Government favoritism in public procurement: evidence from ...1430016/...Daniel Pustan Godkänt 2019-12-31 Examinator Matti Kaulio Handledare Kristina Nyström Uppdragsgivare Kontaktperson

DEGREE PROJECT IN TECHNOLOGY AND ECONOMICS, SECOND CYCLE, 30 CREDITS

STOCKHOLM, SWEDEN 2019

Government favoritism in

public procurement:

evidence from Romania

DANIEL PUSTAN

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

This page intentionally left blank

Government favoritism in public procurement: evidence from Romania

by

Daniel Pustan

Master of Science Thesis TRITA-ITM-EX 2019:728

KTH Industrial Engineering and Management

Industrial Management

SE-100 44 STOCKHOLM

Regeringens favorisering inom offentlig upphandling: fallet Rumänien

Daniel Pustan

Examensarbete TRITA-ITM-EX 2019:728

KTH Industriell teknik och management

Industriell ekonomi och organisation

SE-100 44 STOCKHOLM

Master of Science Thesis TRITA-ITM-EX 2019:728

Government favoritism in public procurement: evidence from Romania

Daniel Pustan

Approved

2019-12-31

Examiner

Matti Kaulio

Supervisor

Kristina Nyström

Commissioner

Contact person

Abstract

In Romania, the consideration that politicians use their influence to control the public

procurement market is axiomatic. It is no surprise that the country ranks high in perception-

based surveys or the low participation of firms on the procurement market. The more difficult

task is to demonstrate the existence of restrictions to procurement contracts in order to benefit

preferred companies. That is, to measure the extent to which the market is captured by favored

companies.

Employing data on all public procurement contracts in Romania for the period 2009 – 2015,

this paper examines government favoritism in public procurement exerted by political parties.

Using a dynamic panel data approach (Dávid-Barrett and Fazekas 2019), the companies are

classified based on their winning pattern with respect to government change. Favoritism is

observed if winning companies within the government period are also associated with a higher

risk of corruption measured by two alternative approaches.

The findings confirm that procurement market is captured in a low to moderate proportion

(24%) and that the market display patterns of systematic favoritism. This may signal certain

progress registered by Romania to combat political corruption. Arguably, the insensitivity of

perception indicators with respect to this progress is, at least partly, due to media coverage of

the on-going corruption investigations related to the past.

Key-words

Grand corruption, High-level corruption, Political corruption, Clientelism, Favoritism,

Governance, Corruption measurement, Public procurement, Romania

Examensarbete TRITA-ITM-EX 2019:728

Regeringens favorisering inom offentlig upphandling: fallet Rumänien

Daniel Pustan

Godkänt

2019-12-31

Examinator

Matti Kaulio

Handledare

Kristina Nyström

Uppdragsgivare

Kontaktperson

Sammanfattning

I Rumänien, finns det en allmän uppfattning om att politiker använder sitt inflytande för att

kontrollera den offentliga upphandlingsmarknaden. Det är ingen överraskning att landet rankas

högt i perceptionsbaserade undersökningar rörande korruption eller att företags deltagande

inom upphandlingsmarknaden är lågt. Ett svårare uppdrag är dock att påvisa och bevisa

förekomsten av begränsningar kring upphandlingskontrakt med syfte att gynna vissa företag.

Med andra ord så föreligger en utmaning att, genom mätning, påvisa i vilken utsträckning

marknaden inkluderar favoriserade företag kontra hur den exkluderar övriga företag.

Med hjälp av uppgifter om samtliga kontrakt gällande offentlig upphandling i Rumänien under

åren 2009 – 2015, undersöker denna avhandling regeringens och dess politiska partiers

favorisering. Företagen i upphandlingarna klassificeras med hjälp av dynamiskpaneldata

(Dávid-Barrett och Fazekas 2019), baserat på des vinnande mönster kopplat till regeringsbyte.

Favorisering kan observeras om vinnande företag inom regeringsperioden även är förknippade

med en högre risk för korruption som mätts genom två alternativa metoder.

Resultaten bekräftar att upphandlingsmarknaden fångas i en låg till måttlig andel (24%) av

favoriserade företag och att marknaden visar mönster av systematisk favorisering. Resultaten

kan dock signalera om visst framsteg som Rumänien har uppnått för att bekämpa politisk

korruption. Det kan argumenteras att perceptionsbaserade indikatorer fångade inte upp dessa

framsteg, åtminstone delvis, på grund av mediatäckningen av pågående korruptionsutredningar

i Rumänien relaterade till det förflutna.

Nyckelord

Stora korruption, korruption på hög nivå, politisk korruption, klientelism, favorisering,

styrning, korruptionsmätning, offentlig upphandling, Rumänien

Government favoritism in public procurement: evidence from Romania 7

Table of Contents

1. Introduction 13

2. Literature review 18

2.1 Definition and classification 18

2.2 Measures of corruption 23

3. Institutional framework 27

3.1 Legislative setting 28

3.2 Public procurement institutions 30

3.3 Government characteristics 31

4. Method 34

5. Data 41

6. Results 44

7. Robustness analysis 54

8. Conclusions 58

References 61

Annexes 70

KTH ROYAL INSTITUTE OF TECHNOLOGY

8 Daniel Pustan

List of figures

Figure 1 The bi-dimensional clientelistic structure

Figure 2 The comprehensiveness of regulation in Romania

Figure 3 Timeline of cabinets and parties for the period 2009 - 2015

Figure 4 Mean regression residuals by two-percentiles of relative procedure period

(contract date - call for tender date), linear prediction

Figure 5 Mean regression residuals by two-percentiles of relative notification of award

period (award notice date - contract date), linear prediction

Figure 7 CRI-arithmetic by company group, half-year period

Figure 8 Combined market shares of company types, half-year period

Figure 9 CRI arithmetic compared to perception-based surveys (CPI, WGI, GCI)

Figure 10 CRI-PCA compared to perception-based surveys (CPI, WGI, GCI)

Figure 11 Single bid compared to perception-based surveys (CPI, WGI, GCI)

Figure 12 Outliers for relative contract value < 2

Figure 13 CRI-PCA patterns by company group, half-year period

Figure 14 CRI-arithmetic patterns by company group, one year period

Figure 15 CRI-PCA patterns by company group, one year period

Figure 16 CRI-arithmetic patterns by company group, changes in legislation

Figure 17 CRI-PCA patterns by company group, changes in legislation



List of tables

Table 1 Amendments to GEO 34/2006 during the studied period

Table 2 Romania’s government cabinets during 2009-2015

Table 3 Main characteristics of the dataset

Table 4 System GMM linear dynamic panel regression explaining company market

success

Table 5 Logit regression results on contract level, 2009-2015, number of winners 2≥

Table 6 CRI component weights

Table 7 Bivariate Pearson correlation of CRI and single bid with perception survey

indicators

Table 8 Linear regression explaining relative contract value and CRI

Table 9 Summary of selected studies using objective indicators of corruption

Table 10 Corruption techniques and indicators

Table 11 Summary of selected studies on firm’s political connections

Table 12 Descriptive statistics


10 Daniel Pustan

Abbreviations

ANAP National National Public Procurement Agency

ANRMAP National Authority for Regulating and Monitoring Public Procurement

BI Business International

CIN Company Identification Number

CNSC National Council for Solving Complaints

CPI Corruption Perception Index

CPV Common Procurement Vocabulary

CRI Corruption Risk Index

CRI Corruption Risk Index

CVM Cooperation and Verification Mechanism

DLAF Department for the Fight Against Fraud

DNA National Anticorruption Directorate

FDI Foreign Direct Investments

GCI World Economic Forum

GEO Government Emergency Ordinance

ICRG The International Country Risk Guide

IMD The Institute for Management Development

NAPP National Agency for Public Procurement

NGO Non governmental organization

NUTS Nomenclature of Territorial Units for Statistics

PC Conservative Party

PCA Principal Component Analysis

PDL Democrat Liberal Party

PNL National Liberal Party

PSD Social Democratic Party

SEAP Electronic System for Public Procurement

TED Tenders Electronic Daily

UCVAP Unit for Coordination and Verification of Public Procurement

UDMR Democratic Alliance of Hungarians in Romania

UNPR National Union for the Progress of Romania

USL Social Liberal Union

WCY World Competitiveness Yearbook

WEF World Economic Forum

WGI Global Competitiveness Index



Acknowledgements

On the very outset of this paper, I would like to extend my sincere and heartfelt obligation

towards all the personages who have helped me in this endeavor. Without their active guidance,

help, cooperation and encouragement, I would not have made headway in the project.

I would like to express my gratitude to my supervisor, Kristina Nyström for her patience and

valuable advice for structuring the ideas of this paper. I am extremely thankful to Mihály

Fazekas - co-author of the paper on which the methodology of this study relied on - for the kind

responds to my questions.

I extend my gratitude to Sebastian Buhai for the precious guidance related to the topic of this

research and to Hayk Sargsyan for his useful comments and encouragement when most needed.

I am deeply indebted to my former manager, Larisa Lăcătuș, for creating the conditions to

expand my knowledge alongside my professional development. On the same note, I gratefully

acknowledge the assistance of Karin Asplund with the Swedish translation.

Lastly but not least, my highest appreciation goes to my wife, Maria, who experienced firsthand

this long and strenuous process. Your limitless understanding and unconditional support made

it possible for this journey to end successfully.

Any omission in this brief acknowledgment does not mean lack of gratitude.

Stockholm, December 31st, 2019

Daniel Pustan


12 Daniel Pustan

1. Introduction

Corruption is estimated to cost the European Union economy €120 billion per year which is

almost the equivalent of the EU annual budget (European Commission 2014). The true cost of

corruption, however, includes arguably inestimable social costs such as: poorer income

distribution, environmental protection as well as undermining trust in democratic institutions

(European Commission n.d.).

Thirty years after communism fell in Romania, it seems that the fight against corruption

remains an ongoing battle which affects the country on many levels. In terms of country

perception, Romania ranks amongst the most corrupt countries in the European Union (Hall

2019) while favoritism appears to be the rule for the majority of government decisions (Schwab

2016). Additionally, Romania is held back from joining the free movement Schengen area

because of the lack of progress in combating corruption as reflected by the Cooperation and

Verification Mechanism (CVM) . 1

The civil society became harshly conscient of the fatal consequences of corruption following

Colectiv nightclub fire in November 2015. A fire that tragically ended the lives of many, only

because, allegedly, corrupt practices led to systematic violation of the rules which, if respected,

could have saved lives. Protests that followed denouncing political elites’ tolerance or

association to corruption led to Victor Ponta’s resignation, the first Prime Minister accused of

corruption while still in charge (BBC 2015).

Considering the external and internal pressure - by means of European Commission reports and

civil protests - everything indicated that the message was understood and corruption would be

irremediably defeated in a battle with the government as protagonist.

Nothing could be further from reality. Liviu Dragnea, chief of the ruling Social Democratic Party

(PSD) and president of the Chamber of Deputies in Parliament who already had a suspended

sentence, was risking prison sentence in his ongoing corruption trial. (France24 2018). In an

attempt to spare the PSD chief and other high profile politicians from the ongoing investigations

or prison sentences, the Romanian Government issued an Emergency Ordinance 13/2017 (to

enter into effect 10 days after publication in the Official Gazette).

1 The Cooperation and Verification Mechanism is the mechanism that European Commission

uses to monitor reform implementation in the area of justice and rule of law.



Massive protests followed against what was perceived as legalization of corruption acts inducing

the government to abolish the controversial Act prior to its entry into force. Should the

Emergency Ordinance came into effect just for one day, Liviu Dragnea could have escaped the

three and a half years of prison that he was sentenced to in May 2019 (BBC 2019).

Unfortunately, political corruption does not represent just a few isolated cases. Only in 2017,

other 12 high level officials were prosecuted among which: ministers, deputies, Secretary of

State, Senators and government Secretary-General National (Anticorruption Directorate Activity

Report 2017).

In this context, public procurement is especially vulnerable to corruption (European

Commission 2014; OECD 2016). Perception based reports advance that public procurement is

manipulated in proportion of 80% (Euractiv 2014). This severe trust deficit is probably one of

the main causes of the low share of firms taking part in the bidding process. In fact, the number

of unique bidding companies during the seven years period of observation was 15,568 (1.3%) out

of a 1.2 million active companies (Oprea 2017). By comparison, in Sweden the number of

companies participating in public procurement bids during 2013 was 212,000 representing

18.80% of the total number of registered companies (Lukkarinen, Morild and Jönsson 2015).

Furthermore, the low number of referrals received from public administrations (9% in 2017) as

well as the rare occasions in which local authorities decide not to become claimants in order to

recover incurred losses (National Anticorruption Directorate 2017), might be an indicator of

participation in a larger network of favoritism or the resignation that such networks exist and

there is not much to do against them.

The overflow of the anti-corruption theme outside the institutional framework into society has

caused deep rifts among many walks of life. In these last years, many have seen in the Social

Democrat Party (PSD) the very essence of corruption, a party that was ripping the resources of

the state shamelessly. At the same time, for others it represented the ultimate guardian for the

ordinary citizen ripped off by multinationals in their run to make profits at any cost.

Perceptions became less reliable while sliding towards a binary, oversimplified division of reality

- between “us” and “they” (Bogdan 2018). Polarization deepened and fact based arguments lost


14 Daniel Pustan

prominence in an ideological battle, arguably, between traditional and modern values (David

2015).

The necessity to regain a common ground between the polarized sides became obvious. This, in

turn, required a measurement tool that focuses on objective behavior rather than opinions and

perceptions.

For a long time, measurement of corruption relied mostly on perception based indicators.

However, perception may not be of much help in this scenario. Furthermore, the little variation

of indicators makes them susceptible of not being responsive to change, in spite of significant

changes in Romania from one year to another. (Oman and Arndt 2006).

Companies associated with high profile politicians are often successful on the procurement

market, earning the label of “companies subscribed to contracts with the State'' (Befu 2015;

Marcovici 2015; Andrei 2018). In fact, Mungiu-Pippidi (2015) highlights that political

connections are the main indicator of favoritism in Eastern Europe. But do high level politicians

use their influential capacity to divert contracts towards favored companies?

According to Gherghina and Volintiru (2017) clientelistic model, parties are interested to engage

in reciprocal relationships by redirecting public contracts towards favored private contractors in

exchange for benefits - which in most cases are monetary. Parties will use this financing to

sustain or increase the electoral support and consequently consolidate their power.

Figure 1 The bi-dimensional clientelistic structure

Source: Gherghina and Volintiru 2017, p. 122



Therefore, if the answer is affirmative as anticipated, what size of the procurement market is

captured by the political elites?

To my knowledge, the only research on Romania’s procurement market with objective indicators

were performed by Doroftei (2016) and Doroftei and Dimulescu (2015). They examined

contracts over €1,000,000 in the construction sector for the period between 2007 and 2013.

Both papers conclude that single bidding and political connections explained 32% of the number

of contracts awarded during the entire research period and 44% in pre-election years.

Using a more sophisticated measurement of corruption, Dávid-Barrett and Fazekas (2019)

found that 50%-60% of procurement market in Hungary suffer from systematic favoritism while

only 10% of companies were politically favored in the United Kingdom. Their approach

consisted of two steps: (1) analyze the influence that government change has on companies

procurement outcome and (2) investigate the risk of corruption associated with the winning of

contracts. The sample comprised period 2009-2012 and was limited to contracts: (i) awarded by

national authorities, (ii) above the mandatory EU-threshold and (iii) only applied to competitive

markets.

Following Dávid-Barrett and Fazekas (2019) method, this paper explores empirically to what

extent the governmental political parties in Romania exert control over the public procurement

process; in other words, it quantifies the level of government favoritism in Romania. The

analysis extends to all public procurement contracts in Romania during 2009-2015 period. The

sample was not restricted to contracts awarded only by national central administration. This

decision was taken considering the centralization of power in Romania both at the level of public

administration as well as that of party organization.

The contribution to the research literature is fourfold. First, to my knowledge it is the first time

to apply Dávid-Barrett and Fazekas (2019) innovative methodology on public procurement in

Romania. Secondly, it expands the method by performing a cross-governmental analysis with a

focus on the main party in government. Thirdly, this study introduces the Principal Component

Analysis as an alternative to simple average method for weighting the indicators composing the

Corruption Risk Index. Lastly, it provides a corrected large dataset suitable for scientific

research.


16 Daniel Pustan

The method follows three steps. First, companies are classified based on the residuals after

estimating the panel regression using Arellano-Bond system GMM transformation. Second,

Corruption Risk Index is calculated both as: (1) as a weighted average (Dávid-Barrett and

Fazekas 2019) and (2) using the Principal Component Analysis approach. Thirdly, the market

share of aggregated suspicious companies is calculated and companies are aggregated by group

to identify patterns of systematic favoritism.

This paper analyzes a dataset of over 650,000 observations and finds that 24% of the Romanian

procurement market is associated with systematic government favoritism. However, it is

noteworthy, that the study only captures specific clientele for each of the two parties. It does not

reflect favoritism in situations where companies buy influence from both parties or public

administration officials that withstand government change.

Furthermore, highly innovative companies may be mislabeled as “surprise winners” as both

exhibit similar pattern behavior (Dávid-Barrett and Fazekas 2019). This risk is reduced for this

specific dataset by the presence of “lowest price” award criterion in 94% of the procedures

during the analyzed period. Typically, due to the high costs of research, innovative companies do

not compete for price and they are not expected to take part in these tenders. Two other

limitations are discussed further.

The Common Procurement Vocabulary (CPV) codes are the basis for the CPV division (2 digit

level of CPV codes) and CPV product group (3 digit level of CPV codes) which in turn serve as

proxies for the market. Although generally accepted as a market indicator, it is not a perfect

instrument due to flaws in the CPV code structure such as: (1) inconsistent hierarchical

tree-structure, (2) inconclusive classification levels, (3) unmatching between subordinate and

superordinate level and (4) not mutually exclusive codes (Attström et al. 2012).

Aggregating the data based on a half year period “is optimal for retaining a high level of

granularity while also taking into account the erratic character of many public procurement

markets” (Dávid-Barrett and Fazekas 2019, p. 13). However, the random length of the cabinets’

period does not perfectly match the half year aggregation. To circumvent this problem, an

alternative model was analyzed where the cabinet variable was replaced by variables consisting

in legislation changes. Both models yield comparable results.



In view of the ensuing inefficient distribution and allocation of resources, corruption and in

particular clientelism constitutes, perhaps, the antithesis of sustainable development.

Clientelism skews the distribution of state resources (Ferreira Rubio and Andvig 2019) with

potential negative consequences in all three aspects: environmental, social and economic.

Among the social consequences are the increase of inequality (Ferreira Rubio and Andvig 2019)

and erosion of confidence in institutions. Corruption affects the most vulnerable groups as

infrastructure projects may be vastly financed to steer money towards favorite companies in

detriment of education and health (Mungiu-Pippidi 2015).

Corruption is also negatively correlated to the growth in genuine wealth per capita. This is

clearly illustrated by Aidt (2010) with Denmark. If levels of corruption in Denmark would

suddenly increase to those of the countries that make it to the bottom of Corruption Perception

Index, the development would become unsustainable.

Narrowing it further down, the implications of this paper on sustainability - described as making

decisions in the present so as not to impair future real incomes (Markandya and Pearce 1988) -

are highly practical. The economic impact is given by the importance of public procurement

which rises to almost 50% of public spending in developing countries. Within the social aspect,

it represents an invitation for society to focus on facts in an attempt to find a common ground

where this was lost. It promotes the equality and importance of individuals in the social system

while promoting the ranking of ideas. The positive impact on the environment stems from the

implicit responsible and efficient use of resources as alternative to particularism.

The remaining of the paper is structured in seven sections. First, the literature is reviewed with

special emphasis on the various definitions of corruption and the challenges encountered in

measuring it. Second, the institutional framework is discussed including the changes in

legislation during the period, the specific institutions related to public procurement and the

composition of government. Following, the method section describes the process of suspicious

companies identification based on their market success and the subsequent analysis of the

conditions under which they as shown by the Corruption Risk Index. The remainder of the paper

is organized as follows: data, results, robustness analysis and conclusions.


18 Daniel Pustan

2. Literature review

2.1 Definition and classification

Corruption is a term used to cover a wide range of different practices. It appears that to some

extent the term is framed by prevalent cultural norms (Werner 1983) as well as depending on

the context (Johnston 1996). The difference is obvious when comparing the more developed

American democracy with the Romanian mindset of government. The American view strives to

limit the authority and divide the power (Huntington 1968). By contrast, the Romanian

dominant psychological profile is that of a collectivist person and concentration of power (David

2015).

In an interview at the Romanian Business Leaders Summit (Roșca 2018), David argues that

favoritism and nepotism are embedded in the Romanian culture and what is considered

corruption from a Western perspective, it is simply perceived as a proof of loyalty. David

continues by exemplifying with the implementation of a Western institution in Romania - that of

a parliamentary cabinet - a parliamentary with a collectivist profile will hire a person who they

can trust most - a relative - instead of the best expert (Roșca 2018).

This shows that conflict of interests as a specific form of corruption, may also be the

unintentional result of institution implementation within a culture different than the one where

the institution was born. However, nepotism as a cultural accepted practice is not specific only

to Eastern Europe. In India and Africa it is not only acceptable but expected (Leys 1965; Bayley

1966).

Huntington (1968) goes as far as to suggest that corruption is a hardly avoidable externality of

modernization [of institutions]. Scholars supporting the “grease the wheels” hypothesis

highlight potential malfunctions or inefficiencies that corruption may help circumvent the

functioning of the institutions and government policies. It can (1) increase the quality of the

service (Leys 1965; Bayley 1966) - by keeping the underpaid talented employees in the public

service, (2) speed-up processes like waiting in the queue (Huntington 1968; Lui 1985) and (3)

foster efficiency as an imperfect auction system in competitions for a limited number of

resources or instead of non optimal mechanical quota based systems (Leff 1964; Bayley 1966),

(4) act as an insurance against bad public policies ensuring predictability and give a sense of

being in control during transition period (Leff 1964; Bayley 1966), (5) non-violent manner to

access the establishment and (6) provide political representation or unity in spite of



irreconcilable ideologies or interests, potentially higher and better investment (Leff 1964; Bayley

1966; Werner 1983).

Most empirical research, however, does not support the claim of positive effects of corruption.

Dependency on grafts acts as an incentive to deliberately slow down or limit the provision of the

service and to block the access to other competent workers in order to secure a large stream of

cash (Lui 1985; Kurer 1993). Furthermore, in order to compensate for the bribe, the firm can

provide products of lower quality (Rose-Ackerman 1997).

The “sand in the wheels” effect of corruption is more obvious when looking at its impact on

economic growth. Various studies confirmed the negative relationship between corruption and

growth consequence of a decrease in private investment (Mauro 1995) or the quality of the

infrastructure (Tanzi and Davoodi 1998) even in countries with a weak rule of law and inefficient

government (Méon and Sekkat 2005). Mo (2001) estimated that 1% increase in corruption led to

a decrease by 0.72% in private investment.

So, what is corruption? Most often, corruption is associated with bribery in the public eyes.

However, it embraces a large palette of forms (bribery, embezzlement, theft and fraud, graft,

revolving door, clientelism, nepotism, cronyism, conflict of interests, patronage, etc.) targeting

different levels of authority and adapting the appropriate methods of sophistication to context

and culture. It is no wonder that there is no consensus among scholars on the definition of

corruption.

The general accepted definition, promoted by international organizations such as Transparency

International, refer to corruption as “the misuse of entrusted power for private gain”

(Rose-Ackerman 2008, p. 551; Andersson and Heywood 2009, p. 746; Xu 2016, p. 86). As

already anticipated, this definition is generally associated with bribery in a bureaucratic context

(Fazekas, Cingolani and Tóth 2016; Fazekas 2017; Mungiu-Pippidi 2013 acc. Mungiu-Pippidi

2015). Scholars classify this type as petty, bureaucratic or low-level corruption.

Another category of corruption - object of this study - takes place at the higher-levels in public

positions, benefits of vast discretion and lacks explicit detailed rules (Fazekas, Cingolani and

Tóth 2016). This type of corruption, defined as a deviation from the universalism norm (Kurer

2005: Mungiu-Pippidi 2015) is known in the literature as high-level corruption, government

favoritism, institutionalised grand or political corruption, (Jain 2001; Graycar 2015; Fazekas,


20 Daniel Pustan

Cingolani and Tóth 2016; Fazekas, Tóth and King 2016; Fazekas 2017). Thus, government

favoritism in public procurement “refers to the allocation and performance of public contracts

by bending universalistic rules of open and fair access to government contracts in order to

benefit a closed network while denying access to all others.” (Fazekas, Cingolani and Tóth 2016,

p. 3; Fazekas, Tóth and King 2016, p. 370).

Ultimately, high-level corruption may deviate into state capture if it reaches the stage where

institutions or laws are expressly created to benefit a private group. At this point corrupt

decisions are implemented in a corrupt-free manner (Søreide 2002; Fazekas 2017). This stage is

more characteristic to economies in transition from communism to capitalism and to some

extent through lobby groups in developed economies (Graycar 2015).

Among the different types of political corruption, clientelism distinguishes as a special form of

particularism because of its contingent or “quid pro quo” nature. Clientelism is the distribution

of all kinds of public goods in exchange for political support (Hicken 2011). The major risk areas

of capture by political elites are the appointments in the public sector, privatization projects,

subsidy programmes and public procurement contracts.

Hicken (2011, p. 291) highlights that the difference between clientelist practices and other types

of exchange consists in the fact that the former “always comes with strings attached”. These

benefits aim to consolidate or increase political power and are typically in the form of voting.

However, it may come in the most diverse forms such as: donations, favorable media coverage,

revolving door and public procurement contracts. It is in such a scenario where principal-agent

models prove their limitations.

Until recently clientelism was understood as a type of principal-agent relationship. However, the

reciprocal character of clientelism does not allow for a clear distinction between the role of the

principal and that of the agent. The relationship may reverse to the point where politicians end

up holding voters accountable for their votes (Stokes 2005).

Instead, Gherghina and Volintiru (2017) propose a bidimensional model of clientelism. Thus,

political elites use state resources, in particular procurement contracts, to favor or promise

favors to private companies in exchange for support, typically donations. These funds are then

used as a competitive advantage against other parties to attract votes (Gherghina and Volintiru

2017).



Complementing Gherghina and Volintiru (2017) model, Dávid-Barrett and Fazekas (2019)

highlight three stages in which politicians can bend the universalistic norms: formation of the

law, implementation by bureaucracy and monitoring of implementation or evaluation. This

paper focuses on the implementation stage, one that requires a new violation of law with each

transaction (Dávid-Barrett and Fazekas 2019).

One direction of literature sheds light on the important role of political connections on

companies’ performance. Political connections are identified as one or more of the three

indicators: (1) political campaign donations, (2) board seat connections or family ties and (3)

stock holdings by politicians.

With few exceptions, companies benefit greatly of political connections in three main areas: 2

stock value, financing and procurement contracts. Fisman et al. (2012) finds no influence of the

stock value related to Chimney’s personal ties. Furthermore, Jackowicz, Kozłowski and Mielcarz

(2014) study reveals a negative correlation between political connections and firm profitability

which, they suggest, is likely due to the instability of Poland’s politics.

More connected with the subject of this paper, the studies related to public procurement reveal

that political connections have a major impact on contract allocation even in countries with a

solid institutional setting such as Denmark (Amore and Bennedsen 2013) or the United States

(Goldman, Rocholl and So 2013, Luechinger and Moser 2014).

The increase in the value of procurement contracts varies. In the United States, campaign

contribution results in additional procurement contracts of about 2.5 times the value of

donations (Bromberg 2014). The same figure in Brazil is at least 14 times (Boas, Hidalgo and

Richardson 2014) while in the Czech Republic it can reach 100 times (Titl and Geys 2019).

Gürakar and Bircan (2019) shows that closed bids are more expensive than open bids especially

if the winner is a company political connected to the party in power.

Scholars conclude that politically connected firms are on average, less productive and more

likely to win procurement contracts (Zhuravskayay 2012; Brugués F., Brugués J. and Giambra

2018). In particular, donating firms receive more generous contracts or a higher amount of

2 For a detailed list of the studies, please refer to Annex C


22 Daniel Pustan

smaller contracts, flexible completion dates and less bidding competition (Brogaard, Denes and

Duchin 2015; Titl and Geys 2019; Canayaz, Martinez and Ozsoylev 2015) suggesting a potential

clientelistic link between politicians and companies.

In a more targeted study, Hyytinen, Lundberg and Toivanen (2018) identifies favoritism

towards the in-house units of municipalities in public procurement for cleaning services in

Sweden. Nonetheless, they associate favoritism with municipalities’ efforts to support local

employment rather than “outright corruption” (p. 422). This conclusion indicates that

favoritism is viewed from the petty corruption perspective rather than as a form of

particularism.

The only available studies on favoritism with objective indicators on Romania is the paper of

Doroftei and Dimulescu (2015) confirmed by the work of Doroftei (2016). The analysis drew on

public procurement contracts over €1,000,000 on construction sector in Romania between

2007 and 2013. Both studies conclude that single bidding and political connections explain the

number of contracts awarded in proportion of 44% in pre-election years.

It is worth noting that in spite of the clarity of the single bidding indicator, its simplicity makes it

vulnerable to circumvention (Fazekas 2017). One basic technique to adopt is for the connected

firm to agree in advance on the content of the bid with one or two other firms at their

participation in the procurement process (Fazekas, Tóth and King 2016). A corruption case on

the airport runway extension in Cluj-Napoca illustrates such a situation. In this case, the

politically connected business man admitted it was a bid rigging. In fact, to create the

impression of legality, he went so far as to ask one of the participants to dispute the bid. Only

because he knew that the company would not stand any chance as it did not qualify (Bacalu

2018).

Dávid-Barrett and Fazekas (2019) propose a more complex measurement for favoritism. For a

company to be considered favored, it has to have both suspicious winning patterns and higher

risk of corruption as measured by CRI. The CRI is a composite indicator that besides the single

bid includes another 5 input indicators. Their analysis revealed that favored companies control

50%-60% of public procurement market in Hungary, a large number compared to 10% in the

United Kingdom.



2.2 Measures of corruption

Measuring corruption is a challenging task. Until recently, the corruption measures consisted

mainly in surveys/indicators of attitudes, perception and experience of corruption or a mix of

these, namely composite indices. The respondents are mostly experts or firms and to a lesser

extent citizens (Fazekas, Tóth and King 2016). Reviews of institutions, scientific analyses, audits

and investigations of particular cases complement these measures.

Examples of surveys based on expert opinions are the International Country Risk Guide (ICRG)

indices by Political Risk Services and Business International Corporation (BI). The first was

used by Tanzi and Davoodi (1998) while the former was used for the first time by Mauro (1995).

The Global Competitiveness Index assembled by the World Economic Forum (WEF) and the

Institute for Management Development (IMD) in the World Competitiveness Yearbook (WCY)

are constructed based on business executives opinion. One cross-country experience based

survey among citizens is the Global Corruption Barometer constructed by Transparency

International. Transparency International developed the most famous composite index,

Corruption Perception Index followed by the Corruption Index from the World Bank.

In favor of these indicators stands their extensive use and the strong correlation with each other

as well as with other economic and development indicators (Blackburn, Bose and Emranul

Haque 2010). Nonetheless, in spite of their broad use, critics abound with respect to surveys and

perception indicators.

Andersson and Heywood (2009) emphasize the remarkable similarity in survey questions,

respondents’ group sample and type of corruption targeted (most often bribery) These

characteristics may explain in good measure the high degree of correlation between these

indicators.

Further, perception does not imply having directly experienced corruption in any way. It may be

driven by different factors such as media attention (Golden and Picci 2005). It may also be

inconsistent in time due to a later discovery and broadcasting of past corruption acts (Kenny

2009) or unconsciously using previous reports on corruption as benchmark (Golden and Picci

2005).


24 Daniel Pustan

Other researchers add the fact that answers are shaped by factors such as diverse as individual

characteristics, ethnic diversity, engagement in social activities (Olken 2009), the phrasing of

the questions (Xu 2016), cultural differences between the interviewee and interviewer (Bayley

1966), or expats impressions on their host country (de Maria 2008).

Most interviews are focused on executives or opinion leaders which may indicate a

representativeness bias, reflecting the views of a specific group not the one of the general

population (Fazekas, Tóth and King 2016). In addition, they may reflect only certain types of

corruption while omitting others (Svensson 2005).

Most indices are weakly correlated with reality (Weber Abramo 2008). In a study comparing

villager's perception on corruption with a more objective measure in Indonesia, Olken (2009)

finds that villagers were only able to predict a small part (only 8%) of the real level of corruption

in the project. This was due to the fact that they were better at detecting corruption in

marked-up prices then inflated quantities which generated most corruption in the project.

The main drawback with composite indices is the conceptual imprecision. They are defined by

various indices that compose them which in turn can vary in time (Knack 2007). Their

reductionist approach makes them insensitive to variations in sectors or regions and oblivious to

underlying phenomena indicated by outliers (Lancaster and Motinola 2001).

The Corruption Perception Index (CPI) receives particular criticism as having a Western

business-orientation (de Maria 2008, Andersson and Heywood 2009) and that it is more

reliable to reflect corruption levels in developed rather than developing countries (Golden and

Picci 2005). In spite of these disadvantages, the stakes are high for developing countries by

conditioning financial support on the country’s ranking taken as a proxy for the country's reform

progress (Oman and Arndt 2006). Misinterpretation of the CPI can lead these countries into a

“corruption trap” (Andersson and Heywood 2009, p. 747).

The risk of misunderstanding the true measurement increases exponentially with indices where

corruption is only a part of a larger group of measurement. This is the case of the broadly use

International Country Risk Guide (ICRG) which measures “the political risk involved in

corruption” (Lambsdorff 2006, 82) not the level of corruption.



Unfortunately, survey of experiences do not provide a higher degree of reliability. Part of the

firms in the sample group may be on the market as a consequence of corruption and may have

no interest to report or even misreport (Svensson 2005; Kurtz and Schrank 2007). Even when

they do report, they mostly reflect petty corruption as citizens do not come in contact often with

high-level corruption (Kenny 2009; Fazekas, Tóth and King 2016).

The reviews of institutions lack measurement precision and are based upon implicit

understanding regarding the functioning of the observed institutions (Fazekas, Tóth and King

2016; Fazekas and Kocsis 2017). The precision of the measurement increases with scientific

analyses, audits and investigation of individual cases. However, their resource-intensive

character makes them available in a limited number and even less useful for comparison

because of their specific goals and risk of state capture (Fazekas, Tóth and King 2016; Fazekas

and Kocsis 2017).

The risk of state capture is circumvented by Escresa and Picci’s (2017) more objective index of

corruption cases based on the Foreign Corrupt Practices Act. Nevertheless, Stephenson (2014)

warns about more or less implicit assumptions that may compromise the validity of this index,

such as: (1) conflict of interest (state ownership in the economy of the targeted country), (2)

biased indicator because of the mix of exports and Foreign Direct Investments (FDI), (3) the

number of investigations in a country being influenced by the USA politics (e.g. high risk

country are investigated more often) and (4) the endogenous character of exports.

Indeed, in the latest years researchers drawn their attention towards the so-called “objective”

indicators of corruption capable to measure behavior instead of relying on perceptions or

experiences. One direction of research analyzed the differences in prices for standardized

products or services (Di Tella and Schargrodsky 2003; Bandiera, Prat and Valletti 2009;

Hyytinen, Lundberg and Toivanen 2018). Other researchers contrasted official reports with

surveys and independent audits to highlight corruption in the award of benefits, grants and

projects in infrastructure (Reinikka and Svensson 2004; Olken 2006; Olken 2007). A different

measure was suggested by Golden and Picci (2005) which consisted in the ratio between the

physical infrastructure and the corresponding cumulative expenditures in 20 regions of the

country. Finally, as we mentioned earlier, using proxies such as donations or connected board of

directors, political connections constitute in a good measure the focus of recent research.


26 Daniel Pustan

Kaufmann, Kraay and Mastruzzi (2011) warns with regards to the two problems that stem from

any kind of data when measuring corruption. First, the “noise” in the measurement, which

represents the difficulty or impossibility to distinguish between corruption, incompetence,

ignorance or other source of noise. This is exemplified in Golden and Picci (2005, p. 37)

statement when describing corruption: “... money is being lost to fraud, embezzlement, waste,

and mismanagement; in other words, corruption is greater.” Second, the imperfect relationship

between specific measures of corruption and overall corruption. There is a risk of generalizing

corruption based on a certain measured type of corruption.

In particular, these indicators reliability is shadowed by their dependency on the context, high

costs to replicate and the use of a single proxy for corruption which only in the most optimistic if

not fanciful scenario coincides with the main driver of corruption (Fazekas, Tóth and King

2016).

To address these shortcomings, Fazekas, Tóth and King (2016) created the Corruption Risk

Index (CRI). This composite indicator of corruption in public procurement is constructed as the

weighted average of various indicators of corruption. The indicators, also called red flags, are

identified through the review of the literature and qualitative interviews. Then, they are

validated through regression analysis (only indicators which prove to be powerful and

significant become part of the index). Further, they are divided between corruption outcomes

such as “single bid” and corruption inputs such as “the procedure type” or “the weight of non

price evaluation criteria”.

The main benefits of this index are not only that it derives from objective data consisting of

more than a single indicator but it is also comparable across time and countries. Furthermore, it

follows directly from the definition of high-level corruption in public procurement, namely, the

bending of prior, explicit rules for the benefit of a limited network (Mungiu-Pippidi 2006;

Rothstein and Teorell 2008) and it can be validated with other proxies (Fazekas and Kocsis

2017).

An important inconvenience is that the indicators are not ranked based on their importance.

Hence, the resulting score is not as explicit as it should be. In fact, the weighted average CRI

score may only reflect the number of indicators composing the CRI, in the absence of categorical

and continuous variables which receive special treatment.



It might be adventurous to read a higher score as more prone to corruption as it is not evident

that a higher number of indicators equals a higher risk of corruption. The presence of a single

indicator - single bidding - may be sufficient to achieve the corrupt goal while the presence of

other numerous indicators may be insufficient to steer the contracts for the favorite company.

To address this issue, the paper introduces an alternative approach to weighted average,

specifically the Principal Component Analysis.


28 Daniel Pustan

3. Institutional framework

The crucial role that institutions play for the prosperity of a country was demonstrated by

Acemoglu and Robinson (2012). In the same note, most suggested measures to combat

corruption refers to features of institutional environment. Hence, a larger body of research

developed under the umbrella of “good governance”, a notion closely related with that of

institutional environment (Andersson and Heywood 2009, p. 750).

Transparency, meritocracy or the separation of bureaucrat’s careers from political influence as

well as the quality and quantity of regulation are some important areas of investigation (Islam

2006; Lindstedt and Naurin 2010; Bauhr and Grimes 2013; Bauhr et al. 2019).

3.1 Legislative setting

Prior to its adhesion to the European Union as a full member in 2007, Romania transposed the

public procurement Directives 2004/17/EC and 2004/18/EC of the European Parliament and of

the Council into national legislation. The main legislation governing public procurement in

Romania up until 2016 was represented by Government Emergency Ordinance (GEO) 34/2006.

The Ordinance has suffered numerous and substantial modifications while into force. Only

during the analyzed period I identified 17 changes.

Table 1 Amendments to GEO 34/2006 during the studied period

2009 2010 2011 2012 2013 2014

GEO 19 GEO 76 Law 279 Law 76 GEO 4 GEO 51

Law 84 Law 284 Law 114 GEO 44 Law 9

GEO 72 Law 278 GEO 77 GEO 31

GEO 35

Law 166

Law 193

In addition to the primary legislation, secondary and tertiary legislation apply to public

procurement. The secondary legislation refer to Government Decisions consisting of the detailed

arrangements on the application of the primary legislation. The orders and instructions issued

by the National Agency for Public Procurement (NAPP) compose the tertiary legislation. The

existence of this extra layer of legislation highlights the problem of red tape in Romania’s public

procurement (Doroftei and Dimulescu 2015).



Figure 2 The comprehensiveness of regulation in Romania

Source: Europam.eu

On May 23rd, 2016 a new package of public procurement laws consisting of three laws 98/2016,

99/2016 and 100/2016 replaced the Government Emergency Ordinance (GEO) 34/2006 and its

subsequent modification acts. These laws are meant to transpose Directives 2014/23/EU,

2014/24/EU, 2014/25/EU.

Mungiu-Pippidi (2015) shows that two thirds of the contracts awarded to favored companies

come from non-EU funds. Corruption with EU funds is foreseeable less prevalent considering

that most potential irregularities with these funds are covered by the Law 78/2000 on

prevention, discovering and sanctioning of corrupt acts. The national funds do not benefit of

such a level of protection. In this case, most prosecutions are undertaken based on the law

punishing the abuse or malfeasance of office (Euractiv 2014).

This law was amended later, after the period of analysis, specifically on June 15, 2016, by the

Constitutional Court decision number 405 which specified that the syntagm “performs faulty”

should be interpreted as “performs by breaking the law”. By “the law” is understood only laws,

simple Government Acts or Emergency Acts. This way secondary regulations, such as

Government Decisions or Ministerial orders, are excluded (Toma 2016).


30 Daniel Pustan

To decrease the period of delays in contract execution, on June 28, 2014 the government issued

an Emergency Act to increase the obstacles for submitting complaints. It requires a “legitimate

interest” as well as a guarantee of 1% and up to 100,000 EUR to be paid by complainer. The

guarantee was forfeited in the event that the complaint were rejected (regardless of the reason:

in substance, inadmissible or by way of a plea) (Avocat.net 2014; Euractiv 2014). On March 19,

2015, the Constitutional Court decided that the Emergency Act 51/2014 was unconstitutional.

3.2 Public procurement institutions

Until mid 2015, the institutional framework consisted of three institutions with competences

exclusively in relation to public procurement (Aprahamian and Pop 2016). The main responsible

body was the National Authority for Regulating and Monitoring Public Procurement (ANRMAP)

with a role in policy and regulatory supervision, controls the manner of the contract award

(ex-post) and applies legal sanctions where necessary. The second public body was the Unit for

Coordination and Verification of Public Procurement (UCVAP) with attribution in verification of

the procedural aspects related to the contract award process (ex-ante control). On May 20, 2015

both agencies were merged into one, namely, the National National Public Procurement Agency

(ANAP).

The National Council for Solving Complaints (CNSC) is the third public body specific to public

procurement system and represents the administrative, non-judicial first instance to solve

complaints lodged under public procurement jurisdiction. As a final remedy procedure, CNSC

decision can be challenged before the Court of Appeal located within the same range as the

contracting authority’s headquarters (Directorate-General for Regional and Urban Policy

(European Commission) and PWC 2016).

The Court of Accounts represents the main oversight body. It conducts ex-post audits related to

all types of funds reporting their findings to ANRMAP and suspicious irregularities under the

criminal law to the National Anticorruption Directorate (DNA). Other entities involved in

monitoring public procurement are: the Management Authorities and the Department for the

Fight Against Fraud (DLAF), the Competition Council, the Authority for Certifications and

Payments and the National Integrity Agency (Doroftei and Dimulescu 2015; Directorate-General

for Regional and Urban Policy (European Commission) and PWC 2016)

It is also worth mentioning the Electronic System for Public Procurement (SEAP) operated by

the Romanian Agency for Digital Agenda (formerly known as the National Management Centre



for the Informational Society). It is an electronic procurement system with the aim to increase

efficiency and transparency in the public procurement system. The platform was upgraded and

integrated with relevant institutions in public procurement on April 2, 2018 (Romanian Digital

Agenda Agency 2018).

3.3 Government characteristics

During the seven years period, Romania had eight Cabinets and three Prime Ministers (though

one of them for a very short period of time). The first party, the Democrat Liberal Party (PDL),

governed between December 22, 2008 and April 26, 2012 when it was dismissed through a

no-confidence motion. The second party, the Social Democratic Party (PSD), was in power

between April 27, 2012 and November 17, 2015.

Following parliamentary elections, Boc I cabinet, made of PDL, PSD and the Conservative Party

(PC), started its mandate on December 22, 2008 and was dismissed by Parliament through a

no-confidence motion on October 13, 2009 (Mediafax 2009). This was soon after PSD and PC

alliance left the government on October 1, 2009. Prime Minister Emil Boc and ministries

belonging to PDL party continued to ensure the interim until December 23, 2009.

On the aforementioned date, Boc II cabinet was invested by the Parliament as Traian Basescu

was re-elected president (Politica Românească 2011). Thus PDL was able to form a

parliamentary majority with the National Union for the Progress of Romania (UNPR) - a party

formed later by independent parliamentarians who either left or were dismissed from their own

party after 2008 parliamentary elections (Ciubotaru 2013).

For clarity purposes, it should be mentioned that 2004 was the last year when Presidential and

Parliamentary elections coincided. This is due to the fact that the Presidential mandate began to

last 5 years while parliamentary continued to be of 4 years. Hence Parliamentary elections took

place in December 2008 while presidential elections were held in December 2009 (Centrul

pentru Integritate în Business 2013).

Cabinet Boc 2 ended its mandate on February 6, 2009 after Emil Boc resigned as prime minister

following street protests. After a three days period of interim, cabinet Ungureanu was invested.

It had an independent prime minister but the ministers were occupied by the National Liberal

Party (PNL), UNPR and the Democratic Alliance of Hungarians in Romania (UDMR). This


32 Daniel Pustan

cabinet only lasted for three months as it was dismissed by the Parliament on April 27, 2012

through a non-confidence motion.

After the dismissal of cabinet Ungureanu, Victor Ponta became prime-minister being in charge

of four consecutive governments between May 7, 2012 until November 5, 2015 when he resigned

following street protests claiming that the tragedy from “Colectiv” club could have been avoided

with more efficient measures to combat corruption (Tran 2015).

The first Ponta cabinet, formed by the Social Liberal Union (USL) formed by PSD and PNL was

in charge until parliamentary elections in December 2012. Cabinet Ponta 2 was formed after the

elections and co-opted UNPR in the government. PNL leaving the government lead to

government restructuring and co-opting the hungarian minority party, UDMR, which resulted

into cabinet Ponta 3. PNL decided to break the coalition suspecting that their ally will not keep

the agreement and will launch their own candidate for presidency instead of supporting PNL’s

candidate.

Cabinet Ponta 4 was formed due to UDMR’s departure from the government as a result of a

massive vote in the presidential election of their electorate in favor of Klaus Iohannis,

contra-candidate of prime minister in charge to presidency.

After a short interim period generated by the prime-minister resignation as consequence of

street protests following Colectiv nightclub fire, the independent Ciolos cabinet was invested.

The investment by the parliament took place on November 17, 2015. Ciolos government was in

power until December 2016 parliamentary elections.

A summary of government’s succession in presented in table 2.



Table 2 Romania’s government cabinets during 2009-2015

Cabinet Period Prime

minister Party Other parties

Type of

ending

Boc I 2008,

Dec 22

2009,

Dec 23 Emil Boc PD-L PSD+PC (PUR)

No-confidence

motion

Boc II 2009,

Dec 23

2012,

Feb 9 Emil Boc PD-L UDMR, UNPR

Government

resignation

Ungureanu 2012,

Feb 9

2012,

May 7

Mihai

Razvan

Ungureanu

Independent PDL, UNPR,

UDMR

No-confidence

motion

Ponta I 2012,

May 7

2012,

Dec 21

Victor

Ponta PSD PNL, PC (PUR)

Parliamentary

elections

Ponta II 2012,

Dec 21

2014,

Mar 5

Victor

Ponta PSD

PNL, PC (PUR),

UNPR

Government

restructuration

Ponta III 2014,

Mar 5

2014,

Dec 17

Victor

Ponta PSD

PC, UDMR,

UNPR, PLR

Government

restructuration

Ponta IV 2014,

Mar 5

2015,

Nov 17

Victor

Ponta PSD

UNPR, ALDE

(PLR + PC)

Government

resignation

Figure 3 presents the two government periods together with the party composition of

government cabinets during the observation period.

Figure 3 Timeline of cabinets and parties for the period 2009 - 2015


34 Daniel Pustan

The average period time of governing for a cabinet was 12 months during the 7 years observation

period. The shortest period in power was that of cabinet Ungureanu (3 months) while the

longest running cabinet was cabinet Boc 2 (25 months). For simplicity, the interim periods were

considered jointly with the cabinet period that preceded them. Furthermore, this paper

proposed to capture the influence of the main parties (PDL and PSD) along a

multi-governmental period which included various cabinets.



4. Method

This paper uses a quantitative approach defined as “entailing the collection of numerical data, a

deductive view of the relationship between theory and research ... and an objectivist conception

of social reality” (Bryman 2012, p. 149). In this study, quantitative research is preferred over

qualitative research as it allows testing hypotheses based on directly observable behavior. The

increase in government transparency and availability of big data on public procurement

contracts made this type of research possible.

Further, a particular form of longitudinal analysis, namely, panel analysis was performed after

aggregating the data on a half yearly basis. The main benefits of using panel data are that it

accounts for heterogeneity and allows to study dynamic factors such as the spells of contract

value won by companies (Hill, Griffiths and Lim 2011).

Following Dávid-Barrett and Fazekas (2019) two hypotheses are verified. First, it is identified

for which companies the change in government affects the value of contracts won. Second,

corruption indicators or “red flags” are analyzed to determine whether the change in companies’

winning tender is also associated with a higher risk of corruption.

Companies benefit of favoritism where both conditions are met. Additionally, favoritism is

considered systematic if the patterns of winning contracts in the presence of more red flags than

the rest of the market can be observed at the level of company type.

Fine shifts in government spending or more sophisticated methods of corruption are the most

reasonable explanation when only the first hypothesis is true. If only the second hypothesis is

true, it is probably because firms buy influence from both parties or other authorities capable to

influence the process (Dávid-Barrett and Fazekas 2019).

Controlling for changes in spending preferences as a result of differences in policy priorities

(CVMti), little variation is expected in contract winning tenders as a result of the change in

government.

To identify the companies for which the value of contracts won was affected by the change in

government, I use the Arellano-Bond system GMM transformation of the following dynamic

panel regression model:


36 Daniel Pustan

CVti = α + β

1 CVt-1i + β

2 CVt-2i + β

3 CVMti + β

4 MMi + β

5 Cabt + U

i + Ɛit

The abbreviations of the model stand for the following concepts:

CVti : the contract value won by company in period t

α : the constant term for the sample

CVt-1i : the contract value won by company i in past period (t-1)

CVt-2i : the contract value won by company i in past period (t-2)

CVMti: the contract value spent in main market (CPV level 3)

MMi : the sectoral dummy for the main market (CPV level 3)

Cabt : dummy representing the cabinet in period t

Ui : the company fixed effect

Ɛ it : the error term

I ran the two versions of the above model - fourth root and natural log contract value - on a

subsample of companies winning contracts in at least two different periods (removing 17,819

observations). The reason for using fourth root is the ability to rescale the data while keeping the

observations with zero contract value during the observed period. The two models reveal favored

companies at different stages of maturity. The fourth root version explore the case of both well

established and incumbent companies. The natural log version analyzes only the more

experienced companies in the procurement market as it removes the companies with zero

contract value observations during the analyzed period (Dávid-Barrett and Fazekas 2019).

During the first nine months, PDL and PSD governed together in cabinet Boc I. However, as

important as it may be, given that the period of governing together took place during the first

two half-year periods and the model uses two lags, these observations are not used. They are

automatically dropped because of perfect collinearity as consequence of the fact that all the

observations are zero after the first two periods.

Based on the regressions company-specific error term, companies having ties to government 1

are labeled “surprise losers” and those with government 2 are called “surprise winners”.



Specifically, companies that have an above period average regression error under one

government and below period average regression error under the other government are labeled

as “surprise losers” and “surprise winners”. Companies with stable residual patterns (above or

below period average in both governments) are considered “stable companies”.

The same terminology as Dávid-Barrett and Fazekas (2019) is used in this paper: “surprise

loser” and “surprise winner”. These terms are defined in relationship to the second government

period and because their use might give rise to misinterpretation, clarification may be necessary.

Thus, the reader is cautioned not to extend the negative and positive connotations of these terms

to the entire period of study. “Surprise losers” while possibly discriminated during the second

government period, might have benefited from preferential treatment during the first

government period. Therefore, it should be borne in mind that “surprise losers” are the

companies suspicious of having ties with government 1 while the “surprise winners” are the

firms suspicious of links with government 2.

The regressions set-up controlled for the following: (i) government spending patterns resulting

from different policy preferences or priorities measured by contract value spent in main market,

(ii) market and technological specifics measured by CPV division (2-digit CPV) and (iii)

regulatory as well as government specificities (e.g. parties composing the government) measured

by cabinet.

Once identified the companies with a suspicious winning pattern, I analyze whether they won

under conditions with a higher potential for corruption. In order to do so, the surprise winners

and losers are cross checked with the Corruption Risk Index (CRI).

The Corruption Risk Index is a relatively recent tool developed by Fazekas, Tóth and King

(2016) aiming to measure the degree of corruption in a more objective manner as a reliable and

cost efficient alternative to current measurements. The index is based on previous research

(Charron et al. 2017, Fazekas and Kocsis 2017, Dávid-Barrett and Fazekas 2019) referring to not

less than 30 quantitative indicators and 20 corruption techniques in public procurement were

identified. A summary table with all these indicators is presented in annex B (Fazekas, Tóth and

King 2013).


38 Daniel Pustan

The model includes one outcome indicator - single bid - and various input indicators. The

elementary corruption risk indicators (red flags) that constitute the CRI are identified based on

the literature review:

1. Single bid (outcome indicator)

2. Call for tender publication (submission phase)

3. Procedure type (submission phase)

4. Weight of non-price evaluation criteria (assessment phase)

5. Procedure period (assessment phase)

6. Notification of award period (assessment phase)

7. Contract type (assessment phase)

Because “submission deadline” was not present in the dataset, it is not possible to build the

continuous variable indicators “length of advertisement period” and “length of decision period”.

However, two similar indicators are used, namely “procedure period” and “notification of award

period”. The “procedure period” indicator represents the number of days between the call for

tender and contract date. The “notification of award period” computes the number of days

between the contract date and notification of award date.

Binary logistic regression is used to determine the suitability of each composite index in the

index excluding those indicators that do not prove to be significant and strong predictors of

single bidding (Fazekas and Kocsis 2017). The logistic model is:

Pr (single bidder = 1) = 11+e−zi

Zi = α + β

1 Sij + β2 Aik + β

3l Dil + β

4m Cim + Ɛi

Where:

Zi - the logit of a contract being a single bidder contract

α - the constant of the regression

Sij

- the matrix of j corruption inputs of the submission phase for the ith contract such as length

of submission period



Aik

- the matrix of k corruption inputs of the assessment phase for the ith contract such weight

of non-price evaluation criteria

Dil

- the matrix of l corruption inputs of the delivery phase for the ith contract such contract

lengthening

Cim

- the matrix of m control variables for the ith contract such as the number of competitors

on the market

Ɛ i - the error term

β1, β2, β3l, β4m - the vectors of coefficients for explanatory and control variables

Afterwards, component weights for each variable are derived. The outcome indicator - with no

regression coefficient - receives the weight of 1. Also, each significant corruption input receives

the weight of 1 while those variables or variable categories that do not prove significant are

discarded (weight zero).

The categories within the categorical variables received component weights ranked according to

their regression coefficient. For instance, in the case of two significant categories, the one with

the highest coefficient would receive the weight of 1 while the other one receives the weight 0.5.

For continuous variables, threshold were identified using linear regression analysis and the

residual distribution graphs to determine discrete jumps in residual values. The control

variables represent the most important alternative explanation for the outcome.


40 Daniel Pustan

Figure 4 Mean regression residuals by two-percentiles of relative procedure period (contract

date - call for tender date), linear prediction

Figure 5 Mean regression residuals by two-percentiles of relative notification of award period

(award notice date - contract date), linear prediction



Both figures display three different period-groups. One which overestimates the probability of

single received bid, one which underestimates this probability and, lastly, one that approximate

it.

Moreover, each corruption input was introduced step by step starting with those from the

earliest phases. The inputs that prove to be insignificant were not introduced in the CRI

composite indicator.

The control variables in the regression are: (i) institutional endowments measured by type of

issuer (national, county or municipal), (ii) market and technological specifics measured by CPV

division (2-digit CPV), (iii) contract size measured by log of contract value in euro, (iv)

legislation measured by year and (v) contracting authorities specificities measured by activity

type.

Finally, CRI is calculated in two ways. First, as the arithmetic average of the individual risk

indicators - red flags (Fazekas, Tóth and King 2016):

CRI = j w

j * CIjiΣ

Secondly, CRI is computed using an alternative technique, namely the Principal Component

Analysis (PCA).

The validation of the Corruption Risk Index was done both at macro and micro level. At the

macro-level I compared the Corruption Risk Index with three notorious perception surveys: The

Corruption Perception Index (CPI) from Transparency International, World Bank’s World

Governance Indicators (WGI) and the Global Competitiveness Index (GCI) developed by The

World Economic Forum.

This last indicator, GCI, covers a much broader area besides corruption (e.g. efficiency of

government) for which reason only the most relevant indicator for this purpose was selected,

specifically, favoritism in decisions of government officials (1.07).

The validation of the Corruption Risk Index at micro-level is done by looking at the correlation

of the relative contract value and CRI. The relative contract value is calculated as the ratio

between the actual contract value and the estimated contract value (Fazekas, Tóth and King

2016). A higher ratio of relative contract value represent a higher risk of corruption.


42 Daniel Pustan

A second micro-level indicator, namely, winner’s country of origin (Fazekas and Kocsis 2017)

was not applicable to this study. The purpose of this indicator is to analyze the correlation of

companies’ main office located in a tax haven country related to the CRI. However, no winner of

public procurement contracts during the 2009-2015 reported their main office in a country

included on the EU non cooperative tax jurisdictions and high risk third country list (European

Parliamentary Research Service 2018).



5. Data

The public procurement data contains all public procurement contracts in Romania between

2009 and 2015. The original datasets were published on data.gov.ro in the form of 11 .csv files by

the Romanian Agency for Digital Agenda (Mihalache 2013). Alternative data was available from

The European procurement journal, Tenders Electronic Daily, which comprised of the contracts

above a certain threshold.

Nevertheless, the data published by the Romanian national agency was prefered because it

allows a better understanding of procurement market in its entirety. Moreover, the complete

dataset is suitable to offset the corruption technique of fragmenting big contracts into smaller

contracts. This technique is used to conceal irregularities or simply to benefit of more permissive

rules.

Unfortunately, the poor quality of the data required a substantial work in order to reach an

adequate quality for scientific research. One obstacle consisted in the fact that the elementary

variable, the Company Identification Number (CIN), included multiple inaccurate observations.

Keeping these observations in the analysis required to manually correct those instances. In spite

of the time consuming process, this effort was carried out with the hope that other researchers

may also benefit from it.

Over 5,000 companies were searched by name on the Ministry of Finance and other specialized

websites (e.g. listafirme.ro and totalfirme.ro). The entities that were still not found, most of

them consisting of self-employed persons, were assigned a unique identification comprising

their names without spaces. In addition, all foreign companies’ CIN was verified. Kompass.com

website together with the official companies registry websites of their countries of origin were

employed for this purpose.

Once the correct CIN was assigned for the winning companies, the names and addresses were

extracted from the dataset published by the Ministry of Finance on the website data.gov.ro. This

dataset consists of all economic operators and public entities. For the operators without a CIN as

well as for foreign companies, the name and address were extracted from a dataset created for

this purpose.


https://www.listafirme.ro/

https://www.totalfirme.ro/

https://kompass.com/

44 Daniel Pustan

Inconsistencies between the CIN and companies’ name in the dataset were tackled by

triangulating with company’s address variable. Whichever between the two - CIN or company

name - matched the address was considered correct while the other was changed accordingly.

Companies that changed their name for branding or legal purposes were assigned the new name

during the entire observation period. Examples of such companies are Engie Romania SRL,

previously GDF Suez Energy and Veolia Energie Prahova SRL which changed from Dalkia

Thermo Prahova SRL due to the acquisition by the former.

The same process was conducted for the contracting authorities related variables. First, the

validity of their CIN was verified. In the event of a mismatch between the entities’ CIN and the

name variable, these were triangulated with the address variable. Subsequently, the name and

address were imported from the Ministry of Finance dataset (Ministry of Finance 2018).

To ensure the quality and integrity of the database, other correction measures were taken for the

existing errors and inconsistencies in all variables. Some examples of the issues encountered

are: broken lines, leading and trailing spaces, consecutive spaces, use of diacritics and non

alphanumeric characters, etc.

The association of companies was treated similarly to Doroftei (2016) and Doroftei and

Dimulescu (2015). Specifically, the contracts awarded to a consortium were assigned to the

leader. This is consistent with the thesis that most frequently it is the leading company with

political connection that enables the consortium to win the contract. Nevertheless, this approach

may generate inflated contract values for some companies.

The regressions were fitted on a subsample of competitive markets. Competitive markets are

defined by product group (CPV group - level 3) and regions (NUTS1) with at least three

participants (Fazekas, Tóth and King 2016). Non-competitive markets under normal

circumstances, such as those for specialized services or defense, were excluded to reduce false

positive due to the underlying assumption that the lack of competition signals corruption.

(Dávid-Barrett and Fazekas 2019, Fazekas and Kocsis 2017).

Albeit the purpose of this paper is to measure governmental favoritism, the sample was not

limited only to government related contracts for two reasons. First, in a centralized country like

Romania the government has significant influence on the local governing authorities through



the allocation of funds. Secondly, most of the vote buying performed by local party organizations

is carried on behalf of the national party organization.

In any case, the distribution of procurement by the type of contracting authority is as follows:

the government and its related agencies is the main contractor with 45% of the total value of

acquisitions. It is followed by the local administration authorities with 35% and the county level

administration with 11%. The remaining 9% consists of non-governmental organizations

(NGO’s) and companies.

Company level analysis was performed on a half-yearly aggregated database comprising the

companies that won contracts in at least two periods between 2009 and 2015 period. Half-year

period “is optimal for retaining a high level of granularity while also taking into account the

erratic character of many public procurement markets” (Dávid-Barrett and Fazekas 2019, p.13).

The government periods were defined in relation with the government party, the party providing

the political support to the prime minister and holding the majority of the ministerial positions

in the cabinet. Apart from this, one year period of transition covering Ungureanu and Ponta I

cabinets was allowed as a means to capture the differences between two established parties in

power (Dávid-Barrett and Fazekas 2019).

Table 3 Main characteristics of the dataset

2009 2010 2011 2012 2013 2014 2015 Total

Total number of contracts awarded 115,868 121,170 103,342 106,507 65,413 72,560 67,662 652,522

Total number of winners 9,890 10,208 9,144 7,933 7,041 6,509 6,255 16,010

Total number of contract. authorities 3,983 3,647 3,331 3,189 2,814 2,813 2,775 6,826

Total value of awarded contracts (mil. EUR)

11,100 10,400 15,200 13,400 13,900 16,300 16,400 96,700


46 Daniel Pustan

6. Results

The analysis reveals moderate evidence of partisan favoritism in Romanian public procurement

market. Specifically, 24% of the public procurement market in Romania appear to be

dominated by favored companies of which 15% surprise losers and 9% surprise winners. This is 3

above the 10% market capture in the United Kingdom’s but lower than 50%-60% results in

Hungary (Dávid-Barrett and Fazekas 2019).

Outcomes similar to Hungary were expected given the comparability between the two countries

in many aspects including the geographical proximity, the communist background and the

equivalent rankings in the corruption perception index. However, there is a resemblance,

specifically in that, like Hungary,the Romanian public procurement market is characterized by

systematic favoritism. That is to say that favoritism patterns are identifiable at the level of

company type.

Using Arellano-Bond system GMM transformation of the dynamic panel regression, I find

moderate evidence of company performance through the analyzed period (Table 4). The first 4

model has a moderate explanatory power while the second model has a relatively low

explanatory power: 0.69 and 0.38. 5

In Dávid-Barrett and Fazekas (2019) using a smaller sample on Hungary (log model), the

predicting power drops dramatically by almost 90%. This can be a sign of a non-representative

sample for the population or there can be influential observations which are lost in the natural

log model. The fact that in my case the sample decreases much less is one of the reasons for the

reduced decrease in predicting power for the log with respect to the fourth root model.

3 Results for both “fourth root” and “log” contract value regressions were taken into account

4 Sargan test of overidentification of instruments show that the null-hypothesis could not be rejected

5The linear correlation coefficient between predicted and observed outcome was used



Table 4 System GMM linear dynamic panel regression explaining company market success

fourthrootvaloareeur lnvaloareeur

fourthrootvaloareeur = L1 0.0274***

(0.00453)

fourthrootvaloareeur = L2 0.0220***

(0.00431)

lnvaloareeur = L1 0.0732***

(0.0171)

lnvaloareeur = L2 0.120***

(0.0148)

Control variables

fourthrootCVmainmarket Y N

lnCVmainmarket N Y

main market Y Y

cabinet Y Y

Constant 14.32*** 6.814***

(1.421) (2.054)

Observations 192,120 28,331

Number of panelID 16,010 6,592

Robust standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

I identified 3,860 suspicious companies (2,359 surprise losers and 1,501 surprise winners) out of

the total of 16,010 companies. These company groups have a CRI trajectory consistent with

favoritism. Surprise losers win in the presence of more red flags than the rest of the market

during the first government period and surprise winners have a higher CRI than the rest of the

market during government 2 period.

There is a significant positive persistence effect. In particular, a one percent increase in contract

value increases by 7% after one period and by 12% after two periods. For the fourth root model,

significant positive effects with one or two lags are also found in the fourth root model.


48 Daniel Pustan

Figure 7 CRI-arithmetic by company group, half-year period

Figure 8 displays the overall market share trends for each company group. This shows clear

contrasting market outcome trajectories between surprise losers and surprise winners. The

difference is lower during the period in which the two parties governed together.

Figure 8 Combined market shares of company types, half-year period



In the following, results for the Corruption Risk Index are presented.

First, not publishing the call for tenders strongly predicts single bidding. The average probability

of a single received bid increases by 135%.

Second, contrary to expectations, only the “accelerated negotiation” and “negotiation without

call for tender” procedure types display higher risk of corruption signs. The contracts with a

single bid in an accelerated negotiation tend to be awarded to winning companies 141% more

often as compared to open bid contracts. The “negotiation without call for tender” award 92%

more single bid contracts than open procedure does. All the other non-open procedure types

predict decreases in single bidding. However, “competitive dialogue”, “limited accelerated

tendering” and “limited tendering” are insignificant and according to the theoretical framework

will not be included in the composite indicator.

Third, the indicator “weight of non-price criteria” is a powerful predictor but with a weak to

moderate impact in explaining single bidding contracts. In the complete model the probability

of single bidding decreases by about 13%.

Fourth, an extremely short procedure period (up to 24 days) is a very strong predictor of single

bidding with 218% compared to those of over 77 days. The contracts awarded between 24 and 77

days from the call for notice date have a more moderate predicting capacity of single bid with

63%.

Fifth, the notification of award period is associated with a low corruption risk. The contracts

where the award notice was published within 8 days after the signing of the contract have a 17%

higher risk of corruption while there is a lower risk of corruption roughly by the same

percentage for the contracts signed over 27 days as compared to contracts between 8 and 27

days.

Sixth, compared to supplies, the works/construction contract type decreases the risk of

corruption by 54% whereas the services increases the risk with 35%.


50 Daniel Pustan

Table 5 Logit regression results on contract level, 2009-2015, number of winners 2≥

singleb

Procedure type

Ref. cat: 3, Open tendering

1, Competitive dialogue -0.475

2, Invitation to tender -0.916***

4, Limited tendering -0.130

5, Limited accelerated tendering -0.0659

6, Negotiation -0.336***

7, Accelerated negotiation 1.412***

8, Negotiation without contract notice 0.918***

No call for tenders 1.345***

Weight of nonprice

Ref. cat:Lowest price

Most economically advantageous tender -0.129***

Procedure period

Ref. cat: 3, 77 days - max

1, 0 - 24 days 2.179***

2, 24 - 77 days 0.631***

4, missing 0.223***

Notification of award period

Ref. cat: 2, 8 - 27 days

1, 0 - 8 days 0.176***

3, 27 days - max -0.194***

Contract type

Ref. cat:Supply

2, Works -0.540***

3, Services 0.351***

Control variables

Type of issuer Y

CPV division Y

lnContractValue Y

Year Y

Activity type Y

Constant -1.663***

Observations 362,815

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1



Table 6 presents the component weights derived based on the regression results.

Table 6 CRI component weights

Variable Component weight

Single bid 1

Procedure type

Ref. cat: 3, Open tendering

7, Accelerated negotiation 1

8, Negotiation without contract notice 0.75

6, Negotiation 0.5

2, Invitation to tender 0.25

No call for tenders 1

Weight of nonprice

Ref. cat:Lowest price

Most economically advantageous tender 1

Procedure period

Ref. cat: 3, 77 days - max

1, 0 - 24 days 1

2, 24 - 77 days 0.66

4, missing 0.33

Notification of award period

Ref. cat: 2, 8 - 27 days

1, 0 - 8 days 1

3, 27 days - max 0.5

Contract type

Ref. cat:Supply

3, Services 1

2, Works 0.5


52 Daniel Pustan

The molding of the CRI on the definition of grand corruption and the fitted regression models

are the main supporters for the validity of the Corruption Risk Index. The macro-level

perception surveys only reinforces this validity.

Table 7 Bivariate Pearson correlation of CRI and single bid with perception survey indicators

Indicator CRI-arithmetic CRI-PCA single bid years

CPI -0.7388 -0.8357 -0.4885 7

WGI (1.07) -0.6862 -0.8321 0.0045 7

GCI -0.4205 -0.3414 -0.3741 7

Figure 9 CRI arithmetic compared to perception-based surveys (CPI, WGI, GCI)

Figure 10 CRI-PCA compared to perception-based surveys (CPI, WGI, GCI)



Figure 11 Single bid compared to perception-based surveys (CPI, WGI, GCI)

The main reason for the moderate level of correlation stems from the fact that perception-based

indicators are more adequate to compare the levels of corruption across countries for longer

periods of time (Escresa and Pici 2016 acc. Fazekas and Kocsis 2017) as they capture mostly

pitty corruption and they are not sensitive to change.

A more reliable indicator is the relative contract value. However, the estimated contract value

variable present various inexactitudes that may bias the analysis. The following are some of the

observed inaccuracies: (1) use of commas to separate decimals incorrectly, (2) Unrealistically

low estimated contract values (below 20€), (3) the estimated contract value appears for the

entire project but the real contract value is divided in various batches and (4) Total contract

value is repeated because some companies won contracts together but each company is

presented with the total amount as if each of them has won the total amount. The outliers with a

relative contract value below five are presented in figure 12.

Figure 12 Outliers for relative contract value < 2


54 Daniel Pustan

For lack of a better solution, considering the time constraints of this study, the regression was

run on a subset of companies with relative contract value lower or equal to 1 (one) in order to

minimize the number of erroneous observations captured in the analysis.

Table 8 Linear regression explaining relative contract value and CRI

(1) (2) (3) (4) (5) (6)

relativeCV relativeCV relativeCV relativeCV relativeCV relativeCV

cri_arithmetic 0.284*** 0.418***

-0.00323 -0.00408

cri_pca 0.373*** 0.677***

-0.00477 -0.00669

singleb 0.147*** 0.124***

-0.000977 -0.00114

Control variables

TipActivitateAC N Y N Y N Y

AuthorityType N Y N Y N Y

Year N Y N Y N Y

CPV division N Y N Y N Y

lnCV N Y N Y N Y

Constant 0.643*** 0.699*** 0.713*** 0.377*** 0.462*** 0.528***

-0.00121 -0.000791 -0.00044 -0.00841 -0.00868 -0.00823

Observations 289,637 224,711 289,637 212,008 169,687 212,008

R-squared 0.026 0.027 0.072 0.154 0.159 0.16

Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

The results confirm the expectations. One additional red flag (1/7 points higher) increases the

price by 4% - 6%. Similar values are obtained for CRI-PCA where the price increases by 5.3 -

9.7%. Single bidding contracts have on average 12.4 - 14.7% higher prices than contracts with at

least two bidders.



7. Robustness analysis

In this section robustness checks are presented. The CRI-PCA approach and the alternatives in

time periods and the change in control variables confirm the robustness of the model.

First, the CRI weighted using Principal Component Analysis (PCA) display similar results to the

CRI calculated as an arithmetic average.

Figure 13 - CRI-PCA patterns by company group, half-year period

Secondly, changing the aggregated period for the company analysis from half-year to one year

yields similar results. According to this specification, the number of favored companies remains

almost identical with that for the half year period. Specifically, 13.22% surprise losers and 11.15%

surprise winners are identified. In addition, the CRI trajectories are consistent with systematic

favoritism.


56 Daniel Pustan

Figure 14 CRI-arithmetic patterns by company group, one year period

Figure 15 CRI-PCA patterns by company group, one year



Thirdly, variables representing all the changes to the procurement law during the observation

period were introduced replacing the variable cabinet as a control for changes in legislation.

Again, the results confirmed the robustness of the base model. A very similar number of favored

companies was identified (24.15%) and systematic favoritism was evident in these cases, as well.

Figure 16 CRI-arithmetic patterns by company group, changes in legislation

Figure 17 CRI-PCA patterns by company group, changes in legislation


58 Daniel Pustan

8. Conclusions and suggestions for further research

This paper has examined the political influence at the highest level on public procurement

market in Romania during 2009-2015. Drawing on the most recent research and novel

measurement tools of corruption the study reveals that 24% of the Romanian procurement

market is determined by systematic partisan favoritism.

This number is below the 50% market associated with partisan favoritism in Hungary

(Dávid-Barrett and Fazekas 2019). Considering the geographical proximity as well as cultural

and institutional similarities, a similar score to Hungary was expected.

There are a few possible explanations why that was not the case. First, for comparison purposes,

this paper followed Dávid-Barrett and Fazekas (2019). However, the methodology differed in

two aspects. Specifically, Dávid-Barrett and Fazekas (2019) sample was restricted to contracts

above the EU-threshold and only purchases made by the government and its agencies.

On the other hand, this study included all contracts during the observation period at all levels of

public administration. Not restraining the study only to government level contracts is justified

by the high level of centralization both administratively and politically in Romania. A very

efficient control over the local administration is carried out through budget allocation.

Secondly, the more vibrant and engaged civil society in Romania may have had a significant

impact in mitigating corruption. The analyzed period was especially troubled in which two

governments had fallen as a consequence of street protests.

A third explanation is related to the high frequency of party switching by Romanian politicians,

famous for their lack of ideological attachment. The migration from one party to another also

moves the clientele between parties as the favors will be offered through the party in which the

politician is currently found. This situation hinders the possibility to identify groups of clientele

specific for parties. A good illustration illustration of mass migration from one party to another

occurred in August 2014 with the GEO 457/2014. Then, thousands of mayors were able to

migrate to PSD party without ceasing to hold office as the law stipulated (Bărbulescu 2014).



The result is more in line to Doroftei (2016) and Doroftei and Dimulescu (2015) findings. Their

study discovered 32% favoritism in construction procurement sector in Romania. It should be

noted though that construction sector is amongst the most susceptible of corruption.

In conclusion, the results of this paper indicate some progress made by Romania in the fight

against high-level corruption in public procurement. However, it invites to moderate optimism

since a decrease in one form of corruption does not necessarily imply a drop in the overall

corruption. The method applied in this paper cannot rule out alternative practices such as

companies buying influence from different parties.

In any case, these results open the way for debates to combat corruption through good

government practices that are based on objective proxy measures. In addition, this study

provides empirical support for EU anti-corruption policies. It may help focus the attention on

specific and quantifiable forms of corruption entailing more precise recommendations by the

European Commission’s in the effort to support the fight against corruption in Romania within

the Cooperation and Verification Mechanism.

In order to be able to fully exploit the potential benefit for the society of this field, further

research is needed. First, as it was shown objective proxy measures are promising tools in

understanding the underlying mechanisms of corruption. Yet they are still in an early phase of

development. Reliable indicators not only in extension but also in depth, ranking the

importance of corruption risk indicators, should be developed.

Another direction of research is represented by the analysis of corruption at local level as well as

the interaction between central and local government related to favoritism. Finally, party

switching in Romania is a far too common practice. Studying political favoritism at a more

granular level, that of the individual politician, may bring important insights on the drives of

political favoritism.


60 Daniel Pustan

References

Acemoglu, D. and Robinson, J. (2012). Why nations fail: the origins of power, prosperity, and poverty. 1st

ed. New York, NY: Crown Business.

Aidt, T. S. (2010). Corruption and Sustainable Development. https://doi.org/10.17863/CAM.5249

Amore, M. and Bennedsen, M. (2013). The value of local political connections in a low-corruption

environment. Journal of Financial Economics, 110(2), pp.387-402.

Andersson, S. and Heywood, P. M. (2009), The Politics of Perception: Use and Abuse of Transparency

International's Approach to Measuring Corruption. Political Studies, 57: 746-767.

Andrei, A. (2018). Topul firmelor care lucrează cu banii statului. Contracte cât bugetul Iaşului. Ziarul de

Iași. [online] Available at:

https://www.ziaruldeiasi.ro/stiri/topul-firmelor-care-lucreaza-cu-banii-statului-contracte-cat-bugetu

l-iasului--181634.html [Accessed 19 Dec. 2019].

Aprahamian, A. and Pop, G. (2016). Building landmarks, smoothing out markets: An Enhanced

Competition Framework in Romania. World Bank Studies. Washington, DC: World Bank. License:

Creative Commons Attribution CC BY 3.0 IGO

Arvate, P., Barbosa, K. and Fuzitani, E. (2018). Party Expertise, Campaign Donation and Government

Contracts: Evidence from an Electoral Quasi-Experiment. INSPER Working paper 377/2018.

Attström, K., Kröber, R., Jähne, J., Bergschneider, J., Thomä, J., Swysen, T., Scholzen, D. and

Trautenmüller, T. (2012). Final Report: Review of the Functioning of the CPV Codes / System.

[online] Berlin: Ramboll. Available at:

https://ec.europa.eu/docsroom/documents/21583/attachments/1/translations [Accessed 9 Dec.

2019].

Auriol, E., Straub, S. and Flochel, T. (2016). Public Procurement and Rent-Seeking: The Case of Paraguay.

World Development, 77, pp.395-407.

Avocatnet.ro. (2014). Garantia de buna conduita obligatorie pentru contestarea licitatiilor publice.

[online] Available at:

https://www.avocatnet.ro/articol_38023/Garantia-de-buna-conduita-obligatorie-pentru-contestarea

-licitatiilor-publice.html [Accessed 20 Dec. 2019].

Bacalu, M. (2018). Uioreanu şi Bene merg la închisoare. Sentinţă finală în dosarul de corupţie aeroport

[online] Muzicainstantelor.ro. Available at:

https://muzicainstantelor.ro/index.php/2018/01/31/uioreanu-si-bene-merg-la-inchisoare-sentinta-f

inala-in-dosarul-de-coruptie-de-la-aeroportul-cluj/ [Accessed 23 Nov. 2019].

Bandiera, O., Prat, A. and Valletti, T. (2009). Active and Passive Waste in Government Spending:

Evidence from a Policy Experiment. American Economic Review, 99(4), pp.1278-1308.

Bărbulescu, O. (2014). Ordonanţa de Urgenţă privind migraţia primarilor a fost adoptată de Guvern.

Mediafax. [online] Available at

httpswww.mediafax.ropoliticordonanta-de-urgenta-privind-migratia-primarilor-a-fost-adoptata-de-g

uvern-13160000 [Accessed 1 Jan. 2020].


http://norayr.am/collections/books/Why-Nations-Fail-Daron-Acemoglu.pdf

https://doi.org/10.17863/CAM.5249

https://doi.org/10.1016/j.jfineco.2013.06.002

https://doi.org/10.1016/j.jfineco.2013.06.002

https://doi-org.focus.lib.kth.se/10.1111/j.1467-9248.2008.00758.x


https://www.ziaruldeiasi.ro/stiri/topul-firmelor-care-lucreaza-cu-banii-statului-contracte-cat-bugetul-iasului--181634.html

http://dx.doi.org/10.1596/978-1-4648-0620-9

http://dx.doi.org/10.1596/978-1-4648-0620-9

https://www.researchgate.net/profile/Paulo_Arvate/publication/301494814_Party_Expertise_Campaign_Donation_and_Government_Contracts_Evidence_from_an_Electoral_Quasi-Experiment/links/57166ee608ae497c1a570020/Party-Expertise-Campaign-Donation-and-Government-Contracts-Evidence-from-an-Electoral-Quasi-Experiment.pdf

https://www.researchgate.net/profile/Paulo_Arvate/publication/301494814_Party_Expertise_Campaign_Donation_and_Government_Contracts_Evidence_from_an_Electoral_Quasi-Experiment/links/57166ee608ae497c1a570020/Party-Expertise-Campaign-Donation-and-Government-Contracts-Evidence-from-an-Electoral-Quasi-Experiment.pdf

https://ec.europa.eu/docsroom/documents/21583/attachments/1/translations

https://doi-org.focus.lib.kth.se/10.1016/j.worlddev.2015.09.001

https://www.avocatnet.ro/articol_38023/Garantia-de-buna-conduita-obligatorie-pentru-contestarea-licitatiilor-publice.html

https://muzicainstantelor.ro/index.php/2018/01/31/uioreanu-si-bene-merg-la-inchisoare-sentinta-finala-in-dosarul-de-coruptie-de-la-aeroportul-cluj/

http://doi.org/10.1257/aer.99.4.1278

http://doi.org/10.1257/aer.99.4.1278

https://www.mediafax.ro/politic/ordonanta-de-urgenta-privind-migratia-primarilor-a-fost-adoptata-de-guvern-13160000


Bauhr, M. and Grimes, M. (2013). Indignation or Resignation The Implications of Transparency for

Societal Accountability. Governance, 27(2), pp.291-320.

Bauhr, M., Czibik, Á., Fine Licht, J. and Fazekas, M. (2019). Lights on the shadows of public procurement

Transparency as an antidote to corruption. Governance, pp.1-29.

Bayley, D. (1966). The Effects of Corruption in a Developing Nation. The Western Political Quarterly,

19(4).

BBC (2015). Romania PM Ponta resigns over Bucharest nightclub fire. 4 November, [online] Available at:

https://www.bbc.com/news/world-europe-34720183 [Accessed 17 Dec. 2019].

BBC (2019). Romania corruption: PSD chief Liviu Dragnea jailed. [online] Available at:

https://www.bbc.com/news/world-europe-48421430 [Accessed 26 Dec. 2019].

Befu, D. (2015). Mama lui Ponta: ”Suntem o mică firmă de familie. Nu avem nimic de ascuns”. România

Curată vă prezintă principalii clienți ai firmei familiei lui Ponta, la care au ajuns azi procurorii DNA.

România Curată. [online] Available at:

http://www.romaniacurata.ro/mama-lui-ponta-suntem-o-mica-firma-de-familie-nu-avem-nimic-de-

ascuns-romania-curata-va-prezinta-principalii-clienti-ai-firmei-familiei-lui-ponta-la-care-au-ajuns-a

zi-procurorii/ [Accessed 19 Dec. 2019].

Blackburn, K., Bose, N. and Emranul Haque, M. (2010). Endogenous corruption in economic

development. Journal of Economic Studies, 37(1), pp.4-25.

Boas, T., Hidalgo, F. and Richardson, N. (2014). The Spoils of Victory: Campaign Donations and

Government Contracts in Brazil. The Journal of Politics, 76(2), pp.415-429.

Bogdan, S. (2018). De ce, în România, ne urâm de moarte pe tema anticorupției. Vice. [online] Available

at:

https://www.vice.com/ro/article/gy3g33/de-ce-in-romania-ne-uram-de-moarte-pe-tema-anticorupti

ei [Accessed 25 Dec. 2019].

Brogaard, J., Denes, M. and Duchin, R. (2015). Political Connections, Incentives and Innovation:

Evidence from Contract-Level Data. Working paper.

Bromberg, D. (2014). Can Vendors Buy Influence? The Relationship Between Campaign Contributions

and Government Contracts. International Journal of Public Administration, 37(9), pp.556-567.

Brugués, F., Brugués, J., & Giambra, S. (2018). Political connections and misallocation of procurement

contracts: evidence from Ecuador. Caracas: CAF.

Bryman, A. (2016). Social Research Methods - 5th Edition. 5th ed. Oxford: Oxford University Press.

Canayaz, M., Martinez, J. and Ozsoylev, H. (2015). Is the revolving door of Washington a back door to

excess corporate returns?. Working paper 1507

Center for Insights in Survey Research (2018). Public Opinion in Romania. [online] International

Republican Institute. Available at:

https://www.iri.org/sites/default/files/final_romania_poll_presentation.pdf [Accessed 25 Dec.

2019].


https://doi.org/10.1111/gove.12033




https://dx-doi-org.focus.lib.kth.se/10.2307/445147

https://www.bbc.com/news/world-europe-34720183

https://www.bbc.com/news/world-europe-48421430

http://www.romaniacurata.ro/mama-lui-ponta-suntem-o-mica-firma-de-familie-nu-avem-nimic-de-ascuns-romania-curata-va-prezinta-principalii-clienti-ai-firmei-familiei-lui-ponta-la-care-au-ajuns-azi-procurorii/

http://www.romaniacurata.ro/mama-lui-ponta-suntem-o-mica-firma-de-familie-nu-avem-nimic-de-ascuns-romania-curata-va-prezinta-principalii-clienti-ai-firmei-familiei-lui-ponta-la-care-au-ajuns-azi-procurorii/

https://doi.org/10.1108/01443581011012234

https://doi.org/10.1108/01443581011012234

https://www-jstor-org.focus.lib.kth.se/stable/10.1017/s002238161300145x

https://www-jstor-org.focus.lib.kth.se/stable/10.1017/s002238161300145x

https://www.vice.com/ro/article/gy3g33/de-ce-in-romania-ne-uram-de-moarte-pe-tema-anticoruptiei

https://www.unlv.edu/sites/default/files/webform/events/files/Ran_Duchin__SeminarPaper.pdf

https://www.unlv.edu/sites/default/files/webform/events/files/Ran_Duchin__SeminarPaper.pdf

https://doi-org.focus.lib.kth.se/10.1080/01900692.2013.879724

https://doi-org.focus.lib.kth.se/10.1080/01900692.2013.879724

http://scioteca.caf.com/handle/123456789/1394

http://scioteca.caf.com/handle/123456789/1394

http://hdl.handle.net/10419/129362


https://www.iri.org/sites/default/files/final_romania_poll_presentation.pdf

62 Daniel Pustan

Centrul pentru Integritate în Business (2013). Mandatele politice. [ebook] Available at:

https://www.businessintegrity.ro/sites/default/files/fileuploads/newsletter/91%20-%20Mandatele%

20functiilor%20publice.pdf [Accessed 22 Dec. 2019].

Charron, N., Dahlström, C., Fazekas, M. and Lapuente, V. (2017). Careers, Connections, and Corruption

Risks: Investigating the Impact of Bureaucratic Meritocracy on Public Procurement Processes. The

Journal of Politics, 79(1), pp.89-104.

Ciubotaru, B. (2013). Organizarea şi funcţionarea Guvernului României: legislație, doctrină și practică

politică. Iaşi: Lumen.

Coviello, D. and Gagliarducci, S. (2017). Tenure in Office and Public Procurement. American Economic

Journal: Economic Policy, 9(3), pp.59-105.

David, Daniel (2015), Psihologia poporului român, Editura Polirom, ISBN 978-973-46-5478-9

Dávid-Barrett, E. and Fazekas, M. (2019). Grand corruption and government change: an analysis of

partisan favoritism in public procurement. European Journal on Criminal Policy and Research.

de Maria, B. (2008). Neo-colonialism through measurement: a critique of the corruption perception

index. Critical perspectives on international business, 4(2/3), pp.184-202.

Di Tella, R. and Schargrodsky, E. (2003). The Role of Wages and Auditing during a Crackdown on

Corruption in the City of Buenos Aires. The Journal of Law and Economics, 46(1), pp.269-292.

Directorate-General for Regional and Urban Policy (European Commission) and PWC (2016).

Stock-taking of administrative capacity, systems and practices across the EU to ensure the compliance

and quality of public procurement involving European Structural and Investment (ESI) Funds.

[online] Luxembourg Publications Office of the European Union, pp.180-188. Available at

https://op.europa.euenpublication-detail-publication60bf8ebc-0204-11e6-b713-01aa75ed71a1langua

ge-en [Accessed 7 Dec. 2019].

Doroftei, I. M. (2016). Measuring Government Favouritism Objectively: The Case of Romanian Public

Construction Sector. European Journal on Criminal Policy and Research, 22(3), pp.399-413.

Doroftei, M., and Dimulescu, V. (2015). Romanian Public Procurement in the Construction Sector:

Corruption Risks and Particularistic Links, Deliverable D8.1.5 ANTICORRP. Bucharest: Romanian

Academic Society.

Escresa, L. and Picci, L. (2017). A New Cross-National Measure of Corruption. The World Bank Economic

Review, 31(1), pp.196–219.

Euractiv (2014). Public procurement corruption in Romania. [online] Euractiv.com. Available at:

https://euractiv.eu/wp-content/uploads/sites/2/special-report/euractiv_special_report_public_pro

curement_corruption_in_romania.pdf [Accessed 2 Dec. 2019].

European Commission (2014). EU Anti-corruption report. Report from the Commission to the Council

and the European Parliament. [online] Brussels: COM(2014) 38 final. Available at:

https://ec.europa.eu/home-affairs/sites/homeaffairs/files/e-library/documents/policies/organized-c

rime-and-human-trafficking/corruption/docs/acr_2014_en.pdf [Accessed 26 Dec. 2019].


https://www.businessintegrity.ro/sites/default/files/fileuploads/newsletter/91%20-%20Mandatele%20functiilor%20publice.pdf

https://doi.org/10.1086/687209

https://doi.org/10.1086/687209

http://doi.org/10.1257/pol.20150426

http://doi.org/10.1257/pol.20150426

https://doi.org/10.1007/s10610-019-09416-4

https://doi.org/10.1007/s10610-019-09416-4

http://doi.org/10.1108/17422040810870079

http://doi.org/10.1108/17422040810870079

http://doi.org/10.1086/345578

http://doi.org/10.1086/345578

https://op.europa.eu/en/publication-detail/-/publication/60bf8ebc-0204-11e6-b713-01aa75ed71a1/language-en

https://op.europa.eu/en/publication-detail/-/publication/60bf8ebc-0204-11e6-b713-01aa75ed71a1/language-en

http://dx.doi.org/10.1007/s10610-016-9319-9

http://dx.doi.org/10.1007/s10610-016-9319-9

http://anticorrp.eu/publications/report-on-romania/

http://anticorrp.eu/publications/report-on-romania/

https://doi.org/10.1093/wber/lhv031

https://euractiv.eu/wp-content/uploads/sites/2/special-report/euractiv_special_report_public_procurement_corruption_in_romania.pdf

https://ec.europa.eu/home-affairs/sites/homeaffairs/files/e-library/documents/policies/organized-crime-and-human-trafficking/corruption/docs/acr_2014_en.pdf


European Commission (2017). On Progress in Romania under the Co-operation and Verification

Mechanism. Report from the Commission to the European Parliament and the Council. [online]

Brussels: Com(2017) 44 final. Available at:

https://ec.europa.eu/info/sites/info/files/com-2017-44_en_1.pdf [Accessed 19 Dec. 2019].

European Commission. (n.d.). Corruption. [online] Available at

https://ec.europa.euhome-affairswhat-we-dopoliciesorganized-crime-and-human-traffickingcorrupti

on_en [Accessed 27 Dec. 2019].

European Parliamentary Research Service (2018). Listing of tax havens by the EU. [online] European

Union. Available at:

https://www.europarl.europa.eu/cmsdata/147404/7%20-%2001%20EPRS-Briefing-621872-Listing-t

ax-havens-by-the-EU-FINAL.PDF [Accessed 27 Dec. 2019].

Fazekas, M. (2017) Red tape, bribery and government favouritism: evidence from Europe. Crime Law Soc

Change (2017) 68: 403.

Fazekas, M. and Kocsis, G. (2017). Uncovering High-Level Corruption: Cross-National Objective

Corruption Risk Indicators Using Public Procurement Data. British Journal of Political Science,

pp.1-10.

Fazekas, M., Cingolani, L. and Tóth, B. (2016), A Comprehensive Review of Objective Corruption Proxies

in Public Procurement: Risky Actors, Transactions, and Vehicles of Rent Extraction.

GTI-WP/2016:03, Budapest: Government Transparency Institute.

Fazekas, M., Tóth, I. and King, L. (2013). Corruption Manual for Beginners: 'Corruption Techniques' in

Public Procurement with Examples from Hungary. SSRN Electronic Journal.

Fazekas, M., Tóth, I. and King, L. (2016). An Objective Corruption Risk Index Using Public Procurement

Data. European Journal on Criminal Policy and Research, 22(3), pp.369-397.

Ferraz, C. and Finan, F. (2008). Exposing Corrupt Politicians: The Effects of Brazil's Publicly Released

Audits on Electoral Outcomes. Quarterly Journal of Economics, 123(2), pp.703-745.

Ferreira Rubio, D. and Andvig, E. (2019). Serious about sustainability? Get serious about corruption.

World Economic Forum. [online] Available at:

https://www.weforum.org/agenda/2019/09/serious-about-sustainability-get-serious-about-corrupti

on/ [Accessed 29 Dec. 2019].

Fisman, D., Fisman, R., Galef, J., Khurana, R. and Wang, Y. (2012). Estimating the Value of Connections

to Vice-President Cheney. The B.E. Journal of Economic Analysis & Policy, 12(3).

France24 (2018). Romania party leader denies corruption in trial over fake jobs. [online] Available at

https://www.france24.comen20180321-romania-party-leader-denies-corruption-trial-over-fake-jobs

[Accessed 29 Dec. 2019].

Gherghina, S., & Volintiru, C. (2017). A new model of clientelism: Political parties, public resources, and

private contributors. European Political Science Review, 9(1), 115-137.

Golden, M. and Picci, L. (2005). Proposal for a new measure of corruption, illustrated with Italian data.

Economics and Politics, 17(1), pp.37-75.


https://ec.europa.eu/info/sites/info/files/com-2017-44_en_1.pdf

https://ec.europa.eu/info/sites/info/files/com-2017-44_en_1.pdf

https://ec.europa.eu/home-affairs/what-we-do/policies/organized-crime-and-human-trafficking/corruption_en

https://www.europarl.europa.eu/cmsdata/147404/7%20-%2001%20EPRS-Briefing-621872-Listing-tax-havens-by-the-EU-FINAL.PDF

https://doi-org.focus.lib.kth.se/10.1007/s10611-017-9694-2

http://dx.doi.org/10.1017/S0007123417000461

http://dx.doi.org/10.1017/S0007123417000461

http://dx.doi.org/10.2139/ssrn.2891017

http://dx.doi.org/10.2139/ssrn.2891017

https://dx.doi.org/10.2139/ssrn.2333354

https://dx.doi.org/10.2139/ssrn.2333354

http://doi.org/10.1007/s10610-016-9308-z

http://doi.org/10.1007/s10610-016-9308-z

https://doi.org/10.1162/qjec.2008.123.2.703

https://doi.org/10.1162/qjec.2008.123.2.703

https://www.weforum.org/agenda/2019/09/serious-about-sustainability-get-serious-about-corruption/

https://doi-org.focus.lib.kth.se/10.1515/1935-1682.3272


https://www.france24.com/en/20180321-romania-party-leader-denies-corruption-trial-over-fake-jobs

https://doi-org.focus.lib.kth.se/10.1017/S1755773915000326



https://doi.org/10.1111/j.1468-0343.2005.00146.x

64 Daniel Pustan

Goldman, E., Rocholl, J. and So, J. (2013). Politically Connected Boards of Directors and The Allocation of

Procurement Contracts. Review of Finance, 17(5), pp.1617-1648.

Graycar, A. (2015) Corruption: Classification and analysis. Policy and Society, 34(2), 87-96.

Gürakar, E. and Bircan, T. (2019). Political Connections and Public Procurement in Turkey. Crony

Capitalism in the Middle East, pp.236-262.

Hall, B. (2019). Romanian government on brink of collapse. Financial Times. [online] Available at:

https://www.ft.com/content/3a22e218-c8d6-11e9-a1f4-3669401ba76f [Accessed 29 Dec. 2019].

Hicken, A. (2011). Clientelism. Annual Review of Political Science, 14(1), pp.289-310.

Hill, R., Griffiths, W. and Lim, G. (2011). Principles of Econometrics. 4th ed. New York: Wiley.

Huntington, S. (1968). Political order in changing societies. New Haven, CT: Yale University Press.

Hyytinen, A., Lundberg, S. and Toivanen, O. (2018). Design of public procurement auctions: evidence

from cleaning contracts. The RAND Journal of Economics, 49(2), pp.398-426.

Islam, R. (2006). Does more transparency go along with better governance?. Economics and Politics,

18(2), pp.121-167.

Jackowicz, K., Kozłowski, Ł. and Mielcarz, P. (2014). Political connections and operational performance of

non-financial firms: New evidence from Poland. Emerging Markets Review, 20, pp.109-135.

Jain, A. K. (2001), Corruption: A Review. Journal of Economic Surveys, 15: 71-121.

Johnston, M. (1996), The search for definitions: the vitality of politics and the issue of corruption.

International Social Science Journal, 48: 321-335.

Kaufmann, D., Kraay, A. and Mastruzzi, M. (2011). The Worldwide Governance Indicators: Methodology

and Analytical Issues. Hague Journal on the Rule of Law, 3(02), pp.220-246.

Kenny, C. (2009). Measuring Corruption in Infrastructure: Evidence from Transition and Developing

Countries. The Journal of Development Studies, 45(3), pp.314-332.

Knack, S. (2007). Measuring Corruption: A Critique of Indicators in Eastern Europe and Central Asia.

Journal of Public Policy, 27(3), pp.255-291.

Kurer, O. (1993). Clientelism, Corruption, and the Allocation of Resources. Public Choice, 77(2),

pp.259-273.

Kurer, O. (2005). Corruption: An Alternative Approach to its Definition and Measurement. Political

Studies, 53(1), pp.222-239.

Kurtz, M. and Schrank, A. (2007). Growth and Governance: Models, Measures, and Mechanisms. The

Journal of Politics, 69(2), pp.538-554.

Lambsdorff, J., (2006). The validity and precision of subjective indicators (CPI). In: Sampford, C.,

Shacklock, A., Connors, C., Galtung, F. (Eds.), Measuring Corruption. Ashgate, Aldershot, pp. 81–99.

Lancaster, T. and Montinola, G. (2001). Comparative political corruption: Issues of operationalization and

measurement. Studies in Comparative International Development, 36(3), pp.3-28.

Leff, N. H. (1964). Economic Development Through Bureaucratic Corruption. American Behavioral

Scientist, 8(3), pp. 8–14.


https://doi-org.focus.lib.kth.se/10.1093/rof/rfs039

https://doi-org.focus.lib.kth.se/10.1093/rof/rfs039

https://doi-org.focus.lib.kth.se/10.1016/j.polsoc.2015.04.001

https://www-oxfordscholarship-com.focus.lib.kth.se/view/10.1093/oso/9780198799870.001.0001/oso-9780198799870-chapter-9

https://www-oxfordscholarship-com.focus.lib.kth.se/view/10.1093/oso/9780198799870.001.0001/oso-9780198799870-chapter-9

https://www.ft.com/content/3a22e218-c8d6-11e9-a1f4-3669401ba76f

http://dx.doi.org/10.1146/annurev.polisci.031908.220508

https://projects.iq.harvard.edu/gov2126/files/huntington_political_order_changing_soc.pdf



https://doi.org/10.1111/j.1468-0343.2006.00166.x

https://doi.org/10.1016/j.ememar.2014.06.005

https://doi.org/10.1016/j.ememar.2014.06.005

http://doi.org/10.1111/1467-6419.00133

http://doi.org/10.1111/1468-2451.00035

https://doi.org/10.1017/S1876404511200046

https://doi.org/10.1017/S1876404511200046

https://doi.org/10.1080/00220380802265066

https://doi.org/10.1080/00220380802265066

http://doi-org/10.1017/S0143814X07000748

http://www.jstor.org/stable/30026836


http://doi.org/10.1111/j.1468-2508.2007.00549.x

https://books.google.se/books?hl=sv&lr=&id=WScfDAAAQBAJ&oi=fnd&pg=PA81&dq=Lambsdorff,+J.,+(2006)&ots=Gw397L7eWm&sig=rqzzdp6Ablfxdxj-DForAI0ehPs&redir_esc=y#v=onepage&q&f=false

http://doi.org/10.1007/BF02686202

http://doi.org/10.1007/BF02686202

http://dx.doi.org/10.1177/000276426400800303


Leys, C. (1965). What is The Problem About Corruption?. The Journal of Modern African Studies, 3(2),

pp.215-230.

Lindstedt, C. and Naurin, D. (2010). Transparency is not Enough: Making Transparency Effective in

Reducing Corruption. International Political Science Review, 31(3), pp.301-322.

Luechinger, S. and Moser, C. (2014). The value of the revolving door: Political appointees and the stock

market. Journal of Public Economics, 119, pp.93-107.

Lui, F. (1985). An Equilibrium Queuing Model of Bribery. Journal of Political Economy, 93(4),

pp.760-781.

Lukkarinen, J., Morild, K. and Jönsson, S. (2015). Facts and figures on public procurement in Sweden.

The Swedish Competition Authority’s report series 2015:9. [online] Stockholm: The Swedish

Competition Authority and the National Agency for Public Procurement. Available at:

http://www.konkurrensverket.se/globalassets/publikationer/rapporter/rapport_2015-9_english.pdf


Marcovici, T. (2019). Firma cumnatului premierului Ponta abonată la contracte cu statul. Cotidianul.


https://www.cotidianul.ro/firma-cumnatului-premierului-ponta-abonata-la-contracte-cu-statul/


Marinas, R. and Dragin, S. (2017). Tens of thousands rally in Romania against planned emergency

decrees. Reuters. [online] Available at:

https://www.reuters.com/article/us-romania-protests/tens-of-thousands-rally-in-romania-against-p

lanned-emergency-decrees-idUSKBN15D118 [Accessed 19 Dec. 2019].

Markandya, A. and Pearce, D. (1988). Natural environments and the social rate of discount. Project

Appraisal, 3(1), pp.2-12.

Mauro, P. (1995). Corruption and Growth. The Quarterly Journal of Economics, 110(3), pp.681-712.

Mediafax (2009). Guvernul Boc 2 a fost demis. [online] Available at:

https://www.mediafax.ro/politic/guvernul-boc-2-a-fost-demis-video-4985350 [Accessed 27 Dec.

2019].

Méon, P. and Sekkat, K. (2005). Does corruption grease or sand the wheels of growth? Public Choice,

122(1-2), pp.69-97.

Mihalache, D., (2013) Achizitii publice 2007-2015 (Contracte). Romanian Digital Agenda Agency. [data]

Available at: http://data.gov.ro/dataset/achizitii-publice-2007-2016-contracte6 [Accessed 17 Dec.

2019]

Ministry of Finance (2018). Date de identificare plătitori actualizate august 2018. [data] Available at:

http://data.gov.ro/dataset/informatii_fiscale_august_2018 [Accessed 17 Dec. 2019]

Mironov, M. and Zhuravskaya, E. (2012). Corruption in procurement and shadow campaign financing:

Evidence from Russia. MIMEO.

Mo, P. (2001). Corruption and Economic Growth. Journal of Comparative Economics, 29(1), pp.66-79.

Mungiu, A. (2006). Corruption: Diagnosis and Treatment. Journal of Democracy, 17(3), pp.86-99.


http://dx.doi.org/10.1017/S0022278X00023636

https://doi.org/10.1177/0192512110377602

https://doi.org/10.1177/0192512110377602

https://doi.org/10.1016/j.jpubeco.2014.08.001


https://dx-doi-org.focus.lib.kth.se/10.1086/261329

http://www.konkurrensverket.se/globalassets/publikationer/rapporter/rapport_2015-9_english.pdf

https://www.cotidianul.ro/firma-cumnatului-premierului-ponta-abonata-la-contracte-cu-statul/

https://www.reuters.com/article/us-romania-protests/tens-of-thousands-rally-in-romania-against-planned-emergency-decrees-idUSKBN15D118

https://www.reuters.com/article/us-romania-protests/tens-of-thousands-rally-in-romania-against-planned-emergency-decrees-idUSKBN15D118

https://doi.org/10.1080/02688867.1988.9726647

http://dx.doi.org/10.2307/2946696

https://www.mediafax.ro/politic/guvernul-boc-2-a-fost-demis-video-4985350

http://dx.doi.org/10.1007/s11127-005-3988-0

http://data.gov.ro/dataset/achizitii-publice-2007-2016-contracte6

http://data.gov.ro/dataset/informatii_fiscale_august_2018

http://chaire-eppp.org/files_chaire/ssrn-id1946806.pdf

http://chaire-eppp.org/files_chaire/ssrn-id1946806.pdf

https://doi-org.focus.lib.kth.se/10.1006/jcec.2000.1703

http://doi.org/10.1353/jod.2006.0050

66 Daniel Pustan

Mungiu-Pippidi, Alina (editor) 2015, Government Favouritism in Europe: Anticorruption report 3,

Barbara Budrich Publishers

National Anticorruption Directorate (2017). Raport privind activitatea desfășurată. [online] Bucharest.

Available at: https://www.pna.ro/bilant_activitate.xhtml?id=40 [Accessed 19 Dec. 2019].

OECD (2016). Preventing Corruption in Public Procurement. [online] Secretary-General of the OECD.

Available at: http://www.oecd.org/gov/ethics/Corruption-Public-Procurement-Brochure.pdf


Olken, B. (2006). Corruption and the costs of redistribution: Micro evidence from Indonesia. Journal of

Public Economics, 90(4-5), pp.853-870.

Olken, B. (2007). Monitoring Corruption: Evidence from a Field Experiment in Indonesia. Journal of

Political Economy, 115(2), pp.200-249.

Olken, B. (2009). Corruption perceptions vs. corruption reality. Journal of Public Economics, 93(7-8),

pp.950-964.

Oman, C. and C. Arndt (2006), Uses and Abuses of Governance Indicators, Development Centre Studies,

OECD Publishing, Paris.

Oprea, A. (2017). Registrul Comerţului: În România există 1,4 milioane de firme, din care doar 1,2

milioane sunt active. Cele mai multe insolvenţe sau radieri se înregistrează în comerţ, construcţii şi

transporturi. News.ro. [online] Available at:

https://www.news.ro/economic/registrul-comertului-in-romania-exista-1-4-milioane-de-firme-din-c

are-doar-1-2-milioane-sunt-active-cele-mai-multe-insolvente-sau-radieri-se-inregistreaza-in-comert-

constructii-si-transporturi-1922400517302017051116986291 [Accessed 27 Dec. 2019].

Politica Românească (2011). Guvernul Emil Boc (2). [online] Available at:

http://www.politicaromaneasca.ro/guvernul_emil_boc_2-2638 [Accessed 22 Dec. 2019].

Reinikka, R. and Svensson, J. (2004). Local Capture: Evidence from a Central Government Transfer

Program in Uganda. The Quarterly Journal of Economics, 119(2), pp.679-705.

Romanian Agency for Digital Agenda (2018). Lansarea noului Sistem Electronic de Achizitii Publice.


https://www.aadr.ro/-lansarea-noului-sistem-electronic-de-achizitii-publice_227_0.html [Accessed

7 Dec. 2019].

Rosca, A. (2018). Daniel David despre psihologia poporului român. Ce ne definește, ca nație, și cum ne

putem schimba. [Blog] Andreea Rosca Playground. Available at

https://www.andreearosca.ro/daniel-david [Accessed 1 Dec. 2019].

Rose-Ackerman, R. (1997). The political economy of corruption. In: Elliott, K.A. (Ed.), Corruption and the

Global Economy, Institute for International Economics, Washington, DC, pp. 31–60.

Rose-Ackerman S. (2008) Corruption. In: Readings in Public Choice and Constitutional Political

Economy. Springer, Boston, MA.

Rothstein, B. and Teorell, J. (2008), What Is Quality of Government? A Theory of Impartial Government

Institutions. Governance, 21: 165-190.


https://www.pna.ro/bilant_activitate.xhtml?id=40

http://www.oecd.org/gov/ethics/Corruption-Public-Procurement-Brochure.pdf


http://doi.org/10.1086/517935


https://doi.org/10.1787/9789264026865-en

https://www.news.ro/economic/registrul-comertului-in-romania-exista-1-4-milioane-de-firme-din-care-doar-1-2-milioane-sunt-active-cele-mai-multe-insolvente-sau-radieri-se-inregistreaza-in-comert-constructii-si-transporturi-1922400517302017051116986291



http://www.politicaromaneasca.ro/guvernul_emil_boc_2-2638

https://doi.org/10.1162/0033553041382120

https://doi.org/10.1162/0033553041382120

https://www.aadr.ro/-lansarea-noului-sistem-electronic-de-achizitii-publice_227_0.html

https://www.andreearosca.ro/daniel-david/

https://www.andreearosca.ro/daniel-david/

https://play.google.com/books/reader?id=7qyxDQAAQBAJ&hl=en&pg=GBS.PA3

https://doi.org/10.1007/978-0-387-75870-1_30

http://doi.org/10.1111/j.1468-0491.2008.00391.x

http://doi.org/10.1111/j.1468-0491.2008.00391.x


Stephenson, M. (2014). Can Foreign Anti-Bribery Enforcement Statistics Help Us Measure Corruption

Levels Objectively?. [Blog] The Global Anticorruption Blog. Available at:

https://globalanticorruptionblog.com/2014/04/15/can-foreign-anti-bribery-enforcement-statistics-h

elp-us-measure-corruption-levels-objectively/ [Accessed 29 Nov. 2019].

Stokes, S. (2005). Perverse Accountability: A Formal Model of Machine Politics with Evidence from

Argentina. American Political Science Review, 99(3), pp.315-325.

Svensson, J. (2005). Eight Questions about Corruption. Journal of Economic Perspectives, 19(3),

pp.19-42.

Schwab, K. (2016). The Global Competitiveness Report 2016–2017. [online] Geneva: the World Economic

Forum. Available at:

http://www3.weforum.org/docs/GCR2016-2017/05FullReport/TheGlobalCompetitivenessReport201

6-2017_FINAL.pdf [Accessed 29 Dec. 2019].

Søreide, T. (2002) Corruption in public procurement: Causes, consequences and cures, Chr. Michelsen

Institute Development Studies and Human Rights

Tanzi V., Davoodi H. (1998) Corruption, Public Investment, and Growth. In: Shibata H., Ihori T. (eds) The

Welfare State, Public Investment, and Growth. Springer, Tokyo

Timu, A. and Zega, R. (2017). Romanian Premier Refuses to Quit as Protests Grow Despite U-Turn.

Bloomberg. [online] Available at:

https://www.bloomberg.com/news/articles/2017-02-04/romanian-government-to-scrap-anti-graft-d

ecree-amid-protests [Accessed 17 Dec. 2019].

Titl, V. and Geys, B. (2019). Political donations and the allocation of public procurement contracts.

European Economic Review, 111, pp.443-458.

Toma, V. (2016). CCR despre abuzul in serviciu: "în mod defectuos" inseamna "incălcare a legii." Revista

22. [online] Available at:

https://revista22.ro/actualitate-interna/ccr-despre-abuzul-in-serviciu-n-mod-defectuos-inseamna-in

clcare-a-legii [Accessed 20 Dec. 2019].

Tran, M. (2015). Bucharest nightclub fire: PM and government resign after protests. The Guardian.


https://www.occrp.org/en/daily/4567-romania-prime-minister-ponta-resigns-amid-protests-over-ni

ghtclub-fire [Accessed 22 Dec. 2019].

Weber Abramo, C. (2008). How Much Do Perceptions of Corruption Really Tell Us?. Economics: The

Open-Access, Open-Assessment E-Journal, 2(2008-3), p.1.

Werner, S. (1983). New Directions in the Study of Administrative Corruption. Public Administration

Review, 43(2), p.146-154.

Xu, X. (2016). Corruption and economic growth: an absolute obstacle or some efficient grease?. Economic

and Political Studies, 4(1), pp.85-100.


https://globalanticorruptionblog.com/2014/04/15/can-foreign-anti-bribery-enforcement-statistics-help-us-measure-corruption-levels-objectively/

https://globalanticorruptionblog.com/2014/04/15/can-foreign-anti-bribery-enforcement-statistics-help-us-measure-corruption-levels-objectively/

https://doi.org/10.1017/S0003055405051683

https://doi.org/10.1017/S0003055405051683

http://doi.org/10.1257/089533005774357860

http://www3.weforum.org/docs/GCR2016-2017/05FullReport/TheGlobalCompetitivenessReport2016-2017_FINAL.pdf


http://doi.org/10.1007/978-4-431-67939-4_4

https://www.bloomberg.com/news/articles/2017-02-04/romanian-government-to-scrap-anti-graft-decree-amid-protests

https://doi-org.focus.lib.kth.se/10.1016/j.euroecorev.2018.11.004

https://revista22.ro/actualitate-interna/ccr-despre-abuzul-in-serviciu-n-mod-defectuos-inseamna-inclcare-a-legii

https://www.theguardian.com/world/2015/nov/04/romanian-government-resigns-nightclub-fire-victor-ponta

https://doi.org/10.5018/economics-ejournal.ja.2008-3

http://dx.doi.org/10.2307/975428

http://doi.org/10.1080/20954816.2016.1152097

68 Daniel Pustan

Annexes

Annex A

Table 9 Summary of selected studies using objective indicators of corruption

Source: Fazekas, Tóth and King 2016, p. 2

paper indicator used Country year sector

Auriol, Straub and

Flochel (2016)

Exceptional procedure type Paraguay 2004-200

7

general

procurement

Bandiera, Prat,

and Valletti (2009)

Price differentials for standard goods

purchased locally or through a

national procurement agency

Italy 2000-200

5

various

standardized

goods (e.g.

paper)

Coviello and

Gagliarducci

(2017)

Number of bidders

Same firm awarded contracts

recurrently

Level of competition

Italy 2000-200

5

general

procurement

Di Tella and

Schargrodsky

(2003)

Difference in prices of standardized

products such as ethyl alcohol

Brazil 1996-1997 health care

Ferraz and Finan

(2008)

Corruption uncovered by federal

audits of local government finances

Brazil 2003 federal-local

transfers

Golden and Picci

(2005)

Ratio of physical stock of

infrastructure to cumulative spending

on infrastructure

Italy 1997 infrastructure

Goldman, Rocholl

and So (2013)

Political office holders' position on

company boards

USA 1990-2004 general

procurement

Hyytinen,

Lundberg and

Toivanen (2018)

Number and type of invited firms

Use of restricted procedure

Sweden 1990-1998 cleaning

services

Olken (2006) Difference between the quantity of

in-kind benefits (rice) received

according to official records and

reported survey evidence

Indonesia 1998-1999 welfare

spending

Olken (2007) Differences between the officially

reported and independently audited

prices and quantities of road

construction projects

Indonesia 2003-200

4

infrastructure

(roads)

Klasnja (2015) Single bidder procedures

Non-open procedure types

Romania 2008-2012 general

procurement

Reinikka and

Svensson (2004)

Difference between block grants

received by schools according to

official records and user survey

Uganda 1991-1995 education



Annex B

Table 10 Corruption techniques and indicators

Source: Fazekas, Tóth and King 2013, p. 63

No. Name Direct indicator Indirect indicator

1 Defining

unnecessary

needs

- -

2 Defining needs to

benefit a

particular

supplier

A) Prevalence of avoiding centralised

procurement

3 Tinkering with

thresholds and

exceptions

A) Proportion of non-open

procedures

D) Contract value according to Public

Procurement Law/total procurement

contract value

B) Average corruption risk score of

procedures followed

C) Frequency of actual contract value

above estimated contract value

4 Tailoring

eligibility criteria

A) Length of eligibility criteria -

5 Abusing formal

and

administrative

requirements

A) Length of eligibility criteria -

B) Proportion of excluded bids

6 Tailoring

evaluation criteria

A) Length of evaluation criteria -

B) Weight of non-price criteria

7 Using long term

complex contracts

A) Combined value of framework

contracts and PPPs / total contract

value

-

B) Average contract duration

8 Tinkering with

submission period

A) Proportion of tenders with

accelerated submission periods

-

B) Proportion of tenders with

extremely short submission periods

C) Average contract value per

weekday available for submission

9 Selective

information

provision

- A) Proportion of tenders with extremely

short submission periods

B) Proportion of procedures with call for

tenders modified within all procedures

10 Avoiding

publication of call

A) Proportion of tenders without call

for tenders in the official journal

-


70 Daniel Pustan

for tenders

11 Strategically

modifying call for

tenders

A) Proportion of procedures with call

for tenders modified within all

procedures

-

12 Excessively pricey

documents,

difficult access to

documents

A) Price of documentation/estimated

contract value

-

13 Deliberate errors

in document

publication

A) Prevalence of extremely erroneous

contract award announcements

-

B) Hiding or erroneously reporting

the final contract value

14 Strategically

annulling

procedures

A) Proportion of annulled procedures

re-launched subsequently

B) Decrease in the number of bids

received in subsequent rounds

15 Repeated

violations of

public

procurement rules

A) Repeated court rulings against the

issuer within the same procedure

-

16 Unfair scoring - A) Average contract value per weekday

available for decision

B) Length of evaluation criteria

C) Weight of non-price criteria

17 Abusing unit

prices in the

contract

A) Proportion of contracts using unit

prices

-

18 Modifying

contracts

strategically

A) Proportion of modified contracts -

B) Difference between the awarded

and final contract value

C) Difference between originally

planned and final completion period

19 Abusing add-on

contracts

A) Proportion of add-on contracts -

B) Proportion of contracts exhausting

the planned reserves

20 Performance

violating contract

- -



Annex C

Table 11 Summary of selected studies on firm’s political connections

Paper Country Sample

period

Political power

level

Proxy for political

connection

Public procurement

Mironov and

Zhuravskayay (2012)

Russia 6 years regional elections illegal donations to parties

Goldman, Rocholl and So

(2013)

USA 1990-2004 House and Senate

(Parliamentary)

revolving door

Bromberg (2014) USA 2001-2006 Federal

government /

House of

Representative

donation

Straub (2014) Paraguay 2004-2011 government /

parliamentary

*member or sympathisant

Boas, Hidalgo and

Richardson (2014)

Brasil 2004-2010 Chamber of

Deputies

donation

Luechinger and Moser

(2014)

USA 1993-2003 Department of

defense

revolving door

Brogaard, Denes and

Duchin (2015)

USA 2000-2012 federal

government

donation

Doroftei(2016) Romania 2007-2013 donation + connections

Doroftei and Dimulescu

(2015)

Romania 2007-2013 donation + connections

Auriol, Straub and Flochel

(2016)

Paraguay 2004-2007

Titl and Geys (2019) Czech

Republic

2007-2014 13 regional

governments

donation

Arvate, Barbosa and

Fuzitani (2018)

Brasil 2007-2010 state legislatures donation

Brugués F., Brugués J. and

Giambra S. (2018)

Ecuador 2006-2018 bureaucrat (incl.

mayors and local

counselors)

revolving door

Johnson and Roer (2018) USA 2007-2016 House

Appropriations

subcommittee

"relocation" and "selection"

influence

Ferris, Houston and

Javakhadze (2019)

USA 2006-2013 donation

Gürakar and Bircan (2019) Turkey 2004-2011 Parliament and

local

governmental

bodies

revolving door


72 Daniel Pustan

Stock market

Roberts (1990) USA 1983-09-01 congressional

seniority

donation

Fisman (2001) Indonesia 1995-1997 President Suharto Suharto Dependency Index

Faccio (2006) 47

countries

1991 see the proxy for

political

connection

revolving door

Jayachandran (2006) USA 2001-05 donation

Cooper, Gulen and

Ovtchinnikov (2010)

USA 1979-2004 House and Senate donation

Fisman et al. (2012) USA 2000-01-03 -

2001-04-30

revolving door

Amore and Bennedsen

(2013)

Denmark 2002-2008 local municipality family ties

Gropper, Jahera Jr. and

Park (2013)

USA 1989-2010 Congress banks' headquartered in the

same state with the

committee chair

Jackowicz, Kozłowski and

Mielcarz (2014)

Poland 2001-2011 central

government /

national

parliament / local

authorities

revolving door

Coulomb and Sangnier

(2014)

France 2007-01-01 -

2007-05-06

presidential Sarkozy Index (media)

Luechinger and Moser

(2014)

USA 1993-2003 Department of

defense

revolving door

Akey (2015) USA 1998-2010 Senators and

Representatives

donation

Canayaz, Martinez and

Ozsoylev (2015)

USA 1990-2012 government

positions

revolving door

Financing

Khawaja and Mian (2005) Pakistan 1996-2002 national elections if firm's director participates

in an election

Fraser, Zhang and

Derashid (2006)

Malaysia 1990-1999 government political patronage +

informal ties

Faccio, Masulis and

McConnell (2006)

35

countries

1997-2002 revolving door+friendship

Claessens, Feijend and

Laeven (2008) Brasil 1998 & 2002

federal deputy

candidates

donations - contribution to

campaigns

Bliss and Gul (2012) Malaysia 2001-2004 government informal ties



Blau, Brough and Thomas

(2013)

USA 2008 federal

government

revolving door + lobbying

expenditures

Yeh, Shu and Chiu (2013) Taiwan 1998-2006 revolving door

Infante and Piazza (2014) Italy 2005-09 -

2009-03

national

parliament and all

local councils

revolving door

Houston, Maslar, and

Pukthuanthong (2018)

USA 2004-2012 White House,

Senate and

Congress

donations

Disaster relief

Nikolova and Marinov

(2017)

Bulgaria 2004-2005

Annex D

Table 12 Descriptive statistics

year mean sd min max p50 sum

2009 97276.44 1777334 0 2.79e+08 3751.26 1.19e+10

2010 87914.86 1308800 0 1.78e+08 3610.09 1.09e+10

2011 157587.2 2387378 0 2.61e+08 6346.42 1.66e+10

2012 141756.1 2609191 0 3.07e+08 5766.765 1.53e+10

2013 215950.2 3063701 0 2.91e+08 5611.643 1.44e+10

2014 239164.9 2658020 0 2.30e+08 6515.018 1.76e+10

2015 247728.3 3747858 0 7.76e+08 7246.912 1.73e+10

Total 155311.2 2463122 0 7.76e+08 5093.455 1.04e+11


Documents

Government favoritism in public procurement: evidence from ...1430016/...Daniel Pustan Godkänt 2019-12-31 Examinator Matti Kaulio Handledare Kristina Nyström Uppdragsgivare Kontaktperson