Stephen S Poloz: Toward 2021: the power - and limitations - of policy

Remarks by Mr Stephen S Poloz, Governor of the Bank of Canada, to the Chamber of Commerce of Metropolitan Montreal, Montreal, Quebec, 21 February 2019.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Central bank speech  | 
22 February 2019


As the country's central bank, our responsibilities are spelled out in the Bank of Canada Act. Its preamble says that our job is to "promote the economic and financial welfare of Canada"-a lofty assignment.

Few mention the phrase that comes right before that: "so far as may be possible within the scope of monetary action." Even in 1935, the drafters of our legislation recognized that monetary policy cannot do everything. There are limits to its power. I will talk about three of these limitations today.

Our understanding of these limits has evolved with experience. Forty years ago, it was widely believed that monetary policy had very limited power. Today, especially after the global financial crisis, people seem to have a pretty high level of confidence in the power of monetary policy.

Every five years, the Bank undertakes a wide-ranging research and public consultation process to recommend to the government how we should carry out our responsibilities. We will make our next recommendation in 2021. So now is a good time to take stock of our understanding of the power-and limitations-of monetary policy. That is what I aim to do today.

The power of inflation targeting

For more than 25 years, the Bank has focused its monetary policy directly on keeping inflation low. For someone who remembers well the inflation of the 1970s, this focus makes perfect sense. But fewer and fewer people remember how difficult those times were. Inflation was not only high-much higher than today-it also jumped around from year to year. For both households and businesses, making financial plans was a very risky exercise.

The transition from an inflation-prone world was not an easy one. Central banks needed to break inflation psychology by raising interest rates dramatically during 1979-81. In 1981, when I arrived at the Bank, the average posted rate for a five-year mortgage in Canada peaked at over 20 per cent. Can you imagine paying 20 per cent for a five-year fixed-rate mortgage? Today, most mortgages of the same terms have an interest rate under 4 per cent.

To prevent future outbreaks of inflation, in 1991, the Bank of Canada became one of the first central banks to adopt the policy framework known as inflation targeting. Many other major central banks soon followed.

To put things very simply, the Bank influences inflation by lowering or raising interest rates to heat up or cool down the economy. The idea is to adjust rates to balance total demand and supply, thereby stabilizing inflation. Low interest rates stimulate demand by encouraging people and businesses to borrow, spend and invest. High interest rates do the opposite.

In practice, inflation targeting is much more complex and difficult than this. Many forces that affect demand and supply in the economy lie beyond our control. What is more, our policy actions take time to have their full impact-up to two years. This means that we are always working with forecasts of where the economy and inflation will be in two years, and attempting to influence those things in the future.

A crucial feature of this framework is that we have just one policy instrument: our influence over interest rates. This represents the first major limitation on the power of monetary policy. With only one instrument, we can aim at only one objective. If inflation is our objective, we cannot use interest rates to target other things, such as the exchange rate or the unemployment rate, at the same time. Of course, that is not the same as saying that interest rates do not temporarily affect other things, such as economic growth and employment, as we go about targeting inflation. But ultimately, inflation is the sole target of the policy.

In the late 1980s, when inflation targeting first came under active consideration, many were skeptical that it could work at all. As it turns out, it has worked very well. Inflation expectations have become firmly anchored on our 2 per cent target. A symptom of our success is the fact that many people do not appreciate how problematic high and variable inflation and interest rates can be. This is a gift to the next generation, if you will. My children will never pay anything like the kind of interest rates I have paid in my lifetime.

It is not just that inflation and interest rates have been lower and more stable. With inflation expectations well anchored, monetary policy is more effective. The economy adjusts more quickly to shocks than in the past. Firms and households can make longer-range plans. Wage negotiations are much simpler. Economic cycles are less severe. Unemployment is lower on average and less variable. These observations are based on global experience from the early 1990s to the mid-2000s, a 15-year tranquil period called the Great Moderation.

All that success has led many to conclude that keeping inflation low, stable and predictable is the best contribution monetary policy can make to the economy. It sounds less ambitious than the preamble to the Bank of Canada Act, but the results have been very powerful.

Nevertheless, the widespread belief that keeping inflation low and stable would keep the economy out of trouble proved to be misplaced. The global financial crisis laid bare the second important limitation of monetary policy: low inflation can still lead to a buildup of dangerous imbalances in the economy. In effect, prolonged economic and financial stability can lead to excessive risk-taking and rising leverage. Regulatory complacency also played an important contributing role.

Imbalances can make the economy fragile. When a shock hits the economy, its effect can be amplified by these imbalances, and the central bank may have difficulty achieving its inflation target for some time. In other words, low and stable inflation may be necessary for sustainable economic growth, but it is not sufficient on its own. More on this later.

The perils of low for long

Policy-makers, both monetary and fiscal, have worked extremely hard for more than a decade to nurse their economies back to health.

By 2014, we appeared to be within a year of returning "home"-by which I mean the point where the economy is running at full capacity, with supply and demand in balance and inflation at target. When the economy is truly home, it can grow sustainably without policy stimulus.

Unfortunately, the economy hit a major setback late in 2014, when oil prices collapsed. The Bank lowered its key policy interest rate twice in 2015, to a very stimulative level of 0.5 per cent, to help the economy work through the shock of lower oil prices. The government also introduced new fiscal stimulus. Over the next two to three years, the economy gradually made its way back home, absorbing excess capacity and seeing inflation return to target. In this context, the Bank began withdrawing its extraordinary stimulus, raising interest rates by a total of 1.25 percentage points, to 1.75 per cent. But with that rate still lower than inflation, it is clear that monetary policy continues to deliver stimulus to the economy today.

Indeed, monetary policy has been stimulating the economy for much longer than anyone expected when the crisis began back in 2008. And we have seen the natural results of leaving interest rates very low for a long time. For one thing, this has been hard for people, such as retirees, who rely on interest from their savings for their income. Further, people have taken on a lot of debt, mostly in the form of mortgages and home equity lines of credit. By 2017, the ratio of household debt to disposable income had hit a record-with the average household owing more than $1.70 for every dollar of disposable income. If we remove households that do not have mortgages, the ratio becomes much higher-close to $3 for every dollar of disposable income. And house prices were rising extremely quickly in some of Canada's biggest cities.

This brings us back to the second major limitation on the power of monetary policy I mentioned above. When low interest rates persist, debt can reach levels that become risky for the borrowers and for the entire economy. When households carry a lot of debt, they become less able to manage through a temporary period of unemployment. The effect of shocks can be magnified when they interact with elevated debt. This can cause the economy to underperform for extended periods.

Let me illustrate by describing what might have happened had we not started to raise interest rates over the past couple of years as the economy approached home. If we had kept our interest rate at 0.5 per cent from mid-2015 until now, our models tell us that we would have seen stronger economic growth-no surprise there. By now, the level of GDP would be about 2 per cent higher than it is today. That sounds good, but inflation would also be higher-it would be close to the top of our 1-3 per cent target range, and very likely would be heading even higher. Faced with that prospect, we would need to raise interest rates forcefully to guide inflation back to its target over the next year or two.

Forceful increases in interest rates would pose problems for households carrying a lot of debt, of course. Indeed, if we had left rates down at 0.5 per cent, that debt load would be even greater by now, over $60 billion higher, or about $2,000 in additional debt for every Canadian. And it is certain that house prices would be even higher today-our models suggest by around 5 per cent nationally.

This scenario shows how monetary policy has the power to affect not just growth and inflation in the broader economy, but also financial vulnerabilities. In other words, in pursuing our inflation target, we may create side effects that make the economy vulnerable to new shocks. This is what happened during the Great Moderation, and it is what has happened during the recovery from the global financial crisis. Of course, the fact that we have only one policy instrument means that we cannot independently try to manage those side effects without putting our inflation target in jeopardy.

The implication is that policy-makers need additional instruments to address these side effects. To this end, Canada has developed a number of what we call "macroprudential policies." These measures include new rules for mortgage borrowing, the so-called B-20 guideline, for example, which was implemented by the Office of the Superintendent of Financial Institutions. The aim of these measures has been to reduce the economy's vulnerability by ensuring borrowers will be able to manage their debt, even when interest rates rise. Let me stress that the goal of these measures has always been to improve the quality of new household debt, not to slow down housing markets. That said, we have also seen various provincial and municipal governments put rules in place that are aimed specifically at containing house price growth.

To help guide the use of macroprudential tools, we need a way to measure the economic importance of financial vulnerabilities as they rise and fall. This would allow us to account for these side effects of monetary policy.

Recently, the Bank has started using a framework that estimates the economic growth that is at risk from financial vulnerabilities. When we forecast economic growth, there is always a range of possible outcomes around that projected path. Rising financial vulnerabilities make it more likely that an economic shock will cause the economy to follow a much weaker growth path, as the shock interacts with high levels of debt, for example. The gap between our most likely scenario and the possible adverse scenario where financial stability risks are realized is called "growth at risk." We can estimate how much economic growth is at risk with our models.

Let us go back to the hypothetical scenario where we simulated keeping our policy rate unchanged at 0.5 per cent from 2015 until now. As I said before, economic output would be higher, but inflation would be approaching
3 per cent by now and likely continuing to climb. Household debt and house prices would also be higher, which would mean the economy would be more vulnerable to a new economic shock. The downside risk from these vulnerabilities to economic growth-and therefore to inflation as well-would be rising throughout the scenario.

Fortunately, we did not follow this scenario. Both macroprudential and fiscal policies were also put in place, and this produced better outcomes.

It is important to understand the role that expansionary fiscal policy played. Our models show that if fiscal policy had remained unchanged following the collapse in oil prices in late 2014, the Bank would have had to set interest rates as much as 50 basis points lower during the 2015-18 period to get the economy back home. Instead of approaching full capacity in mid-2017, it would have taken until the end of 2018 for the economy to get home with inflation on target. This demonstrates that, at low interest rates and high debt levels, monetary policy has less power than fiscal policy in stimulating the economy.

Meanwhile, that lower track for interest rates would have led to even higher household debt and house prices. Our growth-at-risk framework shows that this would have meant an even greater risk that a new economic shock would cause the economy to fall into a low-growth scenario with inflation below target.

The bottom line is that the mix of monetary, fiscal and macroprudential policies matters. In certain circumstances, relying less on low interest rates to bring the economy home can mean a more resilient economy.

Knowledge limitations

Even though it can be hard to imagine counterfactual scenarios, these model simulations demonstrate some of the trade-offs that policy-makers face. They also highlight the third major limitation of monetary policy-how uncertainty limits the ability of the policy-maker.

Complex economic models are indispensable tools for economists, particularly those who work at central banks. However, these tools are all based on historical averages and embrace several simplifying assumptions that may not hold in reality. Every relationship in the Bank's models is meant only to approximate how part of the economy will behave, on average, over time.

The importance of this uncertainty is magnified by the fact that our policy operates in the future. Actions we take today will have their ultimate effects only in the next couple of years. In hockey, we must pass the puck not at the feet of our teammate, but to where he or she will be when the puck arrives. The sense of motion in the economy adds to the uncertainties we face as modellers and forecasters.

The reality is that conducting monetary policy requires a lot of judgment. Rather than the finely tuned mechanical process many people imagine, policy is much more akin to an exercise in risk management. As we said at the Bank's most recent interest rate announcement, we judge that we will need to move our policy rate up into a neutral range over time, to a point where it is not stimulating or constraining economic growth. However, the path back to that neutral range is highly uncertain. We will watch the data as they come in, and use judgment to deal with the uncertainties and manage the associated risks.

One important uncertainty that we are dealing with today is the impact of higher interest rates on highly indebted Canadians. Rising interest rates will mean these people will have to spend more of their income servicing debt, leaving less for other goods and services. Clearly, given these elevated levels of debt, raising rates will have more of an impact on the overall economy than in the past. This is one reason why we have been gradual in our approach to raising interest rates.

Housing markets are adjusting not only to higher interest rates, but also to new mortgage guidelines and rules aimed directly at cooling certain housing markets. Given the unique situation, we are monitoring the impacts carefully. Housing activity has been a little weaker than we expected recently. Mostly it is housing resales that have been soft, suggesting there may have been more froth in certain housing markets than previously thought. Housing markets that were not experiencing bidding wars appear to be adjusting in line with our expectations. However, more data will help us better understand the full situation in Canada's housing market.

A second area of intense interest is the outlook for business investment, which has been less robust than our models indicated for the past couple of years, mainly due to uncertainty about the future of the North American Free Trade Agreement. Although uncertainty remains around ratification of the new Canada-United States-Mexico Agreement, we expect investment spending to regain momentum in 2019, especially in light of the government's new accelerated capital depreciation rules. However, we must acknowledge that the future of the global trade environment is highly uncertain right now. An escalation of the US-led trade war would, of course, be a negative for the outlook, but a resolution would be a source of new lift for the global and Canadian economies.

Given these uncertainties, we have kept interest rates unchanged at 1.75 per cent since last October. We will remain decidedly data-dependent as the domestic and international situations evolve.


It is time for me to conclude.

We have learned over time just how powerful monetary policy can be. Canada has more than a quarter century of experience with inflation targeting. And this experience has taught us that using monetary policy to bring about low, stable and predictable inflation is the best way for the Bank to meet its responsibilities.

Of course, there is more than one way to keep inflation low and stable. Between now and 2021, when we next renew our inflation-control agreement with the government, we will be doing in-depth research and wide-ranging consultations on alternative policy frameworks.

Meanwhile, we must never lose sight of the fact that there are limitations on the power of monetary policy. We have only one instrument at our disposal. History proves that even a highly successful monetary policy can generate harmful side-effects. And uncertainty intrudes everywhere in monetary policy practice. The development, and future refinement, of macroprudential policies shows considerable promise in addressing some of these limitations. The rest may just be a matter of hard work and ingenuity.

As we head toward 2021, we will keep working hard on deepening our understanding of the power-and the limitations-of monetary policy. We are gaining important experience in seeing how various types of policies interact. I know we have some of the sharpest economic minds in the world, both inside and outside the Bank, working on this issue. And we will have lots more to say about it ahead of the next renewal.

I would like to thank Jing Yang for her help in preparing this speech.