XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov.

2014

Contemporary public enterprises in global perspective: theory and evidence

Massimo Florio
1. Introduction

After three decades of privatizations, public enterprises are still a feature of contemporary economy. The
environment in which they operate is, however, different from the one prevailing in most of the XX
century and a new inquiry is needed. According to the Economist (2014b):

«In China companies in which the state is a majority shareholder account for 60% of stockmarket
capitalisation. In Russia and Brasil companies in which the state has either a majority or a significant
minority stake account for 30-40% of capitalisation. Even in such bastions of economic orthodoxy as
Sweden and the Netherlands state-owned enterprises (SOEs) account for 5% of market capitalisation.»

This paper offers a discussion of positive theories of the public enterprise (PE) in the global perspective,
i.e. without reference to specific country or region, and presents some fresh evidence potentially
supporting or rejecting alternative theories. It also concludes with some policy implications.

By PE in this paper we understand organizations ultimately owned by governments, providing economic

goods, with a budgetary autonomy, some managerial discretion, and for which privatization in principle
is possible. We also claim that a ‘true’ PE is an organization for which some public mission can be
identified, at least in principle, see CIRIEC (2012). As we shall see, aproper discussion of the public
mission is an essential aspect of the analysis, as suggested by Del Bo and Florio (2013).

This definition is rooted in history (Millward 2011), but it alsohas an analytical value, as this paper will
show. While in some of the previous literature the label State Owned Enterprise (SOE) was common to
identify our topic, I prefer here the more general term public enterprise (as for the Journal of Economic
Literature classification code L32). In many countries the state is articulated in different layers of
government, and the variety of forms of control is such that ‘state ownership’ does not capture, e.g. mixed
ownership of listed multinationals (Alonso et al.2013) or, at the other endof the scale, consortia of small
municipal utilities (Warner 2011). Thus, while in other contexts, see e.g. Kowalskiet al. (2013) a SOE is
identified by a share in equity of more than 50% owned by a government entity, a PE can be ultimately
owned, i.e. de facto controlled by governments by different ownership arrangements. In some of the
evidence reported below we shall focus on the top-shareholder, whatever his share in the equity. While
the perspective adopted in this paper may seem excessively relativistic, there areadvantages in adopting
an approach where ownership is treated as a control mechanism that can take different, flexible, forms.
For empirical analysis one has to use sources of data that use conventional definition of ownership, and
all of them have their limitations, but in an analytical perspective it seems better to refer to PE in a more
general way, as economic organization for which some key powers are retained directly or indirectly by
governments, at national or local level, and in some case even at intergovernmental level.

When the perspective of ownership as a control mechanism is adopted, even a minority share in equity
can guarantee keydecisions, such as the appointment of top management. Correlated to the power to
appoint managers by governments there is the power to provide instructions to them. Again, this does
not need to be a formal mechanism, as it can be based on informal procedures, including a normally
passive or tacit attitude by government departments or other public sector bodies, then becoming
occasionally a proactive stance.

1
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

In this context, budgetary autonomy and managerial discretion simply mean that the PE is managed
outside government, i.e. is not formally a part of the public administration, it works with own capital,
and has certain reporting obligations on its accounts (which often one would see better developed, see
Greiling, Grüb and Huber 2014). The focus of this paper is not, however, on the legal mechanisms that
concretely ensure budgetary autonomy and managerial discretion to PE, or on internal external
governance arrangements, as these are very different across the world, but on the broad economic and
policy issues surrounding these specific features.

Morevoer, the definition of PE adopted in this paper recurs as a test to the possibility of privatization to
understand whether what we observe is actually an enterprise or not. This is not very restrictive. While
it is evident that organizations providing manufactured goods, mining, electricity, telecommunications,
transport, water, housing, banking, etc. can be and have been extensively privatized in the last decades
in many countries, in principle many other activities can be organized by the private sector, including the
provision of health, education, prisons, and other ‘law and order’ related services. In the UK some local
police serviceshave been recently franchised to contractors, and the same has happened to supporting
services for the defense in the US. Thus, when a government-owned hospital has budgetary autonomy
and managerial discretion, in principle it can be privatized or go bankrupt, hence it is a PE (see Lethbridge
2014). This would not be the case, however, if the hospital is still part of the the Department of Health,
directly linked to the hyerarchy and procedures of the public administration.

It seems essential, then, that in order to define what a PE enterprise is in the real world, we do not stick
to a static definition based on some sectors, or technologies, or types of service. There are mobile
boundaries between governement itself and its controlled enterprises, and we shall not deal in this paper
with the specific advantages or disadvantages of creating the PE. In other words, we simply observe the
fact that a certain class of activities, in certain countries and sectors, have been delegated by government
to ‘agents’ with certain features, the PE, and then we explore these features at an abstact level, in a
theoretical perspective, and we show some evidence supporting the theoretical analysis.

As we shall see, also the boundaries of what is a public mission cannot be considered in a static way.
Millward (2011) has argued that strategic and even specifically national defense arguments can explain
PE in several Western economies, while social cohesion arguments were more important in other
contexts and times. This leads to our main research question:

Why do governments still own enterprises, and in certain cases even create new ones?

We will see that the dominant answer to this question by many economists is that governments keep the
control of PE mainly for the self-interest of politicians, or at least for some inertia. I will label this
explanation the ‘public choice’ perspective aboutthe positive theory of PE. However, a different
explanation is possible: PE exist because thay play a public mission role that is not performed by market
mechanism. I will label this the ‘public mission’ perspective. The two explanations are not mutually
exclusive, but my tenet is that the importance of the public choice interpretation in the real world has
been exaggerated, while the public mission one has been unduly neglected.

The paper is structured as follows: After this Introduction, section 2 presents the two main alternative
theories of PE and evaluates the arguments for and against each of them, section 3 discusses some
evidence, and section 4 concludes with a brief discussion of the policy implications.

2
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

2. Positive theories of public enterprise

The discussion will deliberately set aside many issues frequently encountered in regulatory economics
and in the theory of the firm, in order to exclusively focus on somegeneric ingredients that can be helpful
in trying to answer our research question about the survival of PE in contemporary economy. A more
detailed discussion is given in Florio (2004, 2013).

2.1 The public choice perspective

A sketch of the public choice perspective 1 applied to the analysis of PE is the following one. A
government is assumed to be controlled by self-interested politicians, who have a private agenda. In other
words, politicians are individuals having a standard direct utility function, where the arguments are
consumption goods, or equivalently an indirect utility function where the arguments are income and
prices of goods.

In an intertemporal setting and when politicians need to be elected to get or to preserve office, they also
want to maximize the probability of re-election. Even if the country is not a democratic one, politicians
need to achieve a certain critical mass of consensus in order to achieve the desired amount of power and
to preserve it against social upheaval or even a revolution.

This intertemporal setting, that includes the objective of preserving power, leads to a modification of the
standard utility function to allow for some interest of the politicians to buy popular support through their
political agenda. Here the objective function of the politician becomes a combination, linear in its
simplest form, of self-interest and social welfare. In turn, social welfare is described as the sum of
consumer and producer surplus of the median voter, or of a representative voter.

While a wide literature, theoretical and empirical, has utilized this framework for discussing the size of
government, i.e. of public expenditure and taxation (see Mueller, 2003 for the general public choice
perspective), the same framework may be used to explain the rationale of PE. In this literature they are
seen as another way by the politicians either to extract rents for them or to achieve political support by
voters.

We discuss in turn the two aspects. Rent extraction by the PE to the advantage of the politicians would
see these organizations as simply another form of bureaucracy, where the agents, i.e. the managers, are
instructed by their political principals to exploit the consumer, possibly through monopoly power, and
transfer the rentsto them.

As this would be illegal under most constitutional systems (because the personal assets and income of
the politicians are formally separated by those of government) rent extraction would happen in other
ways, not necessarily illegal from a formal perspective, but de facto amounting to a variety of forms of
corruption. For example, a politician may appoint as manager somebody who is instructed to manipulate
the procurement process of the PE in such a way as to favour firms related or even directly owned by the
politician. The manipulation can take the form of artificially increasing the cost of the PE, creating extra-
profits for private suppliers, possibly then transferred abroad in banking accounts or real estates or other
assets owned by the politicians. Other forms of manipulation may include the appointment of relatives
of the politicians and thenpay them overcompensation for their services, if any, and this also to the
advantage of the politician or their household or extended family.

1
We use the label ‘public choice’ in a loose way, as in fact there are different strands in the privatisation literature, see Florio
(2004).
3
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Manipulation can also happen on the sales side, for example in franchising by the PE generous contracts
to retail supply, exporting agents, or other commercial activities to relatives or agents of the poltical
principal. All these practices amount to a certain extent to a distortion of the economic activity of the PE
and to hard or soft forms of corruption as a mechanism to extract rents.

The second mechanism, under the public choice interpretation of the PE, can be seen as distributing rents
to potential supporters of the politicians. This can be achieved basically under the following headings:
Excess employment and wages unrelated to labour productivity; excessive prices paid to suppliers in
order to enlarge political support; excessive investment for the same reasons; too low prices for some
users of the service or consumers of goods provided by the PE, or excessive service coverage in terms of
effective demand; possibly even excessive quality of the service to please some users; coverage of losses
by exploitation of some types of users through cross-subsidies and/or through excessive taxation of
certain types of tax-payers.

The consequence of the public choice perspective on PE is straigthforward. As they are just another form
of bureaucracy (Bureaucrats in Business, World Bank 1999, was the title of a well-know report by World
Bank economists), and create inefficiency because of the generation and distribution of rents to
politicians and their political supporters, privatization must always be a welfare enhancing policy. Faith
in the efficiency of market mechanisms and the assumption that politicians are more or less always self-
interested, would then be combined to achieve the dogmatic view that whenever and wherever possible
it is just beneficial to social welfare to sell PE.

Before entering the empirics of this perspective, it is worth assessing its value in analytical terms.I have
elsewhere (Florio, 2004) remarked a logical contradiction in the public choice theories of PE and of
privatization. If politicians are intrinsically self-interested, why should they be more interested in keeping
the PE than in selling them? Or in other words, what kind of genetic transformation happens to the
political élite of a country when instead of continuing the inefficient control of PE turns to privatizing
them? Should we conclude that the politicians in favour of PE are self-interested while those who
privatize are a completely different type of person, as they suddenly become altruistic, not interested in
their private income and in probability of re-election?

Good theory cannot be so ad hoc that it changes its assumptions according to circumstances. Either you
think that politicians are always self-interested, and then you have to stick to this assumption also when
you analyse privatisation, or you relax the assumption and then admit that to a certain extent politicians
may be altruistic in both circumstances.

When you are consistent in your theory, however, surprising results may appear. A good example is the
postumous book by the late Jean-JacquesLaffont (Regulation and Development, 2005). In that book by
one of the leading theorist of public economics, if a utility function of the politician is introduced
consistently ‘before’ and ‘after’ privatization, i.e. in cosidering the option of keeping or selling PE, it
turns out that privatization can happen for good but also for bad reasons.A corrupted government may
well be interested in privatization, an issue I have discussed elsewhere with the example of Russia (Florio
2002).

Let me briefly recall how a self-interested politician can manipulate privatization. First, the PE can be
underpriced and sold to relatives or cronies. The difference between the expected net present value of
future profits and the actual privatization proceeds for the government is a lump-sum rent, and this can
be either directly appropriated by the politician or distributed.

4
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Even the distribution in the form of cheap mass share-holding, the so-called popular capitalism, in this
case is nothing more than another way to try to gain political support by beneficiaries of the underpricing.
Second, managers appointed by politicians under government ownership can be confirmed in their
positions and instructed to continue to manipulate the company to extract rents for the politicians, perhaps
in an easier way because of the greater budgetary autonomy of the privatized firm. Thus many of the
mechanisms that were in place before privatization will be in place under private ownership, and crony
capitalism will substitute the PE.

The countervailing force to these mechanisms, in the transition from the PE to the privatized firms should
be the functioning of markets. A privatized PE mismanaged to favour politicians and to win political
support will be costly and if there is market entry, at a certain point, rivals will get the control of the
company or it will simply collapse. In a competitive markets the tolerance of mismanagement, i.e. of
deliberate inefficiency, to favour politicians, would be minimized, while it is maximized under PE.

This is however a second weak point in the theory of privatization in the public choice perspective, as it
assumes market competition as linked to privatization. In fact, there is no strict relation per se between
private ownership and competition, as there is no strict relation between public ownership and monopoly.
In most cases privatization does not lead to perfect competition as the prevailing market form. It usually
leads to oligopoly, and particularly to regulated oligopoly, because of the persistence of ingredients of
natural monopoly in the industries (such as fixed networks for energy and transport services, time slots
in airports, spectrum in mobile communication, capital rationing and other market failures in credit
markets, etc.). In other cases, it leads to franchised monopoly in the form of concessions. The theory of
regulated oligopoly based on private ownership is often strikingly unbalanced. It usually assumes that
firms are profit maximizers but the regulator is benevolent, even if there may be an issue of asymmetrical
information between the regulator and the regulated firms. Any inefficiency is then often just related to
this specific problem.

But the regulator is just an office of the public administration. Even if not usually civil servants, the
members of a regulatory body are appointed directly by governments or in some cases by parliaments,
i.e. they are appointed by the same politicians who would have appointed the managers of the PE. Thus,
why the self-interested politician who may have privatized a PE to extract rents should not at the same
time appoint regulators instructed to protect the rents of the privatized company that it is still under the
indirect control of the politician or his supporters? If the government is made by self-interested
politicians, who appoints the benevolent regulators?

Interestingly, all the mainly US literature about regulatory capture and rent-seeking by the regulated firm,
a literature that was motivated by a context where it is obviously a very profitable investment to lobby
(if not to buy) regulators and legislators to protect certain vested interests, has not been fully transplanted
in the discussion of the regulation of privatized firms in Europe.

Until now I am not making any claim about the empirical evidence supporting or rejecting the public
choice view of PE, of privatization and of regulation. I am just showing the assumptions on which it
often rests, and what I consider two logical contradictions emerging from an examinantion of the
literature. To sum up: these contradictions are (a) the unexplained reversal of the assumption about the
self-interest of politicians when they nationalize or keep PE versus when they privatize, particularly as
far as the control of firms and regulators is concerned; (b) the unjustified assumption that privatization
leads to competition, while PE is associated to public monopoly, with the supposed inherent efficiency
of the former and inefficiency of the latter.

5
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

If you stick to a consistent public choice perspective, you would conclude that either PE or privatization
are policy alternatives that the politicians will pick-up according to their self-interested perspective. Thus
in certain countries, industries, and time the incumbent politician will nationalize private firms, or will
support the PE, and in other circumstances he will shift to privatization, but in both cases nothing can be
said in general about the social welfare effects of the two alternatives, as in both cases politicians want
to maximise their utility, and not social welfare. It will just be an empirical matter to see whether here or
there a policy option or the opposite one was more or less beneficial to a country.A consistent public
choice perspective, in contrast to a biased privatization retoric, should be mute about a generic prediction
that privatization is more efficient than PE. It is hence surprising to see how the supporters of public
choice have been so often in favour of privatization. It is perhaps less surprising to understand the
positions of those economists who used the liberalization argument, and occasionally the close
relationship between privatization and liberalization, but this is clearly a different path of analysis, one
that will be further discussed in the next section.

2.2 The public mission perspective

If we assume that politicians are mainly interested in social welfare, the perspective on PE (and on
privatization) would radically change.

It is important here to introduce a dose of realism in the discussion. A recurrent expression in public
economics literature to describe a government aiming at maximizing the social welfare is the slightly or
unvoluntary derisory term of a “benevolent” government. Obviosly governments, parliaments, regulatory
bodies, management boards, and any collective entity areultimately made up of human beings. Any
human being empirically observed is a mixture of motivations, beliefs, preferences, cultural attitudes.
Even the toughest version of the public choice perspective would concede that finding somebody who is
intrinsically self-interested in the very narrow meaning of wanting to maximizing his utility by the
optimal consumption of goods is - to say the least- absurd. A greedy dictator may include in his self-
interest a certain belief that he is the father of his people, or at least of a part of it (the white men, or the
Arians perhaps), and conceive himself as altruistic with just some occasional interest for luxury goods.
But economics is not about the psychology of individuals. It is about taking assumptions that are
appropriate in modeling the behavior of agents, in abstract, or in a statistical sense, and then testing if a
specific setting is confirmed by data.

Thus, the opposite extreme of the public choice perspective on a continuum of assumptions is the
hypothesis that politicians do not care at all about their income and consumption of goods, while they
are interested in having the power to enforce their vision of what the social welfare should be, given
certain constraints (of feasibility, ofinformation, of political viability, etc.). In this perspective, PE exist
because politicians think that a certain delegated organization for the supply of economic goods is more
efficient than the ‘regular’ public adminstration in terms of a public mission to be achieved.

Before exploring the latter point, it seems worth to clarify a possible misunderstanding. The typical
treatment in regulatory economics of the social welfare function (SWF) as the objective of the
government (or the social planner more in general, including the regulator) is often at variance with the
theory of applied welfare economics, and with common sense. A typical model of a benevolent
government or regulator is such that the latter is an agent that wants to maximize a SWF given by the
sum of consumer surplus of a representative user of the services and the producer surplus. Occasionally
in such a SWF a shadow price of public funds is included if, for example, a transfer is needed from the
government to the PE (or the regulated firm with e.g. universal service obligation) and as usual lump-
sum, i.e., non-distortionary taxes are not available to cover the funding gap.

6
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

There is no reason to make this very particular assumption on the SWF, and it is important to understand
this methodological point because it is the origin of much confusion about the welfare properties of PE.
A SWF that simply sums the welfare of consumers is of the Bentham type. Moreover,if there is just one
representative consumer in the economy, social welfare maximization amouts to optimal consumption
of such individual and nothing more. But if there are two consumers with different welfare initial levels,
you cannot actually escape the issue of including welfare weights in the SWF, that would reflect the
politicians preferences for equity. This leads to Bergson-Samuelson SWF, or to Rawlsian ones, or to
other more complex types (Florio 2014). When you introduce welfare weigths in social welfare
maximization, the views about what is efficient for the PE to do change accordingly.

But when you sum producer surplus and consumer surplus in a SWF you risk to introduce a misleading
notion. If the producer surplus (defined as the integral of the difference between marginal revenues and
marginal costs, i.e. profit) is the expression of monopoly power, this would be a social cost, not a social
benefit, see Laffont and Tirol 1993, for example who explains why leaving a profit to a firm by a regulator
is costly in terms of social welfare). As in fact privatization very often leads to regulated oligopoly or
franchised monopoly, and often with an incumbent in a dominant position in the long term, adopting a
SWF made up of the representative consumer surplus+producer surplus is theoretically highly
questionable, because prices (and quality) will not have the desired welfare properties. We do not need
to think in such restricted terms to what a SWF is in the perspective of a politician committed to optimize
on it.

More typically, the ingredients of a reasonable SWF, one that we can assume as heuristically valid for a
developed and democratic country, would include such objectives as a welfare weighted sum of
individual utility, i.e. an objective of efficiency but also an objective of equity, and in addition quite often
some Musgravian merit goods. Again these SWF are often described by the slightly derisory term of
government ‘paternalism’. Compulsory universal education and health coverage are in this category of
merit goods, hence virtually all governments in the world have a dose of paternalism in the SWF they
adopt. It can be formally shown (see Drèze and Stern 1987) that such macroeconomic objectives as the
minimization of inflation and unemployment, maximization of output and of equity can be easily
included in the standard macroeconomic setting of a Bergson-Samuelson SWF. Add some merit goods,
as the protection of environment for the future generations or as an ethical value, and some concerns for
national security, and most of the typical agenda of many governments is appropriately depicted. It is not
necessary to thing that politicians are saints. It is sufficient to assume that the political system is such
that whatever the psychology of the individual politicians, the society compels them, to a certain extent,
to adopt a social welfare function that includes a given set of objectives more or less shared by the
majority of the voters. This is still a positive theory of government, not a normative one about what is
desirable in the perspective of an indipendent welfare ecomomist.

Let us turn to the implications for the PE of this setting. Here the PEsare no more mainly a mechanism
to extract rents for the politicians and to maximize their probablity of re-election, but PEsare rather
instruments of the government to achieve its objectives, whatever they are.

This implies the following:

a) The government assessed that neither the regular public administration nor private firms are able
to efficiently reach some of the objectivesincluded in the SWF;
b) and that organization as the PE may have certain advantages relative to the above mentioned
organizations;

7
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

c) consequently, the managers of the PE are appointed and instructed to perform a public mission,
which is the concrete manifestation of the SWF objectives delegated to them by the government as part
of a ‘production plan’, i.e. plan to deliver goods not delivered by the markets;
d) thePE management is intrinsically exposed to the risk of being privatized or alternatively to be
fired or relocated in the regular public administration. The latter is in fact one of the important features
of PE management as opposed to top civil servants.

Given this setting, the public mission (which can include different and even contradictory objectives) can
be internalized by the PE managers in appropriate forms. There are basically two mechanisms, which are
symmetrical to similar problems in the public choice perspective, where the greedy politicians need to
instruct the PE managers (or the managers of privatized firms) to transfer them rents. First, managers
who share the same objectives of the government are selected. These types are known in a growing
literature as intrinsically motivated managers. Second, professional managers are incentivized by leaving
them some rents extracted from consumers and taxpayers, reaching an equilibrium between the social
cost and benefits or doing so. Grönblom and Willner (2014) discuss the two mechanisms, Polidori and
Teobaldelli (2013) review some of the literature.

Thus, in the best of possible worlds, PE exist because in a democratic society a majority of citizens selects
polticians who care for aSWF that reflects citizens’ preferences, which include goods not adequately
provided by private markets, and in turn these elected politicians select intrinsically motivated managers
of PE, thus circumventing the need to leave them rents arising from asymmetric information or other
inefficiencies.

It is easy to see some weak points in this representation. First, nowhere the democratic process is such
that politicians are simply the passive agents of citizens. There is a complex interplay between the two
sides of the democratic game, voters and politicians. It would be naïf not to see that preferences of citizens
may be influenced by deliberate actions of politicians, including capture of media to convey a message,
etc. Second, the accountability of politicians to citizens is less than perfect, the agenda of the government
is multidimensional and highly technical on many issues that only a tiny share of the population can
understand. Thus citizens preferences can be betrayed or misinterpreted at a later stage and the public
missions be watered down. Third, something may go wrong in the process of evaluating whether a PE is
better than alternative forms, for example because bad evaluators are selected. Fourth, intrinsically
motivated managers may be difficult to find, and the alternative mechanism or recruitment may hirein
the PE self-interested managers who earn excessive rents, manipulate information and public mission
and eventually can become so powerful to capture the politicians, so reversing the principal-agent
relationship. It is possible to build a wide historical or anecdotalnarrative of several of such occurrences,
without pretending that the politicians were self-interested in the first place.

Before concluding this section we need to elaborate about the theory of public monopoly versus market
opening. As mentioned in the previous section, conceptually there is no need to associate public
monopoly with PE, even if, historically, this association has been often empiricallyobservable. But here
we want to discuss a theoretical setting.

There are three market forms in which one can think that PEsare embedded. The received doctrine of PE
was reflected in textbooks such as Bös (1987), or Rees (1976), and concisely summarized again by
Laffont and Tirole (1993). Without going here in the details of that setting, which was often focusing
mostly on pricing rules of the PE, the standard assumption was legal monopoly, i.e. the prohibition of
entry.

8
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

However, there is very large theoretical literature on mixed oligopoly, where typically a PE competes
with a private rival. This literature is particularly relevant today. Mixed oligopoly is the normal feature
of many industries in many countries. This is not the place to review this literature (see De Frajaand
Delbono1990) but its main message is that it cannot be said in general that a mixed oligopoly, with a
welfare maximizing PE and a profit maximizing rival is inferior or superior in social welfare terms (albeit
restrictively defined in this literature) to public monopoly, private oligopoly, or even private monopoly
(in exceptional cases of very high inefficiency of the PE under public monopoly, see Ceriani and Florio
2011).

Eventually, and perhaps surprisingly for some, there is no reason to excludethat a competitive market
can be based on PE. As ‘competitive’ in economics simply means that firms take prices as parametric,
you may think that a large number of small PE supply goods at a given prices determined by a planner
or regulator or by a market arrangement. For example, hundred of municipally-owned PEs can generate
electricity and their supply cleared by demand in a regulated market. Moreover, in some countries there
are PEs or mostly PE-based oligopolies, with such firms competing among them in markets (or in quasi-
markets in the lexicon of the New Public Management).

Competition among PEs may be desirable or not, this is not our main concern here, but the fact that it is
possible to design realistic economic models where PEs are not operating under a regime of monopoly
is enough to break any dogmatic association of the notion that PEs are necessarily enemies of
competition.

A very loose argument often mentioned in some regulatory economics literature is that regulators tend
to be lenient with PEs because after all they are both public sector organization. One may make a number
of anedoctal counter-examples. Clearly a theory cannot be built on narratives, either about regulators
who are ruder with the PE because they are the incumbent and the regulator want them to allow entry of
private rivals; or the other way round.

2.3 Summing up
Neither the public choice nor the public mission perspective on PE lead to general predictions and a sure
answer to our research question. The public choice view as an argument for privatization was found to
be contradictory for both the shifting hypotheses about the nature of the politicians and the unwarranted
tenet that PEs are linked to public monopoly. The public mission case is a symmetric one that needs to
assume that politicians are the servants of the public interest, are able to evaluate the best organizational
form to pursue it, and to appoint the right managers for the PE.

Given the dominant policy paradigm of the last thirty years, which was largely supportive of
privatization, the public choice view has offered a sort of intellectual justification to the divestiture of
PE. For those who have doubts about the realism of this narrative, it may be interesting to consider that
an alternative theory can be built, a theory of PE,not necessarly more realistic than the dominant one, but
logically consistent and supportive of the PE as expression of a public mission of the governnent and
ultimately of the preferences of the citizens.

In any case PEsare compatible with mixed oligopoly and even with competition, and no claim can be
made in general about the attitudes of regulators towards PEs in such markets as compared to their
attitudes towards private rivals.

Finally, it is also rather obvious that a PE without a public mission is better intepreted as an accident, or
with the tools of the public choice perspective.
9
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Given the uncertainty surrounding the theory, we turn now to the empirical evidence.

3. Evidence

To answer our research question, a possible strategy would be to test the evidence supporting one of the
two alternative hypotheses. This implies the need to identify some features of real world PE more
correlated to one of the hypotheses but not the other one.

Only a detailed analysis country by country and within each country would be able to satisfactorily test
the hypotheses. This cannot be attempted here. A project launched by the CIRIEC International Scientific
Commission Public Services/Public Enterprises2 is attempting to do so, combining detailed case studies
at firm level and thematic analyses: see the contributions in recent special issues of the Annals of
Cooperative and Public Economics (2011), International Review of Applied Economics (2013), Journal
of Economic Policy Reform (2014). The Commission is also planning to collect a set of country studies
about the status and evolution of major public enterprises. Looking at the global South, it is also worth
mentioning the activities of the Municipal Services Project3 and of the Trasnational Institute4 and recent
books by Mc Donald (2014) and by Chavez and Torres (2014). It should also be added that the topic of
the continued relevance of public enterpises has been recently noticed by some international institutions
that have supported privatization in the last decades, particularly the OECD and the World Bank. In the
rest of this section we shall mention some of this evidence, and some additional research currently going
on at the University of Milan, at the DEMM, Department of Economics, Management and Quantitive
Methods5.

3.1 Looking at the major global players

According to Musacchio and Lazzarini (2014)6, there are three forms of contemporaty state capitalism:

«In the first form, which is particularly popular in China, the state submits an SOE to the governance
standards and investor scrutiny that come with a stockmarket listing while retaining the bulk of the shares.
In the second, which accounts for about half of SOEs, the state retains just enough influence, through its
minority stake, to swing some important decisions. In the third, the state seeks to invest in companies –
including ones not previously government-linked – through public development banks (of which there
are currently 286 in 117 countries), sovereign-wealth funds, pension funds and other vehicles.»

A possible research strategy is looking at the performance of the major global players: Is the performance
of the PE in this category worse than their private counterparts? And, second, are these PE concentrated
in countries where the quality of government is weak?

Under the public choice view, if PEsare fundamentally a mechanism to extract rents for the politicians
and their clients, you would expect that their financial performance is poor if compared with their rivals,
where the management is instructed by private owners to minimize costs. At the same time, you would
expect that in countries where corruption is minimized, there should be less PEs than elsewhere.

2
CIRIEC International Scientific Commission Public Services/Public Enterprises
(http://www.ciriec.ulg.ac.be/en/pages/3_2recherches.htm)
3
www.municipalservicesproject.org
4
www.tni.org
5
SUPER, http://www.publicenterprise.unimi.it/?page=super-what_is_super
6
Cited by the Economist (2014b).
10
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

It is important to stress that here we are not considering the public missions of the PE, which could to a
certain extent distract the management from cost minimization. In principle, you may find that the
financial performance of a PE is not as good as it could be because it is contributing to other government
objectives, for example service provision or employement in disadvantaged areas, sustaining research
and technological innovation where the private sector is not willing to risk capital for uncertain returns,
or security of supply in energy and raw materials. Thus, the profitability of a PE could in principle be
lower than that of its private rivals for good or bad reasons, i.e. because of the dominance of respectively
mechanisms related to the public mission or the public choice perspectives.

But if the financial performance is not different between the PE and the private firms, the rent-seeking
mechanisms in place should not be so strong. In other words the incentive for managers to minimize
costs are similar, and –while we could not directly conclude in favour of the public mission view, because
we have not an evidence directly related to the performance in terms of public missions (other than
providing revenue to the government), we must also think that the public choice mechanism is not so
strong as to disrupt the performance of the PE.

In a recent research sponsored by the OECD, Kowalski et al. (2013) study companies included in the
Forbes Global 20007 list, and use the Orbis database8 to collect finacial data of very large multinational
groups. The financial performance variables they consider include sales, profits, assets, and market value.
In orderto create a ranking of global firms, they attribute the state-owned enterprise (SOE) label to those
companies for which according to Orbis the ultimate ownership is more than 50,01% of the shares.

The Forbes Global 2000 list is in fact excluding some major PEs because of their legal status, for example
railways and post offices which have not been corporatized, and mainly focused on listed comapies. This
limitation of scope is acknowledged by Kowalski et al. (2013, fn 15) but in a sense it has the consequence
of comparing private and public firms which are more similar form an organizational perspective. A
second limitation of the scope of the classification is thata share of more of 50% is not needed to control
a firm. In fact, manyglobal PEs in the world are effectively controlled by governments with a stake of
20-25% (see the Economist (2014), p. 18), to be compared with much more dispersed equity owned by
other shareholders. For example in Italy the government has a firm control of the two core energy
companies, both listed and both operating internationally, ENI (mainly oil and gas) and ENEL (mainly
electricity), with a share of around 30% of the equity.

Despite this methodological limitations, the OECD paper, however, provides interesting evidence for our
research question as it covers firms whose combined market value would represent 81,9 % (in2010) of
the global capitalization of listed companies. These large groups in turn control around 330,000 domestic
or foreign subsidiaries.

Around ten per cent of the Forbes Global 2000 list (204 groups) are SOEs as defined by the authors.
Sectors where the share of SOE in the domestic economy is high (more than 10% of GDP) include coal
mining, land transport, transport via pipelines, oil extraction, electricity and gas, but also
telecommunications, financial intermediation, engineering, warehousing, manufacturing, air transport.
The ‘Forbes’ SOEs would represent between 11% and 16% of total sales, profits, assets, market value of
the Forbes Global 2000 aggregate, see Table 1. In this table we use global GDP as a normalization factor
(for consistency of sources, GDP data are again from Kowalski et al. 2013).

7
Available at http://www.forbes.com/global2000/list/
The top SOEs are shown in Annex 1.
8
http://www.bvdinfo.com/en-gb/products/company-information/international/orbis-(1)
11
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

In this way, looking at the first column of Table 1., the total sales of the sample is worth more than 51%
of the world GDP, its market value (as mentioned, these are listed companies) is 58% of GDP, and the
total assets combined, as recorded in the ORBIS database, are worth 2.18 times the GDP. The
corresponding figures are given for the SOE. For example, their total profits in 2010 were 0.5% of global
GDP. In the third column of Table 1 a simple ratio between the values for the SOE and the total sample
are given. As we can see, the share of SOE on the total Forbes 2000 list is 11.5% of sales, more than 13%
of both profits and market value, and more than 16% of of assets.

Table 1. Aggregate financial indicators, % of GDP (2010)

Forbes Global SOE SOE/Total
2000
Total sales 51.1 5.7 11.15
Total profits 3.7 0.5 13.51
Total assets 218.6 35.8 16.38
Total market value 58.4 7.8 13.36
Source: Florio (2014) based on Kowalski et al. (2013, tables 1 and 2)

Table 2. Selected financial ratios

Forbes Global SOE SOE/Total
2000
ROS (profits/sales) 7.24 8.77 1.21
ROA (profits/assets) 1.69 1.40 0.82
ROE (profits/market value) 6.33 6.41 1.01
Source: Florio (2014) based on Kowalski et al. (2013, tables 1 and 2)

Figure 1. ROS compared of Forbes 2000 companies

10
8,77
9
8 7,24 7,07
7
6
5
4
3
2
1
0
Average 2000 SOE Private Firms

Source: Florio (2014) based on Kowalski et al. (2013, tables 1 and 2)

This amounts to say, given the number of SOE in the sample, that they perform around 11% more in
terms of sales per firm; 35% more in terms of profits per firm; own 64% more assets per firm; and have
a 34% higher market value per firm.

12
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Table 2 expands this simple reading of the data. Asthe share of SOEs in the list is 10.02 %, in terms of
number, the striking conclusion is that the average SOE performs better than the average Forbes company
in the list9. The first column of Table 2 shows that the average ratio profit/sales (ROS) in the sample is
7.24%, while it is 8.77% for SOE (second column), i.e the profitability of SOE is 21% higher than the
average (third column); profits/market value (ROE) is aligned with the average; while the ratio
profits/assets (ROA) is 18% lower for the SOE than for the average, an underperformance which is driven
by the much larger assets owned by SOE relative to the average.

The direct comparison of public-private subsamples would obviously confirm that the SOEs outperform
their private counterparts on several indicators: as the total sample is roughly composed by 90% of
‘private’ firms and 10% of SOE, and the latter perform better than the average in terms of ROS, which
is the most relevant profitability indicator (and the least affected by valuations of assets and equity), the
average SOE must perform better than its private counterpart in terms of ROS: this is shown by Fig. 1.

The picture emerging from this global perspective on major public enterprises is far from the usual mantra
on the inefficiency of government ownership, and it does not support the view that PEsare loss-makers
exploited by politicians and insiders. This evidence on major playersdoes not identify either a public
mission for the PE, but - to say the least- one may suggest that governments earn net revenues from them,
and this should have some beneficial effects on public budgets. The benefit side of the story is however
unexplored until now, and cannot be understood at this aggregate level of analysis.

Let us turn now to the second possible research strategy, mentioned at the beginning of this section. Is
there an inverse correlation between the relevance of PE in the economy and the quality of government?
Again this is not the place to suggest a detailed analysis at country level, which would also be affected
by the poor state of statistical information about the incidence of PE in national economies. However,
some hints can be obtained considering together the distribution of major PEs by country and the
classification of such countries in terms of quality of government.

Figures 2 and 3 may seem to offer some support to the view that public enterpises are most often located
in countries with low quality of goverment. The ICRG index is a composite indicator, provided by the
International Risk Guide, which provides a synthetic measure of the Quality of Governance. It is made
up of three indeces, namely corruption, law and order, and bureaucracy quality. The range of the
aggregate ICRG index is 0-1, with higher values indicating higher quality of institutions. The correlation
between the number of SOE as percentage of Forbes 2000 firms in the country and the ICRG is - O.40.

The correlation between the ICRG indicator and market value of Forbes 2000 SOEs as a percentof GNI
is negative, with a Pearson correlation coefficient of -0.319, although it is not statistically significant.
This figuresare based on just 24 observations. Future research should enlarge this perspective and control
for possible confounding factors (including trivially GDP/per capita).

9
Taking the ratio of SOE/Total financial indicators with their share (10,02%) in terms of numbers.
13
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Figure 2. % number of SOEs by country out of Forbes 2000 and institutional quality (ICRG)

Source: Own elaboration on data from World bank and Kowalski et al. (2013).

Figure 3. Assets as % GNI of Forbes 2000 SOEs and institutional quality (ICRG)

Source: Own elaboration on data from World bank and Kowalski et al. (2013).

14
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Going back to the Forbes Global 2000 list, the countries that stand out more significantly in terms of
number of major SOEs aresome emerging economies, notably China with 70 out of the 204 identified
major SOEs, followed by India (30), Russia (9), Brazil (7), Indonesia (6), Korea (4). Among the most
developed countries there areSwitzerland (6), Poland (6), France (5), Poland, the US (3). The importance
of the SOEs would be better representedin terms of contribution of PE to GNP. Japan, New Zealand,
Italy, France, US, Britain, Norway, Canada, Germany10, stand out in this perspective. IMF data show
that the government of these countries owns non-financial assets worth at least 40% of GDP. Other
evidence on network industries mentioned by Florio (2013) shows that PEs in Europe are a persistent
feature of the Scandinavian countries.

Theremarkof a weak association between quality of government and the role of PEs in the economy
would be even clearer if we look back at PE in a historical perspective, i.e. if we study in which countries
the PE were born and developed in the XX century. These would include countries under an array of
entirely different political regimes from Soviet Union, fascist Italy, the empire of Japan, as well as
democratic and developed capitalist economies as the post war UK, the New Deal US, the social-
democratic Sweden.

Summing up, looking at major global players, PEs perform better than their private counterparts in terms
of the most significant financial ratio, the return on sales, and there is a weack negative correlation
between the quality of government and the importance of PE in the economy (albeit imperfectly both
variables are measurable). While this is not enough to conclude that PE mainly exist because government
give them a public mission, it seems enough to consider at least exaggerated the view that PEsmainly
exist to be exploited by politicians, otherwise we would have observed that they underperform relative
to their private counterparts and were concentrated in countries where the quality of government is bad.
At least looking at the major global players, the evidence does not support the public choice perspective
on PE.

3.2 The market for corporate control

A novel research strategy to answer the question is currently being developed by our team at the
University of Milan (Clò et al. 2014). Instead of focussing on the stock of existing private and public
enteprises we focus on deals involving the market for corporate control. We want to understand the
characteristics of firms operating in this market as acquirers, targets, vendors. There are four cases which
are of interest: the usual transaction, on which a wide literature exists, is that of a private firm acquiring
another private firm. There are however other three cases for which the evidence until now has not been
explored in a systematic way: privatization, i.e. when a PE is acquired by a private firm; publicization,
i.e. when a private firm is acquired by a PE; and restructuring of government control of enterprises, i.e.
when a PE is acquired by another PE. Some preliminary results have been reported by Clò et al. (2013),
but we are now working on a wider dataset, and I will report some preliminary findings based on this
new research are detailed below:

The standard theory for the market of corporate control is that efficiency would require that acquirers
perform better than targets. In this way, inefficient managers would be displaced by better managers. A
variant of this theory is due to Rhodes-Kropf and Robinson (2008). They suggest that on average
acquirers perform better than targets, but also that “like buys like”. Mergers and acquisitions would stress
complementarity among merging firms. In the perspective of the acquirer there is a trade-off. Clearly it
is beneficial for the acquirer to merge with a well performing target firm, but if the vendor and the
acquirer can freely negotiate, the bargaining power of the vendor is stronger the better the performance

10
See The Economist (2014a, p. 19)
15
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

of the firm it owns. Thus the acquisition price will be higher, and this should be discounted by the net
present value of the stream of future profits of the target firm.

How does this relate to our research question? Under the public choice perspective one wold expect that
the market for corporate control is efficient when private enterprises are on the acquirer side and on the
vendor side, because the two parties in the transaction will on average share a profit maximization
objective. Also, privatization would be efficient, because inefficient PEs will be acquired by well
performing private acquirer. In the public choice perspective, the rent-seeking behavior of the politicians
and of the managers of the PE would not contribute to market efficiency for the corporate control when
a publicization is involved.

To test this we use Zephyr, a database managed by the Bureau Van Dijk, reporting more than 860,000
completed deals that took place worldwide over the period 2000-201211. Differently from Clòet al. (2013)
we do not try to combine the ownership of acquirer, vendors and targets for each deal, as this would lead
to a too small sample. We focus however on ownership of the acquirer. Zephyr reports information about
the year of the deal, the name, country and sector of both the acquirer and target companies. Then we
match the data about enterprises involved in the deal on both sides provided by Zephyr with information
provided by Orbis, also a dataset managed by the Bureau Van Dijk, which reports selected financial
accounts and corporate governance information on companies worldwide. The years we consider are
2004-2012. We determine the ownership of each enterprise by looking at the nature of the top shareholder
the year before the deal. In detail, we consider an enterprise to be public only if in the year before the
deal its top shareholder was a PE or if it is directly controlled by a public institution (government, public
authority, etc.). This approach is much more flexible than the one involved in the previously mentioned
OECD study which considered SOE only firms whose equity is owned by a government entity by more
than 50%. As we want to include a large set of firms’ characteristics for future econometric testing (not
reported here) we have selected a sample of around 25,000 deals, of which around 2,500 involve a PE as
acquirer. Annex 1 gives some examples of the lessor.

We are particularly interested in examining the difference between the ROS of acquirer and targets,
because this difference is a crucial one first to confirm the generic theory of the efficiency of the market
for corporate control, and second, to confirm the hypothesis that PE behave inefficiently in this market.

Table 3 shows the following:

a) Private acquirers on average are bigger (on several dimensions, such as assets, sales, employees)
and perform better than the target they acquire. This can be seen as confirming that for the market for
corporate control works as expected by the standard theory.
b) On average public acquirers as well are bigger (over the same dimensions mentioned above) and
perform better than their targets. This would reject the view that PE are different in a market efficiency
perspective relative to their private counterparts
c) Confirming the findings with the much smaller sample based on Forbes Global 2000, on average,
the public acquirer is bigger (again in terms of employees, assets, operating revenues, added value) and
performs slightly better than the private acquirer. The mean of ROS of public acquirers is 15% against
12.5% of private acquirers.

11
Sources come from reports, international financial journals, company press release, electronic publications, company
websites, stock exchange information, https://www.bvdinfo.com/en-gb/products/economic-and-m-a/m-a/zephyr.
A small number of illustrative examples are reported in Annex 2.
16
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

d) The difference of ROS between acquirer and target is higher when the acquirer is public, which
would imply that they do not buy the like.The mean of this difference of ROS is 6.2 points for deals with
private acquirers and 8.5 points when the acquirer is public (with similar standard deviations).
e) Targets acquired by PEcan be compared with the targets acquired by private enterprises. The
targets of PE have on average; less total assets; lower Ebit; lower sales; lower operating revenues and
lower added values than targets of private enterprises. But the firms acquired by PE have on average a
higher ROS than those acquired by their private counterparts.

Table 3. Descriptive statistics on the Zephyr-Orbis sample 2004-2012

Public/priv
Private Public ate
A/ A/
Acquirer Target A-T T Acquirer Target A-T T A T
mea
Ros n 12.5 6.3 6.2 2.0 15.08.5 6.5 2.3 1.2 1.0
sd 22 23 29 24 31 22
freq 22,838 22,838 22,838 2,494
2,494 2,494
Operat mea 1,336,10 1,692,2 5,767,76
ing n 3,028,384 3 81
2.3 6,628,988 861,225 3 7.7 2.2 0.6
Reven 11,563,36 6,329,71 10,140,
15,652,97 3,973,87 15,658,6
ues sd 3 9 263
4 2 90
freq 22,838 22,838 22,838
2,494 2,494 2,494
mea 1,006,85 11.
Ebit n 527,525 353,849 173,676 1.5 1,103,613 96,756 7 4 2.1 0.3
2,083,25 1,264,9 3,517,45
sd 2,326,791 9 14 3,513,331 709,918 3
freq 22,838 22,838 22,838 2,494 2,494 2,494
mea 2,160,89 2,339,9 12,629,21 1,154,28 11,474,9 10.
n 4,500,860 3 67 2.1 9 3 36 9 2.8 0.5
Total 19,168,48 10,892,7 16,660, 32,050,93 5,295,70 31,808,4
Asset sd 9 43 410 6 0 26
freq 22,838 22,838 22,838 2,494 2,494 2,494

Source: Clò et al.2014

The following figures4 and 5 break down the sample by country and by sectors. In Western Europe there
is some evidence that the difference of ROS between public and private acquirers and their respective
targets while it is large in Russia. In terms of sectors, in services of general interest (transport and
utilities), energy and mining the difference of ROS between acquirer and target is higher for PE than for
private firms, while in other sectors it is similar or in favour of private acquirers as in the financial sector.

17
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Figure 4. ROS of acquirers and targets by

selected regions
MIDDLE EAST CHINA

35 25
PRIVATE PUBLIC PRIVATE PUBLIC
30
20
25

20 15

15 10
10
5
5

0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
NORTH AMERICA RUSSIA
25 25
PRIVATE PUBLIC PRIVATE PUBLIC
20 20

15 15

10 10

5 5

0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
EASTERN EUROPE WESTERN EUROPE
25 25
PRIVATE PUBLIC PRIVATE PUBLIC
20 20

15 15

10 10

5 5

0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
REST OF THE WORLD
25 The sample includes a total of 25,382
PRIVATE PUBLIC observations, of which the distribution across
20 countries is the following:

15 Private Public

10 Middle East 145 15

China 196 147
5
North Am. 1,441 6
Russia 1,961 665
Western E 13,331 1,093
0
Eastern E 1,784 170
Acquirer target A-T Acquirer target A-T
RoW 3,980 398
REST OF THE WORLD

18
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Source: Own elaboration of data from Clò et al. 2014.

Figure 5. ROS of acquirers and targets by industries

MANUFACT. ENERGY SU
25 25
PRIVATE PUBLIC PRIVATE PUBLIC
20 20

15 15

10 10

5 5

0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
OTHER SGI FINANCIAL
25 25
PRIVATE PUBLIC PRIVATE PUBLIC
20 20

15 15

10 10

5 5

0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
MINING OTHER SECTOR
45 25
PRIVATE PUBLIC PRIVATE PUBLIC
40
35 20
30
15
25
20
10
15
10 5
5
0 0
Acquirer target A-T Acquirer target A-T Acquirer target A-T Acquirer target A-T
Source:Own elaboration of data from Clò et al. 2014.

The sample includes the following observations:

Private Public
Manufact. 6,878 603
Energy su 237 377
Other SGI 2,353 368
Financial 5,944 488
Mining 943 250
Other sector 6,483 408

19
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Preliminary econometric testing rejects the view that when controlling for the dimensional characteristics
of the acquirer the results would change. To be conservative, one can conclude that, in the perspective of
the global market for corporate control, the public choice perspective on the inefficiency of PE is not
confirmed, as the public acquirers are more efficient than their targets.

4. Conclusions and policy implications

Public enterprises exist, they have not been wiped away by three decades of sustained privatization
policies. Our understanding of why they have survived an unprecedented attempt by the media,
economists, financial advisors, international organizations to convince governments to divest their
productive assets is still limited. Also our understanding of the performance and mechanisms in place for
PE is constrained by lack of suitable data, but also by a surprisinglack of scholarly interst for what is a
research topic with important policy implications.

In this paper we have discussed two opposite views about PE. We have (somewhat arbitrary) attributed
the ‘public choice’ label to a view that in nutshell see the PE as a way for politicians to circumvent the
constraints to rent-seeking associated with the regular funtioning of public adminsistration. Also
somewhat arbitrarily, we have attributed the ‘public mission’ label to a view that suggests that the
delegation of tasks of government to organizations with a certain budgtary autonomy and managerial
discretion is linked to a set of policy objectivesof general interest. The predictions of the two hypotheses
are different, and we have discussed some possible weak points of both theories.

It ispossible to empirically test these views, even if in a very tentative way. Under the public choice view
you would expect that PEs are in general less efficient in conventional terms than their private
counterparts, and that PEs should be more frequently observed in countries where the quality of
government in terms of democracy, rule of law, repression of corruption, etc. is low. Also, if the public
mission hypothesis were true, profitability of PE may be hindered by objectives different from just
providing revenues to government.

The result of our discussion of the evidence is that, at least for major players in the Forbes Global 2000
List around 2010, and for firms active in the worldwide market for corporate control as recorded by the
Zephyr data between 2004 and 2012, the public choice view does not find much support.

If the PEs are instruments of politicians for their own private agenda or for their willingnes to buy support
to be re-elected, whatever the rents extracted from such firms, these are not enough to lead them to
underperform relative to their private counterparts. If, instead, the PE are instrument of governments to
pursue a public mission (beyond the trivial one of supporting public finance by dividens accruing to the
Treasury) this public missionis not such as to disrupt the overall efficiency of the PE in the most
conventional terms of profitability.

Only further research may establish whether objectives in the public mission, such as supporting
employment, investment in infrastructure, research and technological development, social and territorial
cohesion are still in place for the PE, or if the public mission is gradually eroded and limited to a quasi-
fiscal perspective of generating funds for the public budget. There is, however, a clear policy implication
of our reading of the evidence. Any claim that privatization is always and everywhere needed to increase
efficiency of a country because PE are loss-makers and disturb the functioning of the market is not
supported by the data. At least at aggregate and global level, i.e. not focussing on individual case histories
of badly managed firms (there are many, of course), public enterprises in the global arena survive simply
because they perform well.
20
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Hence, the temptation for a government to sell a well performing PE to private interests and to cash some
immediate proceeds may be an emergency measure to save the government budget in the short term. But
emergency nationalizations of major banks in the UK and even of automotive firms in the US during the
recent global crisis show that the opposite may also be true. In fact, the same international institutions
that for decades have been advocating privatization in Latin America and in the global South, have had
no objections to nationalizations when these were acted in the US and the UK.

If, however, PEssurvive only because they ensure a welcome non-tax revenue to governments, or
emergency measures, in the end they will be very similar to private enterprises. For example, their
management will have no intrinsic motivation to internalize more deserving public missions.This may
become privatization in another form. Thus, a more general policy implication is that a healthy financial
condition is probably a necessary but not a sufficient condition to justify theexistence of public
enterprises. The ultimate historical test of their future role will be the advent of a renewed public mission.

Acknowledgements
I gratefully acknowledge financial support by CLAD in the occasion of the conference “XIX
CongresoInternacional del CLAD sobre la Reformadel Estado y de la Administración Pública”, Quito,
Ecuador, 11-14 November 2014.I am particularly grateful for the statistical work on Zephyr and Orbis
data to my collegues at the DEMM, University of Milan: Stefano Clò, Matteo Ferrari, Carlo Fiorio, and
to Chiara Del Bo for the analysis of ICRG data. Section 3.2 draws from an unpublished paper with them.
I am also grateful to Rosa Carmosino and Alessandra Cavada for editorial assistance.

ANNEX 1

Top 25 global SOEs

Forbe
Ran Count Sector according to NACE rev.
s Company ROS ROA ROE
k ry 2
rank
1 6 PetroChina CN 0.095 0.084 0.066 Extraction of crude petroleum
and natural gas
2 7 ICBC CN 0.272 0.011 0.078 Financial intermediation
3 8 Petrobras-Petróleo BR 0.175 0.068 0.089 Extraction of crude petroleum
Brasil and natural gas
4 15 Gazprom RU 0.260 0.093 0.149 Land transport and transport via
pipelines
5 17 China Construction CN 0.268 0.011 0.069 Financial intermediation
Bank
6 21 Bank of China CN 0.241 0.009 0.083 Financial intermediation
7 22 Sinopec-China CN 0.038 0.073 0.101 Mining support activities
Petroleum
8 25 Agricultural Bank of CN 0.192 0.007 0.071 Financial intermediation
China
9 29 GDF Suez FR 0.055 0.025 0.073
Electricity, gas, steam
10 34 China Mobile CN 0.247 0.137 0.092
Telecommunication
11 51 Banco do Brasil BR 0.103 0.015 0.146
Financial intermediation
12 60 Statoil NO 0.072 0.059 0.078
Extraction of crude petroleum
and natural gas
13 61 General Motors (*) US 0.046 0.045 0.124 Manufacture of motor vehicles
21
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

14 68 China Life CN 0.100 0.027 0.050 Insurance, reinsurance and

Insurance pension funding
15 77 Rosneft RU 0.226 0.111 0.122 Mining support activities
16 95 Saudi Basic SA 0.141 0.068 0.070 Manufacture of chemicals and
Industries chemical products
17 100 EDF Group FR 0.016 0.004 0.018 Electricity, gas, steam
18 136 State Bank of India IN 0.089 0.008 0.072 Financial intermediation
Group
19 144 Cnooc CN 0.296 0.191 0.079 Mining support activities
20 145 China Shenhua CN 0.251 0.112 0.070 Mining of coal and lignite
Energy
21 157 China Telecom CN 0.071 0.037 0.049 Telecommunication
22 171 PTT PCL TH 0.044 0.068 0.087 Wholesale trade
23 172 Oil & Natural Gas IN 0.190 0.096 0.081 Mining support activities
24 178 Sberbank RU 0.025 0.003 0.011 Financial intermediation
25 179 Ecopetrol CO 0.192 0.117 0.050 Extraction of crude petroleum
and natural gas
Source: Own elaboration from Kowalski et al. 2013

ANNEX 2

Selected examples of deals with a PE as acquirer.

Acquirer Target Vendor
Year 2005 Arriva PLC (GB) Sapav Spa (IT) Savda (IT)
Sector(*) Land Transport Land Transport (Nace 49) Land Transport (Nace 49)
(Nace 49)
Employee 33,186 210
Revenue(**) 2,494,182 14,026
Ros 7.26 2.3
Year 2005 Arriva PLC (GB) Sadem Spa (IT) Savda (IT)
Sector Land Transport Land Transport (Nace 49) Land Transport (Nace 49)
(Nace 49)
Employee 33,186 146
Revenue 2,494,182 13,810
Ros 7.26 -2.29
Year 2006 Ceza S. (CZ) Elektrocieplownia PSEG Global Llc (US)
Chorzow Elcho SP (PL)
Sector Electricity supply Electricity supply (Nace Fishing & aquaculture (Nace 3)
(Nace 35) 35)
Employee 29,905 160
Revenue 4,419,102 88,951
Ros 24.61 40.41
Year 2006 Electricite De Alpiq Holding A.G. (CH) UBS A.G. (CH)
France S.A. (FR)
Sector Electricity supply Electricity supply (Nace Finance &PensionFund (Nace
(Nace 35) 35) 64)
Employee 161,560 8,658
Revenue 52,617,588 5,617,588
22
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Ros 16.74 6.34

Year 2007 Electricite de Electricite de Strasburg Ceva Logistic LLC (US)
France S.A. (FR) S.A. (FR)
Sector Electricity supply Electricity supply (Nace Air transport (Nace 51)
(Nace 35) 35)
Employee 156,524 1,060
Revenue 61,161,000 478,058
Ros 17.55 12.07
Year 2007 Green Cargo AB Ahrens Akeri AB (SE) Ahrens Holding AB (SE)
(SE)
Sector Land Transport Transportation service Land Transport (Nace 49)
(Nace 49) (Nace 52)
Employee 3,115 18
Revenue 652,219 2,395
Ros 4.8 2.2
Year 2007 Petroleos de C.A. La Electricidad de AES Corporation (US)
Venezuela S.A. (VE) Caracas S.A.C.A. (VE)
Sector Extraction oil & gas Fishing & aquaculture Energysupply (Nace 35)
(Nace 6) (Code 3)
Employee 52,816 n.a.
Revenue 75,788,768 570,725
Ros 22.03 n.a.
Year 2008 GDF (FR) SUEZ (FR) SUEZ (FR)
Sector Energy supply Water supply (Nace36) Water supply (Nace 36)
(Nace35)
Employee 47,650 192,821
Revenue 27,793,000 47,543,100
Ros 16 10.66
Year 2008 Hunan Nonferrous China Tundsten& High- Zingong Cemented Cabride
Metals Corporation Tech Metarials Ltd (CN) Ltd (CN)
Limited (CN)
Sector Mining ofmetal ores Manufacture of metals Manufacture of metals (Nace
(Nace 7) (Nace 24) 25)
Employee 22,570 1,770
Revenue 2,017,376 140,869
Ros 4.02 -2.76

Year 2009 Electricite De EdfLuminus (BE) Centralica PLC (GB)

France S.A. (FR)
Sector Energy supply (Nace Energy supply (Nace 35) Mining (Nace 61)
35)
Employee 155,931 953
Revenue 65,831,000 2,594,359
Ros 15.27 2.69
Year 2009 ENEL Spa (IT) ENDESA SA (ES) ENDESA SA (ES)

23
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Sector Energy supply(Nace Energy supply(Nace 35) Energy supply(Nace 35)

35)
Employee 75.981 26.587
Revenue 61.184.000 22.836.000
Ros 16 25.14
Year 2009 Sociedadpara la Alcorta Forcing Gourp Mahle GMBH (DE)
Trasformacioncomp SA (ES)
etitive SA (ES)
Sector Construction of Manufacture of metal Manufacture of vehicle (Nace
building (Nace 41) (Nace 25) 29)
Employee 195 166
Revenue 140,923 32,118
Ros 2.82 1.82
Year 2010 Edison Parco Eolico San GamesaEnergiaSociedadAnon
EnergieSpeciali Spa Francesco Srl (IT) ima (ES)
(IT)
Sector(*) Energy supply(Nace Energy supply(Nace 35) Manufacture ofmachinery
35) (Nace 28)
Employee 60 n.a.
Revenue 80,820 486
Ros 27.04 73.28
Year 2010 EmpresasPublicas GeneradoresHidroelectri IberdrolaEnergia SA (ES)
de Medelin (CO) cos SA (GT)
Sector Water collection Energy supply(Nace 35) Energy supply(Nace 35)
(Nace 36)
Employee n.a. n.a.
Revenue 1,482,276 n.a.
Ros 35.19 n.a.
Year 2011 CassaDepositi e Trans Austria Gasleitung Eni Spa (IT)
Prestiti (IT) GMBH (AT)
Sector Financial & Land Transport (Nace 49) Extraction of oil & gas (Nace 6)
insurance (Nace 66)
Employee 3,916 15
Revenue 1,477,500 283,825
Ros n.a. 48.48
Year 2011 SociedadEstatal de Ebro Food SA (ES) Group Torras SA (ES)
ParticipationeIndus
triales (ES)
Sector Finance & Pension Manufacture offood (Nace
Fund (Nace 64) 10)
Employee 30,759 4,592
Revenue 3,746,298 1,736,431
Ros -3.83 12.95
Year 2012 Interfin (BE) Sibelga (BE) Electrabel (BE)
Sector Finance& Pension Energy supply(Nace 35) Energy supply(Nace 35)
Fund (Nace 64)
Employee 1,061 n.a.
Revenue 367,139 361,922
24
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Ros 20.12 18.5

Year 2012 KFW European Aeronautic Daimler AG (DE)
Bankengruppe (DE) Defence & Space
Company EADS NV(NL)
Sector Finance& Pension Transport equipment Manufactureof vehicle (Nace
Fund (Nace 64) (Nace30) 29)
Employee 4,765 133,115
Revenue 2,670,000 49,506,000
Ros n.a. 3.28
(*) For a description of the sector related to the Nace Code see
http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/KS-RA-07-015/EN/KS-RA-07-015-
EN.PDF;
(**) Thousand of Euro; Source: own elaboration data from Bureau van Dijk

REFERENCES
Alonso J. M., J. Clifton, D. Díaz-Fuentes, M. Fernández-Gutiérrez, and J. Revuelta (2013), ‘The race for
international markets: Were privatized telecommunications incumbents more success full than thier
public counterparts?’ in International Review of Applied Economics 27 (2): 215-236.
Bernier L. (forthcoming 2014-15), Public enterprises today: mission, performance and governance -
Learning from fifteen cases, Book project in the CIRIEC series « Social Economy & Public
Economy » Peter LANG publisher.
Bernier L. (2014), ‘Public enterprises as policy instruments: the importance of public enterpreneurship’
in Journal of Economic Policy Reform (special issue).
Bös D. (1987), ‘Public sector pricing’ in A. Auerbach and F. Feldsteid (eds.) Handbook of public
economics, Elsevier Sciences publishers B. V.(North Holland), pp. 129-212.
Boycko M., A Shleifer and R.W. Vishny (1996), ‘A theory of privatisation’ in The Economic Journal
106 (435): 309-319.
Ceriani L. and Florio M. (2011), “Consumer’s surplus and the reform of network industries: a primer”,
Journal of Economics, 102(2), 111-122.
Chavez D. and Torres S. (eds), 2014, Reorienting Development: State-owned Enterprises in Latin
America and the World, Transational institute.
CIRIEC (2012), The future of public enterprise mission, performance nadgovernante: Learning from
success and failures. A research agenda 2012-2014. Paper draftedbyMassimo Florio and adopted
by the Steering Committee of the Project: available at
http://www.ciriec.ulg.ac.be/fr/telechargements/CIRIEC_project_FUTURE%20of%20PE2012-
2014_%20with%20template.pdf
Clifton J., F. Comín and D. Díaz-Fuentes (2006), ‘Privatizing public enterprses in the European Union
1960-2002: Ideological, pragmatic, inevitabile? In Journal of European Public Policy 13 (5): 736-
756.
Clò S., M. Ferraris, C. Fiorio, M Florio (2014), ‘Ownership and the market for corporate control’, Mimeo,
DEMM, University of Milan.
Clò S., C. Del Bo, M. Ferraris, C. Fiorio, M. Florio and D. Vandone (2013), Publicization versus
privatization: Preliminary findings 2000-2012. CIRIEC working papers.
http://www.ciriec.ulg.ac.be/fr/telechargements/WORKING_PAPERS/WP14-03.pdf
De Fraja G. and F. Delbono (1990), ‘Game theoretic models of mixed oligopoly’ in Journal of economic
Surveys, Volume 4, Issue 1, pages 1–17, March.
Del Bo C. and Florio M. 2012, ‘Public enterprises, planning and policy adoption: three welfare
propositions’ in Journal of Economic Policy Reform Volume 15, Issue 4.

25
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Drèze J. and N. Stern (1987), ‘Theory of cost-benefit analysis’ in A. Auerbach and F. Feldsteid (eds.)
Handbook of public economics, Elsevier Sciences publishers B. V. (North Holland).
The Economist (2014a), Setting out the store. State owned Assets. January 11.
The Economist (2014b), Leviathan as capitalist. State capitalism continues to defy expectations of its
demise.June 21.
Finger M. and C. Rentsch (2014), ‘Yes, no, maybe: The Ambiguous relationships of State-owned
enterprises with the State’, Mimeo paper presented at CIRIEC Workshop, Brussels 2013.
Florio M. (2014a), Applied Welfare Economics. Cost-Benefit Analysis of projects and policies,
Routledge.
Florio M. (2014b), ‘Contemporary public enterprises: innovation, accountability, governance’ in Journal
of Economic Policy Reform, published online, 2 June.
Florio M. (2013a), Network Industries and Social Welfare.The Experiment that Reshuffled European
Utilities, Oxford University Press.
Florio M. (2013b), ‘Rethinking on public enterprise: Editorial introduction and some personal remarks
on the research agenda’ in International Review of Applied Economics 27 (2): 135-149(15).
Florio M. and F. Fecher, eds (2011), ‘The future of public enterprises: Contributions to a new discourse’
in Annals of Public and Cooperative Economics 82: 361-373.
Florio M. (2004), The great divestiture. Evaluating the welfare impact of the British privatizations 1979-
1997, The Mit Press.
Florio M. (2002), “Economists, privatization in Russia and the waning of the “Washington Consensus”
in Review of International Political Economy, n°2/2002, pp. 359-400.
Greiling D., Grüb B, Huber A. (2014), ‘Sustainability Reporting in Public Enterprises: Basic Ideas and
their Development’ in Journal of Economic Policy Reform.
Grönblom S. and Willner J. (2014), ‘Organisational form and individual motivation: public ownership,
privatisation and fat cats’ in Journal of Economic Policy Reform.
Haney A. B. and M. Pollitt (2013), ‘New models of public ownership in energy’, in International Review
of Applied Economics 27 (2): 174-192.
Hefetz A. and M. Warner (2014), ‘Privatization and its reverse: Explaining the dynamics of the
Government contracting process’ in Journal of Public Administration Research and Theory 14 (2):
171-190.
Hughes O. E. (2003), Public management and administration: An introduction, Palgrave Macmillan.
Kowalski P., et al. (2013), "State-Owned Enterprises: Trade Effects and Policy Implications", OECD
Trade Policy Papers, No. 147, OECD Publishing.
Laffont J-J. (2005), Regulation and Development, Cambridge University Press.
Laffont J-J. and Tirole J. (1993), A Theory of Incentives in Procurement and Regulation,The Mit Press.
Lethbridge J. (2014), ‘Public enterprises in the healthcare sector – a case study of Queen Elizabeth
Hospital, Greenwich, England’ in Journal of Economic Policy Reform (special issue).
McDonald D. (2014), ‘Rethinking Corporatization and Public Services in the Global South’, Zed Books,
London.
Millward R. (2011), ‘Public Enterprise in the Modern Western World: An Historical Analysis’ in Annals
of Public and Cooperative Economics, Vol. 82, Issue 4, pp. 375-398.
Millward R. (2005), Private and public enterprise in Europe, Cambridge University Press.
Mueller D. C. (2003), Public choice III, Cambridge University Press.
Musacchio A. and S. G. Lazzarini(2014), Reinventing State Capitalism. Leviathan in Business, Brazil
and Beyond, Harvard University Press.
Polidori P. and D. Teobaldelli, (2013) ‘Prosocial behaviour in the production of publicly provided goods
and services: an overview’ in International Review of Applied Economics, Special Issue: March
2013.
26
XIX Congreso Internacional del CLAD sobre la Reforma del Estado y de la Administración Pública, Quito, Ecuador, 11 – 14 nov. 2014

Rees R. (1976), Publicenterpriseeconomics, London School of Economics handbooks in economic

analysis, Handbooks in Economic Analysis.
Rhodes‐Kropf M. and D. T. Robinson (2008), ‘The market for mergers and the boundaries of the firm’
in The Journal of Finance, Blackwell Publishing Inc., 63:3, pp. 1169-1211.
Thomas S. (2013), The Seven Brothers: Then there were three, PSIRU. Business School, University of
Greenwich.
Warner M. E. (2011), ‘Competition or cooperation in urban service delivery?’ in Annals of Public and
Cooperative Economics, Vol. 82, Issue 4, pages 421–435.
World Bank (1999), Bureaucrats in Business: the economics and politics of government ownership. A
World Bank Policy Research report
Massimo Florio
University of Milan
Department of Economics, Management and Quantitative Methods
Via Conservatorio 7
20122 Milano

27