CENTRAL BANK MODERNISATION

Central bank operational

efficiency: meaning and
measurement

Vern McKinley, BearingPoint

and King Banaian, St Cloud State University

CENTRAL BANKING PUBLICATIONS

This is an extract from Central Bank Modernisation (ISBN 1 902182 38 3),

Published by Central Banking Publications Ltd
Central bank operational
5 efficiency: meaning and
measurement

Vern McKinley and King Banaian

“Not all Germans believe in God, but all believe in the Bundesbank.”
Professor Otmar Issing, quoting Jacques Delors
St Edmund’s College millennium year lecture, Cambridge, 26 October 2000.

Central banks are generally among the most trusted and respected of public
institutions. One reason for this respect is that many central banks, especially over
the last quarter of a century, have been seen as highly effective in controlling
inflation. However, while many central banks are seen as effective, as a class of
institution they are rarely described by outside observers as efficient.1 Furthermore,
while there is an extensive body of literature examining the institutional
arrangements most likely to make central banks effective, there is very little analysis
of how to make them efficient. In part, this reflects the very real methodological
difficulties of comparing such diverse institutions. However such analysis is
overdue and requires a multi-disciplinary analysis that involves application of legal,
policy, econometric and accounting principles.

This chapter will begin by considering why we should care about central bank
efficiency. We will then present a basic definition and examine how that relates to
central banks and their activities. We will consider the broad variety of functions
that central banks do in fact perform and look at how central banks might be prone
to inefficiency as a result of taking on too many functions. Finally, we provide an
index of central bank operational efficiency and make some suggestions about how
central banks might go about reviewing their operations to become more efficient.

Why measure central bank efficiency?

Before we consider how to measure central bank operational efficiency, we should
first consider whether it is worth measuring. Central banks are primarily held

45
Central Bank Modernisation

accountable for meeting the objective of monetary stability, not whether they do so
efficiently. After all, central banks are granted a legislative monopoly in many of
their operations, which naturally reduces pressure to be efficient. The value of an
effective central bank – one that controls inflation, and helps avoid banking or
currency crises – is extremely large. The costs of running a central bank, even a
wasteful one, are small by comparison. If a central bank is judged as effective in
meeting its objective, its relative degree of efficiency may seem less worthy of
analysis. Following this logic, any attempt to pursue efficiency gains is misguided
if it threatens even marginally the central bank’s effectiveness.

We believe this analysis is misconceived. First of all, it should not be assumed that
there is a trade-off between efficiency and effectiveness, so that efficiency gains
necessarily threaten effectiveness. A central bank can improve the efficiency of its
operations, and by doing so improve its effectiveness as well. One obvious example
would be action to improve the governance of a central bank. If the board of a
central bank is wasting time by overseeing unimportant functions or activities, then
the board’s (and central bank’s) performance will be sharpened by shedding these
functions even if this is done in the interests of efficiency gains.

A second justification for focusing on efficiency is to ensure that a concentration of

power within the central bank does not develop. If a central bank takes on too many
roles, and thus too many aspects of economic policy depend on the management of
the central bank, the fate of the entire economy is placed at risk. If there is truly a
case of market failure, whereby a private-sector solution does not efficiently address
a market need and government must intervene, a government agency outside the
central bank might be a more appropriate place to house such functions.2

Thirdly, if a central bank takes on functions where there is no evidence of market

failure, the central bank may crowd out private sector providers of the same
services. Funding or other advantages may give a central bank capacity that is
difficult for a private entity to overcome and thus to compete effectively. Overall
the efficiency of the economy will suffer as a result.

Fourthly, it should be noted that central banks are in almost all cases public
institutions whose profits usually contribute towards the national budget. So
improved efficiency can lead to an increased flow of funds to the government as a
whole.3

Finally, it should be remembered that the current model of independent central

banking, under which central banks are directly accountable to parliaments, makes
central banks vulnerable to accusations of mismanagement of resources. As central
banks adopt more transparent financial reporting arrangements, central bank
policymakers can expect their management to come under increased political
scrutiny. Perceived inefficiency makes the central bank vulnerable to criticism by

46
McKinley and Banaian

politicians, which can undermine central bank independence and interfere with
policymaking.

Defining and measuring efficiency

Having established that it is important for central banks to think about the efficiency
of their operations, a definition of the term is in order.4 Efficiency is best understood
as the cost per unit of output, describing the relationship between the goods and
services produced by a program or activity (outputs) and the resources used to
produce them (inputs).5 Put differently, an activity generating a given output can be
said to be efficient if there is no alternative method of generating the output using
less input.

Effectiveness is often used as a synonym for efficiency, but they are not the same.
Effectiveness relates to achieving an objective ,6 while efficiency relates to the cost
or effort to achieve that objective. So, in comparison to effectiveness, which is
focused solely on outputs, efficiency is focused on both outputs and inputs.

Commercial bank indicators of efficiency

An extraordinary amount of research has been done on efficiency in commercial
banks and financial institutions (much, in fact, by staff of the US Federal Reserve
System).7 In this context, efficiency is defined in terms of cost minimisation or
profit and output maximisation in comparison to best practice. Despite the fact that
central banks are merely specialised, government-controlled financial institutions,
researchers have not committed similar efforts to analysing central bank operational
efficiency.8

Government indicators of efficiency

The analysis of governments and their efficiency cannot help but differ from the
analysis of private-sector entities. Many government entities are monopolists, so
there is no competitive environment to fully stimulate efficiencies. Thus, it is not
surprising that studies indicate that government entities are less efficient than private
entities.9 In addition, it is difficult in many instances to achieve efficiencies given the
political pressures on government agencies. Some researchers argue that the
incentive structure of many government entities is biased more toward expanding
responsibilities and power as opposed to achieving operational efficiency.10

A definition of central bank operational efficiency

The challenge in developing a definition of central bank operational efficiency is in
taking the definition of efficiency in common usage, assimilating how efficiency is

47
Central Bank Modernisation

analysed in relation to financial and governmental institutions generally, and then

applying this to central banks and those activities that a central bank engages in.
Undertaking this analysis leads to some general conclusions about defining central
bank operational efficiency:

Inputs and outputs: the required output for a central bank is determined by its
particular objective(s). A central bank’s inputs are the resources (labour and capital)
used to attain this objective: staff, office space and data processing equipment.

However, unlike competitive financial institutions there are generally no relevant

peers for comparison within a country.11 Thus to determine efficiency in a
comparative sense we need to look to other central banks abroad. To do this,
however, requires the imposition of controls for the state of economic development,
country size and the like, that allows us to act as if two central banks were
identically situated (see section on specifying a central bank index below).

Applying these observations, the definition of central bank operational efficiency

that will be used throughout the course of this chapter is as follows:

The measure of how central banks use resources or “inputs” (labour and
capital) to implement their various granted functions in pursuit of their objective
or “output,” as compared to peer central banks.

Identifying inputs and outputs

Based on our definition, a number of documents must be referenced to attempt to
measure central bank operational efficiency. Obviously, if we cannot identify
required outputs, we cannot measure efficiency, and for this reason the most
important document will be the central bank’s legal mandate which determines what
a central bank’s ultimate goal should be and what it can and cannot do in pursuit of
that goal. As can be deduced from a review of these objectives, most are focused on
price and monetary stability.12 Typically, central bank laws also detail the specific
functions that a central bank can carry out to meet its objective. Often the more
important of these functions are described in some detail (monetary policy, foreign
exchange management, and bank supervision).

Once the legislative structure and objective are identified, the central bank is
generally given wide discretion as to how it independently develops the underlying
operational processes to achieve its objective. If the management of the central bank
is diligent in developing efficient processes and undertaking activities directly
linked to its objective, then this broad grant of discretion is well advised. However,
too much latitude can lead to pursuit of a broad range of activities where the link to
the ultimate objective is unclear (see box 1).

48
McKinley and Banaian

Box 1: An invitation to mission creep?

While many central bank laws commit an entire chapter of provisions to
monetary policy,13 it is also remarkable how brief are the descriptions of
monetary policy in some statutes. Some provide so little detail as to the powers
or functions delegated that they seemingly put almost no limitations on the
pursuit of price and monetary stability. For example, some monetary policy
provisions are contained in a single sentence,14 others in one or two articles,15
while another simply states that the central bank must act as a central bank.16
Other central banks are granted sweeping powers to undertake activities at their
own discretion. For instance the Kazakhstan central bank statute provides for
the bank to:

“have the right to participate in the creation and activities of organizations

that assist the National Bank of Kazakhstan in the performance of its
assigned functions and/or are part of the infrastructure of the financial
market;”17

In the case of Kazakhstan, such a broad authority has allowed the central bank
to engage in a wide variety of activities, including holding stakes in a deposit
insurance corporation and a mortgage crediting and secondary market
company.18 Were it not for a set of relatively detailed and transparent financial
statements, it would not have been possible to determine that the central bank
was engaged in such wide-ranging functions.

Identifying inputs is more difficult. In order for external observers to assess how
efficiently a central bank operates, it must be transparent in its accounting and
reporting of how it uses its powers and resources to meet its objective. This
transparent presentation of operations should include annual financial statements,
notes to those statements and underlying information on employees. In the past, the
opaque nature of many central bank financial statements made this difficult.
However, as central banks move towards more transparent financial reporting (a
majority of central banks, according to recent survey data, employ some version of
IAS 19) this comparison has become easier.

Risks to efficiency
Inefficiencies at central banks can be traced to two primary sources: legislative
inefficiencies and managerial inefficiencies. Legislative inefficiencies result from

49
Central Bank Modernisation

the granting of an inappropriate legal mandate. This could occur where a legislature
gives the central bank too much latitude in how to perform its objectives (see box
1). Or the legislature may require the central bank to undertake too many, or
conflicting, roles. This is a case of an exogenous factor beyond the control of the
central bank, as legislatures may impose duties on a central bank that are unrelated
to their objective.

Managerial inefficiencies occur when an appropriate function is implemented in a

wasteful fashion. Examples might be the maintenance of an unnecessarily extensive
branch network, or maintenance of high staffing levels due to a lack of outsourcing
or insufficient leveraging of technology.

Identifying legislative and managerial inefficiencies

Clearly, identifying legislative and managerial inefficiency is not a straightforward
task. Each central bank operates within a particular national economy, and within a
particular political environment. In small and developing economies where financial
resources may be limited and highly skilled human resources are in great demand
there may be great pressure to load new functions onto the central bank, especially
if the central bank is perceived to be more capable or independent than other
branches of the government. Making judgements about which functions a particular
central bank should ideally perform is always going to be problematic. However,
there is an existing body of research into what in fact central banks are generally
tasked to do, and this may prove helpful in identifying which functions are at the
core of central banking, and those that are not. Table 1 summarises previous
research.

Most interesting for our discussion is to note how varied central banks’ main
functions are, despite the apparent consensus about what central banks exist to do.
Presumably, the justification for granting each of these powers and responsibilities
to a central bank is that they are necessary for the achievement of its main objective.
If powers or responsibilities are not directly related to the central bank's objective,
then there is a strong argument that it should not be burdened with these
responsibilities. Scarce management time, and the central bank's other resources,
are better spent pursuing the central bank's main objective. In recent years a number
of central banks have made concerted efforts to refocus the direction and improve
the efficiency of their operations, and in fact these attempts usually do begin with
this kind of re-examination (see boxes 2, 3 and 4 below).

50
McKinley and Banaian

Table 1: What do central banks do?

Study Tasks/Roles/Functions

Pollard 20 Define and implement monetary policy (referred to as the “primary function”
of a central bank)
Issue banknotes
Conduct foreign exchange operations
Hold and manage official reserves
Act as the fiscal agent for the government
Promote stability of the financial system
Supervise and regulate banks (and the related power as a lender of last
resort)
Implement consumer protection laws
Promote the smooth operation of the payment systems
Collect statistical information
Participate in international monetary institutions

White 21 Issuer of currency

Bankers’ bank
Regulator of commercial banks
Lender of last resort
Conductor of monetary policy

Fischer 22 Monetary policy – managing supply of credit and money and determining
market interest rates (referred to as the “essential” central bank function)
Determining exchange rate and managing foreign exchange reserves (fully or
jointly)
Hold reserves of commercial banks
Managing the payment system
Supervising banks and other financial institutions
Lender of last resort
Administering deposit insurance
Government’s banker
Administer foreign exchange controls
Manage all or part of the national debt
Policy research
Development banking

Green 23 Fiscal services to the government

Creditors’ agent
Facilitate settlement of interbank claims
Lender of last resort support to banks in crisis

51
Central Bank Modernisation

Box 2: Federal Reserve Bank of Cleveland

In the late 1990s, the Federal Reserve Bank of Cleveland undertook an ambitious
review of the bank’s major processes.24 This review categorised the Cleveland Fed’s
work into eight discreet processes to analyse whether the bank was deriving value
from its operations.25 Teams staffed by officers and managers from across the bank,
and aided by an external consultant, reviewed functions and made recommendations
for improvements.

Initial recommendations targeted those “quick hits” that could be implemented in a

short period of time. Long-term changes involving significant reengineering were also
considered. One result of this review was the relocation of the bank’s cash-related
functions from Pittsburgh, Pennsylvania to Cleveland, Ohio resulting in lower labour
costs and expanded use of automated vault processes in the new Cleveland base.

In addition, as part of the long-term changes in operations, several managers and

employees were trained in process mapping and improvements, and in approaches to
quality. Benchmarking performance also became a regular practice and strategy
implementation was tracked through the use of a balanced scorecard. Some operations
were also outsourced as part of the effort. One conclusion of the review was that
industry-standard process improvement methodologies could be applied to central
bank operations with favourable results.

Core and non-core functions 26

Our own research of over 100 current central bank laws points to the following
seven “core functions” of central banks (“core” in the sense that nearly all central
banks presently undertake some or all of these unique central bank activities):

• Monetary policy management;

• Foreign exchange and reserves management;
• Lender of last resort;
• Supervision and regulation of commercial banks;
• Payment and settlement systems;
• Currency and coin management; and
• Fiscal agent

Central bank non-core functions

Alongside these seven core functions, there are a number of other duties that only a
small number of central banks perform and which do not appear closely linked to
the objective of a central bank. Many of these activities are meant to be temporary

52
McKinley and Banaian

Box 3: Sveriges Riksbank

Sweden’s central bank, the Sveriges Riksbank, provides another example of a
broad-ranging campaign to improve efficiency in a developed market. The
Riksbank undertook a long-term self-examination process in the mid-1990s,27
involving a review and evaluation of every area of operations by the staff of the
bank. This review examined each area of the Riksbank’s work and asked the
following three questions:
• What are the real responsibilities of the central bank?
• For whom do we perform our services?
• What level of performance should we expect?

This analysis was done within the context of a more defined focus on the central
bank’s primary objectives: price stability and financial stability.28 Ongoing analysis
of operations over the course of a decade has seen the following changes:
• A strict focus on developing analysis and competence regarding inflation targeting
and financial stability.
• Decisions to outsource/close some main operations including outsourcing
production of some financial statistics, and closing all but one branch as the cash
distribution process was reorganised.
• Dramatic reductions in staff from 1,136 in 1995 (including those employed in
wholly owned subsidiaries) to approximately 430 today,29 with much of the
reduction attributable to the changes in cash distribution. The bank is aiming for
another 10% reduction, to approximately 400 employees, by the end of 2006. By
one measure (number of central bank employees per 100,000 inhabitants) Sweden
has become one of the most efficient central banks in Europe. 30
• A sharpening up of the bank’s strategic planning with intensified focus on
establishing measures and indicators for performance in the different processes
undertaken by the bank.

in nature. For example, developing a credit bureau or engaging in mortgage

activities, with the expectation of passing the activity to private sector operators
once the activity is sufficiently developed and capable of being financially
independent.
• Non-bank financial institution supervision;
• Management of deposit insurance system;
• Unified supervisor of financial institutions;31
• Resolution and liquidation of failing banks;32
• Financial intelligence unit for anti-money laundering;33
• Credit bureau;34
• Mortgage or housing activity;35

53
Central Bank Modernisation

• Open bank assistance;36

• Agricultural related activity;37
• Small and medium enterprise loans;38
• Development banking;39
• Financial institutions development fund;40
• Loan to non-banks (corporations);41
• Export credit;42
• Ownership of a bank or state bank/loan activity; and 43
• Consumer protection.44

Box 4: Central Bank of Sri Lanka 45

Since mid 2000, the Central Bank of Sri Lanka (CBSL) has been undertaking an
examination of its objectives, functions and efficiency. The conclusion of this analysis
was that the CBSL should focus on achieving price stability and financial system
stability. During early 2002, working primarily with the World Bank, the CBSL
announced a $42 million program 46 designed to:
• Build the professionalism of CBSL staff and move the institution towards a merit-
based system of incentives and rewards;
• Provide hardware and software to ensure that large value and time-sensitive
payments within the economy can take place in “real time” rather than through
end-of-day clearing between the banks;
• Enhance the quality of key data to improve management of monetary policy,
exchange rate and foreign reserves; and,
• Support computerisation of the general ledger.47

As a result of these efforts 48 the CBSL statute was amended to redefine the bank’s
objectives as the pursuit of price and financial stability, replacing its previous multiple
objectives.49 The CBSL designed a new organisational structure in line with these new
objectives, and implemented other changes resulting in:
• Divestiture of the Sri Lanka Automated Clearinghouse, which handled the
operations of cheque clearing and off-line funds transfer system;
• Transfer of ownership of the CBSL’s shares in the regional development banks
above 20 percent; and
• Outsourcing of many support services.

These reforms led to some categories of employees being made redundant. The bank
set a target reduction in work force of roughly 50%, and by early 2002 the CBSL
offered a voluntary retirement package resulting in 690 employees retiring from the
bank. This package was largely funded by the World Bank (about $25 million) and
followed previous unsuccessful efforts at restructuring within the Sri Lankan
government, which were circumvented by political interference in the process,
resulting in some staff being rehired after receiving a retirement package.50

54
McKinley and Banaian

Linking functions to central bank objectives

As detailed above, the various functions that central banks undertake are wide-
ranging. In some ways this is not surprising given the very different economic and
political environments in which central banks operate. However, it is not clear when
these powers were granted that consideration was given to limiting activities to
those directly needed to meet the statutory objective. For example, central banks that
often have the same overall objective of price and monetary stability engage in very
different underlying functions to meet that objective. Table 2 details how two
sample central banks have similar objectives, but different functions in pursuit of
those objectives (For a more detailed analysis of objectives and activities, see
appendix 1 regarding the activities of the Federal Reserve).

Table 2: Similar objectives, but different functions

National Bank of Hungary Reserve Bank of India

• Monetary policy • Monetary policy

• Lender of last resort • Lender of last resort
• Issuance of currency • Issuance of currency
• Foreign exchange and • Foreign exchange and reserves management
reserves management • Payment and settlement activities
• Payment and settlement • Bank supervision
activities • Fiscal agent
• Stock exchange (Budapest • Non-bank supervision
Stock Exchange-minority • Deposit insurance (Deposit Insurance and
stake) Credit Guarantee Corp.)
• Credit bureau
• Mortgage and housing (National Housing Bank)
• Agricultural activity (National Bank for
Agricultural and Rural Development)
• Securities trading (Securities Trading
Corporation-minority stake)
• Primary dealer (Discount and Finance House of
India-minority stake)
• Infrastructure finance (Infrastructure
Development Finance Company-minority stake)

Note: the objective of the National Bank of Hungary is to achieve and maintain price stability and that
of the Reserve Bank of India is to regulate the issue of banknotes and the maintenance of reserves with
a view to securing monetary stability. If an ownership stake is not noted, the entity is wholly owned.

55
Central Bank Modernisation

Table 3 is an effort to place the various functions that central banks undertake on a
continuum ranging from those that are closely linked to the objective of price and
monetary stability, to those activities that are not so closely linked. Taking on a
broad and possibly conflicting set of duties carries the risk of wasted and unfocussed
effort.

Table 3: Relation between central bank activities and objective

Mortgage or housing activities

Ownership of banks
Supervision of non-banks
Resolutions and liquidations
Open bank assistance
Small value payment
participant Foreign exchange mgt.
Mgt. deposit insurance Supervision of banks.
Consumer protection Lender of last resort
Anti-money laundering (FIU) Large value payment activity Monetary policy mgt.
Development banking Fiscal agent
Credit bureaus Currency operations
Loans to corporations Supervision of payments
Export credit
State banks or state loans
Small, medium enterprise loans
Agricultural activity

<<Not related to objective Directly related to objective >>

Quantifying legislative and managerial inefficiencies

To measure and compare central bank operations some type of cross-country
comparison, and thus cross-country index, is needed. The rationale for developing
such an index for central bank operational efficiency is that it provides a useful
standard against which to judge central bank activities, just as indices or models for
central bank independence and accountability have been useful with regard to
central bank effectiveness. It should incorporate elements of previous research in the
areas of commercial bank efficiency, as well as the fledgling research in the area of
central bank operational efficiency. The process should allow central banks to
compare their activities to peer central banks that undertake similar functions and
responsibilities. If a central bank undertakes a refocusing of its activities that truly
moves it toward more efficient operations, such as central banks in the US, Sweden

56
McKinley and Banaian

and Sri Lanka have attempted, the index should capture efficiency improvements
through time-series analysis.

Econometric indexing of efficiency

Determining the efficiency of any organisation involves either understanding what
level of outputs could be produced with a fixed set of resources, or what is the
cheapest set of inputs one could use to produce a fixed amount of output.
Econometrically this problem has been analysed through two approaches: data
envelopment analysis (DEA) or stochastic frontier regression (SFR).51 However,
because these approaches often give conflicting results, a choice between them must
be made.

Both approaches generate a measure of the distance between the combination of

inputs a firm or other entity is using and the estimated “best” combination of inputs
that could produce a given level of output. In DEA the entire difference is defined
to be inefficiency. But this need not be so. SFR allows for some of that distance to
be explained by measurement errors, random chance, or inefficiency. The researcher
may be able to describe the nature of the errors as well, but DEA does not give this
opportunity as SFR does. For example, the use of proxies for the quality of the
outputs introduces a source of error that is not properly included in a measure of
inefficiency. SFR gives the opportunity to control for this, while DEA does not.

However, the SFR estimation approach is not without its risks. Because SFR is
parametric, it carries the risk of specification bias, which DEA does not.
Specification bias arises when a researcher chooses incorrect independent variables
for a model or fails to specify the correct functional form for the model. Clearly,
under an SFR approach, it is important to get the functional form correct, so that
misspecification does not introduce an additional element to the errors. Previous
literature on bank efficiency/costs has used a translog production or cost function for
bank operations.

Specifying a central bank index

Given the difficulty of measuring inputs and outputs, the assumptions needed to
employ DEA are probably not met with the data we propose to use for central banks,
so SFR will be used.52

So, to begin to specify an index for central bank efficiency, we need to analyse the
inputs central banks use in terms of labour and capital. For our model we have
included a parameter for the cost of a member of staff (calculated by dividing direct
personnel expenses by the number of employees) and for the level of physical
capital (in the form of fixed assets reported on the central bank balance sheet).

57
Central Bank Modernisation

Considering outputs, we have made the simplifying assumption that central banks
produce two types of output: monetary policy and bank supervision. The model
requires that outputs generated by the central bank can be quantified and, to some
extent, that their quality can be measured. As a proxy for the quantity of bank
supervision produced we use the number of on-site exams (normalised by
population). As a proxy for the quality of the output of the banking supervision
function we have used two indicators. The first is the banking and finance index
produced by the Heritage Foundation/Wall Street Journal which measures the
openness of the banking system and how conducive it is to banking and finance.
The second is the share of non-performing loans as a proportion of total banking
assets. As a proxy for the output of monetary policy we have used the monetary
policy index produced by the Heritage Foundation/Wall Street Journal. Details of
all the source data for the model are provided in appendix 2.

We use data on 32 central banks for the year 2001. The countries were selected
primarily based on data availability. Several are central banks formed in the
transition economies of eastern and central Europe and the former Soviet Union.
Several others are central banks of smaller EU countries. We included the United
States under the assumption that we could aggregate the expenses of all twelve
reserve banks in the system. Only a few of the 32 are from developing countries with
a longer history as a market economy. With only 32 central banks in our sample for
which we have sufficient data for analysis, and with several environmental factors
to consider, the number of parameters to be estimated needed to be economised. For
this reason, we chose to use a linear specification of the cost function.53 The results
of these estimates are shown in table 4.

Analysis of the results of the index

The results are as we might expect. The first column in table 4 shows the
inefficiency estimates that come from our original SFR model. These estimates give
a percentage score for the distance each central bank is from the hypothetical “most
efficient” central bank which would have a score of 0%. The percentages express
the proportion of extra resources each central bank is expending to produce its
particular output over those the “most efficient” central bank would require (were
that central bank to face the same wage costs and population size). Inefficiency thus
increases as scores rise above zero.

How reliable are these estimates? The troubling aspect of the first inefficiency
scores (those with quality measures) is the high level of inefficiency assigned to
countries in Western Europe and the United States. We expected to find one or two
industrialised countries that are inefficient, but not the majority of them. In contrast,
Russia appears in that measure to be on the efficiency frontier. One reason for this

58
McKinley and Banaian

seems to be the inclusion of the quality measures. These are found to be jointly
insignificant in the estimation of cost from the OLS regression,54 implying that there
is no difference in the cost of producing higher- or lower-quality outputs. The
second inefficiency measure, which strips out the quality measures in the model,
generates more plausible results, though we note that Austria and Spain seem more
inefficient than we would expect. The estimation results and significance scores
appear in appendix 3.

Table 4: Inefficiency estimates for a sample of central banks

Country Inefficiency (%)

With quality Without quality
measures measures

Albania 141.3 151.2

Armenia 89.6 88.8
Austria 179.5 212.9
Azerbaijan 117.7 143.2
Belgium 0.1 10.1
Bosnia 179.1 157.0
Bulgaria 153.2 217.5
Canada 129.5 115.3
Croatia 75.2 61.9
Czech Rep. 0.1 0.8
Denmark 68.0 41.4
Estonia 99.4 101.3
Finland 55.4 78.9
Georgia 162.0 118.3
Greece 102.5 154.4
Hungary 132.9 145.3
Iceland 0.2 0.8
India 220.7 252.4
Italy 24.8 69.2
Jamaica 180.3 196.7
Kyrgyz Rep. 132.4 98.9
Latvia 0.8 25.1
Lithuania 0.1 18.5
Malta 0.1 0.8
Oman 127.5 126.6
Poland 128.8 162.8
Portugal 111.3 130.6
Romania 247.2 276.7
Russia 0.1 10.4
Slovenia 121.7 99.3
Spain 152.2 163.8
United States 82.4 4.3

59
Central Bank Modernisation

Comparing the two different formulations of the model, there is a high degree of
consistency in the results in terms of rankings, though the size of the inefficiencies
is reduced in many cases (notably regarding the United States) when quality
measures are ignored.55 While the fit of the more parsimonious second regression is
not quite as good, it has the advantage of avoiding the bias inherent in using the
judgmental factors of the Heritage Index scales. The lesson to be learned here is that
the measures are sensitive to specification, but relative rankings of central banks
may be more robust to choice of specification.56 Graph 1 illustrates this ranking.

Graph 1: Ranking of central bank inefficiency estimates

300%

250%
X-inefficiency (% of budget)

200%

150%

100%

50%

0%
p. d lta es m ia ia ia rk tia ly d ia p. ia ia a ia an al an r y ia ce ia d in ca ria ria ia ia
Re elanMa tat lgiu uss uan atv ma roa Itainlanmenz ReovenstonanadeorgOm r tugbaij ngalbanree osnolanSpamai ust lga Indman
ech Ic ed S Be R Lith LDen C F Ar gy Sl E C G
r Po zer Hu A G B P Ja A Bu Ro
Cz Un
it Ky A

Caveats and directions for future research

Several caveats apply to these results. First, our model assumes that all of these
countries’ central banks share a common efficient frontier, or “most efficient”
central bank. Some similar studies have argued that this may not be the case under
certain circumstances. For instance Bos and Schmiedel57 point out in their study of
European commercial banks, that failure to account for metafrontiers – which in our
case would allow for central banks in different countries to draw on different
technologies – leads to overstatement of the size of the inefficiencies. With our
limited sample, implementation of their alternative technique in our case was
impossible.

60
McKinley and Banaian

Second, in constructing input costs for capital we assumed that the cost of capital
for central banks is the same, so that all central banks face a common real rental
price of capital. Clearly this is not necessarily true, but it seems to be a justifiable
assumption.

Third, as noted above, we have made use of a simpler, log-linear cost function rather
than a translog cost function.58 Future research could attempt to take into account
these caveats, and could draw on more recent central bank financial statement data,
which is improving as time advances. Additionally, as financial statement data
quality increases, a larger sampling of central banks could be subject to analysis.

Methodology for improving operational efficiency

Any effort to improve central bank operations through increases in efficiency should
involve a complete re-examination of what the central bank’s objective is, what
functions it undertakes, why they are done and how they should be done going
forward. It should further incorporate the approaches of previous efforts to improve
efficiency, such as those demonstrated by central banks in the US, Sweden, Sri
Lanka and others. A proposed methodology for improving central bank operational
efficiency follows:
• Undertake an analysis of the current objectives of the central bank and refocus it
toward the objectives most appropriate for a central bank: price and monetary
stability. A strategy document should be developed that details how the central
bank will improve operations and meet objectives over a finite period, such as
three years.
• Review current functions of the central bank based on the refocused objective by
breaking down current operations into discrete processes and placing them under
the analysis of an appointed team. Avoid team structures that reinforce the status
quo.
• The core functions detailed in this chapter were largely based on what central
banks came to acquire in the form of roles and powers as central banks have
evolved, not necessarily on what was logical from a standpoint of meeting
objectives. The central bank should redefine its core operations by narrowing
their focus and tailoring functions to its revised objectives. Initial steps to
improve efficiency would involve elimination of non-core functions through sale
to the private sector or transfer to other parts of government.
• The next steps are more challenging than addressing the “low hanging fruit” of
eliminating non-core functions, but as detailed in the examples of efforts to
improve efficiency, they can be addressed. Review remaining core functions and
consider improving or outsourcing by analysing the function’s contribution to
meeting the central bank’s objective and assessing whether it could be done more

61
Central Bank Modernisation

efficiently outside the central bank, while maintaining the current level of
effectiveness.
• With regard to effectiveness, the proposed review of core functions should be
undertaken with an awareness that there are both good and bad efficiency
changes. For example, a central bank which gives up supervision of commercial
banks only to see this function undertaken less effectively by a politically-driven
agency will not be contributing to financial system stability.
• Determine methods to introduce improved information technology and
outsourcing within remaining core functions. Other support functions, such as
human resources should also be examined.
• Use the results of an efficiency model, as outlined in this chapter or as developed
individually for a central bank, on an ongoing basis, or find peer-group central
banks to compare and benchmark operations against. Communicate with other
central banks to determine potential areas of increased efficiency and develop
performance benchmarks.

Conclusion
The task of maintaining price and monetary stability, the most common objective of
central banks, is an immense task. The mindset needs to be broken that all the
functions that central banks currently perform in meeting this objective are
necessarily those they should continue to perform going forward. Achieving
efficiency of operations dictates that central banks perform only those underlying
functions necessary to achieve this objective. Those central banks that have
undertaken a detailed review of their operations in an effort to improve the
efficiency of their operations, many of which are detailed in this chapter, should be
lauded for taking on such a challenging task. A number of issues raised in this
chapter, such as improving the modelling of central bank efficiency, quantifying
central bank efforts at improving efficiency and researching how functions outside
of monetary policy management contribute to price and monetary stability are
obvious topics for future study.

62
McKinley and Banaian

Endnotes

1 A number of central banks themselves do set forth efficiency as a goal of their operations. For example, the
Central Bank of Nigeria’s website notes that its vision is “[t]o be one of the most efficient and effective world’s
central banks in promoting and sustaining economic development.” www.cenbank.org. The State Bank of
Pakistan in its Annual Report notes that its vision is to “transform SBP into a modern and dynamic central bank,
highly professional and efficient, fully equipped to play a meaningful role, on a sustainable basis, in the
economic and social development of Pakistan.” State Bank of Pakistan, Annual Performance Review, 2003 –
2004, http://www.sbp.org.pk/reports/annual/arFY04/vol2/Performance_Review.pdf. Other central banks have
efficiency as a requirement of their operations as detailed in their enabling statute. Bank of Japan Law, Article
5 (“In light of the public nature of its business and property, the Bank of Japan shall endeavour to conduct its
business in a proper and efficient manner”). Act on the Bank of Finland, Section 14 (“The activities of the Bank
shall be organised in an efficient and cost-effective manner”).
2 Obviously, there may be cases where housing a function in an independent central bank is preferred over having
it performed by a politically motivated entity elsewhere within the government.
3 It should be pointed out that this is usually not a material amount of government revenues in total.
4 Nor is it easy to define precisely what a central bank is. For purposes of this article, and consistent with the
developing common usage of the terminology and the divergent forms that it takes, “central bank” will be
applied very broadly to denote one or more of a country’s or independent territory’s or a group of countries’
monopoly-granted, monetary authorities.
5 Allen, Richard and Tomassi, D. editors, “Managing Public Expenditures-A Reference Book for Transition
Countries,” Organization for Economic Cooperation and Development (2001), p. 450.
6 Ibid.
7 One of the lead researchers on the topic of commercial bank efficiency is Alan Berger of the Federal Reserve
Board. Berger, A.N., Humphrey, D. B., 1997. Efficiency of Financial Institutions: International Survey and
Directions for Future Research. European Journal of Operational Research 98(2). Berger, A. N., Mester, L. J.,
1997. Inside the Black Box: What Explains Differences in the Efficiencies of Financial Institutions? Journal of
Banking and Finance 21. [Hereinafter, Berger and Mester (1997)]
8 There is one exception to this general statement as it is applied to Federal Reserve Bank researchers. Mester,
Loretta J., “Applying Efficiency Measurement Techniques to Central Banks” Prepared for the Workshop on
Central Bank Efficiency, Sveriges Riksbank, Stockholm, Sweden (May 20, 2003) [hereinafter, Mester].
9 For a summary in the context of a central bank, see Cavalluzzo, Ken, “Competition, Fee-for-Service
Requirements, and Government Performance: Evidence on the Federal Reserve,” June 1999, p. 4 [hereinafter,
Cavalluzzo].
10 In general see the body of work known as public choice. For example, Gordon, Tullock, et al Government
Failure—A Primer in Public Choice, (Cato Institute, 2002), in particular Chapter 5 on Bureaucracy.
11 Except for the cases of countries like Serbia and Montenegro (which has three central banks) and Cyprus
(which has two central banks), but it would not be said in all cases that these various central banks are directly
competitive with one another. Central banks in some cases may compete with other government or private
sector service providers in individual functions.
12 In reviewing central bank objectives, references are made to a number of different types of stability. Price,
monetary, financial, economic and currency stability are the most prominent. As these different types of
stability are all intertwined, this chapter will use price and monetary stability as a means to collectively refer
to all of the different types of stability.
13 Chapter IV of the Statute of the European System of Central Banks and of the European Central Bank is
dedicated to Monetary Policy (Article 17-Accounts with the ECB and the national central banks; Article 18-
Open market and credit operations; Article 19-Minimum reserves; Article 20-Other instruments of monetary
control; Article 20-Other instruments of monetary control; Article 21-Operations with public entities; Article
22-Clearing and payment system; Article 23-External operations; Article 24-Other operations).
14 “The Bank is responsible for the stability of the value of the currency of Aruba and shall determine the
monetary policy directed towards maintaining said stability.” Article 10, National Ordinance for Establishing a
Statute for the Centrale Bank van Aruba.
15 Articles 19 and 32, On the Bank of Albania.
16 “Reserve Bank to act as a central bank. The Reserve Bank (a) is the central bank of Australia; (b) shall carry

63
Central Bank Modernisation

on business as a central bank; (c) subject to this Act and to the Banking Act 1959 shall not carry on business
otherwise than as a central bank.” Section 26, Reserve Bank Act 1959 (Australia).
17 Law of the National Bank of the Republic of Kazakhstan, Article 8(m).
18 National Bank of Kazakhstan, Annual Report 2001, Almaty 2002, Note 1 to the Financial Statements, p 58. “As
of December 31, 2001, the National Bank made investment in the following companies, which allowed it
carrying out control and affect these companies considerably:…Guarantee of individuals’ deposits 100%;
Development of mortgage crediting and secondary market for mortgage securities to ensure fast return rate of
banks’ credits…”
19 The majority of respondents (59%) employ IAS in full or in some version thereof. This changeover has been
relatively recent, as a majority of those adopting international standards have done so only since 2000. For a
detailed survey of central bank accounting practices and the movement towards use of international standards
see Kurtzig, Joshua and Mander, Benedict, “Survey of Central Bank Accounting Practices,” in Courtis, Neil
and Mander, Benedict (eds), Accounting Standards for Central Banks, pp. 21 – 46 (London: Central Banking
Publications, 2003).
20 Pollard, Patricia S., “A Look Inside Two Central Banks: The European Central Bank and the Federal Reserve,”
Federal Reserve Bank of St. Louis, January/February 2003, p. 18. The precise description of these activities is
“Tasks of the Federal Reserve System and European System of Central Banks.” The author describes the article
as an examination of modern central banking, but chooses to limit its focus to the world’s most prominent
central banks—the Federal Reserve System and European Central Bank.
21 White, Lawrence H., “In What Respects Will the Information Age Make Central Banks Obsolete,” Cato
Journal, Vol. 21, No. 2, Fall 2001, p 219. The precise description of these activities is “Central banks today play
five major roles…”
22 Fischer, Stanley, “Modern Approaches to Central Banking,” Working Paper 5064, National Bureau of
Economic Research, 1995. The precise description of these activities is “Central banks around the world
perform a variety of functions…”
23 Green, Edward J., “What Tasks Should Central Banks Be Asked to Perform?” Prepared for the Workshop on
Central Bank Efficiency, Sveriges Riksbank, Stockholm, Sweden, May 2003.
24 Although there is no available summary of the Cleveland Fed effort in the public domain, the information
detailed in this section was drawn from a conversation with Larry Cuy, Senior Vice President of the Cleveland
Fed.
25 The processes included cash, check clearing, savings bonds, economic research, bank supervision, treasury
services, information technology and human resources.
26 The authors reviewed approximately 120 central bank laws in compiling this summary of core functions and
the summary of other functions in the section that follows. Other related laws were also reviewed in an effort
to determine if the central bank has undertaken other functions. Efforts were undertaken to use current
legislation. Translations of these laws were received from reliable sources (including oftentimes, the central
bank’s website) and are presumed accurate. Additional materials used in compiling this information included
the central bank’s financial statements, general information on the central bank’s website and other available
sources of information such as available summaries of specialty functions, such as deposit insurance.
27 Once again no summary of this process is in the public domain. The information in this section was drawn
together through communications with Pether Burwall, Deputy Head of the Administration Department for
Sveriges Riksbank. Also, see Heikensten, Lars, “How to Promote and Measure Central Bank Efficiency,”
Workshop on Central Bank Efficiency—Stockholm, Sweden, 23-24 May 2003 [hereinafter, Heikensten
speech]; and for a summary of the overall methodology applied to the operations, see Blix, Marten; Daltung,
Sonja; Heikensten, Lars, “On Central Bank Efficiency,” Economic Review, Sveriges Riksbank, 3:2003, p. 81.
28 Financial stability here is defined as promoting a safe and efficient payment system.
29 Excluding wholly owned subsidiaries the change was from 755 in 1995 to 430 today.
30 See Heikensten speech, p. 4.
31 Central Bank and Financial Services Authority of Ireland Act 2003 renaming the central bank as the Central
Bank and Financial Services Authority of Ireland and creating the Irish Financial Services Regulatory Authority
an autonomous body within the central bank.

64
McKinley and Banaian

32 National Bank of Kazakhstan. Law of the Republic of Armenia on Bankruptcy of Banks, Article 1. Organic
Law of Georgia on the National Bank of Georgia, Article 59(5).
33 The Executive Service of the Commission for the Prevention of Money Laundering and Financial Crime is
housed in the Bank of Spain, http://www1.oecd.org/fatf/Ctry-orgpages/ctry-es_en.htm. Islamic Republic of
Afghanistan, Anti-Money Laundering and Proceeds of Crime Law, Article 19 establishes a financial
intelligence unit under the authority of Da Afghanistan Bank.
34 Palestine Monetary Authority Law, Article 53.
35 The National Bank of Kazakhstan holds 100 percent of the stock of a company responsible for developing a
secondary market for mortgage securities. National Bank of Kazakhstan, Annual Report 2001, Note 1 to the
Financial Statements, Almaty (2002).
36 Also known as “Too Big to Fail” Assistance. Law of the National Bank of Albania. Article 16(2) granting the
authority to extend credit to banks in conservatorship if needed to protect the integrity of the banking system.
37 Central Bank of Nigeria Decree, Article 27. Reserve Bank of India Act, Article 17.
38 Central Bank of Nigeria, www.cenbank.org/aboutus/coremandate.htm which describes the small and medium
enterprises loan scheme.
39 Central Bank of Malaysia Act, Article 30. Bank of Namibia Act, Article 29.
40 Bank of Thailand Act, Section 29 bis.
41 National Ordinance of the Centrale Bank van Aruba, Article 13.
42 State Bank of Pakistan Act, Article 17.
43 The Central Bank of the Russian Federation holds a controlling interest in the Savings Bank of the Russian
Federation. Article 8, Russian Federation Law on the Central Bank of the Russian Federation. The bank
dominates the Russian banking sector as it holds three quarters of the banking system’s individual deposits and
a quarter of the gross loans and assets. World Bank, Building Trust—Developing the Russian Financial Sector,
September 2002, page 167.
44 The Federal Reserve System, Purposes and Functions, 1994, p. 90 –92.
45 A number of such reengineering projects have been undertaken in South Asia in combination with the World
Bank, including in Bangladesh, Pakistan and Nepal. As many largely involved the same approach the Sri Lanka
reengineering project, which is the most advanced of the group, is detailed here.
46 The bulk of the funding was provided by the International Development Association (IDA), the World Bank’s
concessionary lending affiliate, which committed to providing $30.3 million equivalent interest-free credit on
standard IDA terms, with 40 years maturity and a ten-year grace period. The government of Sri Lanka provided
an additional $10.7 million and the Swedish International Development Agency provided $1 million.
47 World Bank, “Sri Lanka Modernizes Central Bank With Support From the World Bank,” June 20, 2001.
48 The following sections are detailed in Central Bank of Sri Lanka Annual Report—2002, Part II, “Modernisation
of the Central Bank of Sri Lanka,” p. I – IV and Central Bank of Sri Lanka Annual Report—2003, Part II,
“Modernisation of the Central Bank of Sri Lanka, p. I - III.”
49 The former objective of the CBSL was wide-ranging and lacked focus and clarity. It included the stabilization
of domestic monetary values, the preservation of the par value or stability of the Sri Lanka rupee and the free
use of the rupee for current international transactions; promotion and maintenance of a high level of production,
employment and real income in Sri Lanka; encouragement and promotion of the full development of the
productive resources of Sri Lanka.
50 World Bank, “Project Appraisal Document on a Proposed Credit to the Democratic Socialist Government of Sri
Lanka for the Sri Lanka Central Bank Strengthening,” Report No: 22178, May 23, 2001, p. 18.
51 Bauer, Paul W., “Recent Developments in the Econometric Estimation of Frontiers.” Journal of Econometrics
46(1-2), October-November 1990, 39-56 provides a useful explanation of the techniques.
52 The stochastic cost frontier has the following general form:

Ci = ln f (yi, wi, qi, zi, hi)+ui+vi i=1,...,N,

where C denotes variable costs, y is a vector of outputs, w is a vector of prices of variable inputs, q is a vector
of variables characterising the quality of output, z indicates fixed netputs (inputs or outputs), h is a set of

65
Central Bank Modernisation

environmental variables, vi is a two-sided error term representing statistical noise, ui is a one-sided error term
representing inefficiency factor.

The random error vi is assumed to be normally distributed with parameters N(0, σv2) and the inefficiency term
ui is assumed to be half normally distributed N(0, σu2). A central bank’s inefficiency estimate is derived as
follows:

XEi=E(vi/(ui+vi)) = σ*λ/(1+λ2)* [ϕ((ui+vi)* λ/ σ)/φ((ui+vi)* λ/σ) - (ui+vi)* λ/σ)],

where XEi is X-efficiency of bank i, E ( ) is the expectation operator, λ is the ratio of the standard deviation of
v to the standard deviation of u, σ = σv2+ σu2 and ϕ and φ are the standard and cumulative normal density
functions, respectively.

The inputs include labour c and physical capital z, which is included as levels in our specification. The price of
labour w is proxied by the ratio of personnel expense to the number of employees.
53 We would have required 16 parameters estimated; there would have been three constraints that would have
raised the degrees of freedom to 19, which still seems insufficient. We explored the possibility that the cost
function was Cobb-Douglas, but the constraints are rejected at a 99% confidence interval, so we chose instead
to use the unconstrained version. This may be due to different levels of technology available to different central
banks.
54 With probability of 0.722.
55 While some rankings of inefficiency changed, the correlation between the two measures is 0.923.
56 Another potential use of the measures is to determine whether inefficiency can be controlled by use of
regulations. For example, we regress the second set of inefficiency scores on the banking and finance index,
the ratio of non-performing loans to total assets, and the number of branches operated by the central bank. All
of these are insignificant.

Inefficiency score = -4.543 + 0.081*Banking and finance index – 0.019*(NPL/Assets) + 0.0006*branches

(0.46) (1.37) (0.09)
R = 0.06, t-stats in parentheses
2

There is insufficient data at this point to measure whether or not inefficiency is increased or reduced by
increasing amounts of central bank independence or accountability, or responsibility of the central bank for
bank supervision.
57 Bos, J.W. and Schmeidel, H., “Comparing Efficiency in European Banking: A Meta Frontier Approach,” De
Nederlandsche Bank, Research Series Supervision no. 57 (May 2003).
58 This is as posited by Mester. We looked at a translog version of the second regression and found that, while one
cannot reject the null hypothesis of a translog specification, the results are not changed significantly while
removing four degrees of freedom from an already small sample. Other studies use a Fourier-flexible transform,
but Berger and Mester (1997) show that this makes little difference from the translog.

66
McKinley and Banaian

Appendix 1

A case study of the Federal Reserve System

In order to more deeply assess the operations of a central bank and determine how
its operations might be inefficient, a case study would be useful. This analysis of the
Board of Governors of the Federal Reserve System (hereinafter, Federal Reserve)
will begin with its legislative objective and then address how its functions in pursuit
of that objective might be inefficient.

Pursuit of its objective

The Federal Reserve and its monetary policy arm, the Federal Open Market
Committee, have an objective as codified in the Federal Reserve Act:

“The Board of Governors of the Federal Reserve System and the Federal Open
Market Committee shall maintain long-run growth of the monetary and credit
aggregates commensurate with the economy’s long-run potential to increase
production, so as to promote effectively the goals of maximum employment,
stable prices, and moderate long-term interest rates.” 1

Reviewing history over the past twenty years, the US economy has experienced two
relatively mild recessions,2 with inflation approximately in the range of two to five
per cent.3 Although it would not be unanimous, most observers would agree that the
Federal Reserve has been effective in meeting this objective, at least over the past
two decades.4

Efficiency in pursuit of its objective

There is no specific requirement in the Federal Reserve Act that the Federal Reserve
has to reach this objective in an efficient manner. However, there are two legislative
acts that the Federal Reserve is either subject to through mandate or through
voluntarily compliance that address efficiency in its operations: the Government
Performance Results Act of 1993 and the Inspector General Act of 1978.

The Government Performance Results Act of 1993

The Government Performance Results Act of 1993 (GPRA) imposes on any
government agency,5 as defined in the act, a requirement to develop and submit to
the Office of Management and Budget a strategic plan to be updated every three
years. This requirement is imposed for the purpose of improving congressional

67
Central Bank Modernisation

decision making by providing more objective information on achieving statutory

objectives, and on the relative effectiveness and efficiency of federal programs and
spending. The legislation was based on the finding that:
“1 waste and inefficiency in Federal programs undermine the confidence of the
American people in the government and reduces the federal government’s
ability to address adequately vital public needs;
2 federal managers are seriously disadvantaged in their efforts to improve
program efficiency and effectiveness, because of insufficient articulation of
program goals and inadequate information on program performance; and
3 congressional policymaking, spending decisions and program oversight are
seriously handicapped by insufficient attention to program performance and
results.”

Because the Federal Reserve is not an agency as defined in the act, it is not required
to file these reports. However, it voluntarily subjects itself to the act by filing the
reports required under the act and participating in the process developed under the
act.6 The most recent of these reports, a biennial performance plan for 2004-05 ,7
notes: “the plan is centred on the monetary policy function, the operations of the
Board in overseeing the activities of the system, and management actions to
improve effectiveness and efficiency”; “to maintain budgetary independence, the
Board believes that it must demonstrate effective and efficient use of its financial
resources”; and “the Board’s planning is oriented toward achieving effectiveness
and efficiency specific to the functions it serves.” 8

The Inspector General Act of 1978

The Inspector General Act of 1978 applies to the Board of Governors of the Federal
Reserve. The act provides for the establishment of an office of inspector general
whose purpose is, in part, to “conduct and supervise audits and investigations
relating to the programs and operations” and “provide leadership and coordination
and recommended policies for activities designed to promote economy, efficiency
and effectiveness in the administration of, and to prevent and detect fraud and abuse
in, such programs and operations.”9 The inspector general for the Federal Reserve
undertakes audits, reviews, investigations and special projects on the activities of the
central bank and files a semi-annual report with the Congress on its activities.10 Most
of the audits undertaken are at the request of a board governor or pursuant to law,
but there is no overall general strategy to move toward efficiency in operations.

Despite this lack of an overall strategy to move toward efficiency in operations, the
board’s inspector general (BOIG) has reviewed efforts to comply with the GPRA
reporting requirements and determined that the board has not achieved its objective
of voluntarily complying with the Results Act.11 In its summary report, the BOIG
report noted that this compliance has not been accomplished, primarily because of a
lack of support from senior management and a perception that GPRA activities were

68
McKinley and Banaian

a paper-intensive way to produce reports.12 Four primary recommendations were

made for the board to achieve the committed compliance:
• establish a framework to adopt a more results-oriented performance management
approach;
• revise its strategic planning process to develop a long-range, board-wide
strategic plan;
• establish specific, quantifiable, results-oriented performance measures aligned
with the strategic plan; and
• revise its internal performance reports to include comparisons of actual program
results with established performance goals and measures.

In its response to the BOIG report, the staff director for the board noted that the
process for establishing specific, quantifiable results-oriented performance
measures would be difficult and that previous efforts at developing performance
measures met with limited success.13

A review of Federal Reserve operations

The Federal Reserve can be used as a case study of how a central bank provides its
services in an inefficient manner. These inefficiencies may not in all instances be
attributed to managerial inefficiency. As discussed previously for central banks
generally, there may be legislative inefficiencies that mandate that the Federal
Reserve takes on certain roles unrelated to its objective.

Most of the attention on the activities of the Federal Reserve is focussed on

monetary policy, probably because of the direct relationship between monetary
policy and the Federal Reserve’s primary objective. However, in terms of
employees, of the more than 20,000 employees of the Federal Reserve, less than
10% work on monetary and economic policy issues and these activities represent
about 10% of total expenses.14 Most undertake other activities that are not directly
related to the objective of the Federal Reserve. These activities often bring them into
direct competition with private sector competitors that do not have the competitive
advantages of the Federal Reserve.15

Overall structure of the Federal Reserve

“The Federal Reserve System has a structure designed by Congress to give it a
broad perspective on the economy and on economic activity in all parts of the
nation. It is a federal system, composed basically of a central, government
agency – the Board of Governors – in Washington, DC and twelve regional
Federal Reserve Banks, located in major cities throughout the nation. These
components share responsibility for supervising and regulating certain
financial institutions and activities; for providing services to depository

69
Central Bank Modernisation

institutions and to the federal government; and for ensuring that consumers
receive adequate information and fair treatment in their business with the
banking system.” 16

This shared responsibility leads to overlap and duplication. Additionally, with the
multiple separate operating boards at the reserve banks it makes a broad-based
review and assessment of operations difficult. The structure of the Federal Reserve
has changed little since its creation 100 years ago. It is comprised of 12 reserve
banks and 25 branches distributed throughout the US. According to congressional
debates at the time, the location of these banks was decided, at least in part, so that
“no bank be more than an overnight’s train ride from its reserve bank.”17 There has
not been a major reassessment of the reserve bank structure since these days.
Commenting on the structure of the system, a General Accounting Office (GAO)18
report notes “except for minor boundary changes, the geographical structure of the
Federal Reserve has remained unchanged since 1914, while the nation’s population
has shifted dramatically (see table 1).”19 As also noted in the GAO report, the
Federal Reserve has the authority to close branches, but has only done so once and
that was 65 years ago.20

Table 1: Population breakdown of Federal Reserve districts over time

Federal Reserve District

Federal Reserve 1910 population % of US 1990 population % of US
District (millions) population (millions) population

Boston 6,307 7 12,378 5

New York 11,329 12 20,514 8
Philadelphia 5,953 6 11,506 5
Cleveland 8,375 9 16,108 7
Richmond 8,479 9 23,305 10
Atlanta 9,094 10 31,833 13
Chicago 13,114 14 30,601 12
St. Louis 8,273 9 12,528 5
Minneapolis 4,423 5 7,574 3
Kansas City 6,899 7 13,541 6
Dallas 4,539 5 18,467 8
San Francisco 5,434 6 46,500 19
Total 92,228 244,855

Source: GAO report, “Federal Reserve System: Current and Future Challenges Require System-wide
Attention,” June 1996, p 84.

Beyond the issue of population rationality, given technological advances there is a

question of why 12 Federal Reserve banks and 25 branches are necessary. The major
justification for branches was that they had to be close to those banks that the

70
McKinley and Banaian

services were provided to. But many of the services are now provided without the
need for direct personalised service.

The following is the response by the Federal Reserve to the suggestion that
efficiencies could be achieved through an overall review of the branch system
followed by the closure of certain banks or branches:

“While the location and territory of some of the 12 Federal Reserve Bank head
offices and their 25 branches would likely be different if they were established
today, relocation of facilities or substantial realignment of office territories
generally is very costly, and the long-term savings would have to be substantial
to offset the transition costs.” 21

This response reveals a lack of willingness to undertake an analysis of the system to

improve efficiency, and merely recognises the obvious fact that there would have to
be substantial savings to undertake such changes.

Functions of the Federal Reserve

The following sections detail a number of the major functions of the Federal
Reserve, and discuss some of the potential inefficiencies in each of these activities.

Supervision and regulation

“In addition to its money and credit responsibilities, the Federal Reserve has
broad supervisory and regulatory authority over the activities of state-
chartered member banks and bank holding companies, including their foreign
activities and Edge corporations, and foreign banks operating in the United
States. It is also charged with writing regulations for the major federal
consumer credit laws.” 22

Of all of the activities that it engages in, one that clearly does not have a direct link
to the objectives of the Federal Reserve of maximum employment, stable prices, and
moderate long-term interest rates is that of enforcement of consumer laws. The
central bank is responsible for a full range of consumer credit laws, including
writing, interpreting and verifying compliance with regulations, investigating
complaints from the public and enforcing the laws.23 Enforcement of these laws
would more logically fall under the responsibility of other agencies in the
government that are not burdened with responsibility for price and monetary
stability.24

With regard to more general financial supervision and regulation of commercial

banks, dozens of central banks, especially throughout Europe, have jettisoned their
responsibilities for supervision of banks, with a unified supervisory authority taking

71
Central Bank Modernisation

over the responsibility for supervising banks on a consolidated basis. The Federal
Reserve has generally taken the position that the power of supervision should
remain with the central bank. Consistent with that stance, the research department
of one of the Federal Reserve banks argues that the power to supervise banks can
actually enhance the conduct of monetary policy by providing confidential bank
supervisory information and notes that it is surprising that many countries have
sought to reduce their central bank’s involvement in bank supervision.25 The study’s
authors argue against separation of monetary policy and supervision because “while
the timely sharing of information between other bank supervisors and the central
bank is certainly possible, the difficulties in sharing highly confidential information,
much of which may not be easily quantifiable, might make such arrangements
difficult at best.”26 But the analysis does not address why this same issue is not
equally a challenge for the separate departments within the Federal Reserve
(monetary policy and supervision) or between the Federal Reserve and the other
agencies that share supervisory authority for Federal Reserve member banks.

There are not only intra-agency inefficiencies, but also inter-agency inefficiencies
among the various supervisory authorities. The structure and differing
responsibilities of regulation between the Federal Reserve, the primary federal
regulator of Federal Reserve member banks, the Office of the Comptroller of the
Currency, the primary federal regulator of nationally chartered banks, the Federal
Deposit Insurance Corporation, the primary federal regulator of state non-member
banks and the individual states as chartering authorities lead to additional areas of
overlap and inefficiency. This structure is not directly attributable to the agencies,
although regulators have historically protected their “turf” from infringement.

Payment systems

“The Federal Reserve serves as a central cheque-clearing system, handling

approximately 18 billion cheques a year… The Federal Reserve banks and
about 7,800 depository institutions are linked electronically through the
Federal Reserve Communications System, a network through which depository
institutions can transfer funds and securities nationwide in a matter of
minutes… Federal Reserve banks and their branches operate automated
clearing houses, computerised facilities that allow for the electronic exchange
of payments among participating depository institutions.” 27

The Federal Reserve, like many central banks, plays a number of roles in relation to
payment systems – “policy maker, regulator, service provider and dominant services
provider” as one banking industry group summarised it.28 As service provider, the
Federal Reserve engages in a wide variety of payment services. Some of the markets
for provision of services are very competitive, with the Federal Reserve facing
competition from private sector providers, while others are not highly competitive.29

72
McKinley and Banaian

In response to recent rapid changes that have occurred with regard to payment
systems and technology, and possibly to respond to those who face the difficult task
of competing with the Federal Reserve in payment system markets, the Federal
Reserve appointed a Committee to examine their role in the payment system. The
Committee on the Federal Reserve in the Payments Mechanism (the payments
committee) examined the Federal Reserve’s role and issued a report in 1998
detailing a variety of recommendations.30

The payments committee looked at five potential scenarios for the Federal Reserve’s
involvement in payment operations:
• liquidation – announce its intention to withdraw from the provision of cheque
collection and Automated Clearing House (ACH) services as of some date with
a planned transition; a transition period would provide for a smooth transition to
commercial providers;
• privatisation – privatise cheque collection and ACH operations by placing them
under a newly chartered, special purpose entity that would eventually become a
commercial entity with no privileged ties to the Federal Reserve;
• continuity and access – continue to provide cheque collection and ACH
services, with the limited goal of universal access for depository institutions;
• promoting efficiency – use its operational presence and influence in cheque
collections and ACH markets to enhance the efficiency of the payments system;
or
• leading to electronic payments – expedite the movement to an electronic-based
retail payment system, replicating the universal acceptance and access that
characterises the current paper-based system.31

The payments committee concluded that the proper course of action was to remain
a provider of both cheque collection and ACH services, with the goal of enhancing
the efficiency, effectiveness and convenience of both systems while ensuring access
for all depository institutions. Additionally, it concluded that the Federal Reserve
should play a more active role in enhancing efficiency and in moving to the next
generation of payment instruments, working closely and collaboratively with
providers and users of the payments system.32 The payments committee came to that
conclusion after receiving input from Federal Reserve personnel and payment
system participants.

In assessing the alternative future roles for the Federal Reserve, the payments
committee set forth the following principles to guide its analysis of the payments
system: the integrity of the payments system – its safety and reliability; the
accessibility of the payments system – that it is available to all depository
institutions so that they can provide for the payments needs of their customers; and
the efficiency of the system – that the cost of making payments is reduced as much
as possible.33

73
Central Bank Modernisation

The noted definition of efficiency, which focusses entirely on the cost of payment
services to the end user, is worth particular scrutiny. It is further detailed in the
following passage from the report:

“The committee addressed the fundamental question of whether the Federal

Reserve should remain a provider of cheque collection services or should exit
the market in favour of private sector providers. It focussed on whether
removal of the Federal Reserve from the market could be expected to make the
provision of cheque processing services more or less efficient and on whether
exit of the Federal Reserve would hasten or retard the evolution of electronic
check processing. Since the Federal Reserve plays such an important role in
processing checks for small and remote institutions, the possible impact of
Federal Reserve withdrawal on the access and costs of those institutions is an
important consideration.

Some assert that private sector providers are inherently more efficient than
quasi-public ones and that the Federal Reserve’s withdrawal would tend to
enhance the efficiency of the system over time. But the Federal Reserve is now
competing with those private sector providers and covering costs comparable
to the ones they incur. Unless the Federal Reserve is benefiting from a hidden
financial subsidy (and the committee does not believe it is), it is not evident
that removing a major competitor from the market place would enhance
efficiency. In sparsely settled areas with few competitors in the cheque
collection market, removing the Federal Reserve might reduce efficiency.” 34

This analysis defines efficiency in the context of a goal of universal access to cheque
collection services, where prices must be the same for all users across the
geographic spectrum, in both densely and sparsely populated geographic areas,
without relation to cost of service to the provider. It reflects the view that somehow
these services are a public good that deserve some form of cross-subsidy. This is
further reflected in the payments committee’s summary of comments from
community banks and small depository institutions, in which it notes that they “are
concerned that the Federal Reserve’s withdrawal would require them to purchase
correspondent services from commercial banks that in many cases are direct
competitors for local deposit and lending business… More generally, community
banks are reluctant to be the source of profits for direct competitors… Thus, small
depository institutions feel strongly that the Federal Reserve should continue to be
an active provider of cheque collection services.” 35 The related point is made in the
payment committee’s criticism of the privatisation option for check collection and
ACH services: “many participants stated that privatisation would likely result in a
large entity motivated by profit rather than the public interest.” 36

74
McKinley and Banaian

The collective definition of efficiency and efficient markets presented in the above
passages diverges from what is generally accepted, as well as from the definition set
forth in this chapter. The public good argument is that the services provided are
worthy of cross-subsidy through the actions of the Federal Reserve, even if it means
that pricing is done inefficiently outside of market mechanisms. By making a value
judgment as to what the pricing should be, the Federal Reserve is actually pricing
services inefficiently, leading to overdemand of services in sparsely populated areas
and reduced demand for services in densely populated areas. Leaving the pricing to
market mechanisms would be more efficient than having the Federal Reserve cross-
subsidise in this manner.

The conclusions of the payments committee are not derived on an independent basis
as they are largely based on the opinions of those with a vested interest in
maintaining the current system: those who currently receive a cross-subsidy under
the current system and those employees of the Federal Reserve who would face
uncertainty under the liquidation or privatisation scenarios. The liquidation and
privatisation scenarios were not given serious consideration. Recent calls for a more
independent review of the Federal Reserve’s role in the payment system have been
ignored .37

Currency and coin

“The Federal Reserve banks distribute currency (paper money) and coin to
depository institutions to meet the public’s need for cash…Unfit currency and
coin are destroyed and replaced with new currency and coin obtained from the
Treasury department’s Bureau of Engraving and Printing and Bureau of the
Mint.” 38

In carrying out its analysis, the payments committee ignored the issue of currency
and coin processing as part of the payments system. It did this, because this is “a
service normally expected of a central bank”.39 Additionally, the board argues that
the Monetary Control Act, which dictates pricing of Federal Reserve services,
applies to currency and coin services such as transportation and coin wrapping, but
not to services “of a governmental nature, such as the disbursement and receipt of
new or fit coin and currency”.40 It does not base this on the actual legislative
language, which merely notes that the act applies to currency and coin services, but
bases it on obscure legislative history, a statement by then Senator Proxmire,
chairman of the Senate banking committee at the time.41

However, recent moves by the Federal Reserve have made progress towards
improving the efficiency of the activities by pricing some of these services. These
services of the Federal Reserve have historically been provided without charge.
Reserve Banks have acted as intermediaries between depository institutions and

75
Central Bank Modernisation

depository institutions have acted as intermediaries between customers, holding

currency to provide this service. Recently, in an effort to limit their holdings of
currency, which are a non-income-earning asset, financial institutions have adjusted
their role as intermediaries between customers. They have done this by increasing
the size and frequency of their deposits of currency to and orders of currency from
Reserve banks, in essence becoming an intermediary between the Reserve banks
and their customers. This can result in a financial institution depositing currency of
a particular denomination one day and within a few days ordering currency of the
same denomination, so called cross-shipping. This allows the institution to avoid
having to hold stocks of currency on hand.42

In an effort to minimise the cost of providing currency, the Federal Reserve recently
issued for comment notice of a proposed policy that would involve the development
of a custodial inventory program combined with a fee assessed on such cross-
shipped currency. The custodial inventory program would allow depository
institutions to transfer into custodial inventories currency that might otherwise be
cross-shipped. These would be inventories owned by the Federal Reserve bank but
located and segregated within a depository institution’s secured facility. The fee
program would involve a recirculation fee of between $5 and $6 per bundle on
cross-shipped currency to discourage the practice. This move by the Federal
Reserve is a step in the right direction toward efficiency in operations, but does not
match the extent to which some countries have privatised these functions. These
countries – including Australia, New Zealand, Brazil, Malaysia, Canada, Norway,
Sweden and the UK – are much further along in privatising and outsourcing these
functions.

The Treasury department’s Bureau of Engraving and Printing and Bureau of the
Mint retains a monopoly position in provision of currency and coin to the Federal
Reserve. Provision of these services is not open to competitive bidding; outside the
US, many countries, including those in developing markets, have a competitive
process. Opening up the process of providing these services to competitive bid
would allow for improved quality of service and expose the process to a more
market-oriented pricing of services.

Research

“Each Federal Reserve bank has a research staff to gather and analyse a wide
range of economic data and to interpret conditions and developments in the
economy.” 43

Each of the reserve banks has a research department that presumably provides some
insights into regional economic activities, but the Board of Governors also has a
large research department. Most reserve banks issue a journal with a small amount

76
McKinley and Banaian

of regional analysis, but most of the issues addressed are not limited to regional
issues. It is also difficult to see how many of the articles published in these journals
and other research outlets are even remotely related to the primary objectives or
functions of the central bank. For example, articles and papers on education issues,
which are published regularly by reserve banks, have a relatively strained link to
banks, price and monetary stability and would appear to be more suited for
researchers at the department of education or for academia generally.44

Efficiency of operations of the Federal Reserve

The Federal Reserve’s operations range from monopoly provider of a stable

currency and monetary environment to competitor for private sector companies in
providing services to individuals and banks. The standard for whether a central bank
engages in a range of activities should not be whether it is traditionally considered
of a “government nature” or “a service normally expected of a central bank” or that
undertaking the function is in the “public interest”. The test should be whether or not
it is closely related to its primary objective and thus whether or not its board and
senior management should focus its attention on managing that function. The
function of monetary policy is clearly related to the objectives of the Federal
Reserve. The function of heavy involvement as a service provider in the payments
system where private sector providers are fully capable of providing the same
services is clearly not so closely related. This issue was summarised well in an
article on central bank efficiency by a trio of senior staff of the Sveriges Riksbank,
the central bank in Sweden:

“In the area of payment systems operations and the pricing of various payment
instruments, efficiency is relatively easy to study. But it does not seem to be
clear why central banks should be directly involved in this area in the first
place. For example, why should central banks be operationally responsible for
the clearing and settlement of large-value payments, or why should they be
directly involved in the business of clearing cheques? Perhaps an efficient
payment system policy would call for the outsourcing of these activities.” 45

Those functions that fall between the two extremes should also be scrutinised. The
structure of the Federal Reserve, both from the standpoint of the mixture of
government and quasi-government entities as well as the number of banks and
branches, should also be thoroughly examined to improve the efficiency of the
system.46 Improved performance management and strategic planning, as envisioned
under the GPRA, should be looked upon as an opportunity to improve efficiency
rather than as a paper-intensive way to produce reports.

77
Central Bank Modernisation

Endnotes

1 12 USC 225a (Federal Reserve Act, Section 2A).

2 A recession is defined as “a significant decline in economic activity spread across the economy, lasting more
than a few months.” In the US during the past twenty years, one recession occurred from July 1990 to March
1991, while another occurred from March to November 2001. The National Bureau of Economic Research
(NBER) is an organisation independent of the US government that acts as the arbiter of when recessions
begin and end. www.nber.gov.
3 http://oregonstate.edu/dept/pol_sci/fac/sahr/sahr.htm contains historical data on inflation rates.
4 One of the more glowing appraisals of the Federal Reserve is from economist Allen Sinai: “The Greenspan
Fed is the all-time champion in American History”, Church, George J., “Alan Greenspan,” Time, volume
150, No 28 (December 28, 1997/January 5, 1998).
5 An executive agency defined under section 105, but does not include the Central Intelligence Agency, the
General Accounting Office, the Panama Canal Commission, the United States Postal Service or the Postal
Rate Commission.
6 Testimony of Vice Chair Alice M. Rivlin on the Government Performance and Results Act, before the
committee on banking and financial services, US House of Representatives, July 29 1997.
7 Board of Governors of the Federal Reserve, Government Performance and Results Act, Biennial Performance
Plan 2004-05 (hereinafter, Fed Biennial Report 2004-05), August 2004.
8 Fed Biennial Report 2004-05.
9 5 USCA appx 1.
10 Reports of the inspector general can be found at the following website: www.federalreserve.gov/oig.
11 Letter from Barry R Snyder, inspector general, to Vice Chairman Roger W Ferguson Jr, July 11 2001.
12 Board of Governors of the Federal Reserve System, “Report on the Audit of the Board’s Efforts to Implement
Performance Management Principles Consistent with the Results Act”, July 2001, p 3.
13 Letter from Steve Malphrus, staff director for management, to Barry R Snyder, inspector general, June 14
2001.
14 Board of Governors of the Federal Reserve, memo from Mark W Olson, 2004 final Reserve Bank budgets
show roughly four per cent of employees and 10% of expenses committed to monetary and economic policy.
This includes the roughly 22,000 employees at the reserve banks, but apparently does not include staff of the
board of governors of about 1,500. Combined data is not readily available. Also see statement of Member of
Congress Carolyn Maloney, House of Representatives, subcommittee on domestic and international monetary
policy, committee on banking and financial services, September 16 1997. (“Instead, I would like to examine
another aspect of the Fed’s operations. Most people do not realise that of the Fed’s 25,000 employees, only
1,600 of them play a role in formulating the nation’s monetary policy. This is only 6.4% of the Fed’s
employees. The other 23,400 federal employees are devoted to other Fed projects.”)
15 These would include ready access to low-cost funding.
16 The Federal Reserve System, Purposes and Functions, 1994, p. 3.
17 GAO report, “Federal Reserve System: Current and Future Challenges Require System-wide Attention”,
GAO/GGD-96-128, p 83 (hereinafter, GAO report), June 1996.
18 The General Accounting Office is the auditing arm of the US Congress; it undertakes reviews and audits of
government agencies and operations.
19 GAO report, p 83.
20 GAO report, p 84.
21 GAO report, p 139.
22 www.federalreserve.gov, “The Structure of the Federal Reserve System.”
23 These laws include the Community Reinvestment Act of 1977, Consumer Leasing Act of 1976, Electronic
Funds Transfer Act, Equal Credit Opportunity Act, Expedited Funds Availability Act, Fair Credit and Charge
Card Disclosure Act of 1988, Fair Credit Billing Act, Fair Credit Reporting Act, Fair Debt Collection
Practices Act, Fair Housing Act of 1968, Federal Trade Commission Improvement Act of 1980, Flood
Disaster Protection Act of 1973, Home Equity Loan Consumer Protection Act of 1988, Home Mortgage

78
McKinley and Banaian

Disclosure Act of 1975, Real Estate Settlement Procedures Act of 1974, Right to Financial Privacy Act of
1978, Truth in Lending Act, Truth in Savings Act and the Women Business Ownership Act of 1988. See
Federal Reserve System, Purposes and Functions (1994), p 90-92.
24 For example, the Justice Department, one of the other federal financial supervisory agencies or state
agencies.
25 Peek Joe, et al. “Is Bank Supervision Central to Central Banking?” (hereinafter, Bank Supervision), March
30 1999, p 2, p 21.
26 Bank Supervision, p 20.
27 www.federalreserve.gov, “The Structure of the Federal Reserve System”.
28 Richard Whiting, executive director, Financial Services Roundtable, letter to Louise Roseman, director,
division of operations and payment systems, July 16 2003.
29 For example, cheque clearing is the most competitive of the services the Federal Reserve engages in as they
hold about a third of the market, with dozens of private sector competitors participating. In most other
services the Federal Reserve maintains 100% or nearly 100% market share. For a discussion of the
competitiveness of the payment markets the Federal Reserve is active in, and the provisions of the Monetary
Control Act which regulate pricing of services, see Cavalluzzo, p 9.
30 Committee on the Federal Reserve in the payments mechanism, The Federal Reserve in the Payments
Mechanism (hereinafter Payment Systems Role), January 1998.
31 Payment Systems Role, p 1-2.
32 Ibid, p 3.
33 Ibid, p 8.
34 Ibid, p 18-19.
35 Ibid, p 19-20.
36 Ibid.
37 See exchange of letters between George Thomas, president and chief operating officer, electronic payments
network, December 18 2002 and Louise L. Roseman, federal reserve division of bank operations and
payment systems, January 17 2003. The exchange of letters was in relation to aggressive pricing strategies by
the Federal Reserve that contributed to the demise of two private sector operators. The review undertaken
cannot be considered independent as the payments committee was made up of governors of the Federal
Reserve, presidents of the reserve banks and a staff director who was a first vice president of one of the
reserve banks.
38 www.federalreserve.gov, “The Structure of the Federal Reserve System”.
39 Payment Systems Role, p 1.
40 Federal Reserve system, Federal Reserve Proposed Currency Recirculation Policy (hereinafter, Currency
Recirculation Policy), October 8 2003. As of Spring 2005 the policy had not been finalised.
41 126 Cong Rec S3168, March 27 1980, statement of Senator Proxmire, as quoted id, p 6.
42 Discussion largely drawn from Currency Recirculation Policy.
43 www.federalreserve.gov, “The Structure of the Federal Reserve System”.
44 Baier Scott L, et al “Income and Education of the States of the United States: 1840-2000”, of Atlanta,
Working Paper, 2004-31, November 2004. Michele Boldrin and Ana Montes, “The Intergenerational State:
Education and Pensions,” of Minneapolis, staff report 336, May 2004. Sahin Ayseful, “The Incentive Effects
of Higher Education Subsidies on Student Effort,” Federal Reserve Bank of New York, staff report 192,
August 2004.
45 Blix, Marten; Daltung, Sonja; Heikensten, Lars, “On Central Bank Efficiency,” Economic Review, Sveriges
Riksbank, 3:2003, p 87.
46 One potential option would be for each of the individual reserve banks to become privately-owned
corporations with publicly-traded stock that act as bankers’ banks, free from the benefits of governmental
status, allowed to compete against private competitors. The individual reserve banks would be free to merge
with one another to achieve efficiencies. Benston, George J. and Humphrey, David B, The Case for
Downsizing the Fed, Banking Strategies, January/February 1997.

79
Central Bank Modernisation

Appendix 2

Variables for the central bank operational efficiency index

Variable Description Source

Personnel costs Direct personnel expense divided by number of Annual financial statements of
employees central banks

Physical capital Fixed assets Annual financial statements of

central banks

Price of labour Personnel expense/ Number of employees Annual financial statements;

Morgan Stanley Central Bank
Directory 2001

Number of on-site Number of on-site exams multiplied by a population World Bank research database
exams normalised factor http://econ.worldbank.org/
by population

Monetary Policy A subindex of the Economic Freedom Index. The Heritage Foundation, The and The
Index index is derived from the weighted average inflation Wall Street Journal
rate from 1992 to 2001. The authors created the
monetary index score by first weighting inflation rates Available at
for each of the past 10 years so that the year farthest http://humandevelopment.bu.edu/
from the present has the least weight and the current
year has the greatest weight. Then the authors
calculated an average of these weighted rates. The
index takes values in [1:5] range. A lower value
indicates better monetary policy.

Banking and A subindex of the Economic Freedom Index. The Heritage Foundation, The and The
Finance Index banking and finance index measures the relative Wall Street Journal
openness of a country's banking and financial system,
and the authors score this factor by determining the Available at
openness of a country's banking and financial system: http://humandevelopment.bu.edu/
specifically, whether foreign banks and financial
services firms are able to operate freely, how difficult
it is to open domestic banks and other financial
services firms, how heavily regulated the financial
system is, the presence of state-owned banks,
whether the government influences allocation of
credit, and whether banks are free to provide
customers with insurance and invest in securities (and
vice-versa). Thus, this index is derived from these
individual factors:

Government ownership of banks

Restrictions on the ability of foreign banks to open
branches and subsidiaries
Government influence over the allocation of credit
Government regulations
Freedom to offer all types of financial services,
securities, and insurance policies

The authors grade each country based on these

factors. The index takes values from 1 to 5. A lower
value indicates a more conducive environment to
banking and finance.

Non-performing World Bank research database

loans / total assets http://econ.worldbank.org/

80
McKinley and Banaian

Appendix 3

Estimation results for the central bank operational efficiency index

Dependent variable is sum of personnel and fixed asset costs; OLS = ordinary least
squares regression, SFR= stochastic frontier regression. T-stats in parentheses.
Independent variable OLS SFR OLS SFR

1.0008 0.9254 1.0289 0.9609

Wages (4.67) (5.31) (124.6)
(11.78)
-0.6354 -0.6226
Monetary Policy Index
(1.02) (0.78)
0.1088 -0.6890
Non-Performing Loans to Assets Ratio
(0.09) (0.41)
-0.0025 0.2337
Banking and Finance Index
(0.00) (0.43)
-0.1863 -0.1868 -0.1412 -0.3150
Number of Bank Supervision Audits
(1.86) (1.60) (1.69) (11.81)
-.01644 0.1609 0.3272 0.6048
Price stability 1996-2001
(0.04) (0.27) (1.92) (31.66)
0.6570 0.7759 0.6948 0.8051
Population
(7.00) (12.48) (9.12) (95.32)
16.970 1.1073 -4.2974 -5.1438
Constant
(2.13) (0.27) (1.77) (31.85)
Standard error — total 0.7073 0.6852
Standard error — statistical noise 0.0059 0.0035
Standard error — inefficiency variability 1.2131 1.3196
Log-likelihood function -29.723 -29.572 -30.591 -32.022

The model was estimated using the nonlinear estimation algorithm in the statistical
package SHAZAM, version 10. Ordinary least squares estimates were used as
starting values for the SFR; the nonlinear routine needed 74 iterations to generate a
solution. We found the iterative solution was sensitive to the starting value of the
ratio of statistical noise to inefficiency variability (λ/σ in fn. 52); too high a ratio
created nonsensical values for the inefficiency measures. The results provided show
the estimation of the parameters of the model. For sake of comparison we include
the OLS estimates of the parameters. In parentheses are the t-values for the
estimation.

81