Massimo Filippini e Paola Prioni
The influence of Ownership on the cost
of bus service provision in Switzerland.
An empirical illustration
Quaderno N. 01-01
Decanato della Facoltà di Scienze economiche
Via G. Buffi, 13 CH-6904 Lugano. Tel. 091 912 46 09 / 08. Fax 091 912 46 29
2
THE INFLUENCE OF OWNERSHIP ON THE COST OF BUS SERVICE
PROVISION IN SWITZERLAND–AN EMPIRICAL ILLUSTRATION-
by
Massimo Filippini and Paola Prioni
Department of Economics, University of Lugano
and
Switzerland Federal Institute of Technology Zürich
ABSTRACT
In deregulated transport markets, a firm’s ownership status and management system represent
an important issue. Economic theory suggests that productivity and performance are higher in
the private than in the public sector. In Switzerland, providers of bus transportation are
traditionally corporations, though a large part of their equity shares are still held by the public
sector (federal government, cantons, municipalities). In this paper we examine the potential
impact of ownership on the cost of bus service provision for a sample of private, public and
mixed bus companies in Switzerland. We have considered the estimation of a translog cost
model for 34 bus transit companies observed over 5 years (1991-1995). The results only
partially confirm that if the private sector holds shares in the company’s capital, efficiency is
enhanced. In addition, measures of economies of scale and density are derived and discussed
within the actual public transport policy.
1. Introduction
In recent decades the public transport policy of most industrialized countries has
been characterized by deregulation and/or privatization policies
1
.
Following the changes in the transport policy of the EC and the increasing subsidies
for operations and infrastructures, the Swiss public transport policy has experienced several
1
See Berechman (1993) for an exhaustive overview on the deregulation policies pursued in different
industrialized countries.
3
unavoidable regulatory reforms. These reforms were also induced by the findings of studies
by Filippini et al. (1992) and Filippini, Prioni (1994), which empirically demonstrated the
relation between the poor performance of Swiss bus operators in the pre-reform years to the
subsidies, public obligations and the compensatory payments for the approximated tariff
2
.
Since 1996 the Swiss
3
government has been introducing important reforms in the
regulation of the regional public transport system in order to encourage competition for the
market
4
. The new regional transport policy recognizes the earlier regulatory failures. With the
implementation of a bidding system for regional transport services and the removal of the
compensations for revenue reductions and approximated tariffs, the new act introduced
greater market incentives for competitive behavior and efficient production.
Although the studies by Filippini et al. (1992) and Filippini, Prioni (1994) provided
evidence on the economic consequences of regulation, due to lack of information no attention
was given to the effect of ownership type on performance. In Switzerland, private, public and
mixed firms characterize the regional public transit industry.
5
The main function of these
companies is to link mainly rural areas to an urban transport network or to the intercity
railway line. The majority of these companies are stock companies created by private
investors or by municipalities.
This mixed economy within the regional public transport market raises the
interesting issue of the effects on costs of the different ownership structures and institutional
organization form. Hence, the following question becomes central for policy makers: are
private bus companies more efficient than public bus companies?
In a changing competitive environment such questions become increasingly relevant,
first, for transport operators in terms of an improved knowledge about efficient organizational
forms, and second, for transport regulators as input into establishing new competitive rules.
The paper is organized as follows. Section 2 reviews the theoretical background on
the effects of ownership on efficiency. Section 3 discusses the cost model; section 4 presents
the data, while section 5 illustrates the empirical results. Section 6 concludes the paper.
2
See Filippini et al (1992) for a comprehensive description of the regulatory setting and subsidization system in
the regional public transport system in Switzerland.
3
Switzerland is a confederation composed of 26 cantons and approximately 3000 municipalities. This federal
state is characterized by a high degree of decentralization in the provision of public services. For instance, each
canton in cooperation with the municipalities has a high degree of autonomy in the organization, planning and
regulation of the local public transport.
4
See “revised Railway Act (revidierte Eisenbahngesetz), 1996.
4
2. The influence of ownership on the cost of bus service provision
In comparing the relative efficiency between public ownership and private
ownership, property rights theory has been often employed. Following Alchian (1965) and
Demsetz (1967), property rights theory analyses how the different incentives created by the
system of property rights affect organizational behavior and company performance. A right of
ownership comprises several rights: the rights to define the goals of a firm, to appoint the
managers, to influence the conduct of a firm, to keep the extra profits and to sell the property
rights. In a private firm, managers will make the day-to-day decisions for the owners. In
reality, these agents probably acquire some independence and have the opportunity to act as
imperfect agents. In this case we have a typical agency problem, where the principal cannot
perfectly monitor the actions of the agent. Thus, the agent could have incentives to deviate
from the pure profit-maximizing goal and try to attain some non-pecuniary benefits such as
pleasant office space and extra staff.
In the case of a public firm, the taxpayers are the owners of the firm and control the
way the firm conducts its activities through their representatives in the parliament and in the
government. These representatives are charged with managing the company in such a way as
to maximize social welfare. Public representatives in turn delegate control to a commission,
which oversees the management of the company. In comparison to private firms, public firms
experience a radical attenuation of property rights resulting in a more significant reduction of
incentives for the management to minimize costs and to follow the owner’s interests.
Moreover, the principal-agent problems within state-owned companies are more complex than
those in the private sector. In the case of a private-sector company, the management itself
answers only to the shareholders, and the employment relationship involves the management
and the employees. Within the public sector, the chain of command from the electorate to the
management go through the parliament, the government and the commission. For instance,
Niskanen’s model of bureaucrats’ behavior assumes that bureaucrats attempt to maximize
their own utility rather than that of the electorate they serve.
6
In a natural monopoly, however, the regulation of a private-sector firm can be more
difficult than that of a public-sector firm, because of more asymmetric information between
5
“Private firms” refers to those firms whose shares are held by the private sector; “public firms” refers to those
firms whose shares are held by public bodies or governments, for example municipalities; and “mixed firms”
refers to those firms whose shares are held by both sectors.)
6
See Niskanen (1975).
5
the private firm and the regulator than between the public firm and the regulator. Therefore, in
a situation of natural monopoly we can identify a trade-off between two sources of
inefficiency in private and public companies: inefficiency from regulation of private firms and
inefficiency from the attenuation of property rights in public firms.
A number of recent studies have compared the cost efficiency of public and private
bus transit companies. The empirical results do not provide conclusive evidence regarding the
efficient form of ownership.
7
In this paper we test the effect of ownership on the production cost using a sample of
Swiss bus companies. The test is performed by introducing in a cost model specification a
dummy variable to differentiate the two forms of ownership: private and public.
8
3. Specification of the Total Cost Function
Cost functions in the bus industry are well documented in empirical research (see
Berechman, 1993 for a good overview). Cost specifications traditionally assume operating
cost as a function of output and input prices (capital, labor and energy price). However,
several authors go beyond this specification and identify other exogenous variables, which
can further explain cost differences among the observations. For example, several studies
recognize output heterogeneity by adding to the cost function specification a series of output
and/or network characteristics such as the length of the network, the number of stops and the
frequency
9
. Another group of factors likely to influence operating cost includes regulatory
features, ownership form and environmental characteristics. Obviously such variables will be
considered in the cost function specification provided that they respond to the research
interest, that they are measurable and finally that a certain variability among the observations
exists.
It is worth noting that, although such variables are theoretically justified, their
inclusion in the cost function specification notably increases the number of explanatory
variables, which in turn reduces the degrees of freedom, which requires more data for the
econometric estimation.
7
See for an overview of 13 early studies Perry et al. (1986) and Bhattacharyya et al. (1995) and Cowie, J.,
Asenova, D. (1999) for more recently published work.
8
The sample of companies used in this study covers the period 1991-1995. In this period the bus companies
operated as franchised local monopolies, assigned without a competitive tendering process.
9
See for instance Spady and Friedlaender (1978) and De Borger (1991).
