Romania, Feb 2016

Memo 1

Proposed methodology

(for consultation)

To make it more convincing for the donor, audience and the external evaluator, our methodology and data collection must take into account / make reference to / follow as much as possible the official strategies and existing practice on the subject, meaning for example (but not exclusively):

  • The broader policy frameworks on the subject (OECD, EU)
  • Our national governments’ strategies and action plans, to the extent they exist
  • The Pareto principle: 80% of results with 20% of effort

In the same time, we also have to ensure as much comparability as possible between our national reports and datasets, in order to have a meaningful overall views of the SOEs problems and policies. The right balance between these requirements is something we must pay attention to, both in the initial planning phase but also in execution.

Third, we must decide, probably in the fisrt meeting when we discuss indicators, how we actually operationalize “clientelism”, which is actually the main promise we made in this project. In practice, we will decide if we:

(a)stick close to a narrow definition (clientelism = corruption and probably some nepotism), or

(b)enlarge it (clientelism = bad governance more broadly, perpetuated by the mechanisms used to run these entities).

The two may look the same, but are not: the first hints more as deviations from an otherwise good set of rules; the second to a systemic failure of the very set of rules. From our previous experience the reality of research will push and pull us towards one or the other interpretation at different moments. It is therefore better to discuss it explicitly upfront and at least be aware that such a dimension exists.

Framework of reference

From all preliminary consultations it looks like we will use, or at least start from, the OECD methodology and classification of the SOEs. In the next stage we will also explore other sources, some general (World Bank, WTO), other country-specific. All partners should get and briefly assess their governments’ strategies and policies towards SOEs before the kick off meeting.

Two countries in this project are OECD members (Italy, Czech Republic); two are not (Romania, Bulgaria). In the latter cases we could include among the project objectives some advocacy for joining the OECD’s working group on SOEs, which is not conditional on full membership and apparently encouraged. This would represent both a nice deliverable for us and a help in activity, because being there may force governments to produce more data and action plans for the sector in a benchmark format.

Selection of sectors

This is something we should more or less finalize during the kick off meeting, with the important aims to have good coverage in terms of size, diversity of sectors and comparability among countries, so please give it some thought in advance. We can follow the OECD classification, which is rather subtle and follows the logic of microeconomics analysis, or alternatively devise one of our own.

OECD:

1. Natural monopolies, ex infrastructure: roads, energy distribution, water-sewage etc.

2. Imperfect contracts: high externalities and unpredictable trends in activity, ex airports, hydropower, forestry etc, which have not just a commercial but also interconnectivity, balancing or protection functions, etc, so are hard to privatize because it is difficult to define and/or enforce in practice the public service obligation when you have a private operator.

3. Incumbent operators (ex postal services). Even though there are private operators on these markets, the state company alone still fulfils the public service obligation of universal coverage.

4. Industrial policy, when the state owns a company in order to promote a certain social goal in terms of economic structure, employment, regional development, innovation promotion, etc. One may like it or not – in general liberals are sceptical; socialists love it - but at least in the EU it is legitimate as long as it is transparent, deliberate (so not by default) and approved according to the existing rules.

Two observations here:

  • For 1-3 it is crucial to separate in our analysis the commercial and the public service parts, which is not always easy.
  • And it is in relation to (4) when the difference in the interpretations of clientelism (a and b above) become more visible in practice: one man’s legitimate policy may be other man’s clientelistic rent seeking at the expense of the wider community.

EU:

The Union has a specific regulatory framework and a lot of practice, falling into the area of competition law, on a specific class of economic agents: SGIs / SGEIs.

Services of general interestare not the ordinary services provided in the market, but services which exhibit special traits. They are usually performed by the state, regions or municipalities in the public interest and their special trait is that they would be under-supplied in the market or they would be supplied in lower quality or geographicalscope if there was no financial compensation provided by public authorities. Legitimacy of such financial aid for services of general interest works on the assumption that the market fails, or partly fails, in these sectors. The market failures which are usually mentioned in this argumentation are the basic four from microeconomic analysis. Services of general economic interest(“SGEI”) are the subgroup of services of general interest, with a more pronounced economic character.

In this project we will of course come across SGEIs. What we must be aware of is that SGEIs overlap with SOEs, but not fully, both ways: there may be non-public, private SGEIs which we will not cover; and there may be SOEs which are not SGEIs.

On the other hand, we have our own theory, sketched in the original application, towards dividing the SOEs into two broad conceptual categories.

A. Those which are more like public agencies, usually at the central level, which are financed from the budget through lump sums of assigned taxes, exert some sort of legal monopoly in activity. What they provide is mostly paid by the taxpayer, not by the direct clients of services. In Romania, this would be the situation of CNADNR, the national roads&highways “company”.

B. Companies which are purely profit-based and carry out 100% economic activity, just that they happen to be state owned (Oltchim, Romgaz).

Of course there are many cases of companies falling in-between, operating for both profit and for public service (Hidroelectrica, Romsilva), as discussed above, which makes our work more complex and beautiful!

In practice, we first need to define the nature of the public service performed: public good that would not be supplied (or would be supplied under different conditions in terms of objective quality, safety, affordability, equal treatment or universal access) by the market without public intervention.

Then we separate SOEs by the nature and complexity of the public service provided (they may serve several public goals at once) and the weight of this obligations. The logic is that these obligations must be compensated for by the state which imposes them – and not opaquely cross-subsidized from the for-profit activity. In a nutshell, this is the starting point for our economic analysis of SOEs.

All these frameworks for analysis described here must exist as a reference for all the partners in the project. Even if we may follow slightly different paths with the country-specific analysis, they will ensure we speak a common language and are able to compare the data.

In Romania the current government is in course of devising a meaningful categorization of SOEs for purposes of monitoring and policy making, this being a declared priority for 2016. There is a history of efforts and even legislative initiatives in this respect, on assessing losses incurred by central and local SOEs and regulating their corporate governance (appointing boards and management), mainly driven by the IMF conditionality. This being the case, it may be true for Bulgaria, too. The Romanian initiatives were not a success: the financial situation of most SOEs (and even their total number !) is still relatively opaque and the attempts to regulate – i.e. depoliticize – the management turned into a fiasco.

Currently the Romanian government has preliminarily defined four categories of SOEs: economic; strategic (of course!); utilities; structural. It is not yet clear how they are precisely defined and in what way relate to OECD ones; and they may change anyway. The point is that we should be (a) aware of what our governments are doing; (b) refer to their plans; (c) try to engage with them in the most helpful manner possible, and if we succeed this will be also a measure of achievement in the project.

Ownership: how much state

SOEs, central, regional or local, may be fully or partly owned by the state. We must decide where we draw the line for inclusion: do we stay with the 100% owned entities? Go down to 51%? Other limit?

If we go for less than 100% - i.e. SOEs with minority private shareholders – additional EU frameworks may apply: Shareholder Rights Directive and its (proposed?) revisions.

How do we analyse clientelism

There will be a matrix with two broad sections: (1) governance; (2) corruption indicators

1. Governance.To what extent the practice in the respective country follows the OECD guidelines, slightly adapted to the local risk profile. The risk profile will be more specifically defined after we do (2).

  • Rationale state ownership
  • state role as owner
  • SOEs in marketplace
  • Equitable treatment shareholders
  • Stakeholder relations
  • Disclosure / transparency
  • Board responsibility

What are the methods we use here (quantitative, qualitative)?

2. Corruption indicators.Here we have to be really creative, along two big directions:

  • Money stolen: manipulated procurement (when they buy goods and services); selling whatever they produce under the market
  • Nepotism / party / clan / etc clientelism in HR.

They will probably be sector-specific, so we can start to explore them once we have done the selection of entities covered in the project.

Total number of SOEs included

Maybe 10-15 per country, selected so as to be diverse and relevant in terms of turnover? 2/3 central and 1/3 regional/local?

Bibliography (including legal or guiding framework):

  • General, applicable in all 4 countries
  • Country specific

See in the attached Excel file the companies selected tentatively for analysis in Romania. This is a maximal list, we will probably cut the number down during the kick off meeting.

The criteria we used for selecting the companies, which you may also want to consider:

1. Domain diversity - the selection of companies should encompass mostrelevant domains of activities e.g transport, energy, industry
2. Central or local SOEs
3. Listed or non listed SOEs (on stock exchange)
4. Ownership - Fully or partial owned by the state; we will establishcertain intervals for the state(or municipality) shares
5. Risk Profile / Corruption indicators - previous frauds, case law,fraudulent public procurement track record, political appointed board ofmanagers/presidents (Sources; Audits/ Court of Auditors' reports,investigations/courts decision, mass media investigations)
6. Companies with commercial/public service parts.Need to separate inour analysis the commercial and the public service parts

1