Determinants underlying incidence and amount of bribery, an evidence from manufacturing firms in vietnam

Using a panel data set from the Small and Medium Enterprise SME Surveys from 2005 -2013, this study investigates the factors which influence the incidence of bribery and thesize of bribe

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS UNDERLYING INCIDENCE AND AMOUNT OF BRIBERY: AN EVIDENCE

FROM MANUFACTURING FIRMS IN

VIETNAM

BY

VU THI THUONG

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

HO CHI MINH CITY, November 2015

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES

VIETNAM - NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

DETERMINANTS UNDERLYING INCIDENCE AND AMOUNT OF BRIBERY: AN EVIDENCE

FROM MANUFACTURING FIRMS IN

VIETNAM

A thesis submitted in partial fulfilment of the requirements for the degree of

MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By

VU THI THUONG

Academic Supervisor:

DR LE VAN CHON

me with essential skills for doing research, especially the skills of using econometric tools.Besides, I truly appreciate the time he devotes to this thesis to carefully review my finalthesis draft and help me rearrange thesis structure reasonably as well as correct errors andinappropriate words usages It must take me a longer time to complete this project withouthis invaluable sources of help.

From my sincere heart, I would like to thank all lecturers in VNP who equipped me withuseful knowledge to conduct this thesis as well as to serve my job in the future Especially,

I would like to express my special thanks to Prof Nguyen Trong Hoai, Dr Pham KhanhNam and Dr Truong Dang Thuy who always travel with me during the two years of myMaster program, share their practical profound insights and inspire the love of scientificresearch

Next, I would also like to thank all my friends here at VNP, especially Vu Thi Khanh, LeHuu Nhat Quang for their enthusiasm of sharing data processing techniques with me; BuiMinh Khoi, Nguyen Huong Nguyen for their spiritual encouragement during thesestressful and difficult moment I also want to say thanks to VNP officers for their restlessassistance

I would like to express my special thanks to Da Nang University Campus in Kon Tum whohas given me advantages to accomplish my Master program I also want to thank mycolleague – Nguyen Thi Phuong Thao for her great support me during my writing of thisthesis

I am deeply and forever indebted to my parents and my sister for their love, support andencouragement throughout my entire life

Ordinary Least Squares

Provincial Competitiveness Index

Small and medium enterprises

Transparency International

The World Bank

United Nations Development Programme

Using a panel data set from the Small and Medium Enterprise (SME) Surveys from 2005

-2013, this study investigates the factors which influence the incidence of bribery and thesize of bribe payment among formal and informal firms in Vietnam Due to censored nature

of the data on bribes and sample selection bias, this paper applies Heckman two - stepprocedure which was proposed by Heckman (1979) to correct these problems This studyfinds strong evidence that the propensity to bribe as well as the variation in the amount ofbribe is highly positively correlated with the interaction level with public officials, thefirm‟s ability to pay and the burden of regulation that firms face In this paper, theinteraction with public officials is looked at different types of interaction Besides, thisstudy also points out that company without official business registration licenses are morelikely to avoid paying the informal costs These results are robust when the lagged values

of profit are used as instruments for profit

Keywords: firms, bribery, Vietnam

Table of contents

Chapter 1: Introduction 1

1.1 Problem statement 1

1.2 Research objectives 3

1.3 Research questions 3

1.4 The scope of the study 3

1.5 The structure of the study 4

Chapter 2: Literature review 5

2.1 Corruption 5

2.1.1 The definition of corruption 5

2.1.2 The Forms of Corruption 6

2.1.3 Measuring corruption 7

2.2 Empirical studies 11

2.2.1 Factors influencing the propensity to bribe 11

2.2.2 How much must graft-paying firms pay? 16

2.3 Basic framework to estimate the incidence and level of bribe 16

2.4 Conceptual framework 18

2.5 Chapter summary 18

Chapter 3: Corruption in Vietnam 20

Chapter 4: Data and Econometric Model 26

4.1 Data 26

4.1.1 Data source 26

4.1.2 Data description 27

4.2 Descriptive statistical analyses 28

4.3 Econometric model 33

Chapter 5: Empirical results 37

5.1 Factors influencing the propensity to bribe 39

5.2 Factors influencing the size of bribe payment 42

5.3 Robustness 43

Chapter 6: Conclusion 47

6.1 Main findings 47

6.2 Policy implications 47

6.3 Limitations and further research 48

6.3.1 Limitations 48

6.3.2 Directions for further studies 49

References 49

Appendices 55

List of tables

Table 2 1: Definitions of corrupt activities 6

Table 2 2: Summary of features of measures of corruption 9

Table 3 1: Vietnam‟s annual CPI result 21

Table 3 2: Average cost of bribes paid, by sector 23

Table 4 1: Panel data structure 26

Table 4 2: Data definition 27

Table 4 3: The descriptive statistics of the size of bribe payment by location 28

Table 4 4: The descriptive statistics of the size of bribe payment by the legal ownership form and sector 29

Table 4 5: The purpose of bribe payment 30

Table 4 6: The summary statistics of key variables 30

Table 4 7: Pairwise Correlation 32

Table 4 8: The Expected Variables in Heckman two step 36

Table 5 1: Heckman two-step regression analyses the incidence of bribery and bribe amount 38

Table 5 2: Heckman two-step regression analyses the incidence of bribery and bribe amount, using instrument variable 44

List of figures Figure 3 1 The most serious Economic & Social Issues for Vietnam 20

Figure 3 2 Perceptions of the prevalence of corruption across sectors 21

Figure 3 3 Key Indicators of Informal Charges (2006 to 2014) 22

Figure 3 4 The purpose of bribe payment 23

Figure 3 5 Awareness about the government's anti-corruption efforts 24

Figure 3 6 Willingness to report an incident of corruption (Southeast Asia) 25

Chapter 1: Introduction 1.1 Problem statement

Corruption occurs in both developed and developing countries with various degreesand has impacted on almost all parts of society (Lawal, 2007; Rohwer, 2009) Amundsen(1999) described corruption as “a disease, a cancer that eats into the cultural, political andeconomic fabric of society, and destroys the functioning of vital organs” Besides, WorldBank (2009) identified corruption as “among the greatest obstacles to economic and socialdevelopment”

In Vietnam, this problem is seriously alarmed for the government‟s failure toreduce corruption over past years In particular, the Corruption Perceptions Index (CPI)

2014, conducted by Transparency International ranked Vietnam at 119 out of 175 in globaland at 18 out of 28 in the Asia Pacific region It is worthy to note that Vietnam‟s CPI score

is unchanged in three consecutive years from 2012 to 2014 whereas a positive change wasrecorded in neighboring countries Besides, according to a report of ProvincialCompetitiveness Index (PCI) 2014, Vietnam witnessed a significant decline in controllinginformal cost that firms have to pay for running a business smoothly More specially, theproportion of companies paying bribe jumped from 41% in 2013 to 66% in 2014 10% offirms showed that more than 10% of their revenue spent on bribes These pieces ofevidence reflect the seriousness of corruption in Vietnam despite the effort of reducing itover the past years

A massive body of academic research on corruption in the past decade divided intotwo segments: the determinants of corruption and the effect of corruption on growth1.Regarding the determinants of corruption, some typical studies include Ades and Di Tella(1997, 1999), Svensson (2000), Persson, Tabellini, and Trebbi (2003) Related to the effect

of corruption, a list of study can be seen Mauro (1995), Wei (1997) and Johnson et al.(1997) Except for few studies, most of preceding studies share three common features.First, these studies have primarily based on cross-country analyses Second, they useperceptive data rather than quantitative data due to the higher cost of collecting the later

1

For a review and summary, refer to Bardhan (1997), Jain (2001), Reinikka, Svensson (2002) and Aidt (2003)

one Third, the interpretation of corruption was based on a function of macro-factors such

as countries‟ policy or institutional environment

This branch of the literature with above-mentioned characteristics such asKaufmann and Wei (1998) and Hellman et al (2000) Kaufmann and Wei (1998) examinedthe nexus between informal cost in the form of time spent by managers with bureaucratsand cost of capital To do this, they employed the firm-level data from three differentsurveys covered around 48 - 73 countries In this study, information about corruption isbased on perception indices constructed from questions related to country characteristics.Also, Hellman et al (2000) used a sample of 3300 firms in 20 countries that were obtainedjointly by the World Bank and the Office of the Chief Economist in 1999 They explainedthe validation of corruption based on the function of the political-institutional environment,including protection of property rights and civil liberties

These existing pieces of literature have provided important knowledge about theaggregate determinants of corruption However, the interpretation based on cross-countryanalyzes as well as using perceptive data caused some drawbacks Firstly, the exploitingperception data on corruption can suffer from bias In particular, certain studies suggestthat small firms are more likely to perceive their environment to be more corrupt than theircounterparts (Batra, Kaufmann, & Stone, 2003; Bennedsen et al., 2009) Similarly, moreproductive companies tend to complain about their business environment more than lessproductive companies do (Malomo, 2013) Secondly, macro- determinant of corruptionconstrains the interpretation of variation in corruption within the country In the otherword, country-level research cannot explain the variation of the level of corruption acrossfirms within a country (Svensson, 2002)

To avoid these problems, Svensson (2003) exploited the quantitative data oncorruption, derived from the Uganda enterprise survey in 1998 which was designed torepresent the population of private enterprises, operating in manufacturing and processingsectors However, there are two striking features in such data set Firstly, not all firms in thesample reported that they need to pay the informal cost Secondly, there has a significantvariation in the size of bribe payment across firms under a similar policy To explain suchvariance, this paper developed a simple bargaining model, in which firms can

choose to either pay the bribe or exit the market when facing a request for a bribe of apublic official.

In similarity to Svensson (2003), this paper investigates the corruption in Vietnamfor the period from 2005 to 2013 Based on the quantitative information on corruption andthe detailed financial information from Small Medium Enterprise (SMEs) surveys in thisperiod, this study attempts to answer two questions: who must pay a bribe and how muchthey pay To answer these questions, this paper primarily uses control right hypothesis andbargaining hypothesis In which, control hypothesis suggests that the interaction level withpublic official will determine whether or not a firm has to pay a bribe The bargaininghypothesis indicates that firm‟s bargaining power which measure by sunk cost componentand ability to pay will interpret how much a firm pays

1.2 Research objectives

This study is to find out the factors that influence the incidence of bribery and thesize of bribe payment of firms in Vietnam In other words, this study has two followingobjectives:

(i) Specify the factors that affect the propensity to pay a bribe of formal and informal firms in Vietnam

(ii) Specify the factors that influence the variation in bribe amount across bribe-reporting firms in Vietnam

(iii) Give some policy implications in order to reduce corruption in Vietnam

1.3 Research questions

This thesis aims to answer the following questions:

(i) Do the factors related to firm characteristics such as firm size, informal status, profit, firm‟s choice of technology, etc have influence the incidence of bribery?

(ii) Why do some firms have to pay more bribes than others?

(iii) What should be done to reduce corruption in Vietnam?

1.4 The scope of the study

The study uses the available data of SMEs surveys in Vietnam These surveys wereconducted in the period of 2005 to 2013, covering around ten provinces and cities inVietnam Except for foreign-owned enterprises, the surveyed samples include non-state

manufacturing enterprises with and without legal business registration licenses The former

is called formal firms while the latter is named informal firms Although the surveys covermore than twenty manufacturing industries, this thesis will focus on 14 largest sectors interms of number of enterprises, including food products and beverages, fabricated metalproducts, wood and wood products, furniture manufacturing, tanning and dressing leathertextiles, wearing apparel, paper and paper products, publishing - printing, chemicalproducts, rubber and plastic products, non-metallic mineral products, other machinery andequipment, water treatment

1.5 The structure of the study

This paper consists of six chapters Following chapter 1 of introduction, chapter 2describes a general picture about corruption, including definition, forms and themeasurement of corruption This chapter also presents the empirical studies and theoreticalmodels that examine the incidence and level of bribe Chapter 3 gives a brief overview ofcorruption in Vietnam Chapter 4 outlines the data and presents the descriptive statisticalanalyses In addition, the econometric models, as well as the variables, included in thesemodels also presented in this chapter Then, the econometric results are discussed inchapter 5 Finally, the main findings, limitations, policy implications and suggestions forfurther researches are presented in chapter 6

Chapter 2: Literature review 2.1 Corruption

2.1.1 The definition of corruption

The term of “corruption” originates from the Latin word “corruption”, meaning

“moral decay, wicked behavior, putridity or rottenness” (Milic, 2001) However, it isdifficult to give a globally accepted definition because corruption is a complex social,legal, economic and political phenomenon (Rohwer, 2009)

In reality, there are many different definitions of corruption The most popular andsimplest definition is given by World Bank (1997), defining corruption as “the abuse ofpublic power for private benefit” Corruption is also widely understood as „„the acts inwhich the power of public office is used for personal gain in a manner that contravenes therules of the game‟‟ (Jain, 2001) Transparency International, 2009 defined that corruption

as „„misuse of entrusted power for private gain‟‟ This definition is similar with the WorldBank‟s definition; however it can cover the corruption activities in both public and privatesector Besides, the Oxford Advanced Learner's Dictionary (2000, p 281) suggested thatcorruption is the association of two important elements: authority and morality Inparticular, corruption is described as “the dishonest or illegal behavior, especially of people

in authority” and “the act or effect of making somebody change from moral to immoralstandards of behavior” The Vietnamese Law on anti-corruption, 2005 defined corruption

as “the acts committed by persons with positions and/or powers of abusing such positionsand/or powers for self-seeking interests”

In the aspect of economics field, (Klitgaard, 1988) explained corruption through anequation: Corruption = Monopoly Power + Discretion – Accountability Besides, UnitedNations Development Program (2004) uses another equation to define corruption, asfollows: corruption= (Monopoly Power + Discretion) – (Accountability + Integrity +Transparency)

Awartani, H (2009) suggests that corruption activities are conducted by two partiesconsisting of demand-side supply-side Demand-side corruption involves corrupt publicofficials who asked individuals or companies to pay a bribe to receive public services or

implement regulations smoothly Supply-side corruption is individuals or firms who corrupt officials because of their benefit.

In general, corruption in the public sector is now widely accepted as “an illegal actthat involves the abuse of a public trust or office for some private benefit‟‟, or “themisuse of public office for private gain.” (Fantaye, 2004)

2.1.2 The Forms of Corruption

As the broad definition suggests, corruption can occur in various forms Inparticular, Transparency International (TI, 2009) shows some popular forms of corruptionsuch as bribery, embezzlement, fraud, extortion, cronyism, nepotism, patronage and graft.The definitions of these most common forms are given in table 1 as follows:

Table 2 1: Definitions of corrupt activities

Bribery The payment (in money or kind) that is given or taken in a corrupt

relationship Equivalent terms to bribery include, for example, kickbacks,commercial arrangements or pay-offs These are all notions of corruption

in terms of the money or favours paid to employees in private enterprises,public officials and politicians.They are payments or returns needed ordemanded to make things pass more swiftly, smoothly or morefavourably through state or government bureaucracies

Collusion A secret agreement between parties, in the public and/or private sector, to

conspire to commit actions aimed to deceive or commit fraud with theobjective of illicit financial gain The parties involved often are referred

Fraud The act of intentionally deceiving someone in order to gain an unfair or

illegal advantage (financial, political or otherwise)

Gifts and Gifts and hospitality (e.g vacations, luxury dinner, etc.) that could affectHospitality or be perceived to affect the outcome of business transactions and are not

reasonable and bona fide

Lobbying Any activity carried out to influence a government or institution„s

policies and decisions in favour of a specific cause or outcome Evenwhen allowed by law, these acts can become distortive ifdisproportionate levels of influence exist – by companies, interest groups,associations, organisations and individuals

Revolving An individual who moves back and forth between public office andDoor private companies, exploiting his/her period of government service for

the benefit of the companies they used to regulate

Trading in The situation where a person is selling his/her influence over the decisioninfluence process involving a third party (person or institution)

Patronage Patronage refers to favouring political supporters, for example with

government employment

Source: Transparency International (2009)

Besides, according to Rohwer (2009) and TI (2015) another popular criterion toclassify corruption is based on scales of corruption In this way, corruption is classified aspetty and grand corruption TI (2009) defined petty corruption as “everyday abuse ofentrusted power by low- and mid-level public officials” when they interact with ordinarycitizens who are trying to connect to public services such as hospitals, schools, policedepartments and other agencies with the small scale of monetary transaction On the otherhand, “grand corruption consists of acts committed at a high level of government thatdistort policies or the central functioning of the state, enabling leaders to benefit at theexpense of the public good” (TI, 2015)

2.1.3 Measuring corruption

Corruption is a complicated social, political and economic phenomenon that cannot

be measured directly However, the past decade witnessed a rise in public awareness of thisproblem along with an extension of corruption measurement (Rohwer, 2009) The keymode of measurement since the mid-1990s is perceptions - based Some indices at theaggregate level in this period can be listed such as the Corruption Perception Index (CPI),

the Bribe Payers Index (BPI), the Global Corruption Barometer (all produced byTransparency International), the Business Environment and Enterprise PerformanceSurveys (BEEPS) Another aggregate indicator may be mentioned is the Control ofCorruption element in the World Bank Group‟s Worldwide Governance Indicators (WGI)(Heywood, 2014) Table 1 will summary some features of these measures.

It is noted that due to the lack of global consensus on the definition of corruption,the multi-face of this issue and its nature, it is almost impossible to give an accurate andobjective measure for corruption (Rohwer, 2009) However, it is necessary to distinguishthree types of corruption indicators: perception – based indicators, experience-basedindicators, and proxy indicators The perception-based indicators base on the opinions andperceptions of corruption of citizens and experts The experience-based indicators base onthe actual experience of citizens or firms The proxy indicators measure corruptionindirectly by trying to aggregate many opinions and signals of corruption or estimatethough opposite aspects such as anti-corruption, good government, etc (UNDP 2008, 8 ff.)

Table 2 2: Summary of features of measures of corruption

Statistical summary of expert

Corruption Perceived corruption assessments (e.g expatriate business Almost global depending on having Cross-sectional ranking of

Perceptions (composite) and some measures executives, senior business leaders, sufficient sources Annual (though perception of corruption focusing on

Index (CPI) of corruption control assessment by the US, regional, and not all data sources annual) business environment

in-country experts )

Similar sources to CPI but with some sufficient sources.

sources annual or biannual)

Index (PII)

uncertain.

prevalent

Global Bribe payments by households Public opinion surveys and partial 69 countries in 2005, Comparative prevalence and

household surveys

(GCB) and many cases needs validation

Enterprise Bribe payments by firms Surveys of businesses 62 countries, various Quantitative comparisons of

(WBES)

Governance and

Diagnostic Bribe payments by households Household surveys 16 countries

bribe prevalence and cost

Surveys

(GCDS)

Source: Oxford Policy Management (OPM), 2007

2.2 Empirical studies

2.2.1 Factors influencing the propensity to bribe

Previous firm-level studies have examined the drivers of bribery and the size of abribe, indicating that these two things are affected by the interaction level with publicofficials and several firm characteristics2 Besides, Rand and Tarp (2012) also suggests thatthe informality status of firms is a good indicator to interpret the incidence of bribery

These studies examine the incidence of corruption in a situation where the firm withsome probability faces a request for a bribe from the officials for running a businesssmoothly The corrupt officials who have discretionary power may take actions aim to eitherbenefit or hurt the firm This section reviews the previous empirical studies investigating thefactors influence the propensity to pay a bribe These factors are divided into three groups,including (i) control right of public officials over firms, (ii) bargaining power and (iii)visibility

(i) Control right: includes several indicator variable capturing the degree of interaction level

with public officials and the regulatory burden that the firm faces

Regulations

Tanzi (1998) suggested that the bribery may derive from the burden of regulations.There is no denying that the regulations such as licenses, permits, etc are of importance forgovernments to manage society and economy However, the existence of these regulationsalong with the monopoly power of the government officials in controlling these activitiesgives officials a good opportunity to extract a bribe from those who need the authorizations

or permits The author also showed that the emergence of these regulations requires afrequent and direct contact between government officials and citizens Citizens have to spend

a large amount of time to deal with these regulations This time amount can be reducedconsiderably through the payment informal cost

In line with Tanzi (1998), Svensson (2003) also demonstrates the influence of controlright on the incidence of corruption, in which control right refers the interaction betweenfirms and government In this study, the extensive interaction with public officials ismeasured by two variables, including regulations which reflect the percentage of

2 See, Svensson (2002), Lee, S H., Oh, K K., & Eden, L A., (2010); Rand, J., & Tarp, F (2012) and Malomo,

F (2013)

administrator‟s time spent dealing with a government regulations each month and themonthly cost for accountants, lawyers who deal with regulation and taxes Meanwhile, Randand Tarp (2012) use the number of government inspectors which strongly correlate with thetime use of management to deal with the government to measure the degree of interactionwith public officials.

Svensson (2003) finds that the probability of paying bribe positively related to thedegree of interaction with public officials More specially, enterprises usually have to pay theinformal cost when they deal with public officials who have the power to affect theirbusiness They are more likely to spend more time dealing with public regulation Besides,they also face an increase in payment for accountants and specialized service providers todeal with regulations and taxes

The burden of regulation that firm faces (Tax)

Svensson (2003) and Malomo (2013) suggested that the burden of regulation causedthe firm to facing a higher risk to pay a bribe To measure the burden of regulation, Svensson(2003) use the type of taxes the firms have to pay while Malomo (2013) use the percentage ofsales declared for tax purposes Svensson (2003) fail to demonstrate the effect of the type oftaxes on the incidence of corruption due to the multicollinearity problem However, whenovercoming this issue by using principal components analysis, the author points out thesignificant and positive relationship between two variables Similarly, Malomo (2013) alsofind evidence that companies that spent increasing percentages of sales for tax purpose aremore likely to pay a bribe

Export dependence

Lecraw (1984) and Luo (2007) suggest that certain firms mainly sell their product todomestic customers while other firms primarily generate their income by exporting theiroutput to global market The former are more likely to be required to interact more intenselywith local supplier, customers, labor force and public official when they operate theirbusiness which caused their exposure to corrupt practices This can be explained by the factthat the intensification of exposing to the internal environment will make firms contact with

to a wider range of potential legal vulnerabilities related business' regulatory requirements

On the other hand, Kobrin (1987), suggest that exporting firms characterized by highertechnical and managerial capabilities which are strengthened over time through the learningand innovation when they business in the global market Their profitability is less likely to

suffer from the vulnerability to government corruption than their counterparts This enablesthem to raise the firm‟s bargaining power vis-à-vis government.

Another argument for the negative relationship between export and the drivers ofcorruption as follows In developing countries and especially in countries with balance ofpayments problem, export activities are highly appreciated because of their contributions toforeign exchange and employment (Grosse, 1996; Vernon, 1971; UNCTAD, 2006) Thispoints out that export-oriented firms have high national salience and public officials are morelikely to dampen their bribe demands in order to avoid punishment (Lee, Oh & Eden, 2010)

In addition, Lee, Oh, and Eden (2010) also suggest that the competition among nationalgovernments in attracting and supporting export-oriented firms will strengthen the firm‟sbargaining position With above arguments, Lee, Oh, and Eden (2010) give the hypothesisthat the export-oriented firms are less likely to face a risk of to pay the informal cost as well

as the smaller the bribes have to pay to government officials

The dependence of firm’s profitability on government

Pfeffer and Salancik (1978) showed that there is the difference in firm-levelprofitability dependence on the government among firms The companies obtaining a largeshare of their revenue from government tend to more beholden to the whims of the publicofficial Such dependence comes to be problematic when the corruption dominates the award

of state contracts (Dela Rama, 2012) Because the sales from such contracts might beimportant for companies‟ survival, the firms‟ strategies might aim to induce politicaldecisions in favor of them This subsequently requests firms to develop more internalresources to quickly obtain information on their government characteristics (Hillman & Hitt,1999; Hillman, 2005)

Hansen et al (2009) indicated that when state-owned enterprises have become firm‟skey client will have a positive and significant impact on firm performance It may be arguedthat there is an informal split benefit between the firm and the government official Rand andTarp (2012) and Malomo (2013) also use a dummy variable that indicates whether thegovernment is a firm‟s customer, to examine the bribe incidence They give hypothesis thatwhen firm‟s business depends on government contracts or government is a firm‟s customer,the firm is more likely to corrupt public official

The dependence of firm’s input on government

According to Tanzi (1998), in most countries, sate sector engages in providing ofgoods, services, and resources at a lower price than market prices The list of these public

goods and services can be listed as foreign exchange, credit, public housing, public land,electricity, water or some public services such as such as health care, education, and so on.With access to a large amount of these public goods at a below-market price, some groups ofpeople can considerably benefit from trading these goods Sometimes, the limited supplycauses the shortages of these products; therefore, a solution like ration coupon becomesunavoidable Because the responsibility of allocating the limited supply is often of publicemployees, people who would like take advantages of access would be willing to pay theinformal cost to the public official.

Svensson (2003) and Malomo (2013) also examine the relationship between thedegree of using public services such as electricity, water, telephones and waste disposal andthe propensity to pay a bribe They point out that the companies receiving public services thave a higher risk of paying a bribe

In line with Svensson (2003) and Malomo (2013), Rand and Tarp (2012) prove thatwhen the government as firms‟ main supplier, they face a higher probability of paying abribe

Government assistance

According to Rand and Tarp (2012), over the past years, to promote the development

of SMEs, Vietnamese Government has implemented many supporting policies, divided intotwo groups: financial and technical assistance However, there has been a significant decline

in the proportion of firms receiving assistance from government authorities since the 1990s.This trend is in line with the Vietnamese government‟s strategy of promoting private sectordevelopment However, this also caused of an increase in exposing to potentially corruptactivities This hypothesis was proved when these authors find that the firms receivinggovernment assistance face a higher probability of paying the informal cost

In line with Rand and Tarp (2012), Stensson (2002) also suggests that tax exemption

is one of the firms‟ incentives to pay a bribe However, this study fails to prove thisrelationship

(ii) Bargaining hypothesis

Svensson (2003) constructs a bargaining hypothesis suggesting that firms‟ ability topay and firms‟ refusal power influence the probability of having to pay a bribe In whichfirms‟ ability to pay is measured by firm‟s profitability and firms‟ refusal power is measured

by firm‟s choice of technology The author expects that companies with high profit are more

cost existing current industry will strengthen firms‟ refusal power However, the result of thisstudy shows that there is no link between firms‟ ability to pay and firms‟ refusal power andthe incidence of corruption.

In line with Svensson (2002), Rand and Tarp (2012) and Malomo (2013) also test therelationship between these variables Malomo (2013) points that profit and sunk cost don‟tinfluence on whether or not companies have to pay informal cost whereas Rand and Tarp(2012) find a significant and positive relationship among them These results prove that thetest of bargaining hypothesis is still ambiguous

(iii) Visibility

Stensson (2002) hypothesizes that firm visibility will affect the propensity to pay abribe Rand and Tarp (2012) give two visibility proxies: firm size and the informal status offirms

With regard to firm size which is usually measured by the total number of firm‟semployees or the logarithm of the total number of firm‟s employees at the end of the mostrecent fiscal year3 There is not consensus about the effect of firm size on the propensity tobribe More specially, Beck et al (2005) point that smaller firms are less likely affected byrent-seeking bureaucrats and other institutional problems compared to their largercounterparts However, Hellman and Schankerman (2003) show that relative to their largerpeers, smaller firms pay a higher proportion of their income as bribes and the frequency isasked to pay informal cost also more permanent In addition, weak political resources ofsmaller firms make them be not able to shape regulation in their favor This caused thembecome more vulnerable to government predation (Bennedsen et al., 2009; Harstad &Svensson, 2011) Besides, Rand and Tarp (2012) find a positive and statically significantrelationship between firm sign and the incidence of bribe whereas Svensson (2003) failure toprove this relationship

Related to firm‟s informal status, Rand and Tarp (2012) suggests that smaller andinformal firms seem to hide more easily from government regulation and easily avoid bribe.Sharing the similar view with these authors, Dabla-Norris and Koeda (2008) argued that the-the main drivers of the increase in the number of such firms are to reduce or eliminate taxburden as well as avoid complex administrative procedures and bribe incidence

However, Rand and Tarp (2012) also points out an opposite effect of informality onbribe incidence In particular, if the firms may seek the benefits of informality, they are

3

Svensson, 2002; Rand, J., & Tarp, F., 2012, Galang, R M N., 2013.

willing to offer an informal cost to maintain or achieve the informal status Therefore, thespecification of the net effect of informality is also an empirical question.

2.2.2 How much must graft-paying firms pay?

Svensson (2003) construct a simple bargaining model to explain the variation in thesize of bribe across firms This model supposes that a large number of private firms operateunder control of public official in a territory These firms must choose between paying bribes

to run their business smoothly and rejecting paying bribe and exit current market If the yieldfrom relocating their capital to another industry is greater than net profit from bribery, thechoice of exiting market is optimal In an opposite direction, firms must bargain about thesize of bribe Their bargaining power depends on firm‟s ability to pay a bribe (profitability)and firm‟s ability to refuse to pay a bribe (choice of technology) Svensson (2003) and Randand Tarp (2012) show that a firm using technology with lower sunk cost (lower K/L ratio)will strengthen the firm‟s bargaining power

Svensson (2003) empirically find a significant and positive relationship betweencurrent and expected future profit and bribe amount Additionally, the author points out anopposite direction between sunk costs which measure firm‟s capital mobility and firm‟srefusal power This means that the firms with higher profit, the more it has to pay and thefirms with lower cost exiting the current industry will strengthen their refusal power and payless bribe These results are robust when industry-location averages of the profit rate are used

as an instrument variable for profit

In line with Svensson (2003), Malomo (2013) provides strong evidence of a positiverelationship between profit and sunk cost with bribe payment

2.3 Basic framework to estimate the supply of bribes

This study aims to examine the factors influence the bribe incidence and the factorscontribute to the bribe amount For this purpose, this study follows the model constructed bySvensson (2003), considering an economy consist of a large number of firms and publicofficials, in which each firm comes within one bureaucrat‟s jurisdiction The Official isassumed to maximize his profit However, he faces two constraints Firstly, he does notcollect any bribes in the event that the firm exits the market Secondly, he may get caughtwith a bribe and be punished The size of bribe depends on the officials‟ control rights overthe firms‟ business activities Therefore, bribery tends to occur in the public sector

characterized by high control right Assume that there are two-sector in an economy,including a relatively clean sector (s1) and a fairly corrupt sector (s2).

The probability of randomly picking a firm i who has to pay bribe is:

p(i) = σ(i∈s 2 ) * ρ (1) where σ(i∈s 2 ) is the probability of firm i operate in sector s 2

ρ: is the probability of randomly drawing an official in sector s2 that tend to take a bribe

Assume that the firms manufacture two goods: x1 and x2 that are traded on the worldmarket The world market price of x1 and x2 is respectively θ and 1 and the host country is as

βΠ(αk, |s1) ≥ Π(k, θt, |s2) – b(θt) + Etβ[Π(k, θt+1, |s2) – pb(θt+1)] (2)

where:

Π: is denotes profit

βΠ(αk, |s1): is discounted profit when firm sells and reinvest its partly sunk capital in

elsewhere, α indicates the degree to which capital is sunk cost

Π(k, θt, |s2): is current net profit

Etβ[Π(k, θt+1, |s2) – pb(θt+1)]: is expected future profits, in which p: is the probability of firmface a request for a bribe

Firms cannot borrow to pay bribes so firm‟s net profit in each period must be non-negative:

gained if firm exit industry and reallocate its capital in another industry It means that higherability to pay lead to a decrease the firm‟s bargain power while lower sunk cost of capitalwill increase its bargain power.

2.4 Conceptual framework

As mentioned in literature review, Svensson (2002), Hansen et al (2009), Rand andTarp (2012) and Malomo (2013) show that the degree of interaction between company withpublic official, the regulation burden, the visibility proxies, the refusal power and ability topay will influence the bribe incidence and the bribe amount This relationship can besummarized through the following diagram:

1 Visibility

- Firm size (The log of total employment)

- Informal (Not registered officially)

- Regulations – real time

- Received government assistance

results of these studies reveal that the interaction level with a government official and theregulatory burden are positively correlated with the propensity to pay a bribe Besides, thevisibility which is captured by firm size and firm‟s informal status is a good indicator tointerpret the incidence of bribery The previous studies also prove that the variation in thebribe amount can be explained by the firm‟s ability to pay and their refusal power The thirdpart introduces the basic framework to estimate the incidence and the level of bribe.

Chapter 3: Corruption in Vietnam

The war against corruption in Vietnam has been implemented over the past years,marked by the Law on Anticorruption in 2005 However, after a decade, corruption hasbecome a threat to the sustainable development of the country

A Sociological survey by World Bank (2012) showed that corruption is one of threemost serious issues drawing the attention of population This survey was conducted in tenprovinces and cities, focusing on three groups of respondents: citizens, firms, and officials Itindicated their perceptions of corruption in the list of ten serious problems which attractconcern from social In particular, officials suggest that corruption is a most serious issue onthe list A group of enterprises admit that corruption became only the second cause ofconcern, behind the cost of living Citizens rank corruption at the third place after the cost ofliving and traffic accidents Indeed, figure 1 depicts clearly about the serious issues thatVietnam faces

Figure 3 1: The most serious Economic & Social Issues for Vietnam4

Source: World Bank (2013)

Additionally, this survey proves that corruption in Vietnam occurs in various formswith different levels in many sectors More specifically, three surveyed groups includingpublic officials, enterprises and citizens reach the high consensus of corruption prevalenceamong sectors More than 75% of the total respondents agree that four most corrupt sectors in

Vietnam comprise traffic police, land administration, customs, and construction Meanwhile,post and telecommunication, media, treasury, and the ward/commune police were recorded asthe four least corrupt sectors (Figure 3.2)

Figure 3 2: Perceptions of the prevalence of corruption across sectors5

Source: World Bank (2013)

Corruption in Vietnam has become more and more serious when CorruptionPerceptions Index 2014 (CPI 2014) reports a low rank for Vietnam In particular, on a scalefrom 0 to 100, where 0 is highly corrupt, and 100 is very clean, Vietnam score 31, ranking

119 out of 175 countries worldwide and 18 out of 28 countries in the Asia-Pacific region It isnoticeable that despite the effort to reduce corruption, there has no change in the population‟sperceptions about corruption in three consecutive years (2012 – 2014) This reflects the lack

of citizens‟ trust in the effectiveness of public administration

Table 3 1: Vietnam‟s annual CPI result

Source: Transparency International (2014)

5 Perceptions of the prevalence of corruption across sectors according to public officials, enterprises, and citizens (% saying prevalent, among those with opinions)

Moreover, the PCI 2014 Survey points out a worrisome decline in the score ofinformal cost criteria Comparing this to PCI, 2006 shows an increase in surveyed firms‟pessimism To be specific, in 2008, 66% of respondents have said that they usually pay theinformal cost to run the business smoothly After many years of decline, the trend of firmspaying the informal cost reverse and turned to be over a threshold in 2014 This report alsoindicated that 10% of businesses paying bribes spend more than 10% their revenue on bribes.

In addition, increasing number of firms say that they usually suffer the harassment frompublic officers in administrative procedures (66% in 2014, up from 41% in 2013) The sameresult also was recorded in the PCI-FDI Survey

Figure 3 3: Key Indicators of Informal Charges (2006 to 2014)

in informal payments

Source: The Provincial Competitiveness Index (PCI), 2014

Figure 3.4 indicates the main purposes of informal payment As can be seen from this,the bribe payments are mainly for dealing with tax collectors and easily accessing publicservices In particular, there is about 20 percent of such payment for the former reason, and30% of payment for latter one Comparing to previous years (2009 and 2011), tax servicesshow improvement while public service access has poorer performance

Figure 3 4: The purpose of bribe payment

To get To get licenses To deal with To gain To deal with Other

connected to and permits taxes and tax government customs

public services collectors contracts

public procurement

2009 2011 2013

Table 3 2: Average cost of bribes paid, by sector

Source: Transparency International (TI), 2013

Several studies point out the cause of corruption in Vietnam CIEM (2005) indicatesfour major causes, include: (i) the abuse of public power; (ii) the arbitrary decisions involved

to policies and administration; (iii) the weaknesses in transparency and accountability of

officials and government agencies, (iv) the weakness in implementation and monitoring ofstate Furthermore, the report of IT also emphasizes that the differences in corruption acrosscountries can be explained by the effectiveness of the judicial system, the rule of law, and thepublic governance personnel (Nguyen & Van Dijk, 2012).

Despite government‟s efforts in combating corruption, communities still show anegative perception of the state of corruption in Vietnam The findings of TI survey, 2013show that only one-fourth (24%) of total respondents agree that the government's anti-corruption efforts are efficient Meanwhile, 38% of replies assess the government's anti-corruption efforts is inefficient or very inefficient The remaining 38% of surveyed peoplethought that the efficiency of anti-corruption programs is not clear This report also revealsthe increase in public mistrust of the government‟s anti-corruption efforts To be morespecific, in 2013, 60 percent of questioned people give negative responses in comparisonwith 35% of those in 2010

Figure 3 5: Awareness about the government's anti-corruption efforts

39%

Source: Transparency International (TI), 2013

About the public apprehension about corruption, TI (2013) shows Vietnameseunwillingness of denouncing corruption over time In 2013, up to 62% of respondents areunwilling to report corruptive incidents while this figure for urban citizens is even higher,66% A contrasting result was recorded in 2010 when 65% of respondents were willing toreport occurrences of corruption

Comparing such results among Southwest Asia shows that Vietnamese citizens appear to

be the least voluntary to reveal cases of corruption The TI (2013) points out that the percentage

less than those of Malaysian responses Vietnamese figure is even lower than the averagenumber of 63%.

Figure 3 6: Willingness to report an incident of corruption (Southeast Asia)

Source: Transparency International (TI), 2013

In general, these countries - level data depicts a brief overview of corruption inVietnam There is no doubt that such hazard in Vietnam is severe, especially in traffic police,land administration, customs and construction The abuse of power by public officials and thearbitrary decisions related to policies and administration are the major causes of corruption.Vietnamese government‟s efforts against corruption over the past years seem to beinefficient These result public skepticisms of anti-corruption program

Chapter 4: Data and Econometric Model

This chapter consists of three main parts The first part outlines the data source anddata description As mentioned above, this study uses the available data of SMEs in Vietnamfrom the period of 2005 to 2013 that provide quantitative data on bribe payment and detailedfinancial information of surveyed firms The second part presents the descriptive statisticalanalyzes This part aims to provide a brief look over the key variables and initial relationshipsbetween the initial relationship between explanatory variables and response variables Thefinal part specifies the econometric models as well as the variables included in these models

4.1 Data

4.1.1 Data source

This study uses the available data of SMEs Surveys in Vietnam from 2005 to 2013that have been conducted by several parties Those are the Central Institute for EconomicManagement (CIEM), The Ministry of Planning and Investment (MPI), the Institute ofLabour Science and Social Affairs (ILSSA) - the Ministry of Labour, Invalids and SocialAffairs of Vietnam (MoLISA), Department of Economics - Copenhagen University, andEmbassy of Demark in Vietnam These surveys covered around 2500 small and mediumenterprises, including non-state manufacturing ones in ten cities and provinces in Vietnam.The formal and informal private companies have different sizes of samples To answer theresearch questions, this study exploit information on firm‟s financial performance, theinteraction between firms and public officials, as well as the information on bribe payment

Table 4 1: Panel data structure

Number

Source: Author‟s calculation

Table 4.1 indicates the panel data structure All the firms in cleaned sampled are absolutely