Technical ReportPDF Available

INSTITUTIONAL DRIVERS OF EFFICIENCY IN THE PUBLIC SECTOR

Authors:

Abstract and Figures

This technical paper has been prepared as a contribution to the Government at a Glance publication. The major objective of Government at a Glance is to present relevant and reliable data on how governments work, what resources they use and what they achieve with those resources, offering valuable international comparisons. Providing public services efficiently is a stated goal of all OECD member country governments. This paper explores the available empirical evidence on the institutional drivers of government efficiency, aiming to show which institutional arrangements work and under what circumstances. It also creates a good foundation for the inclusion of selected institutional variables in Government at a Glance.
Content may be subject to copyright.
Unclassified GOV/PGC(2007)16/ANN
Organisation de Coopération et de Développement Economiques
Organisation for Economic Co-operation and Development 17-Jan-2008
___________________________________________________________________________________________
English - Or. English
PUBLIC GOVERNANCE AND TERRITORIAL DEVELOPMENT DIRECTORATE
PUBLIC GOVERNANCE COMMITTEE
INSTITUTIONAL DRIVERS OF EFFICIENCY IN THE PUBLIC SECTOR
Wouter Van Dooren (University of Antwerp), Zsuzsanna Lonti (OECD GOV Directorate), Miekatrien
Sterck and Geert Bouckaert (University of Leuven)
36th Session of the Public Governance Committee
16-17 October 2007
For more information, please contact Zsuzsanna LONTI:
Tel: +33 1 45 24 16 39; Email: zsuzsanna.lonti@oecd.org
JT03238855
Document complet disponible sur OLIS dans son format d'origine
Complete document available on OLIS in its original format
GOV/PGC(2007)16/ANN
Unclassified
English - Or. English
Cancels & replaces the same document of 04 October 2007
CONFIDENTIAL GOV/PGC(2007)16/ANN
2
TABLE OF CONTENTS
INSTITUTIONAL DRIVERS OF EFFICIENCY IN THE PUBLIC SECTOR.............................................3
1.Introduction ..........................................................................................................................................3
2.What is efficiency and why is it important?.........................................................................................3
3.Measurement challenges in the public sector.......................................................................................4
4.Theoretical and empirical evidence on the likely institutional drivers.................................................6
4.1.Results orientation.......................................................................................................................6
4.2.Increased flexibility.....................................................................................................................9
4.3.Strengthening competitive pressures.........................................................................................14
4.4.Workforce issues.......................................................................................................................16
5.Conclusion..........................................................................................................................................20
BIBLIOGRAPHY .........................................................................................................................................22
ANNEX I: MEMBERSHIP OF STEERING GROUP.................................................................................29
ANNEX II: DATA DESCRIPTION OF THE REVIEWED LITERATURE..............................................30
This paper has been prepared under the guidance of a Steering Group (see membership details in Annex I),
Nick Manning (World Bank), Dirk-Jan Kraan, Theresa Curristine, Barry Anderson, Espen Erlandsen
(OECD Economics Directorate). Special thanks for preparing the summary tables to Maarten Luts, for the
compilation of the Annex II: Data Description of the Reviewed Literature.
CONFIDENTIAL GOV/PGC(2007)16/ANN
3
INSTITUTIONAL DRIVERS OF EFFICIENCY IN THE PUBLIC SECTOR
1. Introduction
This technical paper has been prepared as a contribution to the Government at a Glance publication.
The major objective of Government at a Glance is to present relevant and reliable data on how
governments work, what resources they use and what they achieve with those resources, offering valuable
international comparisons. Providing public services efficiently is a stated goal of all OECD member
country governments. This paper explores the available empirical evidence on the institutional drivers of
government efficiency, aiming to show which institutional arrangements work and under what
circumstances. It also creates a good foundation for the inclusion of selected institutional variables in
Government at a Glance.
The evidence on the institutional drivers of efficiency in the public sector is surprisingly limited.
Available research provides a very limited assessment of the impact on efficiency of varying the mix of
inputs used (more staff on contract, less permanent employees, for example) or of changing structural and
managerial arrangements.
The paper starts by discussing the definition of efficiency and assessing whether and how the concept
can be used in the public sector, focusing on some of the distinctive features of the public sector. It then
summarises some of the key problems in the measurement of efficiency, reviewing several methods to
measure inputs and outputs. The review of institutional arrangements is divided into four main groups:
i) results orientation; ii) strengthening competitive pressures; iii) increased flexibility; and iv) workforce
issues. For each institutional driver the report set outs the major theoretical propositions concerning its
impact on efficiency and the empirical evidence available from the literature to assess these claims.
2. What is efficiency and why is it important?
Productivity and efficiency are regularly used interchangeably in colloquial language. Yet there is a
distinction between them in economic theory. This paper follows the standard definitions of productivity
and efficiency from economic theory.
Productivity is interpreted as a ratio of a volume measure of output to a volume measure of input
(OECD, 2001). If there is only a single input and a single output, the calculation is obvious. If there are
multiple inputs and outputs, then a method for aggregating inputs and outputs into a single index of inputs
and outputs are needed to create a ratio measure of productivity. When referring to productivity we mean
total factor productivity, which is a productivity measure where all factors of production are considered.
Productivity can also be measured for separate inputs. In that case the contribution of a single input to the
outputs is isolated. For instance (average) labour productivity defines the amount of output for each unit of
labour (Coelli et al, 1999). However, partial measures of productivity can be misleading when considered
in isolation. We have to take other production factors into account when we compare partial measures over
time or between countries.1
CONFIDENTIAL GOV/PGC(2007)16/ANN
4
In layman’s terms, efficiency means doing the right thing the right way. In economics, efficiency has
two dimensions: i) technical or operational efficiency refers to the output-input ratio compared to a
standard ratio , which is considered optimal or ideal (and so can never exceed 100%) . Both output- and
input-oriented efficiency can be defined. Output efficiency focuses on the maximization of output for a
given set of inputs, or alternatively, input orientation aims at the minimizations of inputs for a given set
outputs. ii) allocative efficiency refers to the use of inputs in optimal proportions given their respective
prices and production technology. For example, allocative efficiency in input selection involves selecting
the mix of inputs (e.g. labour and capital) which produce a given quantity of output at minimum costs,
based on prevailing input prices. Economic (or cost) efficiency is the product of technical and allocative
efficiency. Further when we refer to efficiency in the text we mean economic efficiency.
Efficiency and productivity measures are in essence concerned with delivering more outputs for the
resources consumed. Their definition means that they do not capture the intrinsic value of those outputs –
an organisation can be technically efficient in producing the wrong services. There is however, a strong
case for focusing on the production side – as they are very relevant to issues faced by policy makers and
managers.
Efficiency can also be improved by exploiting economies of scale. However, this is not necessarily a
viable option for public sector organisations, as output and/or input levels are seldom their own decision,
as they are constrained by political principals and legal requirements.2
Preset output levels. In most welfare states, citizens have a legal right on a number of public services.
Reducing output is usually not an option for public services such as hospitals, courts and unemployment
agencies. These institutions however can attempt to minimize inputs. In other words, they should be input-
efficient.
Preset input levels. Input levels are often also predetermined in advance. In the private sector, a
budget is an estimate. Companies can deviate from the budget, amongst others in order to exploit
economies of scale. In the public sector, the budget is not only an estimate, but also an authorization. It is
an expression of a political desire that a specific amount of resources is used to attain societal outcomes as
well as a guarantee that overall public budget is in balance. For instance cultural institutions, agencies for
economic promotion and environmental agencies usually have a fixed input. These institutions however
can attempt to maximize outputs. In other words, they should be output-efficient.
3. Measurement challenges in the public sector
The quantification of efficiency is the most critical challenge in efficiency analysis. There are many
reasons why public sector organisations have difficulties in collecting performance information. We
discuss some of these difficulties and suggest how to deal with them. In some cases, the problems of
measurement will be hard to overcome. It is an illusion to think that everything is measurable. However,
even if meaningful quantification is not achievable, it will usually be worthwhile to raise the efficiency
questions on a conceptual level.
Measurement of efficiency requires quantitative information on both input or costs and output (or
volume) of public service provision. Measuring output requires capturing the quantity and quality of
products and services. Quality is to be understood as the degree to which outputs contribute to outcomes,
consequently , it needs to be incorporated in any measures of outputs. Micro level measures capture the
outputs of individual organisations, intermediate measures capture outputs at the sectoral level, and macro
measures at the government-wide level.
CONFIDENTIAL GOV/PGC(2007)16/ANN
5
Macro efficiency measurement may be useful to distinguish trends but is less useful for managerial
purposes. Typically, macro level data encompass a diffuse set of activities of numerous units. However,
managers should only be accountable for concrete outputs that are controlled by the units they manage.
Moreover, macro level data do not show where an efficiency problem occurs, and thus do not tell where
managers need to intervene. Micro level measurement has more to offer for public managers. A micro
level focus is required if we want to gain understanding of the underlying causes.
Micro level measurements are the building blocks and higher order measures require aggregation,
which poses several problems (Murray, 1992). First, only the final outputs (the products and services that
are produced for the public or business) should be counted, as the inclusion of intermediate outputs will
lead to double-counting. However, in the public sector it is much more difficult to identify that final
output. First, in the private sector the output of a firm is roughly everything that has a price tag. In the
public sector because of the absence of price information this is much harder to do. Second, aggregation
requires weights. In the absence of price information there is no obvious way for adding up, for example
road patrols and detective work of a police agency. The third issue concerns how quality of goods and
services should be incorporated also in the absence of price information.
Another challenge is that the output mix of many public sector organisations includes intangibles.
Hackman and Oldham (1980) distinguishes between three types of outputs with progressively less tangible
outputs: routine based tasks, human relations based activities, and process based activities. A diplomatic
service, for example, has many process based activities (e.g. negotiating, interest representation), some
human relations based (e.g. issuing immigration permits) and some routine processes (e.g. security
checks). Although such outputs can be captured, the risk is of focusing on the measurable outputs and to
neglect other outputs. Experience has shown that measurement can concentrate too much on the
measurable proxies of public service provision to the detriment of less tangible aspects (Shepard, 1990;
Smith and Rottenberg, 1991).
A related risk is that the measuring of the outputs creates incentives to “game” or cheat, particularly
when output measures directly affect monetary or career incentives (Bouckaert and Balk, 1991; Smith,
1995). Incentives can be monetary as well as reputational. This generally happens in one of two ways (Van
Dooren, 2006). Data can be deliberately captured in a misleading way. In the United Kingdom, for
example, ambulance response-times were misrepresented by using incorrect definitions of when the clock
has to start and stop running (UK House of Commons, Public Administration Select Committee, 2003).
Alternatively, organisational behaviour can be adapted specifically in order to change the output measures
regardless of other perverse consequences.
Measuring inputs is more straightforward. Accounting systems can be regarded as institutionalised
measurement systems for the input side. Efficiency analysis requires cost accounting, and cost accounting
requires an accounting system that registers costs and not only cash flows. A cash-based system is
insufficient.
Cash-based accounting is a method of bookkeeping that records financial events based on cash flows
and cash position. Revenue is recognised when cash is received and expense is recognised when cash is
paid. Accruals-based accounting records financial events based on events that change the net worth. In an
accrual based system, a transaction is registered when the activity generating revenue or when resources
are consumed, regardless of when the associated cash is received or paid. Accrual based systems record
transactions both on an accruals basis and on a cash basis (Blöndal, 2003).
The benefits of accrual accounting are twofold. Efficiency and productivity measures can only be
calculated when costs can be allocated to outputs. Secondly, accrual accounting systems report information
CONFIDENTIAL GOV/PGC(2007)16/ANN
6
that reflects better the economic reality of the organisation. Nearly one-third of OECD member countries
have adopted full accrual accounting.
For costs to be imputed to outputs, it is not only necessary to maintain a system of accrual accounting
but also to maintain a management accounting system that splits out all costs to separate outputs. This
requires a financial administration at the agency level that records all costs of all outputs produced.
Only the governments of Australia, New Zealand and the United Kingdom have adopted full accrual
budgeting (Sterck & Scheers, 2006). Accrual budgeting is less popular among OECD member countries as
it is believed to risk budget discipline and is considered too complex.
However, even input information in monetary terms is not always available. In order to circumvent
the absence of monetary input information, proxies of the input are used in the same way researchers make
use of proxies to assess output volumes. . Most commonly, labour (number of full time equivalents) is used
as a proxy for the input. In some sectors, such as education, labour costs make up the bulk of the total
inputs. In these cases, the use of a proxy is a good strategy. In other sectors, such as hospitals, labour costs
are a considerably lower portion of the total input mix. In this case, other (additional) proxies will be
needed.
It is simply not possible to apply private sector accounting systems in the public sector. An important
issue is the presence of assets and liabilities that do not exist in the private sector, such as heritage assets,
military assets and social insurance programs. The definition of valuation rules (historic cost method
versus current costing) is an important choice that has to be made. There is a need for independent
development of harmonized public sector accounting standards and the International Public Sector
Accounting Standards is an important step in this direction. Implementing accrual accounting and
budgeting is a costly business that needs to be supported by ICT reforms, training and communication
efforts. Overall, accrual accounting and budgeting is not a “magic bullet” of public sector performance
improvement, but only a tool for better financial information that must be used in combination with other
management instruments (Blöndal, 2003: 55).
4. Theoretical and empirical evidence on the likely institutional drivers
As there is a large amount of literature on the subjects in this review we included primarily studies
that contain a meta-analysis, meaning that they summarise the empirical evidence from a large number of
related studies. The literature is discussed under four main headings as mentioned in the introduction.
4.1. Results orientation
4.1.1. A broader view: performance measurement arrangements
Theoretical proposition
Bouckaert and Peters (2002) argue that performance measurement is the Achilles heel in public sector
reform. It is a critical success factor for many reforms to succeed. Public managers increasingly have to
make use of performance information in performance contracts, performance budgets, performance related
pay, etc. (Van Dooren, 2006) some of which will be discussed later in section 4.2.4 under Human
Resources Management (HRM) practices.
A crucial feature of performance measurement arrangements is the coupling of performance
information to decision making (Van Dooren et al., 2006). Decisions can be of two types; planning or
accountability and control. Planning decisions are about future action, while accountability and control
decisions settle on the consequences of past performance. Systems that induce yardstick competition are an
CONFIDENTIAL GOV/PGC(2007)16/ANN
7
important example of a performance arrangement for improved accountability (Dawson, Goddard and
Street, 2001).
A tight coupling implies that measurement leads to the decision in a direct way. Decisions are driven
mainly by measurement. Other sources of information play a negligible role. The underlying reasoning is
that the tight relationship between measure and decision will put external pressure on institutions to
perform better. Target setting is a crucial element in this respect.
In case of a loose coupling, output measurement is one source of information to be incorporated with
others. Other sources of information are used to interpret the output measurement data and decisions are
informed by output measurement, but also by other sources of information such as experience, qualitative
information, etc. The underlying reasoning usually is that institutions are staffed by professionals who have
an intrinsic motivation to perform. Performance arrangements have to provide these professionals with
information in order to better do their job.
Looking at the relationship between performance measurement and decision making, a tight coupling
for planning seems not feasible because of political and technical constraints. Politicians do not want their
hands tied by measures. Technically, it is hard to predict future developments by means of indicators of
past performance. Tight coupling for accountability has a strong enforcement effect, but will more often
than not be undermined by gaming. Loose coupling offers more prospects for planning, but the impact of
the measures can be diluted. In any case, the direct enforcement effect of a loose coupling will be lower.
However, when measures are agreed upon by principals and agents, there can be an indirect effect of
persuasion
Empirical evidence
Researchers point to the fact that performance arrangements are not neutral devices and may have
negative consequences. Yet, their positive effects are seldom investigated. The question whether
performance measurement and management does lead to better performance remains largely unanswered.
There is growing empirical evidence about the negative effects of performance arrangements (Van
Dooren et al., 2006). The potentially negative effects occur because organizations act tactically in order to
obtain better scores. Bevan and Hood (2005) document three important gaming problems.
First, the ratchet effect refers to the consequences of central resource managers basing next year’s
targets on this year’s performance. The impact of this practice is that managers have an incentive to reduce
their output increases to a modest increment so that expectations and future targets will be set at a low
level. Secondly, the threshold effect describes the tendency to focus agency attention on those outputs that
are near to the required level of output. This leads to the concentration of effort on outputs that are just
below the required level at the expense of others, ignoring the best (on the basis that these outputs will
meet the test without effort) and the worst (on the basis that the effort required is outweighed by the cost of
improving these to the minimum standard). Thirdly, distortion refers to the achievement of output
improvements in areas that are measured at the expense of unmeasured aspects of performance.
It seems reasonable to assume that the impact of performance arrangements on performance will
depend on the nature of the performance arrangement. As we have argued, a crucial feature of a
performance arrangement is its coupling with decision-making. Negative effects will mainly occur in tight
systems. Gaming can be mitigated by loosening up the performance arrangements. This was also the
recommendation of the Public Administration Select Committee of the House of Commons (2003). They
reviewed the United Kingdom’s experience with performance arrangements and suggested to move from a
measurement culture to a performance culture.
CONFIDENTIAL GOV/PGC(2007)16/ANN
8
In many OECD countries, performance arrangements that are more loosely coupled to accountability
and planning decisions are in place. These systems are set up to inform professionals and politicians in
public organizations rather than to punish or reward them. Askim (forthcoming) for instance found that
Norwegian counselors’ do use performance information. Yet, performance information is not tightly
coupled to decisions, but is used to set decision agendas, to map out consequences of alternative policies,
to signal rationality, and to win debates.
4.1.2. Budget practices and procedures
Theoretical proposition
Modern budgeting practices aim to provide more financial autonomy for managers and better
information. The use of accrual accounting helps to make the true cost of government activities more
transparent. This will enhance financial management practices and increase efficiency. The modernisation
of the financial cycle is not merely a technical exercise. New forms of financial and non-financial
information have to meet the needs of various users. Moreover, budgeting, accounting and audit should
optimally support the policy and management cycle.
To this end, governments initiated accrual accounting, cost accounting, multi-year planning and more
performance information in budgeting. A good, medium-year system of budget estimates and hard
budgetary constraints may also provide further incentives for efficiency. This may go hand in hand with
increased budget transparency and reduced budget fragmentation, by monitoring extra-budgetary funds and
contingent liabilities (e.g. loan guarantees and public pension schemes) (Joumard et al., 2004: 110).
Empirical evidence
In order to enhance the efficiency of the budget process, different countries have strengthened fiscal
rules since the latter half of the 1990s. Examples are limits on budget deficits, expenditures, tax or debt.
However, it appears to be difficult to enforce these fiscal rules throughout a medium-term cycle and
political willingness to respect the fiscal objectives is crucial. Fiscal consolidation efforts carried out within
a short time-may cause cuts in investments. On the other hand, the small initial amounts required for many
investment programs may protect them from disproportionate impact from fiscal rules. Expenditure
ceilings may also lead to gaming, creating incentives to use off-budgetary funds, public-private
partnerships and loan guarantees which may hide spending commitments (Joumard et.al., 2004).
Seventy-two percent of OECD member countries include non-financial performance data in their
budget documentation but only 18% of the countries specifically link expenditure to all or most of their
output or outcome targets. In 46% of OECD member countries there are no rewards or sanctions applied if
a target is met or not met (Curristine, 2005b: 134).
Public managers have more autonomy in their spending decisions. Carry-over-provisions should
avoid inefficient end-year spending splurges. In a large number of countries the line-item approach has
been gradually abandoned and managers receive a single appropriation for all their operating costs.3
Various researchers have posed critical remarks regarding the implementation of results-oriented
financial management (Rubin & Kelly, 2005). The lack of high-quality performance data and well
formulated objectives is one of the most crucial causes for the failure of reforms (Carlin, 2006). Too strong
an emphasis on the efficient production of outputs can lead to compartmentalization and requires an
adaptation of the system towards more horizontal coordination mechanisms. Some authors question
whether New Public Management type reforms can really improve management (Carlin & Guthrie, 2000).
CONFIDENTIAL GOV/PGC(2007)16/ANN
9
It tends to be even more difficult to use results and cost information in allocation decisions.
Performance data seem to play a more important role in the preparation of the budget than in the budgetary
decision-making (Jordan & Hackbart, 1999; Melkers & Willoughby, 2005). Further, a major challenge is
strengthening transparency and accountability in order to better serve councillors and citizens. When
implementing an accrual accounting system with preservation of the traditional cash budget, the accrual
information is mainly used as a managerial instrument rather than as an instrument to support budgetary
decision-making (Paulsson, 2006).
The research on the direct formula performance budgeting is more straightforward. Robinson and
Brumby analyzed the system of output-based funding in hospitals. They concluded that this system led to
strong productive efficiency gains without major unintended effects. “This case showed that performance
measurement, management accounting and the relaxation of input controls are crucial underpinnings of
successful performance-based funding” (Robinson & Brumby, 2005: 44).
Very little evidence exists that performance information is used in the political budgetary decision-
making process. The major impact of results-oriented budget reform is situated in the internal management
of departments and agencies. Performance information is used to allocate resources and set priorities.
“While the shift of focus from control of inputs to outcomes may deliver efficiency gains, it is important to
recognise that this can also create an upward drift in public spending” (Joumard et al., 2004: 130). At the
level of legislative oversight, there is no information whether members of Parliament are extensively and
systematically using the information on outputs, outcomes, and costs.
4.2. Increased flexibility
4.2.1. Devolution/decentralization
Theoretical proposition
The devolution of functional responsibilities is accelerating. In this paper we confine the concept of
devolution to the transfer of responsibilities from central to sub-national governments (i.e. regions and
local governments). In principle, devolution of functional responsibilities, if accompanied by appropriate
fiscal and political decentralization, provides incentives for sub-central governments to deliver locally
preferred services more efficiently.
The core logic of fiscal and political decentralization is that “governments closest to the citizens can
adjust budgets (costs) to local preferences in a manner that best leads to the delivery of a bundle of public
services that is responsive to community preferences” (Ebel and Yilmaz, 2002: 4). The burden and the
benefits of public service delivery both accrue in the communities. In other words, the allocative optimum
is best determined at a decentralized level because spending and taxing decisions go hand in hand.
It is also argued that decentralization has an impact on the size of government. The Leviathan
hypothesis argues that federalism and decentralization will strengthen competition between governments
and therefore will reduce the size of government (Brennan and Buchanan, 1980).
There are limits to fiscal decentralization, mainly because of a too strong tax and service competition
between sub-national governments. Therefore, most countries have systems (e.g. municipal funds) that
fund governments with a weak tax base (socio economic indicators), with outputs that produce benefits
beyond the local boundaries (the so-called regional function) or with exceptional demands on outputs
(e.g. rural areas with a disproportionate road infrastructure). There is clearly a tension between increasing
efficiency through decentralization and its potentially negative distributional effects.
CONFIDENTIAL GOV/PGC(2007)16/ANN
10
Empirical evidence
Decentralization seems to be beneficial. Swiss cantons provide robust evidence that more
decentralization is associated with higher educational attainment (Barankay and Lockwood, 2006). Huther
and Shah (1998) apply several governance indices on the level of fiscal decentralization in developed and
developing countries. They conclude that “the relationship between the level of decentralized expenditures
and governance quality appears to be strictly increasing but clearly there must be some form of “Laffer
Curve” – it is easy to construct cases where complete decentralization of expenditures would lead to lower
quality governance than where there is a mix of national and sub-national expenditures” (p.17). In addition,
it has been argued that the high performance of Scandinavian countries on quality of life indices is partly
caused by the high level of local autonomy (Social and Cultural Planning Office, 2004).
Most studies seem to support the Leviathan hypothesis that decentralized taxation reduces the size of
government (Feld et al., 2003). However, support for this hypothesis is derived primarily from studies that
compare sub-national governments within federal countries (mainly by comparing US states). Evidence on
the comparison of nations is inconclusive.
4.2.2. Delegation through agencification
Theoretical proposition
The two main arguments for creating agencies are the need for more management and policy
autonomy. Our focus is on increased management autonomy as a driver for improved performance. More
management autonomy would lead to better management due to clear, specific objectives and incentives.
The assumption is that the performance of public agencies can be enhanced only if more managerial
autonomy (i.e. less input controls on financial and human resource matters) is devolved to agencies; if they
receive incentives to align their interests with those of the government; and if they have to report on their
performance to the government (Verhoest, 2005: 236). Autonomy is considered a necessary condition to
motivate managers to improve their performance by rewarding them for accomplishing the goals or
expected results. When agencies are allowed to retain surpluses, then they will strive to improve efficiency.
There are different theoretical schools which predict positive effects of agencification on efficiency
and effectiveness. Agency theory states that actors are motivated to perform well to the extent that they are
considered to be residual claimants. Property rights theory assumes that the more property rights are
centralised in one actor, the more this actor will be induced to perform in an efficient way. Network
theories state that increasingly complex environments require more delegation. In the case of UK agencies
the argument is that networks offer a way of preserving some of the original efficiency objectives of
agencification whilst reducing the fragmentation of the public sector.
Empirical evidence
Different scholars have pointed at the lack of systematic evidence on the relationship between
agencification and performance (Pollitt et al., 2001; Pollitt, 2004,). “Most contemporary studies on the
effect of autonomization and control of public agencies on performance seem to show ambivalent
results“(Verhoest, 2005: 236). An immediate and overall improvement in the efficiency and effectiveness
of policy implementation is not found. “One should not exaggerate the situation: there are certainly case
studies and performance indicator data showing how specific agencies have improved specific aspects of
performance. However, it is usually difficult if not impossible to connect such improvements to agency
status per se” (Pollitt et al., 2001: 278).53. In Sweden, it is proven that the efficiency of the agency-
CONFIDENTIAL GOV/PGC(2007)16/ANN
11
ministry system is based on the strength of informal networks. Informal mechanisms of policy advice and
coordination have proven critical to the system (Pierre, 2004).
There is some scientific evidence that a reduction of input controls combined with steering on results,
financial incentives and competition will lead to savings and increased efficiency. However, the impact on
the quality of service delivery and policy effectiveness is less clear. There is some scientific evidence that
service characteristics (e.g. measurability) influence the effect of agencification.
At the same time, several scholars have pointed at the negative effects of agencification. They have
raised the issue of fragmentation caused by agencification and have questioned the feasibility of directing
agencies through contracts. There are different possible explanations for the negative effects of
agencification: difficulties of managing by the numbers; gaming; lack of competent parenting (as agencies
can consistently outgun the understaffed, unqualified ministries as their personnel rules allow them to hire
experts); lack of internal management capabilities in agencies; and budgetary instability. The major risks of
agencification that can threaten overall government efficiency and efficacy and accountability are: the risk
of losing control of agency operations; the abrogation of political accountability; evasion of general rules
for staffing and budgeting; exposure of government to financial and employment risks; and opportunities
for political patronage and corruption (Laking, 2005).
4.2.3. Intra-governmental coordination
Theoretical proposition
The need for a more coordinated approach in the public sector arose from a necessity to deal with a
growing number of interconnected policy issues and the challenge to manage an administration which has
experienced a fairly large degree of decentralization. There is also a need to counterbalance the
performance management systems that are mainly focusing on individual organisations and to take into
account the growing importance of specific client groups demanding a coordinated range of services
(Peters, 2003). It can also be explained from more general, recent tendencies of interconnectedness and
mutual interdependence in the modern, networked society (Boelinck, 2007).
Theoretically, one can distinguish between three major types of coordination mechanisms:
i) hierarchy-type mechanisms that focus on objective- and rule-setting, on allocation of tasks and
responsibilities, and on lines of control (e.g. creation of mega departments); ii) market-type mechanisms,
focusing on the creation of incentives to enhance the performance of public actors; and iii) network-type
mechanisms, focusing on the establishment of common knowledge, common values, and common
strategies between partners (e.g. linking info systems, consolidating financial documents, chain
management).
Empirical evidence
Verhoest and Bouckaert (2003) analysed coordination mechanisms in six countries. Countries have
followed the three trajectories identified above. There is little evidence on the effectiveness of coordination
mechanisms. Moreover, the relationship between coordination mechanisms and efficiency is unclear. The
efficiency argument is mainly used as an argument for reorganisation and specifically for mergers.
Merging organisations would lead to efficiency gains through economies of scale and reduction of
duplications. However, merging different organisations into mega departments may hide the coordination
problems as these may still occur within this mega department.
4.2.4. HRM arrangements
Theoretical proposition
CONFIDENTIAL GOV/PGC(2007)16/ANN
12
Management literature emphasises that people are the most important organisational resources and
that human resource management is a key driver for performance (Osborne & Gaebler, 1992). Civil service
systems are, however, often depicted as rigid and rule bound. Public managers complain that personnel
systems impede their ability to manage and make critical personnel decisions. Employees are frustrated
because they are not adequately compensated and do not receive sufficient recognition (Selden, Ingraham
& Jacobson, 2001: 598). The public sector would not be able to compete with the private sector to attract
high-performing staff. Human resources management reforms try to cope with this issue. The flexibilility
of the civil service systems is increased both through individualisation of HR policies and delegation of
HR decisions. Moreover, incentives for high performance are created by linking individual objectives to
the organisational mission and objectives.
Traditionally, civil service systems have been classified as either position-based or career-based
systems. Position-based systems are believed to allow for more individualisation and decentralisation than
career-based systems. As flexibility could lead to more efficiency, this would imply that position-based
systems are more efficient than career-based systems.
Delegation of HR practices can take different forms, for example by introducing more flexible
working-time arrangements, increased mobility, easier deployment, simplified/flexible recruitment
arrangements, open recruitment, flexible terms of employment, simplified employee termination/laying
off, more flexible and less complex classification, flexible pay arrangements, introducing operating costs,
and central HR bodies oriented towards more strategic concerns.
Theories on human resources distinguish between hard and soft styles of human resource
management. The former treats personnel as instruments that can be manipulated to obtain organisational
ends. Hard incentives are needed to motivate employees. The latter approach thinks that effective
personnel management takes care of the needs and aspirations of individuals. A satisfied and motivated
workforce will lead to higher performance.
Empirical evidence
The strict division between career-based systems and position-based systems does not reflect reality.
There are many countries that fall in between. They are characterised by relatively high level of delegation
to line ministries and a relatively low level of individualisation (lifelong careers and minimum lateral
entry). These hybrid systems are named department-based systems (OECD, 2005a). There are also
countries with a high level of individualisation and a low level of delegation.
Individualisation has mainly taken place in the selection process, the terms of appointment,
termination of employment, performance management and pay. Staff can, increasingly, be treated
differently according to the changing needs of organisations and depending on their performance.
An OECD (2005e) study showed that performance-related pay initiatives appear to have a low impact
on staff motivation. “Staff surveys showed that only a small percentage of employees thought their existing
performance pay schemes provided them with an incentive to work beyond job requirements and in many
cases they found it divisive” (OECD, 2005e: 1). Staff are less motivated than might have been expected by
the prospect of more money for working better. This study corroborates earlier findings on the failure of
the performance system in the United States in the 1980s, both in terms of organisational effectiveness and
motivation (Dailey, 1992; Halachmi and Holzer, 1987). This is not only the case in the public sector.
Ingraham (1993) argues that the success of performance pay in the private sector is also unconvincing. Yet,
she argues that rewarding performance does make sense. Therefore, systems that are adapted to public
sector context need to be designed. Group incentives are a good example.
CONFIDENTIAL GOV/PGC(2007)16/ANN
13
A 2002 survey of 17 states and 284 local governments in the United States showed that the
performance bonuses accounted on average for 6% of the income (Lewin, 2003). Even where
performance-pay schemes allow for variable payments, most employees tend to receive similar ratings as
managers appear unwilling to differentiate among their subordinates (World Bank). However,
performance-related pay can help improve performance as it may facilitate organisational changes by
linking bonuses to new objectives. Performance-related pay can also act as a recruitment incentive. “In
Denmark for 57% of managers, performance-related pay leads to better opportunities for recruitment.
Similar experiences occurred in attracting and retaining top quality school teachers in England and Wales”
(OECD, 2005e).
Brewer (2005) studied human resources management in 22 United States’ federal agencies. The study
shows that human resources management matters a great deal. High-performing agencies tend to have
skilful upper-level managers, strong cultures that value employees and emphasise the importance and
meaningfulness of the agency’s work, and policies that empower those employees. Boyne (2003) analysed
eight quantitative studies that looked at the impact of human resource management on performance.
Boyne’s overview suggests that soft aspects of human resource management, such as employee satisfaction
and morale, are more important drivers of performance than hard aspects of personnel management, such
as performance-related pay and job security. These findings are consistent with studies in the 1980s that
stress the importance of soft HRM. Brewer and Selden (2000) showed that human capital building and
retaining high performing people correlate positively with organisational performance. The level of
training did not explain the level of performance
Zigarelli (1996) found a positive correlation between staff morale and student achievement in
American schools. This study also supported the hypothesis that a higher level of autonomy in personnel
decisions (such as hiring and firing teachers) correlates positively with student achievement.
To conclude it is safe to put that human resource management does matter. The soft aspects of human
resource management appear to be more important than the hard aspects. “The greater the ability of the
organisation to reorganise processes and match individuals to the right position within a mission-based
organisation, the greater the ability of the organisation to achieve performance” (Moynihan & Pandey,
2005: 427).
4.2.5. E-government
Theoretical proposition
There is a general tendency in the public sector towards automating bureaucratic procedures and
processes, and electronic interaction with citizens. However, e-government is a very broad concept,
ranging from electronic communications, via online services to e-democracy and e-participation. The
OECD defines e-government as “the use of information and communication technologies, and particularly
the Internet, as a tool to achieve better government”.
There are different stages in the development of e-government (Moon, 2002: 426): i) one-way
communication (billboard stage); ii) two-way communication (request and response); iii) transaction:
service delivery and financial transaction (e.g. tax online); iv) vertical and horizontal integration: inter-
governmental and intra-governmental integration (e.g. data sharing); v) political participation (e.g. voting
online).
The basic assumption is that e-government leads to better government, by enabling better policy
outcomes, higher quality services, greater participation of citizens, and by improving efficiency,
contributing to economic policy objectives and advancing the public reform agenda. Investment in e-
CONFIDENTIAL GOV/PGC(2007)16/ANN
14
government can lead to more efficiency and savings due to a decrease in the number of civil servants.
Trimming down the back-office, it would release significant extra resources for high priority areas.
In the discussion on the impact of e-government on efficiency there is a need to differentiate between
the efficiency gains expected from innovations in ICT in trimming administration and its substantial
impact on the efficiency of service delivery.
Empirical evidence
There has been little empirical evidence to test the claims about e-government. The evidence of the
benefits of e-government is rather anecdotic than based on empirical research. Some studies are based on
perceptions of the impact of e-government and not on actual results. For example, a survey with chief
information officers of 50 states and 38 major federal agencies in the United States reported that 83% of
respondents believed that e-government had made government more efficient and 63% claimed that e-
government has reduced government costs (West, 2004:17). A survey on of e-government among
American municipalities was, however, less optimistic (Moon, 2002). This study concluded that e-
government has been adopted by many municipal governments, but is still at an early stage and has not
obtained many of the expected outcomes such as cost savings and downsizing. Few cities have experienced
cost savings, procurement cost savings or reductions in the numbers of staff, while many cities have
observed changing roles of staff and changes in business processes. It appears that e-government practices
reduce time demands but increase task demands on staff. E-government initiatives save time but require
more technical skills from staff members.
Until now, the expected efficiency gains have been the major driver for e-government reform. The
current movement towards improvement of service delivery through the use of ICTs is very technology-
driven (Snijkers, 2005: 2). There is a need to consider the impact of these innovations on public
administration values such as openness, transparency, accountability and social equity. The OECD E-
government project concluded: “The initial impressive visible results of e-government (government
websites, a number of sophisticated transactional services, development of portals) contrast with the next
stage of e-government which requires the development of hidden infrastructure, connected back office
arrangements and more complex services. Greater collaboration across levels of government, higher
funding levels and deeper organisational change will be needed” (OECD E-Government Task Force,
2003).
4.3. Strengthening competitive pressures
4.3.1. Privatisation
Theoretical proposition
The input categories for public service production in the National Accounts are labour, procurement
of goods and services, gross capital investment, social benefits in kind and subsidies. These input
categories are building blocks of the ‘mode of production’ (Manning et al., 2006), a potentially important
driver of efficiency.
The mode of production issue leads to a central theme in the efficiency literature: public versus
private ownership. The claim of the literature is that publicly owned organizations will be less efficient
than privately owned ones (Boardman and Vining, 1989). This is attributed to an inherently different
incentive structure that exists for public managers compared to private managers. Wilson (1989) argues
that “whereas business management focuses on the 'bottom line' (that is, profits), government management
focuses on the 'top line' (that is, constraints)”. Public managers, and, to a lesser extent, non-profit
managers, have to cope with three constraints that private managers do not have. They are not able to
CONFIDENTIAL GOV/PGC(2007)16/ANN
15
allocate the production factors in accordance with the managers’ preferences; they cannot lawfully retain
and devote the earnings of the organisations to the private benefit of their members; and they must serve
goals that are set by principals outside the organisation.
Empirical evidence
Evidence on the impact of ownership on efficiency is inconclusive.4 Some studies found that public-
owned institutions are more efficient than privately owned ones. Hollingsworth et al. (1999) reviewed 35
studies of hospitals mostly in the United States, but also in Europe. They conclude that public hospitals are
more efficient than private hospitals. Mobley and Magnusson (1998), who did a comparative study of
Norwegian and Californian hospitals, explain this higher efficiency by the beneficial effects of regulation.
Through regulation, Norwegian hospitals had better capital utilisation compared to Californian hospitals.
Hammerschmidt and Staat (2000) found that non-profit hospitals in Germany are less efficient than public
hospitals.
However, other studies conclude the opposite. In education, non-profit owned schools seem to
outperform both public and private schools. Dronkers and Robert (2004) assessed the effectiveness of
public and private schools based on the PISA data. After controlling for student intake, social background
of parents and social composition, they concluded that private, government-dependent schools have a
higher effectiveness than purely public and purely private schools.
Finally, there is evidence from the refuse collection sector showing that contracting with private
companies is more efficient than public provision. Savas (2000) reviews studies from the United States,
Canada and the United Kingdom and finds that private contracting is 35% cheaper than municipal
collection without any loss in quality of services. In addition, equity of service delivery is not affected
since local governments still pay for the service. He finds comparable evidence from bus services. Public
services were found more costly than private bus services.
4.3.2. Competition
Theoretical proposition
The dominant theme of public management reform in the 1980s and 1990s was not ownership, but
competition. The pressure of competition in public organisations is expected to induce efficiency gains in
their operations. Savas (2000) argues that the issue is not public versus private ownership, but monopoly
versus competition in service provision. Pollitt and Bouckaert (2004) demonstrate that switching to private
ownership (privatization) is only one of the strategies for injecting competition into public service delivery.
Competition can be introduced within government as well. Internal (quasi) markets are an example of this
strategy. At the same time, a shift from public to private ownership is not necessarily a shift towards
competition. The main hypothesis is that increased competition leads to efficiency improvements.
Empirical evidence
The empirical evidence for the impact of competition hypothesis is also inconclusive. Boyne (2003)
reviewed 18 studies on the impact of the market structure on efficiency and found positive, negative and
insignificant results. Most studies on competition assess the relationship between the degree of competition
and efficiency. Seldom is a competitive case compared with a non-competitive case. The studies also
struggle with the operationalisation of competition in the public sector. For public schools, the number of
children in private schools is often used as an indicator of competitive pressure (Arum, 1996; Duncombe et
al., 1997; Dee, 1998). It can be argued, however, that children in private schools are another segment of
the education market for which public schools never intended to produce.
CONFIDENTIAL GOV/PGC(2007)16/ANN
16
The divergence of the findings of the empirical studies points to the conditional nature of the potential
benefits of competition. Low asset specificity and low information costs are in this respect requirements
for competition (Williamson, 1975). An asset is specific if it makes a necessary contribution to the
production of a good and it has much lower value in alternative uses. Information costs are high when it is
hard to assess the performance level of the service. One can observe immediately when refuse is not
collected properly; the streets are littered. However, the benefits of education will only show in the longer
term, i.e. when students are entering the labour market.
In a sector with high asset specificity and high information costs, such as hospitals, the introduction of
competition will encounter problems. Andersen and Blegvad (2006) describe how competition in the
Danish dental care system for children (which is public) is nullified by the professional self-regulation by
the dentist profession. Yet, in a sector with relatively low asset specificity and information cost, such as
refuse collection and bus services, it is a lot easier to exploit the benefits of competition.
4.4. Workforce issues
4.4.1. Workforce size
Theoretical proposition
There are two, contradictory hypotheses regarding the impact of workforce size on efficiency.
According to the first hypothesis, larger organisations are more efficient than smaller ones. This effect is
attributed to economies of scale which result from savings in overhead costs and fixed costs in tangible
assets (Scherer and Ross, 1990). The reverse hypothesis states that size leads to diseconomies of scale
through excessive rigidity or conservatism/resistance to change. Slack and bureaucratic behaviour can be
found in all large organisations. Specifically for the public sector, public choice theorists such as Downs
(1967) and Niskanen (1971) argued that bureaucrats strive towards budget maximization, even though
these resources are not required for achieving the organisational objectives. In a more moderate approach
Dunleavy (1991) argues that bureaucrats are primarily seeking to improve the prestige of their office
(bureau shaping). The second hypothesis asserts that for efficiency and effectiveness “small is beautiful”,
as smaller organisations are more easily monitored and are more responsive to principals and citizens.
Empirical evidence
Boyne (2003) concludes from a reading of 26 studies that there is a positive effect of resources on
service performance. These studies are mostly about the impact of workforce size on output and outcome
levels, and not about the impact of workforce size on the input levels relative to the output levels, which is
the definition of efficiency. Most of the studies are thus not directly related to technical efficiency in the
strict sense. These studies, however, refute the radical public choice arguments that bureaucratic
dysfunctions will dissolve most of the benefits of growth in public sector spending.
If we focus on efficiency in the strict sense (input versus outputs relative to a production frontier),
several studies seem to suggest that increasing the size of organisations does have a positive influence on
efficiency. Ruggiero and Duncombe (1995) found that larger schools are more technically efficient.
Similar conclusions were drawn from studies of the Belgian public sector for secondary schools (de
Brabander and Vos, 1992), day care centres for children (Donni, 1993) and refuse collection (Lawarree,
1986). McCallion et al. (1999) analysed the efficiency of the Irish hospital services and concluded that the
Irish strategy of concentrating hospital services in six large hospitals supported by medium-sized hospitals
is appropriate.
The empirical evidence seems to support the hypothesis that efficiency gains can be obtained by
increasing the scale of operations. However, the potential losses on other public service values (equity,
CONFIDENTIAL GOV/PGC(2007)16/ANN
17
access to services, robustness) as well as on quality need to be taken into account when adopting these type
of strategies for efficiency improvements.
4.4.2. Workforce composition
Theoretical proposition
Two streams of research theorize on the impact of workforce composition on performance
considering the impact of representative bureaucracy and diversity of the workforce (Pitt, 2005).
Representative bureaucracy attempts to match the population on indicators such as ethnicity, linguistic
group, or gender (Selden, 1997). The central assumption is that because all groups in the population are
represented in the bureaucracy, policy formulation and implementation will benefit the interests of diverse
groups. Representation then would lead to better results for all groups in society.
The second stream of research focuses on the effects of diversity. The argument is that members of a
diverse workforce will draw from the differences in the group and take the best of each others’ background
(Adler, 1997) thereby producing better results. However, other scholars argue the opposite. They assert
that individuals look for sameness. The more one’s self is reflected in other group members, the more one
will trust others and value their work (Pitt, 2005).
Empirical evidence
Little work has considered the impact of representative bureaucracy on performance. Rather, studies
examine whether a particular group benefits from its representation. An exception is the study of Andrews
et al. (2005). They found that ethnic diversity is negatively associated with consumer perceptions of the
performance of English local governments. However, they also found that representative bureaucracy is
not negatively related to audited performance, which is a better proxy of objective performance than
subjective perceptions.
There is not much research on the impact of diversity on workforce performance either. A public
administration study is offered by Brewer (2005), who asserts from a survey of front line supervisors that
high performing agencies tend to strive towards workforce diversity. Pitt (2005) however, found a negative
relationship between teacher diversity and two performance indicators of student achievement, but he also
found a positive relationship between diversity and a third indicator of student achievement. There was no
relationship between manager diversity and performance. This study proves again the importance of the
choice of the proxy for outputs or outcomes. Different indicators may lead to substantially different results.
In social psychology, more studies on the effects of diversity are conducted. Yet, the findings are
inconclusive too (Pitt, 2005). Some studies point to higher creativity and implementation ability in diverse
organisations. Other studies find no link between diversity and performance. Furthermore, there are studies
that find negative effects of diversity, such as increased absenteeism. Baugh and Graen (1997) found that
ethnically diverse teams are more likely to receive negative evaluations than homogeneous teams.
Almost no studies exist on the impact of workforce diversity and representation on efficiency.
Representation and diversity are typical examples of values which public employers attach importance to
because of their social, non-monetary effects. For example, representation of different linguistic groups in
government will have a negative impact on efficiency. Translation and communication costs are
necessarily higher. Yet, the outcomes of such a linguistic policy in terms of legitimacy and trust of citizens
in government may far outweigh these inefficiencies.
CONFIDENTIAL GOV/PGC(2007)16/ANN
18
4.4.3. Extent and nature of unionisation
Theoretical proposition
Unions play an important role in the labour relations in the public sector. Public sector union density
surmounts private sector union density with approximately 30% in most developed countries (Visser,
2000). Unionisation in German public sector was 56%, compared to 22% in the private sector (1997). In
the United States, 37% of public employees were members of a union compared to 9% of private sector
employees (1999). It should also be mentioned that overall union membership is on the decline in most
developed countries with the exception of the Nordic countries and Belgium. In some countries, this
decline can be attributed solely to private sector. In Austria, Canada, Norway, Spain, and the United States,
public sector unionisation remains constant or has increased between the 1980s and the end of the 1990s
(Visser, 2000).
In one view, trade unions can be expected to have a positive impact on efficiency. The concept of the
“collective voice” expresses the underlying mechanism (Freeman and Medoff, 1984). Unions are expected
to improve productivity by various means, such as reducing staff turnover, increasing job security, and
reducing grievances. They can be a vehicle for reconciling employers’ and employees’ interests. Also,
unions may play a constructive role during public sector reforms by expressing the often legitimate
concerns of the staff, and pointing to shortcomings or inconsistencies.
Craft (2003) predicts that management – union cooperation in the United States will be reinforced,
since they are increasingly interdependent. Both managers and unions will continue to be subjected to
pressure from the public and legislators to cut costs and improve service delivery. From this perspective,
union density would be a facilitating factor for change and productivity, provided unions and managers
succeed in establishing a constructive relationship.
In the second view, high unionization is associated with negative performance (Menezes-Filho and
Van Reehnen, 2003). Firstly, unionization is believed to lead to higher compensation and thus a relative
inefficient cost structure. Secondly, unions may lower performance through featherbedding, striving for
more personnel than strictly necessary. Thirdly, it is argued that union density can lead to more antagonism
in the workplace.
Empirical evidence
Research on the impact of unions on public sector performance is scant compared to research on the
private sector. There is some empirical evidence from local school districts and fire services in the United
States that suggests that high levels of unionization constrain both flexibility and productivity
(Donahue et al., 2000). Methé and Perry (1980) found that collective bargaining in local government in the
United States led to increased municipal expenditures. They add, however, that the impact on issues of
efficiency and effectiveness is unclear.
The scarcity of research on public sector unions is all the more remarkable because, in all probability,
the role of unions in the public sector will differ from the private sector substantially as public sector
unions are more prominent; bargaining is not strictly a managerial, but also a political affair and many of
the public services are considered essential. In addition, the lesser ability to measure public sector output
may shift the focus of unions towards the definition of output. Craft (2003) even argues that in the next two
decades, union membership in the public sector in the United States will increase mainly as a defence
against the on-going pressure on the public sector for cost-saving and efficiency gains.
CONFIDENTIAL GOV/PGC(2007)16/ANN
19
4.4.4. Attractiveness of the public sector
Theoretical proposition
A competent and highly motivated workforce is essential for government performance. The central
question is how to recruit and retain high-skilled staff. Due to ageing, many countries will be forced to hire
a high number of new employees to replace current employees as they reach retirement age. How can the
public sector compete with the private sector and attract and retain well-educated personnel?
Many scholars have examined the determinants of job satisfaction. Individual factors such as gender
or age hardly determine job satisfaction, but the job context appears to be very important. Herzberg (1966)
identifies intrinsic factors, such as job content, as “satisfiers” and extrinsic factors, such as remuneration,
as “dissatisfiers”. This means that job content can make a real difference to job satisfaction, while adequate
wages are considered as a basic condition that do not act as a motivator, but acts as a demotivator in case
of perceived inadequate wage levels. Hackman and Oldham (1975; 1980) developed a model of job
satisfaction in which not only objective job characteristics, but also the needs and expectations of the
individual employee play an important role.
The basic assumption is that the public sector should pay attractive salaries in order to recruit and
retain high-skilled staff. Another proposition is that a high status and/or prestige accorded to civil servants
can increase the attractiveness of the public sector. Moreover, the introduction of performance-related pay
systems could provide additional incentives for people to choose the public service as employer. Another
issue is whether there exists a specific motivation to serve the public interest opposite to the typical self-
interest that is assumed in rational choice theories
Empirical evidence
There is an extensive literature on wage differences between public sector and otherwise comparable
private sector workers covering many OECD member countries (Gregory and Borland, 1999; Lee, 2004;
Melly, 2002; Mueller, 1998; Postel-Vinay and Turon, 2005). In most countries there exists a substantial
public sector premium (Gregory and Borland, 1999) although its size varies over time and across countries.
At the same time significant differences are found in the differential by various worker characteristics,
such as gender and occupation. For example, in Germany wages for men were lower in the public sector
than in the private sector, but the opposite was found for women. In addition, the public sector wage
premium was the highest at the lower end of the wage distribution and decreased monotically as one
moved up the wage distribution (Melly, 2002) While higher public sector wages could help public sector
organisations to attract and retain good quality workers, it could also lead to a higher costs and
consequently inefficiencies in public service provision if those wages are not matched by correspondingly
higher job performance by public sector workers.
The OECD has studied how 11 OECD member countries cope with the issue of recruiting and
retaining highly skilled civil servants. Canada, Denmark, Finland, and Sweden reported recruiting
problems, while Austria, Germany, and Norway said that they soon expect to face this problem. Canada,
Italy, Korea, Poland, Portugal, and Spain stated that they suffer from critical skills shortages. There are
especially shortages of certain groups of professionals, such as IT specialists. Difficulties in retaining staff
emerge especially in the 20-34 age group. Wages have proven to be a crucial factor in the retention of staff,
especially after two-to-four years of service (Aijälä, 2001). Some governments have adjusted their wages
in order to better reflect market conditions (e.g. in Finland and Ireland). Sustained budget constraints may
also affect the attractiveness of the public sector. Schick (1996) has argued that sustained budgetary
constraints employed by successive New Zealand governments led to serious retention and attraction
problems.
CONFIDENTIAL GOV/PGC(2007)16/ANN
20
Another problem may be the image of the public sector. In the United States, the sustained assaults on
the public service encourage many of the best and brightest students to pursue careers in the private sector.
The denigration of the public sector and public servants can produce a self-fulfilling prophecy that drives
out the most able (Holzer & Rabin, 1987).
There is also a need for governments to identify incentives other than salary if they want to be able to
recruit and retain high quality employees. The OECD cited Germany as a positive example, since it has
implemented flexible working hours, flat hierarchies and the use of the latest technical equipment. Other
non-monetary incentives cited in the report include: co-operative leadership, open communication;
sufficient freedom to display initiative and make decisions; good working conditions; good opportunities
for training and personal development; telecommuting; family-friendly personnel policies; job rotation;
assignments to private sector companies, other public services or international organisations; and
opportunities for educational leave or leave for other personal reasons (Reichenberg, 2002).
5. Conclusion
Overall, the evidence on the institutional drivers of efficiency in the public sector is surprisingly
limited. Available research provides a very limited assessment of the impact on efficiency of varying the
mix of inputs used or of changing structural and managerial arrangements.
Summarising the empirical evidence reviewed, relatively strong findings emerged in three areas.
First, it seems that efficiency gains could be obtained by increasing the scale of operations, based on
evidence collected mainly in the education and health sectors. However, the impact on other public service
values, such as equity, access to services and quality of services needs to be considered when adopting
these kinds of strategies for efficiency improvements. Second, functional and political decentralization to
subnational governments also seems beneficial for efficiency.
Third, human resource management practices also matter a great deal. The soft aspects of human
resource management such as employee satisfaction and morale are the most important drivers of
performance. While wages are still important for staff, non-monetary incentives are essential. High wage
levels – compared to similar work in the private sector – could lead to inefficiencies, although governments
often are model employers and their wage policies reflect equity concerns as well. Wages are also
important for attracting and retaining qualified staff, especially in case of skill-shortages.
Findings are inconclusive on the impact of ownership, competition and agencification. While private
ownership is not a guarantee for efficiency, public ownership does not necessarily lead to inefficiencies
either. It is more probable that it is not ownership but competition that drives efficiency. However, there is
a need to further explore for what and with whom public organisations compete. The nature of service
delivery, such as low asset specificity and low information costs, is crucial for successful competition in
public services.
Regarding agencification there is some evidence that a reduction of input controls combined with
steering for results, financial incentives and competition could lead to increased efficiency. However, the
impact on the quality of service delivery and policy effectiveness is unclear. The literature also calls
attention to the major risks of agencification, including the exposure of government to financial and
employment risks and opportunities for political patronage and corruption.
There is very little evidence on impact of workforce diversity and representativeness on efficiency.
Relatively uncharted territory is the assessment of union’s role in public sector efficiency as well, although
union representation is relatively high in the public sector in most countries. The effects of new intra-
CONFIDENTIAL GOV/PGC(2007)16/ANN
21
governmental co-ordination mechanisms are also not known. Surprisingly, the impact of e-government has
not been thoroughly evaluated either by researchers.
There is empirical evidence about the unintended negative effects of performance measurement/
management. However, it is related to a very specific type of performance management. The evidence
mostly comes from England and from those applications with a very tight coupling between measurement
and decisions. More positive effects are found in performance arrangements with a loose coupling between
measurement and decisions. These performance arrangements inform rather than punish professionals in
the public sector. The evidence on the use of performance information in the political arena is diffuse. .
Their major impact is situated in the internal management of departments and agencies.
In summary, while there have been a plethora of public sector reforms in many OECD member
countries the research evidence shows fewer success stories that have been claimed by practitioners. There
are several reasons for this contradiction. First, research in this area is extremely complicated due to data
availability issues, measurement difficulties, and the potential effect of many external factors on efficiency
and productivity (the attribution problem). Second, reforms are often driven by ideological considerations
and management fads rather than by efficiency concerns. Third, practitioners often have a vested interest in
the success of the reforms and may over-claim their impact. Finally, there could substantial differences
between short-term and long-term effects of these reforms, such as efficiency gains dissipating over time.
While this review of the empirical evidence provided an assessment of the current state of the
literature, it also helps selecting areas where further research might be fruitful in understanding the various
drivers of efficiency in public services. In this we should not rely solely on academic choices of research
topics but must also build on the experiences and needs of practitioners.
The lack of robust empirical evidence also points to the importance of collecting good quality
comparative data on a regular basis that could contribute to better understanding the impacts of
institutional arrangements on the efficiency of the public sector. The proposed publication “Government at
a Glance” intends to include information on all institutional variables described in this paper. This might
entail the analysis of existing information on new ways – e.g. expenditure and employment data – that
could be carried out in the near future, as well as the collection of new data by the OECD – e.g. on the
distribution of responsibilities across different levels of government; agencies – that will take place over
the longer term.
NOTES
1. For instance, an increase in an organisation’s labour productivity does not necessarily tell that labourers
work harder than before. It can be the result of an investment in capital too. Equally, a high labour
productivity in a country does not necessarily imply that labourers work harder than in other countries, as
is sometimes suggested in the media.
2. Wilson (1989) argues that dealing with these constraints is the main task of public sector managers. He
states that whereas business management focuses on the 'bottom line' (that is, profits), government
management focuses on the “top line” (that is, constraints).
3. For example in Australia, Canada, Denmark, Finland, Iceland, the Netherlands, New Zealand, Norway and
Sweden.
CONFIDENTIAL GOV/PGC(2007)16/ANN
22
4. The reason of the inconclusiveness of the findings may be the robustness of Data Envelopment Analysis
(DEA) – the most commonly used research method. DEA calculates a production frontier base on an input
and output mix. Changes in the input and output mix may substantially alter results. In other words, the
results depend on which inputs and outputs are selected to be included in the analysis.
BIBLIOGRAPHY
Adler, N. J. (1997). International Dimensions of Organizational Behavior. Cincinnati, OH: South-Western
Press.
Aijälä, K., (2001). Public Sector – An Employer of Choice, Paris: OECD.
Andersen, L.B. and M. Blegvad (2006). ”Does Ownership Matter for the Delivery of Professionalized
Public Services? Cost Efficiency and Effectiveness in Private and Public Dental Care for Children in
Denmark”. Public Administration. 84(1): 147-164.
Andrews, R., G.A. Boyne, K.J. Meier, L.J. O’Toole, Jr. and R.W. Walker (2005). “Representative
Bureaucracy, Organizational Strategy, and Public Service Performance: An Empirical Analysis of
English Local Government”. Journal of Public Administration Research and Theory 15(4): 489-504.
Arum, R. (1996). “Do Private Schools Force Public Schools to Compete?” American Sociological Review
61:29-46.
Askim, Jostein (2007). Determinants of Performance Information Utilization in Political Decision Making.
In: Van Dooren, Wouter and Van De Walle, Steven (2007). Utilising performance information:
bureaucracy, politics and society. Palgrave McMillan, London.
Barankay, Iwan and Lockwood, Ben, (2006) Decentralization and the Productive Efficiency of
Government: Evidence from Swiss Cantons (December 2006). IZA Discussion Paper No. 2477.
Baugh, Gayle, and George Graen. (1997). “Effects of Team Gender and Racial Composition on
Perceptions of Team Performance in Cross-functional Teams”. Group and Organizational
Management 22 (3): 366–84.
Bevan, G. and Hood, C. (2005). What’s Measured Is What Matters: Targets and Gaming in the English
Public Health Care System. The Public Services Programme. London: Economic and Social
Research.
Blöndal, J. (2003). “Accrual Accounting and Budgeting: Key Issues and Recent Developments. OECD
Journal on Budgeting. 3(1): 43-59.
Boardman, A. E., and A.R. Vining. (1989). “Ownership and Performance in Competitive Environments.”
Journal of Law and Economics 32 (April): 1-34.
CONFIDENTIAL GOV/PGC(2007)16/ANN
23
Bouckaert, G. and W. Balk (1991). “Public Productivity Measurement: Diseases and Cures”. Public
Productivity & Management Review, 15:2. pp. 229-235.
Bouckaert, G. and B.G. Peters (2002). “Performance Measurement and Management. The Achilles' Heel in
Administrative Modernization”. Public Performance and Management Review. 25:4. pp. 359-362.
Boyne, G. (2003). “Sources of Public Service Improvement: A Critical Review and Research Agenda.”
Journal of Public Administration Research and Theory 13 (2): 767-94.
Brennan, G. and S. Buchanan (1980). The Power to Tax Analytical Foundations of a Fiscal Constitution,
Cambridge University Press.
Brewer, G. A. (2005) “In the Eye of the Storm: Frontline Supervisors and Federal Agency Performance.”
Journal of Public Administration Research and Theory 15: 505-527.
Brewer, G. A. and S. Selden (2000). “Why Elephants Gallop: Assessing and Predicting Organizational
Performance in Federal Agencies.” Journal of Public Administration Research and Theory. 10: 685-
711.
Beuselinck, E. (2007). The Concept of Coordination: A Literature Review. Unpublished manuscript.
Katholieke Universiteit Leven, Belgium.
Carlin, T. (2006). “Victoria’s Accrual Output Based Budgeting System – Delivering as Promised? Some
Empirical Evidence.” Financial Accountability & Management, 22(1), 1-19.
Carlin, T.and J. Guthrie (2000). “A Review of Australian and New Zealand Experiences with Accrual
Output Based Budgeting.” Paper presented at the third bi-annual conference of the International
Public Management Network. Sydney Australia, 4-6 March.
Coelli, T., D.S. Prasada Rao and G.E. Battese (2001). An Introduction to Efficiency and Productivity
Analysis. Kluwer Academic Publishers.
Craft, J.A. (2003). “Future Directions in Public Sector Labor Relations: A 2020 Perspective.” Journal of
Labor Research,14(4). pp. 545-560.
Curristine, T. (2005). “Government Performance: Lessons and Challenges”, OECD Journal on
Budgeting, Volume 5, No. 1.
Daley, D.M. (1992). “Pay for Performance, Performance Appraisal, and Total Quality Management.”
Public Productivity & Management Review, 16(1). pp. 39-51.
Dawson, D., M. Goddard and A. Street (2001). “Improving Performance in Public Hospitals: A Role for
Comparative Costs?” Health Policy, 57: 235-48.
De Brabander, K. and M. en Vos (1992). Kostprijsanalyse van scholen in het lager en secundair
onderwijs: vrij gesubsidieerd ondervijs, (HIVA, Leuven).
Dee, T. (1998). “Competition and the Quality of. Public Schools”. Economics of Education Review,
vol. 17, pp. 419-27.
CONFIDENTIAL GOV/PGC(2007)16/ANN
24
Donahue, A. K., S.C. Selden and P.W. Ingraham (2000). “Measuring Government Management Capacity:
A Comparative Analysis of Human Resources Management Systems.” Journal of Public
Administration Research and Theory 10 (2): 381-411.
Donni, O. (1993). Efficiency of Day Care Centers in Belgium. Université de Liège, CREPP, Liege.
Downs, A. (1967). Inside Bureaucracy. Glenview, Ill.: Scott, Foresman and Co.
Downs, G (1986). The Search for Government Efficiency: From Hubris to Helplessness. Temple
University Press, Philadelphia.
Dronkers, J. and P. Robert (2004). The Effectiveness of Public, Private Government Dependent and Private
Independent Schools: A Cross-national Analysis. Florence: European University Institute.
Duncombe, W., J. Miner and J. Ruggiero (1997). “Empirical Evaluation of Bureaucratic Models of
Inefficiency.” Public Choice. 93 (1). 1-18.
Dunleavy, P. (1991). Democracy, Bureaucracy and Public Choice. Hemel Hempstead: Harvester
Wheatsheaf.
Ebel, R.D. S. and Yilmaz (2002). On the Measurement and Impact of fiscal Decentralization.
Washington, D.C.: The World Bank.
Feld, L., G. Kirchgässner and C.A. Schaltegger (2003). Decentralized Taxation and the Size of
Government: Evidence from the Swiss State and Local Governments. CESifo Working Paper
No 1087.
Freeman, R., and J. Medoff (1984). What Do Unions Do? New York: Basic Books.
Gregory, R.G. and J. Borland (1999). “Recent Developments in Public Sector Labour Markets”, in
Handbook and Labour Economics, Vol. 3C, edited by O.Ashenfelter and D. Card pp. 3573-630.
Hackman, J. and G. Oldham (1980). Work Redesign. Reading, MA: Addison-Wesley.
Halachmi, A. and M. Holzer (1987). “Merit Pay, Performance Targeting, and Productivity.” Review of
Public Personnel Administration, 7(8). pp. 80-91.
Halachmi, Arie (2002). “Performance Measurement: A Look at Some Possible Dysfunctions”. Work Study
Volume 51 Number 5.
Hammerschmidt, M. and M. Staat (2000). Benchmarking the Health Sector in Germany – An Application
of Data Envelopment Analysis .Mannheim; Universität Mannheim.
Herzberg, F. (1966). Work and the Nature of Man. New York: The World Publishing Co.
Hollingsworth, B, P. Dawson and N. Maniadakis (1999). “Efficiency Measurement of Health Care: A
Review of Non-parametric Methods and Applications.” Health Care Management Science, 2:3,
pp. 161-172(12).
Holzer, M. and J. Rabin (1987). “Public Service: Problems, Professionalism and Policy
Recommendations.” Public Productivity Review. 11(1): 3-13.
CONFIDENTIAL GOV/PGC(2007)16/ANN
25
Huther, J., A. Shah (1998). Applying a Simple Measure of Good Governance to the Debate on Fiscal
Decentralization. World Bank Working Paper 1894.
Ingraham, P.W. (1993). “Of pigs in Pokes and Policy Diffusion: Another Look at Pay-for Performance.”
Public Administration Review. 53(4). pp. 348-356.
Jordan, M. and M. Hackbart (1999). “Performance Budgeting and Performance Funding in the States: A
Status Assessment.” Public Budgeting & Finance, 19(1), 68-88.
Joumard, I., P.M. Kongsrud, Y. Nam and R. Price (2004). “Enhancing the Cost Effectiveness of Public
Spending: Experience in OECD Countries.” OECD Economic Studies No. 37, 2003/2. Paris: OECD.
Laking, R. (2005). “Agencies: Their Benefits and Risks.” OECD Journal on Budgeting. 4(4): 7-25.
Lawarree, J. (1986). « Une comparaison des performances des secteurs privé et public: le cas des collectes
d’immondices en Belgique ». Cahiers Économiques de Bruxelles. 25,109,3-31.
Lewin, D. (2003). “Incentive Compensation in The Public Sector: Evidence and Potential.” Journal of
Labor Research, 14(4). pp. 597-619.
Manning, Nick, D.J. Kraan and J. Malinska (2006). “How and Why Should Government Activity Be
Measured in Government at a Glance?”. OECD: Paris.
McCallion, G., D.G. McKillop,J.C. Glass and C. Kerr (1999). “Rationalizing Northern Ireland Hospital
Services Towards Larger Providers: Best-Practice Efficiency Studies and Current Policy.” Public
Money and Management 19 (2), 27-32.
Melkers and Willoughby (2005). “Models of Performance Measurement Use in Local Governments:
Understanding Budgeting, Communication and Lasting Effects.” Public Administration Review,
65(2), 180-190.
Melly, B. (2005). “Public-Private Sector Wage Differentials in Germany: Evidence from Quantile
Regression.” Empirical Economics, 505-520.
Meneezes-Filho, N, and J. Van Reenen (2003). Unions and Innovation: A Survey of the Theory and
Empirical Evidence. Discussion paper No 3792. London: Centre for Economic Policy Research.
Methé, D.T. and J.L. Perry (1980). “The Impacts of Collective Bargaining on Local Government Services:
A Review of Research.” Public Administration Review, 40(4). pp. 359-371.
Mobley, L.R. and J. Magnussen (1998). “An International Comparison of Hospital Efficiency: Does
Institutional Environment Matter?” Applied Economics. 30. 1089-1100.
Moon, M.J. (2002). “The Evolution of E-government Among Municipalities: Rhetoric or Reality?” Public
Administration Review. 62(4): 424-433.
Moynihan, D.P.and S.K. Pandey (2005). “Testing How Management Matters in an Era of Government by
Performance Management.” Journal of Public Administration Research and Theory. 15: 421-439.
Mueller, R.E. (1998). “Public –Private Sector Wage Differentials in Canada: Evidence from Quantile
Regressions.” Economic Letters, Elsevuir, 60(2): 229-235.
CONFIDENTIAL GOV/PGC(2007)16/ANN
26
Murray, R. (1992). “Measuring Public Sector Output: The Swedish Report.” In: Griliches, Z. Output
Measurement in the Service Sectors. Chicago: The University of Chicago Press.
Niskanen, W. (1971) Bureaucracy and Representative Government. Chicago: Aldine-Atherton.
OECD (1997) Managing Across Levels of Government. Paris: OECD.
OECD (2001) Measuring Productivity – OECD Manual: Measurement of Aggregate and Industry-level
Productivity Growth, OECD: Paris.
OECD (2003). “The Case for E-government: Excerpts from the OECD Report ‘The E-government
Imperative’”. OECD Journal on Budgeting. 3(1): 61-92.
OECD (2005a), Modernising Government: The Way Forward, OECD: Paris.
OECD (2005b). “Paying for Performance: Policies for Government Employees.” OECD Policy Brief.
Osborne, D and T. Gaebler (1992). Reinventing Government: How the Entrepreneurial Spirit is
Transforming the Public Sector. New York: Addison-Wesley.
Paulsson, G. (2006). “Accrual Accounting in the Public Sector: Experiences from the Central Government
in Sweden.” Financial Accountability & Management, 22(1), 47-62.
Peters, B. G. (2003). “The Capacity to Coordinate”. Paper presented at the workshop on Policy Capacity
in October 2003, Hong Kong, 39.
Pierre, J. (2004). “Central Agencies in Sweden: A Report from Utopia”. In Pollitt, C. &C. Talbot (2003), A
Critical Analysis of the Global Trend to Agencies, Quangos and Contractualisation, Routledge.
Pitt, D. W. (2005). “Diversity, Representation, and Performance: Evidence about Race and Ethnicity in
Public Organizations.” Journal of Public Administration Research and Theory. 15: 615-631.
Pollitt, C. and G. Bouckaert (2004). Public Management Reform. A Comparative Analysis. Oxford: Oxford
University Press.
Pollitt, C. (2004). “Theoretical Overview.” In Pollitt, C. & C. Talbot Unbundled Government. A Critical
Analysis of the Global Trend to Agencies, Quangos and Contractualisation. Routledge Studies in
Public Management. 1. London and New York: Routledge (pp. 319-342).
Postel-Vinay, F.and H. Torun (2005). The Public Pay Gap in Britain: Small Difference that (don’t) Matter.
Centre for Market and Public Organisation: University of Bristol no. 05/121.
Reichenberg, N. (2002). “Branding the Government as an Employer of Choice”. International Recruitment
and Selction Strategies. Washington, D.C.: IPMA-HR.
Robinson, M. & J. Brumby (2005). Does Performance Budgeting Work? An Analytical Review of the
Empirical Literature. IMF Working paper. WP05/210.
Rubin, I. and J. Kelly (2005). “Budget and Accounting Reforms.” In E. Ferlie, L.E. Lynn Jr. & C. Pollitt
(Eds.), The Oxford Handbook of Public Management (pp. 639-590). Oxford: Oxford University
Press.
CONFIDENTIAL GOV/PGC(2007)16/ANN
27
Ruggiero, J. and W. Duncombe (1995). “On the Measurement and Causes of Technical Inefficiency in
Local Public Services.” Journal of Public Administration Research and Theory. 5:4. 403-429.
Savas, E. S. (2000). Privatization and Public Private Partnerships. London: Chatham House Publishers.
Scherer, F. M. and D. Ross. (1990). Industrial Market Structure and Economic Performance. 3rd ed.
Boston: Houghton Mifflin.
Schick, Allen (2003). “The Performing State: Reflection on an Idea Whose Time Has Come but
Whose Implementation Has Not”. OECD Journal on Budgeting Vol. 2, No. 2, OECD, Paris.
Selden, S. (1997) The Promise of Representative Bureaucracy: Diversity and Responsiveness in a
Government Agency. Armonk, NY, M.E. Sharpe.
Selden, S.C., P.W. Ingraham and W. Jacobson (2001). “Human Resource Practices in State Government.
Findings from a National Survey.” Public Administration Review. September/October. Vol. 61, n° 5.
Shepard, L.A. (1990). “Inflated Test Score Gains: Is the Problem Old Norms or Teaching the Test.”
Educational Measurement. Issues and Practices. 9, 15-22.
Smith, M. L. and C. Rottenberg (1991). “Unintended Consequences of External Testing in Elementary
Schools.” Educational Measurement: Issues and Practice, 10(4), 7-11.
Smith, P. (1995). “On the Unintended Consequences of Publishing Performance Data in the Public
Sector.” International Journal of Public Administration, 18. pp.277-310.
Snijkers, K. (2005).” E-government. ICT from a Public Management Perspective.” Paper presented at the
13th annual NISPAAcee Conference, 19-21 May 2005, Moscow State University.
Social and Cultural Planning Office (2004). Public Sector Performance. An International Comparison.
Den Haag: SCP.
Sterck, M. and B. Scheers (2006). “Trends in Performance Budgeting in Seven OECD Countries.” Public
Performance & Management Review, Vol. 30, No. 1, September, pp. 47-72.
U.K. House of Commons, Public Administration Select Committee (2003). On Target Government by
Measurement. Fifth report of Session 2002-2003 (vol1). London: Stationery Office.
Van Dooren, W. (2006) Performance Measurement in the Flemish Public Sector: A Supply and Demand
Approach. Leuven: University of Leuven. Faculty of Social Sciences:
Van Dooren, W. et al. (2006) “Issues in Output Measurement”, OECD: Paris.
Verhoest Koen (2005). “Effects of Autonomy, Performance Contracting and Competition on the
Performance of a Public Agency.” Policy Studies Journal. 33(2).
Verhoest Koen and Geert Bouckaert (2003). Machinery of Government: Shifting from Effectiveness of
Coordination to Coordination of Effectiveness. International Workshop on State Policy Capacity,
Hong Kong. October. p. 32.
Visser, J. (2000). Trends in Unionisation and Collective Bargaining. Geneva: International Labour Office.
CONFIDENTIAL GOV/PGC(2007)16/ANN
28
West, D.M. (2004).” E-government and the Transformation of Service Delivery and Citizen Attitudes.”
Public Administration Review. 64(1): 15-27.
Williamson, O. E. (1975). Markets and Hierarchies: Analysis and Antitrust Implications. Free Press, New
York.
Wilson, J.Q. (1989). Bureaucracy: What Government Agencies Do and Why They Do It. New York: Basic
Books.
Zigarelli (1996). “An Empirical Test of Conclusions form Effective Schools Research.” The Journal of
Educational Research. 90: 103-111.
CONFIDENTIAL GOV/PGC(2007)16/ANN
29
ANNEX I:
MEMBERSHIP OF STEERING GROUP
COUNTRY NAME TITLE/POSITION MINISTRY
AUSTRIA Elisabeth DEARING Director Federal Chancellery
Administrative Development, Div
III/7
CANADA Lee MCCORMACK Executive Director Treasury Board of Canada
FINLAND Heikki JOUSTIE Government
Counsellor Ministry of Finance
Public Management Department
FRANCE Xavier HURSTEL Sous-Directeur Ministère de l’Économie et des
Finances
Direction du Budget
FRANCE Bernard BLANC Chef de la mission des
relations
internationales
Ministère de l’Économie et des
Finances
GERMANY Marga PROEHL Deputy Head of Public
Administration
Management Unit
Federal Ministry of the Interior
(BMI)
ITALY Pia MARCONI Director General Department of Public
Administration
NETHERLANDS Koos ROEST Adviser on Strategic
Policy Ministry of the Interior &
Kingdom Relations
Directorate-General for
Management of the Public Sector
NETHERLANDS Peter VAN DER GAAST Head of International
Civil Service Affairs
Division
Ministry of the Interior &
Kingdom Relations
Directorate-General for
Management of the Public Sector
NORWAY Morten STROMGREN Senior Advisor,
Economic Analysis
Unit
Ministry of Government
Administration & Reform
SWEDEN Richard MURRAY Chief Economist Statskontoret
UNITED KINGDO
M Alison SHARP Data Analyst & Data
Manager Cabinet Office, Prime Minister’s
Strategy Unit
CONFIDENTIAL GOV/PGC(2007)16/ANN
30
ANNEX II:
DATA DESCRIPTION OF THE REVIEWED LITERATURE
Table A1.1 Cash and accrual in central/federal government whole-of-government financial statements and
budget appropriations (Blöndal, 2004)1
Use of accrual accounting
Whole-of-government financial statements Use of accrual budgeting
Budget appropriations
Australia Accruals Accruals
Austria Cash basis Cash basis
Belgium Cash basis Cash basis
Canada Accruals Cash basis and certain transactions on accruals
basis
Czech Republic Cash basis Cash basis
Denmark Cash basis and certain transactions on
accruals basis Cash basis and certain transactions on accruals
basis
Finland Accruals Accruals, except for capitalization and
depreciation of assets
France Cash basis and certain transactions on
accruals basis Cash basis
Germany Cash basis Cash basis
Greece Cash basis Cash basis and certain transactions on accruals
basis
Hungary Cash basis Cash basis
Iceland Accruals, except for capitalization and
depreciation of assets Accruals, except for capitalization and
depreciation of assets
Ireland Cash basis Cash basis
Japan Cash basis Cash basis
Korea Cash basis and certain transactions on
accruals basis Cash basis and certain transactions on accruals
basis
Luxembourg Cash basis Cash basis
Mexico Cash basis and certain transactions on
accruals basis Cash basis and certain transactions on accruals
basis
Netherlands Cash basis Cash basis
New Zealand Accruals Accruals
Norway Cash basis Cash basis
Poland Cash basis and certain transactions on
accruals basis Cash basis
Portugal Cash basis and certain transactions on
accruals basis Cash basis and certain transactions on accruals
basis
Slovak Republic Cash basis Cash basis
Spain Cash basis and certain transactions on
accruals basis Cash basis
Sweden Accruals Cash basis and certain transactions on accruals
basis
Switzerland Cash basis Cash basis
1 Countries are classified as full accrual irrespective of whether heritage and military assets, and non-
exchange revenue (taxes), are treated on cash basis. Countries are classified as full cash basis irrespective
of whether they have an obligations/commitments system in place.
CONFIDENTIAL GOV/PGC(2007)16/ANN
31
Turkey Cash basis Cash basis
United Kingdom Accruals Accruals
United States Accruals Cash basis and certain transactions on accruals
basis
CONFIDENTIAL GOV/PGC(2007)16/ANN
32
Table 1. Literature related to results orientation
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
Carlin 2006 Budget
practices and
procedures
Does the accrual output-
based budgeting system
result in the creation of
measurably improved
performance-related
information and thus
facilitate continuous
improvement in a way
that traditional public
financial management
systems are alleged to
be less capable of?
n.a. Victorian
Department of
Treasury and
Finance
Public sector
Time series
analysis Australia
(state of
Victoria)
Eight
departments
of the
government
in the state
of Victoria
(Australia)
1998-
2002 a. Without adequate historical data reference points against
which to benchmark current performance, it is arguable that the
difficulty of management is compounded, not reduced in the way
that advocates of AOBB systems have typically claimed.
b. It is not at all clear that even if we were certain of the benefits
claimed to be associated with the use of AOBB technologies to
come to fruition, they would outweigh the costs of initial change
and subsequent operation.
c. Thus, rather than it adding credibility to claims relating to the
virtues of AOBB systems, the evidence set out in this paper
lends itself to the conclusion that accrual output based
budgeting, like so many financial management techniques
implemented within public sector entities before, may well
represent the latest passing public financial management
panacea.
Carlin,
Guthrie 2000 Budget
practices and
procedures
Does accrual output-
based budgeting system
help in the improvement
of public management
practices?
n.a. Finance and
Administration
Department of
Australia, New
Zealand
Treasury
Public sector
Comparative
country
analysis
Australia,
New Zealand 1995-
1997 Output Based Budgeting (OBB) Systems can provide welcome
assistance in the effort to improve public management practices,
but the authors caution that at present, as a result of the types of
implementation weaknesses they have identified, OBB systems
will not constitute the panacea suggested by some central
agency rhetoric.
Curristine 2005 Budget
practices and
procedures
a. What are the different
approaches by OECD
member countries to
enhance public sector
performance?
b. Which new levers and
approaches to
management,
budgeting, personnel
and institutional
structures did OECD
member countries adopt
in this perspective?
Theories on
performance-
based
governance
OECD/World
Bank Budget
Practices and
Procedures
Database
Survey
Public sector
Cross-
country
comparison
on the base
of self-
reported
surveys
27 OECD
member
countries
2003 a. The majority of OECD member countries are implementing
performance management and performance budgeting, although
the extent and the approaches vary widely across countries.
b. Most countries continue to struggle with changing the
behaviour of public servants and politicians.
c. The common assumption that the performance information
that is useful for the executive would also serve the legislature
remains unproven.
d. The combined experiences of OECD member countries
highlight the importance of taking a long-term approach and
having realistic expectations about the capacity of performance
management and budgeting to improve performance and
accountability.
CONFIDENTIAL GOV/PGC(2007)16/ANN
33
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
Jordan,
Hackbart 1999 Budget
practices and
procedures
a. What is the current
status of state
performance budgeting
efforts?
b. What is the perceived
impact of performance
budgeting on budget
decision making?
c. What is the
relationship between
performance budgeting
and various aspects of
state budgeting
practices?
n.a. Survey of state
executive
branch budget
officers (Martin
School of
Public Policy
and
Administration
Survey)
Public sector
Cross
tabulations,
regression
analysis
United
States
46 states
1997 a. The predominant form of state level performance budgeting
involves the identification of performance indicators.
b. Almost two-thirds of the responding states indicate that the
governor's executive budget recommendation could be impacted
by the achievement of performance standards.
c. States seem reluctant to use performance directly as an
allocation tool.
d. The regression analysis of state characteristics reveals the
importance of organisational capacity.
d. The budget process is unlikely to be changed substantially
until and unless decision makers demand and use information on
program performance when making decisions about allocating
resources. Having this information is a necessary, but not
sufficient, prerequisite to changing the policy process.
Joumard,
Kongsrud,
Nam,
Price
2004 Budget
practices and
procedures
What is the best way to
enhance the cost
effectiveness of public
spending?
n.a. OECD data
(OECD
individual
country
Economic
Surveys)
Public spending
Cross-
country
comparison
from public
finance data
30 OECD
member
countries
1997-
2002 In most OECD member countries, public spending rose steadily
as a share of GDP over the past decades to the mid-1990s, but
this trend has since abated. The spending pressures stemming
from the continued expansion of social programmes have been
partly compensated by transient or one-off factors. Pressures on
public spending, however, appear likely to intensify, in particular
as a consequence of ageing populations. Since most OECD
member economies have very little scope for raising taxation or
debt to finance higher spending, reforms to curb the growth in
public spending while raising its cost effectiveness are now
required. Based on detailed country reviews for over two-thirds of
OECD member countries, this paper identifies three main areas
for action: the budget process; management practices; and the
use of market-type mechanisms in the delivery of public services.
CONFIDENTIAL GOV/PGC(2007)16/ANN
34
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
Melkers,
Willoughby 2005 Budget
practices and
procedures
a. What is the effect of
performance-
measurement
information on
budgetary decision-
making, communication
and other operations of
U.S. local governments?
b. How pervasive is
performance-
measurement use in
local governments?
c. What types of
measures are perceived
as most useful in these
governments?
d. Where in the budget
cycle are performance
measures documented?
e. How important are
performance measures
at various phases of the
budget process?
f. Are budgeting and
other effects of
performance
measurement evident in
local governments?
g. What factors influence
the effects of
performance
measurement?
n.a. Survey data
from national
survey of city
and county
administrators
and budgeters
+ community-
level data from
International
City/county
Management
Association
Municipal
Yearbook 2002
Local
governments
(cities and
counties)
Multiple-
regression
analysis
United
States
277
responding
city and
county
administrator
s and
budgeters
2000-
2002 a. The use of performance measurement is pervasive, although
survey respondents are less enthusiastic about measurement
effectiveness.
b. The consistent, active integration of measures throughout the
budget process is important in determining real budget and
communication effects in local governments.
c. The implication of performance measurement adds value to
budgeting decisions by providing relevant information about
results, as well as costs and activities.
d. Respondents are less enthusiastic about the effectiveness of
using performance measures to influence budgeting processes
and outcomes in particular.
e. Performance measurement's greatest applicability seems to
be in the budget-development phase.
Paulsson 2006 Budget
practices and
procedures
What does the Swedish
case of the use of
accrual accounting
information tell us about
the prospects of using
that kind of accounting
in public organisations?
New Public
Management
reform
Survey data,
collected by
authors
Central
government
agencies and
government
offices
Survey: 2
questionnair
es + 2 series
of interviews
Sweden
124 financial
managers of
central
government
agencies and
60 senior
advisors of
government
offices
2003 a. Accrual accounting information is used less for both budgetary
politics and policy making in public sector organisations where
the budget is still based upon a cash principle, as is the case in
Sweden.
b. The accrual accounting information is used mostly as a tool in
the management of the agencies.
CONFIDENTIAL GOV/PGC(2007)16/ANN
35
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
Robinson,
Brumby 2005 Budget
practices and
procedures
a. What light does the
empirical literature shed
on the efficacy of
performance budgeting?
b. Does that literature
support the contention
that efforts to link
funding to results in
government budgeting
have failed?
c. What are the effects
of output-based funding
in hospitals?
n.a. Empirical
literature
Hospitals
Literature
review United
States 1970-
2002 a. The empirical literature on government-wide performance
budgeting is disappointingly limited in scope and methodology,
and does not provide a basis for strong conclusions about the
efficacy of these systems. Nevertheless, it does appear to
provide some support for the proposition that, where the
necessary investment in the development of performance
measurement and other performance information has been
made, it is possible to use that information in budgeting to
improve both allocative and productive efficiency.
b. There is quite strong evidence of the efficacy of the sectoral
system chosen —“casemix” funding of hospitals.
Rubin,
Kelly 2005 Budget
practices and
procedures
a. What kind of public
budgeting reforms have
been proposed?
b. What are they
intended to accomplish?
c. How are they working
out?
New Public
Management Data collected
by authors
Public
budgeting
Survey of
budget and
accounting
reforms and
analysis:
cross-
country
comparative
analysis
Australia,
Japan, New
Zealand,
Sweden,
United
Kingdom,
United
States,
China, South
Africa
2002 a. There is a story of ongoing change and adaptation. New
Zealand, Australia and the United Kingdom may be seen as
leaders in the field of public budgeting but these countries
continue to struggle with problems that many claim the NPM
resolves.
b. Reforms have been adopted differentially, emphasis placed on
different elements; they have evolved differentially in the
countries that have adopted them; they have abandoned in
places, in others they accomplish different goals than intended
initially; and they have sometimes set in motion unintended
consequences such as increases in the legislative role to
compensate for the overwhelmingly administrative or executive
branch focus of some of the reforms.
c. In the end NPM seems to have settled into the public
administration landscape as one model that has strengths and
weaknesses just like all other approaches.
CONFIDENTIAL GOV/PGC(2007)16/ANN
36
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
Ammons 2002 Performance
measurement
arrangements
Does performance
measurement lead to
service improvement?
n.a. n.a.
Public sector
n.a. n.a. n.a. a. The performance improvement proposition, that measurement
is a catalyst for performance improvement, deserves a serious
test by practitioners and researchers. It deserves a test that is
fair, one that excludes systems built solely to address
accountability objectives, often in a minimal manner.
b. The key to design measurement systems that will produce
performance gains, especially in organisations with empowered
departments and employees, is the careful development of
measures that cause supervisors and operating personnel to
reflect thoughtfully on the adequacy of services and to consider
strategies of service improvement. In short, measures that
inspire managerial thinking. This means that the measures must
address dimensions of service that operating personnel consider
important, focusing on the efficiency, quality, and effectiveness of
these services and providing regular and reliable performance
information.
Bevan,
Hood 2006 Performance
measurement
arrangements
How robust is the
regime of targets and
terror? Is it performing?
Theory of
governance
by targets
National Audit
Office (2001;
2004)
English Public
Health Service
“Before” and
“after”
comparisons
(United
Kingdom)
United
Kingdom
600 National
Health
Service
organisations
1997-
2005 a. The implicit theory of governance by targets requires two sets
of heroic assumptions to be satisfied: of robust synecdoche, and
game-proof design. The data suggest that these assumptions
are not justified.
b. There are three important gaming problems:
1. ratchet effects refer to the consequences of central resource
managers basing next year’s targets on this year’s performance;
2. threshold effects lead to the concentration of effort on outputs
that are just below the required level at the expense of others,
ignoring the best and the worst;
3. distortion refers to the achievement of output improvements in
areas that are measured at the expense of unmeasured aspects
of performance.
c. Although there were indeed dramatic improvements in
reported performance, we do not know the extent to which these
were genuine or offset by gaming that resulted in reductions in
performance that was not captured by targets.
Halachmi 2002 Performance
measurement
arrangements
Are the same schemes
that measure
performance as a
means of improving
accountability as
suitable for facilitating
better performance or
greater value for
money?
Management
theory on
performance
measurement
n.a. n.a. n.a. n.a. a. Performance measurement can involve the siphoning of
resources from "production" to overhead (under noble pretences)
and that this may not foster greater value for money.
b. The possible dichotomy between the attributes of performance
measurement and the attributes of scorecards that are meant to
enhance performance may resemble Mintzberg's (1994)
dichotomy between strategic thinking and planning.
c. Performance reports should be able to grade organisations on
their ability to meet the challenges as they are carrying out their
mission. Unfortunately most performance measurements still fall
short of meeting this challenge.
CONFIDENTIAL GOV/PGC(2007)16/ANN
37
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source
and sector Methodology Country
Number of
observations
Time
frame Main conclusions
House of
Commons 2003 Performance
measurement
arrangements
Is there a proper
measurement culture? n.a. a. The
Comprehensive
Spending
Review and the
Public Service
Agreements
b. Data
collected by the
authors
throughout 11
evidence
sessions and
two visits
Public sector
(especially
health;
education, local
government
and criminal
justice system)
Qualitative
assessment United
Kingdom
11 evidence
sessions (39
witnesses,
leading to 63
memoranda)
and two
observations
visits
1998-
2002 a. The increase in accountability and transparency which targets
have brought with them has been valuable.
b. But because of the lack of proper integration between the
building of an organisation's capacity through what the authors
call “the performance culture” and tracking quantitative
achievement in the public services through the “measurement
culture”, the laudable aims the government set for its targets, are
in many cases not being fulfilled nor widely recognised as such
by those on the front line whose job is to deliver them.
c. The perverse effect of gaming can be mitigated by loosening
up the performance arrangements.
d. There must be a move from a “measurement culture” to a
“performance culture”.
Van
Dooren 2006 Performance
measurement
arrangements
a. Why are
organisations measuring
performance?
B. Does administrative
supply meet political
demand?
C. What are the system
requirements for
different purposes of
performance
information?
D. What are the effects
of performance
information?
Organisational
theory Flemish
government
officials and
documents
Regional
government
departments
Unstructured
interviews;
content
analysis,
survey, semi-
structured
interviews
Belgium
155 sections
2000-
2005 a. The impact of factors on adoption (having PM) and
implementation (doing PM) of measurement is as follows:
- Measurability of output has a positive correlation with both
adoption and implementation.
- Political interest does not affect adoption or implementation of
performance measurement. Yet, many organisations see it as a
potential hindrance.
- Large organisations have a higher adoption and
implementation.
- The reduction of discretion plays a role in the implementation of
measurement systems, not for adoption.
- The lack of resources does not affect adoption or
implementation. Many sections experience a lack of resources.
Some cope with it, while others do not.
- Coupling measurement with goals is neutral to adoption, but
significant for implementation.
b. Political interest in PM is substantial, but differs significantly
between policy sectors.
c. Perverse effects of measurement are a consequence of the
way PM is used.
- Accountability use leads to many perverse effects.
- Management use leads to moderate effects.
- Learning use leads to minimal effects.
CONFIDENTIAL GOV/PGC(2007)16/ANN
38
Table 2. Literature related to increased flexibility
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Barankay,
Lockwood 2006 Intra-
governmental
structures and
functional
responsibilities
a. Does fiscal
decentralization
increase the efficiency of
delivery of government
services?
Theories of fiscal
federalism Swiss Federal
Statistical Office
Education
Panel
regression
(regression
analysis)
Switzerland
26 Swiss
cantons
1982-
2000 a. More decentralisation is associated with higher
educational attainment.
b. Expenditure decentralisation is more beneficial
when central governments are less competent.
Brewer 2005 Workforce
composition a. What is the role of
frontline supervisors in
the 22 largest American
federal agencies?
b. What about their
contributions to
organisational
performance and
effectiveness?
Boyne's (2003)
model of
government
performance and
service
improvement
(supplemented
by other
sources)
The 2000 Merit
Principles Survey,
U.S. Merit
Systems
Protection Board
Federal Agencies
Descriptive
statistics United States
22 US federal
agencies, and
6 958 Federal
Agency
respondents
2000 a. Public managers must motivate and empower
employees; create sustainable organisational
cultures; emphasise performance at the individual,
work unit, and organisational levels.
b. Frontline supervisors are an important part of the
management team. These employees play a crucial
role in improving federal agency performance.
c. Human resource management matters a great
deal. High-performing agencies tend to have skilful
upper-level managers, strong cultures that value
employees and emphasise the importance and
meaningfulness of the agency’s work, and policies
that empower those employees.
De Borger,
Kerstens 1994 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
a. What are the
differences in productive
efficiency of the
municipal governments
in Belgium?
b. How can we explain
these variations?
Non-
parametrical
theoretical
literature
Data collected by
authors
Municipalities
A non-
parametric
method
based on the
Free
Disposal Hull
reference
technology
Belgium
589 Belgian
municipalities
1985 a. The scale and the fiscal revenue capacity of
municipalities are important determinants of
efficiency.
b. The financing mechanism of local public service
provision and the political characteristics of municipal
governments were estimated to affect inefficiencies.
Laking 2005 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
a. What benefits and
risks are there from
“genetic engineering” of
public organisation
structures?
Theories on the
effects of the
creation of
agencies
n.a.
Agencies
Literature
review n.a. 1980
s and
1990
s
a. In most OECD member countries, agencies work
generally well and meet important needs of good
governance. There is indeed some evidence that in
the appropriate roles they have improved both the
quality and the credibility of public performance.
b. In the developing world and transitional economies,
some appear to have created benefits, others have
caused significant problems for public governance.
c. The major risks of agencification that can threaten
overall government efficiency and efficacy and
accountability are:
1. loss of control of agency operations;
2. abrogation of political accountability;
3. evasion of general rules for staffing and budgets;
4. exposure of government to financial and
CONFIDENTIAL GOV/PGC(2007)16/ANN
39
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
employment risks
5. opportunities for political patronage and corruption.
Pollit 2004 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
What is the usefulness
of the existing theories
about agencies, how are
they used, and for what
purposes?
Theories on
agencification Data (theories
and literature from
other authors)
collected by
author
Literature
review/
theoretical
overview
n.a. n.a. a. One might say that there are three broad families of
theories assessing the phenomenon of agencification:
1. “traditional” mainstream social science: innovation
diffusion theory, some network theory, contract
specification and design, Kingdon's model of the
policy process.
2. economic approaches: rational choice theory;
3. interpretive/ constructivist theories: historical
institutionalism, institutional isomorphism, some
(other) network theory, Latour's translation theory.
b. The different theories have been more or less
successful in answering different questions.
Pollit,
Bathgate,
Caulfield,
Smullen,
Talbot
2001 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
What are the reasons
for, and depth of, the
trend towards
"agencification"?
a. Path-
dependency
theories
b. Historically
informed
institutionalism
Data collected by
author
Public sector
Literature
review Finland, the
Netherlands,
Sweden, and
the United
Kingdom
1996 a. General evidence of agencification as the cause of
greater efficiency, higher quality, etc. is conspicuous
by its absence.
CONFIDENTIAL GOV/PGC(2007)16/ANN
40
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Verhoest 2005 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
What effects do new
control mechanisms
have that governments
use to monitor the
performance of quasi-
autonomous public
agencies?
Principal-agent
theory Data collected by
author
Vlaamse
Openbare
Instellingen'
(VOI's)
Case study
Flemish
Employment
Service
(FES):
a. An
extensive
document
analysis and
time series
analysis with
quantitative
data.
b. A
"subjective"
analysis of
the
perception of
the involved
key-actors,
of the
evolution of
the variables
Belgium
(Flanders)
a.32 'Vlaamse
Openbare
Instellingen'
(VOI's)
b. In-depth
case study of
two
embedded
cases (the job
brokerage
decision and
the vocational
training
division) within
the critical
case of the
Flemish
Employment
Service (FES)
c. 23 semi-
structured
interviews with
the top
management
of the FES
1992-
1999 a. Most contemporary studies on the effect of
autonomization and control of public agencies on
performance seem to show ambivalent results.
b. This case shows that the four control mechanisms
– managerial autonomy, performance contracts,
financial incentives, and competition – seem to induce
the public agency to performance-enhancing
behaviour through ways assumed in the theoretical
model, but only under very specific conditions.
c. Both FES-divisions were motivated to performance-
enhancing behaviour by an urge to strengthen or
restore its legitimacy towards its customers,
stakeholders, and, most importantly, its political
principals (i.e. ministers; cabinet, and Parliament).
Verhoest,
Bouckaert 2003 Intra-
governmental
structures and
functional
responsibilities:
coordination
a. How does the basic
structure of government
change over time: is
specialisation and
fragmentation a
common evolution
(descriptive research
question)?
b. How does the
coordination strategy of
government change
over time (descriptive
research question)?
c. How can eventual
variations in strategies
and trajectories be
a. Path
dependency of
policy capacity
building
b. Trajectory
choice of policy
capacity building
Data collected by
authors
Public sector
Comparative
static
analysis
Belgium,
France,
Netherlands,
New Zealand,
Sweden, and
the United
Kingdom
(France, New
Zealand,
Netherlands
are described
more in detail)
1980-
2000 There has been a growing awareness for the issue of
coordination since the second half of the 1990s.
Countries have followed three trajectories to expand
their policy capacity:
1. renewing existing capacity using hierarchy type
mechanisms: e.g. creating holding ministries and
mega departments,
2. creating new capacity using network type
mechanisms: e.g. linking information systems
(Strategic Results Areas in New Zealand),
consolidating financial documents (consolidated
budgets, consolidated accounts, consolidated audits)
according to policy portfolios (see Canada), (value
added) chain management (see the Netherlands),
3. focusing on present capacity. Market-type
mechanisms are mainly used for support services and
CONFIDENTIAL GOV/PGC(2007)16/ANN
41
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
explained theoretically
(explanatory research
question)?
logistic activities and less for core business.
Verhoest,
Peters,
Bouckaert,
Verschuere
2004 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
a. Do scholars mean the
same thing when they
refer to “autonomy” (or
independence or
discretion) of public
organisations?
b. Are the inconclusive
results of the reviewed
research on autonomy
partly a function of
different
conceptualisations,
operationalisations and
measurements of
'autonomy' by the
involved researchers?
Public
management
theories on the
“autonomy”
concept (of
public
organisations)
A range of studies
concerning the
link between
organisational
autonomy of
public agencies
and their
performance.
b. Survey
conducted by the
authors in 2002-
2003
Government
organisations
a. Literature
overview
b. Survey
along
Flemish
public
organisations
Belgium
(Flanders)
84 Flemish
government
organisations
2002-
2003 a. Contemporary research on the influence of
organisational autonomy on performance in public
organisations uses a diverse and a too restrictive
conceptualisation of “autonomy”.
b. The authors propose a six dimensions-concept of
“autonomy” in public organisations.
c. The formal-legal status of a public organisation is
not the appropriate indicator of its autonomy. the
reason why is because there exists a substantial
heterogeneity of organisations with the same formal-
legal status on each (of the six presented) dimension
of autonomy.
Yamamoto 2006 Intra-
governmental
structures and
functional
responsibilities,
agencification,
delegation
a. What is the
relationship between
autonomy and
performance?
b. The more operational
autonomy agencies
have, the more their
performance will
improve?
c. Will performance
increase by downsizing
the agency?
d. What is the effect of
competition on agencies'
performance?
e. Does the level of
dependency on
government influence
the efficiency of the
agencies?
Kingdon's Policy
Process Model Mail survey of
chief executives,
directors and
senior managers
in 57 semi-
autonomous
public bodies in
Japan, in 2003
Semi-autonomous
public bodies
Regression
analyses Japan
181
responses
received from
57 semi-
autonomous
public bodies
2003 a. There is a causal relationship between operational
autonomy and performance.
b. There is a partial causal linkages between other
types of autonomy like structural or legal autonomy
and performance.
c. The majority of respondents stated that autonomy
in operational management had increased. This,
however did not imply a decrease in accountability.
d. The autonomy in budget execution and
organisational structure was also high.
e. In general, employees agree that agencification
has led to improved efficiency. Employees say that
they have become more cost conscious and customer
oriented to earn more income.
CONFIDENTIAL GOV/PGC(2007)16/ANN
42
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Boyne 2003 Workforce size What are the sources of
public service
improvement?
Theories on
service
improvement
n.a.
Included are
studies of states,
counties,
municipalities
school districts,
schools, police
and fire
departments,
hospitals,
libraries, drug
abuse agencies,
other agencies
Meta
analysis;
literature
study of
statistical
studies
explaining
public
performance
Studies are
mainly from
the United
States (54)
and the United
Kingdom (6)
n=65
1970-
2002 a. Quality of evidence is in general very meagre;
underlying theoretical models are vague, analyses are
static, independent variables are not related to each
other, evidence is mainly from the United Kingdom
and the United States.
b. There is no evidence of a link between financial
resources and service improvement (most studies are
flawed however, because they control for real
resources rather than seeing financial resources as a
way of purchasing real resources).
c. The impact of regulatory reforms on performance is
not sufficiently substantiated.
d. The impact of competition on performance is
mixed; positive, negative and insignificant results are
found.
e. The impact of size on performance shows no
pattern; all kinds of relations are found.
f. The impact of management on performance is
positive; management seem to matter.
g. The mixed results are cases in point for developing
a research agenda on determinants of performance.
h. Soft aspects of human resource management,
such as employee satisfaction and morale, are more
important drivers of performance than hard aspects of
personnel management, such as performance-related
pay and job security.
Brewer,
Selden 2000 Human resource
management
arrangements
What are the key
variables of
organisational
performance in federal
agencies?
n.a. The 1996 Merit
Principles Survey
(U.S. Merit
Systems
Protection Board)
Public sector: 23
largest federal
agencies
Regression
analysis United States
23 federal
agencies
(9 710 federal
agencies
employees
were included)
1996 a. The variables that most affect organisational
performance are efficacy, teamwork, building human
capital, structure of task/work, protection of
employees, concern for the public interest, and task
motivation: all elements of a high-involvement
workplace strategy.
b. The relationship between agencies and their
employees and clients is important: employees want
to be valued by their agencies and involved in
management and decision-making processes.
CONFIDENTIAL GOV/PGC(2007)16/ANN
43
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Moynihan,
Pandey 2005 Human resource
management
arrangements
Which external
environmental
influences and internal
management factors
combine to create
performance?
n.a. National
Administrative
Studies Project
survey of state
government
health and human
services officials
State-level
primary health
and human
services agencies
Descriptive
statistics,
ordinary
least squares
tests,
multivariate
tests
United States
274 managers
(involved in
information
management)
2002-
2003 a. Management matters to performance, it is
reasonable to expect managers to undertake actions
to improve performance.
b. Among the external environmental variables we
find that the support of elected officials and the
influence of the public and media have a positive
impact on effectiveness.
c. Among internal management choices, the ability to
create a developmental organisational culture,
establish a focus on results through goal clarity, and
decentralise decision-making authority are all
positively associated with organisational
effectiveness.
OECD 2005b Human resource
management
arrangements
What are the trends in
performance-related pay
in the OECD member
countries?
n.a. Empirical country
studies (OECD
member
countries) and
extensive staff
surveys in some
OECD member
countries
Public services
Comparative
country
analysis
OECD
member
countries
1980-
2005 a. Staff are less motivated than might have been
expected by the prospect of more money. Extensive
staff surveys, conducted notably in the United
Kingdom and the United States, showed that despite
broad support for the principle of linking pay to
performance, only a small percentage of employees
thought their existing performance pay schemes
provided them with an incentive to work beyond job
requirements and in many cases they found it
divisive.
b. Performance-related pay can help improve
performance when it is applied properly in the right
managerial context, if not because of the financial
rewards then indirectly through the changes in work
and management organisation needed to implement
it.
c. Some common trends are emerging:
1. long-running standardised PRP schemes have
evolved into more decentralised systems;
2. a notable development in recent years has been
the extension of PRP from senior management to
non-managerial staff;
3. several countries have strongly encouraged the
move to a more collective approach to PRP over the
past five years (e.g. Finland, Korea, the United
Kingdom).
d. Overall, the most important criteria for assessing
civil servants’ performance are the following: outputs
achieved which include the pre-identified objectives;
competencies and technical skills; interpersonal skills
and teamwork; leadership and management skills.
CONFIDENTIAL GOV/PGC(2007)16/ANN
44
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
OECD 2005a Human resource
management
arrangements
What are the trends in
human resource
management policies in
OECD member
countries?
Theories on
human resource
management
The OECD
Survey on
Strategic Human
Resources
Management
Public sector (civil
service)
Cross-
national
comparisons
29 OECD
member
countries
1980
s and
1990
s
a. While we continue to have a group of countries with
traditional career-based systems and a group with
position-based systems, there are now many
countries that fall in between; there are countries with
a relative high level of delegation and a relative low
level of individualisation. These countries' HRM
systems tend to emphasise lifelong career in the civil
service with minimum lateral entry (career based).
However, there is a high level of delegation to line
ministries and to lower administrative levels in terms
of hiring numbers and policies, promotion, and pay to
a certain extent (position-based). The authors call this
hybrid system “department-based system”.
b. Austria, Belgium, the Czech Republic, Germany,
Italy, Mexico, the Netherlands, Poland and Portugal
tend to fall into this new category.
Weisman,
Nathanson 1985 Human resource
management
arrangements
What is the relationship
between the aggregate
job satisfaction level of
nursing staff in family
planning clinics and two
outcomes:
1) the aggregate
satisfaction level of
teenage clients with
contraceptive services
obtained in the clinic?
2) the subsequent rate
of client compliance with
contraceptive
prescriptions?
n.a. Longitudinal study
of the professional
staff and clients of
78 county health
department family
planning clinics
Family planning
clinics
Regression
analysis,
path analysis
United States
(the state of
Maryland)
77 family
planning
clinics
(interviews
with 344
family
planning and
community
health nurses
+ 2 900
clients)
1980-
1982 a. The job satisfaction level of nursing staff is the
strongest determinant of the aggregate satisfaction
level of clients.
b. Client satisfaction level, in turn, predicts the rate of
clients' subsequent contraceptive compliance.
Zigarelli 1996 Human resource
management
arrangements
What are the correlates
of effective schooling? n.a. The National
Educational
Longitudinal
Study (1988, 1990
and 1992),
initiated by the
Department of
Education, the
National Opinion
Research Centre
and the
Educational
Testing Service
Regression
analysis United States
24 599
students
(participants in
the base
survey of
1988)
1988-
1992 The most important characteristics of effective
schools are an achievement-oriented school-culture,
principal autonomy in hiring and firing teachers, and
high teacher morale.
CONFIDENTIAL GOV/PGC(2007)16/ANN
45
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Schools
Moon 2002 E-government
preparedness a. Examination of the
rhetoric and reality of e-
government at the
municipal level.
b. What is the current
state of municipal e-
government
implementation and its
perceptual
effectiveness?
c. Which institutional
factors contribute to the
adoption of e-
government among
municipalities?
Hiller and
Bélanger's
Electronic
Government
Framework
a 2000 E-
government
Survey
(conducted by
International
City/County
Management
Association and
Public
Technologies Inc.)
Municipal
governments
E-
government
surveys of
municipal
governments
+ descriptive
statistics
United States
1 471
municipalities
2000 a. Overall, this study concludes that e-government
has been adopted by many municipal governments,
but it is still at an early stage and has not obtained
many of expected outcomes (cost savings,
downsizing, etc.) that the rhetoric of e-government
has promised.
b. There are some widely shared barriers (lack of
financial, technical, and personnel capacities) and
legal issues (such as privacy) to the progress of
municipal e-government.
c. City size and manager-council government are
positively associated with the adoption of a municipal
Web site as well as the longevity of the Web site.
CONFIDENTIAL GOV/PGC(2007)16/ANN
46
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
OECD 2003 E-government
preparedness a. What is the potential
and impact of e-
government?
b. What are the changes
(budgetary barriers)
required to maximise the
benefits of e-
government?
n.a. Deliberations and
guidance of the
OECD E-
Government
Working Group
(12 OECD
member
countries); a
series of project
seminars,
involving
commissioned
papers and
seminar
deliberations;
papers and
information
prepared by
member
countries; and
analyses of
countries' e-
government
strategies
Public sector (e-
government)
Cross-
country
comparison
12 OECD
member
countries
n.a. The initial impressive visible results of e-government
(government websites, a number of sophisticated
transactional services, development of portals)
contrast with the next stage of e-government which
requires the development of hidden infrastructure,
connected back office arrangements and more
complex services. Greater collaboration across levels
of government, higher funding levels and deeper
organisational change will be needed.
Snijkers 2005 E-government
preparedness What can we say about
ICT and e-government
from a public
management
perspective?
Public
management
perspective
n.a.
Social security
sector (public
sector)
Case study Belgium 1990-
2005 a. In the Belgian case of the Crossroad Bank for
Social Security, ICTs were successfully used to
improve public administration in all its aspects.
b. Many e-government benchmarks score the front-
office use of ICTs, but fail to understand, analyse or
score the improvement of back-office processes and
the impact of these process innovations on public
administration values.
CONFIDENTIAL GOV/PGC(2007)16/ANN
47
Reference year institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
West 2004 E-government
preparedness What are the
consequences of e-
government for service
delivery, democratic
responsiveness, and
public attitudes?
b. Is there a potential for
governments using e-
government, to influence
citizens' views about
government and their
confidence in the
effectiveness of service
delivery?
Hiller and
Bélanger's
Electronic
Government
Framework
a. The National
Association of
State Information
Resource
Executives (1998,
1999, 2000)
b. Two detailed
content analyses
of U.S. state and
federal
government Web
sites performed
by the authors.
c. Survey data
from the polling
firm of Peter Hart/
Robert Teeter of
Washington (on
behalf of the
Council for
Excellence in
Government)
Federal and state
government
Content
analysis
public sector
web sites +
national
public
opinion
survey
regression
analysis) +
interviews
with state
officials
United States
a. 27 states
b. 1 813
government
Web sites in
2000, 1 680
government
Web sites in
2001
c. 1 003
Amercian
citizens
1998-
2001 a. In many respects, e-government is developed
incrementally, although in a few cases, technology is
transforming the type of material available to people.
b. While technology has the potential to alter
government service delivery in the future and to
enhance overall system performance, its current
usage has not produced dramatic changes or much
evidence of the fourth stage of e-government-
interactive democracy.
c. In both surveyed years, technology was being used
to enhance democratic performance and
responsiveness to citizen questions.
d. With educational effort about e-government, citizen
beliefs can be transformed in a positive direction for
beliefs about government effectiveness, but not about
spending priorities.
CONFIDENTIAL GOV/PGC(2007)16/ANN
48
Table 3. Literature related to strengthening competitive pressures
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Dronkers and
Robert 2004 Privatisation a. What is the degree
of effectiveness
differences in reading
(and partly
mathematical)
performance of
individual students in
public and private
secondary schools in
19 modern countries,
controlling for the
characteristics of
students and parents
and compositional
characteristics of
schools?*
b. How can
differences be
explained?
c. Are these
differences in
effectiveness similar
in these 19 modern
countries?
Theories on
school
effectiveness
OECD PISA 2000
Survey
Education;
secondary schools
Multi-level
analysis 19 countries,
Austria, Belgium,
Czech Republic,
Denmark, Finland,
France, Germany,
Hungary, Ireland,
Italy, Netherlands,
New Zealand,
Poland, Portugal,
Spain, Sweden,
Switzerland, United
Kingdom, and the
United States
(selection of
countries where
both the public and
private sectors of
education are
developed, and
which represent a
wide-range of
comparable
societies)
2000 a. The main differences in the gross effectiveness of
private and public schools in these 19 modern
countries can be explained by differences in their
student intake and by the related differences in school
composition.
b. Private government-dependent schools have a
higher net effectiveness in reading (and math) than
comparable public schools with the same students,
parents and social composition.
c. The explanation of this higher net effectiveness is
the better school climate in the former, in comparison
to the latter.
d. Private independent schools are less effective than
public schools with the same students, parents and
social composition.
e. These effects of private independent and private
government-dependent schools are more or less equal
in the different modern countries.
Hollingsworth,
Dawson,
Maniadakis
1999 Privatisation a. Are DEA
approaches for
measuring the
performance of health
care appropriate?
b. What is the
efficiency of health
care institutions?
Production
frontier Studies on hospital
efficiency + studies
on general health
organisation
efficiency + studies
on productivity
analysis
Health care;
hospitals, nursing
homes, physicians
Meta
analysis;
literature
study of Data
Envelopment
Analysis
studies
United States and
European Union
n=91 of which 35
studies on
hospitals
1983-
1997 a.DEA techniques work better when the product is
homogeneous and one-dimensional (for example,
kilowatt/hour in the electricity industry) and not multiple
and heterogeneous as in health care.
b. Little statistical testing and sensitivity analysis has
been done in the DEA studies.
c. Efficiency measurement will be more accurate if
researchers concentrate on homogeneous and small
segments of the health care system.
d. In the United States, public hospitals outperform
private hospitals.
e. European hospitals outperform American hospitals.
f. DEA is appropriate, but it should not be relied upon
as the sole decision-making mechanism.
CONFIDENTIAL GOV/PGC(2007)16/ANN
49
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Mobley,
Magnussen 1998 Privatisation Do institutional
environment and level
of market competition
significantly affect the
degree of productive
efficiency in
hospitals?
The Property
Rights
School and
Public
Choice
Theory of
Organisa-
tional
Behaviour
Data collected by
authors
Hospitals
Data
Envelopment
Analysis
United States
(California) and
Norway
228 hospitals of
which 50
Norwegian
hospitals and 178
Californian
hospitals
1991 There is no support for the hypotheses that hospitals
operating in an environment that is private and largely
unregulated perform better than hospitals operating in
one that is public and heavily regulated.
Staat and
Hammer-
schmidt
2000 Privatisation a. Are DEA
approaches for
measuring the
performance of health
care appropriate?
b. What are the
explanations for the
observed productivity
differences?
Production
frontier Krankenhaus
Report 1995, 1996
Health care;
hospitals providing
basic care without
any large-scale
technical facilities
or with some minor
technical facilities
DEA Germany
n=160
1994-
1995 a. The most inefficient hospitals could half their cost by
applying an efficient technology.
b. Ownership seems to play a role for the efficiency of
a hospital, as the hospitals run by non-profit
organisations seem to be less efficient than other
hospitals.
c. DEA is useful as a technique to evaluate hospitals.
Andersen,
Blegvad 2006 Competition a. Does private
provision of child
dental care lead to
higher efficiency in
terms of lower costs
per child, given the
professional
regulation of both
public and private
dentists?
b. Does the
effectiveness of child
dental care services
differ between private
and public providers,
again given the same
professional
regulation?
Theories on
professional-
ism and
efficiency
a. National SCOR
which registers oral
health
b. Statistics
Denmark for cost
data
c. Local Authority
Key Data of the
Danish Ministry of
the Interior for
socio economic
data
d. Organization of
General Dental
Practitioners and
Association of
Public Health
Dentists for
ownership data
Municipal-ities,
organising dental
care for children
Regression
analysis Denmark
n=271
1996-
2002 a. Ownership per se has no bearing on the delivery of
Danish dental care services to children; cost-efficiency
and effectiveness almost identical in public and private
provision.
b. Regulation, in particular from the professions,
nullifies the effect of ownership on performance.
CONFIDENTIAL GOV/PGC(2007)16/ANN
50
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Arum 1996 Competition a. Does the size of
the private school
sector affect student
performance in the
public schools?
b. Are the effects on
student outcomes a
result of changes in
the peer climate,
changes in school
resources, or simply
varying rates of
organisational
efficiency brought
about by market
competition?
c. Does competition
from private schools
affect the
organisational
structure of public
schools?
d. Does completion
from the private
school sector affect
student outcomes in
the public schools, by
changing the
organisational
structure of the public
schools?
Organisa-
tional theory a. Statistical
Abstract of the
United States
1982-1983 (U.S.
Bureau of the
Census 1983)
b. Biennial Survey
of Education in the
United States
1948-1950 (U.S.
Department of
Health, Education,
and Welfare 1954)
c. Digest of
Educational
Statistics 1982
(National Centre for
Education Statistics
1982)
d. High School and
Beyond dataset
Public schools
Regression
analysis United States
a.10 673 public
school students
b. 50 states and
the District of
Columbia
a.
1950
b.
1980
a. State-level student/teacher ratios differ substantially
between public and private school sectors in states
with relatively few private schools, and are closer to
equality in states with a larger private school sector.
b. The student/teacher ratios affect student outcomes
and are responsible for much of the difference in public
school student performance related to competition from
the private school sector.
c. Competitive forces do change the behaviour of
public secondary schools, but they improve public
schools through increased funding, not increased
efficiency.
Dee 1998 Competition Does competition
have a positive effect
on quality?
Theories on
school
effectiveness
National Centre for
Educational
Statistics (NCES)
Common Core of
Data (CCD).
School districts
Two-stage
least-square
regression
United States (18
states)
n=4 488
1993-
1994 a. Exogenous variation in competition from private
schools does appear to improve the quality of public
schools.
b. The success of private schools can proxy for omitted
determinants of student achievement in public schools.
c. The demand for private schools and the quality of
public schools are simultaneously determined.
CONFIDENTIAL GOV/PGC(2007)16/ANN
51
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Duncombe,
Miner and
Ruggiero
1997 Competition What explains
inefficiency of
schools?
Public choice a. Standardised
test scores
b. administra-tive
data on expenses,
salaries
School districts
Data
Envelopment
Analysis
United States (New
York)
n=585
1990-
1991 a. Efficiency is negatively related to school district size,
percent tenured teachers, district wealth, non-
residential property values and labour intensity, and
positively related to the percent of adults who are
college educated.
b. Contrary to expectations, efficiency is found to be
negatively associated with the relative number of
private school students and percent of households with
school-age children.
Lawarree 1986 Competition a. How efficient is
garbage collection? n.a.
Municipal-ities
Cost function Belgium (Brussels
and Wallonia)
118 municipalities
1983 a. Private garbage collection is cheaper.
b. There are increasing returns to scale for
municipalities that organise their garbage collection
together.
Feld,
Kirchgässner;
Schaltegger
2003 Devolution a. What is the effect
of different federalist
institutions on the size
and structure of
government revenue?
Leviathan
model Swiss Federal Tax
Administra-tion,
Swiss Federal
Finance
Administra-tion,
Statistical
Yearbook of
Switzerland
Regression
analysis
(using a
pooled
cross-section
time-series
model)
Switzerland
26 Swiss cantons
1980-
1999 a. Tax exporting has a revenue expanding effect.
b. Tax competition favours a smaller size of
government.
c. Fragmentation has essentially no effect on the size
of government revenue.
d. Overall revenue decentralisation favours a smaller
size of government revenue and shifts government
revenue from taxes to user charges.
Huther, Shah 1998 Devolution Is there a relationship
between the level of
fiscal decentralisation
and governance
quality?
Theories of
fiscal
federalism
World Bank,
Transparency
International,
United Nations
Development
Program,
International
Monetary Fund and
some data
collected by author
Public sector
Regression
analysis 80 developed and
developing
countries
1995-
1996 a. the use of an index of governance quality allows
them to reach unambiguous conclusions regarding the
net positive effects of fiscal decentralization on public
sector performance in a majority of countries.
b. the relationship between the level of decentralized
expenditures and governance quality appears to be
strictly increasing but clearly there must be some form
of "Laffer Curve" – it is easy to construct cases where
complete decentralization of expenditures would lead
to lower quality governance than where there is a mix
of national and sub national expenditures.
c. highly centralized countries can improve their
governance quality through more decentralized
expenditures without the risk of engaging in excessive
decentralisation.
CONFIDENTIAL GOV/PGC(2007)16/ANN
52
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and
number of
observations
Time
frame Main conclusions
Social and
Cultural
Planning
Office
2004 Devolution a. To what extent do
countries differ in
terms of public
performance in the
fields of education,
health care, law and
order and public
administration?
b. To what extent do
differences exist
between countries in
terms of productivity,
quality and
effectiveness in the
delivery of these
public services?
c. To what extent are
public services taken
up by target groups of
government policy?
d. To what extent can
the differences in
public performance be
ascribed to national
institutions?
n.a. Eurostat, OECD,
World Bank and
Council of Europe
Public sector
Regression
analysis 25 EU member
states + 4 non-EU
countries which are
OECD member
countries
(Australia, Canada,
New Zealand,
United States)
2004 a. The score on various government functions can be
combined in one overall index of public sector
performance. the index consists of: i) stabilisation and
growth of the economy; ii) distribution of welfare;
iii) allocation of public services; and iv) quality of public
administration.
b. The same clusters of countries repeatedly emerge in
analyses of public sector performance, regardless of
the policy area reviewed, or the level of analysis:
northern European countries, western European
countries, southern European countries, central
European countries, Anglo-Saxon countries.
c. In many cases, more money does not guarantee
more effective policy outcomes.
d. In many instances there is no one-to-one
relationship between the quality of public sector
services as perceived by citizens or the business
community and quality measured by objective
performance standards.
e. High performance in Scandinavian public sectors is
partly caused by the high levels of local autonomy.
CONFIDENTIAL GOV/PGC(2007)16/ANN
53
Table 4. Literature related to workforce issues
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and number
of observations Time
frame Main conclusions
de
Brabander
and Vos
1992 Workforce
size a. How efficient are
secondary schools?
b. What are the
determinants of
efficiency in the
educational system?
Micro-
economical
theory of cost
function
Two mail
surveys
(performed by
authors) and
administrative
data sets
Primary and
secondary
schools
Descriptive
and
econometric
analysis
Belgium
209 schools
1988-
1989 a. School size has a negative and dominant effect on
the average cost/ student.
b. Vocational and technical education is more
expensive than secondary education.
c. Amalgamations of schools with a comparable offer of
studies is reduces costs because of the increasing
returns to scale of staff use.
Donni 1993 Workforce
size a. How efficient are day
care centres for
children?
b. Do the size or the
profitability of the day
care centres have an
effect on efficiency?
n.a.
Day care centres
for children
DEA Belgium (Wallonia)
115 Belgian day
care centres (71
public day care
centres, 37 private
non-lucrative day
care centres , 7
private commercial
day care centres)
1992 a. Public day care centres are as efficient as the
private non-lucrative day care centres.
b. Private day care centres (for profit) are overall more
efficient than the others.
c. Size has a positive effect on technical efficiency.
McCallion,
McKillop,
Glass, Kerr
1999 Workforce
size a. Is there a difference
in efficiency between
small hospitals and
large hospitals?
b. What is the source
of the difference
between small
hospitals and their
larger counterparts?
n.a. Data collected
by authors
Hospitals
Data
Envelopment
Analysis
Northern Ireland
23 hospitals (five of
which could be
classified as large,
10 as medium sized
and eight as small)
1989-
1992 Larger hospitals display higher cost efficiency, higher
allocative efficiency and higher technical efficiency than
their smaller counterparts.
Ruggiero,
Duncombe 1995 Workforce
size What explains
inefficiency of schools? Public choice a. Standardised
test scores
b. Administra-
tive data on
expenses,
salaries
School districts
Data
Envelopment
Analysis
Tobit
regression
United States (New
York)
n=636
1990-
1991 a. Theories of inefficiency are not supported.
b. Competition hypothesis is not supported; private
schooling correlates positively with inefficiency.
c. Size hypothesis of monitoring costs is not supported;
larger districts and larger schools tend to be more
efficient.
d. Fiscal stress hypothesis is not supported; wealthier
districts are more efficient.
e. Tenure hypothesis (more tenure is more inefficiency)
is supported.
Andrews,
Boyne, 2005 Workforce
composition a. Are representative
bureaucracies a.
Representative Survey of
authorities Regression
analysis Great Britain
2001 a. More representative bureaucracies are associated
with lower levels of citizen satisfaction (perceptions).
CONFIDENTIAL GOV/PGC(2007)16/ANN
54
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and number
of observations Time
frame Main conclusions
Meier,
O'Toole,
Walker
associated with higher
levels of organisational
performance?
b. Is the relationship
between representative
bureaucracy and
organisational
performance
contingent on strategic
stance?
bureaucracy
b.
Organisational
strategy
Local
government
(major English
local authorities)
314 authorities,
2 355 informants b. Representative bureaucracies is unrelated to the
actual level of performance, as measured
systematically and audited.
c. As one managerial strategy is pursued aggressively,
the negative link between representative bureaucracy
and performance can be overcome.
Baugh,
Graen 1997 Workforce
composition What effects do team
gender and racial
composition have on
objective and
subjective team
effectiveness?
n.a. Data collected
by the authors
State regulatory
agency
a. Literature
review
b. Tests in a
field setting
(observation
+ interviews)
c. Univariate
and
multivariate
analyses of
covariance
on the data
United States
31 teams in a
medium-sized state
regulatory agency
n.a. a. Members of cross-functional project teams that vary
with respect to gender or racial composition rate their
team as less effective than members of homogeneous
(all-male or all-white) teams.
b. Ratings of external evaluators show no differences
based on team composition.
Pitts 2005 Workforce
composition a. Does racial diversity
increase or decrease
organisational
performance?
b. Does racial
representation –
matching agency
employees to
characteristics of the
target population –
increase or decrease
performance?
c. Are these
relationships different
for street-level
bureaucrat diversity
and manager diversity?
a.
Representative
bureaucracy
b. Diversity of
the workforce
State records
(reported by all
school districts
on an annual
basis)
School districts
Multivariate
regression
analysis
United States
(Texas)
2 500 school
districts in the state
of Texas
1995-
1999 a. Diversity among managers is unrelated to the three
performance outcomes tested.
b. Diversity among teachers has a diffuse impact,
being negatively related to one and positively related to
two performance outcomes.
c. Representation among managers is positively
related to all three performance outcomes.
d. Representation of teachers is negatively related to
one of the outcomes.
Donahue,
Selden,
Ingraham
2000 Extent and
nature of
unionisation
a. How do major city
governments in the
United States vary in
terms of their human
resource management
Authors' model
of government
management
performance/
capacity
Government
Performance
Project (survey)
Cities
Descriptive
statistics,
bivariate and
partial
correlations
United States
29 (largest) cities (by
revenue) (6 561
observations)
1999 a. Higher capacity governments are able to achieve
better human resource outcomes.
b. More unionised governments and those that lack a
senior professional administrative officer generally
have lower human resource management capacity.
CONFIDENTIAL GOV/PGC(2007)16/ANN
55
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and number
of observations Time
frame Main conclusions
systems and capacity?
b. Is variation in human
resource capacity
associated with
variation in critical
environmental
contingencies facing
city governments?
c. Is variation in human
resource capacity
associated with
variation in the
outcomes of human
resource management
systems in the cities?
d. Is variation in critical
environmental
contingencies
associated with
variation in the
outcomes of human
resource management
systems in the cities?
e. What do these
findings suggest about
the linkage between
management capacity
and government
performance
generally?
Menezes-
Filho, Van
Reenen
2003 Extent and
nature of
unionisation
What are the effects of
unions on innovation ? n.a. Empirical studies
of the impact of
unions on
innovation
Establishment,
firm and industry
level
Literature
review
(survey of
studies)
United States and
Europe
40 empirical studies
of the impact of
unions on innovation
1986-
2001 a. North American results find consistently strong and
negative impacts of unions on research and
development.
b. European studies (mainly in the United Kingdom)
generally do not uncover negative effects of unions on
research and development.
c. There is no consensus of the effects of unions on the
other main measures: technological diffusion,
innovation or productivity growth even in the North
American studies.
d. Unions have a clear positive effect on wages and a
clear negative effect on profitability.
CONFIDENTIAL GOV/PGC(2007)16/ANN
56
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and number
of observations Time
frame Main conclusions
Visser 2000 Extent and
nature of
unionisation
What are the trends in
union membership and
union density,
bargaining coverage,
levels of bargaining,
role of governments in
bargaining, practices of
national social
dialogue, works council
and bargaining
coordination?
n.a. International
Labour Office's
World Labour
Report, national
statistical offices
and the DEUS
database
Public sector
Cross-
country
comparison
100 countries (25
African, 24
American, 20 Asian
and 31 European
countries)
1985-
1995 a. In all but a few industrialised countries market
economies, there unionisation is declining.
b. In OECD member countries, union decline most
notably affected the private sector.
c. The downward trend in unionisation is related to the
rise in unemployment during these years and to
structural changes in labour market participation and
employment.
d. Another trend is the introduction of an additional
level of bargaining, usually in the firm, articulated within
national or – more common – industry level bargaining.
e. In all but a few European countries at least two out
of three workers in the market sector are covered by a
collective agreement.
Aijälä 2001 Attractivene
ss of the
public
sector
a. How to enhance the
competitiveness of
public employers?
B. How to solve critical
skill shortages?
n.a. OECD
Secretariat
(using data from
OECD Country
Reports)
Public sector
Country
reports Austria, Canada,
Denmark, Germany,
Korea, Italy, Norway,
Poland, Portugal,
Sweden, and Spain
11 OECD member
countries
2000 a. Improving the image of the public sector is one of
the most important challenges faced by countries.
b. Surveys are a good way to identify critical issues for
further development.
c. Highlighting the advantages of being a public servant
compared to private employment.
d. Improving human resource management systems is
becoming more and more important and includes
ensuring high quality leadership, a professional
recruitment system, a fair personnel policy etc.
e. Developing pay systems and other incentives may
be necessary.
f. Putting human resources to full use is one solution to
the forthcoming shortage of workforce and critical
skills.
g. Knowledge management remains a big challenge.
h. Regular monitoring and evaluation of strategic
actions is a crucial part of the process and is a tool for
continuing development and improvement.
Gregory &
Borland 1999 Attractivene
ss of the
public
sector
Public-private sector
wage differentials Theories of
human capital Journal articles
Federal
government
Meta study United States n.a. Wage gap is 10 to 20% greater for federal government
workers than otherwise comparable private sector
workers.
Holzer,
Rabin 1987 Attractivene
ss of the
public
sector
What are the effects of
the sustained assaults
on the public service,
on the recruitment of
public servants?
n.a. n.a.
Public sector
Literature
review United States 1980
s a. Sustained assaults on the public service encourage
many of the best and brightest students to pursue
careers in the private sector.
b. Sustained assaults on the public sector lower the
morale of the public service and thus increase attrition.
c. Sustained assaults on the public service encourage
CONFIDENTIAL GOV/PGC(2007)16/ANN
57
Reference Year Institutional
driver Research question Theoretical
framework (if
any)
Data source and
sector Methodology Country and number
of observations Time
frame Main conclusions
elected officials to minimise pay raises for career
officials and thereby discourage recruitment and
retention of the most able public servants.
Lee 2004 Attractivene
ss of the
public
sector
What is the impact of
unobserved
heterogeneity and
selectivity among
sectors on the public-
private sector wage
gap?
Theories of
human capital National
Longitudinal
Survey of Youth
Cohort
Federal
government
Ordinary lest
squares
regression
United States
12 686 employees
a) For men the wage differential is reduced from 17%
to 1% with selectivity correction.
b) For women it is reduced from 22% to 4%.
Melly 2005 Attractivene
ss of the
public
sector
How large is the public
pay gap? Human capital
theories German Socio-
economic Panel
Public sector
Quantile
regression Germany
4 770 wage earners
1984-
2001 a) Wages are lower in the public sector for men but
higher for women.
b) Wage premium is highest at the lower end of the
wage distribution.
Mueller 2000 Attractivene
ss of the
public
sector
Wage differentials
between public and
private sectors
Theories of
human capital Labour Market
Activity Survey
Federal,
provincial and
municipal levels
Fixed-effect
pooled
cross-
sectional
regression
Canada
38 360
1988-
1990 a) Females in the public sector are paid a wage
premium, with federal employees enjoying the highest
rents, followed by local and provincial employees.
b) Results for males are inconclusive.
Postel-Vinay
& Turow 2005 Attractivene
ss of the
public
sector
How large is the public
wage gap? Theories of
income and
employment
British
Household Panel
Survey, United
Kingdom
Public sector
Panel data,
log-likehood
estimation
United Kingdom
3 791 workers
1996-
2002 a) Positive average public premium exists both in
income flows and present discounted sum of future
income flows.
b) Income inequality is lower but more persistent in
public sector.
c) When taking job mobility into account lifetime public
premium disappears for “high-mobility” individuals.
Reichenberg 2002 Attractivene
ss of the
public
sector
How can governments
both compete
successfully for
talented applicants and
retain high performing
employees?
n.a. OECD
Public sector
Review of an
OECD study
on difficulties
in recruiting
and retaining
civil servants
in 11
member
countries
Austria, Canada,
Denmark, Germany,
Korea, Italy, Norway,
Poland, Portugal,
Sweden, and Spain
11 OECD member
countries
2000 If governments are both to compete successfully for
talented applicants and retain high performing
employees, they need to brand government as an
employer of choice that provides challenging work,
progressive human resource policies, and opportunities
for promotion and career development.
CONFIDENTIAL GOV/PGC(2007)16/ANN
58