Fighting Corruption in on the Transnistrian Border: Lessons from Failed

Anti-Corruption Programmes and New Successful Programmes




Bryane Michael, Linacre College (Oxford) and

Mariya Polner (EUBAM)




Preprint version of Review of Eastern and Central European Law 34(1).





Overview of the Transnistrian Region.

Estimating the Level of Corruption.

Problems of Anti-Corruption Legislation and Action Planning.

Absolute Integrity and Risk Management

The Optimal Location of Anti-Corruption Authority in an Executive.

Conclusions and Unresolved Questions.






In 2007, both Moldovan and Ukrainian authorities took firm steps to reduce corruption along the Transnistrian border. This paper – aimed mainly at anti-corruption practitioners and scholars in public administration – discusses the background and underlying principles guiding the anti-corruption work being adopted by both governments in order to facilitate discussion about optimal anti-corruption programme design. This paper presents a set of tools used during the planning phase of the anti-corruption programme -- outlining the methodology used to assess the extent of corruption on the Transnistrian border, the problems of legislative transplants, a “contract test” for defining corruption offenses, a method of risk analysis, and a model of optimal anti-corruption programme organizational design.






Fighting Corruption in on the Transnistrian Border: Lessons from Failed

Anti-Corruption Programmes and New Successful Programmes

Bryane Michael, Linacre College[1]




Government services in both Ukraine and Moldova are widely perceived to be corrupt and inefficient. These perceptions are broadly supported by data compiled by the World Bank.[2] Figure 1 shows the results of a survey which measures the effectiveness of government to engage in public sector activities. The Figure shows that Ukrainian and Moldovan government services are, according to the survey data, more effective than most Former Soviet countries -- but much less effective than their Eastern European counterparts in countries such as Poland and the Czech Republic. One implication of relatively low government effectiveness has been an inability to control corruption. Looking at control of corruption indicators shown in Figure 2, Ukraine and Moldova are roughly as capable (or slightly more capable) in fighting corruption than other former Soviet Union countries -- but much less able to control of corruption than the countries of Central and Eastern Europe.




This article attempts to provide an academic background on practitioner work in designing an anti-corruption programme at the Transnistrian border – allowing the reader to “look under the hood” of an anti-corruption programme and see some the academic considerations in anti-corruption programme design.[3] The first part of this article will provide an estimate of the scale of corruption along the Transnistrian border – showing an economic methodology used to estimate corruption in two particular services – border guard and customs -- in Ukraine and Moldova. The second part of the article presents the problems with current anti-corruption programme design (particularly those relying on legislative changes and action plans). To prepare for a greater role of administrative sanctions (instead of criminal sanctions) against corruption, the paper discusses the “contract test” for corruption. The third part of the article discusses a risk-management approach to fighting corruption – showing how authorities can fight corruption without increasing the regulatory burden which actually increases corruption. The fourth part of the article discusses the optimal “location” (within the public sector) of anti-corruption work in a public sector – namely the decision taken by government authorities and advisors where to base legal responsibility for fighting corruption. The final section concludes by raising questions for further research.


This paper discusses the fight against corruption focusing on two specific government services – the border guard service and customs service – in Ukraine and Moldova; in a very specific geographical region – the Transnistrian region. At the time of this writing, the political situation surrounding the geographical area between the Dniestra River and the Ukrainian border remained uncertain (as will be briefly discussed in this article). For the purposes of this article, Transnistria shall refer to the region in along the Ukraine-Moldovan border and not to the political entity occupying the Transnistiran area.

Overview of the Transnistrian Region


            The Transnistrian area is situated along a significant proportion of the Moldovan-Ukrainian border. In 2006, more people and goods flowed from Ukraine to Moldova than visa versa. In 2006, roughly 7.5 million people crossed the border from Ukraine into Moldova, while 5.8 people crossed the border from Moldova into Ukraine. Figure 3 shows the level of trade between these two countries – and goods are either transported in trucks and cars across the joint border, by plane, or by sea. Both countries also serve as important crossing through points for goods in transit on route to other countries.[4] Trade with both Ukrainian and Moldovan businesses is highly regulated, creating incentives to pay bribes in order to circumvent trade restrictions as will be discussed later in this paper. As can be seen from Figure 4, the number of regulations required to engage in international trade is higher in these two countries than in Central and Eastern Europe in general (though lower than in many Former Soviet countries).




            The red area in Figure 5 represents the area disputed between the Moldovan government and the authorities of the Trannistria based in Tiraspol. This existence of this semi-automonous area has created a unique situation – from an anti-corruption point of view – because the region has often been referred to as a “black hole” for the trafficking of people, arms, illegal narcotics, and goods for resale. Reliable estimates put black market activity in the region at roughly $250 million (CISR, 2003). [5]  Reliable information also suggests that Transnistrian companies are engaging in fraudulent re-export – importing products from outside Moldova and then falsifying documents of Moldovan origin in order to take advantage of lower tariff rates between Moldova and Ukraine.[6] Anedotal evidence suggests that the extent of administrative corruption in the Transnistrian territory is extremely high, compared to corruption levels in either Ukraine or Moldova – in effect representing a pool of individuals and resources willing and able to corrupt officials on either side of the territory. In one view, the Transnisitrian authorities have facilitating the traffic of contraband in order to finance its military activities aimed at independence and bolster a highly regulated economy still based on central planning and Communist based public administration practices. However, this view has recently been contested by independent observers who note that previous estimates of black market activity have been exaggerated . [7]  


            Irregardless of Transnistria’s diplomatic position, the existence of this semi-regulated territory wedged between Ukraine and Moldova has resulted in several distortions in the trade environment. First, a cordon sanitaire has been established in Moldova, along the Dniestrian River, consisting on internal customs inspectors responsible for inspecting goods and people coming into other parts of Moldova from the Transniestrian region. Second, in order to ensure that Transnistria based companies are regulated according to international standards, third-party countries such as Ukraine insist that Transnistrian companies exporting abroad must have a Moldovan certificate of registration – obtainable in Chisinau (the capitol of Moldova). Third, and partially as a continuation of Soviet practice, both Moldovan and Ukrainian authorities engage in 100% inspection of goods and individuals crossing the border are searched – resulting in significant economic losses in time spent at the border, administrative delays and disincentives for engaging in international trade.

Estimating the Level of Corruption


Corruption, particularly the payment of bribes, occurs when a private individual interacts with a government official.[8] Such interaction with government officials in the Transnistrian region primarly occurs as border crossers interact with customs officers or border guards.[9] Customs services in Moldova and Ukraine are responsible for collecting taxes on particular types of goods, as well as ensuring that importers and exporters follow import/export procedures. Customs officials are also responsible for ensuring that contraband (such as weapons, drugs, and generally prohibited goods and materials) does not enter the country. The border guard service checks passports (to ensure that the individual entering the country has the legal right to enter) as well as patrols border crossing points and the “wild” border (known as green borders on land and blue borders at sea) to ensure that individuals are not illegally entering the country. Any time an importer or an individual entering the country wishes to contravene national regulations, they can bribe customs and border guard officials.


From the reports and survey information available, corruption affecting the particularly affecting the Ukrainian/Moldovan border is relatively modest by national standards (though excessive by European Union standards). While direct measurement of corruption is difficult, indirect measurements are often used to point out obvious problems with customs and border guard service delivery – as corruption is often either a symptom of poor service delivery, or concomitant with poor service delivery. Slow and unhelpful administration often results in corruption. As shown in Figure 5, roughly 30% of Moldovan respondents in a business survey reported by Carasciuc (2005) noted that they had to give a bribe every time they crossed the border (with roughly the same amount of respondents noting that they never need to pay bribes). In exchange for those payments, as shown in Figure 6, the majority of border crossers received faster processing times, and one-third received a “better” classification of goods. More worryingly, in 14% of the cases, business representatives faced coercion – having “no other choice” but to pay bribes.



Survey data covering trade-related corruption among Ukrainian business people are much more difficult to obtain.[10] Figure 8 shows popular perceptions of corruption in a number of government services.[11] As shown in the Figure, 64% of respondents think that corruption represents a “serious” problem (as stated in the survey) whereas almost 80% of respondents thought corruption in hospitals represented a serious problem.[12]


Corruption has a negative impact on both public revenue and economic development. According to the Carasciuc (2005) study, in Moldova, the average size of a bribe paid to a customs official is $32 (staying the same amount in 2002 as 2006). Border guards asked for – on average -- $17 (remaining also at the same level from 2002 to 2006). According to Government of Moldova data, 5.8 million people crossed the border from Moldova into Ukraine. Based on international evidence, at least 5% of border crossers may be required to pay bribes. Thus, if Moldova conforms to other countries, $4.9 million in bribes were paid to Moldovan border guard. Assuming that the bribe on the Ukrainian side is the same, and that 7.5 million people crossed from Ukraine into Moldova, then the total amount paid in bribes is $6.4 million. 


Indirectly, the level of corruption can be estimated from the level of trade between Moldova and Ukraine as well as estimates from businesses about their “bribe tax” (or the proportion of the value of their trade and/or sales) which is collected in the form of bribes – 20% of firms estimate that 2% of the value of their trade is subject to a “bribery tax.”[13] Given a base of $480 million in exports from Ukraine to Moldova, $1.9 million in customs is paid in bribes in Moldova for Ukrainian-origin goods. According to World Bank (2005c), 15% of firms note that they pay an average of 1.5% of the value of their merchandise on a “bribery tax.” As also shown in Figure 3, for a base of $100 million in trade between the two countries, the amount of customs bribes is a minimum of $225,000 in Ukraine for Moldovan-origin goods.[14]


These estimates, however, underestimate the extent of corruption along the Ukrainian-Moldovan border, because they do not take into account unreported activities (either persons crossing the border to the value of goods not reported in the official statistics). In order to estimate corruption involving these unreported activities, corruption “risk” needs to be calculated as the probability of each unreported event occurring multiplied by the number of times it occurs.[15] Taking Carasciuc’s (2005) estimates of the average bribe level for customs and border guard officials, and other estimates related to the proportion of under-valuation of goods at least $30 million in trade taxes is not collected by Ukrainian officials for goods from Moldova in route to Ukraine.[16] Only particular reflecting Moldova’s much larger import values, roughly $110 million in trade taxes is not collected by Moldovan officials for goods from Ukraine heading to Moldovan markets.


Some corruption is most likely tied to illegal activity.[17] Figure 10 provides a brief rough estimate of the scale of corruption tied to black market activities related to the traffic in illegal narcotics, weapons, and the humans. According to publicly available sources, the total estimated amount of heroine flowing through Ukraine in destination to Western European markets is 23.1 metric tonnes at a market price of $78 per gram results (with roughly 1 million grams per ton) in a market value transited through Ukraine of $1.8 billion.[18] If the bribery tax previously referred to covers illegal transit of narcotics, then the value of these bribes equals $18 million. With regard to arms, estimated that, in the 1990s, roughly $5.3 billion in arms leaves the Ukraine. If the same bribery tax applies to guns, then the value $106 million. If the actually is even 10% of the 1990s value, then the likely value is $10.6 million.[19]


Figure 10: Contraband and Estimated Corruption Value

                          See print (offline) version for estimates

Sources where cited. Estimated bribe revenue derives from World Bank estimates of the “bribe tax” previously mentioned in the text. Data are unavailable for Moldova.


            The economic logic underpinning these specific calculations is described in Figure 11. However, the economic impacts are much larger in a number ways. First, the reader having taken a course in international economics will immediately recognize that the bribe serves as a trade tax – decreasing wages in Moldova and the return on capital in Ukraine (assuming Ukraine is the relatively capital-intensive trading partner). Second, the reader versed in Keynesian macroeconomics – and particularly the Keynesian multiplier -- will immediately recognize that a bribe which equals 5% of the value of the goods will reduce overall economic activity by paid at the border can reduce the level of economic activity in a sector by 20%. Third, a reader versed in public economics will recognized that a 5% bribe will result in 25% loss in welfare in a particular market.


Figure 9: An Economic Calculation of Corruption along the Transniestran Border


Estimates involving the value of corruption in international trade are always difficult to obtain because parties to corruption have little incentive to disclose their activities to third parties. However, approximations can be made.[20] For example, as previously noted, survey data (from World Bank (2003), can reasonably reliably estimate the value of traders’ goods paid in bribes. A “bribe tax” (denoted by b) may affect a value of normal consumer goods flowing across a border in any particular year (Xt).[21] While the payment of bribes is an obvious economic harm to traders (though from an economist’s point of view, bribes often represent a simple redistribution of national income). Bribes are often paid, following economic incentives, to provide traders with an economic advantage – such as lower taxes which result from under-valuing or mis-categorising goods. If Xt represents the true value of goods and Xr represents the reported value of goods, then the State budget clearly loses the tax (t as expressed as a percentage) of the difference between the reported value and the true value.[22]


However, the real harm to corruption at the border occurs for two reasons. First, the value of production may be affected as traders divert a fraction of their overall production into other activities (d) – as businessmen substitute out of activities where bribes are sought and into other economic activities. Offsetting this loss are the gains from faster trade and lower taxes which encourages producers to increase by a fraction their production (c). Second, corruption in legal goods (like potatoes) often encourages – or at least occurs in parallel with -- trade in illegal goods (like contraband). If q represents fraction of contraband goods which are traded in relation to normal consumer goods, then the total loss to budget (as these goods as are not taxes) is qXt.[23]


Combing these terms results in an overall estimate for the loss due to corruption at:


                         Corruption Loss = bXt + t(Xt – Xr) + (d-c)(X*-Xt) + qXt


Of course, this estimate does not incorporate the social and other harms often referred to by academics and policymakers.

Problems of Anti-Corruption Legislation and Action Planning


In Central and Eastern Europe, a large amount of activity has focused on the adoption of a comprehensive anti-corruption law which criminalizes bribery and the elaboration of a national anti-corruption action plan. Both Ukraine and Moldova have anti-corruption laws which criminalize bribery and have adopted multi-year anti-corruption action plans. In Ukraine, in the Ministry of Interior has been charged to elaborate an Anti-Corruption Action Plan covering the entire executive, for implementation by 27th of February 2007.[24] Moldova has adopted a multi-year National Anti-Corruption Action Plan (until 2008) which covers all the major services. In both Ukraine and Moldova, the customs and border guard services have been tasked with elaborating department specific instructions aimed at implementing the anti-corruption law and the national anti-corruption action plan.[25]


A large amount of donor assistance has been committed by foreign donors – particularly USAID and the OECD – aimed at improving the quality of anti-corruption legislation in both countries.[26] Judging the quality of legislation is difficult – particularly in the area of anti-corruption where no set of practices has been shown definitely to be more effective than other practices.[27] However, an obvious metric by which to assess the quality of legislation and/or implementing regulation must depend on its purely technical aspects. Following Michael (2004), one way of assessing the quality of anti-corruption legislation and regulations depends on its specificity and relevance. Specificity refers to the extent to which legislation refers to concrete practices instead of abstract principles and answers the 5 journalist questions (who is affected, what is involved, when are actions to be carried out, and how are these activities to be undertaken). Second, in both the practitioner and academic literature, a very wide array of factors have been linked to the level of corruption in an administration (from the motivation of staff to national culture). Yet, some factors relate more directly to fighting corruption – as opposed to other areas of activity such as civil service reform – and thus are classified by their relevance to the anti-corruption programme.  


Figure 11 provides an assessment of the national anti-corruption action plans. As shown, the quality of the actual legislation seeking to prevent corruption is relatively immaterial because the plans aiming to implement the law are highly non-specific and irrelevant. In the case of Ukraine, its national plan is significantly more abstract than Moldova’s plan. However, the international advisors previously working with these countries are partly to blame. As also shown in Figure 11, the recommendations for reform offered by organizations such as the OECD and the Asian Development Bank in their action plans for the region are also relatively general.[28]


     Figure 11: Assessing the Quality of Anti-Corruption Action Plans


Region Specificity
Ukraine 2.3
Moldova 3.9
Kosovo 1.8
Romania 2.6
OECD/Istanbul 1.4

         Source: author. (one is lowest score and five is highest). See Michael  for a detailed

         computation of these scores and a further discussion of the methodology.


            An obvious – though wrong – implication of this discussion is that anti-corruption legislation and implementing regulations need to be strengthened. However, the adoption of anti-corruption conventions and work on national anti-corruption programmes do not correlate to more effective action against corruption. Stevens and Rousso (2003) – using econometric analysis – investigate whether the adoption of anti-corruption legislation affects the level of corruption in country (as proxied by individual’s perceptions of the level of corruption). They specifically define three variables aimed at measuring the legislative environment surrounding the fight against corruption. First, they measure the adoption of “omnibus” activities (the adoption of a national anti-corruption which involved NGOs and multiple branches of the executive), the development of a national anti-corruption action plan, and the creation of an anti-corruption agency. Second, they measure the ratification of legal frameworks, specifically a civil service law, a financial disclosure law for civil servants and/or politicians, a freedom of information law, a law on political party finance, and an anti money laundering law. Third, they measure membership to international conventions including the Stability Pact, the OECD Convention and the Council of Europe’s four conventions against corruption. Figure 12 shows some of the many variables in their analysis – pointing to the fact that legal adoption has a statistically significant correlation with reductions in perceived corruption—though the adoption of national action plans or work in international consultative committees does not.






Figure 12: Selected Regression Coefficients from Stevens and Rousso (2003)

 See print version for coefficients


            Such results are worrying because the adoption of national anti-corruption laws, especially those which criminalize corruption offenses as recommended by conventions promulgated by the OECD and the Council of Europe are often used as a measure of the strength of anti-corruption legislation. These conventions call for the establishment of criminal rather than simply administrative or civil liability for corruption offenses by civil servants. Ostensibly, the more severe criminal remedies -- proposed by model laws such as the OECD Convention -- provide greater deterrence for engaging in corruption. While implementation of the criminal responsibility foreseen in the OECD Anti-Corruption Convention (and thus the acquis communautaire) is laudable, criminal responsibility for corruption may reduce the incentive to investigate and prosecute cases of corruption in Ukraine and Moldova for two reasons.[29]


However, the adoption of the OECD Anti-Bribery Convention by Ukraine and Moldova poses problems.[30] First, the Ukrainian and Moldovan Ministries of Justice can not be relied upon to successfully prosecute criminal cases of corruption given severe resource constraints and corruption endemic within their own service.[31] Second, as superiors are legally and administrative responsible for the corrupt activities of their sub-ordinates, they have been unwilling to strenuously investigate – in order to avoid prosecution themselves! Third, the level of proof required for a successful criminal conviction reduces greatly the range of cases which may be pursued. Administrative sanctions against corruption based on “balance of probabilities” standard for successful conviction can – in certain circumstances – provide greater deterrence against corruption than the ostensibly strong criminal standard which requires proof “beyond a reasonable doubt.”[32]


In both Ukraine and Moldova, administrative sanctions should be strengthened against corruption – starting with work on teaching civil servants to recognize corruption offenses. Figure 13 provides an administrative test which may help civil servants work on the difference between corruption and gift-giving.


Figure 13: The Definition of Corruption and the “Contract Test”


            The definition of corruption has been debated.[33] While in the 1990s, the standard definition of corruption was “the (mis)use of public power for private gain”, more recent definitions – particularly promulgated by the international non-governmental organization Transparency International -- have encapsulated “the misuse of entrusted power” either in the public or private sector. Such a definition has gained a large amount of credibility in policy circles given Transparency International’s large PR activity and close relationship with a number of international organizations and national governments.


Such a definition of corruption should be avoided for three reasons. First, such a wide definition of corruption subsumes every possible form of deceit, deception, or conflict between persons – thus providing no definition at all! Second, the nature of the principal-agent relationship is completely different in the public as opposed to private sector. A civil servant is delegated authority by the entire society, through a political process – making infractions crimes against the body politik. A private individual enters into a formal or informal contract in a company or NGO, whereas the agent serves one or more principals – making infractions torts against private persons. Third, a definition of corruption should serve a specific legal function. Adequate legal definitions exist for the misuse of entrusted power in the private sector – as defined under fraud, theft, and other tort and/or criminal offenses. Until the mid-200s, the field of anti-corruption treated matters relating to the use of public power for private gain and the field of corporate governance treated matters relating to the misuse of entrusted resources in a private setting.  


If corruption is defined as the use of public power for private gain, the obvious question is whether gift-giving does (or should) constitute a corruption offense. In developing countries, gift-giving is an entrenched part of many types of interactions between civil servants and private individuals. Such gift-giving can help augment low civil servant salaries and provide incentives for more efficient service delivery (in a highly regulated civil service environment where civil servants often have few high-powered incentives to work hard). In order to help civil servants move away from a dogmatic definition of corruption, a simple test may help civil servants – and the members of an Internal Security Department or Ombundsman institution– define the difference between a gift and a bribe. The test consists of two parts:


1. Has an extra payment been made for the quantity, quality, speed, friendliness and informativeness of a service which the service user has an administratively defined right to?


2. Was the ex-post delivery of the public good or service made on the ex-ante expectation of an extra-payment?


If the answer to both questions is affirmative, then the transaction involves corruption and if the answer to both questions is negative, then the transaction involves a legitimate gift.



Absolute Integrity and Risk Management


The response to fighting corruption in many countries has been to increase the amount of anti-corruption regulations. These regulations increase the amount of reporting which civil servants do – creating an “audit culture” which serves to demonstrate probity instead of supply public goods and services.[34] Both Moldovan and Ukrainian border guard and customs services have adopted a wide range of measures to fight corruption including the dissemination of rules against taking bribes, psychometric testing of staff to test for psychological pre-dispositions toward corruption, bans on pocket money and the use of a personal mobile phone during working hours, and increased oversight of all staff by newly formed departments for Internal Security. With anti-corruption aid to both countries from the United States (of almost $50 million!), the amount of control will increase dramatically.


Yet, the pursuit of “absolute integrity” in Ardy's  words, may harm the public administrations of both countries more than it helps them. First, each regulation increasingly distorts the action of civil servants and creates incentives to contravene these regulations. Second, these regulations costs money to create and to enforce – resulting in an often repeated case where grants of equipment and regulations (“developed through consensus with key stakeholders in order to promote participation and build political will”) lie unused after the donor team returns to headquarters. Third, regulations of this kind of deincentivise staff – resulting in work effort losses which are greater than the harms from petty bribery. Simply put, an anti-corruption regulation should be put in place when:


             marginal                                          marginal                     marginal                            marginal

     loss from corruption          >         creation of rents   +    cost of enforcement   +    reduced work effort . [35]


Instead of seeking to establish rigid rules, a system of guiding principles (such as the “contract test” previously referred to) and a system of random audit should be conducted for four reasons. First, controlling all goods goes against the basic tenants of economics (particularly those involving the gains from trade) as well as the principles enshrined in the European treaties. Second, a properly designed random audit will detect the same proportion of offenses as 100% inspection. Third, random audit frees up resources to be used for investigation. Indeed, an inverse relationship normally exists between inspection breadth and quality – namely the higher the percent of people or goods inspected, the lower the quality of the inspection.[36] Fourth, (as previously mentioned), regulations potentially result in the creation of rents – thus reductions in regulations often result in decreases in incentive to seek bribes.


Such a stratified random sampling procedure divides people or goods into similar groups representing similar risk levels -- for example all consignments from Russia may have a statistically significantly different level of risk of under-valuation than those from France. Usually initial random sampling is used to arrive at these estimates.[37] Repeated sampling results in probabilities changing as the underlying risks change. [38] Such a random audit balances the expected gains from searching one group of people or goods versus randomly inspecting another group – or strata. Figure 14 shows the logic of differential random sampling. As shown in the graph, the first inspections of group A will have a significant benefit as obvious cases of under-valuation or contraband are found. As more inspection occurs, these marginal costs decrease. A similar logic applies to the inspection of group B – thus both marginal benefit curves slope downward. Clearly inspections could occur on group A to the point where the expected return to search group B (defined as the probability of detection multiplied by the value of the contraband or legal infraction found).



In order to detect cases of corruption, service staff can also undergo stratified random sampling and audit. As shown in Figure 15, the border guard or customs service staff may randomly audit differing strata of border crossers. Similarly, the Internal Security department may randomly audit service staff – based on the expected probability that they might be involved in corruption offenses. Naturally, the risk of service staff being corrupt is tied in some way to the probability that the people they are controlling are involving in legal infractions. Thus, the two tiered system described in Figure 15 addressed both regulatory risk as well as corruption risk.[39]



            Such a risk management approach is increasingly common practice in the European Union. However, as government efficiency is higher in other countries, Ukraine and Moldovan Internal Security services need to play a developmental as well as supervisory function – by auditing for better performance as well as the possibility of corruption…as an Internal Security Department which seeks only to control corruption will have difficulties generating enough public value to continue.[40] Such activities could include Ombundsman’s functions, the regular conduct of service delivery surveys and training.

The Optimal Location of Anti-Corruption Authority in an Executive


            Since the late 1990s, the international donors have financed work on the establishment of anti-corruption co-ordinating councils or agencies in a number of countries including Georgia, Latvia, Lithuania, Bolivia, Nicaragua, Nigeria, Kenya and others. These Anti-Corruption Co-ordinating Councils (and in many cases anti-corruption agencies) were created because the often complex nature of corruption cases. Figure 16 shows the process of a corruption case.  


Yet, the large variation we observe around the world in centering anti-corruption activity suggests that the optimal organizational structure for fighting corruption remains an open question. Figure 17 shows the various possible organizational forms for fighting corruption – and stylized facts do not point to the efficiency of one form over another.


Figure 17: Various Anti-Corruption Organisational Arrangements


Lead Ministry

Work as Lead



Ministry of Interior

In many countries, investigators are responsible for ensuring successful prosecutions of corruption.

Strong competencies in enforcement, can ensure case management

Excessive law enforcement

Ministry of Justice

Inspired by Italian experience, the courts and judges as main bulwark against corruption

Can change laws and prosecute cases as own-competence.

Excessively legalistic perspective

Ministry of Economy/Development

Responsible for overall framework and incentives

Ensures root economic incentives for corruption removed

No competencies in investigation or prosecution

Office of Prez/VP

In corruption ridden administrations, president’s office takes AC as political promise and ensure clean government.

Ensures co-ordination and senior level political will

Work is politicized and changes with each administration


Army upholds the interests of the State even when Government weak 

Strong power potentially independent of the government

History of human rights violations.

Each service responsible for self

Each service, often with Internal Security department, ensures own integrity.

Each service responsible for self

Knows own problems best and can tackle at lowest level

Significant duplication of resources and overlap


These bodies centralize work on AC and look only at public servants.

Ensures centralization of AC knowledge and competencies

Shifts responsibility to “someone else”

Co-ordinating structure

Consists of representatives of several agencies meeting together reguularly

Balances specialization with integration

Everyone (thus noone) responsible

Private company or NGO

Works as “island of integrity” completely outside of a corrupt government.

Can ensure integrity when whole public service corrupt

Authority delegated by government…no legitimacy

            However, theory provides a guide for thinking about the optimal organization structure for anti-corruption work.[41]  Transaction cost economics argues that integration should occur between organizations – or organizational units – as the cost of writing, performing on and enforcing contracts between organizational units increases.[42] Such an effect militates for organizational integration because marginal benefits increase (to a point) as the centralization of anti-corruption activity occurs – cases do not need to cross departments or ministries and information can be shared more easily. Such an effect is shown as an upward sloping line in Figure 18a and labeled as a transactions cost effect.[43] On the other hand, several salient results from organizational theory point to the benefits from specialisaton of labour and tasks. Authors such as Kogut and Zander (1992) representing this tradition argue that integration serves only to combine competencies – and the development of these competencies in themselves should be left to specialized organizational units. In this view, as anti-corruption work becomes more integrated, the specific learning from each ministry decreases and the marginal benefits of such integration decrease as integration occurs. As shown in Figure 17a, such a capabilities effect is traced as a downward sloping line. The degree of integration of anti-corruption work should be at the point at which the marginal benefits from integration equal the marginal benefits from decentralization results in the optimal organizational structure for anti-corruption work.





Naturally the factors underlying these transaction cost and competencies effects may change over time – or vary from country to country. Looking at the resulting changes to the marginal benefits of integrating anti-corruption work as underlying factors change results in changes to the optimal degree of centralization of anti-corruption activity. Figure 17b shows the results when underlying conditions change/differ. For example, if transactions costs change (political factionalism increases or a data protection act is passed which hinders the sharing of data across departments), then anti-corruption work should become more centralized. If competencies effects change – such that each department becomes better at developing and using anti-corruption knowledge), then the optimal amount of anti-corruption activity integration decreases.



            While Figures 17 and 17b are purely illustrative, they provide intuitions needed to solve the problem of organisatonal structure in Ukraine and Moldova. Ukraine has relatively fractious and highly politicized services, prone to centralized corruption. Thus, concentration of anti-corruption activity in a corruption free ministry is preferred to creating an anti-corruption agency or centralizing work in the office of the President or Prime Minister. Given extreme weaknesses in all law enforcement bodies, co-ordination of relatively efficient anti-corruption work across services would not be likely. Moldova having primarily disarticulated corruption has a co-ordinating committee – probably the most efficient organizational structure given existing costs and benefits of fighting corruption.[44] Given this model, over time the optimal anti-corruption structure will probably be a cross-border informal anti-corruption committee – covering border guard and customs services -- in which representatives from each service (as well as external donors and advisors) would meet. Ukraine and to a lesser extent Moldova already have such a structure at the national level. In support of the national anti-corruption co-ordinating committee, a similar structure should be established for border guard and customs to review activity against corruption.

Conclusions and Unresolved Questions


            In some ways the experience over the last 20 years in Central and Eastern Europe provided useful cases of what not to do in Ukraine and Moldova.[45] The design of an anti-corruption programme on the Transnistrian border can not rely on more action planning and the adoption of international conventions by Ukrainian and Moldovan governments. A new system of anti-corruption regulations at the departmental level are also contra-indicated. Instead, standard remedies from public sector reform offer the greatest remedy against administrative corruption on the Transnistrian border. The implementation of a system of risk management and a coherent organizational form which is able to investigate and successfully prosecute corruption remains the bulwarks of effective anti-corruption. In Ukraine and Moldova, the border guard and customs services are relatively autonomous entities who can enact reform quickly.


            Yet, the Transnistrian situation poses a particularly new situation for anti-corruption programme design for two reasons. First, the literature lacks models of public administrations in which one or two services can achieve dramatic decreases in corruption when the entire public sector management environment is corruption-ridden. Thus, work on the Transnistiran border will provide a useful case study for further work looking at department-specific dynamics of corruption. Second, the literature lacks convincing models of corruption which spills over from foreign territories – as corruption appears to do from the Transnistrian region into both Ukraine and Moldova. The “import” of corruption is a tragically under-researched topic in the anti-corruption literature.   



Black, D. and R. Blue. (2005). Concept Paper: Rule of Law Strengthening and Anti-Corruption in Ukraine. Available at:


Carasciuc, L. (2005). Analytical report: Exploring the perceptions and experiences with the customs control, services, and border officers in 2002-2005. Chisinau: Transparency International – Moldova. Available at:


Centre for Investigations and Strategic Reforms -- CISR. (2003). Research Paper on Transnistria", Chisinau. Available at:


Democracy International, IFES and DPK Consulting. (2006). Moldova Anti-Corruption Assessment Final Report. Available at:


Drew, K. (2003). Whistle Blowing and Corruption: An Initial and Comparative Review. London: Public Services International. Available at:




Galtung, F. and C. Stamford. (2003). Measuring Corruption. London: Ashgate.


Gorodnichenko, Y. and K. Peter. (2006). Public Sector Pay and Corruption: Measuring Bribery from Micro Data. IZA Discussion Paper No. 1987 Available at:


Hellman, J., G. Jones, D. Kaufmann and M. Schankerman. (2000). Measuring Governance, Corruption, and State Capture: How Firms and Bureaucrats Shape the Business Environment in Transition Economies. World Bank Working Paper 2313. Available at:


Hors, I. (2001). Fighting Corruption in Customs Administration: What Can We Learn From Recent Experiences? OECD Development Centre Working Paper 175. Available at:


IFES. (2005). Survey: Public Opinion in Ukraine November 2005. Available at:


International Crisis Group (2006). Moldova’s Uncertain Future. Available at:


Kaufmann, D., A. Kraay, and M. Mastruzzi. (2006). Governance Matters V: Governance Indicators for 1996–2005. Available at:,,contentMDK:21045419~menuPK:1976990~pagePK:64168445~piPK:64168309~theSitePK:1740530,00.html


Kiev International Institute of Sociology. (1998). Questions on National Integrity.


Kouzmina, L. (2006). Russian Customs Reform Follow-up survey, 2006. Stockholm: SITE. Available at:


Kogut, B. and U. Zander. (1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science 3: 383-397.


Layne, M. (2001). Estimating the Flow of Illegal Drugs Through Ukraine. Washington: US-Ukraine Partnership. Available at:


Movchan, V. (2004). Influence of Tariffs and Non-Tariff Barriers on Ukraine’s Imports: Would the WTO Membership Make Change? Paper prepared for the 6th Annual Conference of the European Trade Study Group. Available at:


OECD. (2004). Ukraine: Summary of assessment and recommendations. Downloaded from the World Wide Web on 26 March 2007. Available at:






SPAI. (2002). General Assessment Report for Moldova. Available at:


Spector, B., S. Winbourne, J. O’Brien and E. Rudenshiold. (2005). Corruption Assessment Ukraine. Washington DC: USAID Report. Available at:


Trade and Transport Facilitation in Southeast Europe Program (TTFSE). Available at:


UN (2004). United Nations Handbook on Practical Anti-Corruption Measures for Prosecutors and Investigators. Vienna: UN. Available at:


UNODC. (2005). Annual prevalence of drug abuse. Available at:


World Bank. (2005a). Guidelines For Risk Management in Customs. Available at:


World Bank. (2005b). Enterprise Surveys. Available at:




[1] The following article draws upon the author’s experiences while working for the European Union Border Assistance Mission (EUBAM) in Odessa, Ukraine in 2007 as an independent expert. All the data are drawn from public sources. The views in this paper remain my own and do not represent the views of the EUBAM or the Ukrainian and Moldovan governments. None of the material in this paper can be construed as representing a position on the diplomatic situation surrounding Transnistria (as I have no view on this situation).

[2] These data, like other data such as Transparency International’s cross-country index of corruption perceptions represent a computed summary measure from several perceptions surveys. While the World Bank estimates use more complicated statistical procedures, these numbers still only represent levels of corruption as perceived by survey respondents.  See Kaufmann et al. (2006) for more on the methodology. See Galtung and Stamford (2003) for more on the strengths and weaknesses of various empirical methods of measuring corruption.

[3] As a contribution to the anti-corruption literature, the material in this paper naturally does not represent all or the considerations of the Transnistrian programme or suggest that the advisors slavishly followed theory in designing the programme.

[4] Due to the difficulty in estimating the volume or value of this trade, such traffic is not considered in this paper.

[5] The CISR (2003) estimates that, given the economic sanctions being imposed on Transnistria, up to 50% of officially reported GDP is involved in the black market. In 2005, according to the same source, GDP in the region was $500 million. 

[6] These schemes are both variegated and often complex. One example of such a scheme is the illegal transport of goods to duty-free zones in Moldova were these products are labeled as duty-free production and then re-exported to their original country of origin (often Ukraine)! The EUBAM website ( provides this and other information.

[7] For an unbiased and highly readable analysis of Transniestrian situation (and the possible responses by the EU and the EUBAM, see International Crisis Group (2006). 

[8] The canonical definition of corruption involves the “use of public power by a government official for a private gain” A defense of this definition is presented later in this paper.

[9] In practice, several other government departments may be present at the border, including health and safety inspections and veterinary services. 

[10] Several USAID financed reports attempt to assess the level of corruption in Ukraine (Spector et al., 2005; Black and Blue, 2005). As the report authors are known experts in the anti-corruption field, the abstract, generalized and highly jargon-laid nature of these reports can probably be attributed to the funder (USAID). See SIGMA (2006) for a comprehensive governance assessment. SPAI (2002) provides an assessment of Moldova’s anti-corruption programme.

[11] According to a survey taken in 2003 under the Partnership for a Transparent Society Program, 75% of respondents believed corruption to be very widespread in the Ukrainian central government, while 62% indicated they had actual personal encounters with corrupt officials over the previous five years. This corresponds roughly with perceptions surveys from other EU countries. For example, Rudzitis (2002) cites survey evidence showing that 73% of survey respondents thought the Lithuanian customs service was dishonest but only 5% having actually paid a bribe to the survice. Kouzmina (2006) provides data from the Russian Federation showing that customs clearance delays are roughly on par with those of Ukraine and Moldova.

[12] These data contrast resoundingly with data from most Central and Eastern European countries. For example, in recent data looking at corruption perceptions in Romania, the largest number of respondents (66%) thought that in the customs service, “all or almost all officials in the agency are corrupt” – as opposed to 55% for police and 54% for the health service.

[13] See World Bank (2005b) for estimates for the bribe tax in Ukraine.

[14] The higher level of bribes for Moldova, which is a much poorer country than Ukraine, may seen unusual at first. However, the higher level is due to the much larger absolute level of imports into Moldova from Ukraine. As a check, the estimated levels of corruption affecting Moldova and Ukraine were compared with other corruption estimates related the level of overall corruption to GDPP – see for Moldova and Gorodnichenko and Peter (2006) for Ukraine.

[15] The calculation involves the most basic math. A probability (p) of an event, such as someone paying a border guard to allow an illegal crossing across a green border can be estimated, however rudely. The estimated amount of involved denoted as X, multiplied by the probability of the occurrence p results in a risk of a single unreported event as pX.  For the reader with a background in mathematics, the computation of the total yearly risk is equal to the probability of an occurrence of corruption at any particular site is ΣjΣi(piXi)where i=number of sites and j=the number of time periods or trails. 

[16] For a similarly minded (though more technically complicated) calculation of the size of bribery in Ukraine, see Gorodnichenko and Peter (2006) who estimate the total value of bribery in Ukraine at between 460 and 580 million USD.

[17] The relationship between corruption and contraband is complicated – with several officials noting that no direct evidence of corruption related to contraband can be found:

[18] Heroin Market Value Figure:  United   Nations, 2005. World Drug Report. Vienna: Office on Drugs and Crime, June 2005, pg. 132,

[19] Sources about contraband in Moldova provide much less reliable results. According to the UNODC (2005), roughly 0.1% of the Moldovan population uses cocaine (as opposed to 0.8% of the Ukrainian population). Given a population base of 4.5 million people, the average estimated use of cocaine is 45,000 people. 

[20] Rose-Ackermann  represents a useful vade-mecum for the economic rationale behind corruption.

21 These normal consumer goods are contrasted with contraband goods. The previously given formula -- ΣjΣi(piXi) where i=number of sites and j=the number of time periods or trails – can be used as a micro-economic check on this, fundamentally macroeconomic estimate. 

[22] The bribe tax, and the undervaluation estimate, assumes that the civil servant knows the true value of goods. Such an assumption is obvious because if the civil servant did not know the true value, he or she would accept the value claimed by the trader and no need for a bribe would arise.

[23] Indeed, the social harm is higher and can be calculated roughly as the value of earning lost for lives taken by guns and drug addiction.

[24] Indeed, the National Security and Defense Council of Ukraine has elaborated an Action Plan with monitoring mechanisms, actions to prevent corruption focused on the finance of political activity and the approximation of legislation to EU standards (which in this case revolves around compliance with the OECD and CoE conventions).

[25] For the reader unfamiliar with administrative procedure, national laws passed by the parliament – and in some countries by Presidential degree -- often give general instructions. Departments are subsequently responsible to devising internal regulations aimed at implementing these broad legislative projects.

[26] Unlike other expert reports, this report will try to avoid to the extent possible a discussion of regulations – preferring to focus on actual work practices for two reasons. First, as noted by a recent OECD evaluation, “while Ukraine has a rich array of legal instruments and broad strategic documents, efficient coordination, implementation and enforcement remain insufficient” (2). Second, specific regulatory deficiencies and recommendations for improvement are being made directly with the counterpart agencies. 

[27] A number of authors bemoan the lack of success in anti-corruption work. See Kaufmann (2006) for an empirical treatment of the anti-corruption industry’s progress. 

[28] Michael and Bowser (2005) covers the problems with the OECD’s work in the area as part of its Anti-Corruption Network for Transition Economies. Most advisors in the region have indicated that the OECD’s working methods, in the cover of the Network, have resulted in expensive and ineffective assessment. The work of the Network will certainly either be devolved to SIGMA (also in the OECD), to the EU or discontinued.

[29] EU Member States in adopting the acquis communautaire, ratify legislation aimed at fighting corruption. However, the acquis only requires EU Member States to ratify the OECD Convention on Bribery of Foreign Officials in International Business Transactions and the Council of Europe Conventions (ETS 173 and ETS 174 which establish criminal and civil liability for corruption offenses).   

[30] While the adoption of the criminal sanctions embodied in the OECD Convention may negatively impact on Ukraine and Moldova’s ability to detect and prosecute corruption cases, the existence of the Convention itself will certainly help both governments fight corruption. Under the OECD, businessmen from OECD countries who bribe Ukrainian or Moldovan officials are committing

[31] The weaknesses in the judicial systems of both countries is covered in the literature previously cited.

[32] In most EU countries, the accuser of a corruption offense has the burden of proof. In the UK, the burden of proof is reversed – namely the civil servant has the duty to show he or she did not participate in corruption. While the reversal of the burden of proof for allegations of administrative corruption are not appropriate to Ukraine and Moldova, the UK experience shows that reducing the required standard of proof may serve to deter corruption.

[33] See Print source for citation

[34] See Michael (XXXX) for a further description of the audit culture and its impacts of public sector service delivery.

[35] Foreign aid to Moldova and Ukraine aimed at fighting corruption will significantly distort the incentives for right-regulating against corruption as both governments will be paid to implement a number of punitive measures against corruption which will make the programme benefit in the short-term higher than the cost imposed by efficiency reducing regulation.

[36] For an easy to understand instruction sheet for implement a risk analysis approach in Customs, see World Bank. (2005).

[37] Current data, based on 100% inspection is probably unreliable due to poor inspection quality concomitant with large-scale checking.

[38] For more on this approach, see UN (2004).

[39] While not discussed in this article, the careful reader will immediate observe that a hazard function for corruption for each civil servant can be roughly estimated by the sum of the expected risks of legal infraction multiplied by the rent (or bXt as discussed previously).

[40] Because of the increasing adoption of New Public Management concepts in government, public audit is seen like private audit. In a private company, an audit department must frequently show that its activities result in a greater cost reduction or reduction in business risks than the audit ties up in staff and other costs. In the same way, an internal audit in a public sector organization should aim at creating more public value than it costs to conduct the audit. 

[41] This section covers two very extensive literatures very quickly. The reader unfamiliar with these academic approaches should consult the relevant citations given in this paper for a further background.

[42] Transaction cost economics originally concerned itself with the question of why firms integrated instead of using arms-length contracts Williamson (1985). Later work – particularly Jenson and Meckling (1990) -- using a similar logic, sought to derive the optimal number of departments in an organisational unit.

[43]  Print source for citation  notes, political transactions costs often represent a determinant transaction cost – particularly within a public sector context. Such political transactions costs – or the costs of obtaining settlements over differing views of public sector activity – are usually large relative to the cost of ‘contracting’ across organizational boundaries in the public sector.

[44] As an aside, this model explains other countries. Turkey’s public sector is extremely politically fractured, possessing a range of political and institutional interests – explaining why Turkish anti-corruption work has remained decentralized. In Bolivia, the transactions costs are extremely high for co-operating between under-funded ministries and few competencies existed in each service. However, instead of centralizing anti-corruption authority in the office of the Vice President (as the World Bank recommended and funded in 1997), work should have been concentrated in an independent anti-corruption unit.

[45] See  Print source for citation or an overview of the successes and failures of anti-corruption over the last 10 years.