Expectations, Distrust and Corruption in Education: Findings on Prevention through Education Improvement

In this article we present results from research on how education environments may influence the propensity of education participants to engage in corrupt practices. We approached this task with the help of a conceptual framework that draws on rational choice and routine activity theory, and on economic models of human behaviour. The framework guided a collection of evidence through desk research, numerous interviews and focus groups on behalf of authorities and civil society in five countries (Armenia, Kazakhstan, Serbia, Tunisia, Ukraine). Our findings suggest that circumstances in education influence problematic conduct in two major ways – by opening opportunities for the abuse of regular processes in education, and by providing education participants with incentives to engage in it. This contribution offers some insights concerning the incentives of education participants to participate in corruption and discusses why research on corruption can be a valuable source of guidance for systemic improvement in education.


Introduction
Education matters for the wellbeing of individuals and societies alike, but its significance makes it also vulnerable to corruption -a problem important enough to have been singled out as a reason for the failing of targets in the first decade of the global Education for All initiative 1 (World Education Forum, 2000), and also persistent enough to be on the anti-corruption agenda of numerous countries around the world today (Transparency International, 2013, p. xix;. This article discusses selected findings from an ongoing exploration of corrupt practices in education and the systemic circumstances which may provide education participants with reasons to engage in them. The work was initiated under the name of INTES (Integrity of Education Systems) in the context of the Anti-Corruption Network for Eastern Europe and Central Asia (ACN) of the Organisation for Economic Cooperation and Development (OECD) and included the development of a conceptual framework of education integrity, as well as the integrity research presented here. 2 Following an insight from international anti-corruption practice that sector-specific solutions can be more effective than measures that are imposed from the "outside" (Boehm, 2014), the objective was to gather insights into how corruption in education can be addressed from within the sector, through improvements in areas that are within the professional remit of education practitioners and policy makers, such as assessment of student achievement, teacher policies, admission procedures, etc.
The scope of "corrupt conduct" in this research includes both practices for which there is criminal liability as well as softer, sector-specific actions, which are harmful, but may not qualify as corrupt by international standards . "Education" and "education system" refer to all education providers in a country, regardless of their ownership or sponsorship (UIS/OECD/ EUROSTAT, 2002), which are accredited or licensed to grant credentials (certificates, degrees or diplomas) through formal learning and related services. This includes also entities which set the conditions of operation of education providers and monitor their performance, such as line ministries, regional education authorities, education inspectorates, quality assurance/accreditation agencies, testing centers, and similar bodies.
The research presented in this paper was carried out in five countries (Serbia, Tunisia, Armenia, Ukraine, and Kazakhstan) between 2011 and 2018, in which national education authorities and/or civil society organizations volunteered to participate as they were in the process of revisiting education reforms, advocacy campaigns, or national anti-corruption strategies. The work covered school and higher education, as well as pre-school education in Ukraine and Armenia, and followed the principles of OECD peer reviews, as described in (OECD, 2003): a formal request by national counterparts (authorities or civil society); the presence of review criteria derived from national and international sector commitments; clearly defined review/research procedures; and conclusions which are based on a consensus between the peersmembers of the respective research teams comprising national and international experts.
Our research focus was on the connections between actions of education participants which qualify as corrupt, and the systemic conditions in which they have their exposure to the education sector as teachers, students, school and university administrators, or external stakeholders. We inquired whether these conditions (education environments) can be drivers of specific forms of corrupt conduct and if yes, how.
Our starting assumption was that manifestations of corruption can be unequivocally linked to specific policy problems in the respective education systems (OECD, 2012), the resolution of which is possible through education improvement. Our findings suggest that the circumstances in which education takes place can indeed influence propensity for problematic conduct, in two major ways -by opening opportunities for the abuse of regular processes in education such as assessment, admission procedures, recruitment, etc. for corrupt purposes, and by providing education participants with incentives to do that.
In this article we offer a selection of insights concerning the incentives of participants in education to participate in corruption. The paper starts with an outline of the conceptual framework and methodology which guided the research, and then presents an overview of findings on how education environments may influence the motivation of education participants to engage in illicit or illegal practices. 3 The article concludes with an observation that corruption is not just a problem that calls for punitive responses, but that it also holds critical lessons that can guide systemic improvement.

Conceptual framework
Our research was guided by a conceptual framework which connects corrupt conduct, motivation and opportunity as the three variables of a causal model of corruption in education (INTES model). The model presumes that each integrity violation is linked to policy-related circumstances which motivate and facilitate the violation (Figure 1).
The aim of the research was to describe the three variables and explain how they are connected: how circumstances in education create incentives (motives) and opportunities for education participants to engage in integrity violations. The violations are practices in education which are intentional, system-wide, involve education participants, and deliver undue benefits . The INTES model draws on several theories. The first is the routine activity theory of Cohen and Felson (Cohen & Felson, 1979), according to which all offences are situational, caused by a combination of circumstances rather than by some inherent features of perpetrators. A crime can be committed by anyone who finds himself in such circumstances. More recent research confirms that professional environments too can play a role in motivating problematic conduct by affecting the commitment of individuals to the values and principles of their organizations (Doty & Kouchaki, 2015). In other words, the integrity of professionals depends not only on their character, but also on their organizational environment.
Our model draws also on elements of rational choice theory. Like the "homo economicus" who is consistent in making choices that are in his best interest (Friedman, 1966;Gruene-Yanoff, 2011), we posit that the "homo educationis" in focus of our research is consistent in going after his best interest as education participant and stakeholder. Given a choice, parents would always prefer to send their children to a good school, students would work towards successful transition to university or employment, teachers would opt for fair pay, education authorities would secure enough funding and presentable reform results, etc.
That said, the reference to rational choice in our model is not a commitment to delivering an accurate reflection of individual decision-making processes. Newer economic theories (i.e. behavioral economics) argue that this is more complex task than suggested by rational choice approaches, and that the outcomes of such processes can be difficult to predict. The reference is rather meant to point out the origin of our research hypothesis: that participants in education are consistent in their choice of priorities, preferences and beliefs, and that they always try to act accordingly.
In the same vein, we refer to education participants as collective identities (Melucci, 1989) and do not consider factors such as individual character traits, predispositions, or personal circumstances, despite the role they may be playing in motivating integrity violations. The collective identities are those of education professionals, students, external stakeholders (e.g. parents, guardians or equivalent, civil society, media), and representatives of national and regional education authorities.
Finally, to identify which practices in education qualify as integrity violations, the INTES model borrows from international anti-corruption practice in defining anticorruption offences, as reflected in major conventions against corruption such as the United National Convention Against Corruption (UNCAC) (United Nations, 2004) and the Council of Europe Criminal Law Convention on Corruption (Council of Europe, 1999a).
These Conventions forego an overarching definition of corruption and describe specific forms of corrupt conduct instead, i.e. bribery, illicit payments, obstruction of justice, etc., which all signatories agree are undesirable and must be prevented (Council of Europe, 1999b). The offences are distinctly different from each other, but they share common features. For instance, they all involve public sector officials, deliver undue personal benefits, are intentional, and states-signatories of the Conventions should define them as illicit or illegal. We used these features to set the criteria that helped us determine which practices in education qualify as integrity violations.

Methodology
Our inquiry into corruption and its system-wide context was guided by two questions: whether education participants engage in corrupt practices and if yes, how the systemic conditions in which they have (or deliver) their learning experiences may play a role in that. We approached these questions by dividing them in sub-questions and research routines, organized in three interconnected thematic protocols.
The focus of the first protocol was on the identification of sector-specific practices in each country, which qualify as acts of corrupt conduct against criteria such as prevalence, intent, involvement of official positions, and presence of undue personal benefits. The focus of the other two protocols was on the education policy environment in which each of these practices typically thrives, and specifically on the ways in which the education policy context may create incentives for corrupt conduct. We split that last query in two sub-questions: what education participants expect from their involvement in an integrity violation, and what prevents education from meeting their expectations in legitimate ways.
In each country, the implementation of the protocols took place in three phases. In the first phase in each country we relied on background questionnaires, literature review and on scoping missions for initial discussions with national authorities and stakeholders to consolidate a preliminary selection of integrity violations which the research will cover. On that basis we prepared an inventory of violations and factors in the education environment which may be promoting the proliferation of these violations.
The second phase was devoted to analysis of these factors through further desk research and site visits for individual and group interviews with education participants, and in some cases (pre-school education in Armenia) through focus groups. The purpose was to explore the reasons of participants in education to engage in violations and the ways in which they do it.
Finally, in the third phase we prepared a draft of a report with our findings, which stakeholders and national authorities in each country complemented with recommendations and validated before it was finalized and published.
In the second phase in most 4 countries we met between 150 and 200 counterparts, which we selected through purposive sampling. Table 1 provides an overview of their number by role in education and country, and Table 2 gives a break-down by level of education and, where applicable, by type of institution. In total in all countries, we met approximately 760 education participants. Notes: 1) PS: pre-school education; P: primary education; S: secondary education; HE: higher education. 2) I: One-on-one interviews; GRI: group interviews; FG: focus groups. The sampling was based on judgement by national education authorities and researchers, who were asked to propose a balanced selection of education participants according to criteria which included role in the education system (teachers, administrators, students, parents, education authorities), and key features of the education providers they are associated with, such as quality (well-performing/poor performing), education level (pre-school/school/higher education), type (e.g. general/specialised schools), geographical location (urban/rural, central/regional), and form of ownership (public/private). Once into site visits, we also used snowball sampling to complement the initial selection of counterparts, which sometimes became necessary due to the sensitivity of our topic.
The individual and group interviews were semi-structured (Given, 2008). In the first third to half of the time, they would follow a simple protocol with a set of predefined questions for the purpose of establishing comfort with interviewees by explaining the purpose of research, clarifying the role of interview counterparts in the education system, and inviting them to comment on statements collected from peers in previous interviews. The remainder of the interview would then progress as a normal conversation in which we would probe for in-depth information about actions and contexts based on facts collected in the desk research and prior interviews.
At the end of each day, the research team and national experts would hold debriefing meetings to compare the written notes from the interviews, clarify interview statements and agree on possible interpretations, and agree on fact-finding questions to be included in the subsequent interviews. Considering the number of participants in each country, this was time consuming, but it allowed us to apply the research protocols in a contextually sensitive way.
In some countries (Armenia for pre-school education), the evidence collection was complemented by focus groups in which we used a funnel approach (Strauss & Corbin, 1998) to converse with parents, civil society, teachers and principals. The discussions lasted three hours with each group and started with broader questions on undesirable practices, before moving on to practices which may qualify as integrity violations. This was followed by an in-depth exploration of which integrity violations emerge as more common and how. The focus group discussions were audio recorded, transcribed verbatim and translated in English.

Education deliverables at the center of corrupt conduct
The involvement of individuals in corrupt conduct could be explained in a number of ways, which may range from personal traits to human nature in general ("opportunity makes the thief"). As far as education is concerned, the findings of our research confirm that conditions in the sector play a major role as well.
Most violations we came across were actions which crime prevention theory would call "routine activities" (Cohen & Felson, 1979) -the final outcomes of processes that are closely related to the environment of perpetrators, a form of highly problematic response to gaps in the routines of that environment. We noted a number of stories like those of students who pay their school teachers for private tutoring because of the low quality of teaching in class (OECD, 2012;Milovanovitch, 2014;OECD, 2017a), of parents who had to bribe neighbourhood schools through "donations" in exchange for admission, because the schools were overfilled (OECD, 2017a), or of teachers who over-mark students to keep their jobs, in settings marked by arbitrary and nontransparent staffing decisions (OSF -Armenia, 2016).
These are not isolated cases, but examples of typical, education system-wide conduct, which took place in circumstances that are similar within and between countries. For instance, in all countries for which we had information, the practice of giving school students better grades than they deserve in exchange for favours, was particularly prevalent in regional contexts, around the end of the school year, and in the last years of secondary schooling.
The examples in our selection and numerous other cases that we recorded have one more common trait. They describe illicit ways of obtaining deliverables which education was designed to provide in legitimate manner, without corruption: good or at least fair school grades, education of acceptable quality, access to a good school in the neighborhood, adequate wages and secure employment, etc. Those who engaged in corrupt conduct in pursuit of such goals seemed to pursue their legitimate interests as education participants -an observation we made in all countries and for most integrity violations.
This doesn't mean that all partakers in education corruption are driven by a wish to secure education deliverables. Some of those who were abusing their professional position in order to provide such deliverables, were interested only in the monetary benefit. Yet, their benefit would always depend on whether they can deliver to unmet demand for an education deliverable. In Tunisia, for example, where at the time of our inquiry (2013) teachers were earning above the national average for occupations requiring similar qualifications, many teachers were providing paid private tutoring to their own students to earn even more, despite having a conflict of interest (Milovanovitch, 2014). The illicit practice was thriving only because the other side of the transaction, the parents of school children, felt that tutoring may provide two deliverables which public education cannot -effective learning in class and a fair chance to progress in school.
The range of deliverables which could motivate problematic conduct can be wide. Examples may include free schooling, fair access to higher education, good quality knowledge, credible graduation credentials, adequate employment conditions for teachers, transparent budgeting and resource allocations, etc. Early on in our research, we therefore summarized them through the lens of past and present global commitments in education, such as the UN Education for All initiative or the Sustainable Development Goals, in three broad areas of stakeholder demand: equity and access to education; quality of education; sound management, in particular of staff and financial resources (OECD, 2012). Based on results from two of the most recently analyzed countries (Armenia and Ukraine), Table 3 illustrates how various violations (Column 1) and the expectations for education deliverables that they serve (Column 2) fall in these four areas of demand (Column 3), as well as who is demanding and supplying the education deliverables (Columns 4 and 5).  Table 3 shows a diversity of expectations for education deliverables behind each violation, but also that the education deliverables at stake in each revolved around similar things: education that is accessible, equitable, of good quality, and provided in education environments that are managed properly, particularly with respect to staff and resources.
The Table also shows that sometimes education deliverables drive the motivation of all sides involved in an integrity violation, including of those who abuse their professional position. In Ukraine, private tutoring was a key source of income for school teachers who would have otherwise earned below the subsistence minimum, while for parents it was a way to ensure that their children master the curriculum and prepare for graduation tests in subjects of importance. In Armenia, teachers were giving grades to students who were absent from school (mostly to attend private tutoring in preparation for university admission) as a favor to parents who were donating for maintenance and teaching materials. The benefit of parents and students in this arrangement was additional time to prepare for the highly competitive university entrance exams.

Violations in response to distrust and failed expectations
If education participants engage in integrity violations to obtain otherwise legitimate education deliverables, we also wanted to understand why they see corrupt instead of integral conduct as the better way to attain these deliverables. We therefore inquired about the circumstances in which an integrity violation appears to be the more promising course of action than respecting the rules.
In the examples described in Tables 3 and 4 (below), those concerned revealed that they did not believe that education will deliver as required. Their education environments were characterized by distrust in the capacity of the sector to address their expectations in legitimate ways and so, they felt compelled to find alternative solutions to secure the education deliverables they need. For some, the distrust and the illicit actions were the result of direct experiences with policy failure and its consequences. It is interesting to note that others who did not experience such failures first-hand, shared the distrust and engaged in integrity violations as well, as a form of pre-emption.
In Serbia and Tunisia, for example, parents believed that payments to school teachers in the form of supplementary tutoring are the best way to ensure student success (OECD, 2012;Milovanovitch, 2014), although none of their children was ever failed because of a refusal to take tutoring classes with a school teacher. In Ukraine, parental donations were a well-established way to secure a place in a sought-after school or pre-school (OECD, 2017a;Milovanovitch & Bloem, 2019). Most of the parents who believed that admission without such "voluntary" donations is impossible, were in fact never asked for such donations by the principals, but offered them nevertheless. In Armenia, where the license of private education providers to operate depends on licensing and inspection processes which are known to be unpredictable and corrupt, principals tend to seek "connections" to the officials in charge and offer them favors of their own accord (Milovanovitch & Bloem, 2019).

What prevents the provision of education deliverables in permissible ways?
The last sub-question in our protocol on incentives inquired how an education system can fail expectations to an extent where for a seemingly critical mass of people who participate in education, illicit conduct appears as a justifiable alternative to integrity.
In countries where, for example, illicit access to education was a widespread remedy for shortages in admission capacity, we inquired about the reasons for the capacity shortages in the school network. In places where, like in Armenia, teachers told us that they inflate student grades to preserve their employment, we asked about their employment conditions and what in those conditions might be fueling their conviction that their jobs are otherwise at risk. In Kazakhstan, where we established that the allocation of vacated public scholarships in universities is plagued by irregularities and integrity violations (OECD, 2017b), we tried to understand why faculties and their members persistently handle decisions about such allocations in an illicit and nontransparent way.
We received numerous answers to such questions. Some of them were on macro level and went well beyond the education sector, pointing to the overall economic situation in a country (Ukraine, Tunisia), the legacy of a recent past (Armenia, Serbia), or the financial crisis and the decline in the price of natural resources, which affected the purchasing power of households (Kazakhstan). Another group of responses were on a more intermediate level of aggregation and provided account of broad problems in the education sector, such as low quality of teaching and low status of the teaching profession, decline in student motivation ("bad students"), or low spending on education. Finally, the remaining responses (the majority, in fact) directed us towards specific shortcomings in the education environment of our counterparts, where obvious shortcomings in policy and practice were limiting the capacity of education to deliver precisely in the areas affected by stakeholder distrust and integrity violations.
For instance, the contextual analysis of intentional grade inflation in Armenia and Ukraine exposed problems in the professional circumstances of teachers, which seemed to explain some of their readiness to participate in the integrity violation. One of these problems was the persistently low level of public funding for school maintenance, which made public schools financially dependent on parental contributions and gave parents a leverage over school staff. Parents used that leverage to pressure teachers into a preferential treatment of students (Milovanovitch & Lapham, 2018). The precarious employment conditions of teachers were another common problem, which put them in a vulnerable position vis-à-vis principals who could fire them at will (OSF -Armenia, 2016;OECD, 2017b). In some countries (Ukraine) teachers engaged in the integrity violation also in the hope of improving their sub-standard income through gifts from grateful parents at the end of the school year (OECD, 2017a). Table 4 shows further examples like these. It lists integrity violations (Column 1), the policy areas affected by shortcomings which lead to these violations (Column 2), and the policy shortcomings themselves (Column 3). Column 4 presents the outcomes of the contextual analysis through which we linked the shortcomings, deliverables, and violations from the point of view of education participants who engage in them. It is interesting to note how similar shortcomings seem to motivate similar violations in different country contexts. Oversized networks of public schools and scarce resources for their maintenance lead to increased financial involvement by parents. Public education providers are ill-equipped to manage the parental donations, which facilitates embezzlement. Shortage of pre-school places can be traced back to lack of planning and foresight by municipalities and regional authorities regarding the provider network, which in turn creates situations conducive to corrupt transactions in exchange for access, especially in large cities and the capitals. Obstacles to effective teaching and learning in class, such as overloaded curriculum and lack of time or else, fuel demand for private tutoring which teachers often provide to their own students from regular education, and so on.

Conclusion
In this article we discussed some of the connections between corrupt practices in education and the conditions in which people partake in them as teachers, students, administrators, and parents. We outlined a conceptual framework and methodology which identifies and describes such practices and then investigates the typical circumstances in which they take place. The results we described drew on findings from research in five countries and illustrated how corrupt actions revolve around education deliverables which are in short supply. We provided examples of how violations can be traced back to shortcomings in education policy and practice, which fuel stakeholder distrust in the ability of education to deliver to expectations in legitimate ways, and motivate various forms of problematic conduct.
Our findings suggest that corruption in education does not necessarily call for measures that differ from ongoing reform processes in education. Corruption problems can be approached and interpreted as specific education policy problems, so that corrupt conduct is addressed through improvement in common areas of education policy, not only through stricter rules and better enforcement. If the motivation of education participants to engage in corruption depends on education deliverables, then this motivation could be influenced through education improvement which targets the expectations at stake, for example by broadening access to the type of education and education providers in demand, by improving teacher policies, increasing transparency in resource allocations, and so on.
Some of the policy areas in focus of our research are sensitive and complex and changing them might be difficult. However, many countries have already embarked on reforms in areas which are difficult and of significance to integrity, such as teacher policies, funding for education, curriculum or assessment. Their reform plans are usually regularly updated, which could offer an opportunity for recalibration on the basis of findings from research like ours, as the INTES methodology that we used is in the public domain.
Designing or updating reforms in this way could make them more relevant for stakeholders by ensuring that they address the professional contexts in which expectations and corruption risks emerge. It would also allow countries to advance against corruption in education as part of their ongoing reform agenda in education, which is likely to be more effective than stand-alone anti-corruption plans. If aimed at the right selection of problems, education reforms can therefore become a viable corruption prevention strategy -one that is within the remit of education professionals and involves also those who participate in the integrity violations.

Mihaylo Milovanovitch is senior policy specialist with the European Training
Foundation, Italy and senior integrity analyst for the Center for Applied Policy and Integrity, Bulgaria. His research and publications focus on education policy, corruption prevention, and integrity of education in countries of Eastern Europe, Central Asia and Northern Africa.
* The views and opinions expressed in this article belong to the author and do not necessarily reflect the official policy or position of the organisations he is affiliated with. The author wishes to thank the expert network at the Center for Applied Policy and Integrity and Dr. Iva Kosutic from the Institute for Social Research in Zagreb for the valuable advice provided in the process of preparing this contribution.

Notes
[1] The Education for All (EFA) movement is a global commitment to provide quality basic education for all children, youth and adults. It is led by UNESCO.
[2] For more information about the Network see https://www.oecd.org/corruption/acn/aboutthenetwork/. We have provided an extended description of the INTES initiative and methodology in (Milovanovitch, 2013) and . Sections of this article draw on these papers, in particular on (OECD, 2018).
[3] For simplicity, in this article we may use adjectives such as "illicit", "illegal", "problematic" and "corrupt" as well as nouns such as "corruption", "integrity violation", "transgression" and "illicit/illegal conduct" interchangeably, although in legal terms they may imply different forms of liability for actions (criminal, civil, administrative, or disciplinary).
[4] Due to time and project constraints, in Tunisia and Kazakhstan the site visits were shorter and the number of interviews more limited.