Characteristics Affecting Anti-corruption Behavior in Indonesia

DOI: 10.21532/apfjournal.v6i1.203 ABTRACT The problem of corruption in Indonesia at this time is nothing new in the midst of Indonesian people. Low integrity and permissive culture of corruption are suspected to be the cause of the development of corrupt practices. The Indonesia Anti-Corruption Behavior Index (ACBI)) compiled by Statistics Indonesia (Indonesia: Badan Pusat Statistik/ BPS) is used to obtain a complete picture of the current situation and conditions of community anti-corruption behavior. This study aims to determine the characteristics that differentiate Anti-Corruption Behavior Index (ACBI)scores. The analytical method used is descriptive and non-parametric analysis to determine differences in ACBI seen from other dimensions and characteristics such as residential area, education level, and age group. In general, ACBI in Indonesia has continued to increase from 2012 and reached the highest moment in 2020, amounting to 3.91 in the category of very anti-corruption. The dimensions and characteristics of the area of residence and the level of education have significant differences in ACBI, while the age groups do not have differences in ACBI. Policies in an effort to improve anti-corruption behavior of the community need to pay attention to dimensions and characteristics that have significant differences so that the policies implemented can be right on target.


INTRODUCTION
Corruption comes from several languages, such as "corruptio" in Latin, "corruption" or "corrupt" in English, "corruption" in French, and "coruptie" in Dutch. Corrupt means bad or rotten, uses his power for his own benefit and likes to accept bribes and other meanings of a bad nature. Thus, corruption is a bad act such as embezzlement of money, receiving bribes, and so on (Setiadi, 2018).
The current issue of corruption is no longer a new thing in Indonesian society. This can be seen from the widespread corruption that has occurred so far. In Law of the Republic of Indonesia Number 20 of 2001 concerning Amendments to Law Number 31 of 1999 concerning the Eradication of Corruption Crime, it is stated that corruption is not only detrimental to state finances, but also a violation of social and economic rights. Therefore, the eradication of the criminal act of corruption must be carried out in an extraordinary way.
Currently, corruption has spread to the lower class society so that it occurs not only in government administrators or public officials and people's representatives, but also but also in the lower levels of society. Low integrity and a permissive culture towards acts of corruption are thought to be the causes of the development of corrupt practices. Every time a corruption case is disclosed, this is considered an achievement in law enforcement. From a cultural perspective, this condition is the tragic side of the corruption mentality which is unstoppable (Santoso. Et al., 2014). Cultural problems are the cause of the never-ending eradication of corruption.
The government has issued Presidential Decree number 55 of 2012 concerning the long-term National Strategy for Prevention and Eradication of Corruption (Indonesia: Stranas PPK) for 2012-2025 and the medium-term National Strategy for Prevention and Eradication of Corruption 2012-2014 in order to accelerate efforts to prevent and eradicate corruption. The National Strategy for Prevention and Eradication of Corruption was later refined into the National Strategy for Prevention of Corruption. The National Strategy for Prevention of Corruption itself contains a focus and targets in accordance with the needs of preventing corruption. Thus, the prevention of corruption can be carried out in a more measured, focused and direct manner so as to create a just and prosperous society (Presidential Regulation of the Republic of Indonesia Number 54 of 2018 concerning the National Strategy for Prevention of Corruption).
Based on the indicators in the fifth strategy, increasing education and an anti-corruption culture, the government, through theStatistics Indonesia (Indonesia: Badan Pusat Statistik / BPS), measures the Anti-Corruption Behavior Index (Indonesia: Indek Perilaku Anti Korupsi / IPAK). This fifth strategy is considered important because the development of corrupt practices is thought to stem from the low integrity of the perpetrators. In addition, the causes of the development of corruption practice, one of which is the persistence of a permissive culture towards corruption. The mindset and culture of society that encourage social structures to behave in corruption must be changed so that society can be free from corrupt behavior and uphold integrity. For this reason, active behavior from the community is needed to prevent corrupt behavior in the environment.

LITERATURE REVIEW AND HYPO-THESIS Corruption
Corruption is a word that has many meanings. Literally, corruption means rottenness, ugliness, depravity, dishonesty, bribery, immorality, deviation from chastity, words or utterances that are insulting or slander and so on. In the Indonesian dictionary, corruption is misappropriation or embezzlement of state or company money and so on for personal or other people's gain. According to the Law of the Republic of Indonesia No. 31 of 1999 as amended by Law no. 20 of 2001 concerning the Eradication of Corruption, corruption is an attempt to enrich oneself or another person or a corporation in a way that is against the law and can harm state finances or the country's economy (Ramli, 2017).
Thus, corruption can be interpreted as an action that can harm many people and even a nation. This is because acts of corruption only benefit a person or group of people, while something that is corrupted is someone else's property. Committing acts of corruption is the same as stealing and seizing other people's rights, which are not only in the form of money but also goods and services.
The definition of corruption can be divided into 3 (three) categories (Kurniawan, 2009): • Public office centered definition, which means deviating from the supposed responsibility as a public officer because of personal interests (family, close friends), because they expect financial gain or status; or breaking the rules by exploiting personal influence. • Market-centered definition, which means a corrupt civil servant considers his office as a business and generates as much income as possible for himself. The office then becomes a unit to be maximized. • Public interest centered definition, which means that a pattern of corruption can occur when a power holder, who has the responsibility to do something, in fact supports or takes actions that are only in accordance with the wishes of the person who gives money, or gifts which are not allowed. This can damage the public interest.
According to Kurniawan (2009), there are at least six direct causes ofcorruption: • Arrangement and authorization; • Taxation; • Expenditure / budget policies; • Provision of goods and services below market prices; • Other discretionary policies; and • Political party financing.
Meanwhile, there are at least six indirect causes of corruption: • Bureaucratic quality; • Salaries in the public sector, • Punishment system; • Institutional supervision; • Transparency of rules, laws and processes; and • Leader's role model According to Shah, corruption in the public sector will depend on a number of factors (Kurniawan, 2009). Efforts to tackle corruption without considering these factors will lead to less profound and unsustainable results. These factors include: • Quality of public sector management; • The nature (condition) of the accountability relationship between government and society; • Legal framework; and • The level of public sector processes which is complemented by transparency and information dissemination.
From the explanation regarding the definition of corruption and the factors that cause corruption, in general, it can be concluded that corruption has an impact on life in society, either socially, economically, politically, or environmentally. Thus, it is necessary to create anti-corruption attitudes or behavior.

Anti-Corruption Behavior Index (ACBI)
In 2012, the government issued Presidential Regulation No. 55 on the long-term National Strategy for the Prevention and Eradication of Corruption 2012-2025 and the medium term 2012-2014 in order to accelerate efforts to prevent and eradicate corruption. Then, in 2018, the National Strategy for the Prevention and Eradication of Corruption was refined into the National Strategy for Prevention of Corruption as stipulated in the Presidential Regulation of the Republic of Indonesia Number 54 of 2018 concerning the National Strategy for the Prevention of Corruption. The National Strategy for the Prevention of Corruption is in accordance with the need for corruption prevention which contains the focus and targets for preventing corruption. Thus, corruption prevention can be carried out in a more measured, focused, and directly impacted manner To meet data needs, from 2012 to 2020 (except in 2016) the Statistics Indonesia (Indonesia: BPS) conducted an Anti-Corruption Behavior Survey (ACBS), which aims to measure the level of public permissiveness towards anti-corruption behavior by using the Anti-Corruption Behavior Index (ACBI). This survey only measures people's behavior in petty corruption and does not cover grand corruption. The coverage of the data collected relates to opinions on habits in society and experiences related to public services in terms of bribery, extortion, and nepotism behavior.
The Anti-Corruption Behavior Index calculated by the Statistics Indonesia (BPS) is built from two dimensions: perception and experience. The perception dimension is built from the sub-dimensions of family, community, and public. Meanwhile, the sub-dimensions of public experience and other experiences are part of the experience dimension. The selection of the Anti-Corruption Behavior Index constituent variables is based on a number of questions using Explanatory Factor Analysis (EFA).
In full, the stages in calculating the Anti-Corruption Behavior Index (ACBI) 2020 are as follows: • Cleaning data.
• Recording variables • Calculating the weight of each indicator with Principal Component Analysis (PCA). • Calculating sub-dimension index, dimension index, and Anti-Corruption Behavior Index.
The index is the weighted average of all the answers to the index constituent variable with a standardized weight each with the following formula: Based on the calculation results using the formula above, the Anti-Corruption Behavior Index has a value range of 0-5. The index value which is getting closer to 5 indicates that the anti-corruption behavior of the community is getting higher. The Anti-Corruption Behavior Index scores can be grouped into 4 categories as follows:

Research Framework
The author developed the research by determining the characteristics that differentiate the ACBI score, which is seen from the dimensions and other characteristics such as residential area, education level and age group. Schematically, the research framework can be seen in the following figure 1.

Research Hypothesis
This research was conducted by developing a hypothesis in general, it is assumed that the ACBI scores are different in each dimension and characteristic so that the policy of understanding anti-corruption behavior can be applied on target with the partial hypothesis as follows: Ha1: Perception and experience dimensions have significant differences in ACBI. Ha2: The characteristics of urban and rural living areas have significant differences in ACBI. Ha3: The characteristics of the education level have significant differences in ACBI. Ha4: Characteristics of age groups have significant differences in ACBI.

METHOD Types of Research and Sources of Data
This type of research is quantitative. Quantitative research is carried out using secondary data sources which are useful for developing mathematical models and testing hypotheses so that research conclusions can be obtained. The secondary data used in the study is the Anti-Corruption Behavior Index (ACBI) from 2012 to 2020 (except 2017) obtained from the Central Bureau of Statistics of the Republic of Indonesia (BPS) based on other dimensions and characteristics.

Data Analysis Method
The analytical method used in this research is descriptive analysis and nonparametric difference analysis with the Mann-Whitney test for characteristics with two groups and the Kruskal-Wallis test for characteristics with more than two groups. Non-parametric difference analysis is used because this study has a small sample size so it will be easier to calculate than using the parametric method. In addition, non-parametric methods are also widely used to identify data in the field of social science (Supranto, 2002) so that it is better used to analyze anti-corruption behavior. Data processing is carried out using data processing tools in the form of SPSS 25 and excel software.

RESULT AND DISCUSSION Overview of the Anti-Corruption Behavior Index (ACBI) Based on Dimensions
Anti-Corruption Behavior Index is prepared based on two dimensions: the Perception Dimension and the Experience Dimension. The perception dimension is in the form of an assessment / opinion of the anti-corruption behavior in society. Meanwhile, the Experience Dimension is in the form of anti-corruption experiences that occur in society (Figure 2). There was an increase in the Perception Index from 2012 to 2018. However, it decreased in 2019 and 2020. This shows a decrease in public understanding and assessment of anti-corruption behavior. On the other hand, in the experience index, there was fluctuation from 2012 to 2019. In line with the experience index, the ACBI score also showed fluctuation. In 2020, the ACBI score was 3.84. This figure is higher than the ACBI 2019 (3.70) (BPS, 2020). From 2012 to 2019, Indonesia had an ACBI score in the anti-corruption category, but in 2020 it increased to the very anti-corruption category, with ACBI score of 3.91 which was the highest moment since the ACBI was first calculated.
Despite the increase, the achievements in 2020 are still quite far from the target. In 2019, the Indonesia's ACBI is targeted to be at a score of 4.00. This shows that there needs to be a lot of improvement from the perspective of the community and government institutions, especially in terms of anti-corruption knowledge. (BPS, 2020)

Based on Living Areas
The living areasare classified into two, urban area and rural area. The following is the ACBI based on the area where the respondent lives: Based on the graph above, in general both urban and rural ACBI scores increased along with the increase in ACBI as a whole. The ACBI scores for urban areas are relatively higher than the ACB scores for rural areas. This indicates that there are differences in anti-corruption behavior for urban and rural communities. In 2020, the urban ACBI score was 3.87, while the rural ACBI score was 3.81.

Based on the Levels of Education
Respondents' education is classified into 3 groups: junior high school and below, senior high school, and above senior high school. The following are ACBI scores based on the levels of education: In general, ACBI for all levels of education increased until 2017. However, in 2018 ACBI for the level of senior high school education or above senior high school education decreased. Meanwhile, ACBI for junior high school education level or lower experienced an increase. However, ICBI for junior high school education level or below was lower that the ACBI for higher education levels. This indicates that there is a difference in ACBI for each level of education of the respondent. The higher the level of education, the better the anticorruption behavior which can be seen from the higher ACBI score.

Based on Age Group
The age group of respondents is classified into 3: less than 40 years, between 40 and 59 years, and 60 years or more. The following are ACBI scores based on age groups: In general, ACBI for all age groups increased along with the increase in ACBI as a whole. In the years prior to 2019, the ACBI for the age group 60 years and over had significant differences compared to the younger age group. However, after 2019, the ACBI for the age group 60 years or older had a relatively the same score as other age groups.
A non-parametric test is then carried out to find out the statistical difference between ACBI scores in terms of dimensions or characteristics, such as area of residence, education level, and age group. The test results will show whether there is sufficient statistical evidence of a difference in the ACBI score or not

Non-Parametric Analysis Based on Dimensions
Since the dimensions of the ACBI consist of two groups, the Mann-Whitney test statistic is used to analyze the difference with the following hypotheses: H01: M perception =M experience or the dimensions of perception and experience do not have a significant difference in ACBI The results of statistical processing using SPSS 25 are as follows (Table 2).

Source: Processed data
Based on the test results above, H0 is rejected because the probability value is 0.009 lower than the 0.05 significance level. So it can be concluded that there is a significant difference between the perception dimension and the experience dimension. This difference is in accordance with the previous descriptive analysis. These differences are signs for policy makers to pay attention to the dimensions of perception and experience in an effort to increase anti-corruption behavior in society.
According to Fatkhuri (2017), survey data 2015 indicates that people are increasingly idealistic against corruption but it is not in line with real behavior in everyday life. This is reflected in the index depicted by the dimension of perception which experienced an increase in 2015, while the index depicted by the dimension of experience decreased. This data shows that basically public awareness of the importance of preventing and eradicating corruption is getting better. However, this positive trend has not been followed by public behavior not to commit corruption. Since 2018, ACBI has actually experienced a different trend so that people today tend to consider corruption a common thing because the index depicted by the perception has decreased, but the real Source: Processed data behavior carried out by the community is in line with anti-corruption behavior. Therefore, to harmonize anti-corruption perceptions and behavior in society, a policy that takes into account this dimension is needed. Anti-corruption behavior needs to be improved from the sub-dimensions of the family, community and public as well as the sub-dimensions of public experience and other experiences, so that the targets of the policies can be achieved.

Based on the Living Areas
Since the living areas consist of two groups, the Mann-Whitney statistical test is used for the difference analysis with the following hypothesis: H02: M urban =M rural or the characteristics of urban and rural living areas do not have a significant difference in ACBI. Ha2: M urban ≠M rural or the characteristics of urban and rural living areas have significant differences in ACBI. The results of statistical processing using SPSS 25 are as follows (Table 3).
Based on the test results above, H0 is rejected because the probability value is 0.004 or lower than the 0.05 significance level. So it can be concluded that there are significant differences in ACBI between urban and rural areas. This difference is in accordance with the previous descriptive analysis. Although based on the ACBI score the urban area is always higher than the rural area, in 2020 the scores for both are almost the same. However, this does not mean that urban communities are more anti-corruption than rural communities or less corrupt than rural communities. This is because there is a trend that shows that petty corruption acts involving bribery, extortion and nepotism have actually increased (Candra, 2017). These differences are signs for policy makers to pay attention to areas where they live in an effort to increase anti-corruption behavior in society. Thus, the provision of material for socialization and introduction to anticorruption behavior needs to be adapted to the area where the community lives, so that policy targets can be achieved.

Source: Processed data
Based on the test above, the decision taken is to reject H0 because the probability or significance is lower than the significance level of 0.05, which is 0.004. So it can be concluded that there are significant differences in IPAK between urban and rural areas. This difference is in accordance with the previous descriptive analysis. Although in terms of IPAK scores for the City area are always higher than that for the Village area, in 2020 the values for both have almost the same value. However, this does not mean that urban communities are more anti-corruption than rural communities or less corrupt than rural communities. This is because there is a trend that shows that petty corruption acts involving bribery, extortion and nepotism have actually increased (Candra, 2017). These differences are signs for policy makers to pay attention to areas where they live in an effort to increase anticorruption behavior in society. Thus, the provision of socialization materials and the introduction of anti-corruption behavior need to be adapted to the area where the community lives,so that the policy will be more targeted.

Based on the Levels of Education
Since the levels of education consist of more than two groups, the Kruskal-Wallis statistical test is used to analyze the difference with the following hypothesis: H03: M ≤junior high school =M senior high school =M>senior high school or the characteristics of the level of education do not have a significant difference in ACBI. Ha3: There is at least one M_i ≠ M_j where i ≠ j and i, j are ≤junior high school, senior high school, and> senior high school The results of statistical processing using SPSS 25 are as follows (Table 4).

Source: Processed data
Based on the test results above, H0 is rejected because the probability value is 0.000 which is lower than the significance level of 0.05. So it can be concluded that there are significant differences in ACBI for each level of education. This difference is in accordance with the previous descriptive analysis. This is also in accordance with the statement of Darto (2012) that the higher the education level, the higher the ACBI score. These differences are the signs for policy makers to pay attention to education levels in an effort to increase anti-corruption behavior in society. Thus, the provision of socialization materials and the introduction of anti-corruption behavior needs to be adjusted to the education level of society, so that the policy target can be achieved.

Based on the Age Group
Since the level of education consists of more than two groups, the Kruskal-Wallis statistical test is used to analyze the difference with the following hypothesis: H04: M <40 years =M 40-59 years =M ≥60 years or the characteristics of the age group do not have a significant difference in ACBI. Ha4: There is at least one M_i ≠ M_j where i ≠ j and i, j are <40 years, 40-59 years, and ≥60 years. The results of statistical processing using SPSS 25 are as follows (Table 5).

Source: Processed data
Based on the test results above, H0 is not rejected because the probability value is 0.066 which is higher than the significance level of 0.05. So it can be concluded that there is no significant difference in ACBI score for each age group. The absence of this difference is in accordance with the previous descriptive analysis which shows that there is no significant difference in ACBI for each age group, especially from 2019 to 2020. This is in accordance with the statement of Darto (2012) that the anticorruption spirit between the old and the young people does not differ significantly.

CONCLUSION
Based on the results of the analysis it can be concluded as follows, 1) from 2012 to 2019 Indonesia had an ACBI score in the anti-corruption category, but in 2020 the score increased to the very anticorruption category, or 3.91 which is the highest moment since the first ACBI was calculated.
2) The results of difference test to the dimensions that make up ACBI show that there are significant differences between the dimensions of perception and the dimensions of experience.
3) The results of difference test to other characteristics are as follows: a) Urban and rural areas have a significant difference in ACBI scores where urban areas have higher ACBI scores than rural areas; b) The education level has a significant difference in ACBI scores where the higher the level of education, the higher the ACBI score; and c) The age group has no a significant difference in ACBI scores. Despite the increase, the achievements in 2020 are still quite far from the target. In 2019, Indonesia's ACBI was targeted to be at a score of 4.00. This indicates that there is still a need for a lot of improvement in both society and government institutions, especially in terms of knowledge about anti-corruption. There are differences in the dimensions of the formation of ACBI and other characteristics, such as the area of residence and level of education. These differences are signs for policy makers to pay attention to the dimensions of perception and experience as well as the area of residence and level of education in an effort to increase anti-corruption behavior in society. From a practical aspect, the results of this study are expected to be of benefit to policy makers in efforts to eradicate corruption by increasing anticorruption behavior in Indonesia starting from the lowest level. The results of this study are expected to provide benefits for policy makers in the field of eradicating corruption in the realm of implementing policies to increase anti-corruption behavior for the community, especially in relation to its role in increasing anticorruption behavior starting from a small level in Indonesia. This study is limited to the availability of research data from 2012 to 2015 and from 2017 to 2020 only, while data for 2016 is not available.