Challenges in Central Banking

15 September 2012

An address at the University of International Business and Economics in Beijing

Ladies and gentlemen,

I very much appreciate your warm welcome to this great university, and the opportunity to again visit this great country.  I first visited China in 1986 and this is my tenth visit.  But the last time I visited was in 2006 and, with the extraordinary rate of growth in your country, much has changed in the last six years.  So thank you for this opportunity.

This is the second of four lectures I’ll be giving while I’m in Beijing on this visit.  Yesterday, I spoke about the causes of the Global Financial Crisis.  This one will focus on the three big challenges facing central banks today; the third will discuss the pros and cons of foreign direct investment; and the fourth will talk a bit about New Zealand and New Zealand’s relationship with China.

The relationship between government and central bank

So what are the three big challenges facing central banks today?   I believe that one of the key challenges is the kind of relationship which should exist between government and central bank, and this is the challenge I propose to spend most of my time on.

Prior to 1989 (and I’ll explain the significance of that date in a moment), there were just two kinds of relationship between government and central bank.  In most countries, the central bank was simply a part of the government apparatus.  It provided advice to ministers and implemented the decisions of ministers, but it did not make decisions itself.  Making monetary policy decisions was the prerogative of ministers.  This model of the relationship between government and central bank I refer to as the old Bank of England model — that was how the Bank of England operated prior to 1997, and that model was prevalent in a great many countries.

At the other extreme, the central bank was, at least in theory, completely independent of the government and the political process.  That was the model in Germany with the Bundesbank, prior to the creation of the European Central Bank; and it was (and still is) the model in Switzerland, with the Swiss National Bank.   It is also the model in the United States, with the Federal Reserve Board.

In my opinion, neither of these models is perfect.  The old Bank of England model, by leaving the control of monetary policy in the hands of elected politicians, created an almost inevitable bias towards higher inflation.  Why?  Because as I’m sure you know, monetary policy works on inflation with what Milton Friedman famously called long and variable lags.  In the short term, easing monetary policy — making more money available and reducing the cost of credit — results in increased economic activity and reduced unemployment, leading to increased popularity for the government.  By contrast, tightening monetary policy — restricting the growth in the amount of money available and increasing the cost of credit — initially results in reduced economic activity and increased unemployment.

So in countries which have elected governments, and indeed in all countries where governments want to maximise their popular support, politicians have a bias towards easier monetary policy.  

Unfortunately, in the longer term easier monetary policy leads only to higher inflation.  As a result, countries which have a real fear of inflation made their central banks very largely independent of the political process and charged them with the responsibility for maintaining the purchasing power of their currency.  The best example of this of course is the Bundesbank.   The German people have a deep-seated fear of inflation, stemming from the hyper-inflation which totally destroyed the value of their money after the First World War.  (Towards the end of 1919, 47 marks bought one US dollar.  By November 1921, it required 330 marks to buy a US dollar, and by December 1923 it required 4.2 trillion marks to buy one US dollar.  Through much of 1923, prices were doubling every two days.)   

The European Central Bank inherited this Bundesbank independence from the political process, and the stresses and strains which this is producing within the Eurozone are clear for all to see.  Some governments of countries within the Eurozone want the European Central Bank to be more aggressive in using its ability to create money to reduce the government bond rates in countries like Spain and Italy, while the governments of countries in the German tradition vigorously resist that pressure.

The Reserve Bank of New Zealand was explicitly created in the image of the old Bank of England.  The Bank provided advice to ministers and implemented the decisions made by ministers.  But it was the Minister of Finance who made the decisions.  And in line with other countries with central banks under tight political control, New Zealand’s inflation was relatively high — not German-type hyper-inflation of course, but much higher than inflation in many other developed countries.  The Minister of Finance was much more concerned about the re-election of his Government than he was about preserving the purchasing power of New Zealand’s currency.

In 1984, a Labour Government was elected to power in New Zealand, and one of its first actions was to direct the Reserve Bank to get inflation under control.  The Bank was effectively told to make its own decisions about how best to do that.
In April 1988, with consumer price inflation dropping into single figures for the first time in some years, the Minister of Finance was asked during a television interview whether he was now satisfied that inflation was under control.  He made it clear that he was expecting the Reserve Bank to get inflation down to zero, or very close to it.

I was appointed Governor of the Bank in September 1988 and it was made clear that my mandate was to get inflation to between zero and 2%, as measured by the consumer price index, by 1992.  The law still required monetary policy to deliver economic growth, low unemployment and low inflation, and allowed all monetary policy decisions to be made by the Minister of Finance.  But the Minister made it clear he wanted me to focus on getting inflation out of the New Zealand economy, and wanted me to make the decisions about how to do this.
In late 1989, after considering a range of alternative arrangements around the world and talking to some of the best monetary policy economists in the world, the New Zealand Parliament passed a new Reserve Bank Act.  Reflecting the latest economic thinking, and in particular the contribution of people like Milton Friedman in this area, the new law required monetary policy to “achieve and maintain stability in the general level of prices”.  It made no mention of economic growth or employment because by 1989 it was recognised that the best contribution monetary policy can make to growth and employment is to maintain the value of money — in other words, to keep inflation very low and stable.

Had that been all the new law did it would have marked a big step forward for New Zealand but it would have had no particular significance for the rest of the world.  But it did much more.  It created a totally new model of the relationship between government and central bank by explicitly providing for a role for the government while creating a high degree of independence for the central bank in implementing policy.

Specifically, the 1989 legislation provided that:

-    The Minister of Finance on behalf of the government would agree what rate of inflation he expected the Governor of the Bank to deliver during his tenure as Governor;

-    The Minister of Finance would have the right to override that agreement during the Governor’s term of office;

-    The agreement between Minister and Governor, and any override of that agreement, must be made public at the time it was made;

-    The Governor would be totally independent in making monetary policy decisions, and would make them without reference to the government;

-    The Governor must conduct monetary policy in a very open and transparent way, publishing regular reports on how he sees the inflation outlook, and therefore the prospect for monetary policy;

-    The Governor was liable to dismissal from office if he failed to keep the inflation rate within the agreed limits without valid reason (as judged by the Bank’s independent directors and by the Minister’s own advisers).

This was revolutionary stuff.  It provided for an explicit political involvement in the process, but required that involvement to be open and transparent, visible for all to see.  It gave the Governor complete freedom to conduct monetary policy targeted at the agreed inflation rate, but required him to conduct policy in an open and transparent way, and made him accountable for the inflation outcome.

Much of the public discussion at the time was focused on the Bank’s new independence from the political process, but in fact the real genius of the framework created by the 1989 legislation was mandatory transparency.

No government wants to announce its intention to debase its own national currency, so the inflation target agreed between Minister and Governor inevitably focuses on a very low rate of inflation.  The Governor in turn is obliged to explain to the public and to financial markets how he sees the economic outlook, how he sees inflation evolving, and what he plans to do about it.

As a result of the 1989 legislation, New Zealand became the first country in the world to have an explicit and public inflation target, and the Reserve Bank of New Zealand pioneered a new tradition of openness in the conduct of monetary policy.  Previously, central banking had been notoriously secretive.  Alan Greenspan famously made a virtue of speaking in ways which his listeners could not understand.   Today, most central banks, at least in developed countries, have an explicit inflation target (the US Federal Reserve Board is an exception, despite Ben Bernanke’s sympathy for inflation targeting1).

Fewer countries have adopted the New Zealand model for the relationship between government and central bank, and none have adopted it precisely.  The three countries which have adopted it in substance are, in order, Canada, Australia, and the United Kingdom.  In each of those countries, the Minister of Finance (or Chancellor of the Exchequer in the case of the United Kingdom) reaches agreement with the Governor of their central bank about the inflation rate to be achieved and this agreement is made public.  Only in New Zealand, however, is the Governor personally accountable for the actual inflation outcome.  In the other three countries, monetary policy decisions are made by a board or committee, so of course the Governor cannot be held personally responsible for the actual inflation outcome.  In New Zealand, it is the Governor’s job which is on the line!

(When the 1989 legislation was being drafted, I questioned the Minister why the proposed agreement about the inflation target was to be between the Minister of Finance on behalf of the Government and the Governor personally.  I was told that that was because, if the agreed inflation target was not achieved, it would not be feasible to dismiss all the staff of the Bank, or even all members of the Bank’s board, “but we sure as hell can fire you!”)

Has the New Zealand model worked?  It’s impossible to be dogmatic in answering this question, because of course it’s impossible to know what might have happened in the absence of the new model.  But personally I have no doubt at all that it worked.

We know that from the time the government gave the central bank the instruction to get inflation under control, and the de facto independence to deliver that result, inflation in New Zealand fell from well above the OECD average to a rate consistent with the original inflation target of zero to 2%.

Yes, there was a cost to that process — economic growth stalled and unemployment reached 11% of the workforce.  But that was only in part a result of the disinflationary process.  At the same time that the Government elected in 1984 had instructed the Reserve Bank to get inflation under control, it had also removed all quantitative import controls, reduced tariffs on imports, removed subsidies from the farming sector, and privatised many government-owned businesses.  All of those moves had adverse effects on employment in the short-term.

What we know is that within months of the inflation target being formally agreed in early 1990, the head of the New Zealand trade union movement campaigned on the need for wage increases consistent with the new inflation target if unemployment was to be minimised.
What we know is that in the early nineties, yields on New Zealand government 10 year bonds fell from levels well above the yields on US Treasuries to levels closely similar to the yield on US Treasuries (and indeed, briefly to a level below that of US Treasuries).

What we know is that over the following 20 years the inflation rate in New Zealand, as measured by the consumer price index (the inflation measure used in the agreement between Minister of Finance and Governor), has been within or close to the agreed inflation target, with inflation expectations as reflected in both surveys and bond yields remaining consistent with the inflation target — despite the impact of occasional price shocks arising from external factors such as oil prices.

What we also know is that the framework provided by the 1989 legislation has protected the Reserve Bank from any public criticism of monetary policy by government politicians — and, to the best of my knowledge, from any private criticism also.    Politicians are not likely to attack central bankers for inflation slightly exceeding the agreed target, because politicians are reluctant to suggest that monetary policy has been too loose.  Politicians are much more inclined to criticise central bankers for having monetary policy too tight — but they can hardly justify doing that if inflation outcomes are within the inflation target previously agreed with the Governor.  They can only complain about monetary policy being too tight if inflation falls below the agreed target, or seems likely to do so, and so far at least that has never happened in New Zealand.

So I am a strong advocate for the kind of relationship between government and central bank provided by the Reserve Bank of New Zealand Act 1989.

Is targeting consumer price inflation sufficient?

The second big issue facing central banks is whether keeping consumer price inflation low and stable is a sufficient task for monetary policy.  There is a fairly widespread consensus that keeping consumer price inflation under tight control is a necessary task for monetary policy, but is it sufficient?

This question has of course become a vitally important issue in the light of the Global Financial Crisis.  To what extent, in other words, was the GFC a result of excessively easy monetary policy?  In my own view, there were many factors which together caused the GFC, among which were:

-    First, the political pressure put on American banks to lend to people who were not really creditworthy.  Such political pressure has been present for at least two decades and explains at least in part the number of poor quality loans made by US banks over the decade leading up to the GFC.

-    Second, the widespread practice of securitising a large number of bank loans.  Many of these loans were of reasonable or good quality, but because many were not, securitisation effectively spread the problem loans around financial institutions all over the world.  Moreover, because securitisation effectively shifted the credit risk of a loan off the balance sheet of the originating institution, it reduced the incentive for loan originators to assess credit as carefully as might otherwise have been the case — especially where bonuses were linked to the amount of loan business written.

-    Third, the curious American tendency of making many home loans on a non-recourse basis.  This tendency inevitably reduced the incentive for borrowers to service their home loans when the value of the security fell below the outstanding loan, and indeed lead to what is colloquially referred to in the US as “jingle mail”, where people who have borrowed against the security of their home simply walk away, and mail the keys to their bank.

-    Fourth, the practice in many countries, including the US, of tightly regulating the supply of residential land, so that once demand for residential property started to go up there was little or no scope for an increase in supply to meet that demand, with the inevitable result that house prices started to rise strongly.  While house prices hardly rose at all in a large and fast-growing city like Dallas where restrictions on the supply of residential land were almost non-existent, they rose very strongly indeed in places like Los Angeles and San Francisco.  This accentuated the boom in US house prices and convinced far too many borrowers — and indeed far too many bankers — that house prices never fall.

-    Fifth, the supervisory practices of bank regulators in many parts of the world.  You may be expecting me to argue that the GFC was caused by lax banking supervision, or perhaps even the complete absence of banking supervision.  But you’d be wrong.  Contrary to public perception, most large banks were subject to intense supervision by government regulators before the GFC.  Indeed, in the largest institutions regulators were permanently camped on site.  What worries me is the effect which that intensive supervision had on the incentives for bank directors and bank managers to take responsibility for their own risk management systems.  That’s the subject of another whole speech, but suffice to say at this point that I believe one of the factors which caused the GFC was the way in which banking supervision allowed bank directors to pass the buck when it came to monitoring the risk in their own institutions.  

But what about monetary policy?  Surely that had some role to play in causing the GFC?   It does seem pretty clear that interest rates in several major economies were in some sense “too low” over much of the time leading up to the GFC, notwithstanding the low rates of consumer price inflation, leading to a huge increase in property prices and a massive increase in household sector debt.  

Ironically, China may have unwittingly played a part here.  By keeping the exchange rate between the yuan and the US dollar at a level which generated very large balance of payments surpluses, China did two things which helped keep US interest rates at very low levels.  First, the flow of inexpensive exports from China to the US helped keep consumer price inflation in the US low, thus seeming to justify the Federal Reserve Board keeping its policy interest rates low.   And secondly, by investing the proceeds of its large current account surplus in US Treasuries, China helped to keep US long-term rates low also.  This suited both countries — it enabled China to generate a high rate of growth in exports, thus helping to employ the tens of millions of people migrating from the countryside into Chinese cities; and it allowed a boom in household sector and government spending in the US.

But this brings us back to the question of whether keeping consumer price inflation under control is an adequate definition of good monetary policy.  This is obviously a debate which is still raging in the central banking community, but there is an increasingly widespread view that central banks should again pay more attention to the rate of growth of money and credit in the economy — even if slowing the growth in money and credit has the effect of pushing the rate of consumer price inflation temporarily below zero.

In the late ‘nineties, an article appeared in the annual report of the Federal Reserve Bank of Cleveland — it was widely assumed that it had been written by the Bank’s then president, Jerry Jordan.  The article purported to be about the 1920s, and noted the very low rate of CPI inflation at a time of strong productivity growth.  The article also noted that the US money and credit aggregates had grown strongly during that period, and that asset prices had exploded out of all relation to the underlying economic fundamentals — eventually ending in the social and economic pain of the Great Depression.  Most readers assumed that the article was really an oblique criticism of the way the Federal Reserve Board had implemented monetary policy during the 1990s — again, well behaved consumer price inflation, strong productivity growth, rapid growth in money and credit aggregates, and a massive increase in share prices.

One of the problems in dealing with this situation is that it’s not clear that even relatively aggressive increases in the Fed’s policy interest rate would have been sufficient to slow that growth in the money and credit aggregates.  It may well be that some additional instrument would have been needed to slow that growth — perhaps a change in the minimum capital that banks must hold against their risk-weighted assets or some other macro-prudential instrument.

So that’s a challenge which central banks are still grappling with — how do they keep consumer price inflation low and stable while simultaneously avoiding the explosive growth in money and credit which too often ends in tears?

And finally, what about “the impossible trinity”?
And finally, the third challenge for central banks, that related to what economists have called “the impossible trinity”.   As many of you will know, the impossible trinity refers to the impossibility of allowing free movements of capital across a country’s borders, while simultaneously maintaining both a fixed exchange rate and an independent monetary policy.  

A country can choose any two of those three, but not all three.  It may, for example, choose to maintain an independent monetary policy and allow free movements of capital into and out of the country.  With that choice, it abandons any ability to control its exchange rate.   That has been the choice made by many developed countries over the last two or three decades.

Alternatively, a country or territory may choose to allow free movements of capital into and out of its territory while maintaining an exchange rate fixed against another country or group of countries.  With that choice, it abandons any ability to run an independent monetary policy, and effectively adopts the monetary policy of the country or region to which its exchange rate is pegged.  That has been the choice made by Hong Kong (vis-à-vis the US dollar) since 1983 and the countries of the Eurozone (vis-à-vis the euro) since each country joined that currency bloc.  

Or a country may decide to maintain an independent monetary policy and keep its exchange rate within tight bounds.  With that choice, it must restrict movements of capital across its borders.  This, of course, is the choice which China has made.

None of the combinations is ideal.  In countries which have chosen to abandon their ability to control their exchange rate, there are frequent complaints from export industries about either the volatility or the level of the currency.  In New Zealand, exporters have been very vocal about both the level and the volatility of the New Zealand dollar.  The New Zealand dollar has moved from a level of around 80 US cents in early 2008 to 50 US cents in early 2009 and back to over 80 US cents today; and the current account balance of payments deficit, which is currently close to 5% of GDP, is expected to reach over 7% of GDP within the foreseeable future — so it’s not hard to understand why exporters are unhappy.

So in countries which have made this choice, the central bank is under constant pressure to “do something” about the exchange rate.  In New Zealand, exporters assume that currency volatility is a consequence of New Zealand’s being a very small economy — like a cork bobbing on a large ocean — and they blame “currency speculators”.  But in fact, the New Zealand dollar is not much more volatile than other floating currencies, and the kind of appreciation experienced by our currency is no different from that experienced by the currencies of much larger economies, as the graph of peak appreciations over the period 1992 to 2009 makes clear.  New Zealand’s biggest currency appreciation over that 17 year period was just on 50% on a trade-weighted basis, almost identical to the biggest trough-to-peak real appreciation of the Canadian dollar, the US dollar and the Japanese yen over that period:

Amplitude of exchange rate cycles (peak to trough, 1992-2009)

Source: Bank of England, effective exchange rates.

Those exchange rate appreciations are very uncomfortable, but the experience of Hong Kong being pegged to the US dollar, or that of some of the countries of southern Europe being part of the Eurozone, doesn’t suggest that fixing the exchange rate while abandoning an independent monetary policy is terribly comfortable either.  

Interest rates in Hong Kong have almost certainly been substantially too low for the economic conditions in that country for much of the period since 1983 with the result that inflation in Hong Kong has been higher than that in the US.  As a result, while Hong Kong’s nominal exchange rate has remained constant against the US dollar, its real (inflation-adjusted) exchange rate — which at the end of the day is what actually matters to exporters and those competing with imports — has appreciated substantially.

In the Eurozone, it is now pretty clear that interest rates were too low in countries such as Ireland and Spain in the early years of that currency union, with the result that some countries experienced a rapid growth in credit and a bubble in asset prices.  Now, when the bubble has burst wreaking terrible damage on banking systems, those countries would benefit from  a lower exchange rate, but the only way in which that can be achieved while remaining part of the currency union is by reducing internal costs through a slow and painful process of deflation.
And the third option, of maintaining an independent monetary policy and a largely fixed exchange rate by means of enforcing controls on capital movements (the policy which China has adopted), has problems too.  To begin with, the currency to which one is pegged may move in ways which are totally inappropriate for the country doing the pegging, as Argentina discovered after some years pegged to the US dollar; the policy does not provide exchange rate certainty against other currencies; it creates strong pressures to get around the controls on capital movements; and it may not avoid an appreciation of the real exchange rate if monetary policy fails to suppress a strong increase in domestic costs.

So there is no “perfect” solution to the impossible trinity.  Perhaps not surprisingly given my own experience, I lean in favour of a floating exchange rate, with an independent monetary policy aimed at delivering low and stable inflation, with no controls on capital movements.  

But I certainly acknowledge the serious disadvantages of a floating exchange rate for those producing exports or competing with imports.  To reduce the amplitude of exchange rate fluctuations, I think we need to look hard for additional policy instruments which can moderate domestic demand pressures without requiring interest rate changes, perhaps involving changes in excise tax on products with low elasticity of demand in the short-term, such as gasoline.  But that’s the subject for another lecture at another time!


1 See Inflation Targeting: Lessons from the International Experience, by Ben Bernanke et al., Princeton University Press, 1999.

2012-09-15 00:00:00.Long.
Back to Top

Built by Fullscreen Interactive | Powered by SilverStripe CMS.