Central Bank Independence

September 28th, 2017

I am very sorry that a recently arising family commitment makes it impossible for me to be with you in person.  Becoming independent for a central bank is I suppose like going off to college for a young person—a moment of validation, maturation and new adventure.  The Bank of England’s performance since becoming independent 20 years ago has in many, if not all ways, borne out the hope of those like me who strongly supported independence.  There is much to celebrate.

I however want to look forward rather than backward and suggest that the case for central bank independence today is much weaker than it was 20 years ago and to argue that institutional evolution in major countries is both likely and desirable over the next couple of decades.

The Case for Central Bank Independence

Gordon Brown’s actions to grant independence to the Bank of England were, as I read the record, based on the desire to insulate monetary policy from politics so as to solve the dynamic consistency problem.

Economic theory and the experience of the 1970s taught a clear lesson. When monetary policy was subject to political control, people anticipated loose policy and raised wages and prices.  The policy was then delivered and the result was stagflation—high inflation and high unemployment.  Bringing down inflation was then costly as policies that were tighter than expected depressed the economy.  The idea then was that credible apolitical policy would permit low inflation without recession.

These ideas were confirmed empirically by cross country comparisons.  For example, Alberto Alesina and I demonstrated in 1991 that countries with more independent central banks had lower inflation without suffering any output or employment penalty.

To be sure there were other arguments for independence as well.  An independent central bank would promote fiscal discipline by reducing the possibility of fiscal dominance and monetization of deficits.

Note that some argument of this type is necessary to justify central bank independence.  It is insufficient to argue that “experts are better than politicians.”  Democracies normally entrust matters ranging from regulating nuclear power to prosecuting criminals to granting patents to managing wars to officials who report directly or indirectly to heads of government.

How the World Has Changed

The dynamic consistency argument was compelling in the 1980s and 1990s and there is little doubt that the granting of independence to the Bank of England has enabled higher activity, with lower inflation, than would otherwise have been possible.

But the dynamic consistency argument seems much less strong today than it has in the past for a variety of reasons.

First, after a splurge of indiscipline following the breakdown of Bretton Woods, politics seem to have internalized anti-inflation norms so insulation from politics is less important.  It is noteworthy that in the US, Europe and Japan political criticism of central banks comes much more from the hawkish side than the dovish side.

Second, the problem of this era in most countries is too little inflation not too much. The United States, Europe and Japan have all fallen short of their inflation targets for close to 10 years now.   And judging by the index bond yields, inflation will average well below target for the next 10 to 20 years.

Third, much of the best contemporary thinking about the liquidity trap emphasizes the credibility issue with respect to central banks being willing to accept inflation after the economy recovers.  Structural insulation of central banks likely reduces this credibility.

Fourth, new institutional arrangements in which it is normal to pay interest on reserves reduce the concern that a non-independent central bank can be forced to enable excessive deficits through monetization.  Money in the current environment is essentially equivalent to floating rate government debt so money finance involves much smaller savings than was once the case.

In sum, central bank independence is defended as reducing inflation expectations by reducing the inflation temptation.  When inflation expectations are too low, this argument boomerangs.  Moreover, given the nature of political criticism of central banks it is not clear that more dependent central banks would in fact be less disciplined.

Costs of Central Bank Independence

If the benefits of central bank independence have diminished, I suspect the costs have increased because of the greater importance of Treasury – Central Bank cooperation and the wider remit of central banks.

Coordination between independent entities is more difficult than between entities that are politically connected.  The range and significance of the issues where Treasury – Central Bank coordination is important has surely increased over the last two decades.  These include:

–Debt management policy where QE operates by shortening the maturity structure of the debt that a country’s creditors have to hold.  The impact of QE is essentially the equivalent of changes in the maturity structure of a country’s debt accomplished through altered patterns of issuance or buybacks.  It makes very little sense for debt maturity policies for   one country to be set separately in two places.

–Fiscal monetary cooperation is a much more significant issue when an economy is in the liquidity trap, near the liquidity trap, or facing the possibility of getting into the liquidity trap at some point in the future.  When monetary policy cannot sterilize impacts of fiscal policy on demand, there surely should be coordination of fiscal and monetary policy.

–Exchange rate policy is another area where given the number of industrial and international interests involved it is hard to imagine complete delegation to an apolitical institution.  And yet exchange rate policy is completely intertwined with monetary policy and prospective monetary policy.  Again Treasury – Central Bank cooperation is essential.

–Crisis prevention and response is now seen after the events of 2008 as more salient than it was when the central bank independence wave swept across the world.  This is inherently political as it involves, for example, deciding who will and who will not get access to government funds and on what terms.  Treasury – Central Bank coordination is again crucial.

Each of these issues can be managed by dialogue between independent central banks and Treasuries.  But this comes with costs in terms of speed of decision, clarity of communication, lack of accountability and democratic legitimacy, or distorted policy choices.

For example, it is likely that American taxpayers paid dealers billions of dollars in the QE period because the Treasury was lengthening debt maturities, while the Fed was, in effect, issuing short and buying long.  Or to take another example, exchange rate policies that do not work through changed monetary policies are not likely to have lasting effects.

Where do we go from here?

It is clear in a way that was not the case prior to the financial crisis that central bank policy is about much more than turning an interest rate dial.  Much of what is new, is by its nature, political and in need of integration with other policies.

I suspect that over time we will see some “drawing closer together” of Treasuries and central banks and perhaps more devices like inflation targets where politically accountable officials establish frameworks in which central banks can act.

And central bankers will find themselves in closer consultation with governments and engaged in more joint endeavors.

The end of debate over central bank mandates, like the debate over the end of history, is destined to go on for a long time.

Remarks by Lawrence H. Summers (as prepared to be read)

Bank of England: ‘Independence – 20 years on’ Conference 

September 28, 2017