Título: The Role of Political Ideology in the Structural Design of New Governance Agencies
Autor: Anthony Bertelli
Fuente: Public Administration Review
The Role of Political Ideology in the Structural Design of New
Department of Public Administration and Policy
School of Public and International Affairs
University of Georgia
204 Baldwin Hall
Athens, GA 30602
Phone: (706) 542-9660
Fax: (706) 583-0610
Forthcoming, Public Administration Review
The author would like to thank Sandra Van Thiel for graciously providing the data and some excellent
suggestions for the empirical portion of this paper. Jamie Carson, Keith Dougherty, Laurence O’Toole,
Mark Stater, Shawn Treier, anonymous reviewers for Public Administration Review, and seminar
participants at the Truman School provided helpful comments. Ellen Rubin provided excellent research
This paper employs theories of structural politics and delegation to develop a set of
theoretical propositions about legislative delegations of authority to quasigovernmental
entities, operating more or less at arms length from political authorities. Legislators have
the incentive to condition their choice of the structure of an organization they charge with
implementing a policy on their own political attitudes toward “good government.” The
quasi-independence of quangos provides credibility for legislators to commit to a process
that takes policymaking out of their hands, while still creating a structure that increases
the expected likelihood of achieving their policy goals. Theoretical implications are
empirically examined using data on the financial autonomy of Dutch public bodies. The
results support the theoretical argument that it is very important to consider politicians’
ideologies directly in governance studies as they form the key component of structural
Anthony Bertelli is assistant professor in the Department of Public Administration and
Policy at the University of Georgia. In addition to a book-length treatment of
quasigovernment, his current research interests lie largely in comparative governance and
the statistical measurement of ideology. He is author (with Laurence E. Lynn, Jr.) of
Madison’s Managers: Public Administration and the Constitution, forthcoming, Johns
Hopkins University Press.
Political Ideology and New Governance, Page 1
Governance is an expansive concept, but one that is increasingly important in
understanding the implementation of public policy. It refers to “regimes of laws, rules,
judicial decisions, and administrative practices that constrain, prescribe, and enable the
provision of publicly supported goods and services” (Lynn, Heinrich, and Hill 2000, 7).
With “provision” as their focus and outputs and outcomes as their data, many governance
studies address important issues of policy implementation. But, governance, defined in such
structural terms, relates directly to policy design. Relationships between political authority
and administration generate incentives for the political choice of administrative entity. Hill
(1997, 383) calls governance a “policy/implementation process in which the ‘top’ (and
particularly the politicians) have sought to inculcate dramatic value shifts at the lower level
through institutional changes particularly directed at changing incentive structures.” Note
Peters and Pierre (2000, 22, emphasis original), “the assumption is that if you want to get
governance ‘right’ you need to manipulate the structures within which it is presumed to be
Structural and organizational design present politicians with critical choices (e.g. Moe
1991; McCubbins, Noll, and Weingast 1987). Nowhere is this clearer than in the realm of
quasigovernment, a collection of organizations also known somewhat pejoratively as
“quangos”. These increasingly prevalent organizations lie along a continuum between
hierarchical public bureaucracy and private sector contractees. Thus, the “politics of
bureaucratic structure” (e.g., Moe 1991) has a natural analog in quasigovernment, and we
should expect a theoretical connection as well. Indeed, Van Kersberger and Van Waarden
(2004, 143-144) have noted that governance serves as “a bridge between [academic]
disciplines” that encourages “comparisons between rather different phenomena, which, when
Political Ideology and New Governance, Page 2
viewed under the more abstract perspective of governance, might be found to have something
This paper employs structural politics to develop a theory the legislative delegation of
authority to quasigovernmental entities. It frames the structure question confronting the
political branches of government with theory from political economics (e.g., Kydland and
Prescott 1977; Persson and Svensson 1989; Tabellini and Alesina 1990; Huber and Shipan
2002). Though the political institutions in a given nation are important in understanding
governance (e.g., Kettl 2000; 2002; Pollitt and Bouckaert 2000), legislators, as popular
representatives with votes, share the incentive to condition their choice of the structure of an
organization that they charge with implementing a policy on their own political attitudes
toward “good government” and the strength of their own position—such as the strength of
their party within a parliamentary government—within the political system at the time they
make the choice. For example, if legislators consider administrative efficiency, a leading
goal of the New Public Management (cf. Lynn 1998; Kaboolian 1998), when they choose an
entity to implement a policy, we should expect to see a relationship between such ideology
and institutional choice even though quangos often deal with routine administration.
The structural definition of governance is certainly not monolithic. Rhodes (1997,
47) notes six different definitions of the concept. On one view, governance has been related
to “the use of markets and quasi-markets to deliver ‘public’ services” (Rhodes 1997, 47). In
this particular sense, the new governance relates directly to the quango problem.
Quasigovernmental organizations exist as hierarchies, within networks, and/or within the
discipline of the market, a flexibility that relates to all three prongs of the “structure and
process” conception of governance noted by Peters and Pierre (2000, 14) (See Hogwood
Political Ideology and New Governance, Page 3
(1995, 209-210 for a discussion of this difficulty in defining quangos in structure and process
terms). But, governance can also be considered in a legalistic sense, as a structure of rules
that constrain organizational activity (Rhodes 1997, 48). The ministerial authority retained
over quangos analyzed in the empirical component of this paper relates directly to this sense
of governance, and evidence exists that political attitudes are strongly correlated with the
amount of authority retained over quangos by parent departments (Bertelli forthcoming
2006). The incentives and managerialism of New Public Management as well as notions of
“good” or “efficient” government, are also part of how scholars conceive of governance
(Rhodes 1997, 48-50), which is an important frame for the delegation incentives and political
attitudes analyzed below.
The ex ante choice of quasigovernmental structure is important to governance in that
the chosen structure effects the organization and operation of networks, which are often
associated with the new governance (Rhodes 1997, 51-52). The choice itself allows the
legislature to reflect its attitudes in an intuitive way (e.g., Majone 2001). The quasiindependence
of quangos provides the credibility for legislators to commit to a process that
takes policymaking out of their hands, while still creating a structure that increases the
expected likelihood of achieving their policy goals. The credible commitment perspective is
an important theoretical tool for understanding quasigovernmental delegations (Gilardi 2002;
Elgie and McMenamin 2004).
Locating the provision of public services within a sophisticated view of the political
system in a democratic society is central to understanding governance, i.e., in a “sociocybernetic”
sense (Rhodes 1997, 50-51). Public management reforms in the new
governance, write Pollitt and Bouckaert (2000, 17) amount to “deliberate changes to the
Political Ideology and New Governance, Page 4
structures and processes of public sector organizations with the objective of getting them (in
some sense) to perform better.” What is “better” is defined by the legislators’ attitudes
toward administration, i.e., administration that is more efficient is better. This attitudinal
model—the term is borrowed from the work of Segal and Spaeth (1993) on the ideology of
U.S. Supreme Court justices and its influence on decision outcomes—is quite distinguishable
from a leading perspective on delegations to quasigovernment attributable to Van Thiel
(2004, 183, emphasis original): “The benefits of reduced ministerial responsibility and
patronage will appeal to all politicians, irrespective of their ideological preferences.”1 The
attitudinal perspective suggests that those governments who have specific ideologies on
administrative efficiency and corruption will make structural choices about
quasigovernmental institutions conditional on their attitudes. Thus, the theory presented here
moves beyond generic left-right notions of ideology toward specific attitudes of relevance to
the new governance.
Theoretical implications are empirically examined using data drawn from a study of
an important class of quangos, public bodies, published in 1995 by the Netherlands Court of
Audit (Algemene Rekenkamer 1995) as well as manifestos (i.e., party programs or platforms)
data on government policy preferences (Budge, et al. 2001). The structures examined relate
to the public financing of a quasigovernmental entity, and results suggest significant
relationships between financing regimes and legislative attitudes toward administrative
efficiency. The results support the theoretical argument that it is very important to consider
politicians’ ideologies directly in governance studies as those attitudes form the key
component of structural politics.
Political Ideology and New Governance, Page 5
This article proceeds as follows. First, an introductory discussion of quangos is
provided. An attitudinal model of quasigovernmental design is then developed from the
political economics of delegation and testable hypotheses are constructed. Descriptions of the
data and statistical methods are then presented, followed by a discussion of the empirical
results. The paper concludes with some brief remarks.
Defining the Quango
Quasigovernment is comprised of a set of “organizations which as their main task, are
charged with the implementation of one or more public policies, and which are funded
publicly but operate at arm’s length from the central government, without an immediate
hierarchical relationship existing with a minister or a parent department” (Van Thiel 2001,
5). An important form of quasigovernmental authority is known as the public body. These
organizations, of which British Non-Departmental Public Bodies (NDPBs) and Dutch
zelfstandige bestuursorganen or independent governing organs (ZBO) are examples, are not
legally a part of the government. Nonetheless, they are publicly funded through budgetary
allocations, taxes, or fees. In the Netherlands, for example, public bodies can be financed
through government payments or by levied tariffs, the latter of which can be set by the state
or the public body (Van Thiel 2001, 25).
Public bodies reveal the flexibility of policy design in the organizational “grey zone”
between markets and hierarchies (Greve 1999). Legislators are not constrained to the choice
of one bureau or another or some combination to be their agents. They may design agencies
for a single policy task, even if it means committing assets to the quango that are “nonrecurrent
and task-specific” (Van Thiel 2001, 46). Quasigovernmental organizations permit
“control and regulation by new bodies of experts or appointed individuals who worked with
Political Ideology and New Governance, Page 6
an element of independence from traditional politics and therefore were [sic] not tainted . . .”
in the eyes of an increasingly distrustful public (Flinders 1999, 30). Comparative public
policy scholars have provided evidence that, consistent with the political economic theories
of delegation, quasigovernmental delegations represent a credible legislative commitment to
specific policies (Gilardi 2002; Elgie and McMenamin 2004). Important among policy
commitments in the New Public Management is the efficiency of administration and service
Administrative independence, however, has serious consequences for accountability.
Quasigovernmental delegations “place public money and government functions in the hands
of unelected persons whose links to the elected bodies that supervise government are tenuous
at best” (Hirst 1995, 341). Much has been written on the corruption and patronage
opportunities available in quasigovernment (e.g. Jolley 1983; André 1985; Payne and
Skelcher 1997; Skelcher 1998). In the Dutch context, Van Thiel (2001, 45) notes as an
example that “interest groups can be charged with policy implementation and thus become a
quango (‘hiving in’) or they can be appointed members of a quango board.” Though interest
groups can influence bureaucracies, they cannot receive this type of patronage. “By
increasing the participation of (groups of) voters in the implementation of policies, their
commitment—and hence their electoral support—to politicians will increase” (Van Thiel
2001, 45). Overcoming this problem has been an issue in delegations to administration for
many years.2 One often used tool for increasing diversity of representation on public
boards—to reduce patronage—is the inclusion of a requirement that members of the board be
paid for their service. Remuneration makes it possible for a wider range of individuals to
afford service in the public bodies.3
Political Ideology and New Governance, Page 7
Governance and Policy Design
In the common conception of democratic government, the polity grants political
authority to a legislature to make laws on its behalf. An accountability check is retained
through the “electoral connection” which creates for legislators certain incentives when
congealing the preferences of represented constituencies into legislation. In the United
States, for example, members of the House of Representatives are elected by plurality in
single-member districts. Individually, they may seek prestige, influence, and the
accomplishment of particular policy objectives (Fenno 1973), but the electoral connection
focuses his or her attention chiefly on reelection (Mayhew 1974). Responding to the
reelection incentive leads legislators to publicly take policy positions and to claim credit for
policy accomplishments supported by their constituencies. Communicating achievements
leads constituents to trust their representatives, and this trust, manifest through security in
office, allows legislators to build influence among their colleagues (Fenno 1978).4 The
localization of electoral incentives is also known to students of quasigovernment: “[T]he
local preoccupations and priorities of the politicians and private actors” figure prominently in
the reforms at the heart of the New Public Management (Pollitt and Bouckaert 2000, 18). As
credit-claimer and complaint-resolver, the legislator has the incentive to be keenly interested
in governance structures and processes as they effect his or her technology for the
accumulation of political capital.
Parliamentary systems likewise have an electoral connection, though it operates quite
differently. The Netherlands, for example, is a closed list, proportional representation
system, and since the entire country is a single constituency it is one of the world’s most
proportional electoral regimes (Farrell 2001, 89). But, this drives it away from regional
Political Ideology and New Governance, Page 8
accountability linkages that characterize seats in the U.S. House (Gladdish 1991, 100-101).
Nonetheless, the closed list5 does induce a connection to a party’s issue niche. Party elites
have the ability to order the lists in any way they choose, and since parties represent “by
extension the ideological groups for which they stand” (Katz 1980, 121), voters gain
substantive representation by issue, rather than geography. Thus, party pressure to avoid
losing one’s place on a list through voting against party ideology creates incentives that hew
MPs to their ideological brands. Collegial trust comes from support of the party’s issue niche
in legislative behavior, which builds security in office.
But, the new governance, with its “hollowing” of the state (Milward and Provan
2000), makes the legislator’s role more difficult, for those who are tied to their constituencies
by issues or geography. If services are to be “hived out” to quasigovernment (or new
services “hived in” from the private market to quasigovernment), the ex ante design of these
delegations is critically important. Legislator ideology should be apparent in these decisions,
just as in the political control literature of deck stacking (e.g., McCubbins, Noll, and
Weingast 1987) and structural politics (e.g., Moe 1991) since similar agency incentives are
operating. This notion also comports with the managerialistic notion of governance as
“nothing less than the steering of society by officials of what are organizationally the
‘commanding heights’ of society” (Goodin 1996, 13).6
The Political Economics of Delegation
In a model of the U.S. government, Epstein and O’Halloran (1999) have shown that
policy uncertainty creates the incentive for legislatures to delegate vague powers to
administration. However, discretion is tempered by conflict between administrators and
legislators over the nature of ideal policies. To account for the effects of differences in
Political Ideology and New Governance, Page 9
political systems, Huber and Shipan (2002) examine the effects of both policy conflict and
legislative capacity on discretion. Unlike in the highly professionalized U.S. Congress, they
argue, parliamentary “politicians may find themselves in the position of wanting to write
detailed legislation (because they distrust bureaucrats [in the sense of a principal-agent
model]) but being ill equipped to do so” due to informational and institutional costs (Huber
and Shipan 2002, 79).
The reasons for delegation have also been the object of study in this literature.
Beyond simple expertise and division of labor, delegations can make legislative
commitments to policy credible, i.e., believable by other concerned actors. In thinking about
monetary policy, Kydland and Prescott (1977) theorized that given the electoral connection,
politicians have a powerful incentive to temporarily stimulate the economy on the eve of an
election, regardless of the long-term economic consequences that such action engenders.
Thus, delegation to an independent authority can overcome this problem of temporal
inconsistency of incentives.
Even assuming a solution to this problem, the limited time horizon of a government—
it is elected for a short period of years—makes commitment to policy over time problematic,
since a new government with a new ideology can come to power. “Government” is used here
in the sense of a parliamentary coalition or, in a separated powers system, legislativeexecutive
balance (divided, unified, and degrees of division). For example, if the next
regime is expected to have different views of a package of government-provided public
goods or public spending levels, the current government has the incentive to spend now and
saddle the next regime with debt to weaken the successor’s ability to implement an austerity
program (e.g., Persson and Svensson 1989; Tabellini and Alesina 1990; Alesina and Roubini
Political Ideology and New Governance, Page 10
1997). Both sets of incentives strengthen with ideological polarization.7 In proportional
representation (PR) systems such as the Netherlands, the issue niches of parties correlate with
polarization on more specific issues than in “left-right” competition in single-member district
plurality systems (Katz 1980; Cox 1997). Thus, on an issue like efficient public
administration, polarization may be more readily discernable in PR systems.
Quangos present a challenge for existing political-economic theories of delegation,
since their tasks are often comprised of routine administration, such as the distribution of
benefits and the collection of user fees, rather than more politically salient policymaking.
Thus, the policy space that legislators face is quite different than that of, say, economic
policy. An election would very likely not be called on account of quango policy, yet since
multi-party, proportional representation systems like the Netherlands are characterized by
issue representation, the “administrative efficiency” conception of the new governance, for
example, can create incentives for parties who have advocated efficiency reforms to make
policy commitments through delegations to quangos. In other words, regardless of the
characteristics of the policy that is ultimately implemented by the quango, the mere act of
delegating to a quasi-independent agent removes the taint of limited time horizon incentives
(Gilardi 2002; Elgie and McMenamin 2004).
The theory of delegation, then, provides an important link between what Lynn,
Heinrich, and Hill (2000, 35-36) call the “institutional level” (policymaking by political
institutions) and the “managerial level” (the “elaboration of strategies” for policy
implementation by administrative actors) of governance. This is the central locus of policy
design in the new governance, and the bridging theory of delegation permits sophisticated
analysis of the role of political ideology.
Political Ideology and New Governance, Page 11
Taking the perspective of the political economy of delegation generates several
implications for the comparative study of governance. Government attitudes, particularly
those regarding administrative efficiency, affect the structural design of quasigovernmental
organizations. More specifically, commitment problems make delegation to an independent
administrative agent useful (temporal incentive incompatibility), but create the incentive to
retain governmental authority over budgets so that burdens can be laid on successors that
prolong the effect of the current government’s policy ideology (limited time horizon).
The task performed by the quasigovernmental organization matters. Delegation is,
after all, granting authority to another to perform a job on one’s behalf. Under the
philosophy of the New Public Management, for example, many tasks are better checked by
the market. The market, the argument goes, provides discipline with better informational
quality than does government oversight. But, market discipline and contractual
accountability work better when organizational outputs are easily observed, and this occurs
when tasks are well-defined rather than given a vague mandate under which agents may
exercise broad discretion (e.g., Williamson 1985).
It is a well-established aspect of spoils systems that patronage weakens the credibility
of commitment (temporal incentive incompatibility). If a board member’s job is obtained
through a political favor, it is more likely that the official will ally him or herself with the
interests of the politicians who made such employment possible.
The political economy of delegation elaborates this phenomenon through the credible
commitment. Temporal incentive inconsistency dictates that the incentive for legislators to
renege on policy commitments for electoral gain is substantial, and they can increase
Political Ideology and New Governance, Page 12
credibility by granting independence to tasks performed by quasigovernment. Legislators
would also like to extend the time horizons of their policy outcomes beyond the life of the
current government coalition. Thus, it is expected that legislators will retain less
governmental control over more observable tasks, permitting the market to discipline their
providers and legislators to claim credit for the efficiency gains of running government like a
business. However, as tasks permit more discretion, the incentives for policy influence and
the ability to renege on particular choices outweigh the returns to credibility. As a
consequence, we expect more quasigovernmental independence at lower levels of discretion,
but that these incentives will gradually diminish as discretion increases.
Of course, this political economy view is not the only theoretical lens for the quango
problem. The view of new governance as a non-self-executing collaboration among state,
citizen and intermediaries produces some additional considerations of the interconnectedness
of quasigovernmental agents (Salamon 2002). Frant (1997) argues that independent
administrative authorities can add to the selection of organizational alternatives, rather than
merely provide government access to markets. In keeping with the framework derived
above, his argument suggests that legislators choose organizational structures and procedures
as well as financing arrangements. In other work (Bertelli forthcoming 2005), I have
examined the legislative choice of accounting rules for quangos by viewing the ex ante
design of quango structures from the perspective of auditing, as in the economic theory of tax
compliance (i.e., Andreoni, et al 1998). By creating such requirements as the requisite
submission of annual plans, reports, and accounts, governments attempt to create a
compliance-inducing scheme much like that present in tax enforcement (Bertelli forthcoming
2005). The purpose of this paper however is to consider credible delegation theories in the
Political Ideology and New Governance, Page 13
context of quasigovernment, and the remainder of the paper constructs an empirical test of
the theoretical implications previously discussed.
Data and Methods
The data are drawn from study of 545 Dutch public bodies conducted by the
Netherlands Court of Audit (NCA) during 1993-94 and published in 1995 (Algemene
Rekenkamer 1995). The NCA, an auditing body with no judicial authority, examines the
regularity and efficiency of expenditures of Dutch public funds, and in fulfilling this mission,
conducted this study of ministerial accountability in “autonomous administrative authorities.”
To be sure, public bodies represent a significant part of Dutch public administration,
comprising, for example, 18 percent of the total expenditures of the Dutch government in
1992 (Algemene Rekenkamer 1995). The study, conducted in 1993-1994, provides a
particularly rich cross-section of quango activity. Summary statistics and variable
descriptions appear in Table 1.
The unit of analysis is the legislation that creates a public body or “cluster” of public
bodies.8 A cluster occurs when multiple public bodies are charged with exactly the same
task. The common case of clustering is a single policy implemented by a formally distinct
public body in each geographic region of the Netherlands (Algemene Rekenkamer 1995; Van
Thiel 2001). For example, legislation governing hunting licensing can set up separate
entities that issue licenses in each of the twelve Dutch provinces. Thus, the legislators’
choice of establishing a cluster of quangos in a single substantive area is represented in the
NCA data as a single organization. Since government choice is the theoretical unit of
analysis, my empirical choice of the unit of analysis as the statute is appropriate.
[TABLE 1 ABOUT HERE]
Political Ideology and New Governance, Page 14
Of interest in this study is the legislative choice of financing regime for a public body.
Chun and Rainey (forthcoming) employ a measure of financial publicness—the proportion of
resources from non-public sources. The NCA data permit the construction of a financial
publicness scale variable as a four-category nominal variable. This variable takes a value of
1 when the public body sets its own tariffs (Revenue Autonomy), 2 when the government
sets the tariffs (Revenue Restriction), 3 when the public body is a government financed
commission or board (Budget Autonomy), and 4 when the budget for the public body is
determined by the government (Budget Restriction).
When a public body has the ability to set the tariffs that generate its operating
revenues (category 1), it has a substantial measure of autonomy. In such cases, within the
mandate expressed in the delegation (i.e., in the language of legislation) the public body is
granted the ability to collect user fees, much like the power of a local government under a
municipal charter (Van Thiel 2001, 25). The legislature also has the ability to retain this
authority for the government (i.e., charging a cabinet ministry with the task), such that the
minister of the quango’s parent department is given the authority to set the relevant tariffs
(category 2). Government financing (category 3) means that a public body receives funds
for its operation from the government to spend within its discretion, allowing it to respond to
market, rather than ministerial, forces. However, the legislature may also grant a parent
department the authority to set the budget for a public body (category 4), restricting the
quango’s financial autonomy since budget size and priority-setting remain fully government
powers, while the quango may spend the funds it gets with discretion limited by the
parameters of the budget.
Political Ideology and New Governance, Page 15
For purposes of the present analysis, then, the key comparisons are between
categories 1 and 2 as well as between 3 and 4. These are stressed in the results and shown in
Political attitudes are operationalized through the use of Manifesto Research Group
(MRG) political party preference estimates, which are available for 10 Dutch parties from
1946-1993 (Budge, et al. 2001). Coded manifestos (i.e., party platforms or programs) “can
be used as a convenient summary of parties’ preferred policy positions as distinguished from
their actions or record in government” (Budge, et al. 2001, 53). Since rational choice
theories such as those that motivate the hypotheses in this study require preference data for
testing, manifestos provide an excellent set of preference measures on specific issues, or in
the lexicon of spatial theories of politics, policy spaces (Budge 2000).
The use of manifestos data as preference estimates is based on a “saliency theory of
party competition,” which holds that “[p]arties by and large endorse the same specific issue
stands but still prioritize them differently” (Budge, et al. 2001, 76). Of specific importance to
this study, the MRG authors found, in support of the saliency theory, that the preference
estimates provided evidence that parties emphasize different policy priority issues, rather than
creating a situation where substantively “everything one side urges must be flatly repulsed by
the other” (Budge et al. 2001, 7). This is particularly important for the present study, as it
allows for a more nuanced measure of salience on the issue of administrative efficiency,
which represents the policy space of primary concern in the new governance.
The quantitative MRG estimates were constructed through a content analysis of party
policy statements “in the shape of election programmes (manifestos)” (Budge, et al. 2001, 2).
Political Ideology and New Governance, Page 16
A party ideology estimate on a policy issue is simply the number of references to a
substantive policy priority9 present in the manifesto(s) for a particular election year.10 The
measure of administrative efficiency used in the present empirical analysis captures
manifestos statements regarding the “need for efficiency and economy in government and
administration; cutting down civil service; improving governmental procedures; general
appeal to make the process of government cheaper and more effective” (Budge, et al. 2001,
224). This is intended for present purposes as a measure of those attitudes central to the New
Single-party estimates are transformed into a government preference estimate by
calculating the sum of the manifestos scores on an issue for all parties in parliament weighted
by their seat share through the following formula.12
“j = pij #
Government ideology on issue j, Γj, is the sum of the preference estimates on issue j for all N
parties in parliament weighted by their seat share.
As Baron (1991) shows, equilibrium policy proposals in parliamentary governments
reflect the preferences of parties both in and out of government.13 This effect is exaggerated
in the case of single-issue ideology, such as the administrative efficiency attitudes presently
utilized for the following reason. Parties, such as those measured in the manifestos data,
condition their manifestos on the issue positions of parties with which they anticipate sharing
coalition power, and consequently may not speak to issues that their partners champion.14
Thus, each party with a positive seat share in the Dutch parliament is included in the
estimates. The administrative efficiency ideology score for the public body is the score in the
year in which it was created. If a quango was created in a year between elections,
Political Ideology and New Governance, Page 17
government preference estimates for the immediately preceding election year were
employed. Note that though this variable is temporal, the dataset remains a cross-section. As
noted below, I do control for unobserved attributes that lead to the creation of public bodies
in a particular year, as they may be correlated with the error term in the multinomial logit
A variable identifying board remuneration takes a value of 1 if the legislation
requires remuneration of board membership and 2 if such payment is not obligated.
The NCA study identifies eight categories of tasks, which have been arranged on an
ordinal scale from least to greatest task discretion according to the public body’s ability,
during the course of implementation, to “fill up the details” unspecified by the legislature.
Paying benefits and collecting fees represent the lowest rank category of task discretion,
followed by registrations, research, advising, judging quality, and licensing. The penultimate
level of discretion is the ability to make regulations and adjudicate claims, while the highest
is supervision, akin to appellate authority. Placement of a public body within a category was
a decision made by the analysts who conducted the NCA study. These decisions were based
on the analysts’ assessments in 1993-1994 of the tasks that a public body performed based on
its enabling legislation and official activity reports (Algemene Rekenkamer 1995).
The size of public bodies, measured through ranges in the number of employees, is
included as a control variable and described more fully in Table 1. Some ZBOs are very
large. For example, four public bodies employ 16,823, 3696, 5336, and 2374 personnel each.
Some are unstaffed; 37 report no personnel and 23 have five or fewer employees (Algemene
Rekenkamer 1995). This variable controls for any unobserved attributes of quangos of
various sizes. Since there is no reason to expect that quango size affects financing
Political Ideology and New Governance, Page 18
arrangements as a continuous linear determinant (i.e., each additional employee creates a
statistically significant effect), the variable is categorized as to capture relative size effects.
Thus, quango size is a categorical variable taking a value of 1 when the public body has no
employees, 2 for 1-24 employees, 3 for 25-49 employees, 4 for 50-74 employees, 5 for 75-99
employees, 6 for 100-999 employees, and 7 for 1000 or more employees.
The theoretical implications presented above can be operationalized as follows:
Hypothesis 1: Governments with stronger administrative efficiency attitudes
are more likely to give public bodies the ability to set their own tariffs, but are
also more likely to grant government the ability to set the budget for a public
Administrative efficiency is strengthened when public bodies are not micromanaged
by ministerial authorities. Legislators who have these attitudes grant tariff-setting authority to
public bodies as a commitment device. Furthermore, legislators make a credible
commitment by retaining government authority over budget-setting. Alienating budgetary
authority is tantamount to deck stacking, and plays against anti-corruption and administrative
Hypothesis 2: Public bodies with lower task discretion are more likely to set
their own tariffs and less likely to have government set their budgets.
However, the likelihood of these regimes becomes indistinguishable from that
of government tariff and budget setting as the public body’s task discretion
Observable tasks, as discussed, are granted more independence, than more observable
tasks. As independence dampens with increasing discretion, the likelihood of each
publicness regime does not go to zero, for each public body must have some such regime.
Rather, the expectation is that each regime will become equally likely.
Political Ideology and New Governance, Page 19
Hypothesis 3: The inclusion of an obligatory board remuneration requirement
by the legislature increases the odds of revenue and budget autonomy.
If credibility is a goal when making delegations, legislators should choose to increase
the diversity of personnel while also increasing the budget and revenue autonomy. They can
increase diversity (weaken patronage) by requiring board payment, which makes it possible
for capable individuals across income brackets to serve the public through positions in the
Since the dependent variable is a four-category nominal variable, I estimate a
multinomial logistic model (Nerlove and Press 1973) to provide probabilities of the
legislature’s choice of each type of financing regime for a public body. I model financial
publicness as a function of discretion, administrative efficiency, board payment, size and an
error term. The specification from passes a Hausman test for the independence of irrelevant
alternatives in the legislature’s choice (Hausman and McFadden 1984). Wald and
Likelihood-Ratio tests suggest that no categories of the dependent variable should be
combined (Anderson 1984).
Huber-White-Sandwich robust standard errors are computed controlling for intracluster
correlation during the year in which the public body was created (Kent 1982). For
robustness, an unreported model including year of creation as an independent variable to
control for trending; no substantive difference in results was shown.
The results support each of the three hypotheses presented in the last section. Table 2
presents regression results for all comparisons of categories of financial publicness. To test
Political Ideology and New Governance, Page 20
for robustness, a model controlling for the policy area in which the public body operates was
estimated (McFadden LRI = 0.286). No sign or significance changes from the results
presented in Table 2 were demonstrated.
The key comparisons (Revenue Restrictions v. Revenue Autonomy; Budget
Restrictions v. Budget Autonomy) show nearly all variables having significant effects. In the
budget comparison, board remuneration is not a significant predictor though it maintains
significance in the revenue assessment. Multinomial logit regression coefficients are difficult
to interpret, so to more closely examine the hypotheses, simulated probabilities of public
bodies being located in each category of reporting requirements are computed. This is done
using the King, et. al. (2000) post-estimation simulation method and their program,
CLARIFY, for STATA. 1000 simulations were drawn of the parameters in the statistical
model from their asymptotic distributions (multivariate normal with mean at the estimates
from Table 2 and variance equaling the variance-covariance matrix of those estimates).
Then, predicted probabilities were generated from the samples based on minimum and
maximum values of administrative efficiency with other values set at their means.
[TABLE 2 ABOUT HERE]
The Attitudinal Model
The results strongly support the notion of an attitudinal model (Hypothesis 1) at work
in quasigovernmental delegations. Under a legislative regime with the minimum level of
administrative efficiency ideology in the sample, the chance by which a public body is able
to set its own tariffs is roughly 44 percent, with a 95 percent confidence band (hereafter
denoted in brackets) holding that likelihood between 27 and 64 percent. By contrast, under
the strongest administrative efficiency ideology, that chance increases to 59 percent [.34,
Political Ideology and New Governance, Page 21
.80]. Thus, stronger attitudes toward administrative efficiency are associated with a higher
probability that quangos have revenue autonomy. Similarly, the expected results are found
for revenue restriction. Under the weakest attitudinal regime, the likelihood of governmentset
tariffs is 8 percent [.01, .26], but decreases to only 1 percent [0, .05] under the most
stringent views favoring efficient administration. With respect to the revenue categories of
financial publicness, stronger legislative attitudes toward administrative efficiency are
associated with greater quango autonomy.
Support for the attitudinal model is also apparent in the budgetary aspects of the
financial publicness measure. The least stringent regime of attitudes favoring administrative
efficiency is associated with a 43 percent [.24, .63] chance that a public body is government
financed, but only a 5 percent [.01, .12] chance that the government sets its budget.
However, under the strongest efficiency orientation, the possibility of government financing
drops to 17 percent [.06, .35] and government budget climbs to 23 percent [.08,.46].
Governments with stronger sentiment toward administrative efficiency are correlated with
greater budgetary restriction on quangos.
[FIGURE 1 ABOUT HERE]
Hypothesis 2 is also strongly supported. Figure 1 shows the effect of task discretion
on financial publicness regimes in the sample. As hypothesized, the curves showing the
probability of own tariff-setting by public boards and government financing are greater at all
levels of discretion, but diminish in that quality, until probabilities are roughly equivalent at
the highest levels of discretion.15
Political Ideology and New Governance, Page 22
Obligatory board remuneration is associated with higher probabilities of revenue and
budget autonomy than the absence of such a requirement. The probability of a public body
setting own tariffs is 56 percent [.30, .79] under a required payment regime, but 50 percent
[.34, .66] under no requirement. The likelihood of government-set tariffs is 16 percent [.03,
.46] when payment is required, but just 2 percent [0, .06] when board payment is not
compulsory. Finally, under a payment obligation, the probability of government budgetsetting
is 12 percent [.03, .32], which falls to 9 percent [.04, .16] when that obligation is not
present. These results suggest that efforts by Dutch legislators to counteract the policy costs
of patronage through board remuneration requirements are coupled with greater revenue and
budget autonomy. Checking patronage, and thus signaling greater agent independence,
allows legislators to strengthen the credibility of policy commitments through
quasigovernmental delegations by granting their agents more financial autonomy.
This article provides theoretical insights and suggestive empirical evidence with
regard to a critical problem in the new governance: arms-length delegations to quasiindependent
agents. Such delegations are at the heart of new governance reforms that would
place administrative agencies more or less within the discipline of the market. The political
economics of delegation, and in particular, the notion of credible commitment through
delegation, affords analytical leverage to understand such delegations. Structural politics,
long considered in delegations to bureaucracy, is also an important theoretical perspective for
understanding quasigovernment. The bureaucracy and new governance problems are not
fundamentally dissimilar in the incentives they impose on legislators. A key difference is in
the range of types that agents may take, which is much larger in the case of quasigovernment.
Political Ideology and New Governance, Page 23
Analyzing quasigovernmental delegations in this way also facilitates comparative study since
political attitudes can be compared using manifestos data, as can regimes (i.e., through
measures of legislative capacity).
Empirical results for public bodies in the Netherlands suggest that political attitudes
have a significant impact on the revenue and budget autonomy of a quango. Interestingly,
legislator concerns over administrative efficiency manifest themselves in the form of the
financing regimes employed. Stronger legislative attitudes toward administrative efficiency
are more likely to grant quangos revenue autonomy, but are also more likely to impose
budgetary restrictions on public bodies (compare Langbein 2000). The observability of tasks
performed by the quasigovernmental authority also matters, with more observable tasks
being associated with less financial restriction. As Thompson (1993) notes, a mismatch
between financial controls over agencies and the tasks those agencies perform can be very
costly, and the results presented here seem to suggest that Dutch lawmakers are cognizant of
such costs when delegating authority to public bodies.
Patronage incentives exist in delegating powers to quangos, but when they are
counteracted with salary requirements to induce the selection of a more professional and
representative directorship, legislators grant quangos greater authority. Checking known
incentives for patronage allows legislators to make their policy commitments stronger.
In closing, it is worthwhile to note the role of the practitioner in this commitment
scheme. Commitment is premised on the responsible exercise of expert judgment by the
quango personnel when performing their duties. This is quite unlike the implications of a
patronage regime, where the primary expectation is loyalty, not independent judgment.
Political Ideology and New Governance, Page 24
Good public administration provides a set of incentives for legislators that is important in
understanding the new governance.
1. Van Thiel (2004, 182) views the creation of a quasigovernmental organization as the transfer
of production rights from the legislature to the public body. This reduces the responsibility
borne by the legislature for administrative performance. It also creates patronage opportunities.
“Based on these suppositions,” Van Thiel concludes, “it would seem logical to expect politicians
at all times to prefer quangos as executive agent, as opposed to government bureaucracy” (182).
Nonetheless, “[e]fficient and effective implementation of policies offers politicians the
opportunity to implement more policies from the same budget or reduce taxes[,]” both of which
play well with voters (182). Diminishing responsibility through quasigovernmental delegations
increases monitoring costs for politicians who “still need information to meet the demands of
their (now limited) political responsibility” (182).
2. For example, in 1916 the U.S. Congress established the Bureau of Efficiency as an
independent unit with a broad general power to survey and investigate organizations and their
methods of policy implementation, and to incorporate their findings in recommendations to the
President. Wrote Emmerich (1950, 21-22) of the bureau, “[d]uring its two decades of existence .
. . as the first federal agency of over-all administrative reforms it had an impact on the spirit of
self-improvement within the executive establishment that continued to be felt long after its
3. For example, a recent report to the U.K. Parliament states that “[t]he weight of evidence that
we have received convinces us that the existing arrangements for remunerating those who serve
on public bodies are a significant barrier to wider participation, especially among people who are
low paid or self-employed. Dame Rennie Fritchie told us: “Remuneration is a diversity issue.
About 80 per cent of public appointments are unpaid and I am told by many people that the lack
of remuneration and the inconsistent level of remuneration across different bodies are real
barriers to enabling a broad cross section of people to participate on public bodies” (House of
Commons 2003, ¶ 154).
4. Assisting individuals and organized constituent groups with access to government services for
which they or their members qualify is also a source of constituent trust (Fiorina 1977), which
rounds out the legislator’s tripartite role as educator, lawmaker, and ombudsman (Price 2000).
5. In a closed list system, each party selects a list of length equal to the number of seats in
parliament allocated to a given district. If the voters give that party, for example, 50 percent of
the vote in a ten-seat district, then the first five names on the list become MPs (Farrell 2001, 82).
In the Netherlands, since only one “district” exists, then a 50 percent outcome would send the
first 75 of the 150 names on the list to parliament, since there are currently 150 seats in the
Tweede Kamer (Second Chamber).
6. Kettl (2000) suggests that the question is open regarding the theoretical successors to
hierarchy and bureaucracy. Hill and Hupe (2002, 94-95) note that public administration had
traditionally been more reliant on organizational sociology than political microeconomics,
though the latter rose to prominence during the “rise of public management.” Indeed, modern
political economics—particularly the theory of delegation—sheds light on this problem.
7. On economic issues, for example, Alesina and Roubini (1997, 141-142) reason that partisan
differences “should be more prevalent in two-party systems where the majority has full control
of executive and legislative powers and less pronounced in where broad-based coalitions are the
rule.” Additionally, with endogenous elections (where the election date may be chosen, rather
than exogenously fixed), “[e]lections may be called early if the economy is doing well to
capitalize on the government popularity or postponed as long as possible if the ruling
government is unpopular” (Alesina and Roubini 1997, 142).
8. Though there were 545 public bodies on which the NCA collected some information, 23
clusters accounted for 406 quangos with 139 single organizations in the sample. That yields 162
observations. Of those, 143 were created between 1946 and 1993, when the MRG policy
preference data were available. The number of employees, a measure of the size of public bodies,
has 137 observations, but only 132 observations on board payment obligations. Thus the sample
size is 132.
9. The unit of analysis in the MRG data is the “quasi-sentence . . . the verbal expression of one
political idea or issue” (Budge, et al. 2001, 217). As the authors’ ultimate goal is the crossnational
study of political ideology, this permits a classification scheme (as in content analysis)
that “contains 56 different categories grouped into seven major policy domains: Each of the 56
categories sums up related issues in a way that changes over time can be measured across parties
and cross-culturally. Thus, the coding procedure comprises a quantification (how many
statements do parties make?) and a classification (what kind of statements do parties make?) of
election programmes” (Budge, et al. 2001, 217). Since my measure is of specific attitudes in the
Netherlands, the cross-national comparability issues do not arise presently.
10. Dutch elections occurred in 1946, 1948, 1952, 1956, 1963, 1967, 1971, 1977, 1981, 1982,
1986, and 1989.
11. Among the 56 available additional MRG measures, the following are relevant in addition to
administrative efficiency: anti-devolution of services from central to local government, prodevolution
(the inverse of the preceding measure), pro-free enterprise initiatives, pro-regulation
of private entities, pro-centralization of political authority, pro-business incentives. Some
measures are highly correlated. Anti-corruption and the devolution measures are correlated at ρ
= .65, and political authority with pro-devolution at ρ = .64. Administrative efficiency and
political authority are correlated at ρ = .85, while both are correlated with the free enterprise
variable at ρ > .62. Thus, it would be difficult to discern the significance of political attitudes
with regression analysis if all of these measures were included in a model given the presence of
such multicollinearity. An unreported model including the political authority, anti-devolution,
regulation and business incentives measures was estimated showing little significant impact of
these attitudes on financing choices. This is a likely result of the measurement error arising
from the use of these variables instead of the most appropriate measure, administrative
12. Seat share is calculated for the Tweede Kamer (Second Chamber). The total number of seats
increased from 100 to 150 in 1956.
13. In a multidimensional spatial bargaining model, Baron (1991) shows formally that
equilibrium policy proposals in parliamentary governments reflect the preferences of parties both
in and out of government, and particularly that “small parties may have a substantial impact on
government policy.” This reflects, for example, the intuition that given the knowledge of the
specific issue preferences of a government partner, a party with larger seat share may refrain
from programmatic reference to an issue that a smaller party has considered, yet that issue may
still have substantial impact on enacted policies. This notion is further borne out by the
experience of the manifestos coders in the saliency theory of party competition. Essentially,
parties write manifestos that appeal to their own niche (issue) constituency, not to crowd in on
other parties’ niches (the “repulsion” idea discussed above). Though all parties are included, the
measure, drawn from Baron’s equilibrium proposition regarding party influence on legislative
outcomes, places greater weight to large parties (and, ipso facto, those in government) and less to
others since it is based on seat share, which is quite intuitive.
14. In the Netherlands, a religious coalition of the Calvinist ARP, Catholic KVP, and Protestant
CHU fused into the CDA in 1980. This group of “confessional parties” dominated Dutch
politics throughout the period covered in the dataset. For example, anti-corruption MRG data
from 1956-89 show no statements from members of the governing coalition during that period.
With the exception of 1948, 1952, and 1971, this is likewise true for the anti-devolution
statements. Other parties carried these issues. Substituting these seat-share weighted coalitional
variables for Γ in the model in an unreported model leads to insignificant results for all but
administrative efficiency attitudes in the comparison between values of 2 and 3 for the dependent
15. Monte Carlo standard errors for constructing confidence intervals around the simulated
means presented in Figure 1 for the discretion categories by financial publicness regimes
(denoted as discretion:publicness) are as follows: (1:1) .08, (1:2) .01, (1:3) .07, (1:4) .02, (2:1)
.07, (2:2) .01, (2:3) .06, (2:4) .03, (3:1) .06, (3:2) .02, (3:3) .05, (3:4) .03, (4:1) .06, (4:2) .03,
(4:3) .05, (4:4) .03, (5:1) .07, (5:2) .05, (5:3) .07, (5:4) .04, (6:1) .08, (6:2) .09, (6:3) .08, (6:4)
.06, (7:1) .10, (7:2) .12, (7:3) .10, (7:4) .08, (8:1) .10, (8:2) .16, (8:3) .11, (8:4) .11.
Alesina, Alberto, and Nouriel Roubini. 1997. Political Cycles and the Macroeconomy.
Cambridge: MIT Press.
Algemene Rekenkamer. 1995. Zelstandige bestuursorganen en ministeriële
verantwoordelijkheid. The Hague: Sdu.
Anderson, J.A. 1984. Regression and Ordered Categorical Variables (with discussion). Journal
of the Royal Statistical Society Series B, 46, pp. 1-30.
André, Rae. 1985. On Beyond Quango: The Evolution of Business-Government Cooperation in
Northern Ireland. Journal of General Management, 11(2): 45-61.
Andreoni, James, Brian Erard, and Jonathan Feinstein. 1998. Tax Compliance. Journal of
Economic Literature, 36(June): 818-860.
Baron, David. 1991. A Strategic Bargaining Theory of Government Formation in Parliamentary
Systems. American Political Science Review, 85: 137-164.
Bertelli, Anthony M. forthcoming 2005. Governing the Quango: An Auditing and Cheating
Model for Quasi-Governmental Public Authorities. Journal of Public Administration Research
Bertelli, Anthony M. forthcoming 2006. Delegating to the Quango: Ex Ante and Ex Post
Ministerial Constraints. Governance.
Budge, Ian. 2000. Expert Judgments of Party Policy Positions: Uses and Limitations in Political
Research. European Journal of Political Research, 37: 103-113.
Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, and Eric Tanenbaum. 2001.
Mapping Policy Preferences: Estimates for Parties, Electors, and Governments, 1945-1998.
New York: Oxford University Press.
Chun, Young Han and Hal G. Rainey. forthcoming. Goal Ambiguity in U.S. Federal Agencies.
Journal of Public Administration Research and Theory.
Cox, Gary W. Making Votes Count. New York: Cambridge University Press.
Elgie, Robert and Iain McMenamin. 2004. Credible Commitment, Political Uncertainty, or
Policy Complexity? Explaining Variations in the Independence of Non-Majoritarian Institutions
in France. Working Paper. School of Law and Government, Dublin City University.
Epstein, David, and Sharyn O’Halloran. 1999. Delegating Powers. New York: Cambridge
Farrell, David M. 2001. Electoral Systems: A Comparative Introduction. Basingstoke: Palgrave.
Fenno, Richard F., Jr. 1973. Congressmen in Committees. Boston: John Wiley & Sons.
Fenno, Richard F., Jr. 1978. Home Style: House Members in their Districts. Boston: Little,
Flinders, Matthew V. 1999. Quangos: Why Do Governments Love Them? in Quangos,
Accountability, and Reform: The Politics of Quasi-Government, Matthew V. Flinders and Martin
J. Smith, eds. Basingstoke, UK: Macmillan Press, pp. 26-39.
Frant, Howard. 1997. Reconsidering the Determinants of Public Authority Use. Journal of
Public Administration Research and Theory, 7(4): 571-590.
Gilardi, Fabrizio. 2002. Policy Credibility and Delegation to Independent Regulatory Agencies:
A Comparative Empirical Analysis. Journal of European Public Policy, 9: 873-893.
Gladdish, Ken. 1991. Governing Europe from the Centre: Politics and Policy-Making in the
Netherlands. London: Hurst & Company.
Goodin, R. 1996. The Theory of Institutional Design. Cambridge: Cambridge University Press.
Greve, Carsten. 1999. Quangos in Denmark and Scandinavia: Trends, Problems, and
Perspectives. in Quangos, Accountability, and Reform: The Politics of Quasi-Government,
Matthew V. Flinders and Martin J. Smith, eds. Basingstoke, UK: Macmillan Press, pp. 83-108.
Hausman, Jerry and Daniel McFadden. 1984. Specification Tests for the Multinomial Logit
Model. Econometrica, 52: 1219-1240.
Hill, Michael. 1997. Implementation Theory: Yesterday’s Issue? Policy and Politics, 25(4): 375-
Hogwood, Brian W. 1995. The ‘growth’ of quangos: Evidence and explanations. Parliamentary
Affairs, 48(2): 207-225.
House of Commons. 2003. Public Administration—Fourth Report. London: United Kingdom
visited on January 3, 2005.
Huber, John, and Charles R. Shipan. 2002. Deliberate Discretion? New York: Cambridge
Hirst, Paul. 1995. Quangos and Democratic Government. Parliamentary Affairs, 48(2): 341-359.
Jolly, R.A. 1983. Quangos and Public Finance. Australian Journal of Public Administration,
Kaboolian, Linda. 1998. The New Public Management: Challenging the Boundaries of the
Management vs. Administration Debate. Public Administration Review, 58(3): 189-193.
Katz, Richard S. 1980. A Theory of Parties and Electoral Systems. Baltimore: Johns Hopkins
Kent, J.T. 1982. Robust Properties of Likelihood Ratio Tests. Biometrika. 69(1):19-27.
Kettl, Donald F. 2000. Public Administration at the Millenium: The State of the Field. Journal of
Public Administration Research and Theory, 10(1): 7-34.
Kettl, Donald F. 2002. The Transformation of Governance. Baltimore: Johns Hopkins University
King, Gary; Michael Tomz; and Jason Wittenberg. 2000. Making the Most of Statistical
Analyses: Improving Interpretation and Presentation, American Journal of Political Science,
Vol. 44, No. 2 (April): 341-355.
Kydland, Finn and Edward Prescott. 1977. Rules Rather than Discretion: the Inconsistency of
Optimal Plans, Journal of Political Economy, 85:3.
Langbein, Laura I. 2000. Ownership, Empowerment, and Productivity: Some Empirical
Evidence on the Causes and Consequences of Employee Discretion. Journal of Policy Analysis
and Management, 19(3): 427-449.
Lynn, Laurence E., Jr. 1998. The New Public Management: How to Transform a Theme into a
Legacy. Public Administration Review, 58(3): 231-237.
Lynn, Laurence E., Jr., Carolyn Heinrich, and Carolyn Hill. 2000. Improving Governance: A
New Logic for Empirical Research. Washington: Georgetown University Press.
Majone, Giandomenico. 2001. Two Logics of Delegation: Agency and Fiduciary Relations in
EU Governance. European Union Politics, 2(1): 103-122.
Mayhew, David. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.
McCubbins, Mathew, Roger Noll, and Barry Weingast. 1987. Administrative Procedures as
Instruments of Political Control. Journal of Law, Economics, and Organization, 3(2): 243-277.
Milward, H. Brinton, and Keith G. Provan. 2000. Governing the Hollow State. Journal of Public
Administration Research and Theory, 10(2): 359-379.
Moe, Terry M. 1991. Politics and the Theory of Organization. Journal of Law, Economics, and
Nerlove, Marc, and S.J. Press. 1973. Univariate and Multivariate Log-linear and Logistic
Models. Santa Monica: Rand.
Payne, Teresa, and Chris Skelcher. 1997. Explaining Less Accountability: The Growth of Local
Quangos. Public Administration, 75(2): 207-224.
Persson and Svensson. 1989. Why a Stubborn Conservative Would run a Deficit: Policy with
Time-Inconsistent Preferences, Quarterly Journal of Economics, 104: 325-345.
Peters, B. Guy, and Jon Pierre. 2000. Governance, Politics, and the State. Basingstoke:
Pollitt, Christopher and Geert Bouckaert. 2000. Public Management Reform: A Comparative
Analysis. New York: Oxford University Press.
Rhodes, R.A.W. 1997. Understanding Governance: Policy Networks, Reflexivity, and
Accountability. Maidenhead: Open University Press.
Salamon, Lester M. 2002. The New Governance and the Tools of Public Action: An Introduction
in The Tools of Government: A Guide to the New Governance, Lester M. Salamon, ed. New
York: Oxford University Press, pp. 1-47.
Segal, Jeffrey and Harold Spaeth. 1993. The Supreme Court and the Attitudinal Model. New
York: Cambridge University Press.
Skelcher, Chris. 1998. The Appointed State: Quasi-Governmental Organizations and
Democracy. Buckingham, UK: Open University Press.
Tabellini, Guido and Alberto Alesina. 1990. Voting on the Budget Deficit, American Economic
Review, 80: 37-49.
Thompson, Fred. 1993. Matching Responsibilities with Tactics: Administrative Controls and
Modern Government. Public Administration Review, 53(4): 303-318.
Van Kersbergen, Kees and Frans Van Waarden. 2004. ’Governance’ as a Bridge Between
Disciplines. European Journal of Political Research, 43: 143-171.
Van Thiel, Sandra. 2001. Quangocratization: Trends, Causes, and Consequences. Ph.D. Diss.
University of Utrecht.
Van Thiel, Sandra. 2004. Trends in the Public Sector: Why Politicians Prefer Quasi-Autonomous
Organizations. Journal of Theoretical Politics, 16(2): 175-201.
Williamson, Oliver. 1985. The Economic Institutions of Capitalism. New York: Free Press.
Description Mean Std. Dev.
1=quango-set tariffs, 2=gov’t
set tariffs, 3=gov’t financed
Task Discretion 1=paying benefits, collecting
fees, 2=registration, 3=
regulations and quasijudicature,
1=obligated to pay board
2=not obligated to pay board
Government declarations in
favor of administrative
Quango Size Scale 1=0 quango employees, 2=1-
24, 3=25-49, 4=50-74, 5=75-
99, 6=100-999, 7=1000+
Date of Creation Date the quango was created. 1973.986 15.019
Table 1: Description of Variables
Financial Publicness† Administrative
†Quango set tariffs (Revenue autonomy), Government set tariffs (Revenue Restriction), Government
financed board (Budget Autonomy), Government controlled budget (Budget Restriction)
p-values in parentheses
N = 132
X2(12df) = 129.16 (0.000)
McFadden LRI = 1-(LLFULL/LLINT) = 0.286
Table 2: Results of Multinomial Logit Regression Analysis of Financial Publicness Regimes
in Dutch Quangos, 1946-1993
Figure 1: Simulated Probabilities of Financial Publicness Regimes by Task Discretion