DOES GOVERNANCE LEAD TOPERFORMANCE?

What we learned about SOEs from Bulgaria, Czech Republic, Italy and Romania

Otilia Nuţu, Sorin IoniţăBerlin, May 2018

www.expertforum.ro

www.statecapture.euOverall report, 4 country cases

“All happy families are alike; each unhappy family is unhappy in its own way.” Lev Tolstoy, Anna Karenina

SOEs are a very diverse and unhappy family

1. Defining & measuring “performance” for SOEs

Are SOEs really comparable to private companies?

Fiscal Council Report EC report

Are SOEs really comparable to private companies?

What are SOEs and why do they exist?1. Natural monopolies (ex infrastructure): roads, water-

sewage, energy mains, etc2. Imperfect contracts: high externalities, unpredictable

trends, effective regulation would be too complex / costly (forests, other nat assets, airports, hydro)

3. Incumbent operators (ex post): burdensome social service obligation, costly regulation

4. Industrial policy (the most fuzzy): Various social goals like employment, regional cohesion, “strategic sector” (?): promote R & D; “infant industry”; in ECE, legacies of inherited socialist industry (Oltchim), other reasons (Italy)

What are SOEs and why do they exist?

SGEIs SOEs

• Different from purely commercial companies• Not exactly entities providing only public goods and services

- SGEI• Continuum between SGEI provision and purely commercial

Several ways around this, but limitations• Sectorial comparisons (IMF, EC) – each sector same SGEIs

Financial Output

• Benchmark before/after privatization

What we found: options• Define SGEI/PSOs clearly in the legislation, finance it

separately from public budgets, avoid cross-subsidies between commercial activities vs for-profit You can then simply benchmark profitability of commercial part with private & commercial companies and monitor cost to budget/output for SGEI part

• Define a set of indicators that are relevant for “benchmarking” considering the SGEI vs commercial activity limitation for each company / sector

Our research approach

1. Researching SOEs: performance (vs governance)

What we did: 1. Sample SOEs – difficulties RO IT BG CZ

CENTRAL / REGIONALRoads CNADNR ANAS MRDPW RSDRailways CFR Calatori Trenitalia BDZ (passenger) CD (passenger)Electricity HE, CEO, ELCEN IREN NEK CEZGas Romgaz IREN BulgargazPorts Constanta VarnaAirports OTP Sofia APForestry Romsilva SW SOE - Blagoevgrad Lesy CeskePost Posta Romana BG Post Ceska Posta

LOCALUrban transport RATB Metro Milanese Sofia Metropolitan DPPDH RADET IREN TOPLOWater Apa Vital AMIU Genova, AQUE Torino Russe waterWaste AMIU Genova

What we did: 2. Define clientelism in SOEs

In theory, corruption / clientelism / capture in SOEs have three main avenues:i. Buy goods and products for own use at above market

prices (fraudulent procurement & equivalents)ii. Offer goods, services or assets preferentially, at below

true market prices, to various individuals/entities (reverse of A)

iii. Nepotism / politicization in hiring, firing, management

What we did: 3. “Grade” the governance

RO IT BG CZCENTRAL / REGIONALRoads CNADNR ANAS MRDPW RSDRailways CFR Calatori Trenitalia BDZ (passenger) CD (passenger)Electricity HE, CEO, ELCEN IREN NEK CEZGas Romgaz IREN BulgargazPorts Constanta VarnaAirports OTP Sofia APForestry Romsilva SW SOE - Blagoevgrad Lesy CeskePost Posta Romana BG Post Ceska Posta

LOCALUrban transport RATB Metro Milanese Sofia Metropolitan DPPDH RADET IREN TOPLOWater Apa Vital AMIU Genova, AQUE Torino Russe waterWaste AMIU Genova

What we did: 4. Place SOEs on SGEI/commercial continuum

Two broad categories:A. SOEs purely profit based, with 100% commercial activity

but which happen to be state owned (for various reasons, but mainly in category 4): freight railways, coal mines, airlines, industrial plants

B. SOEs which operate more like the public administration, get budget money, have taxes assigned, legal monopoly: roads companies, state TV/radio, “research institutes”• They are not paid by clients, but mostly by taxpayers

What we did: 4. Place SOEs on SGEI/commercial continuum

Budget pays (general taxation)Public agency

Performance: delivery @ reasonable cost Clear separation public goods / profit

CNADNR, ANAS RAI Uno Posta Romana Hidroelectrica HPPMRDPW, RSD TVR public television TOPLO Ceska Posta Romsilva forest management

CFR Infrastructure RADET DH BG Post SW SOE - Blagoevgrad Romgaz, Bulgargaz supIREN DH Lesy Ceske

CFR Calatori Port Constanta, Burgas TAROM airlinesTrenitalia Aeroport Otopeni, Sofia Airport IREN gasBDZ Passenger Russe Water IREN electricityCD Passenger AMIU Genova NEK CEZRATB, Metro M CFR Marfa Apa Vital, IREN water CE Oltenia electricitySofia Metro, DPP AQUE Torino

Transelectrica (regulated), Bulgargaz pipes ELCEN

Public services / goods Commercial services / goods

Consumers pay (price)For-profit company

SOEs

Performance: profit maximization

What we did: 5. Define indicators comparable across companies placed on same spot

Identified areas of risk (procurement – of core vs non-core activity items; sales; output)Energy sector / mining: • Percentage of overheads: Overhead/revenues; overhead/installed capacity; overhead/distribution assets (both in $ and in km

of wires and number of transforming stations); overhead/number of consumers; overhead/sales to consumers (as opposed to wholesale market)

• Cost of coal / oil / gas per MWh (procurement)• Price of electricity EUR/MWh in direct contracts vs similar contract on power exchange (underestimated sales)• output (GWh) or turnover (EUR) per employee; arrears as share of turnover, correlation of arrears with electoral cycles• overdue receivables / turnoverTransportation – railways:• Percentage of overheads: Overhead/revenues; overhead/number of clients; overhead/sales to clients• output (person-km, tons) / employee; revenue (EUR) / output (traffic, tons); profitability compared to private counterparts.• procurement: cost per car /locomotive, SOE vs private.• miscellaneous (non-core activity related) procurement / total procurement (highlights if company procures useless supplies)• Freight transport - sales: price per ton of major commodity (e.g. coal), SOE vs private freight transporters• Overdue receivables / turnover, SOE vs private transporter (e.g., SOE allows clients to default on payments)• turnover / employees

What we did: 5. Define indicators comparable across companies placed on the same spotWater-sewage operator• Price / km of newly build sewer main, Price / km of newly build water main• Price per unit of service (usually m3)• overhead/number of consumers; overhead/sales to consumers • miscellaneous (non-core activity related) procurement / total procurement• turnover / employeesLocal public transportation• Price for client served• Number of employees (other than direct operators) / number of clients• Price of insurance per vehicle• overhead/number of clients; overhead/sales to clients• turnover / employee• OPEX / turnover• revenue (tickets + subsidies) / clients• procurement: cost per vehicle (bus; tram; trolleybus)• miscellaneous (non-core activity related) procurement / total procurement

What we did: 5. Define indicators comparable across companies placed on the same spotConstruction contracts / infrastructure (for public authority-like SOEs)• Cost per km of newly built motorway / standard national road• Cost per km of yearly maintenance of motorway / standard national road• Profitability of commercial partners (contractors), national vs foreign; national vs EU funds• miscellaneous (non-core activity related) procurement / total procurement• turnover / employeePostal services:• arrears / turnover; correlation of arrears with electoral cycles• turnover / employee• miscellaneous (non-core activity related) procurement / total procurement• profitability benchmarked with private postal services?Forestry:• arrears / turnover; correlation of arrears with electoral cycles• turnover / employee• miscellaneous (non-core activity related) procurement / total procurement• area of forest managed by SOE/ total country surface (evolution in time)

What we did: 6. Limit – 10 key indicators• Criteria: available data, comparability, relevance• Purpose: measure variations of performance

• Across time (e.g. 2010-2015) vs across countries (Ro-Bg-It-Cz) • Variations: search deeper for explanations (illustrated on Romania)• Correlate with national governance quality in SOEs• List of indicators:

• Roads: Total investment cost / network km• Railways - passenger: cost/passenger*km (total operational costs/passenger*km)• Energy, gas: profitability benchmarked with similarly structured companies and

GWh/employee• Forestry: profitability vs preservation of forest• Ports: traffic per employee (airports – passengers; ports – cargo)• Postal services: benchmarking of profitability (considering all similar SIEG)• Urban transport: cost/passenger-km; water: cost/m3• DH: arrears or subsidies/GCal

Roads: Total road investment cost / network km

OECD data, own calculations. Total investment includes upgrades, new constructions. RO: partly backlog of investments, partly over-expensive procurement. Note: Competition Council highlights motorway costs in Ro = 3x BG

RO: hike in spending in 2009 – crisis year and 2011 (Transalpinacase). Reduction of costs: reduced investment spending, procurement on EU rules

In-depth Roads analysis: CNADNR Romania2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

General frameworkCNADNR governance

ProcurementProfitability partners

Foreign cosRomanian vs foreign less profitablee.g. 2011 (tenders)

In-depth Roads analysis: CNADNR Romania2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

General frameworkCNADNR governance

Romstrade One-off spike:Transalpinasince 2015insolvent

Strabag EU-fundedprojectstake off in 2012

Romstrade Strabag normalvs Strabag rates ofprofitability profitability

-250-200-150-100-50050100150200

0

500

1000

1500

2000

2500

2008 2009 2010 2011 2012 2013 2014

Millio

ns

Millio

ns

Turnover (left axis) Profit (right axis)

-50

0

50

100

150

200

0

500

1000

1500

2000

2500

2009 2010 2011 2012 2013 2014 2015 2016

Millio

ns

Millio

ns

Turnover (left axis) Profit (right axis)

-80-60-40-20

020406080

2009 2010 2011 2012 2013 2014

Romstrade Strabag

Energy: GWh/employee; profitability

To compare exactly the same figures, we used electricity generation vs employment for electricity generation, as some companies – NEK, CEZ, IREN are integrated companies / groups comprising many activities. This is why we cannot benchmark profitability between them directly. Cogeneration is considered jointly (electricity + heat total production in GWh)Note: the better governed, the better the performance. IREN and CEZ are stock-listed; NEK and CEO are notoriously inefficient; Hidroelectrica is a case detailed below.

In-depth Analysis: Hidroelectrica Romania2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

General frameworkHidroE governance

Below market price sales No transactionsoutside competitive market

Foregone profits No foregone profits

Profitability of traders Regular traders fare better than "smart guys"

Smart guys contracts - e.g. 2011, mn RON

Smart guys vs regular traders - e.g. 2011, RON Ex- smart guys vs regular traders, e.g. 2015, RON

No smart guys, e.g. 2016 (except attempt 2013)

2304

1452

852

0%

20%

40%

60%

80%

100%

Value Foregone profits

0

10

20

30

40

Turnover/employee

Mill

ions

Alpiq RI Repower

-

200

400

600

Profit / employee

Thou

sand

s

Alpiq RI Repower

16

17

17

Turnover/employeeM

illio

ns

Alpiq RI Repower

(60)

(40)

(20)

-

20

Profit / employee

Thou

sand

s

Alpiq RI Repower

In-depth Analysis: Hidroelectrica Romania2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

General frameworkHidroE governance

Overvalued inputsGWh/employee Slightly better performance

Still large gapDifficulties to fire excessstaff

Third-party maintenanceand works (EUR/MWh)

High improvementoverachievementpossible underfinancing

Overall profits

HidroE becameprofitable

0

10

20

30

40

50

Hidroelectrica AES Tiete CESP Verbund0

10

20

30

40

50

Hidroelectrica AES Tiete CESP Verbund

0

1000

2000

3000

4000

5000

Hidroelectrica Verbund0

200400600800

1000120014001600

Hidroelectrica Verbund

ConclusionsThe balance between the two sets of goals (profit and social obligations) must be:● Clearly defined in policy and measurable● Written into the contracts● SOEs must be compensated by the state

(central/local) for the public service obligations performed

In general (at least ECE) none of these is true – which creates big problems for defining clientelism

Conclusions• Very few “benchmarkable” indicators, as companies are

very different. However, general indicators highlight general issues, research can be refined to explain variations (“case studies”)

• Needed: DATA TRANSPARENCY+ full implementation of state aid, proper SGEI definition and financing• Note: bad governance, bad performance do not

necessarily mean corruption; but acting on governance and performance limits corruption

ConclusionsBroader definition of clientelism = bad governance in general, perpetuated through mechanisms (i)+(ii)+(iii)

Practical dilemma: In the case of “legacy SOEs” and when there is no economic viability (ex Oltchim) is the very existence of the entity an act of clientelism?

What is the difference between legitimate “industrial policy” (subsidization / support) and outright rent-seeking by particular groups / enterprises? (ex. cheap gas in RO?)