Incentives and Efficiency in the Museum Market The Case of Regulatory Reform of Dutch State Museums

Herman R.J. Vollebergha

Richard Laanb

Research Memorandum 9501

OCFEB Erasmus University Rotterdam Room H 7-25 P.O. Box 1738 3000 DR Rotterdam The Netherlands Tel.: (31) 10 4082430 Fax.: (31) 10 4089173 Email:

[email protected]

WWW:

http://www.eur.nl/few/ocfeb

The authors would like to thank all participants of the conference Kunstwereld ("World of Art") at the Erasmus University Rotterdam, April 23, 1993, and especially B. Frey, R. Goudriaan, M. Montias and W. Grampp for comments on an earlier draft of this paper.

a

Research Fellow at the Research Centre for Economic Policy (OCƒEB) of the Erasmus University Rotterdam.

b

Followed an apprenticeship at the Dutch Ministry of Welfare, Health, and Culture and recently got a Masters degree in Economics at the same University.

Abstract This paper applies principal-agent theory to the case of regulatory reform in the Dutch State museum market. The reform aims at improving the efficiency of museums like the Van Gogh Museum and the Rijksmuseum Amsterdam through incentive regulation. Principal-agent theory makes it possible to focus on the potential efficiency gains and losses to be expected from the reform. The paper finds that efficiency gains arise from financial deregulation, as this policy leaves much more flexibility for the museum management. In addition, however, efficiency losses might easily arise due to overregulation in response to problems inherent in monitoring museum outputs. The overall effect is thus ambiguous and crucially depends on the credibility of the principal and the risk attitude of the museums.

1.

INTRODUCTION

The government in the Netherlands is currently reforming its regulatory approach with respect to several of its most important museums. State-owned museums like the Van Gogh Museum and the Rijksmuseum Amsterdam (State Museum Amsterdam), were judged to have been administered rather inefficiently. The regulatory reform is the final stage of a process in which the Ministry of Welfare, Health and Culture (WHC) has gradually revised its view on public provision of museum services and is now of the opinion that that the common, detailed interference in the production processes of museums by the government itself is the ultimate cause of this inefficiency. The Ministry of WHC therefore decided to change its regulatory approach towards the musuems under its jurisdiction. This Ministry is responsible for only 21 out of the 51 so-called `State museums,' which means that they are owned by the central government. 1 Nonetheless this subset is responsible for almost 20% of all visits to museums in 1989 in the Netherlands, while they got more than 30% of the public budget for all museums (public and private). The reform aims to improve the efficiency of providing museum services. This goal should be met by using output regulation instead of input regulation. In general the idea behind this is to improve efficiency using less rigorous budgetary rules for the museum management. A rise in managerial freedom would improve the museum's services to the public. Output regulation should provide control for the government over its budget. We evaluate whether the regulatory reform in the Dutch State museum market will contribute to this aim. Thus the potential improvement of museum services is our yardstick, rather than the overall goal attainment of art policy of the Dutch government. We evaluate this change in regulatory policy using principal-agent theory (see e.g. Harris and Raviv, 1978 and Sappington, 1991), taking the government as the principal and the museums as the agents. We show that the expected efficiency gain of a regulatory approach for agents with difficult-to-monitor outputs such as museums is likely to be overstated. Section 2 describes briefly the principal-agent framework for the government and museums and provides some background information about the Dutch museum market. Next, section 3 describes occasion to and plan of the regulatory reform by systematically exploring asset constraints, following from the ownership structure, budget constraints and the regulatory constraints. Section 4 evaluates the regulatory reform, given the specific problems of incentive regulation with respect to museums. Finally, we draw some lessons from this case study for both policy and theory.

1

These 21 `State museums' do in fact comprise 17 museums, one agency for advice and control (Rijksdienst Beeldende Kunst), one professional training school for restorers, one bureau for art documentation (Rijksbureau voor Kunsthistorische Documentatie), and one counseling agency for conservation and restoration (Centraal Laboratorium).

The Case of Regulatory Reform of Dutch State Museums

1

2.

PRINCIPAL AGENT THEORY AND THE MUSEUM MARKET IN THE NETHERLANDS

What is the best method to assess the changes in the regulation of the State museums in the Netherlands? As we noted in the introduction, we use principal-agent theory as our basic theoretical framework to evaluate efficiency effects of regulatory reform. Obviously, the government is the principal, while the museums are the agents. In an agency relationship the principal has to design a contract that motivates the agent to maximize the principal's welfare. An important assumption is that agents will be driven by self-interest, so that the actions choosen by the agent do not necessarily reflect the interest of the principal. Now consider a typical principal-agent problem. The principal depends on a specialized agent for producing an output in which the principal is interested. Therefore the principal has to define the output and a compensation or incentive scheme for the agent (the `price' of the output) to produce the output. Principal and agent have to settle on a contract in which they decide on the output, its price and what to do in the case of unforeseen events. The settling of such contracts is usually aggravated due to all kind of problems. In a world of complete and costless information, the principal would not have a problem of monitoring agents pursuing other goals and using discretionary slack for their own benefit. The principal could simply fix an optimal contract ex ante with no efficiency loss provided that he solves the motivation problem appropriately. This is suggested by the `canonical case' solution (Sappington, 1991, p. 47ff). Here, a risk-neutral principal and agent with similar beliefs about a critical random productivity parameter (e.g. rainfall), agree on a contract that makes the agent the residual claimant in the relationship. This ensures the two ideal outcomes for the principal: i) for each realization of the critical productivity parameter the agent chooses its efficient level of effort and thus maximizes the expected surplus; ii) the principal collects the entire surplus. This is realized with a so called `franchise fee' paid by the agent for the right to work for the principal and the fee is set equal to the expected total surplus from efficient operation. However, the world as we know it is less transparent. First of all, knwoledge about performance is not always publicly available. Monitoring is costly, and difficulties can arise if outputs are less tangible or jointly produced. In such cases the principal faces problems of cost observation and might therefore find it difficult to observe performance. Moreover, if agents are risk-averse, as is often supposed, then this would render the first-best contract suboptimal. Under the assumption that the principal is risk-neutral while the agent is risk-averse, principalagent theory shows that an efficiency loss arises if the agent is bearing any risk. On the other hand, such a loss may be necessary to provide the agent with enough incentives to do his share. As a consequence the principal has to pay an extra risk premium to motivate the agent. Furthermore, there might not be a symmetry of pre-contractual beliefs between principal and agent with respect to uncertain events in the future. Finally, commitment might not be perfect, meaning that both principal and agent might, for instance, break promises. 2

Incentives and Efficiency in the Museum Market

In general, any contract diverging from the canonical case will imply some welfare loss for the principal compared to what he would get from the first-best contract. In fact, the principal has to calculate all benefits and costs of regulatory reform and realize that first-best contracts are not always possible. The principal might then aim to minimize his welfare loss by devising incentive schemes that are appropriate for the specific circumstances that apply for particular markets. The principal could, for instance, benefit from investment in additional information, e.g. by comparing agents or organising auctions; he could also introduce more detailed contracts or claim to be involved in strategic decisions about the production process. However, such actions are costly for both the principal and the agent. These so-called `monitoring and bonding costs' (Jensen and Meckling, 1976) suggest that the optimal contract for the principal is the contract in which the marginal cost of monitoring and bonding activities are equal to their marginal return, i.e. the marginal reduction in welfare by an extra monitoring or bonding activity. How does this theoretical literature apply to the museum market in general and to the Dutch State museums in particular? As in most developed countries, the museum market is not insignificant in the Netherlands. The Netherlands is a small though densely populated country of 15 million people. Table 1 shows the structure of its museum market. In 1989 the market as a whole comprised 667 museums, providing work of 5,521 man years, while the museums were visited over 20 million times. Government influence on the supply side of museums is paramount, although only 192 out of 667 (29%) are publicly controlled (State, Federal, Municipal) (CBS, 1991a, p.17). The other museums have an independent legal form. Nonetheless, even those 475 (71%) privately controlled museums depend, on average, for over half of their budget on governmental subsidies (51%). From the 192 publicly owned museums, 51 are State museums, which indicates the legal fact that the central government owns the assets and is responsible for their management. As noted in the introduction, the subset of 21 State museums (on which we focus in this article) are the State museums administered by the Ministry of WHC. Among them are world famous museums like the Rijksmuseum Amsterdam, which exhibit among other artwork many paintings of Rembrandt, and the Van Gogh Museum with its unequaled collection of Van Gogh paintings. Less known but, due to its specialization, also internationally respected are some cultural-historical museums like the Koninklijk Penningenkabinet (history of coins), the Scheepvaart Museum (history of navigation) and the Zuiderzeemuseum (history of the emergence and maintenance of protection by dikes). In general, contemporary art is not considered to belong to the working assets of the greater part of the State museums. Presumably the Kröller-Müller Museum is the only exception. The main characteristics of the State museums of WHC compared to the others, both public and private, is to be derived from table 1. The table shows that the 21 State museums of the Ministry of WHC are few in The Case of Regulatory Reform of Dutch State Museums

3

number, though relatively large, generating a large number of visits, while getting much public support from the government. What does the principal, in this case the government of the Netherlands, aim for by supporting museums with public funds? Generally the government itself refers to their art policy goals, such as the spread of culture among the population and the preservation of the common heritage. Such policy goals, however, are difficult to assess. Moreover, these goals are not at all operational for understanding the production process of the museums. Instead we follow Bradford, Malt and Oates (1969) in their seminal paper on the nature of public outputs. They distinguish between services directly produced ("D-output") and the thing or things of primary interest to the representative citizen-consumer ("C-output"). In the case of museums, exhibitions and the preservation of works of arts via restorations could be understood as Doutputs, while the spread of cultural heritage and its preservation seems best understood as Coutputs. The latter not only depends on D-outputs produced, but also on other things like the weather situation or the so called `random variable' in the principal-agent model. Table 1: Profile of Dutch Museums, 1989 Museums according to ownership

Number Labor force (man year) Visits (mln) Exhibitions Public support (in mln Dfl)3 Total expenses (in mln Dfl) Funding ratio4 (%) Source:

Public Museums State Museums (WHC)

Other Public Museums1

Private Museums

All Museums

Abs

% of total

Abs

% of total

Abs

% of total

21 1156 3.6 n.a. 104 123 85

3 22 18 n.a. 33 28

171 1802 5.5 9432 144 178 81

26 35 27 492 45 41

475 2263 11.0 963 70 137 51

71 43 55 51 22 31

667 5221 20.1 1906 318 438 73

Figures for all museums are derived from CBS (1991b) and for State museums from Ministry of Welfare, Health and Culture (1989a and b)

Notes: 1

2 3 4

4

Excluding all 21 State Museums of the Ministry of WHC but including other State, provincial and municipal museums; Figure for all Public museums, including 21 State museums; Excluding Public support via Agency for State Buildings (ASB); Funding ratio is the share of government subsidies in the total expenses of the museums (also excluding ASB).

Incentives and Efficiency in the Museum Market

Table 2: Developments in the State museum market

Number of visits Index Admission Income (1000 Dfl) Index1 Admission per visit2

1980

1984

1988

1991

2,996,620 100 4,593 100 1.53

3,548,900 118 12,827 227 2.94

3,510,199 117 12,828 221 2.88

3,790,228 126 n.a. n.a. n.a.

Source: Ministry of WHC (1989) and Annual Budget of WHC (1992), page 206 Notes: 1 2

Index of real revenue (deflated by consumer price index); Real admission per visit (real guilders of 1980).

One could infer revealed preferences for the operationalized policy goals of the government from the developments in the State museum market. Table 2 shows a gradual rise in the number of exhibitions and visits in the last decade, which is quite remarkable given that the real price of a museum ticket has doubled in the same period. The number of visits peaked at 4,951,726 visits in 1990 due to two major exhibitions: the well-known Van Gogh manifestation in the Van Gogh Museum celebrating the centennial memorial of Van Gogh's dying-day and a Dinosaur exhibition in the Museum of Natural History. Because the growth in inputs did not keep up with this trend, one could deduce that policy emphasis had been on `exhibitions' and not preservation of works of arts over the last decade. This conclusion can also be confirmed by the growing concern about the current condition of the works of art found in policy papers and the estimated overdue maintenance of the current stock adding up to 4 billion Dutch guilders.

The Case of Regulatory Reform of Dutch State Museums

5

Table 3: Expenditures and Revenues of Dutch Museums, 1989 State Museums (WHC)1

Expenditures - Wage expenditure - Collection investment - Material and investment Total Expenditure Revenues - Government funding - Earned income: * Admissions * Auxiliary income - Private support income - Other Total Revenue Source:

All Museums

Dfl. (mln)

%

Dfl. (mln)

%

67 4 802 151

44 3 53 100

204 21 215 440

46 5 49 100

132

87

318

72.5

14 2 3 151

9 2 2 100

58 20 20 24 440

13 4.5 4.5 5.5 100

Figures for State museums derived from Ministry for WHC (1989) and for All Museums from CBS (1991b)

Notes: 1 2

Only the 21 State museums in the regulatory program are taken into account; Including investment from the Agency for State Buildings at Dfl. 28 million.

Before we go into detail about the regulatory changes in the State museum market, we offer a final insight derived from cost observations of the agents (Laffont and Tirole, 1986). Table 3 summarizes the different income and expenditure sources for the 21 State museums of the Ministry of WHC compared to all museums using the overall balance sheets of the museums.2 According to Table 3, before the regulatory change, the input mix of the State museums is on average only slightly different from other museums. State museums seem to spend less on personnel and collection investment but some more on material and investment. On the revenue side, the State museums have over 87% of their budget financed by the central government, while only 13% is earned in the market. However, other museums in the Netherlands, even the private ones, depend for at least 50% of their budget on public funding (see table 1)! Such simple facts reflect the diversity in financing cultures in the Western world, as, for instance, in the US only the Federal Museums show comparable figures (Rosett, 1991, p.142). 3.

THE REGULATORY SYSTEM AND ITS REFORM

2

These figures must be interpreted with care. The figures on State Museums are based on information provided by the Ministry and represent a recalculation of figures presented in the National Budget. These are, strictly spoken, not cost but expenditure data while they are organized according to public accounting rules. Therefore they might deviate from the cost categories as measured for All Museums (including also the 21 State museums) by the Dutch Central Bureau of Statistics.

6

Incentives and Efficiency in the Museum Market

The regulatory reform of the public museums will have major consequences for museum management. In the current regulatory system of the State museums, a museum manager is like an administrator having little degree of freedom with respect to his managerial policy. This is due to the fact that the State museums are part of the Ministry of WHC. Thus the agents are simply subject to hierarchical control from the government, while the minister of WHC is formally responsible for the management of the museums. Consequently, many budgetary and regulatory constraints for the agents directly follow from general rules concerning the way in which public expenditures are governed by the central government. One could assert as a general rule that the old system more or less follows an implicit governance structure between principal and agent, while the new regulatory approach aims at explicit contracts for governance between the government and its agents, the State museums. This new system substitutes financial control through input regulation for output regulation within an adapted financial framework. Following Frey and Pommerehne (1989, pp. 61-77) we present the regulatory reform with reference to the different budget constraints a museum manager faces now and in the new system. The budgetary constraints come from both the financial support by the government using the budgetting process and the rules of the government with respect to ownership and allowances for obtaining income from third parties. 3.1

Ownership and Asset Constraints

The assets of the State museums in the Netherlands, notably the buildings and the collections, are owned by the State and held in trust by the museums. With respect to this important constraint, no change is provided in the regulatory reform. This continuation of the status quo implies for the buildings that the museum management will remain completely dependent on the Agency for State Buildings (ASB). Here priorities are set at a very general level, weighing funds for investments between, for instance, new prisons or maintenance of light houses, to mention only two extremes. The Ministry of WHC or the administrative agency responsible for the State museums has to negotiate with the ASB about priorities in the allocation of their budget. In short, a particular museum directorate depends for its expansion or maintenance on time consuming negotiations by others. Only one change is provided here in the new regulatory scheme.

3

After privatization the museums no longer get their buildings in kind, but they

actually rent the building from the ASB on the basis of an explicit contract between museum and ASB. This contract regulates payments by the leaseholders (for a total amount of Dfl. 10 mln. a year), and creates obligation for maintenance by the executive office of the owner (ASB). Obviously, this strenghtens the position of the leaseholder.

3

Note, however, that the current system does not prevent particular solutions for space constraints, although they remain an exception. Recently the Van Gogh Museum was allowed to expand using funds provided by Japanese sponsors.

The Case of Regulatory Reform of Dutch State Museums

7

State ownership of the collection also continues. Therefore the directorate must negotiate any sale or destruction of a particular work of art. Moreover, the sale of works of art is allowed only by exception. For acquisitions the museums receive only a small budget (Dfl. 3.3 mln. or 2% of total cost in 1992). This is in accordance with the stated objective by the Dutch government mentioned before that it is now caring more about conservation of the current stock of works of art in those museums because concern has grown with respect to the conservation output of the museums. In any case, the government still assumes that care about the common heritage of the past, whether it is through conservation or presentation, is best served by public responsibility for such outputs. Therefore both outputs will remain subject to ministerial responsibility in the future. 3.2 Budget Constraints: Income and Expenditure Although the Ministry of WHC has promised that the amount of public funding will remain constant in nominal terms over the next few years, the budget restrictions facing the museums are subject to considerable revision. Changes in the rules governing the sources of income and the allocation of the budget over different sources of expenditure will become less rigid then before. The different sources of income to be discussed here are government funding, earned income and income from private support income, respectively. In order to understand the regulatory changes properly, one should realize that the State museums are simply part of the Ministry of WHC in the old Dutch regulatory system. Therefore, with a few exceptions that will be discussed later, any revenue or expenditure of the State museums should be accounted for at the central level. Direct government funding According to the prevailing budget rules in the Netherlands, laid down in the so-called `Comptabiliteitswet', the principal allocates public money to the musuems using earmarked budgetted funds. Governance is organised by way of ex ante control of spending through input regulation and ex post checks on the legal justifiability of expenditures. In fact, this implies that the overall budget of a specific agency is funded by different Ministries, while there is no managerial freedom with respect to the allocation of those budgets. In general, earmarked government funding comprises three elements: i)

budget for wages as accounted for at the Ministry of WHC (P-budget);

ii)

nonwage budget including expenses on materials and acquisitions (M1 -budget);

iii)

nonwage budget for maintenance, renovation and enlargement of buildings (M2 -budget).

Table 4 summarizes this regulatory system decomposed with respect to the different Ministries that are financially involved with respect to the State museums. The figures in table 4 show not only which Ministries are financially involved in the State 8

Incentives and Efficiency in the Museum Market

museum business, but also how the money should be spent by the museum management. The three different elements of the budget ask for three different `channels of decisionmaking.' In the old system both the P-budget and its allocation were decided upon by a central administrative agency for Personnel (DGCZ) in the Ministry of WHC. Decisions about the number of staff, payment of salaries and even recruitments of new personnel were thus centralized, far away from the museums and even from the directorate responsible for the output of the museums. Especially lack of recruitment facilities at the decentralized level often caused misunderstandings and slow fulfillment of vacancies. Moreover, the personnel of museums belongs legally to the Civil Service, which implies that all terms of employment (e.g. salary schedules) are decided upon by the Ministry of Internal Affairs (BIZA). As a result flexibility in personnel management is low, and often museums resorted to employ temporary and even voluntary personnel via the material budget (M1-budget) in the past because that was allowed by the budgetary rules. Table 4: I.

Financial Flows between State Museums and Government in 19891

Financial flows from the central government to state museums

II.

- Ministry of WHC * Wages (P-budget) * Collection investment (M1-budget) * Nonwages (M1-budget)

67 4 34

- Ministry of VROM (ASB) * Building (M2-budget) * Investment (M2 -budget)

18 28

Total

151

Financial flows from state museums to central government

- Ministry of Finance (Treasury) * Admission * Lease

Total

14 P.M.

14

Source: Ministry of WHC (1989) Note: 1

It is impossible to calculate the net financial flow of the government to the museums, for some amount of revenue raised by other sources of income is not shown in the table (compare table 1 and table 2).

One part of the M1-budget includes acquisitions. This was decided upon separately from the other part comprising operating expenses, such as the expenditure on furniture, materials for restoration and the organization of an exhibition. Occasionally museums are allowed to buy specific works of art, such as a recent acquisition of a Rembrandt by the Rijksmuseum Amsterdam. In that case the total budget of acquisition is raised incidentally (Dfl. 7 mln. in 1991). As noted before the budget for acquisition is quite low, and museums often complain about their inability to expand the collection. The last part of the budget, the M2-budget, has been discussed before, because it is closely related to the ownership structure of the buildings. In the new system both the P-budget and the M1-budget will be put together in one lump sum The Case of Regulatory Reform of Dutch State Museums

9

budget afforded by the ministry of WHC. Moreover, museums become free to allocate their budget (see next section), although the budget for acquisitions is excluded from these new rules. The terms of employment for the museum staff will not be dictated by the ministry of Internal Affairs (BIZA) any more because the personnel will no longer belong to the Civil Service. The M2-budget as such does not change, although, as noted before, the explicit lease contract facilitates claims on the ASB. One exception is allowed in the general rule that the new regulatory scheme does entail higher nor lower budgets. This is a once-only shift of a part of the overhead working at the Ministry of WHC towards the museums and a once-only facility for improving the museum management (Dfl. 4 mln.). The increased museum staff is supposed to be necessary for new tasks such as personnel management. Earned income As shown in table 3 about 11% of total revenue of the museums is an income flow from people visiting the museums, buying particular museum products or using the building for special events. Moreover, table 4 shows that the revenue from admissions has to be remitted to the Treasury. Therefore no explicit incentive exists to maximize the number of visits, although the policy of the principal with respect to the entrance fees could be interpreted in such a way. As a general rule, fees are used as an instrument of art policy by the principal. In charging low fees museums were supposed to contribute to the dissemination of culture among all income strata. Thus a particular museum was not able to decide upon its entrance fees on its own. In the new system, the budget is lowered once-only with the amount of revenues from admission fees, leaving responsibility for the entrance fee and its revenue with the museums. Thus the museum is allowed to experiment with respect to its pricing policy. In the past museums were creative in using auxiliary income from shops and restaurants for their own operating expenses. The establishment of private foundations enabled them to keep control over this part of income. Moreover, the cost of using its own staff providing services was usually not passed on to the foundations. In fact this amounts to internal subsidization, which was well-known and allowed by the Department of WHC deciding on the budget. Such practices are no longer necessary in the new regulatory scheme. Like the revenues of admissions, museums may dispose freely of the revenues from commercial activities. A third source of earned income results from leasing the building to third parties for particular events. Museums have to share revenues with a particular agency of the Ministry of Finance, which could be understood as a payment in return for using State owned buildings. Until now no new contract has yet been decided upon, although it seems that the government is willing to give up its share in the revenue. However, not all activities will be allowed. In particular a Board of Governors will become responsible for preventive control. 10

Incentives and Efficiency in the Museum Market

Income from private support income A last source of income museums acquire from membership dues, funds and sponsors. The total amount of money arising from these sources is still quite low compared with, for instance, museums in the US. On average only 2% flows from these sources in the Netherlands, while private support income of a museum like the Museum of Modern Art in New York accounts for 20% compared to government funds of only 3% of its operating expenditure (Frey and Pommerehne, 1992, p.69). Nonetheless the overall amount of private support income seems to be ont the rise in the Netherlands, although incentive mechanisms such as tax deduction have not yet been envisaged. On the other hand, information on this source of income is unreliable and difficult to obtain in the Netherlands, because sponsors sometimes provide services in kind or provide funds for particular exhibitions with separate balance sheets. The new regulatory scheme will lift all rules with respect to these sources of income, which might provide incentives to exploit these sources of income more explicitly. 3.3

Administrative Constraints

The way in which an agent is able to choose among different possible categories of expenditures is also important for judging degrees of managerial freedom of agents. In the old system the musuem directorate has to follow the earmarked budget rules applying to public government agencies in general. Thus, the directorate could not shift money between budget categories like the P-budget and M-budgets, as mentioned before. Moreover, shifting money from one year to another is prohibited and investment is only considered from an expenditure point of view.4 This is obviously contrary to elementary principles of cost-based budgetting. As touched upon already in the preceding section, the regulatory reform program allows the museums to dispose of their operating budget much more freely. This is mainly due to the fact that the classic budget rules for the public sector no longer apply to these erroneously called `privatized' agencies. Museums are allowed to shift money from one year to another, attract private capital and make reservations for investment. Thus the agent could decide on a much larger part of its own production function such as expenditures on staff and materials. This flexibility facilitates, for instance, decisions on personnel management including task allocation, wage classification, bonuses and education. If the directorate thinks it necessary to improve terms of employment, it could use its own judgement, even if this implies a larger budget for personnel than is paid for by the regulatory agency. The museums are also allowed to use their inputs for producing outputs that are not part of the explicit output contract with the regulatory agency (see below). The new rules, however, do not apply across the board. Exceptions are the budget for

4

Cost-based accounting is uncommon in public agencies.

The Case of Regulatory Reform of Dutch State Museums

11

acquisitions (former part of the M1-budget) and buildings (M -budget). The museum 2 directorate still has no juridiction over its allocation. The budget for acquisitions is set apart and remains at the discretion of the Ministry of WHC. The principal therefore is able to allocate money from this budget differently across museums and between years. On the other hand, museums could ask for the money not to be spent on acquisition, but for other outputs such as conservation. Obviously, the exact amount of public funding and especially its allocation remains part of the negotiation process between the Ministry and its agents, especially with respect to the earmarked budgets. Although input management using financial regulation is deregulated in the system, the principal reregulates through a specific form of output management of the agents. In theory the museums have to define their outputs, calculate their cost and sell them to the principal. In practice output management entails so-called `task budgets,' written down in explicit contracts for four years between the Ministry of WHC and the museums. Depending on the objective function of the principal, the principal's preference for a certain composition of outputs is specified and the agents have to `sell' their output to the principal for funding on the basis of cost calculations of outputs. According to the new explicit contract, the `products' or outputs of the museums are organized in three groups and quantified as much as possible: i) collection management, ii) preservation, iii) presentation. Collection management includes acquisitions, loaning, and alienation of works of art. Preservation includes all activities aimed at maintaining a good quality of the collection. Activities include conservation, restoration, documentation (including scientific research), making inventories and registration. Presentation of works of art for the general public is thought to be the third output category. It is supposed to contribute to the dissemination of culture, realized through the organization of exhibitions (general and specific), maintaining a library, organizing workshops, guided tours and courses. The explicit contract between principal and agent, however, is related not only to the outputs but also to the activities of the agents. It does in fact specify in great detail task forces of the museums with respect to the outputs and activities. The contract specifies numbers and, if possible, quality of certain products and activities.5 In order to get public support, museums have to develop `business plans' in which they specify medium-term policy plans. This is converted into short run contract offers by the museums. If museums do not produce the outputs, the principal may refuse to pay the agreed amount of support. Monitoring will also be reorganized. In the future an inspection officer will check compliance with the management contract. Secondly, external accountants are made responsible for control of annual reports. Finally, a Board of Directors appoints and supervises

5

12

Thus the principal not only regulates outputs but also effort by means of its contracts.

Incentives and Efficiency in the Museum Market

the museum director, although the Minister should be consulted for specific appointments of the director. 4.

EVALUATION

The regulatory reform of State museums in the Netherlands provides an interesting case-study for assessing recent changes in public management. As noted in section 2 a principal should adapt its contractual approach to the specific circumstances prevailing in a particular market. The principal should carefully calculate all benefits and costs of regulatory reform while realizing that first-best contracts might not always be possible. How can we evaluate the new regulatory system for the Dutch State museums to be evaluated, given the lessons of principal agent theory? Can the principal be expected to improve its welfare in the new system? As noted in the introduction, the aim of the reform is to improve the efficiency of the museum services. In order to realize these aims, the reform of the regulatory scheme exploits two principal means. The first one is decentralization in order to improve degrees of freedom for the museum directorate and the other is improvement of value for public money using public funding as incentive regulation. We simply concentrate on the question whether the regulatory reform in its current form will contribute to the aim set by the principal. Decentralized decisionmaking With respect to decentralization the overall balance cannot be but positive. As was demonstrated in section 3, many rules badly geared to the specific situation of the museums will be abolished. The constraints described before reflect the existing complicated governance structure between the different central bureaucracies and the executive agencies in the Dutch public sector. In the case of the State museums, no less then four ministries are directly or indirectly involved. First, the formal responsible ministry, the Ministry of WHC, has to allocate several budgets to the State museums. Note, however, that within this Ministry these budgets are also separated between different internal departments. Second, the museums have to follow the rules concerning employment as notified in the Employment Act for the civil service decided upon by the Ministry of Internal Affairs. Third, the State museums are dependent on the Agency for State Buildings (ASB), which belongs to the Ministry of Housing, for the maintenance and expansion of their buildings. And last but not least, the State museums had to respect the accounting rules concerning all expenses and income that are imposed by the Ministry of Finance. Thus an agency, like a State museum, must report to several principals at the same time, each of whom decide on several of its inputs and outputs separately. The current implicit contract between `the government' and their producing agencies is a governance structure not much different from an agency at the Ministry itself. In contrast, the new regulatory system The Case of Regulatory Reform of Dutch State Museums

13

specifies much more clearly its explicit contract expectations and methods of monitoring of outputs and abolishes the involvement of several other Ministries. Accordingly, the museum directorate gains much flexibility in setting personnel management policies — terms of employment, recruitment, and training facilities. This is further enhanced by budgetary reforms, according to which museums no longer have to follow the rigid overall rules for public expenditures. It becomes possible to shift money between budgetary parts and across years (minimizing the urge of expenditure in the last month of the budgetary year). This is in fact what happens with the museums already experimenting with the new rules.6 After a slow start, they are beginning to deviate from the normative budget. For instance, museums now spend more on personnel and less on materials then the indicative Pbudget of the government. The nine museums made net reservations in both 1990 and 1991 of 6% of their total public receipts (APSM, 1992, pp.13-15).7 Now there is even some discussion to allow museums to build up equity capital and to attract loan capital. Moreover, museums have a clear incentive to generate private income — be it admission fees, auxiliary income, sponsors or other sources of private support — because the Treasury is no longer part of the contract between government and museum. As long as the museum is able to meet the conditions for public support, it has many degrees of freedom left for expansion. For instance, if museums think there is a clear pay-off in raising the price of admission or organizing special exhibitions, no one can prevent them from doing so. Thus it becomes attractive for the directorate to use cost-benefit and risk analysis with respect to outputs and activities in order to optimize its results. Nonetheless, managerial freedom remains restricted in a fundamental sense: the museum directorate is not allowed to decide on its own about the collection and the building. Because both the building and the collection remain publicly owned, managerial freedom remains low (Grampp, 1994) and negotiations with the principal for its most important investment decisions carries on. This might remain a serious drawback as long as the government owns the assets of the museums, and museums only hold those assets in trust. It cannot be expected that an owner

6

As originally envisaged, after an experimental period with nine museums using covenants with specified products and prices), six of them were to have been `privatized' at 1-1-1993. In turn on 11-1994 the other three and another group of three were to have been privatized, with the eight remaining museums to follow on 1-1-1995 (APSM, 1992). However, the first six are still waiting for definite arrangements.

7

Unfortunately, the information contained in the APSM (1992) study is ambiguous, because it is not clear which 'budgets' they present exactly. Moreover, it is impossible to find out what shifts in the balance sheets of the nine museums have occurred. Therefore the observer does not know whether the deviation of budget for personnel is paid for by the M1-budget (except acquisitions) or by other sources of income. Moreover, the so-called `realized products' contain even inputs like `investment.'

14

Incentives and Efficiency in the Museum Market

allows someone else to play with his assets, which leads to a considerable amount of interference by the principal with the agents. Incentive Regulation Although the agent is thus provided with much more managerial freedom, this freedom also enhances the scope for discretionary behavior. As noted before, the principal used to have these rather strict rules in order to have ex ante control over spending of public funding. If agents are self-interested, less ex ante control might easily end in museums investing in their own priorities without much care for the principal's goals. The museum might, for instance, organize more exhibitions instead of looking after preservation. In order to align the goals between principal and agent, the Dutch government, well aware of this problem, has introduced incentive regulation using output management. Unfortunately, limited information on outputs, prices and exogenous factors in the case of the museum market lead to specific situations that make it difficult to provide the adequate incentive scheme and to prevent efficiency losses for the principal. As noted before, knowledge about performance is not always publicly available, and, it might be added, perhaps such knowledge is not even privately available if such outputs are less tangible or jointly produced. In such cases the principal suffers from problems of cost observation and thus performance measurement. Cost calculations are difficult in the museum market because production processes tend to be unit production and comparisons between museums are almost impossible because of their uniqueness. For instance, the organization of specific exhibitions or the restoration of particular works of arts is quite different between museums. Restoring paintings is quite different from restoring old vases. Exhibitions differ because the material exhibited might be compiled of the museums own collection or that of someone else. Outputs of museums are often joint, such as the exhibition and sale of a particular work of art, or the acquirement of public funding and earned income. It is therefore difficult to allocate costs to particular outputs and many degrees of freedom prevail with respect to the accounting rules. In case where the principal finds it difficult to observe performance the agent might easily get a price higher than the efficient price for its services — implying an efficiency loss for the principal.8 Another important problem is the principal's knowledge of the risk-attitude of the museums. The rules of the game seem to be simple: if the museums are not able to produce their contracted outputs, they will not get their agreed-upon public budget (output-price).

8

The only, but difficult way out of this problem might be international comparison of museums in particular areas. Museums, however, are usually information monopolists in their specific area, implying that it always remains difficult to obtain reliable information on performance for a monitoring agency.

The Case of Regulatory Reform of Dutch State Museums

15

However, if the museums are risk-averse, the principal should be careful in allocating risk. Leaving a risk-averse agent with the risk of bad luck in production might cause an adverse result, because such an agent tries to pass on the costs to the principal. It seems reasonable to suppose museums who do not own their collection and buildings are risk-averse. Therefore the principal should allow museums to calculate an amount of extra risk-premium in their prices. The discussion about the authorized level of savings of public money over more than one budget year has to be seen in this light, although it is mainly the principal who is paying for this `capital.' As a consequence of limited information, the principal (understandably) resorts to a second-best strategy, using contracts with relatively detailed tasks in order to provide value for money. This enables the principal, together with the business plans and inspections, to keep track of the public money spent on the museums. The principal should have asked for those plans under the old regime also, but failed to do so (Algemene Rekenkamer, 1988). However, the principal does not seem to realize adequately that detailed output and activity regulation cannot fill the gap caused by the heterogenity and jointness of the D-outputs of the museums. As argued before, some loss will inevitably occur in such cases and if the principal is not aware of this problem it might overregulate. The current detailed contracts illustrate this hypothesis in practice. This is the central pitfall that lies in wait. On the one hand, savings on monitoring and bonding cost are possible due to deregulation: on the other hand, these costs increase if new regulation concerning the museum outputs comes into the game. Overregulation might undo the advantages of the reform. It seems that expectations with respect to the use of output management as incentive regulation have been too optimistic. This optimism might therefore explain why the principal fails to acknowledge that if outputs and prices are difficult to observe some loss is inevitable for the principal. Next we will touch upon problems such as credibility and commitment with respect to the use of public funding as the means of improving efficiency by the government. One remarkable difference with the previous system is that contracting between principal and agent now accounts for the objective function of the principal explicitly, because the principal has to implement its policies through the composition of these contracts. This situation differs from its predecessor because in the administrative system the principal assumes its policy aims were fulfilled by simply stating them publicly. Implementation took place in the negotiation process between museums and the principal mainly on the basis of financial input data as provided by the public administrative system. Moreover, the stated objective was mainly directed at the Coutputs and not the D-outputs, because museums are usually thought to contribute to the dissemination of culture. As noted before, the principal simply used the fees as an instrument to

16

Incentives and Efficiency in the Museum Market

attain this goal — whether or not fee policy this contributes to the stated policy goal.9 The current system, however, has the advantage that the principal has to define its direct outputs much more explicitly in order to attain its policy goals. Thus the principal becomes more aware of the value he gets for its money spent on museum services.10 On the other hand, some pitfalls remain and might even be expected to pose serious constraints on a possible improvement in the performance of the museums. The operation of a museum is still a risky affair. Expansion is possible, but contraction, in the case of failure, is also. Financial setbacks are no longer absorbed by the government, although the government remains a lender of last resort (it is their assets the museums take care of). Moreover, the museums might lose their public support if they do not realize the contractual outputs. But, how reliable is the principal? What happens with the public budget if innovative management results in higher performance or if other circumstances like a budgetary crisis of the central government happens? Undoubtedly the museum directorates stick to their belief that good results lead to budget contraction. Although the principal promised that this would not happen during the regulatory reform and, in fact, the State museums did not face a smaller budget over the last five years (the real budget in 1992 equals that of 1987), the budget continues to be subject to the overall budgetary process within the government as a whole. This raises doubts within museums with respect to potential claims in the future and causes intertemporal inefficiencies because of credibility problems over time, as was illustrated by the recent debate over the new budget. Here the mere (but wrong) announcement of a potential cut in the public budget for the State museums immediately induced the museums to respond by threatening not to comply with the agreed-upon contact. Whether museums will expand their business into new markets does depend on the risk-attitude of the museum directorate with respect to both its public and private funds. Obviously, as long as the museums remain publicly owned and the public budget might be unstable a disincentive on innovative management continues to exist. 5.

CONCLUSION

The regulatory reform in the Netherlands with respect to the State museums illustrates the difficulties of output management if outputs and costs are difficult to observe. Nevertheless,

9

As table 2 shows, the number of visits to Dutch museums grew considerably in the 1980s. Surprisingly, though, the real costs of a museum visit have doubled in the same period. Whether or not this has improved the dissemination of culture has not been studied until now. In fact, it often turns out that such a policy implies a subsidy to the high income strata, because the visitors, in general, do not belong to the lower income deciles (Goudriaan et.al., 1984).

10

The stated preference of the principal with respect to collection management is optimization of collections within and between museums. As noted before, maintainance of the existing stock of works of art is now thought more important than presentation.

The Case of Regulatory Reform of Dutch State Museums

17

the transition of implicit contracting using input regulation to explicit contracting using outputs and activities seems to be an improvement. Nevertheless deregulation seems to succeed: the amount of detailed regulation will be lower and the agents have more managerial freedom than before. Moreover, the available information for evaluation has also improved considerably and the principal now makes its objectives much more explicit and translates them even into contracts. Nevertheless, several drawbacks might be expected from difficulties concerning the use of output management related to museums services. Expectations of government officials tend to be too optimistic with respect to the possibilities of governing agents whose outputs are difficult to measure. This might easily backfire if the most important lesson of principalagent theory is not recognized — that first best incentive schemes are impossible if outputs are difficult to observe and agents are risk-averse.

18

Incentives and Efficiency in the Museum Market

REFERENCES Agency for Privatization of State Museums (APSM) (1991), Annual Report 1990, Ministry for Welfare, Health and Culture (WHC), The Hague. Agency for Privatization of State Museums (APSM) (1992), Annual Report 1991, Ministry for Welfare, Health and Culture (WHC), The Hague.

Algemene Rekenkamer (1988), Rapport Rijksmusea, Tweede Kamer 1987-1988, 20.697, nrs.1-2, The Hague. Bradford, D.F., R.A. Malt and W.E. Oates (1969), `The Rising Cost of Local Public Services: Some Evidence and Reflections', National Tax Journal, 22, 2, 185-196. Central Bureau for Statistics (CBS) (1991a), Sociaal-Culturele Berichten, 91/2. Central Bureau for Statistics (CBS) (1991b), Sociaal-Culturele Berichten, 91/17. Dutch National Budget (1993), Tweede Kamer 1992-1993, 22.800, chapter XVI, nr.2, The Hague. Feldstein, M. (ed.) (1991), The Economics of Art Museums, Chicago (The University of Chicago Press). Frey, B.S. and W.W. Pommerehne (1989), Muses and Markets, Oxford (Basil Blackwell). Goudriaan, R. and G.J. van 't Eind (1985), `To Fee or not to Fee: some Effects of Introducing Admission Fees in Four Museums in Rotterdam', in V.L. Owen and W.S. Hendon (eds.), Managerial Economics for the Arts, Ohio (Akron). Grampp, W.D. (1994), Cultural Heritage Policy: its Purposes and Effects, mimeo. Harris, M. and A. Raviv (1978), "Some Results on Incentive Contracts with Applications to Education and Employment, Health Insurance and Law Enforcement", American Economic Review, 68, 1, 20-30.

19

Jensen, M.C. and W.H. Meckling (1976), "Theory of the Firm: Managerial Behaviour, Agency Costs and Ownership Structure", Journal of Financial Economics, 3, 305-360. Laffont, J.J. and J. Tirole (1986), "Using Cost observations to Regulate Firms", Journal of Political Economy, 94, 3, 614-641. Ministry for Welfare, Health and Culture (1989a), Zoeklicht op Zelfstandigheid (Searching for Independency), The Hague. Ministry for Welfare, Health and Culture (1989b), Profielen voor Rijksmusea en de Overige Rijksmuseale Instellingen (Profiles of State Museums), The Hague. Ministry of Welfare, Health and Culture (1989c), Projectplan 1990, The Hague. Ministry of Welfare, Health and Culture (1990), Externe verzelfstandiging belicht (OEVeR) (A Study of Independency), The Hague. Rosett, R.N. (1991), "Art Museums in the United States: A Financial Portrait", in: M. Feldstein (ed.), The Economics of Art Museums, pp.129-178. Sappington, D.E.M. (1991), "Incentives in Principal-Agent Relationships", Journal of Economic Perspectives, 5, 2, pp. 45-66. Throsby, D. (1994), "The Production and Consumption of the Arts: A View of Cultural Economics", Journal of Economic Literature, 32, 1, 1-29.

20

Research Memoranda Previously published: 9201 Bovenberg, A.L., Broer, D.P., Westerhout, E.W.M.T., Public pensions and declining fertility in a small open economy: An intertemporal equilibrium approach 9300 Sinderen, J. van, Over pre-economen, beleidseconomen en wetenschappers 9301 Knoester, A., Sinderen, J. van, Taxation and the Abuse of Environmental Policies: A supply-side view 9302 Bovenberg, A.L., Mooij, R.A. de, Do environmental taxes yield a double dividend? 9303 Broer, D.P., Westerhout, E.W.M.T., Bovenberg, A.L., The Tax Treatment of Pension Savings in a Small Open Economy 9304 Bovenberg, A.L., Mooij, R.A. de, Environmental policy in a small open economy with distortionary labor taxes: A general equilibrium analysis 9305 Smith, S., Vollebergh, H.R.J., The European Carbon Excise Proposal: A `Green' Tax Takes Shape 9306 Heijdra, B.J., Broer, D.P., Fiscal and Monetary Policy in a Dynamic Model of Imperfect Competition 9307 The Use of Economic Sanctions in Trade and Environmental Policy 9308 Burbidge, J.B., Scarth, W.M., Eliminating Interest Taxation and Tariffs: The Underpinnings for Recent Canadian Policy 9401 Verkiezingen 1994: Politiek en Economie, Economendebat 1994 9402 Cnossen, S., Administrative and Compliance Costs of the Vat: A Review of the Evidence 9403 Lindbeck, A., Uncertainty under the Welfare State: Policy-induced Risk 9404 Toen-Gout, M.W., Jongeling, M.M., Investments in infrastructure and economic growth 9405 Bovenberg, A.L., Mooij, R.A. de, Environmental Tax-Reform and Endogenous Growth 9406 Koutstaal, P.R., Vollebergh, H.R.J., Vries, J. de, Hybrid Carbon Incentive Mechanisms for the European Community

Papers and Proceedings Previously published: The Use of Economic Sanctions in Trade and Environmental Policy (zie Research Memorandum 9307) 9301 Beleidsperspectieven voor de middellange termijn, Economendebat 1993 Verkiezingen 1994: Politiek en Economie, Economendebat 1994 (zie Research Memorandum 9401) 9401 The Economics of Pensions: the case of the Netherlands 9402 Deregulering in Nederland: Economische Uitdagingen voor de Nieuwe Regering 9403 Miljoenennota 1995 9404 Het Paarse Regeerakkoord: Financiële, economische en sociale aspecten 9405 Quantitative Economics for Environmental Policy

Reprints Previously published: 9201 Cnossen, S., Vollebergh, H.R.J., Toward a global excise on carbon 9202 Bovenberg, A.L., Cnossen, S., Het Rapport-Ruding Bovenberg, A.L., Milieubeleid en budgettaire normen 9203 Kremers, J.J.M., Groeidynamiek in Duitsland 9204 Bovenberg, A.L., Fiscus en vergrijzing: Naar een ander fiscaal systeem 9205 Bovenberg, A.L., Petersen, C., Public debt and pension policy 9206 Vollebergh, H.R.J., Energieheffing: Melkkoe of krachtvoer? Vollebergh, H.R.J., Gezocht: Energieke energieheffingen 9207 Vollebergh, H.R.J., Vervuilen schaadt, betalen baat Vollebergh, H.R.J., Het beginsel `De vervuiler betaalt' en publieke heffingen 9208 Vollebergh, H.R.J., Vries, J. de, Financiële gevolgen van een regulerend milieuheffingenpakket 9301 Van Sinderen, J., Tien jaar aanbodbeleid 9302 Vollebergh, H.R.J., Fiscale milieu-instrumenten en bedrijfsmatige vergroening 9303 Goudriaan, R., Groot, H. de, State Regulation and University Behavior 9304 Heijdra, B.J., Yang, X., Monopolistic Competition and Optimum Product Diversity: Comment & Imperfect Competition and Product Differentiation 9305 Vollebergh, H.R.J., Milieubeheer 9306 Cnossen, S., What Kind of Corporation Tax? & Issues in Adopting and Designing a Value-Added Tax 9401 Knoester, A., The Inverted Haavelmo Effect and the Effects of Fiscal Policy in the United States, the United Kingdom, Germany and the Netherlands 9402 Kremers, J.J.M., Pauwels, M., Conjuncturele spanning en structurele convergentie in de Europese economie 9403 Sinderen, J. van, Taxation and economic growth 9404 Knoester, A., Kolodziejak, A., Effects of Taxation in Economic Models: A Survey 9405 Knoester, A., Mak A., Real Interest Rates in Eight OECD Countries 9406 Dalen, H.P. van, Praag, B.M.S. van, Public pension, market power, and intergenerational confidence 9407 Bovenberg, A.L., Mooij, R.A. de, Environmental Levies and Distortionary Taxation 9501 Kremers, J.J.M., Lane, T.D., Economic and Monetary Integration and the Aggregate Demand for Money in the EMS