Advanced economies are now emerging from the aftermath of the great financial crisis. Despite facing the largest shock in the postwar period, central banks have succeeded in maintaining price stability and fulfilling their mandates.
A key factor behind this has been central bank independence. But in spite of its benefits, its desirability and relevance are nowadays being increasingly challenged.
The evolution of central bank independence
The rationale for delegating powers to independent central banks evolved mainly from the experience of the 1970s. At that time, policymakers believed that there was a stable trade-off between unemployment and inflation, whereby monetary policy could achieve permanently higher employment at the cost of slightly higher inflation.
The trade-off rested on the notion that an increase in the money supply could boost aggregate demand and stimulate employment, while keeping inflation expectations well anchored.
This policy was revealed to be time-inconsistent. Over time, commitments to control inflation at a later date lost credibility. The public came to anticipate the behaviour of monetary policy and quickly embedded expectations of higher inflation in their price setting and wage bargaining. The trade-off between unemployment and inflation disappeared.
In the United States, for example, year-ahead inflation expectations rose steadily from around 4% in 1970 to 12% in 1979. Inflation stood between 3% and 12% for the whole decade. But there was no fall in unemployment, which varied between 4% and 9%. In fact, inflation and unemployment had a tendency to rise in tandem.
Therefore, to overcome time-inconsistency, central banks had to regain credibility as institutions capable of delivering price stability. They thus committed to a non-inflationary path while resisting any political pressure to pursue objectives other than price stability.
As Alesina and Summers showed, central banks that acted in this way – such as the Bundesbank – had managed to achieve lower inflation without suffering output or employment penalties.
Annual inflation in Germany from 1970 to 1990 averaged just 3.8%, whereas in France it averaged 7.9% and in Italy 11.8%. At the same time, unemployment in Germany over the same period averaged just 3.6%, whereas in France it was 5.6% and in Italy 6.4%.
Societies gradually converged on a framework whereby monetary policy was removed from the pressures of short-term political agendas and delegated to independent central banks. A clear social consensus had emerged that price stability was a common good under all circumstances, but its achievement depended on the right institutional framework.
This rested on three elements: central banks needed a clear mandate to achieve price stability; they needed independence over the instruments they could use to achieve their mandate; and monetary policy had to be embedded in a strong accountability framework requiring the central bank to explain how its decisions contributed to its goal.
This framework offered the public a guarantee that independent central banks would not exercise power arbitrarily. Their discretion was limited to how they formulated monetary policy. They had no discretion over whether to pursue their goals, and what goal they had to achieve. It is on this basis that the mandate of the European Central Bank was established.
For much of the 1990s and early 2000s, the monetary policy framework went unquestioned. There was a broad consensus that granting independence to central banks had been successful in bringing inflation under control. Cross-country studies found a clear negative relationship between independence and inflation in OECD economies.
Moreover, central bank independence was seen as a key factor in the lower volatility of output and inflation observed over this period – the phenomenon identified as the “great moderation”. The variability of real output growth was found to have declined by half since the mid-1980s, while the variability of inflation had declined by about two thirds.
Importantly, central banks had been able to maintain low and stable inflation using a single instrument – policy interest rates – which caused few concerns. Interest rate policy was perceived to be a normal tool of monetary policy and to create few distributional effects.
Then the great financial crisis erupted, which led to a steep drop in both output and inflation across advanced economies. This presented central banks with two new challenges.
The first challenge was to their objectives. For the first time in living memory, there was a threat of a deflationary spiral taking hold. This would have led not only to central banks failing their mandates, but to the whole economy sinking into a prolonged depression – just as had happened in the 1930s.
The second challenge was to their instruments. Faced with these deflationary threats, central banks responded decisively with conventional policy, but they eventually ran into the lower bound on interest rates. This required them to adopt new, unconventional measures, such as large-scale asset purchases, to stabilise inflation.
The new environment of low inflation and unconventional policies has led to questions about the key tenets of the pre-crisis monetary policy consensus. Some observers now query whether the circumstances that justified central bank independence still exist – and whether the grounds for delegating monetary policy to independent authorities remain valid.
Time-inconsistency under low inflation
The first claim is that, in a lower-inflation environment, the time-inconsistency problem that emerged in the 1970s has become less relevant. As such, independent central banks are no longer a necessity to ensure credibility and keep inflation expectations anchored.
Such reasoning rests on two assumptions: first, that inflationary pressures have, by and large, disappeared in advanced economies; and second, that the social consensus behind price stability is now so well established that elected authorities would not compromise it in the pursuit of other objectives. Both of these assumptions can be questioned.
First, the absence of inflationary pressures in recent years is well-explained by two factors, and neither would justify a change to the monetary policy framework.
One is the stability of the monetary regime, which has meant that economic shocks no longer have lasting effects on inflation, either during slumps or during booms. This is a direct result of the success of central banks over the past twenty years in anchoring expectations and reducing inflation volatility. So, the response to a stable inflation environment should not be to dismantle the framework that has delivered it.
The other factor is the depth of the crisis that struck advanced economies, which created unusually high unemployment and made inflation less responsive as the labour market recovered. But here, too, little would be gained by making central banks less independent. Long-term institutional arrangements which are based on a strong social consensus should not be changed because of cyclically-driven developments in inflation.
But are there ongoing structural changes in the economy, as some argue, that will permanently contain inflationary pressures in the future?
The notion that inflation is a non-monetary phenomenon is not new. In 1971, Federal Reserve policymakers argued that inflation was “a structural problem not amenable to macroeconomic measures”, and similar statements can be found in the records of other central banks throughout the 1970s. The Fed’s view only changed when Paul Volcker became Chairman in 1979 and affirmed the responsibility of monetary policy.
Today we are witnessing major demographic and technological changes, but these do not necessarily imply that structural factors are the main drivers of inflation.
Indeed, the impact of demographic change on inflation is unclear. It might put downward pressure on prices if aggregate demand falls more than aggregate supply. But it might equally create upward price pressures. According to the life-cycle hypothesis, as society ages the elderly will eventually spend their savings and consume more.
The impact of technological change, such as the spread of e-commerce, is also unclear. E-commerce could in principle erode the monopolistic power of suppliers, which would reduce mark-ups and flatten the Phillips curve. In other words, lower levels of unemployment would not necessarily lead to faster wage increases. But it might also result in suppliers changing prices more frequently, which would result in a steeper Phillips curve.
In any event, the effects of structural changes cannot create steady-state disinflation. They can only lead to a potentially long transition to a new steady state. Whether inflation is higher or lower over time will depend on the reaction of monetary policy along the way. And what recent experience suggests is that independent central banks remain the best institutions to anchor inflation expectations through such transitions.
This brings me to the second assumption about the incentives of elected authorities.
The belief that they will no longer compromise price stability in the pursuit of other short-term objectives is not borne out by what we have seen in recent years. In fact, it has become evident that time-inconsistency is a recurring risk that arises in both high- and low-inflation environments.
When inflation is rising, short-term political considerations still create a certain set of incentives to pressure central banks into prioritising economic growth and avoiding monetary tightening. And when inflation is falling, we have seen that there are incentives to prioritise considerations of moral hazard and financial sector concerns and to oppose monetary easing.
Such incentives were visible in the euro area during the crisis. Despite an observable slide in inflation, there were those who opposed responding with unconventional measures, because they feared the possible effects on governments’ willingness to enact reforms, or on banks’ incentives to clean up their balance sheets. These risks were given priority over the certainty that price stability would be sacrificed without action.
And today in other jurisdictions, we see concerns being publicly expressed about whether the central bank should pursue a normalisation path in the face of rising inflation.
These deeds have so far had limited consequences only because central banks are independent and credible. But if central banks were less independent, and the public perceived that monetary policy could be pressured in either direction, it would eventually de-anchor inflation expectations and jeopardise price stability – just as in the 1970s.
The relationship between monetary and fiscal policy
The crisis has also prompted questions about the relationship between monetary and fiscal policies. First, it is claimed that, once central banks use unconventional measures that entail buying large volumes of government bonds, they cross the boundary into fiscal policy. In doing so, they exceed their mandates and the scope of their independence.
Such reasoning ignores key aspects of our institutional framework.
The specific design of central banks’ mandates – which gives them independence over their tools, but not their goals – was put in place because no one could foresee all the challenges that monetary policy might face in the future. Allowing central banks leeway over their instruments was the best way to ensure that, come what may, the social consensus in favour of price stability could be maintained.
The actions of central banks during the crisis not only followed this principle to the letter – they demonstrated precisely why it was necessary in the first place.
With conventional policy no longer sufficient to secure price stability, central banks would have failed their mandates without instrument independence. But by expanding their set of tools, they extended their policy space so that they could deliver their objective.
Though the measures we used were unprecedented, the shift was only in form, not in function. Their function was exactly the same as that of conventional policy: to increase money and credit aggregates, lift demand and stabilise medium-term inflation around our objective.
Indeed, Milton Friedman argued that asset purchases are not only an appropriate tool of monetary policy, they are ‒ under certain circumstances ‒ essential to boost money supply and maintain price stability. He therefore supported their use during deflationary episodes such as the Great Depression or Japan in the 1990s.
The fact that deploying asset purchases was fully in line with our mandate was confirmed by the European Court of Justice. In its judgement on our Outright Monetary Transactions programme, it ruled that purchases of government bonds are legal under the ECB’s statute and a legitimate tool of monetary policy.
Another challenge to central bank independence is related to the effectiveness of monetary policy at the lower bound on interest rates.
When interest rates fall towards zero, it is argued, monetary policy becomes ineffective, since it enters a “liquidity trap” where it can no longer stimulate demand. In these conditions, fiscal policy is needed to stabilise inflation.
For as long as interest rates remain constrained by the lower bound, the argument goes, monetary policy should therefore support fiscal policy to achieve price stability. But central bank independence, it is said, impedes such coordination.
To begin with, the notion that monetary policy is ineffective when the lower bound is reached has been disproved by the experience of the crisis. There is now a wealth of evidence showing that the unconventional policies adopted by central banks have eased financing conditions, boosted output and stabilised inflation.
In the ECB’s case, there was a particular question about whether our asset purchases would be effective in a bank-based economy. But we estimate that the growth rate of bank lending to euro area firms would be more than a third lower today without our package of credit-easing measures. The rates that banks charge for firms to borrow money would have been around 50% higher.
This is not to say that monetary policy could not benefit from greater alignment with fiscal policy when faced with a deep slump. In the euro area, it was clear that the lack of a supportive aggregate fiscal stance was a headwind for monetary policy during the crisis. But for such policy alignment to be successful, it has to be on the basis of every authority fulfilling its own mandate in full independence.
The existence of multiple actors in setting macroeconomic policy requires clear mandates to avoid what Milton Friedman described as the “dispersal of responsibility, which promotes shirking of responsibility in times of uncertainty and difficulty”.
Moreover, if central banks were to enter into a form of coordination with fiscal authorities that reduced their independence, it would ultimately be self-defeating.
While the mandate of the ECB is price stability, fiscal authorities have multiple mandates. So, if the central bank were to submit to political control, coordination with fiscal authorities would be unlikely to be limited to the lower bound.
Fiscal authorities would have an incentive to use monetary policy to achieve other objectives. And this would end up with monetary policy becoming fiscally dominated, which history shows is inconsistent with price stability in the long run.
Distributional effects of unconventional measures
The third claim against central bank independence relates to the distributional effects of unconventional policies. In particular, it is argued that the distributional effects are much larger than for conventional policies. But the delegation of monetary policy to an independent body rests on the premise that such effects are not first order.
Unconventional policies are perceived to have larger distributional effects because they involve active intervention in financial markets. And conventional policies are perceived to have lower distributional effects because they entail passively providing liquidity to banks.
But this is an incorrect perception, for two reasons. First, conventional policies also produce differing effects on creditors and debtors. Second, empirical evidence suggests the effects on distribution of both conventional and unconventional policies are in any case limited.
In the long run, monetary policy is not an important factor for the distribution of resources within societies. Distribution is affected more by low-frequency developments such as structural and institutional factors, or fiscal policies such as tax regimes.
In the short run, monetary policy can influence distribution via two channels: a direct effect related to asset prices and financial income, and an indirect effect related to the macroeconomic impact of expansionary policy on jobs and wages. But the net effect on distribution appears to be balanced, including with unconventional measures.
Across households, research finds that our unconventional measures have barely affected wealth inequality, because they have had a positive impact on housing wealth. This is fairly evenly distributed across households and dominates wealth held in stocks and bonds.
Furthermore, while some savers have seen their financial income reduced by lower interest rates, asset purchases have at the same time triggered both a sizeable reduction in unemployment and wage increases for the employed. In fact, households with the lowest incomes and the fewest liquid assets have benefited in particular from monetary easing.
Estimates suggest that asset purchases by the central bank leading to a 30 basis point decline in the term spread, reduce the unemployment rate among households in the bottom income quintile by 2 percentage points, and contribute to a modest drop in the Gini coefficient on gross income.
So, the positive effect on employment – which has risen by 9 million in the euro area since the trough of the crisis – counterbalances any potential negative distributional effects of monetary policy. And the same appears to be true in other advanced economies.
The distributional effects of monetary policy across member states are similarly balanced.
If we look at how the net interest income of different countries has been affected by falling interest rates since 2008, we do not see a picture of “creditor” countries losing out and “debtor” countries gaining. In fact, despite having a large net international investment position, Germany has gained slightly from low rates – meaning it has received more in interest earnings than it has disbursed in interest payments.
At the same time, all member states have benefited from the economic recovery and the stable macroeconomic environment, which were enabled by our expansionary policies. Since the launch of our measures, the dispersion of growth and unemployment rates have fallen to their lowest level since the start of Monetary Union.
So, there is little to suggest that the use of unconventional measures during the crisis has created first-order distributional effects – and certainly not to an extent that would undermine the rationale for central bank independence.
And perhaps more importantly, the distributional effects of not acting to defend our mandate would have been severe. This would have resulted in a fundamental shift in growth and inflation expectations, triggering a deflationary spiral, with significant implications in particular for the poorer and younger members of society.
The continuing value of central bank independence
These various arguments against central bank independence also miss a more fundamental point, which is the value of an independent central bank that can act decisively without political pressure – and especially in the euro area.
Indeed, we saw clearly during the euro area crisis that coordinated policy responses among governments were difficult to achieve, tended to arise only under severe market pressure, and then often turned out to be insufficient, requiring further responses later.
In this context, the governance structure of the Eurosystem – where each member state is represented by the governor of its central bank in their own personal capacity, who is not bound in their decision-making by their nationality – was essential, since it facilitated an effective systemic response.
A central bank that was both independent and built to serve the whole of the euro area, and not individual member states, was able to create the required policy space in an extremely difficult context.
This brings me to my conclusion.
Central banks are powerful, independent and unelected. This combination can only be squared if they have a clearly defined mandate for which they are held accountable by the public.
It is legislators that define the mandate that establishes the goals of monetary policymaking. And it is legislators that are responsible for holding central banks accountable for the effectiveness of their monetary policy. But for effectiveness, credibility is essential, and therefore legislators must want a credible central bank.
Credibility hinges on independence: the central bank should not be subject to fiscal or political dominance and should be free to choose the instruments that are most appropriate to deliver its mandate. Legislators should therefore protect the independence of central banks, as it is essential for fulfilling the mandate that they themselves have defined.
The distributional consequences of monetary policy, if there are any, and other risks such as moral hazard and financial stability, must be addressed by other tools. These include fiscal policies, macroprudential policies and banking supervision – policies that are more targeted and therefore inherently more appropriate for close scrutiny by legislators.
Faced with future crises, central banks will adhere to their mandates and use their independence to fulfil them. In our case, the ruling of the European Court of Justice has shown that we can use all the tools within our mandate to tackle future challenges to price stability. We cannot foresee now what those challenges might be, but if and when they arise, they may require us to adjust our policy space once more to meet our mandate.