Public Administration & Public Management

The Principal-Agent Perspective

Jan-Erik Lane

• Publisher: Routledge
• Year: 2006
• ISBN:
  9781134199952
• Edition: 1
• Topic: Política y relaciones internacionales
• Subtopic: Política

A perspective on the public sector that presents a concise and comprehensive analysis of exactly what it is and how it operates.

Governments in any society deliver a large number of services and goods to their populations. To get the job done, they need public management in order to steer resources – employees, money and laws – into policy outputs and outcomes. In well-ordered societies the teams who work for the state work under a rule-of-law framework, known as public administration. This book covers the key issues of:

• the principal-agent framework and the public sector
• public principals and their agents
• the economic reasons of government
• public organization, incentives and rationality in government
• the essence of public administration: legality and the rule of law
• public policy criteria: the Cambridge and Chicago positions
• public teams and private teams
• public firms
• public insurance
• public management policy

Public Administration & Public Management is essential reading for those with professional and research interests in public administration and public management.

1 The principal–agent framework and the public sector

Introduction

The principal–agent framework (PAF) emerged in private sector analyses of human interaction that are more complex than the simple games studied in classical game theory (Gintis, 2000). The problem studied in the PAF is the following: If people interact longer than a one-shot game and if one party hires another as an expert of some kind, then how do they behave? This question is relevant to many private sector relationships that go beyond the simple adaptation in oceanic markets, including insurance, lawyers, agriculture (sharecropping), managing directors and health care (Rasmusen, 1994). How about public sector interaction? The application of the PAF to the public sector has not resulted in the same impressive list of new findings as with the private sector analysis using the PAF. The PAF comes up against the following perennial difficulties in approaching the public sector, such as:

• Who is the principal? The population at large or the elected politicians?
• What is maximised? The welfare of society or the private utility of the actors?
• Who is the agent? An entire organisation or single individuals?

It would perhaps be an exaggeration to demand that the PAF should clarify these eternal questions about the state, but one would like to see that the PAF moves the analysis one or two steps ahead when compared with other approaches to the public sector, such as public administration, public policy, public choice or management. The aim in this chapter is to suggest some clues as to how the employment of the PAF advances our understanding of the nature of the public sector, especially the logic of its organisations.

It is often stated that the PAF is merely another right-wing inspired theory advocating downsizing of the public sector. One should, however, distinguish between agency theory and the Public Choice School, as agency theory is neutral in relation to the preference between the public and the private. It argues basically that whether one employs public or private organisation, fundamental agency problems are bound to emerge in contracting.

Government as the principal may face different agents: a service-providing bureau in the soft sector, a public enterprise in the business sector and the regulatory agency. How are the contracting games with these bodies to be modelled?

Why is the principal–agent framework popular?

It would be too simplistic to merely refer to the advances in the application of the PAF to private sector interactions. Although economics is the master science and the directions it goes into spill over into the other social sciences, there must be an isomorphism between private organisation and public organisation for the PAF to work in the study of government. Perhaps it is the growing relevance of the democratic regime which creates this isomorphism? The democratic state may be modelled as a principal– agent interaction in two separate phases. First, there is the principal–agent relationship between the population and its government, channelled through the electoral arena. Second, there is the principal–agent interaction between government and its agencies or public organisations. Thus, government is both agent and principal in this simple model of democratic policy-making, modelled as responsive to its principal, the demos, as well as giving direction to its agents.

Democracy according to the PAF would be a regime under which the population as the fundamental principal of the state instructs government and pays taxes to the politicians as their agents who set up a public sector with a carpet of organisation, staffed by people working as agents of the government. Thus, democracy is simply a double set of principal–agent relationships. Any student of public sector complexity would reject such a model of government as unrealistic at worse or simplistic at best (March and Olsen, 1996; Lindblom and Woathouse, 1994). One does not have to go far into the literature penetrating all the corners and muddy waters of modern government in the advanced democracies to realise that the PAF comes high on parsimony but low on descriptive realism. Yes, its normative appeal is obvious from the point of view of democratic theory.

After the re-engineering of government in the 1990s one needs a conceptual framework for analysing what some scholars call ‘plural government’ and other scholars describe as ‘networking government’. To cut through so much organised complexity one would wish to employ a simple analytical approach which identifies the key players in an interaction and classifies them as either principal or agent. One does not have to agree with all of Milton Friedman’s ideas about government in order to endorse his recommendation to look for model simplicity as long as the predictive power of the model is satisfactory (Quine, 1992).

The popularity of the PAF derives to a considerable extent from the renewed surge in interest for game theory in all the social sciences. The key concept is strategy. It has certainly not been lacking from the writings of the great masters of public administration such as Max Weber and Aaron Wildavsky. However, the attention given to the concept of strategy and its immense elaboration with the Nash solution concept as well as with the idea of subgame perfect Nash equilibria is a new phenomenon in academia (Myerson, 1997; Fudenberg and Tirole, 1991; Ordeshook, 1986). It opens up a new systematic perspective on human interaction in general, the implications of which for government and its organisations in particular remain to be stated.

Suppose we regard government as a nexus of principal–agent interactions, then what new insights may we gain into bureaucracies, public enterprises and social security? It has been suggested that from the election emerges a principal–agent relationship between the voters and the politicians. This contract is, however, far more elusive than the contracts analysed in the PAF. Yet I shall look at the government provision of services in a broad sense as constituting a huge set of principal–agent interactions, stretching from the population to the agents over government as the main principal. I will confine the analysis to well-ordered societies, i.e. countries with democracy and the market economy (Rawls, 1995, 2001).

The agency framework is not an extension of the public choice approach to politics and administration, as the principal–agent model emerged from classical game theory and the economics of information (Laffont and Martimort, 2001). It can take into account the complexities of a public sector under plural government, as well as networking government, and it does not reduce behaviour to merely egoism, brutal or sophisticated. The PAF may be employed to illuminate the key ideas in public sector reform, such as the movement from public administration towards New Public Management (NPM). Why is there such an emphasis on entrepreneurial government and reregulation at the same time? May the PAF be employed to explain the main theories about the reorganisation of government in the early twenty-first century?

Principals and agents: contracting as the essence of interaction

According to the PAF, providing public services to the population is basically a contractual matter concerning the writing and enforcement of agreements between first, the leaders of government and an organisation responsible for service delivery, and second, the leaders of the organisation and single individuals. It makes little sense under the PAF to call the first contract ‘politics’ and the second contract ‘administration’. From an institutional point of view, these contracts may look very different from one country to another. Different forms of law are employed when setting up these organisations and funding them – compare, for example, unitary and federal systems of government. And the organisations responsible for service delivery may employ alternative forms of contracting with individuals – compare bureaucracy with a competitive/tendering regime. Although the importance of public law is great in shaping these contractual relationships between government and public organisations, it remains true that the interaction between them adheres to the logic of contracting. I do not deny that alternative institutional arrangements matter greatly for how government and public organisations interact; this is why it is important to research the consequences of alternative institutions. Public sector reform in the OECD countries has resulted in such a wide variety of public organisations that one has started to use the word ‘plural government’ in order to emphasise that there is no longer any simple structure prevailing in the public sector (Ferlie et al., 1996; McLaughlin et al., 2001).

Whatever form or structure the public organisation has from a legal point of view, it is the case that government cannot get anything done unless there is an agreement between it and the public organisation about service delivery, i.e. a contract. Thus, underneath the paraphernalia of laws and instructions, budgetary appropriations and regulations, there is an agreement between the principal and the agent about what to deliver and how to be paid – a contract. One may perhaps not wish to use such a private law term as ‘contract’, but that is what the interaction between government and its organisations comes to.

The contract is more visible in the second principal–agent interaction, namely, that between the organisation and its employees. Whether the public organisation insources or outsources, it will, in any case, handle lots of contracts with individuals of varying length of time and with different contents. There may certainly be framework legislation that structures these individual contracts, but contracts they still are. Thus, one may look on the public sector as consisting of numerous teams of people working for government under varying institutional arrangements. Instructing and monitoring these teams is what the PAF concentrates its analysis on. To put the matter bluntly: the civil service is a set of teams, namely the professionals who work in schools, universities and hospitals as well as public enterprises and regulatory agencies.

After establishing the fundamental importance of the contract between government and the public organisations in the midst of all institutional details, which may change rapidly when public sector reform is frequent, the PAF concentrates on predicting the basic problems in the contracts that will be forthcoming between the principal and the agent. Alternative contractual arrangements will have different real-life outcomes. The PAF covers simple contracts between government and one agency as in many regulatory programmes, but the PAF may also cover complex contractual arrangements typical of networking where government contracts with several agents. Whether there is insourcing or outsourcing, the basis is the contract between government as the principal and the organisation as the team of agents.

Incentives: how is egoism and social value recognised?

The PAF underlines the role that incentives play in social interaction in general and in principal–agent games in particular (Laffont, 2003). However, it is not restricted to modelling only narrow self-interests as with the Public Choice School. As a matter of fact, one may employ the PAF to pin down a key difficulty in public administration or public management, namely the objectives of the principal. Herein rests the great and still unresolved mystery of incentives in the public sector (Breton, 1996): What is the objective function of the government – public, private, social, selfish? The incentives of the agent are basically mundane, and modelled by two principles:

• The reservation price: The principal must at least pay the agent what he/she can get in other employment.
• Incentive compatibility: The principal must compensate the agent for trying hard.

The agent will be offered a menu of contracts which all satisfy his/her reservation price and some of them will pay a higher wage to the agent if he/she promises a high effort or commitment. The agent is always riskaverse, meaning that he/she prefers employment to being an entrepreneur. The agent maximises his/her utility by combining salary, which is a positive, and effort, which is a negative. What, then, is the principal maximising?

The principal will want to supply a range of public services or public programmes. He/she may do this for a variety of reasons, egoistic or altruistic. Be that as it may, the basic idea in the PAF is that the principal contracts with the agent to get things done. The principal is dependent on the agent to get the job of government done, while the agent is dependent on the principal to get paid for his/her services. Thus, the objective function of the principal definitely comprises public services or public policies. Little is gained by speculation about whether egoistic or altruistic reasons matter most to the principal. Without supplying public services or accomplishing public policies he/she cannot reach egoistic goals such as power, prestige and money.

The supply of public services or public policies is the quid pro quo between the politicians and the population, the citizens accepting to pay taxes or charges in exchange for a menu of public programmes. The principal will use this income to pay the agent for his/her services and to pay him/herself a decent compensation. The capacity of the principal to raise income is related to the quality of services provided, meaning that the value of the public programmes implemented is crucial to the entire equation, starting with the population and ending with the agent. And here is the difficulty: How does one estimate the value of the services in the public sector? This is the crux of the matter in any theory about the output and outcome side of government.

In traditional public administration the principal simply picks up the bill from the agent and tries to cover it through taxation and user fees, if not by borrowing money. In NPM the idea is that the principal should focus on value for money and attempt to maximise the difference between the value of the services to the population and the costs of the operations of the agent. Thus, NPM recommends a whole set of strategies to reveal and estimate the value of public services:

• benchmarking, total quality management;
• results-based budgeting, accrual budgeting;
• user charging;
• privatisation;
• outsourcing.

The value of public services or public programmes is that entity which drives the entire set of principal–agent interactions, from the population over government to its organisations. Yet, it is elusive because it is difficult to measure (Moore, 2003). Principal–agent theory predicts that there will be conflict over the distribution of the social value of the public services. Thus, the agents will want to pocket a large part of the social value, especially if they are instrumental in producing it. And politicians would want their share of the cake, either at the expense of the population or through a reduction in the costs for the agents. This conflict over the social value of the public sector typifies principal–agent games. How recurrent are they in the public sector? Only empirical research can tell. I shall discuss the main games in order to see what they can illuminate about the public sector.

Principal–agent games

In the PAF the key distinction is that between post-contractual opportunism and pre-contractual opportunism, the relevance of which needs to be spelled out in relation to the public sector. I will distinguish between three main games that surface in the public sector of advanced countries like the OECD nations, combining democracy with the market economy.

The bureaucracy game

No governmental institution has been more analysed than the bureaucracy. One may classify the major scholars of gove...