Privatization of network industries in the EU Massimo Florio Department of Economics, Management, Quantitative Methods, Università degli Studi di Milano

Privatization of network industries in the EU

Massimo FlorioDepartment of Economics, Management, Quantitative Methods, Università degli Studi di Milano

EU Industrial Policy

Objective

The reform paradigm that will be discussed includes privatization, vertical disintegration, price regulation and liberalization. The sectors that will be considered include mainly electricity, gas and telecoms.


Reforms in the EU

The key feature of these industries is that they include both a major fixed-cost component, the network, under increasing returns to scale, and several potentially competitive upstream or downstream operations.


Reforms in the EU

This feature leads to natural monopoly for the network services, and potential market dominance of the vertically integrated network's owner.


Reforms in the EU

Wide disappointment with earlier private monopolies or oligopolies, often sheltered by various forms of concessions, lead most European governments in the last century to take control of industries often plagued by collusion, underinvestment, and price discrimination


Former wave of privatization

Historically, vertically integrated public monopoly was the industry structure preferred by most governments.

Networks and operations were nationalised at the same time.


Former wave of privatization

Tariffs were strictly regulated and included cross-subsidies for some segments of the residential users (low-income, rural or other disadvantaged areas, etc).

After World War II, service provision and investment in the networks under public ownership increased significantly in many countries, particularly energy and telecoms


The starting of the new privatization wave In the 1970s, however, several nationalised

SGEI in Europe were under budgetary stress, and widely perceived as inefficient.

The United Kingdom, in the mid 1980s, was the front-runner of reform, while, among the EU member states, France was often regarded as a country adverse to move away from public monopoly.

The starting of the new privatization wave In fact, in the last 15-20 years virtually all the

European countries have adopted dramatic regulatory reforms in the network industries.

Wide variations around a common policy trend, however, can be, observed across countries and sectors.



The starting of the new privatization wave

We treat this story as a natural experiment in policy reform, to study its welfare effects on users.


The typical EU reform

A new paradigm (and legislation) has emerged.

We want to test the paradigm on empirical grounds.

We consider three industries (electricity, gas, telecom), each representing different stages and features of the reform story.

Empirical strategy 1/3

Take advantage of cross-country variability of the reform pattern and trends, and to consider the reform pattern as a quasi-experiment.




Under the quasi-experiment approach, one outcome variable is selected, e.g. productivity, investment, or wages, and –after considering country or industry-specific control variables- a set of predictions about the impact of reforms is tested.



Focus on consumer prices. After all, we think that the main justification of a regulatory reform in SGEI should lie in the potential benefits for users.

Under very general assumptions consumer surplus changes are inversely related to price changes, thus focussing on consumer prices is a sensible shortcut for welfare analysis.


The main data set 1/2

Consider both actual price changes, as recorded by Eurostat (most data start in 1991). These are country-average data.

Perceived price fairness by (individual) users, as recorded by Eurobarometro, a large EU sponsored survey, for which four waves and around 60,000 data points are available.


The main data set 2/2

Other data set are also used including the International Energy Agency (IEA, for electricity and gas), the International Telecommunication Union (ITU, for telecoms), the World Development Indicator (WDI).


Main research questions 1/2

Hence, we can split our research question in two:

1. Are average consumer prices lower (after controlling for non-reform related factors) in countries that implemented the reforms?

2. Are individual consumers happier with the price they pay in the latter countries?

If the answer is 'yes' to both questions, the reform is on a promising track, because it is at the same time supported by objective evidence and it enjoys public support.


Main research questions 2/2

If prices are low, but perceptions are less positive, reformers should carefully consider why people are unhappy.

If there is no evidence that prices are lower and that people are happier in countries that adopted the reforms, something is going wrong.


Electricity


Gas


Telecom


ECTR indicators


ECTR Indicators


ECTR Indicators


ECTR Indicators


The paradigm: three pillars

The first pillar of the paradigm:

the case for privatization

The second pillar: Unbundling

The third pillar: Market opening

OWNERSHIP OF THE INCUMBENT OPERATOR

OWNERSHIP UNBUNDLING OF THE NETWORK

MARKET ENTRY AND COMPETITION

INTEGRATED PUBLIC MONOPOLY

PUBLIC NO NO

INTEGRATED PUBLIC MONOPOLY WITH NETWORK ACCESS

PUBLIC

NO

Accounting Legal / Functional

separation

Public Competitors

YES Private Competitors

Mixed

UNBUNDLED PUBLIC INCUMBENT

PUBLIC

YES

Private Public

Network Network

Public Competitors


Mixed

INTEGRATED PRIVATE MONOPOLY

PRIVATE NO NO

INTEGRATED PRIVATE INCUMBENT WITH NETWORK ACCESS

PRIVATE

NO

Accounting Legal / Functional

separation

Public Competitors


Mixed

UNBUNDLED PRIVATE INCUMBENT

PRIVATE

YES

Private Public

Network Network

Public Competitors


Mixed

Reminder

Monopoly/private oligopoly welfare effects in the absence of (public) government

intervention



Natural monopoly


Natural monopoly


Public Monopoly

p1=c(1+α)

c*

q1


Private monopoly, unregulated


Private monopoly, with a price-cap


Duopoly without costs of unbundling


Duopoly with costs of unbundling


Oligopoly 1/4

After unbundling of the network, the regulator can offer a license to any new entrant

n is the number of providers No price regulation Network service costs covered by the

taxpayer


Oligopoly 2/4

Profit maximization for each provider:


Oligopoly 3/4

Market liberalization in network industries in poorly interconnected regional markets seems to result in oligopoly

Under oligopoly, prices can be higher than under a vertical integrated public monopoly


Oligopoly 4/4


Emergence of a dominant policy paradigm on the reform of network industries in the EU

Divestiture of public ownership, unbundling, liberalization

Telecommunications, electricity and gas markets structures in the last 3 decades in the EU-15 members


Contrast the predictions and prescriptions of the paradigm with a theoretical discussion of the welfare impact of the reform

Comparative static examples and a review of some standard microeconomic assumptions on the role of ownership, economies of scale and scope, governance, and market forms

Actual success of the reform

Main European Energy Directives

Year Name Objective

1996 Directive 1996/92/EC concerning common rules for internal market electricity

The Directive provides for a gradual market opening in 3 steps. Each Member State shall open the market such that they respect at least the minimum market opening; they can go for a further opening including liberalisation

1998 Directive 1998/30/EC concerning common rules for the internal market in natural gas

Its main objective is to create a single natural gas market integrated, competitive, and regulated at EU level. The genrale principle promoted is the third party access.

2003 Directive 2003/54/EC concerning common rules for the internal market in electricity and repealing Directive 96/92/EC

Its final aim is to speed up the integration process and the development of competition in the electricity industry. It identifies a detailed timetable for market opening to all type of customers.

2003 Directive 2003/55/EC concerning common rules for the internal market in natural gas repealing Directive 98/30/EC

It lays down the rules relating to the organisation and functioning of the natural gas sector, access to the market, the criteria and procedures applicable to the granting of authorisations for trasmission, distribution, supply and storage of natural gas and the operation system

2004 Directive 2004/67/EC concerning measures to safeguard security of natural gas supply

It establishes a common framework within which Member States shall define general, transparent and non-discriminatory security policies compatible with the requirements of a competitive internal gas market.

2005 Directive 2005/89/EC concerning measures to safeguard security of electricity supply and infrastructure investment

It establishes framework within which Member States are to define transparent, stable and non-discriminatory policies on security of electricity supply compatible with the requirements of a competitive internal electricity market

Main European Telecommunications Directives

Year Name Objective

1990 Directive 1990/388/EC concerning competition in the markets for telecommunication services

It introduces competition in the telecommunication sector. Memeber States shall withdraw all special or exclusive rights for the supply of telecommunication services and ensure that the operator is entitled to supply such telecommunication services

1996 Directive 1996/19/EC amending Directive 1990/388/EC, concerning the implementation of full competition in telecommunication markets

Its aim is to fully liberalize the telecommunication services from 1998 onwards. The provision of telecommunication services may be subjected only to a general authorization or a declaration procedure

2002 Directive 2002/19/EC concerning access to, and interconnected of, electronic communications networks and associated facilities

The purpose of the Access Directive is to ensure that the relations

between operators are left as far as possible to competitive market

forces, while giving flexible powers to the NRAs to intervene

where market forces alone cannot ensure the fulfillment of the

policy objectives set out in the Framework Directive.

2002 Directive 2002/22/Ec concerning universal service and users’ rights relating to electronic communications network and services

It is founded on the premise that competition and

market forces are the most effective means to satisfy user needs,

but provides national regulatory authorities with the necessary

powers to protect users’ interests where they need to do so.

2002 Directive 2002/77/EC concerning competition in the markets for electronic communications networks and services

In line with the provision established under Directive 90/338/EC, it requires Member States to abolish special rights or exclusive rights relating to electronic communications networks and services and to ensure that any undertaking is entitled to provide them.



A typical 'European-style' reform package has four main dimensions:

1. divestiture of public ownership;2. unbundling of the network from service

operations;3. price regulation by an independent office

(usually in the form of price capping);4. lifting of restrictions to market entry and

eventually full liberalisation.



The EU institutions have been in general strongly supportive of the reforms. While neutral on public ownership divestiture, the EC has proposed over the years a number of important directives on transport and energy sectors that push the Member States towards a homogenous pattern of regulatory legislation, see e.g. CEC (2007).

A new paradigm (and legislation) has emerged.


Testing the paradigm 2/2

We consider the EU15 (low reliability of data from NMS, shorter time series, very dierent institutions up to 1990s)

Government-owned providers in energy were not loss makers, hence their prices covered costs, and comparison with pricing of private firms is more sensible.

EU Industrial Policy - a.a. 2010-2011

Telecommunications market

EU Industrial Policy - a.a. 2010-2011

Electricity market

The electricity industry

The electricity industry can be described as including four different activities:

1. generation,

2. transmission (the high voltage network),

3. distribution (the middle and low voltage network),

4. retail (supply to final consumers).

Only transmission and distribution are natural monopolies, at the national and regional level, because of the high network fixed sunk costs.

Data

The primary source for the average price variables are the International Energy Agency (IEA). The only alternative data source available for a EU15-wide analysis would be the Eurostat, which however is available for a much shorter time series as it starts at best in 1991

The IEA net-of-tax electricity prices for households are expressed in €/unit and present a correlation with household net-of-tax electricity prices

As for the regulatory reform variables, including measures of entry regulation, public ownership, market structure, vertical integration, in varying levels of detail ETCR data set

Results

The following table presents the result of the estimation of various models for electricity prices. The first column provides a test of the reform as a whole estimating a model where the reform variable is the ETCR sector score ranging from 0 to 6. In the following columns, we use as reform variables the set of dichotomous dummy variables. All models are estimated using GMM

The highly significant lagged dependent variable sheds doubt over similar panel data models in earlier literature without a dynamic specification, whose estimates are very likely to be affected by omitted variable bias

As for the electricity industry, it is interesting to notice that the lagged dependent variable is highly significant also after inclusion of year fixed-effects, regulatory variables and other controls

Results

If, instead of the ETCR score, a dummy variable for each of the three dimensions of the electricity industry reform is included, it emerges that only public ownership variable (ERpo_d) presents a consistently significant coefficient, reducing average price by roughly 0.3 (log) points.

Vertical integration, which is strongly correlated with public ownership, is never statistically significant.

Entry regulation is significant at the 10% significance level although it is mainly the freedom of choice of providers that drive this effect

(1) (2) (3) (4) (5)GMM GMM GMM GMM GMM

L.lEAprinet_kw 0.854*** 0.414*** 0.407*** 0.428*** 0.414***(0.000) (0.000) (0.000) (0.000) (0.000)

ER 0.003(0.690)

ERpo_d -0.376*** -0.356*** -0.350*** -0.366***(0.000) (0.000) (0.000) (0.000)

ERen_d -0.106*(0.072)

ERvi_d 0.099 0.079 0.037 0.066(0.233) (0.393) (0.598) (0.462)

ERen1_d -0.043(0.483)

ERen2_d -0.015 -0.024(0.741) (0.596)

ERen3_d -0.098* -0.100** -0.092*(0.073) (0.049) (0.088)

lEAscmbf 0.025** 0.016 0.021 0.023 0.021(0.011) (0.517) (0.408) (0.354) (0.409)

lEAimports 0.000 -0.000 -0.000 -0.000 -0.000(0.720) (0.913) (0.900) (0.936) (0.917)

lEArescons -0.022** 0.002 -0.008 -0.007 -0.005(0.012) (0.948) (0.805) (0.820) (0.870)

lMWgdppc 0.016 0.231*** 0.215*** 0.225*** 0.227***(0.274) (0.000) (0.000) (0.000) (0.000)

Year fixed-effects yes yes yes yes yesConstant -0.292** -0.676** -0.683** -0.612** -0.647**

(0.026) (0.011) (0.011) (0.015) (0.014)Observations 402 402 402 402 402Dep. var.: log average price for electricity supplySource: Authors' calculations using IEA, WDI source data. IEA data used for price series. For exact source and variable definition refer to the label (in

italics in the first column). Notes: ER = electricity reform; ERpo_d=public ownership; ERen_d =entry regulation; ERvi_d =vertical integration:EAprinet = Electricity net-of-tax price for households, submitted to the IEA Secretariat by Administrations (in €/unit) (Source: IEA; (a)); EAscmbf =

Electricity source: Total Combu. Fuels (GWh/Tj) (IEA); EAimports = Electricity import (GWh) (Source: IEA; (a)); EArescons = Electricity residential consumtion (GWh) (Source: IEA; (a)); MWgdppc= Nominal GDP (billion of euro) (Source: WDI; (a))

Robust p-values in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Year dummies included in all models. ar1p and ar2p report the p-values of tests for first-order and second-order serial correlation, asymptically N(0,1) under the null of no autocorrelation. GMM results are one-step estimates with heteroskedasticity-consistent standard errors (in parentheses) and test statistics. Instruments used in all GMM equations include dependent variable at lags t-3, t-4, t-5, the predetermined regulatory variable (at time t-1 and earlier) and exogenous variables.

GMM estimation of dynamic panels for electricity prices

Concluding remarks

The overall impact of the reform package on consumer prices, as summarised in the ETCR industry score, is never statistically significant for electricity prices. This is what one would expect if the adverse allocative effect of privatisation is just counterbalanced by the beneficial effect of liberalisation

Similar results for other industries


Concluding remarks

The overall evidence on the correlation of liberalisation reforms with electricity prices is mixed. Consumers welfare didn’ t increase significantly, while there is some evidence of regressive effects (higher tariffs for the poor)

Main reference: Florio M., Network Industries and Social Welfare. The Experiment that Reshuffled European Utilities, OUP, 2013

Documents

Privatization of network industries in the EU Massimo Florio Department of Economics, Management, Quantitative Methods, Università degli Studi di Milano