[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Time inconsistency and inflation

In May 1997, the chancellor of the exchequer decided that he would no longer meet the governor of the Bank of England to set the bank base rate. He then set an inflation target of 2.5% per annum to be met by a Monetary Policy Committee. The UK gilts market moved to indicate lower inflation expectations. Why was the Bank of England granted such independence? What does this independence mean for future inflation in the UK?

Economic theory of inflation
Inflation has many impacts:
– Inflation erodes the standard of living when price indexing of income is lagged.
– Producers have the costs of reprinting prices.
– Unexpected inflation redistributes resources from creditors to debtors.
– In times of surprise inflation, people may spend resources in gathering information on changing prices.
Inflation is stable (or in equilibrium) when the growth in the money supply is equal to inflation and inflation expectations. For example, if inflation expectations rise, people may bargain for inflationary salary growth.
Economic theory says that in the long run, unemployment is independent of inflation. This relationship is known as the Long Run Phillips Curve (LRPC). Long run unemployment is also known as the natural rate of unemployment and it is determined by demand and supply in the labour market. Those who are unemployed at market wages contribute to the natural rate of unemployment.
The Short Run Phillips Curve (SRPC) depicts a negative relationship between unemployment and inflation. The relationship can be negative because of ’sticky prices’. Prices can be sticky if producers have periodic price lists (menu costs) or if nominal wages do not move downwards (sticky wages). The position of the SRPC is determined by inflation expectations and intersection with the LRPC gives the long run equilibrium. Figure 1 shows that higher inflation expectations lead to higher inflation (in the long run).

Time inconsistency
If low inflation is possible through means of a suitable monetary policy and the driving down of inflation expectations, why have we had episodes of high inflation? Perhaps politicians have not wanted low inflation. The public has then had higher inflation expectations. It may be that politicians have preferred to use the levers of economic policy rather than to announce elections at ‘good times’.
A second story uses economic incentives. In basic consumer theory, utility curves are convex for two goods. Optimal consumption is where the budget constraint is tangential to the utility curve.
Now consider two bads, unemployment and inflation. The loss functions for these curves become inverted. Lower losses are because of lower inflation and unemployment. Each curve, L1 and L2 in figure 2 across, give points of equal loss, and loss curve L2 indicates lower losses than loss curve L1.
We know that in the long run the best the government can achieve is unemployment at the natural rate, U*. For example, equilibrium at point A is superior to the equilibrium at point B because the economy has the same rate of unemployment but a lower rate of inflation.
Suppose that a government announces a low inflation policy and inflation expectations are low. The economy could be at a point like A in figure 2. However, the government has an incentive to lower unemployment (temporarily) and raise inflation. If the economy has sticky prices, a government can temporarily cheat and lower losses by taking the economy to a point like C in figure 2. In the longer run this will lead to an upward revision of inflation expectations. The economy may finally arrive at a point like B. In the long run this is inferior because inflation is higher than it need be.

Policy implications
In figure 2, how can the economy remain at a point like A, which is a superior equilibrium to the equilibrium at point B?
The government could hope to acquire a reputation. In the context of the theory, the government announces low inflation and sticks to it. In practice this is easier said than done. Politicians might not care about their reputations, especially when their time horizons are short. Even the anti-inflation Thatcher government does not provide a good historical example.
The other method is to delegate monetary policy to an independent central bank. There are two forces which help establish anti-inflation credibility.
– The independent central bank does not have the same electoral problems as the government.
– The central bank may put a higher weight on inflation in the loss function than the government.

Central bank independence
What does central bank independence involve? An evaluation would look at the following factors (see Cukierman et al, 1992).
– Appointments at the bank The term of office, who appoints the decision makers and the conditions of dismissal.
– Policy formulation Who formulates policy? For example, is it the bank, is it the government or is it done by joint decisions by the bank and the government.
– Economic objectives Is it exclusively to have price stability, or some other short-term conflicting role such as achieving full employment?
– The conduct of monetary policy The bank’s terms and conditions of lending to the government.
An empirical evaluation involves some subjectivity as various factors have to be weighed up. Moreover, legal independence and actual independence can be different. This is important since the spirit and practice of legislation can differ. Nevertheless, for developed economies, Cukierman et al find an inverse relationship between bank independence and inflation. Alesina and Summers (1993) add that bank independence is not correlated with economic growth (as expected). Hence some economists feel that bank independence comes very close to a ‘free lunch’.

Future inflation
The theory predicts that an independent central bank would have little incentive to cheat, and would deliver lower inflation. The theory also predicts a structural break in the inflation series. To the statistician this is called non-stationarity and many economic time-series behave very differently over time precisely because of structural change in the economy.
If a central bank is given an inflation target and significant political independence we could legitimately predict future inflation to be at the level of the inflation target. We would also be sceptical of results from an econometric model. Relationships in the economy are likely to change upon a shift in economic policy. However, relationships in a statistical model are unlikely to change, given strong dependence on historical data. In economics this is called the ‘Lucas critique’ (after the economist Robert Lucas). The Lucas critique says that parameters in statistical models are not invariant to policy change.
There is no simple link between democracy and good economic policy. Politicians should tie their hands on monetary policy. When there is a significant change in economic policy, the past is a weak guide to the future.

Further references
Alesina, Alberto and Summers, Lawrence (1993), ‘Central Bank Independence and Macroeconomic Performance: Some comparative evidence’, Journal of Money Credit and Banking, volume 25(2), pp151162.
Blanchard, Olivier Jean, and Stanley Fischer (1989), ‘The theory of central bank independence’, Lectures on Macroeconomics. MIT Press, Cambridge. 590pp,
ISBN 0262022834.
Cukierman, Alex, Webb, Steven B and Neyapati, Billin (1992), ‘Measuring the Independence of Central Banks and its Effects on Policy Outcomes’, The World Bank Economic Review, volume 6.3, pp353398.