Monetary policy: theory and practice
by Peter Dawson
Monetary policy has been at the forefront of government thinking about the workings of the economy for the last
30Monetary policy has been at the forefront of government thinking about 7.he workings of the economy for the last
30years. Together with fiscal policy it is one of the main methods governments employ in the pursuit of their economic
objectives of high economic growth, low unemployment and low w-d stable inflation. Traditionally monetary policy
has been conducted by central banks on behalf of governments. This means that although the central bank implements
monetary policy~ it is the government which makes the final decision about the timing and the magnitude of the
change. Recently governments in a number of countries have granted varying degrees of independence to central banks.
In the UK, for example, the Bank of England (BoE) was given 'operational' independence in 1997 granting it a degree
of discretionary power in the setting of interest rates and other monetary variables.
The importance of monetary policy can be found in the increased media interest in monetary policy matters. Barely a
day goes by without some mention of monetary policy Newspapers are filled with speculation about the likely moves
monetary authorities will take in order to stabilise the economy Remarkably there is now broad agreement amongst
economists that monetary policy is the only policy tool capable of reducing business-cycle fluctuations. However, this
does not mean that monetary policy is no longer considered to be controversial. There is still disagreement amongst
economists and central bankers over how it should be implemented and who should control it.
This chapter addresses three fundamental questions. First, what is the role 3Ild purpose of monetary policy; what is it
used for and how does it affect the workings of the economy? Next, how is monetary policy implemented? For
example, what is the procedure by which monetary policy is conducted? Third, and finally should governments or
central banks have ultimate responsibility for monetary policy?
The objectives of monetary policy
To understand the role and purpose of monetary policy requires an understanding of what monetary policy actually
means. The standard textbook definition describes monetary policy as 'actions taken by monetary authorities to affect
monetary and other financial conditions in pursuit of the broader objectives of sustainable growth of real output, high
employment and price stability' (Lindsey and Wallich, 1989).
The monetary authorities in the UK are the Bank of England (BoE) and the Treasury There is little doubt that the
monetary authorities would like sustained economic growth because it increases the economy's output of goods and
services over time and increases the standard of living of its citizens. However, growth is unlikely to be sustained if it is
not stable. The reason for this should be clear: consumers and firms are unlikely to make long-term investment plans
because they will be unwilling to undertake the risks associated with an unstable economy.
Economic growth is closely related to employment. If there is high unemployment then there will be unused resources
(workers, machinery and factories) in the economy resulting in a loss of output. Needless to say firms will not
undertake any capital investment programmes when there is spare capacity Nevertheless, this is not to say that the goal
should be a zero rate of unemployment. Employment should be at a level whereby the amount of labour demanded is
equal to the amount of labour supplied (the full employment or natural rate of unemployment level).
The monetary authorities are also interested in maintaining price stability Price stability in a market economy is
desirable because it conveys information about resource allocation. High inflation in particular makes it especially hard
to plan for the future because high inflation is synonymous with uncertain inflation, and evidence suggests that this has
a detrimental effect on economic growth (see for example, Barro, 1995). As a result it is widely accepted that low and
stable inflation is the key objective of monetary policy. In addition, the belief that it was possible to raise output and
employment by accepting a higher level of inflation as depicted in the celebrated Phillips curve no longer exists (in the
How do the monetary authorities go about achieving price stability? This would be relatively straightforward if they
could control inflation directly Unfortunately they cannot. What they can do is try and manipulate the financial system
by seeking to control one or a number of monetary variables which they believe to have an important influence on price
stability These variables are called the intermediate targets of monetary policy and include such things as the money
supply and the exchange rate. Unfortunately the intermediate targets cannot be controlled directly Consequently, the
monetary authorities ' use operating targets - such as short-term interest rates and the amount of reserves held by
The debate over whether the authorities should pursue price stability or inflation stability has attracted
increasing attention in recent years (see for example the references cited in Cecchetti, 2000). Price stability
implies that the authorities have to address previous failures to meet the target whereas with an inflation rate
target it does not. The implication of this is that a price stability target will create more certainty in the long-run
- because the authorities are obliged to return to the target path for prices - but more variability in the short-run:
if the inflation rate this year is 4% but the target rate is 2.5% then the goal this year will be less than 2.5%.
banks - which are directly controlled by the monetary authorities and are closely related to the intermediate targets.
Finally in order to achieve their operating targets the monetary authorities use the various monetary policy instruments
at their disposal. The most common instruments are often market operations, discount policy. and reserve requirements.
The linkages between the goals, targets and instruments are represented schematically in Figure 1. What matters in
terms of monetary policy is the reliability of these linkages. This diagram and the elements within it form the basis of
the discussions that follow in this chapter.
Monetary policy: process
The monetary transmission mechanism
The formal analysis of how monetary policy influences spending and prices is called the monetary transmission
mechanism. As noted in the previous section, the monetary authorities use policy instruments, which are directly under
their control, to change short-term interest rates and the money supply to. ultimately enable them to achieve their goal
of price stability A more thorough discussion of the instruments and targets at the authorities' disposal will occur later.
The objective of this section is to analyse how changes in the financial system translate into changes to the real
economy or, to put it another way how changes in interest rates and prices of financial assets influence investment and
Money supply and demand
The purpose of using an instrument is that it will create a shock or disturbance in the financial system, in other words it
disturbs equilibrium, by changing either a price variable or a quantity variable. The most common price variable is the
rate of interest (cost of borrowing) and the most common quantity variable :s the quantity of money (supply of money).
These components are modelled within the money market.
In constructing the money market it is common practice in most textbooks to assume the money supply schedule is
directly under the control of the authorities. Diagrammatically this enables the supply curve to be presented as a
vertical line (i.e. perfectly inelastic). In truth the money supply is never fully under the control of the authorities. The
reason for this is it is partly determined by the behaviour of banks and their customers. However, the assumption of a
completely con1rolled money supply schedule does not lose the generality of the analysis presented here.
The demand for money reflects the three different motives for holding money as stipulated by John Maynard Keynes in
his celebrated 1936 book, The General Theory of Employment, Interest and Money. These motives are transaction.
precautionary and speculative. In Keynes' terminology; the demand for money is known as liquidity preference theory
because when an individual is holding money they are expressing a preference for the liquidity that holding money
brings with it.
Transaction demand arises from the function of money as a medium of exchange in that it enables individuals to
purchase every day goods and services. such as food for lunch and petrol for the car. Keynes emphasised that the
transaction motive for holding money increases as income increases.
Because we live in an uncertain world, individuals are likely to hold money as a precaution against unexpected
occurrences. For example, you may hold money as a precaution against an unexpected car repair bill or if a previously
expensive item of clothing that you wanted suddenly becomes massively reduced in a sale. The precautionary
demand, like the transaction demand. increases as income increases.
Keynes' third and crucial motive was based on the fact that as well as a medium of exchange, money is also demanded
as an asset - in other words as a store of wealth. Keynes called this motive the speculative demand. Keynes argued
that the speculative demand too was positively related to income but in addition was negatively related to interest
rates2. He explained this by assuming there were only two types of asset - money (the most liquid of assets), and bonds
(to represent all other financial assets). Individuals would then determine how much money they would hold on the
basis of their expectations of future movements in interest rates. If, for example, individuals expect interest rates to fall
then given that the interest rate payable on bonds issued in the past are fixed, the expected fall in interest rates will
increase the attractiveness of bonds, more people will wish to buy them and their price rises. In addition, those that
already hold bonds will make a capital gain when interest rates fall. If interest rates are above the perceived normal
level, individuals will expect a fall in the near future and so they will buy bonds and hold less money.
Adding these three motives together gives us a downward-sloping money demand schedule (Md0) and if we add this
demand schedule to the perfectly inelastic supply schedule (Ms0) we obtain an equilibrium price of i0 and an
equilibrium quantity of q0 (Figure 2).
If we now consider an expansionary or contractionary shock then the interest rate will change. For example consider an
expansionary shock. The monetary authorities can achieve such an expansion by increasing the money supply; say to
Ms1. This would lead to an excess supply of money, which would require a fall in interest rates from i 0 to i1 in order to
restore equilibrium. This mechanism can also work in the opposite direction. In this case, instead of a change in the
money supply to control interest rates we have a change in the interest rate to control the money supply.
2. Baumol (1952) and later Tobin (1956) argue that the transaction motive and the precautionary motive for
holding money are also negatively related to interest rates. The reasoning behind this was that although by
holding money individuals could avoid the problems of transaction costs, there is an opportunity cost in the
form of interest that could be earned on other assets. As the rate of interest, and therefore the opportunity
cost of holding money increases, individuals will reduce the amount of money they hold for transaction and
Monetary policy and the economy
A critical question for the monetary authorities is how changes in money supply and interest rates affect the real
economy? Only by understanding the mechanisms through which monetary policy affects the economy will the
authorities be successful in the conduct of monetary policy
According to Keynes this fall in interest rates has the effect of increasing business investment through the lower cost of
borrowing. Spending on consumer durables (houses, cars and white goods) also increases because saving is now less
attractive because the opportunity cost of money is lower. In addition the fall in interest rates results in a rise in value of
financial assets (bonds) adding a further stimulus to consumption via its effect on wealth and to investment by lowering
the cost of capital.
The monetarist story is somewhat different. Monetarists argue that tile excess supply of money induced by the
monetary authorities or brought about via a reduction in interest rates means that individuals find they have more
money than they need for normal spending. These consumers may decide to spend some of this money on a much
wider range of financial and real assets than Keynes had originally postulated. In terms of financial assets this would,
as before, include bonds but also shares and other financial assets such as derivatives. The increased demand for these
assets will increase their price, so that those that already own such assets will find that the value of their financial
wealth increases. And if the assumption that individuals wish to spread the amount they spend over time - life-cycle
spending hypothesis - holds they may well increase current consumption. Investment is also boosted, as before, via the
reduction in the cost of capital. Although both approaches come to similar conclusions they differ in their nomination
of the key variable. In the standard Keynesian approach interest rates are key whereas in the Monetarist approach it is
the money supply
As well as the cost of borrowing, wealth and the effect on asset prices, and because the UK conducts a large amount of
trade with other countries, consideration must be given to how monetary policy affects the exchange rate. Looking at
the impact on the exchange rate is important because this will affect the amount of overseas expenditure and investment
relative to domestic expenditure and investment undertaken. Maintaining our discussions in terms of an expansionary
monetary policy consider the following. As before we have a reduction in interest rates. This time consumers and firms
in the UK and overseas will sell UK assets and purchase overseas assets because the relative interest rates on overseas
assets are higher than those currently available on domestic assets (we are, of course, assuming that the reduction in
interest rates in the UK is not matched by a reduction in interest rates overseas). Consequently in order to purchase
these overseas assets consumers and firms will sell pounds in order to buy foreign exchange in order to purchase
overseas assets. This leads to a fall in the value (depreciation) of the pound, which makes UK goods more attractive to
overseas buyers, and therefore increases export sales. However such a mechanism only works when the exchange rate
is flexible (allowed to change). If the exchange rate is fixed, such as when the UK was a member of the European
Exchange Rate Mechanism (ERM), there will still be a net outflow of funds but the exchange will be unchanged.
It cannot be over-emphasised that the mechanisms described above will not affect the real economy instantaneously:
Indeed there are significant time-lags involved. Both Monetarists and Keynesians agree that monetary policy can have
an effect on output in the short-run, but the only lasting effect is on prices. A standard rule of thumb has developed in
the literature which states that policy tends to have its full effect on output within about one year, and on inflation in
around two years, see, for example, Lane and Van Den Heuvel (1998). Some of the problems generated in the past
arose because the authorities incorrectly believed that monetary policy had long-term effects on output.
In recent years these so-called traditional monetary transmission mechanisms have come under attack from two groups.
The first group contend that monetary policy is ineffective even in the short-run because individuals have rational
expectations. Proponents of this view - the New Classical School- developed by Robert Lucas and Thomas Sargent
among others, argue that because individuals use all information available to them they will quickly learn how
monetary policy is being co-ordinated by the monetary authorities. Consequently they will learn to anticipate what the
authorities are likely to do next, thereby neutralising the (short- run) real effects of the policy: As will be demonstrated
later, this has important implications both in terms of how the authorities undertake monetary policy and of who should
be in charge of it.
The second group contend that there are other important channels that traditional theories overlook. These channels
specifically relate to the important role played by banks within the financial system; see Bernanke and Gertler (1995)
and Dimsdale (2001) for lucid expositions of how these channels work. A particularly important example is in the way
firms are financed. For example small and medium sized firms tend to rely exclusively on bank lending whereas larger
firn1s are able to utilise the stock market as an alternative source of finance. If, for instance, following an expansionary
monetary policy banks are unwilling to lend money then there will be less investment opportunities for small and
medium sized firms. This will, in turn, dampen the resulting increase in output, thereby reducing the effectiveness of
the policy: In many ways, these channels act as complements rather than substitutes for the traditional channels and can
be thought of as being part of a broader class of channels which relate to the functioning of the financial system.
Implementation of monetary policy: choosing an intermediate target
As outlined in Figure 1, in order to achieve a particular objective (goal) the monetary authorities often seek to control
an intermediate target variable. An intermediate target variable is a variable which lies between the policy instrument
and the ultimate goal. It is a variable which the monetary authorities believe will directly help them achieve such a
goal. To understand how this could happen consider the following analogy. Imagine you are a cricketer who has just
fielded the ball near the boundary. You have a good throwing arm so will have no problem reaching the stumps (goal).
However, the stumps are obscured by a hill. Without any additional information you would have to guess where the
stumps were before throwing. Suppose you were given some information that would improve your chances of hitting
the target. For example, imagine that a ring could be positioned on top of the hill and throwing the ball through this ring
would give you a better chance of hitting the stumps. In effect this ring is the intermediate target because it gives
immediate feedback - if the ball goes through the ring you know you have a better chance of hitting the target3.
Hubbard (2000) uses a similar analogy.
As demonstrated in the previous section, when monetary policy is undertaken there is a wide variety of variables
affected (interest rates. money supply exchange rate). Each one is a possible candidate for the intermediate target. The
question is which one (or which ones) should the authorities choose? In general there are three aspects to consider
when choosing an intermediate target (Svensson, 1999). First, the intermediate target variable must be measurable.
The measurement of the variable must be accurate and obtained within a short space of time so that it conveys the
appropriate message of whether the policy is on or off track. Second, the variable must be controllable. If the monetary
authorities cannot control the chosen variable then it does not matter whether the policy is on or off track because they
cannot do anything about it. Finally, and most importantly, the variable must have a predictable effect on the goal.
Even if the authorities can measure and control a variable, if it is not closely :elated to the goal then there is no point in
In order to appreciate the types of intermediate targets used and their relative successes and failings, what follows is a
discussion of the intermediate targets that have been used in the UK between 1976 and 1992. This will provide a
backdrop for the most recent developments in this area.
During the late 1960s and early 1970s the monetary authorities began to pay close attention to the growth of the money
stock. This importance attached to the growth of the money stock was officially recognised by the Labour government
in 1976 when the practice of targeting of monetary aggregates was first introduced. The main reasons for the explicit
target of monetary aggregates were due to the higher rates of inflation that were being experienced at the time brought
about by the failings of policy (fixed exchange regime and fine-tuning) and growing approval- at least with policy-
makers - of the Monetarists' assertion of a strong link between money growth and inflation. Milton Friedman, the best
known and most important of the Monetarists, provided much of this evidence. He provided
empirical evidence which, amongst other things, seemed to suggest that changes in the growth of the money stock
cause changes in the business cycle (e.g. Friedman and Schwartz, 1963).
The main advantage of using a monetary target is that it provides the authorities with an immediate signal on the
achievement of the target. The adoption of a monetary target also creates additional benefits in terms of creating an
automatic rule for the authorities. For example, Friedman advocated that the chosen money stock be targeted to grow at
a constant rate. Such rules, it is argued, give the authorities greater credibility in their pursuit of their goals (see below
for a fuller discussion of credibility and an alternative method of achieving it based on giving the central bank greater
One of the fundamental problems encountered when targeting the money stock is definition. Which money stock
should be controlled? Clearly the aim is to choose a monetary .aggregate which is closely related to expenditure. The
trade- off the authorities faced was between narrow definitions of money (i.e. measures which incorporate a narrow
range of financial assets) which were relatively easy to control by the authorities but not reliably related to expenditure,
and broader definitions of money (i.e. measures which include a much broader range of financial assets) which were
more reliably related to expenditure but less easy to control.
A good example of this was the £M3 ('sterling M3') measure which the UK authorities used between 1976 - 1986. £M3
was a broad definition of money because it included all cash in circulation and all sterling bank deposits held by UK
residents in the private and public sectors. It was chosen because it seemed to respond well to changes in expenditure.
Unfortunately the authorities were unable to meet the targets for £M3 with any degree of regularity Not only that, the
targets were being missed by considerable margins as the government attempted to bring inflation under control by
progressively lowering the target rate of sterling M3, a commitment established with the introduction of the Medium
Term Financial Strategy (MTFS) in 1980. In an attempt to revive the fortunes in monetary targeting the government
focused on alternative monetary aggregates. However, fortunes did not change mainly because the targets they could
control, such as MO which consists of notes and coins in circulation outside the BoE and banks' deposits with the BoE,
were too narrow and were likely to be caused by inflation rather than the cause of inflation. In 1986 formal monetary
targeting was officially abandoned.
The main reason for the abandonment of monetary targeting has been attributed to the instability of money demand.
Empirical evidence on the demand for money has occupied a great deal of time for many leading economists. The
literature has generally focused on two questions: is there a relationship between the demand for money and interest-
rate changes, and is the demand for money stable over time? What follows is a consideration of the latter question
because this is central to the use of monetary aggregates in the conduct of policy4.
The former question relates to the discussions of the mechanics of the monetary transmission mechanism. It
is based on whether the demand for money schedule is relatively flat (as Keynesians argue) or relatively
steep (as Monetarists claim).
The stability of the demand function began to break down in the early 1970s. In the UK, for example, studies such as
that conducted by Artis and Lewis (1976) based on data from the 1950s and 1960s were unable to accurately forecast
money stock in the 1970s. Ironically this breakdown occurred around the same time as the monetary authorities were
becoming interested in using monetary targets. The numerous works produced in this era were partly a response to try
to re-establish the stability argument. For an introductory analysis of the development of money demand studies and a
discussion of the main measurement issues, see Laidler (1993).
The impact of an unstable money demand function can be clearly seen if we reconsider Figure 2, which as previously
discussed represents the demand and supply of money This is reproduced in Figure 3. Previously we implicitly
assumed that the money demand schedule was fixed in one position. If we now introduce the idea that the demand
schedule is unstable and will therefore fluctuate in an unpredictable way (key point here being that the fluctuation is
unpredictable), say between Md0 and Md1, then any policy which attempts to target the money supply schedule will, at
best, make monetary policy more difficult. For instance, an expansionary monetary policy which seeks to shift the
money supply schedule from Ms0 to Ms1 will result in an interest rate which could lie anywhere between il and i2.
Because the effect on interest rates is unpredictable, the resulting impact on the economy will also be unpredictable.
The main cause of this instability has been attributed to innovation and deregulation in the banking industry Financial
innovation can affect the demand for money in many ways. For example the introduction of automated teller machines
(ATMs) or 'hole-in-the-wall' cash machines has led to a reduction in transaction costs - the shoe leather costs of having
to go to the bank to change money The most likely impact of this has been a reduction in the precautionary demand for
money More fundamentally; the payment of interest on, amongst other things, sight deposits (i.e. current account
deposits) brought about by deregulation and increased competition in the banking sector has changed the nature of the
relationship between money and other financial assets. With money acting as a means of wealth as well as a means of
payment the analysis presented earlier now only holds up if the interest payments on money do not change, or change
very little, when the market rate of interest changes. If, on the other hand, it increases or decreases as the market rate of
interest increases or decreases then this makes the conduct of monetary policy particularly difficult' because as the
market rate changes the return on money also changes and the relative opportunity cost between holding money and
other financial assets remains the same. Furthermore, the changing relationship between financial assets also has
implications for the choice of the appropriate definition of money
Exchange rate targeting
Exchange rate targeting, in contrast to monetary targeting, has long been a strategy open to the authorities.
Traditionally this has taken the form of fixing the value of the currency to a commodity such as gold in the gold
standard which operated on and off for over 200 years. In more recent times the UK has been involved in two episodes
of exchange rate targeting: informal linking of the pound to the deutschmark (March 1987 to September 1990) and
subsequent membership of the European Exchange Rate Mechanism - ERM (October 1990 to September 1992).
A clear advantage an exchange rate target has over a monetary target is that it ties down the inflation of traded goods.
As mentioned earlier, changes in the money supply and interest rates have implications not only for the demand for
domestic goods and services but also for the demand for overseas goods and services. It also reduces the uncertainty
that occurs when exchange rates fluctuate. Clearly these advantages increase as the amount of international trade the
country conducts increases. Unfortunately the downside of this is that it makes the targeting country more sensitive to
events occurring in the targeted country as was evident in the UK's withdrawal from the ERM in September 1992
following the speculative attacks on the pound.
The breakdown of the relationship between money growth and inflation and the problems encountered with exchange
rate targeting have resulted in a number of countries searching for a target in terms of the monetary policy objective.
Adopting an explicit inflation target is one such method that has become popular with monetary authorities. New
Zealand was the first country to go down this route in 1990. Since then Canada, Sweden, Finland, Australia and Spain
have all followed suit. The UK introduced an inflation target in October 1992.
It was mentioned earlier that the monetary authorities are unable to control inflation directly Although inflation
targeting focuses on a policy objective the authorities still use an intermediate variable. Here the intermediate variable
is the inflation forecast. This is of paramount importance because if, for example, the authorities choose actual
inflation in formulating policy then this clearly fails to take into account the time-lag between policy implementation
and ultimate outcome as previously mentioned. This is because current inflation is determined by previous decisions
and events, and has been likened to the pointless exercise of a dog chasing its tail (Haldane, 1998). What the authorities
should be, and are, concerned with is possible future inflation rather than present inflation. This is described as a
forward-looking view of inflation.
What is involved in a forward-looking assessment of inflation? In forming an inflationary assessment the monetary
authorities take account of many variables. Monetary aggregates and the exchange rate are used but unlike before are
monitored rather than targeted. In this respect it has been argued that both monetary targeting and exchange rate
targeting are limited versions of inflation targeting. Other variables, which also provide useful information about
inflationary trends, include survey data on price expectations and unemployment rates. The weights attached to these
indicators will vary from one time-period to the next depending on the authorities' assessment of their relative impact
on the economy Some would argue that such an eclectic approach could be interpreted as misleading and .designed to
confuse. My own view is that adopting a case-by-case approach is better because it does not discard information useful
for predicting future inflation, information that otherwise would be discarded if monetary targeting or exchange rate
targeting was used.
The complexity of the forecast has meant that the process by which it is determined needs to be transparent. The BoE,
since February 1993, has provided an assessment of progress towards achieving the inflation target in the form of a
quarterly Inflation Report. The centrepiece of the report is the so-called 'fan chart' (first published in February 1996).
Figure 4 shows the fan chart published in the February 2002 edi40n of the BoE Inflation Report and can be best
described as resembling a contour map (BoE Quarterly Bulletin, Nov 1998): the darkest band represents the most likely
projection, lighter bands represent less likely projections. The projections 'fan out' over time to reflect increasing
uncertainty in forecasting inflation many periods ahead.
Bank of England inflation forecast
Percentage increase in prices on a year earlier
Source: Bank of England Inflation Report, February 2002.
Although some of the issues surrounding the construction of the 'fan chart' are quite technical (for details, see Britton,
Fisher and Whitley; 1998) it is worth pointing out that the projection is based on unchanged interest rates. In one sense
this is problematic because it is highly unlikely that interest rates will remain unchanged over the two-year forecast
period. On the other hand, and as noted by Haldane (1998), the unchanged interest rate assumption is helpful in
providing relevant information about the likely effects of current policy
Central bank independence and the work of the MPC
Perhaps the single most important development in monetary policy in recent years has been the legislation that has
taken place in several countries granting greater autonomy in monetary policy-making to their central banks. Much of
this development has taken place around the same time as the movement towards inflation targeting discussed in the
previous section. In the UK, the BoE was granted 'operational independence' in 1997 and similar arrangements have
been set up in New Zealand and Canada, amongst others. Independence is also an integral component of the recently
formed European Central Bank (ECB).
The main theoretical reason for giving control of monetary policy to an independent central bank is the time-
inconsistent problem which governments with full discretion in the conduct of policy are likely to create (Kydland and
Prescott, 1977). It is commonly argued, for example, that a government that wishes to get itself re-elected may pursue
an overly expansionary monetary policy strategy just prior to the election in order to maximise votes and retain power.
Providing wages adjust only slowly; this has the effect of increasing employment and therefore output in the short-run.
In the long-run inflation builds up and interest rates will rise requiring contractionary policies. The cycle continues with
an expansionary policy just prior to the next election. This is the so-called political business cycle.
Crucially this requires expectations of firms and consumers to remain unchanged. If firms and consumers are rational
they will be, or eventually will be, aware of this (i.e. they will expect the politicians to act in this way) and will take
account of this in the prices they charge for their goods and services and the wages they will be willing to work for.
Either way firms and consumers will not believe government statements about their commitment to price stability as
they believe government policies are time-inconsistent. Because of inflationary expectations these policies are, as
emphasised by Barro and Gordon (1983), said to suffer from an inflation bias.
One way this time-inconsistency problem can be overcome is to make the central bank independent. Because central
bankers are not elected in the same way as politicians they are perceived as being trustworthier in their stance against
inflation and hence more credible. Whilst this is perhaps true, McCallum (1995) raises two important issues. First,
making the central bank independent in itself does not overcome the time-inconsistency because invariably the
governments still retain a degree of power - delegation is not absolute - which potentially still gives them the
opportunity to pursue pre-election policies by simply altering the terms of delegation. Second, if the government can
make a pre-commitment to independence why can it not make the same pre-commitment to inflation and avoid having
to give the central bank independence in the first place?
The answer to the first question obviously requires that there are sufficient costs involved in changing the 'rules of the
game'. This can be achieved by creating a legal framework which is specifically designed to limit a government's
ability to interfere. In response to the second question a common argument is that it is easier, from the point of view of
the general public, to determine whether the government has reneged on independence than it is to verify whether it has
reneged on its inflation objective.
As well as the theoretical justification there is also strong empirical evidence for central bank independence. Evidence
is usually provided on the basis of analysing variations in central bank independence across countries. In a study typical
of this area Alesina and Summer (1993) found the following.
. A negative relationship between average inflation and central bank independence.
. A negative relationship between inflation variability and central bank independence.
. No relationship between central bank independence and either average growth or the level of unemployment.
Although of some value, this and other studies like it must be treated with a degree of caution. Two common problems
have been identified with these empirical studies. The first and most important issue surrounds the measurement of
independence. Most studies use legal definitions in the construction of the level of independence. However, there are
numerous factors that contribute to the degree of independence (e.g. statutory guarantees of independence, methods of
appointment of officials on the committee and the range of instruments at the bank's disposal). Each of these factors
must be combined somehow, using appropriate weights, to generate a single measure of independence. As a result there
are enormous differences in interpretation. In a recent paper, Forder (1999), adopting a somewhat extreme position,
suggests that because we cannot determine objectively which factors are important all tests of independence are invalid.
A second problem is that a negative relationship between average inflation and central bank independence does not
necessarily imply that greater central bank independence is the cause of low inflation. In particular there may be
additional factors which are important. One such factor which satisfies this criterion, but at the same time is difficult to
measure, is people's attitude to inflation. Germany provides a striking example of this. The particularly nasty bouts of
hyperinflation - excessively high inflation - between December 1920 and November 1923, when price levels rose by a
factor of 66,342,065,000, have been ingrained within the nation's psyche. So even though the German central bank is
highly independent, the low and stable inflation that has consistently been achieved can be at least partly explained by
the German nation's aversion to inflation. It could also be argued that greater independence was itself driven by the
public's preference for stable prices. This once again indicates quite clearly the importance of expectations in policy
Remit of the Monetary Policy Committee (MPC)
For much of its existence the BoE has played a supporting role to the Treasury While the BoE implemented monetary
policy in a manner as described in the previous section, the elected government made the decisions about policy, such
as the timing of interest rate changes. In May 1997, following the Labour Party's victory in the General Election, the
Chancellor of the Exchequer, Gordon Brown, announced that the way monetary policy was to be conducted would be
changed. The announcement formally established a new organisation, the Bank of England Monetary Policy
Committee (MPC), which would be responsible for the conduct of monetary policy.
The remit for the role and constitution of the MPC is set out in the Bank of England Act 1998 (Rodgers, 1998), the
foundations of which were established by Gordon Brown back in May 1997. Since June 1997 the MPC has been
responsible for the setting of short-term interest rates.
The MPC's power is not absolute. It has officially been granted 'operational independence', which broadly means it is
free to choose an interest rate level subject to it being consistent with the government's objectives. These are:
(i) an inflation target for retail price index excluding mortgage interest payments (RPIX )5 of 2.5%;
(ii) without prejudice to the objective of price stability, the Bank will support the government's economic policy;
including its objectives for growth and employment.
It is clear therefore that whilst the MPC has the ability to use whatever instrument it feels is appropriate, it must do so
in a way in which the target can be met. In other words it has instrument independence but goal dependence. One of the
purposes of not allowing goal independence is that it reduces the potential problem whereby the government finds it
more difficult to coordinate monetary policy with fiscal policy. Specifically; if the MPC is able to set its own goal (i.e.
own inflation target) this may be below the government's target and one that is inconsistent with objective (ii). If, on
the other hand, the government is responsible for setting the target then it is likely to do so in a way that is consistent
with its fiscal objectives. In this case the MPC is simply the agent undertaking monetary policy on behalf of the
government (albeit with responsibility for both the 'timing and magnitude of the change), and as a result there should be
no conflicts in the coordination of monetary policy with fiscal policy The government can take operational control of
monetary policy in extreme economic circumstances but in order to maintain a credible reputation such powers require
the approval of Parliament. Each year the Treasury reviews the inflation target.
The main criticism of this remit has been directed towards its apparent incompleteness. Specifically; it has been noted,
for example by Bean (1998), that there is nothing within the remit which requires the MPC to achieve the inflation
target within a given time period, merely to achieve the target on average over the medium term. In some sense this
does actually give the MPC an element of goal independence by allowing them to specify their own time horizon in
which to be judged. As demonstrated by Artis, Mizen and Kontolemis (1998) this is analogous to the advantage the
British Prime Minister has over the American President in being able to choose an election date. However, by the same
token, choosing the appropriate time horizon may be too prescriptive and be subject to the same kind of criticisms that
have been directed towards rule-based behaviour in general.
RPI (or the 'headline' rate) includes mortgage interest payments whereas ::\PIX does not. Preference is
given to the RPIX for policy purposes because mortgage interest rates are closely related to official interest
rates. For example, a rise in interest rates designed to dampen inflationary pressures leads to an immediate
rise in the RPI. Another common inflation measure is the RPIY In addition to mortgage interest payments
this measure also excludes local authority and indirect taxes. It is designed to measure inflation without the
effect of government driven price changes. Invariably the RPI produces the highest measure of inflation
and RPIY the lowest. Taking 1998 as a typical example. the RPI was recorded as 3.4, RPIX was 2.7 and
RPIY was 2.1. It is important, therefore. to know the type of inflation measure being used in a particular
study There is also a long-running debate about whether more volatile components of the RPI such as food
and energy should be excluded because they are particularly prone to seasonal effects and supply shocks
respectively (see below). In addition, it is readily recognised by these authors that these potential concerns
have been overcome by the introduction of transparency and accountability measures.
Why a target of 2.5%? Why not an inflation target of 0%? One reason for targeting a moderate amount of inflation
which allows for some increase in the price level over time is that the existence of cash means interest rates cannot fall
below zero (if they did, everyone would obviously withdraw deposits from their banks and store it under their beds
etc). What we are referring to here is what is known as the nominal (or actual) rate of interest. The nominal rate of
interest does not take account of inflation. In contrast, the real rate of interest does take account of inflation: it is the
nominal rate of interest minus the inflation rate. If, say; the target is 0% inflation, this necessarily means the real
interest rate also cannot fall below zero. But in some instances, particularly during periods of recession when spending
is low; the authorities may want to stimulate spending by making real interest rates negative (inflation rate greater than
the nominal interest rate). This will entice greater spending today because of the purchasing power of cash, say one
year from now, is lower.
The MPC are not required to hit the 2.5% target all of the time. In fact, the target is allowed to fluctuate 1% either side
of this target (i.e. between 1.5% and 3.5%). Only if it moves outside of these boundaries will necessary steps have to be
taken. The reason for this symmetry is to ensure that the MPC treats excessively low inflation in the same regard as
excessively high inflation. The boundaries themselves are introduced to allow for the effects of temporary shocks and
practical considerations arising from errors in measuring inflation from one time period to the next.
MPC structure and conduct
There are nine members of the MPC - the Governor of the BoE, two Deputy Governors, the BoE official responsible
for monetary policy analysis and the BoE official responsible for monetary policy operations; these latter two positions
are appointed by the Governor after consultation with the Chancellor. The remaining four members, appointed by the
Chancellor, are considered outside professional experts. The current members (as at October 2001) include Professor
Stephen Nickell (School Professor of Economics at the London School of Economics and President-Elect of the Royal
Economic Society) and Kate Barker (Chief Economist at the Confederation of British Industry). A representative from
the Treasury also attends meetings but, unlike other members, has no voting rights. His/her role is to inform the MPC
of developments in matters such as fiscal policy that might be relevant for achieving the inflation target and to inform
the Chancellor and the Treasury on how well the Committee is functioning. This is an added check of the coordination
policy identified previously.
A pre-MPC meeting takes place on the Friday before the regular meeting, the purpose of which is to allow the BoE's
professional staff to brief the MPC members on the prospects for inflation. The MPC meeting itself occupies two days
each month, normally the Wednesday afternoon and Thursday morning following the first Monday of the month. The
only exception to this so far has been the special meeting of the 18 September 2001, following the terrorist attacks in
the USA. Decisions on interest rates are announced immediately after the last day and are determined on the basis of
majority voting. The Governor has the casting vote in the event of a tie.
The over-riding aim of the MPC is to create a reputation for low and stable inflation. In order to achieve this it must
convince firms and consumers that the way it is conducting monetary policy is credible. It is commonly perceived that
in order to achieve credibility the MPC must be both transparent in its operations and accountable for its actions 6. The
main way transparency is achieved is through the publication of the MPC meeting minutes. The 1998 Bank Act had
originally stated that the minutes should be published within six weeks of the meeting. This was reduced to two weeks
in October 1998 in order to keep financial markets better and more fully informed. The minutes contain discussions of
all the views on recent economic developments and the possible consequences these might have on output and prices.
Although the specific views of individuals are not recorded in the minutes, information on how each member voted is.
Some recent statistical work on the voting patterns of the members has found that BoE representatives are less likely to
disagree than the outside experts (H.M. Treasury 2001). Of course this merely confirms the propensity for economists
A further source of transparency is the open letter which the MPC is required to send to the Chancellor if inflation
moves outside its allowed band. Because the target is symmetric, a letter is required if the target goes beyond 3.5% or
falls below 1.5%. The letter must contain an explanation of why inflation has deviated outside the target range, what
actions the MPC intend to take to get it back within the target range, how long this will take, and how this is consistent
with the government's economic policy objectives.
As well as the publication of MPC meeting minutes and the open letter system, transparency can take many other forms
including speeches by MPC members and research and analysis conducted through the recently established MPC unit.
These add to the more established transparency measures, such as the quarterly Inflation Report, already in place and
Each of these transparency measures also acts as a device to ensure public accountability Government accountability is
achieved through a House of Commons Treasury Select Committee and a House of Lords Select Committee. These
committees regularly call upon MPC members to explain their decisions and make speeches detailing their personal
views on the economy Amongst other things, the Select Committees have been influential in requesting the speeding up
of the publication of the minutes and in the publication of information, including a need for the publication of a report
on the MPC's view of the transmission mechanism of monetary policy
One aspect that could seriously undermine the transparency and credibility of the MPC is the revised role of the
Inflation Report. When inflation targeting was first introduced the BoE, acting as an advisor to the government,
produced independent inflation forecasts. Under the present arrangements the Inflation Report represents the views of
the I\I1PC. As a result the committee now not only makes interest rate decisions but also assesses these decisions,
which means the Inflation Report can no longer be considered independent. Goodhart (2001) has suggested a solution
to this problem. It involves making the Governor of the BoE personally responsible for the Inflation Report with the
other members of the I\I1PC giving their approval or otherwise (by having the opportunity to present an alternative
through the MPC unit or some other forum) to that Report. Goodhart argues that such a mechanism would work better
than some alternatives, such as giving the BoE responsibility or certain members of the Bank such as the Deputy
Governor, because this would be widely perceived as giving the Bank members greater influence over the decision.
The Governor, unlike other Bank members. does not get directly involved in the forecasting process and can. therefore,
act as an entirely independent arbitrator between members, and perhaps between members and (BoE) staff (Goodhart.
The use of transparency and accountability is not the only way of achieving credibility The German
Bundesbank. for example, has achieved credibility via its consistent anti-inflationary performance over
the past fifty years.
How MPC decisions influence the interest rate
In order to achieve its inflation target the MPC must be able to control the amount of money in circulation. This does
not simply mean controlling the amount of notes and coins. It also requires the control of money generated via the
banking system (bank deposits, etc).
There is a number of techniques that the MPC can use in order to achieve this. One option is to make sure banks hold a
minimum amount of cash or other very liquid assets. By altering the size of the reserve ratio (the ratio of reserves to
total deposits), banks will have more or less opportunity to lend depending on whether the reserve requirement has
decreased or increased. Alternatively; they could be required to keep a certain amount of deposits with the BoE.
Reserve requirements have been widely used in Germany; Greece and Italy and continue to be used in the USA. The
UK abandoned mandatory reserve ratios in 1981, favouring instead a policy that allows banks greater freedom to set
their own reserve ratios.
A second option to manipulate the money supply requires altering the interest rate charged by the discount houses. The
discount houses play the role of intermediary between the BoE and banks and other financial institutions. The key
function here is the central bank's role as the lender of last resort. This role of the central bank is fundamental to the
efficient functioning of, and maintaining confidence in, the financial system. Because of its lender of last resort role.
the central bank is prepared to provide extra lending to banks. This extra lending :nay be required, for example, when
banks are short of funds. Banks may find themselves short of funds for numerous reasons. For example there may be a
sudden increase in customers withdrawing funds. Alternatively shortages may arise as a result of specific central bank
policy; such as increases to the reserve requirement or open-market sales of government securities (such as Treasury
bills or gilts) to the general public; the latter has the effect of transferring the public's bank deposits to the central bank.
In such instances the central bank can take advantage of its lender of last resort role because the banks effectively have
nowhere else to turn to (the central bank becomes the monopoly supplier of funds). However this only applies if there
is a general shortage of funds. If a single bank is short of funds it could borrow from another bank through a special
market which deals with bank-to-bank transactions - the inter-bank market. AssUI:1ing there is a general shortage the
central bank can alter the rate of interest, in this instance the discount rate, it charges to the banks in accordance with
what it wants to do to the money supply: For example, if the central bank wanted to reduce the money supply it would
increase the discount rate relative to the market rate of interest.
The third, by far the most common, and arguably the most important. monetary policy tool is the use of open market
operations. Open market operations are seen as a more suitable instrument for the purposes of monetary control
because they are a more flexible alternative to reserve requirements. It is through open market operation that MPC
announcements about interest rate changes are achieved. In other words it is through open market operations that the
announced interest rate becomes the equilibrium interest rate. In order to maintain this equilibrium interest rate the
central bank conducts daily open market operations. As with discount policy; the success of open market operations for
monetary policy purposes depends on banks being short of funds.
An open market operation involves purchases or sales of government securities (Treasury bills and bonds) by the BoE
to commercial banks and/or the general public. The discount house once again plays the role of intermediary A recent
addition to the instruments used in open market operations has been 'repos', which is short for sale and repurchase
agreements. In fact, repos are now the standard instrument used by the BoE in its open market operations. A repo
works as follows. One party, say the central bank, buys some gilts or bills from a particular bank which has a legally
binding agreement that it will buy them back from the central bank at some future date (usually two weeks later) for a
specified price. (A reverse repo works in much the same way but this time involves the central bank selling gilts or
bills.) It is this specified price which in a similar way to the discount rate. can be manipulated for monetary policy
purposes. In effect, the difference between the sale and repurchase prices reflects the interest rate announcement made
by the MPC.
Policy rules versus policy discretion
In theory the MPC has full discretion in its choice of interest rates, subject to the achievement of the government's
economic objectives. An alternative would be to use pre-determined rules regardless of the state of the economy A
well-known example is the Taylor rule, developed by the American economist John Taylor. The Taylor rule indicates
that the interest rate should be determined by the amount by which inflation exceeds its target and the amount by which
output (growth) exceeds its target both weighted by one half. The formula also includes the inflation rate itself and the
equilibrium real rate of interest (that is, the real rate of interest consistent with full employment). As an example,
suppose UK inflation stands at 2%, 0.5% below its target of 2.5%, and growth were 0.5% below its potential. If we also
assume, following Taylor, that the equilibrium real interest rate is 2% then the Taylor rule suggests that the interest rate
should be set at 3.5% (2% inflation plus 2% real interest rate minus ½ of 0.5% for the inflation gap and similarly minus
½ of 0.5% for the output gap. Naturally these latter two components would become positive if the inflation and output
were above their respective targets). Such rule-based behaviour can be seen as an alternative solution to the time-
inconsistency problem because it does not require granting greater independence to the central bank.
The consensus within the literature is to favour a more discretion based policy together with greater central bank
independence. The main reason for this is that it is able to combine the benefits of credibility (brought about by greater
central bank independence) with the flexibility to adjust policy in the light of changing economic conditions. Rule-
based behaviour, whilst credible, in many ways requires too much information and too many assumptions in order for it
to be operational. The Taylor rule, for example, requires assumptions about the current level of potential output and the
equilibrium real rate of interest and even simple rules such as the constant money growth rule require the assumption of
a stable money demand schedule.
Although not working to an explicit rule there is some evidence to suggest that the MPC do follow an element of rule-
based behaviour in the setting of interest rates. Of the 54 decisions taken by the MPC between June 1997 and October
2001, 9 (17%) decisions were made to increase interest rates, 13 (24%) decisions to reduce interest rates and 32 (59%)
decisions to leave interest rates unchanged. In general, the changes to interest rates have been quite small (between
0.25% and 0.6%, or 26 and 60 'basis points' as it is more commonly expressed), often in the same direction over several
months. In one case, between October 1998 and February 1999, there were five consecutive reductions - three at 0.5%
and two at 0.25%. The rules they seem to be working on then are (i) not to undertake large increases or decreases in
interest rates and (ii) to avoid interest rate reversals (i.e. one month up followed by another month down, or another
month down followed by another month up).
The policy of small interest rate changes in the same direction is not new: Cukierman (1992) argues that one
explanation for this phenomenon is due to the BoE's concern for maintaining stability in the financial sector. This is
particularly true of banks whose main function is to transfer short-term deposits into long-tern loans. For example, an
unanticipated increase in interest rates may adverse1y affect bank profitability given that on the one hand it may have
made a commitment to offer a loan with a set (fixed) rate of interest, but on the other hand would be forced, through
competition, to offer higher rates on savings. Another explanation could be due to the uncertainty that surrounds the
effects of changes in the policy instrument; many small steps are better because large changes are associated with
greater variability in inflation due to the long and variable lags involved. Finally, Goodhart (1998) offers a simple but
persuasive argument that frequent reversals may call into question the MPC's credibility.
It is undeniable that the MPC has made a good start. Between May 1997 and September 2001 annual inflation has
averaged 2.37%. It also appears to be more stable, lying within a range of 1.4% (minimum inflation during this period
being 1.8% and a maximum of 3.2%). In contrast, during the 1970s inflation averaged 13.5%, 7% during the 1980s,
and 4.5% between 1990 and April 1997 (see Figure 5). Furthermore, the commitment to price stability does not seem to
have had a detrimental effect on either growth or unemployment. On the contrary; the evidence suggests that price
stability has had a positive effect on both; growth rates have remained steady at about the 2% mark and the
unemployment rate as at October 2001 stands at 6.1% (around 1.6 million)7.
RPIX inflation (1976-2001)
Source: Office for National Statistics.
Office for National Statistics.
Nevertheless it is also true that some of this success can be put down to the international climate of the last few years, a
climate which has been conducive to low and stable prices. This is evident in a number of the BoE's press releases. For
example, on 10 December 1998 the MPC reduced interest rates by 0.5% to 6.25% for the following reason:
Since the November Inflation Report, the prospect for global activity appears to have weakened and commodity prices
have fallen further. In the UK. GDP growth in the third quarter was revised down. and surveys of activity have
continued to indicate deterioration across the economy; although the labour market is still tight and monetary and
financial indicators remain relatively strong.
Furthermore, if we were to judge the MPC solely on its achievement in actually hitting the inflation. target we would
have to conclude that it has failed. As highlighted in Figure 6 there have been only five occasions when the target has
been met. the last of which was in November 1998. Since then. thirty-one of the thirty-five rates have undershot the
target. The evidence also suggests that monetary policy has become progressively tighter; the average deviation in 1997
was +0.29, but by 2000 the average had become -0.43. It would, however, be unfair and in many ways short -sighted to
claim this as a failure because although the point target has generally not been met, at no time has the target fallen
outside of its permitted zone (1.5% and 3.5%). The persistent undershooting of the target since June 1998 - the
likelihood of inflation being below the 2.5% target rather than above it - is of far greater concern and one which may; if
it persists, ultimately damage the MPC's credibility
RPIX inflation under the MPC 3.5
Source: Office for National Statistics.
Undoubtedly the most important way to evaluate the performance of the MPC is to analyse what it has done to
expectations. It has been mentioned several times in this chapter that credibility is ultimately determined by
expectations. Although work in this area is still in its infancy a number of preliminary conclusions can be drawn.
The first impact following the announcement of the new monetary framework was an almost instantaneous drop in
inflation expectations as perceived by the financial markets. Haldane (2000) refers to this as an instant credibility
windfall. Further falls have followed since. In contrast, Wadhwani (200l) notes that previous arrangements prior to
May 1997, such as the government controlled inflation target, lacked credibility because the expected inflation rate
exceeded the target inflation rate.
Independent verification, as provided by economic forecasters, also provides evidence for the continued fall in
inflationary expectations. The forecasts also suggest that inflationary expectations have become much more stable. For
example, Consensus Economics survey forecasts for monthly inflation since 1997 have been very close to the target.
Finally the BoE recently commissioned a series of surveys to gauge public opinion and understanding of the new
monetary policy arrangements and found that in general the public believe that the MPC is committed to price stability.
Most of the monetary authorities around the world have price stability as their number one priority In order to achieve
this the authorities cannot simply use a policy instrument and wait and see what happens. They must evaluate the
timing and effectiveness of such changes and have a proper understanding of the ways changes in monetary variables
are transmitted to the economy. Because or the long and variable lags involved most monetary authorities aim for an
intermediate target. The UK authorities have used several different targets in the past but since 1992 have adopted an
intermediate target based on inflation forecasting.
The effectiveness of monetary policy ultimately depends on the expectations of consumers and firms. The central
question is whether the authorities are seen as credible in their stance against inflation. Undertaking some rule-based
behaviour is one possibility in eliminating the credibility problem. A more favoured alternative and one that is
increasingly being used is to give the central bank greater independence in the setting of the policy instrument.
In the UK, the combination of inflation targeting and central bank independence, through the creation of the MFC,
seem to have created an environment of low and stable inflation. Even now, after four years, it is perhaps still too early
to determine whether the regime change has resulted in lower inflation or whether its introduction has coincided with
the downturn in output. However. what does seem to be clear is that increased transparency and accountability have
enabled it to gain a better inflation reputation.