Open Journal of Political Science
2013. Vol.3, No.2, 85-89
Published Online April 2013 in SciRes (
Copyright © 2013 SciRes. 85
The Principal-Agent Approach to Politics: Policy Implementation
and Public Policy-Making*
Jan-Erik Lane
University of Freiburg, Freiburg im Breisgau, Germany
Email: janeriklane34@googlemail.c om
Received December 29th, 2012; revised February 12 th, 2013; accepted March 7th, 2013
Copyright © 2013 Jan-Erik Lane. This is an open access article distributed under the Creative Commons Attri-
bution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the
original work is properly cited.
The principal-agent models may be employed to elucidate central problems in interaction between prin-
cipals and agents in both policy implementation and public policy-making concerning performance and
remuneration. One then hits upon the double principal-agent relationships that are typical of the policy
cycle, from policy-making to policy implementation and back: 1) government as principal for agents in
public service delivery; 2) the population as principal for political agents under various forms of rulership.
Keywords: Principal-Agent Interaction; Asymmetric Information; Moral Hazard; Adverse Selection;
Bureaucracy; New Public Management; Rule of Law; Long-Term Contracting; Short-Term
Contracting; Looting
In a world of politics based upon perfect and symmetric in-
formation, an imaginary world of politics along Downs’ 1957
well-known model of two-party system competition in a presi-
dential type re gime, the de mos as principal would contract with
the correct politicians as agents, promising the making of the
policies that majority group in the principal favours. Politicians
would be paid a decent salary, somehow above their fixed res-
ervation price. But there would be no need for rents or qua-
si-rents. Politicians who performed better in terms of a goal like
for instance affluence or GDP growth would be favoured ahead
of politicians who performed badly, the re-election mechanism
doing the selection of agents for the demos.
Given perfect knowledge, voters would pick political agents
on the basis of the proximity rule, minimising the policy dis-
tance between their own political preferences and those of the
politicians in government and the legislature. Thus, radical
left-wing leaders like Lenin, Stalin or Mao would not be chosen
in free and fair elections, when the electore would know about
their hidden agenda. Ringht-wing leaders like Hitler, Mussolini
and France could only for chosen in the first election before
their hidden agenda had been revealed to the population, hav-
ing then experience the empthiness of national chauvinism.
Principal-agent games under perfect information would re-
semble the referendum democracy or a representative regime
with imperial mandates. In a two-party system in Downs’ mod-
el, the principal would elect the agents they want, pay them in a
straightforward manner as well as receive the policies they
prefer. With no rents, opportunism would not work. Besides,
one may assume that politicians in these games are driven by
vocation, as with Max Weber’s model of the ideal agent politi-
The same—first best solution or perfect equilibria—would
be found in the implementation of policies under perfect and
symmetric information. The government as principal would
hire a set of bureaux to do the job, i.e. provide a set of public
services. These bureaux would be paid a decent remuneration,
hired on long-term contracts with promise of predicatble pen-
sion in order to make sute they develop expertise on policy
matters. Given a sharp distinction between policy-making and
policy implementation, the bureaux would focus their attention
upon the means to achieve goals, maximising efficiency and
effectiveness in policy technology as well as neutrality in poli-
tics, along lines suggested by Weber’s ideal-type model of the
The introduction of asymmetric information changes all of
this, allowing for the search for rents or quasi rents by agents
engaging in opportunism.
The Principal-Agent Model
Two conspicuous features in the politics and policy-making
in the early 21st c en t ury include:
1) Policy implementation: The increased externalisation of
public services provision;
2) Policy-making: The increasing call for clean politics and
restraints upon politicians.
The first is linked with the success of the NPM reform
movement, which was based upon a critique of the classical
model of bureaucracy of Weber (Weber, 1978: pp. 956-1002),
whereas the second involves a rejection to the ideal model of
the politician, also launched by Max Weber in “Politics as a
vocation” (Weber, 1991: pp. 77-129). Both of these major
events in politics today may be analysed in terms of the per-
spective upon politics and policies launched by the princi-
*This article is based upon a talk at the University of Haifa with the
School of Political Sciences
November 2012.
pal-agent model. It asks pertinently:
1) Which agents are to be employed in policy implementa-
2) Can politicians as agents of demos be trusted, under what
rules of the game?
The principal-agent model identifies two major difficulties
when a principal contracts with a set of agents under asymmet-
ric information, namely moral hazard (hidden action) and ad-
verse selection (hidden knowledge). These two difficulties sur-
face whether the contract is explicit and enforceable in court, as
with policy implementation, or the contract is opaque and only
enforceable to a limited extent, as with politics.
The principal-agent model has been applied in various pri-
vate sector settings, such as the remuneration of CEOs, the
choice of contracts in agriculture and the client-lawyer interac-
tion (Rasmusen, 2006). When it is applied to politics, then one
must model a double principal-agent interaction, starting
backwards with first the government choice of agents who will
handle the provision of services and moving then to the choice
of the electorate of political agents with different policy pref-
The most simple principal-agent model analyses the interac-
tion between a risk prone principal and a set of risk avert agents,
where the former hires the latter on the basis of a contract in-
volving work effort, salary plus perks involving a basic quid
pro quo, whereby the agents are paid from the value of the
output they deliver. The agents may deliver low or high effort,
which has implications for the probabilities of low or high out-
put. As the principal aims for high output, he/she wants to write
a contract that elicits high effort. But all contracts are subject to
two basic principles that must be satisfied: the reservation price
of the agents on the one hand and incentive compatibility on
the other hand. With perfect information one may calculate first
best solutions that satisfy these requirements. However, given
asymmetric information—hidden actions and hidden knowl-
edge, one has to face suboptimal solutions. They are actually
well-known in the literature on bureaucracy and comparative
politics, although the language of the principal-agent model has
not been used.
The Niskanen model of bureaucracy with the public choice
school is a principal-agent model where the agents employ their
information advantage to supply a non-optimal level of public
services. In non-democratic politics, the rulers monopolize the
benefits in politics, sometimes reducing the population to a
form of political slavery (Burma), but always restrict the choice
of the electorate, in order to make looting easier. The opposite
solution, exploitation, is also feasible, for instance in agricul-
ture with powerful landlords (zamindars) employing indentured
labour or controlling sharecropping contracting (McLane,
Between these two extreme solutions, exploitation by the
principal versus looting by the agents, one finds all kinds of
varying solutions concerning both the value produced and the
division of the mutual gains from interaction. The output can be
either private or public services and the value of the output may
be calculated with market prices or the willingness of tax pay-
ers to pay. One application of the principal-agent model was
less focussed upon rent seeking and targeted more prestige. The
public choice model of public regulation claimed that it had a
fundamental credibility problem, as regulators would like to
deviate from the original policy intentions behind regulatory
Below, we focus upon the struggle between the principal and
the agents about the division of the monetary gains from coop-
eration. When agents are self-centred and do not refrain from
opportunism with guile, then what strategies can they employ
in order to get an extra payment using asymmetric information?
Agency Costs
The population as principal has to carry two kinds of agency
costs. First, there is the remuneration of political elites, both
pecuniaty and non-pecuniary. The direct costs may go very
high in non-competitive regimes, as looting is much more
probable in them. Second, there are the indirect costs from dire
agent performance or mistakes, which could go as high as the
complete loss of huge national assets or political territory. Po-
litical leaders may promise paradise on Earth, but accomplish
only destruction. “Mission accomplished”, declared President
Bush, only to have to face the strong insurrection in Iraq that
costs so much in terms of peoples’ life, American causalties
and misspent trillions.
First, politicians wish to have discretion on policies in tela-
tion to their voters, which is what asymmetric information pro-
vides. Thus, they claim that they have a general mandate from
the population to search for the policies that are in the national
interest. Politicians have their own agenda, open or hidden. The
election contract tends to be little specific, often general or
ambiguous. It is a fundamentally incomplete agreement, al-
lowing the politicians as agents much space to maneouvre in
relation to unforseen circumstances, or the contingencies. The
politicians may set up a multi-party system, which would make
coalitions possible beyond the horizon of their voters.
Given such opaque election contracts, the politicians would
wish to amass consierable amounts of resources to allow them
to operate freely, either as parties (partitocrazia) or as in-
depenents (political entrepreneurs). And they would wish to
maximise the resources for the conduct of their business, either
by state contributions to parties (Europe) or by means of soci-
ety contributions through e.g. the PAC system in the US. These
resources—spolia—will be used to remunerate or give favours
to the people who helped them win elections, using massive
propagande to convince the principal about their suitability as
political agents. Thus, party government emerges with massive
rent seeking where the political elite has a formidable knowl-
edge advantage over the population.
Second, under asymmetric information the implementation
agents will embark upon various opportunistic startegies that
increase their remuneration, pecuniary or non-cuniary ones.
Thus, bureaucracies and public enterprices exploit the budget
maximising strategy-shirking. The move of government to New
Public Management, i.e. tendering/bidding under a regime of
short-term contracts, will stop moral hazard in the public sector,
but it invites adverse selection as agents bid who are not trust-
worthy or reliable . Managing tournaments and auctions will be
as difficult as to monitor huge bureaux for government as the
principal of the public sector.
Agency costs, like e.g. the spoils of competitive politics are
motivated by the classical Burke theory of politicians as the
guardians of the general interests of the principal. But who
knows the true interest of the population—asymmetric infor-
Sometimes the rent-seeking ambitions of political agents
lead them to eng age in illegal activiti es :
Copyright © 2013 SciRes.
Tax evasion;
Kick backs and commissions on public contracts.
An economic rent is an unearned income, meaning an exces-
sive remuneration compared with what had to be paid, or it is a
payment in excess of the opportunity cost. The theory of eco-
nomic rents has concentrated upon monopoly profits, collusion
gains from oligopolistic competition, as well as rents for lob-
bying government to secure lucrative contracts or favourable
regulation (Tullock, 2005). Rents figure prominently in princi-
-pal-agent interactions, whether as pecuniary or non-pecuniary
advantages. The nature of the political regime—its basic struc-
ture of public law—affects how political agents maximise the
Non-Competitive Regimes: In a traditional or authoritarian
regime, like Saidi Arabia or North Korea, political agents tend
to take such a huge part of total output (GDP) that one is al-
lowed to speak of looting. Thus, the remuneration of the large
Saidi fami ly is immense, co mprising all kinds of income , perks
and capital assets, home or abroad. In North Korea, the entire
country has been mobilised to provide favours and advantages
to the clan of the ruling family. Charismatic regimes like the
Maoist rulership in China involve enormous agency costs, both
direct in supporting Mao and his entourage, as well as indirect
ones, resulting from mistaken policies.
Competitive Regimes: The agency costs for political regimes
that have open access tend to be much lower than those of
closed regimes. Thus, when political agents cause costly mis-
takes to the population, they will probably be voted out of of-
fice. However, also competitive regimes run with economic
rents for the politicians.
Politicians or political parties have an unsatiable need for
resources in order to maintain themselves and their advanta-
geous positions in society, as there are considerable costs of
competition. Thus, they employ all kinds of ways to bolster
their coffers: high salarier for representative positions, political
appointments in the bureaucracy, campaign contributions,
“cumul des mandates”, close connections with support groups
(industry, labour, and agriculture), public contributions to
newspapers affiliated with parties, subsidised educations
courses by affiliated orginisations, tax reductions for trade un-
ion fees and payments to think tanks, etc.
Value of Output, Remuneration and Rent
The agents hired by the principal to deliver a valuable output
must of course be paid somehow for their work and effort. The
payment of the agents comes out of the value of the output,
either directly through market prices or indirectly through taxa-
tion and public fees. All other things equal, the agent wants as
large remuneration whatever its forms as possible whereas the
principal remains the residual claimer, thus being interested in
maximising the difference between the value of output—agent
What the agent is paid in pecuniary and non-pecuniary forms
of remuneration depends upon their effort, the reservation price
and incentive compatibility, given asymmetric information.
The occurrence of conditions for moral hazard or adverse se-
lection opens up for strategies of opportunism on the part of the
agents, attempting to get hold of an extra remuneration, a
so-called rent in economic theory. Long-term contracting in-
vites the option of shirking, whereas adverse selection provides
for opportunities of pretending.
Policy Implementation: From Moral Hazard to
Adverse Selection
The NPM (New Public Management) revolution in public
administration emerged from a background of weariness with
big government, public deficits and government overload in the
mature welfare state. Intellectually, it was much inspired by
public choice theories, especially their criticism of long-term
contracting, as with bureaucracy and the public enterprises.
Both bureaucracy and the public enterprise were institutional
mechanisms that invited one key strategy from the agents,
namely shirking. Given asymmetric information, government
would be in a weaker position, having to take whatever cost
increases the bureau or public enterprise came up with. Both
the bureau and the traditional trading department captured an
excessive remuneration by expanding activities beyond what
was optimal, resulting in excessive number of employees and
costs. In reality, there was here a form of moral hazard, as the
risk ended up one-sidedly with government whereas the bene-
fits would be mostly with the bureau or public enterprise. In
long-term contracting, promises are cheap, and memory is short
concerning what was promised when outcomes turn up that are
NPM recommended short-term contracting as a strategy to
strengthen the position of the principal in relation to the pro-
viders of public services. Instead of bureaucracy and the tradi-
tional public enterprise, the principal would apply the following
mechanisms suitable for short-term contracting:
1) Outsourcing;
2) Tournaments;
3) Auctions;
4) Incorporation.
All these mechanisms involve the employment of tender-
ing-bidding that replaces the authority structure with the capac-
ity of government to exercise authority over its bureaux and
employees. Government gives up its power to plan and direct
the agents responsible for policy implementation in order to
buy specific services in accordance with a private law contract,
specifying performance as well as quantities and quality. The
public law governed budgetary process is undone, as the pri-
vate law contract is supposed to cover the most relevant con-
tingencies, including costs and service quantity and quality.
Government becomes a contractor, employing private law to
arrange for the provision of services. Public ownership is
transformed through incorporation into the holding of assets
through aktien, i.e. a private law institution. Government hires
agents to manage its capital assets in accordance with private
management principles, focussing upon rentability in the first
The move from long-term contracting to short-term con-
tracting entails that government has to struggle with the prob-
lem of adverse selection, i.e. how to figure out which agents
who are forthcoming to bid for all the government contracts
can be trusted. Government is bound to run into massive trans-
action costs when it moves from bureaucracy to tender-
ing-bidding, as conflicts arise about what has been agreed upon.
Contracting presupposes considerable time and effort for nego-
tiation and may involve substantial costs for enforcement as
Copyright © 2013 SciRes. 87
well as dispute settlement in courts or outside.
Nothing guarantees that government will be able to reduce
its overall costs or increase performance in service delivery,
when moving from long-term contracting to short-term con-
tracting. Just as a number of strategies can be used by agents to
shirk under long-term contracting arrangements, so alternatives
ways for agents to engage in opportunism exist under short-
term contracting, all allowing for agents to pretend they are
better than they really are. It may indeed be costly for govern-
ment to correct these pretending strategies, incurring switching
costs when turning to another set of agents or being forced to
pay more for unforeseen costs besides loosing court cases
where agents exploit badly written contracts.
The principal and the agents have common interests in the
agents delivering a big output of value, but agents will only try
hard when incentive compatibility is met by the principle.
Nothing prevents the agent from demanding a huge remunera-
tion for pretending high effort, i.e. to engage in the opportun-
ism of looting. When the traditional public enterprise is incor-
porated and deregulated, then the temptation of the looting
strategy may be irresistible. Thus, with incorporation the num-
ber of employees goes down and profits go up, but the salaries
of the key CEO:s tend to skyrocket, especially when large pub-
lic corporations turn to regional or global strategies outside of
the home country.
Policy Implementation: Hidden Actions and Adverse
When governments set out to deliver a set of public services,
they have to rely upon sets of agents. The classic model of
public administration outlined a set of bureaux with specialised
functions, accumulating expertise over time. However, under
long-term contracting agents have incentives to capture a rent
due to asymmetric information. This rent from shirking may
consist of X-inefficiencies or merely too much employment.
The controller of the bureau—the Ombudsman—would inves-
tigate the occurrence of hidden actions among the agents, i.e.
violations of the public law framework of bureaucracy.
Public sector reform during the last twenty years has turned
to short-term contracting to remove this rent, favouring exter-
nalisation of the delivery of public services, including the in-
corporation of the big public enterprises. However, with ten-
dering-bidding and short-term performance contracting come
adverse selection, which provides the agents with another type
of opportunistic strategies in order to capture a rent from pre-
tending. The principal has to assume considerable transaction
costs in order to handle the implications of adverse selection
Policy-Making and Asymmetric
The agency problems involved in hidden action and hidden
knowledge characterize not only the implementation of policy,
but they figure prominently also in politics as policy-making:
Can the demos as principal trust its politicians as their agents
for the making of public policies?
The distrust in political elites was theorized in the so-called
elite theory with the three Machiavellians: Pareto, Mosca and
Michels. They argued that politicians develop their own agen-
das in order to promote their special elite interests, capturing a
rent by means of opportunistic strategies.
Elite theory was much criticised by pluralist theory, mainly
by means of the argument that political elites tend to be plural-
istic in terms of both origins and composition. However, this
counter-argument does not take into consideration that political
elites from various backgrounds—social and ideological, en-
gage in hidden actions and employ hidden knowledge to in-
crease their remuneration in a broad sense of the term, covering
both pecuniary and non-pecuniary rewards.
Moral hazard pocketing any favourable outcomes—merely
luck, while placing the risk with the country or blaming
—unforeseeable negative results—occurs often in politics.
Similarly adverse selection is omnipresent, especially in elec-
tions where the demos has difficulties in evaluating electoral
promises as realistic or unrealistic ones. The rationale of con-
stitutionalism in politics derives from its contribution to undo
asymmetric information, revealing hidden actions and hidden
knowledge among the political elite.
Thus, the mechanisms against political opportunism are
handed down in the theory of rule of law, including:
1) Judicialisation of politics;
2) Referendum and recall;
3) Parliamentary opposition;
4) Federalism or political decentralization;
5) Civil society involvement;
6) Re-election of politicians: limits on tenure, primarie s, etc.
Constitutional democracy consists of a number of institutions
that counteract the implications of political elitism. These in-
stitutions divide up the set of political agents into competing
sets with the consequences that asymmetric information is re-
duced considerably for the demos. Opportunistic strategies
based upon hidden knowledge—bad politicians—or hidden
actions-illegal manoeuvres—trigger responses or counter
strategies. In a rule of law regime, bad politicians are—some-
times at least—eliminated and illegal practices punished. More
specifically, public law restrains the solutions to the princi-
pal-agent problematic through the following: Predictability:
Public law when properly implemented makes it possible for
people to increase the rationality of behaviour. They know
what rules apply, how they read as well as how they are applied
consistently. This is very important for the making of strategies
over a set of alternatives of action.
Transparency: Societies operate on the basis of norms pro-
hibiting, obligating or permitting certain actions in specific
situations. Rule of law entails that these norms are common
knowledge as well as that they are not sidestepped by other
implicit or tacit norms, known only to certain actors.
Due Process of Law: When conflicts occur either between
individuals or between persons and the state, then certain pro-
cedures are to be followed concerning the prosecution, litiga-
tion and sentencing/incarceration. Thus, the police forces and
the army are strictly regulated under the supervision of courts
with rules about investigations, seizure, detainment and prison
sentencing. No one can take the law into their own hands.
Immunities: People have certain rights against the state,
meaning that government faces definitive duties concerning the
protection of life, personal integrity and property. One may call
them the habeas corpus rights.
Counter-weighling Powers: Under the rule of law regime
there could be no single source of political power, or a hierar-
chical order of command. Instead, it favours multiple centres of
power, or pluralism. In terms of religion, it adheres to a secular
state based upon religious tolerance.
Copyright © 2013 SciRes.
Copyright © 2013 SciRes. 89
Separation of Powers: In order to have respect for the law as
the key instrument for governing society and regulating the
state, legislation, policy-making and implementation as well as
law adjudication must somehow be separated. Under rule of
law this separation of powers targets the political elite active in
the state, with the claim that it has to be divided into three dif-
ferent elites: legislators, governors or governments and courts.
These classic functions in the state cannot be exercised by one
and the same set of political elites. Separation of powers en-
hances checks and balances ingovernment as well as counter-
weighing powers.
Fairness: Public law comprises a number of mechanisms
that promote not only the legal order, or the law, but also jus-
tice, or the right. For ordinary citizens, the principle of com-
plaint and redress is vital, providing them with an avenue to
test each and every decision by government, in both high and
low politics. Here one may emphasize the existence of the
Ombudsman, as the access to fairness for simple people.
Politics is both policy-making and policy implementation.
The demos selects and instructs a set of agents—the politicians
—to come up with a list of policies that the demos prefer. Yet,
the politicians do not have time or expertise to put these poli-
cies into practice, which is why they—as principals—rely upon
a set of agents to deliver public services. Thus, one arrives at
the double nature of principal-agent interactions in politics.
One may interpret the move from bureaucracy to New Public
Management as the search for strategies that reduce the asym-
metric information advantage of bureaux. NPM helps against
shirking but invites another form of opportunism, namely pre-
tending. Adverse selection is not a major problem in long-term
contracting, as the principal has a general authority to direct the
work of the agents. But in short-term contracting, the selection
of bad agents can only be corrected by high transaction costs,
as failure to fulfil a contract will often be contested in court.
Shirking may of course also occur in short-term contracting.
The huge attention given to political scandals in present day
politics fits well into a principal-agent approach to elections
and policy-making. Politicians as elite have incentives to cap-
ture a rent by means of all kinds of opportunism. Only the rule
of law regime can counteract the consequences of hidden ac-
tions and hidden knowledge.
Burke, E. (1999). The portable e dmund burke. London: Pengu in.
McLane, J. R. (1993). Land and local kingship in eighteen century
Bengal. Cambridge: Cambridge University P ress.
Rasmusen, E. (2 006). Games and information. Oxford: Blackwell.
Tullock, G. (2005). The rent seekig society. Washington, DC: Liberty
Weber, M. (1978). Economy and society I-II. Berkeley: University of
California Press.
Weber, M. (1991). From max weber: Essays in sociology. B. S. Turner,
H. H. Gerth, & C. W. Mills, Edited. London: Routledge.