Transparency: The More, The Better?
Transparency is the cornerstone of a well-functioning financial system. It's an issue that has been getting a lot of attention, and deservedly so, as we consider what has gone wrong in the market for asset-backed commercial paper. I'll be happy to take any queries you might have on this topic during the question and answer session following my remarks. But first, I want to talk to you about transparency in central banking and in setting monetary policy. Then, I'll close my remarks with a review of the changing monetary policy outlook over the past six months, which serves to illustrate my key points.
So, what has been done, and what more can we do to improve transparency in monetary policy? Are there any limits? I'll approach these questions from three perspectives: transparency about the Bank's policy framework; transparency about the inputs, the processes, and the reasoning behind monetary policy decisions; and transparency about our assessment of the outlook for the economy and monetary policy.
That last element, disclosing more of the Bank's assessment about the outlook, including forward-looking statements about monetary policy actions, is particularly tricky and really tests the limits of transparency. Nevertheless, it is in this area that there may be the most room to increase transparency. But first, we must figure out if it would be beneficial to provide more information for market participants, firms, and individuals. More fundamentally, would it improve the effectiveness of monetary policy? And if we find that it would be beneficial, how can we convey this information so that it would be readily understood? Or more importantly, how can we convey this information in a way that will not be misunderstood?
At the Bank of Canada, we do not believe in constructive ambiguity, nor in saturating the market with a lot of information that has no clear message. Real transparency involves judgment: communicating what is important, what clarifies, and not what obscures. I'll have more to say on this in a few minutes.
Why Be Transparent?
But first, let's back up one step to look more closely at just why transparency is so important to the conduct of monetary policy. It has become very clear over time and with experience, that monetary policy is most effective when the policy objective is clearly understood and accepted. When consumers and savers, business owners, and financial market players, all understand the Bank of Canada's policy objective – and believe that it is attainable – then they can make better long-term plans and decisions. And when everyone expects this target to be maintained, and acts accordingly, then the target becomes self-reinforcing.
But the benefits of transparency don't stop there. I believe that transparency also helps us to make better decisions. The extra rigour that comes from holding the rationale behind our decisions up to external scrutiny leads to better results.
Finally, the Bank of Canada is a public institution – funded by, and accountable to, the taxpayer. Information and analysis gathered in the context of the Bank's business should be considered a public good – except when the release of such information would compromise the implementation of the Bank's mandate.
Are There Limits to Transparency?
In general, more transparency is better than less. But this does not mean that there are no limits. First, there is the need to protect the confidentiality of some information and analysis that is given to the Bank by outside parties, be they public- or private-sector institutions. Indeed, public release of third-party, confidential information would jeopardize the central bank's ability to get all the information it needs to make good monetary policy decisions.
Second, there is the need to protect the integrity of some internal policy deliberations. For example, the public release of policy advice and recommendations could stifle the free debate and consensus building that is necessary for sound policy-making. We want to hear all aspects of an argument, but it could be hard for staff members to play devil's advocate, knowing that such a position will be made public and could be taken out of context. And certain information should not be released while policy is still being developed, or has not yet had its full effect. In such cases, premature transparency could lead to misinterpretation or be acted on inappropriately, which could derail good policy intentions and could potentially be damaging to Canada's economic interests.
Third, there is the need for good quality information. Providing useful, relevant information is far more important than dumping a large quantity of information of questionable quality. Too much information can actually cloud what is important. As a result, actions and decisions can become less transparent, because the really important, effective, and relevant information gets lost in the minutiae.
Good quality information must not only be accurate, it must also be communicated clearly and simply, so that it won't be misunderstood. This is especially germane when we try to be transparent about our assessment of the outlook by publishing our economic projections, along with the risks and uncertainties surrounding them, and forward-looking statements about possible future policy actions.
Transparency in Setting Monetary Policy
Debates about the limits to transparency for effective monetary policy are certainly not unique to the Bank of Canada. Central banks around the world are still learning the best ways to communicate monetary policy. This is an important element of the art, rather than the hard science, of central banking. I believe that we at the Bank of Canada have made great strides in this area.
So let's now take a closer look at the issue of transparency in setting monetary policy from the three perspectives that I mentioned at the outset: transparency about our policy framework; transparency about the inputs, the processes, and the reasoning behind monetary policy; and transparency about our assessment of the outlook.
The Bank made a major leap forward in increasing transparency about our policy framework when we adopted an explicit inflation target in 1991. This target provides a clear objective for monetary policy, which has helped to anchor financial and economic decisions. It makes it easy to measure the success of monetary policy and to hold the Bank accountable for its actions. Canadian individuals and firms can align their savings, investment, and spending plans with a common inflation-control objective. If inflation persistently deviates from the target, we are committed to explaining the reasons why, what we will do to return it to target, and how long we expect the process to take. Previously, when we targeted monetary aggregates, for example, there was ambiguity about what the bank was trying to achieve, and we were not always clear about the implications of such a target for output, inflation, and interest rates.
Now, as well as being clear about our objective, we are also being transparent about our assessment of the factors that influence inflation and about how we implement monetary policy. We conduct monetary policy in a symmetric way, worrying as much about the trend of inflation falling below target as we do about it rising above target. To keep inflation on target, we try to keep the economy operating near its full capacity. When the demand for goods and services pushes the Canadian economy against the limits of its capacity, and inflation is poised to rise above target, the Bank will raise interest rates to cool off the economy. And when the economy is expected to operate below its production capacity, and inflation is poised to fall below target, the Bank will lower interest rates to stimulate growth. We also factor in shocks that directly affect inflation.
Because we target domestic inflation, we have a floating currency. A central bank cannot successfully control both the domestic and external values of its currency at the same time. And a flexible exchange rate is an important price signal of changing global and domestic circumstances that can help to prompt and facilitate necessary adjustment.
This is our paradigm for the conduct of monetary policy, and we have been so transparent about it that we have devoted entire speeches and published a great deal of staff research on the topic. We do this so that a careful observer can understand just how monetary policy will adjust to changing circumstances.
Our inflation target is established under an agreement with the federal government. When this is reviewed periodically, we look for ways to improve the conduct of monetary policy based on recent experience and research in order to clarify aspects of our framework. In the most recent review, for example, we examined how to deal with asset-price bubbles and looked at the appropriate time horizon for returning inflation to target following an economic or financial shock. In preparation for the next review of our inflation targeting agreement in 2011, we are conducting an intensive research program into possible improvements to our policy framework. Specifically, we are looking into the merits of a lower inflation target and price-level targeting.
Now, let's look at transparency about the inputs, the processes, and the reasoning behind monetary policy decisions. Just a little over a decade ago, the Bank didn't even issue a press release when it made an interest rate decision. Now, our rate announcements are widely anticipated. There is a very high level of interest, stretching from households to office towers, in what the Bank of Canada has to say on interest rates, inflation, and the economy.
This interest is at least partly the result of a series of measures taken by the Bank to increase transparency. After the introduction of the inflation target, we moved in 1994 to become more open about how we implement monetary policy by targeting the overnight interest rate. A further major step towards greater transparency came in 1995 when we began to publish a regular Monetary Policy Report, and later added MPR Updates, to provide a window for financial markets and the general public into the analysis behind our conduct of monetary policy. Soon after that, the Bank began to publicly announce interest rate changes through press releases. These steps helped to increase understanding and acceptance of the Bank's policy objective.
Still, there was a lot of ambiguity and uncertainty in financial markets about exactly when the Bank of Canada might change its monetary policy stance. So, in 2000, we decided to set eight fixed dates each year for interest rate announcements, regardless of whether rates were changed or not. This commitment to a timetable provided more certainty for markets, a better focus for external commentators to develop and expand their own views on the economic outlook, and regular opportunities for us at the Bank of Canada to review the accumulation of data and to update our views about the outlook for inflation and the appropriate course for monetary policy. This extra focus and rigor has, in my opinion, improved our decision-making process. Of course, the Bank retains the option of moving between dates in extraordinary circumstances, an option that has been exercised only once since fixed announcement dates were established.
Most of the information and analysis that the Bank uses when it makes monetary policy decisions are based on data that are also available to the public. There was a time when one such input – our Business Outlook Survey – was not publicly available. But after we conducted research to determine which of our survey questions provided the most useful information, we began to publish the results of those questions. The motivation was not only to help improve the public's understanding, but also to be as open and transparent as we could, given our responsibilities as a public institution. We continue to examine whether we can make public more of the inputs to monetary policy, without compromising the policy-making process. Indeed, in the upcoming January Business Outlook Survey, we will be including responses to two additional questions on past sales and credit conditions.
I'll now turn to the third perspective, which is transparency about our assessment of the outlook for the economy and monetary policy. Because it can take as long as two years for monetary policy actions to have their full effect on inflation, we must always be looking well into the future. So, the questions become: how can we effectively communicate the kind of uncertainty and the many risks that any views about the future must always include? How can we talk openly about possibilities and risks in a way that won't be misunderstood? It would not aid transparency if our assumptions and projections of what might transpire – assessments which are, by their very nature, conditional – were misconstrued as more concrete predictions or commitments.
At the Bank of Canada, the Governing Council sets out its base-case projection for inflation and growth in the Canadian economy four times a year in our Monetary Policy Reports and Updates. This base-case projection reflects Governing Council's best judgment about the most likely outcome, based on a number of assumptions. Over the years, we have become increasingly transparent in describing our projections, and the underlying assumptions. This January, we will include in our Update projection tables on global economic growth – tables that previously had only been included in full Monetary Policy Reports.
We have provided more detail about how we see the economy unfolding, what forces might affect inflation, and what assumptions we have had to make about more volatile or uncertain variables. This, in turn, helps to explain the reasons behind our most recent decisions and provides some insight into possible future actions.
The base-case projection embeds changes in the policy interest rate that would be necessary to achieve our 2 per cent inflation target over the medium term. We describe the direction and magnitude of that interest rate path in a few words, in Chapter 4 of the Monetary Policy Report and when we update the base-case projection in the Update. We may also give an indication of the time horizon for this path.
We have also become more forthcoming in recent years about the risks that we see surrounding the base-case projection and whether we think these risks are balanced or not. Our policy rate statements, in press releases and in the Overview section of the Monetary Policy Report and Update, reflect our best judgment in the context of the overall outlook at that time, including the balance of risks. But that, too, is no guarantee that the future will play out the way we expect.
There are several ways we might consider providing more transparency about the risks and uncertainties around our base-case projections. Some central banks use fan charts showing confidence intervals that can suggest the extent of uncertainty around a particular factor in the overall forecast. Another approach is to publish complete alternative scenarios in addition to presenting a base-case projection. And some academics have called for more transparency by sharing the monetary authorities' probability estimates of different risks. Examining some of these possibilities is in our work plan for the year ahead.
What is not in our work plan is the use of stock words and phrases to signal policy intentions. This isn't our practice at the Bank of Canada. Instead, we try to spell out the situation as we see it. We set out our base-case projection for the economy and the key risks; we make clear the indicators that we're closely following; and by being clear about our paradigm, we try to help a careful observer to better understand how monetary policy could evolve, depending on changing circumstances.
But we count on outside observers to do their own analytical heavy lifting as well. In preparing for our interest rate decisions, we look at market prices, external commentary, and other analysis. We rely on observers to provide their own perspective on the outlook. If our views differ substantially from the consensus of external observers, we might consider either redoing our own analysis, seeking out further information, or communicating more explicitly our views and analysis to help those observers to understand the factors that we consider to be important. As I have said on many occasions, if you are considering placing a bet between what you think we said we were going to do and what you think we ought to do, I'd go with the latter.
Outlook for the Canadian Economy and Inflation
So, let me conclude by looking at the evolution of monetary policy over the past six months. It certainly has been a fluid situation, and an excellent example of how forward-looking assessments can change quite rapidly.
Back in July, we raised our policy rate by a quarter of one percentage point because we judged that the economy was operating, at that time, at a level further above its production potential than had been projected at the time of the April Monetary Policy Report. As we expected both total and core CPI to remain above 2 per cent, we concluded that some modest further increase in the overnight rate might be required to bring inflation back to 2 per cent over the medium term.
By the end of the summer, the situation had changed in many respects. The Canadian economy was operating even further above its production potential than estimated in July, which was putting upward pressure on inflation. Developments in financial markets had led to some tightening of credit conditions for Canadian borrowers, however, which would temper the growth in domestic demand. And the U.S. economic outlook had also weakened. This, together with a higher assumed level for the Canadian dollar, suggested that there would be more drag from net exports in 2008 and 2009 than previously expected. Given these developments, we held our policy rate steady through September and October. We identified several risks to the outlook for inflation and judged them to be roughly balanced, with perhaps a slight tilt to the downside.
In the event, the downside risks prevailed. When we re-examined our monetary policy stance early in December, the Canadian economy was growing broadly in line with the Bank's expectations, reflecting in large part the underlying strength of domestic demand. But both total CPI and core inflation were now below the Bank's expectations, reflecting increased competitive pressures related to the level of the Canadian dollar. In addition, other developments since October suggested that the downside risks to our inflation projection had increased. Global financial market difficulties had worsened, tightening credit conditions further, and there was an increased risk to the prospects for the demand for Canadian exports, since the outlook for the U.S. economy had weakened further. So, on 4 December, we judged that the balance of risks had shifted to the downside, and we lowered our target for the overnight rate to 4 1/4 per cent.
Now, as we prepare for our next interest rate announcement on 22 January, and our Monetary Policy Report Update two days later, we are preparing our regular quarterly economic projection and risk assessments, so that we can fully assess the implications of all economic and financial developments since the October Report, and set the appropriate course for monetary policy.
At the Bank of Canada, we are very interested in any ideas you might have on this topic of transparency in monetary policy – what the limits are and how transparency can continue to be improved. In my view, generally, more is better. Apart from the need to protect confidential information and not compromise the policy-making process, the true limits to transparency in monetary policy lie in communicating effectively in the face of uncertainty, a skill which we are constantly striving to master.