V. Coalition 2000 corruption indices: methodology and THEORETICAL approach
1. General Notes
The corruption indices of Coalition 2000 are a system of synthetic indicators, which present the results of the quantitative surveys of the CMS (S1 and S2) and the media monitoring (S6) in a summarized form. Implicitly the corruption indices also summarize the results of the qualitative surveys (S3, S4 and S5), as the latter are used for the elaboration and the improvement of the methodology of the quantitative surveys.
The main objective in constructing the corruption indices is to reduce the multidimensionality of corruption as a social phenomenon to a reduced set of synthetic indicators. The advantages of such an approach are at least the following:
• synthetic indicators (corruption indices) facilitate the public presentation of the results, thus making analysis easier to perceive;
• the employment of synthetic indicators is a prerequisite for establishing time series and respectively for analyzing and assessing change.
2. Theoretical base of the study of the elements of corrupt behavior
The four types of corruption indices are based on a relatively simple theoretical reconstruction of the elements of the social action:
• social action has its specific prerequisites, among which the more important are social actors’ attitudes, interiorized values and the way actors perceive their social environment;
• social action itself presupposes a specific interrelationship of actors in which they exchange the resources they possess in order to achieve a specific objective;
• action results include certain specific characteristics: 1) they change or preserve the initial (pre-action) prerequisites; and 2) they leave a specific “trace” in the social environment (change or preserve its structural components);
• the fulfillment or non-fulfillment of the objectives generates expectations among actors about the character and the structure of social action in its subsequent cycle. Based on these expectations, each actor constructs or changes his behavioral strategy.
Applied to corruption, this model of social action could be reduced to the following basic components:
Action Components Components of Corruption Behavior
Action Prerequisites Attitudes towards corruption
Include the identification of corruption as a social phenomenon, the assessment of its normative (value) permissibility and the degree of willingness to override the norms of legitimate social behavior
Actor Interaction Corrupt practices
Include the activity of the actors, connected with creating a situation for corrupt practices (the exercising of pressure) and the actual act of corrupt behavior.
Action Results Assessments of the magnitude of corruption
Include the assessment of the level of proliferation (involvement) of public officials in different forms of corrupt behavior as well as the assessment of the levels of transformation of corruption into a behavioral norm (into a socially effective instrument for solving personal problems).
Expectations Assessment of the perspectives on corruption
Include assessments of the capabilities of society (its potential) to combat corruption.
It should be especially noted that the application of the general model of social action (even in its simplified form) to corruption (in constructing the methodology of the quantitative studies) is, for understandable reasons, subject to a multitude of restrictions. In this respect the model on which the study of corrupt behavior is based is much simpler than it could have been. In constructing the set of indicators that was actually employed, the objective was to use the indispensable minimum in a way that would make it possible to maintain professional standards of fieldwork.
3. Theoretical Interpretation of Corruption Indices.
Corruption is a complex and multidimensional phenomenon and is associated with several differing definitions. The operational definition adopted defines corruption as abuse of power (economic, political, and administrative) in the interest of personal or group gain and at the expense of the individual, specific groups or society as a whole. This rather broad definition is warranted by the character of the phenomenon itself as well as by the wide-scope initiatives against corruption on the national and international level, including both citizens and governments.
The reproduction of corruption presupposes the existence of four necessary components: 1) state/public officials; 2) discretionary power; 3) abuse of public power; 4) private gain on behalf of officials. Depending on the forms in which these components actually function, there could exist different forms, levels, spheres and mechanisms of corruption.
Corruption manifests itself mainly in the interactions between the public sector on one side, and citizens and private business on the other. There are two basic types of corruption: Grand corruption, which involves top state officials, politicians and business people and refers to the allocation of substantial resources; and Petit corruption, which usually includes lower-level public officials and refers to the daily interactions between them and citizens and businesses (small and medium size). This second type is more widespread and is associated with smaller payments and/or a system of favors and gifts.
Having in mind the specific objectives of corrupt behavior, two specific cases of corruption should be considered. First, abuse of power for private gain in cases where officials are obliged to provide a certain service by law. This type of corruption, known as “greasing the wheels,” is targeted at the faster or more expedient delivery of services, or greater safety in the resolution of the problem. A second case is when an official provides services/rights to which the citizen (business) is not entitled by law, or even services that are a direct abuse of the law.
From an economic point of view the proliferation of corruption follows the classic market principles of supply and demand: a larger demand and a larger discretionary power of officials produces an environment that facilitates and encourages corruption. The value of the bribe depends on the expected profit or benefit. In this respect corrupt behavior could be regarded as mutually beneficial economic transactions. However, these transactions eliminate the rules of competition and the legal regulations and thus distort market principles and criteria for efficient economic action and decision making.
The definition of corruption as a negative social phenomenon allows several assumptions to be made concerning the interpretation of the indicator included in the CMS:
• in an ideal state of society (the practical absence of corruption), corruption attitudes should assume minimal values; i.e., citizens should perceive corrupt behavior as morally inadmissible and they should not be inclined to compromise their moral values. Intensity of corrupt behavior should also be minimal, and corrupt behavior should be rated low as an effective problem-solving practice.
• it would be logical to assume that the existence of a certain level of tolerance of corrupt behavior (moral admissibility) would have several consequences. First, readiness to compromise would increase as the level of tolerance increases. Second, the frequency of practical acts of corruption would also increase with the level of tolerance (due to the “natural intensity” of social conformity). Third, in cases when the frequency of corrupt behavior surpasses the level of “single isolated cases” it is likely that assessments of the pervasiveness of corruption would substantially surpass the frequency of practical acts of corrupt behavior (only one case of “taking a bribe” would be sufficient to socially label an official as “corrupt”). Fourth, the existence of a social environment where acts of corrupt behavior exceed the level of “single isolated cases” is very likely to produce the perception that corruption is a socially normal environmental component. The chances of that perception being firm and widespread increase with the limited implementation of sanction mechanisms.
One of the basic theoretical assumptions for the construction of the CMS is that it is more important to track the dynamics of corruption in several dimensions than to analyze its initial/current values. Because corruption has been identified as a problem to Bulgarian society, it would be important to assess its gravity. However, it is more important to know dynamics: whether corruption is evolving in the positive or in the negative direction in comparison with its initial baseline values.
Corruption indices provide an approximation about the scope and aspects of corruption based on the assessments of citizens and public officials. These assessments are the starting point for their practical behavior and the way they perceive their social environment. Corruption indices could not be a base for making direct conclusions about the exact level of proliferation of corrupt practices.
Because the index of personal involvement in corrupt practices is based on the anonymous admissions of respondents, it comes closest to indicating the “level of proliferation of corruption.” Also, to a certain extent the specific legal characteristics of corruption (that both sides act illegally) make this index one of the few realistic measures of the actual level of proliferation of corruption.
Currently, in comparison with the information available from law enforcement institutions, the accuracy level of empirical survey estimates of the realities of corruption is (for obvious reasons) substantially higher.
The CMS includes four types of indices:
4. Structure and Conceptual Interpretation of Corruption Indices
Attitudes towards Corruption
1. Principle Acceptability of Corruption
The index reflects the degree to which Bulgarians accept, at the level of values, certain acts of corrupt behavior on behalf of members of Parliament and public officials. Using a 3-point scale the admissibility of 12 types of practices with different degrees of “corruptness” are assessed.
2. Susceptibility to Corruption (general public)
This index measures the inclination to compromise on values, principles and legality in order to perform corrupt acts, such as giving or accepting money and/or gifts for the purpose of solving certain personal problems. The index is based on two independent questions describing situations in which there are two possible types of behavior: giving/taking money or gifts, and refusal.
3. Susceptibility to Corruption (public officials)
This index reflects the existing attitudes among public officials to accept money or gifts (to participate in corrupt practices) under certain conditions: providing additional services and performing duties faster than normal. Assessments are rated on a 3-point scale.
1. Corruption Pressure on the General Public
This index shows the degree to which the citizens are subject to direct or indirect pressure to participate in corrupt practices with public officials. It accounts for cases in which public officials wanted or showed they were expecting corrupt behavior from the citizens and/or their families.
2. Corruption Pressure on Public Officials
This index reflects the frequency of attempts to corrupt public officials with the purpose of solving personal problems. The frequency is assessed on a 4-point scale.
3. Personal Involvement in Acts of Corrupt Behavior (general public)
This index reflects self-confessions about cases in which citizens have offered public officials money or gifts. The frequency of these cases is registered on a 4-point scale.
4. Personal Involvement in Acts of Corrupt Behavior (public officials)
This index reflects the degree to which public sector officials accepted money or gifts when dealing with citizens’ personal problems as part of their professional obligations.
Magnitude of Corruption
1. Spread of Corruption (according to citizens)
This index reflects citizens’ assessment of the spread of corruption in the country. It can also be computed for specific groups of public officials. The spread of corruption is measured by two indicators using 4-point scales.
2. Spread of Corruption (according to public officials)
This index reflects the extent to which citizens attempt to resolve their personal problems by offering money or presents to public officials. The frequency of these cases is estimated using a 4-point scale.
3. Practical Effectiveness of Corrupt Behavior
This index shows the extent to which citizens perceive corruption as an efficient tool for solving personal problems. It is based on the registered probability of citizens offering money and/or gifts in order to successfully resolve their problems.
Expectations about the Future of Corruption
This index reflects the expectations of the general public about the capacity of society to curb corruption in the country. Expectations are rated on a 4-point scale.
5. Method of Computation of Corruption Indices
Two types of indicators have been used in constructing the corruption indices:
(a) Indicators measured on three- or four-point scales
(b) Indicators measured on dihotonomous scales
The method used to construct indices using indicators of the first type includes the following steps:
1. A rank (ranging between 0–2 or 0–3) is assigned to each value of indicator.
2. Each rank is weighted by the percentage of respondents who have chosen the respective answer option (excluding the “don’t know” and “no answer” categories).
3. The value of the indicator is computed as a sum of the weighted ranks.
Example: The index for the spread of corruption is constructed based on the following question:
According to you how widespread is corruption in this country?
One answer only.
Rank Valid percent
1 Almost all state officials are involved in it 3 19
2 Most state officials are involved in it 2 42
3 Only a few state officials are involved in it 1 38
4 Hardly any state officials are involved in it 0 1
9 DK/NA — —
I = 3õ0.19 + 2õ0.42 + 1õ0.38 + 0õ0.01 = 1.79
The value of the index ranges between 0 and 3. The closer this value is to 3, the more widespread is corruption (according to respondent assessments).
With the second type of indicators indices could assume values between 0 and 1. Their specific values are computed in the following way:
1. The two answer options are assigned the ranks of 0 and 1.
2. The value of the index is equal to the percentage of respondents who have chosen the answer option ranked 1.
Example: The index of susceptibility to corruption is constructed based on the following question:
Imagine that you are a state official on a low salary and a citizen who comes to you with a problem offers you money or a gift. Would you:
One answer only.
Rank Valid percent
1 Be tempted to accept and provide 1 40
a better service
2 Be offended by the proposal 0 60
8 Neither/ It depends — —
9 DK/NA — —
I = 1õ0.4 = 0.4
The closer the value of the index is to 1, the higher the susceptibility to corruption.
In order to construct aggregate indices, the values of the individual indices are normalized by adjusting their values to fall into the range 0.00 — 1.00. Normalized values are then summed. For example, the index for the spread of corruption is normalized by dividing its current value (1.79) by its maximum value (3.00), obtaining its normalized value (0.60). The values of all aggregate indices thus range between 0.00 and 1.00. This value is then recalculated to fit into a scale ranging between 0 and 10 (or between 0 and 100).
6. Information base for computing the 1998 corruption indices
The aggregate corruption indices have been constructed based on the information collected in two of the surveys of Vitosha Research: (S1) — the quantitative national representative survey of the population aged 15 and over as of June 1998, and (S2) — the quantitative survey of public officials of July 1998.
The aggregate indices presented have been computed for the population as a whole, but could also be computed for specific population categories. The baseline information and values for the indices are presented in the publication Coalition 2000 Corruption Indices. Baseline values of indices, however, do not reveal the full explanatory potential of corruption indices. Their full potential will be utilized in the next series of surveys when analysis of the dynamics and comparisons will be made possible. In their first issue, corruption indices have only one value and provide the baseline characteristics of public consciousness and corruption behavior in the beginning of the monitoring period (June-July 1998)