure it's necessarily unique to a country's culture. It may be the result—and that's one of the things I want to examine—simply of historical development, or it may be a choice that was made at a particular point in time because of circumstances. But I can say that for the three models you describe, you can find examples of each. Most of the European economies use the first model, which is internal forecasting, for both the economic and fiscal. Some do that but seek outside advice, much like Canada does, and some—the two I mentioned—have more or less completely independent agencies.

º (1645)

Mr. Don Bell: The final question, if I may, is on your fourth term of reference and the way you've defined it: “Consultations with individuals (external to the federal government) on economic and fiscal forecasting procedures and institutions and their relevance to the public interest in the federal budget process.” Are you talking about the way in which other input is received?

Mr. Tim O'Neill: I guess it's on two issues. One is how the forecasts are actually done. I'd already made a public commitment on CBC Radio to consult with Jim Stanford, for example, but there are other people like that who have done forecasting and have a reason to have insight into the process and the techniques and the modelling and so on.

And second, those people would also have informed views—they might be very different from each other—about the other set of issues related to the transparency of the budget forecasting process. So it's really in those two main categories of information ideas and insights that I want to pursue those.

Mr. Don Bell: I'm very pleased you've taken on this assignment and I like your understanding of the challenge that's before you.

Thank you.

The Chair: Ms. Ambrose.

Ms. Rona Ambrose (Edmonton—Spruce Grove, CPC): Thank you, Mr. Chairman.

Welcome to the committee. I want to thank you, Mr. O'Neill, for what I see so far as offering us a lot of straightforward and thoughtful responses. As a brand-new member of Parliament, I have to say that the one thing I've observed so far is that we don't get that very often. So good luck with your endeavours. We look forward to working with you and with this process.

You indicated earlier when you were responding to Mr. McKay that you use the same process in the private sector to advise your own board of directors as you do when you're advising the Minister of Finance, to a certain extent. Obviously when you advise your own board, you have all the numbers you need. My question really goes to the credibility of this process and as well for your industry. My concern is that I would suggest that if you were out by the amount we know the government has been out in their forecasts for the last few years, you would obviously have a serious problem in the private sector with your own organization, the Bank of Montreal.

I want to know whether you can reassure us that you have the same level of access as you would have in the private sector to the numbers you need to make sure you can accomplish what you are suggesting you will be accomplishing, and that's of course because when you do advise the Minister of Finance you have to rely on the numbers given to you by the Department of Finance. I wonder if you could address that issue.

Mr. Tim O'Neill: Because it's come up in several questions, let me just make a very simple, almost arithmetic point. If revenue projections were out by 1% and spending projections were out by 1%, which I think most people would admit is a relatively small number, that would constitute under the current levels of activity close to $4 billion of “error” in the combined estimate, especially if there was the same direction of impact. That of course would constitute a change in the forecast and result in either a surplus or a deficit. In other words, you don't need a large percentage of error in either revenue or spending to have what would appear to most people to be a reasonably significant impact on the final number, and I think it's that context we always have to keep in mind.

Therefore, with respect to the degree of “accuracy” or “error”, clearly, when I'm advising internally, just as when we are advising the minister not about what the numbers ought to be but what he ought to do with the budget itself and whatever surplus may be there, my input is one of many. Within the bank, the bank would use my economic forecast as a critical input in doing its own planning, let's say, for the coming year while acknowledging there are risks associated with the forecast. The exchange rate might be higher or lower than anticipated, which would affect our revenue flows or the value of them from the U.S.; interest rates might be higher or lower than expected, and growth might be higher or lower than expected. That has to be part of the decision making.

Similarly, in looking at the actual point estimates that are in the projections by the government, irrespective of how they are arrived at, we have to understood that there are necessarily risks associated with them even if you're only forecasting out one year, let alone if you're forecasting out two years or five years ahead. That is part and parcel of doing projections.

It seems to me one of the key issues is, can we, as we translate from the economic growth forecast... I suspect we can't make it any more accurate than we do. If we could, then we would have already done it as economists. Given that degree of error, can we made the fiscal forecasts more accurate or can we at least have a clearer understanding of why there can be variances between what we project and what we actually see as an outcome.

º (1650)

Ms. Rona Ambrose: Thank you.

Mr. Charlie Penson: I would like to use the remainder of the time, Mr. O'Neill, to show the difficulty we and the industry have. I agree with you, but I think I read or heard in an interview you gave just recently that the spending side is a difficult side. In a CBC interview you did yesterday you left it dangling—at least, you said the industry has been reasonably accurate in forecasting the economic growth side but the program spending side is maybe not so accurate.

I just want to point out, Mr. O'Neill, that in the minister's economic statement he gave on Tuesday he was continuing to forecast direct program spending growth of 5.7%. Now, this is on a $70 billion amount. However, “The Fiscal Monitor” that was released in September showed the first six months of spending growth at only 1%. This is what I think we're all getting at here, that there's a very serious problem with overestimating the spending, which contributes to that budget surplus, which some people may think is a great thing. There seems to be an example right there of how far out it's been for six months of this year.

Would you have any thoughts on whether we are going to have a tremendous amount of growth in direct program spending? It goes back to what Ms. Ambrose was talking about. Basically, in my view, the projections your industry can do are only as accurate as the information you're provided with. Can you enlighten us a little bit on that?

Mr. Tim O'Neill: If I understand the numbers you've just cited, it sounds to me like it was an overestimation of spending. That's what you're suggesting—

Mr. Charlie Penson: That's correct.

Mr. Tim O'Neill: —and that so far this year the numbers have come in lower than had been projected back in the budgets.

º (1655)

Mr. Charlie Penson: Mr. O'Neill, that is consistent with the pattern of the last seven years, I would suggest.

Mr. Tim O'Neill: Again, because the work isn't in the final stages, I can only tell you what I observe from the data that are currently available. In any given year, if you look at the variance between the projection and the actual, you'll see there does not seem to be a systematic single factor. In other words, in some years what's a more important explanation of the variance is that the revenue growth is higher than projected, which may have to do with economic performance. In other years it's the spending numbers that are out. For example, in the past fiscal year the dominant reason for the gap between projection and actual was revenue growth. In the previous year the dominant explanation was spending growth.

Mr. Charlie Penson: Are you doing that type of analysis in what you're putting forward?

Mr. Tim O'Neill: That will be a central part of this first section I described in my outline. What are the actual numbers, and is there a pattern to them that is clearly evident from the data? And whether the answer is yes or no, we look at each of the years to see what the contributing factors were and then ask, for example, was 80% of the variance over the last seven years the result primarily of the economic forecasts underestimating what growth would be, was it 60%, or was it 90%?

Mr. Charlie Penson: I think the committee would be interested in having that information, once you get it completed, as soon as you can, Mr. O'Neill. That would be helpful to us.

Mr. Tim O'Neill: Of course.

As I say, that's a central part of the analysis. Beyond that it's, what are the implications of that? Can we make changes in the process we use that would help improve that, and if the answer is no, can we at least have it clear why we're faced with this issue? Then all members of Parliament and all Canadians can at least understand where these variances have come from. That's the transparency part of it.

Mr. Charlie Penson: Thank you.

The Chair: Thank you, Mr. Penson.

Monsieur Côté.

[Translation]

Mr. Guy Côté (Portneuf—Jacques-Cartier, BQ): Thank you, Mr. Chairman.

Mr. O'Neill, your presentation is very refreshing compared to that given to us by Minister Goodale.

You mentioned earlier that the average variance between projections and real numbers was $4 billion. Is it since 1994 or 1997?

[English]

Mr. Tim O'Neill: The specific numbers I was citing were since 1997.

[Translation]

Mr. Guy Côté: In your presentation, you said that the massive use of contingency funds have an impact on the budget flexibility of the government. In the Minister's forecasting, we can see that in the next five years, we will have repeated surpluses and contingency funds will increase year after year.

Is the present budget process still appropriate in that context of repeated surpluses or does it not go against democracy as my colleague said earlier? This way, the Minister can use those funds without any public debate, as he sees fit.

[English]

Mr. Tim O'Neill: I think it's fair to say the contingency reserve and the prudence have been important. I personally as an economist have supported their use in the budget—in the budget-making process, as opposed to the forecasting, because obviously it's a decision by the government to include or not include.

There were really two reasons for that. One is, certainly in the last five years we've gone through a period when we've had a number of shocks to the global and domestic economic system that might have been expected to have significant influence on growth and the performance of the economy. I think it was critically important to have a reasonable amount of caution—call it contingency or prudence—in the budget-making process.

I think it's a fair question and one I certainly will be looking at, precisely the question you asked. I don't want to make a final judgment at this point, but the question is, do we need it going forward, and if so, what would be the logic for it? As you know, up to this point the contingency reserve has always been used—because it's still there at the end of the year—for debt reduction. Canada was a high-debt economy, a high-debt country, and I think it served us well to be able to pay down the debt directly.

I'm talking about my personal views now as an economist. I have been supportive as an economist in my discussions with the minister and with his officials of allocating some portion of any surplus revenue to debt reduction, some of it to spending, and some if it to tax cuts. That's my own personal view.

In this analysis, though, the key issue, I think, for any government, under any set of circumstances, is whether it is appropriate to have some caution formally built into the budget, and if so, how much that should be. Part of it has to do with the kinds of shocks we might face, but part of it may also have to do with the fact that if over time we are changing the degree of discretion we have in the budget—the discretion to make changes in the middle of a fiscal year, even—and if we're facing an environment where the shocks may make us fiscally more vulnerable, then it's an open question whether more or less is needed than we're currently using.

My own preliminary view, and I can be persuaded by evidence and argument to the contrary, is that we should have some caution, some contingency, some prudence built in to all budgets. That's simply because I think the vagaries of economic forecasting are such that unless I were confident we could get the numbers precisely right, I'd want the government to be prepared to make, or be capable of making, an adjustment to things they couldn't have anticipated.

» (1700)

[Translation]

Mr. Guy Côté: We can see that the Minister of Finance is using the average of the private sector's projections to base his forecasting but most negative impacts are amplified in those projections. Don't you think it might be one of the main reasons for the huge variances we have seen in the last seven years?

I know that you mentioned it several times, but I still find it very difficult to understand how a $1.9 billion surplus could become a $8.1 billion surplus or, like this year, a $4 billion forecast did translate into $12.8 billion if you add the amounts allocated for health. It is difficult to understand and I must say that when I go back to my riding, I find it very difficult to explain to my constituents.

[English]

Mr. Tim O'Neill: I understand the challenge. In fact, that's really one of the central purposes of the work. What I want to do is provide as much fact and analysis as I can. You know I'm at the early stages of this. You can appreciate that I'm not anywhere close to being able to draw firm conclusions, even about the analysis of why the gaps or variances have occurred. But certainly I want to get that work done before drawing any sort of implications or any conclusions and therefore any recommendations from it.

[Translation]

Mr. Guy Côté: If I have some time left, I'll give it to Mr. Loubier.

The Chair: You have exactly three minutes left.

Mr. Yvan Loubier: Mr. O'Neill, you are at the analysis stage. I Know that you have an extraordinary mandate that will end in February, which does not give you much time. I hope that you will be able to spend some time with your family during the Christmas Holidays.

I would like you to take into consideration, within your mandate, the fact that the independent forecasters in the US or Europe have access to data from the Departments of Finance and Revenue. I wish to know if those independent forecasting offices have access to more refined data than those available to our private sector economists which have been consulted in the past. I wish you can make that analysis and take it into account in your recommendations if you suggest another forecasting structure for Canada.

I also would like you to consider, in your analysis, the possibility of a continuous process all year round. There might be an estimation of revenues and expenditures four times a year. For example, three months after budget, there might be a follow-up on surplus level etc., until the next budget. It would allow us to make some adjustments. It would also allow us to adjust the process and have a clearer idea of new opportunities in the management of public funds.

Finally, I wish you good luck in your mandate. I found your comments very interesting and I hope that you will have enough time to answer all questions, because your mandate seems extremely vast and complex to me.

» (1705)

[English]

Mr. Tim O'Neill: Thank you for your best wishes and for your suggestion about the frequency of reporting.

I would simply note that when the last review was done, one of the suggestions was precisely for the process we have now, which is the minister reporting in the fall update. Certainly one of the things I will want to look at with respect to the transparency side of the issue is what is an appropriate frequency and what is the nature of those reports.

So I thank you for your suggestion. I'll certainly look at that very carefully.

[Translation]

Mr. Yvan Loubier: Thank you very much. And do not forget the access to government data.

M. Tim O'Neill: Yes.

The Chair: Thank you, Mr. Loubier. You had 20 seconds left. I think this is the first time you do not use up all your time. I shall use it for you.

Mr. Yvan Loubier: Our witnesses are very interesting. The Minister is not very interesting, mind you. Mr. O'Neill still has a lot of credibility.

The Chair: You have been very moderate with Mr. O'Neill.

Madame Minna.

L'hon. Maria Minna (Beaches—East York, Lib.): Thank you, Mr. Speaker.

[English]

My questions are very simple and short.

I've been here a while, and the main issue we've been discussing and that keeps coming up at this committee is not so much how forecasts are done, but the accuracy of them or how close you can get to the actual surpluses. Then, of course, the other debate is what you do with them when you have them—but not knowing them is not being able to....

The question is twofold. One is—and to some degree maybe it's not a question to you—when there is a surplus after the year ends, the assumption or understanding is that according to accounting procedures it must go to the debt. My understanding from some things I've read with respect to the Auditor General is that she says that's not really quite true. Would we require regulatory changes to actually do that and use the money for other things than the debt? Or is it a fait accompli, and therefore that we have to be absolutely accurate up front—which isn't always possible?

It's a question I've raised for many years myself, so I do it again.

Mr. Tim O'Neill: Clearly, I won't try to give you a legal opinion about the statutory nature of their crime; I would simply observe that there have been years in which the government has had year-end surpluses larger than those that were to be set aside for debt reduction. They have used them for funds or programs such as the Millennium Scholarship Fund. I'm assuming that the government therefore has in the past at least determined that there are other uses for a year-end surplus larger than anticipated, besides debt reduction.

If you're asking me whether it should be used for something else, I'll take off my hat for this mandate and put on my hat as the chief economist who provides advice to the finance minister. My own inclination is to use it for debt reduction, but I do understand that there are certain types of programs that may be regarded as effective and useful for a purpose that I would strongly support, which is improving the productivity of the Canadian economy. If we can find those kinds of uses, I think there's some value in that. It doesn't have to be done only out of a year-end surplus. It could actually be a part of the operating activity of the government. But I'll perhaps leave that to another day.

» (1710)

Hon. Maria Minna: There's no question that it can be done as part of the operating part of the budget. The only difficulty is of course when you have a projection that is slightly off—and I'm not suggesting necessarily there's any fault—and you end up with more than you anticipated, that the assumption now is the whole amount automatically goes to debt reduction. My assumption was that's not necessarily so. I'm not talking about the $3 billion and $1 billion for prudence and contingency. I'm talking over and above that.

The Auditor General seems to suggest that's not necessarily so either. I'm just wondering whether we actually need a regulatory change for it to happen. The minister says according to proper accounting procedures it should go to debt. The Auditor General seems to suggest, not necessarily so.

That's what we're trying to get at, in other words. It may be that our problem is not so much trying to fine-tune to the penny the surpluses or lack of, because that may never be possible to that extent, but to actually address what we do with them when we get them, and flexibility that we have with them, that would be...

Mr. Tim O'Neill: I understand the nature of your question. I just don't think I have the expertise to tell you whether you need a regulatory change in order to do that.

Hon. John McKay: On a point of order, Mr. Chair, I don't want to interrupt my colleague's line of questioning, but I don't want to get too far down this road either. As I understand it, in order for the government to do anything other than take surplus moneys and apply them to debt, the legislation has to already have been passed by the fiscal year-end, March 31.

You just can't, after the fiscal year-end, dream up some money to spend it to, etc.

The Chair: I'm not sure if that's a point of order, but I'd like to give—

[Translation]

Mr. Yvan Loubier: I rise on the same point of order as Mr. McKay.

That was done before in Quebec. We had forecasted a $900 million surplus. We passed a legislation before March 31st and that surplus was distributed between several priorities. That money was deposited in trust for two or three months and managed by the Quebec's Caisse de dépôt et placement until we could hold a debate on those priorities. Then, that surplus money was spent.

So it is possible, but a bill must be tabled, as you said.

[English]

Hon. John McKay: If it has already been passed.

The Chair: Basically, where there's a will, there's a way.

Ms. Minna.

Hon. Maria Minna: Yes, absolutely. Just a couple more. Sorry, I didn't mean to create a… I understand that.

You mentioned in there comparative analysis that's going to be done with the various OECD groups. What's the timing of that, do you know, in terms of when we'd be able to see some results?

Mr. Tim O'Neill: The IMF group that is working on it hasn't been able to provide a firm commitment to a specific date, but the intent is to have an initial draft sometime in January. It may very well be that their final report will actually be available only after I've completed my work. But at least I'll have access to their preliminary findings, and that will certainly be reflected in the report that I put together with the recommendations to the minister.

Hon. Maria Minna: One thing I've been curious about for some time is the fact that the Centre for Policy Alternatives seems to have come much closer to being on the button than we have in pretty well all of the last forecasts. Are you familiar with their methods, as opposed to the system we use, and can you tell us why they are so much more accurate as to the actual surpluses?

Mr. Tim O'Neill: I'm not certain. I know they have a dramatically different process, and that's why I want to check it out.

Hon. Maria Minna: They've been right on the button almost every time—well, every time, actually. Ever since we've had surpluses, they've been accurate and we haven't.

Mr. Tim O'Neill: And I do understand that.

Hon. Maria Minna: I was just curious if you knew what system or method they used to get there.

Mr. Tim O'Neill: I don't know it in detail, and that's part of the conversation I'm going to have with them.

[Translation]

Mr. Yvan Loubier: We did the same thing.

[English]

Mr. Tim O'Neill: I can't answer the question at this point, but I will be able to answer the question once I've done the consultation work with them, because that's one of the parts of the work I plan to do.

Hon. Maria Minna: So you're talking to them. Very good.

Thank you very much.

» (1715)

The Chair: Thank you. Thank you, Mr. O'Neill.

I just have a quick question, because the committee is also going to do some work on the independent fiscal forecaster. What I'm saying is, I want to make sure we don't double up or do parallel work. I understand you're going to the Netherlands and will look at what the IMF is going to do and what the U.S. is going to do. If I may ask, do you have any suggestions as to what we should be looking at, so we don't double up?

Mr. Tim O'Neill: If that's the sole focus of the work you're going to do, that is, to look only at the issue of whether or not we ought to have an independent budget forecasting agency, then I think you're invariably going to look at the same things I'm looking at. There may be doubling up as a consequence of that, but your focus may be slightly different from mine. Your focus may be on parliamentarians and how they are involved in and, to use Mr. Loubier's question, have access to information. I'll certainly look at that, but my focus will be somewhat more on how the process got started, what value it has had in terms of better forecast accuracy, and in addition to that, what it has done with respect to transparency. If I understand correctly, your focus is more on that broader second issue and less on the first one; it's more on what kind of access it provides us. You may end up looking at the same things but talking to different people.

The Chair: I think we're looking at a mixture of both, because we want parliamentarians to be involved but we also want reliable figures. I don't think we can expect that parliamentarians—this is my opinion again—will be able to get to the numbers unless we have proper information.

Anyway, I thank you for your time.

Monsieur Loubier.

Mr. Yvan Loubier: We could go together to the Netherlands.

Some hon. members: Oh, oh!

The Chair: I'll give everybody a chance to go.

Mr. Penson.

Mr. Charlie Penson: Mr. O'Neill, I think the question the chair never got to was just to have you confirm that your report, when you present it to the minister, will also be available to us. I don't think it was quite clear how you intend to handle that.

Mr. Tim O'Neill: Well, I'll be quite honest with you. The mandate was requested by and given to me by the finance minister initially, and that was my reporting responsibility as far as I knew. But I'm inferring from something he said to you here on Tuesday that he anticipated that I would also be providing the report to you, and presumably, if you then wanted to discuss its contents, its implications, etc., there would be an opportunity for that as well. But I would assume you are driving that process more than I am.

Mr. Charlie Penson: So your mandate as it exists now is only to report to the minister.

Mr. Tim O'Neill: As it was laid out for me, that's correct.

The Chair: But as chair, I can make a request and get a copy of that report.

Mr. McKay.

Hon. John McKay: Again, so we don't get too far down the road on a sidetrack here, I want to make sure we all understand that the Quebec legislation Mr. Loubier referred to would not be acceptable to the Auditor General. The Auditor General has apparently already ruled on this particular form of spending after the end of a budget year. I just don't want members to go off thinking this is a simple thing to do.

The trouble is, records get created, people start believing things, and momentum develops. Well, talk to the Auditor General first.

The Chair: If I may, I'll answer that. At the end of last year we as a government made a promise for $1 billion for BSE. That's the type of thing that could be done at the last minute.

Hon. John McKay: But that's allocated to the Department of Agriculture.

Mr. Charlie Penson: We have a witness here who's said he doesn't have knowledge as to how it works in that area. I don't think we should be discussing this at this committee with him here today, because we can do that at some future date. He's already told us he's not sure what legislative process is necessary to do this.

Ms. Ambrose has a question, I think, for the witness before he leaves.

» (1720)

The Chair: Ms. Ambrose.

Ms. Rona Ambrose: Thank you, Mr. Chair.

Obviously you've been brought in, as you've made allusion, for a very specific job, and that's to review the budgetary process and make recommendations to us and to the Minister of Finance. But the public perception and the assumption are that you've been brought in because there's a problem with the government's fiscal forecasting. I know you've talked about this in many ways, but you've done it in a much more quantitative way, and I wonder if you could address it in a more qualitative way. I would be happy with even just a yes or no answer. I'm wondering if you can tell me to what degree or if you believe there is a credibility problem with the government's fiscal forecasting.

Mr. Tim O'Neill: Those kinds of questions are almost never amenable to a yes or no answer, quite frankly.

Ms. Rona Ambrose: I thought you might say that.

Mr. Tim O'Neill: The key point is whether I think or don't think there's a credibility problem. Clearly, the government has decided it wishes to examine the whole issue of forecast accuracy and has asked me to do that. Part of that analysis is to determine how transparent the process is.

It may well turn out that when all is said and done, after comparing Canada to other countries, looking at our own track record, and looking at the normal difficulties of forecasting, I may conclude that if there's a problem of some sort with the variance between projections and actual, it may not be possible to do a great deal about changing that significantly. But I haven't reached that point yet, and certainly I ask your indulgence until I've had an opportunity to do it and draw appropriate conclusions. Then I will certainly be glad to discuss that with you and come, if it's possible, to some meeting of minds on where we should go from here.

Ms. Rona Ambrose: Thank you.

The Chair: Thank you, Mr. O'Neill. Again, thank you for taking time out of your schedule. The members appreciate it, and I hope we'll be able to come up with good reports, ours and yours, and maybe come to some kind of compromise. Thank you again, and good luck in your endeavours.

The meeting is adjourned.