Social Sciences
Volume 4, Issue 6, December 2015, Pages: 134-141

Determinants of Corruption in Kenya: Born and Bred to Bribe

Odhiambo Fredrick Onyango

ResearchPro Solutions, Department of Research & Evaluations, Nairobi, Kenya

Email address:

To cite this article:

Odhiambo Fredrick Onyango. Determinants of Corruption in Kenya: Born and Bred to Bribe.Social Sciences.Vol.4, No. 6, 2015, pp. 134-141. doi: 10.11648/

Abstract: Corruption is an epidemic in Kenya. Major corruption scandals have been reported since the early 90’s. These include the Turkwel Hydroelectric Power Station scandal (1986 – 1990), the Goldenberg scandal (1990 – 1999), the Grand Regency scandal in 2008, and the Triton Oil scandal in 2009 among numerous others. Despite the attempts to fight corruption, the war has never been won. While a number of studies have examined the determinants of corruption in order to offer policy recommendations to fight corruption, individual-level factors have not been exhaustively examined especially for developing countries like Kenya where international corruption indices paint a grim picture. Moreover, the studies have mostly been based on perception of individuals and not the actual payment of bribe. This study sought to assess the individual factors that influence individuals to pay bribes in Kenya. The study uses survey data from Afrobarometer Round 5 survey. The probit analysis shows that corruption in Kenya is influenced by gender, race, ethnicity, religiosity, employment status, and education while age, religion and location were not significant determinants of corruption. The study therefore concludes that a number of individual-level factors explain the likelihood to be corrupt suggesting that some individuals may be born or bred to bribe. To address corruption in Kenya, policy makers should include individual-level determinants of corruption in policy formulation efforts as they are just as important as other factors in explaining corruption.

Keywords: Corruption, Bribe, Afrobarometer, Religiosity, Kenya

1. Introduction

Corruption is one of the top three constraints to the developing and emerging economies (Shehu, 2005) although it occurs in all countries, regardless of levels of social and economic development (Stapenhurst & Langseth, 1997). It is common to read about corruption cases in Kenya right from the National Government to the County Governments. Kenya has performed dismally in the Corruption Perception Index (CPI) surveys conducted by the Transparency International. In the latest CPI 2014 survey, Kenya was position 147 out of 177 countries in the world making Kenya the second most corrupt country in East Africa after Burundi and 35th in Sub-Saharan Africa (Transparency International, 2014a). An Afrobarometer Perception of Corruption Index (PCI) saw Kenya as the sixth most corrupt country in Africa (Richmond & Alpin, 2013).

With a current population of over 40 million people and over 42 ethnic groups and tribes, Kenya is a collection of complex ethnic, religious, cultural and linguistic nationalities. Kenya has enjoyed a GDP growth averaging 5% for the last few years and with the discovery of oil, the growth is projected to rise. Recently, Kenya’s GDP was rebased catapulting it to a lower middle income economy (Kenya National Bureau of Statistics, 2014). In fact, Bloomberg ranked Kenya’s growth rate third in the world. Despite these glowing statistics, about half of Kenyans still live below the poverty line (Njonjo, 2013).

While public opinion polls show most respondents citing corruption as one of the top problems facing Kenya, there is schizophrenia in voicing of concerns as the same people do not hesitate to be involved in corrupt practices (Bardhan, 1997). For instance, the 2013 Global Corruption Barometer showed that 70 percent of Kenyans believed that corruption is a problem (Transparency International, 2014b) yet most of them still paid bribes (Transparency International, 2013). In a report to the National Anti-Corruption Campaign Steering Committee on the state of corruption in Kenya, it was noted that about 27 percent of Kenyans felt that corruption was the most important issue the country needed address (Strategic Public Relations and Research Limited, 2005).

Surveys have indicated that corruption varies across countries and that, even within countries, some public agencies are more prone to corruption than others (Stapenhurst & Langseth, 1997). In Kenya, the police, the judiciary, and the public health institutions have been cited among the most corrupt institutions in the past surveys (Transparency International, 2013). Some individuals are also more likely to be more corrupt than others. In a recent Afrobarometer Survey, 56% of Kenyans had paid a bribe in the past year – the fourth highest in Africa (Richmond & Alpin, 2013).

Studies have examined general determinants of corruption for countries and for individuals using perception indices yet very few have used the actual bribery as a measure of corruption. Given the findings of (Richmond & Alpin, 2013), this study seeks to investigate what factors explain an individual’s likelihood to pay a bribe. The study focuses on specific demographic factors that relate the individuals in order ascertain whether corruption can be said to be inborn or nurtured. The findings show that corruption is influenced by gender, race, ethnicity, religiosity, employment status, and education. This paper is organised as follows. The next section reviews literature on corruption especially the theoretical and empirical issues. This is followed a research methodology which outlines the data and the model. The results are then presented followed by a discussion and conclusion of the study.

2. Literature Review

There is no single comprehensive and universally accepted definition of corruption as it is a relative concept and varies over time and place (Shehu, 2005). To define corruption, one has to be confined to the purpose for which it is intended. Most scholars who have studied corruption define it as the use of public office for private gains (Bardhan, 1997; Treisman, 2000; & Stapenhurst & Langseth, 1997). This is a fairly broad definition. The United Nations Convention against Corruption (UNCC) suggested that for purposes of criminalisation and enforcement, states adopt two definitions. One of the definitions suggested is "the promise, offering or giving, to a public official, directly or indirectly, of an undue advantage, for the official himself or herself or another person or entity, in order than official act or refrain from acting in the exercise of his or her official duties" (Shehu, 2005, p.70).

Stapenhurst & Langseth (1997) distinguish between two types of corruption – petty and grand corruption. Petty corruption is practiced by civil servants who may be grossly underpaid while grand corruption involves high public officials who make decisions involving large public contracts. The Kenya Anti-Corruption and Economic Crimes Act of 2003 defines corruption as a benefit, that is, an inducement or reward for, or otherwise on account of an agent, the receipt of expectation of which would tend to influence an agent to show favour or disfavour (Strategic Public Relations and Research Limited, 2005). This definition is adopted here as it fits within the purpose of this study.

The motivation to study corruption stems from the consequences of corruption especially for developing countries. While Bardhan (1997) noted that in the context of pervasive and cumbersome regulations, corruption improves efficiency and helps growth, most scholars find that corruption has negative effects on economies. Stapenhurst & Langseth (1997) found that corruption leads to market misallocation and inefficiency. According to McAdam & Rummel (2004), corruption has wide-ranging detrimental effects including lower levels of private investment and growth, capital flows and currency crises, inferior resource allocation, distorted government expenditure and revenues, higher income inequality and poverty, inflation, and lower standards in public life. Shehu (2005) further noted that corruption is a threat to stability of societies and the establishment and maintenance of the rule of law, and obstructs sustained economic development. In a study on Kenya, (Kimuyu, 2007) found that businesses paid an average of 7 percent of their annual sales on unofficial payments including kickbacks on government contracts as others paid as high as 60 percent. Further, the study revealed that the consequences of corruption in Kenya were deleterious to businesses as it undermined their growth and reduced their propensity to export.

The focus of this study is to diagnose the causes of corruption. Generally, a number of reasons have been given by both scholars and practitioners. On why countries have differing levels of corruption, Bardhan (1997) noted that one of the reasons that has been given by liberal economists, though valid but inadequate, is differences in regulatory states of economies. Poverty has also been shown as a determinant of corruption as governments especially in poor countries fail to pay a living wage to public servants (Stapenhurst & Langseth, 1997). The study by Rieckenghem and Weder 1997 (cited in Leite & Weidmann (1999) found that under certain circumstances, higher wages in the public sector deter corruption by increasing the potential loss in case of detection. Leite & Weidmann (1999) also showed that the extent of corruption depends on natural resource abundance, government policies, and the concentration of bureaucratic power.

According to Treisman (2000), religion – protestant traditions – influenced corruption. Shehu (2005) on the other hand found that poverty, unemployment, ethnic and religious dichotomies caused of corruption. A study by Shaw (2009) found that perceptions on corruption influence the actual bribing behaviour. The study also found that women tend to have higher probability of bribing. Rabl & Kuhlmann (2009) noted that rationalisation may possess potential determinant of corrupt behaviour. In yet another study, trust was found to have a mutual causality with corruption (Morris & Klesner, 2010). Personal characteristics such as gender, marital status, education, labour, religion, and perception on democratic state of a nation have also been found to influence individual’s perception on corruption (Melgar, Rossi, & Smith, Perceptions of Corruption in a Cross-Country Perspective: Why are Some Individuals More Corruption Than Others?, 2010). While these factors have been identified in some studies, no study has been conducted on the Kenyan environment to ascertain the determinants of corruption in Kenya. The present study seeks to contribute in this regard by examining the individual factors that may explain the likelihood for individuals to engage in corruption.

Scholars have also offered ways in which corruption can be fought. Salifu (2008) noted that no single institution can be used to control corruption and efforts to control it need to come from multiple fronts. To curb corruption especially in public administration, an incentive pay structure is usually cited as the most effective method (Bardhan, 1997). In some countries like Singapore, a wage premium above the private sector salaries is given in order to deter public officials from engaging in corruption as the potential cost of job loss on detection stiffen their resistance to temptation for corruption. Svensson (2005) also agrees that wage incentives can reduce bribery only under certain conditions. Stapenhurst & Langseth (1997) offers a number of public sector anti-corruption strategies including enactment and commitment to observe clear ethical codes, improved remuneration for civil servants, administrative reforms, disclosure of income/assets/gifts, policy and programme rationalisation, improved procurement procedures, and establishment of watchdog agencies such as anti-corruption agencies, ombudsman, and supreme audit institutions like Auditor General.

3. Methodology

3.1. Sample and Data

The study uses Kenya’s Afrobarometer Round 5 data collected in 2012 by Afrobarometer – an African-led, non-partisan research network that conducts public attitude surveys on democracy, governance, economic conditions, and related issues across more than 30 countries in Africa. The sample size for the Afrobarometer survey was 2,399 respondents spread across all the 47 counties in Kenya. In this study however, the data was cleaned to remove all responses that were coded as "missing", "don’t know" and all the negative values. After this exercise, the final sample used in this study was 2,197 observations.

3.2. Measurement of Variables

Corruption was the dependent variable in this study. While several studies in corruption have used the perception-based measures, there is a string of literature that find perception-based measures of corruption useless (Zaman & Ur-Rahim, 2009) (De Maria, 2008). Such criticisms of measures of corruption based on perceptions informed the use of actual bribery in this study to measure corruption. The specific question in the questionnaire was Q61: "in the past year, how often, if ever, have you paid a bribe, give or gift, or do a favour to government officials in order to [obtain a service." A binary variable was constructed from the responses and if any responded noted that they had indeed paid any bribe, it was coded as 1, otherwise 0.

The independent variables used in the study are education, employment, gender, age, religion, religiosity, location, race and ethnicity. Education measured the quality human capital available in the country. Question 97 asked "what is the highest level of education you have completed?" The ordinal choices in the question were used in the regression model to test the effect on corruption. Employment assessed whether the respondents were employed or not. This was Q96 of the survey and asked for the "employment status" of the respondents. In this study, employment status was measured as a binary variable of 1 if the respondents was employed either part-time or full-time and 0 otherwise.

Gender has also been shown to influence corruption. In this study, gender was used as an independent variable as was found in the questionnaire as Q101. In this study, gender was coded as a binary variable of 1 if the gender of the respondent was male and 0 otherwise. Age of the respondent was Q1 in the questionnaire. This was a continuous variable and was left so in this study. Religion measured the religion of the respondents. This was question Q98a in the survey questionnaire. A number of choices were offered in the questionnaire. In this study, however, religion was broken down into Christians, Muslims, and others. Religiosity measured the religion was important to the respondent. This was question Q98b of the questionnaire.

The study also used location of the respondent as an independent variable. This measured whether the respondents was from an urban or rural primary sampling unit through question. Race was measured as the race of respondent in question Q102 in the survey questionnaire. In this study, race was coded as a binary variable of 1 if the respondent was Black/African and 0 otherwise. Ethnicity was question Q48 in the survey questionnaire. In this study, ethnicity was measured as belonging to the five largest tribes of Kikuyu, Luhya, Luo, Kamba or Kalenjin. Thus, if a respondent belonged to the group, a response of 1 was coded, otherwise 0.

3.3. Model

Given that the dependent variable in this study is categorical, the model falls within the categorical dependent variable regression models (CDVMs). Unlike the ordinary least squares (OLS) models, these CDVMs are nonlinear. There are four CDVMs namely binary response, ordinal response, nominal response, and event count data. All these models use the maximum likelihood (ML) estimation methods. Since the dependent variable in this study was a binary variable, only two models can be used – logit or probit models. Following the practice in corruption studies, I used the probit model which has been used by several other scholars (Barr & Serra, 2006; Melgar & Rossi, 2008; Shaw, 2009; Melgar, Rossi, & Smith, 2010; and Lee & Guven, 2013). Stata statistical package is used to analyse the data in this study.

BRIBE = α + β1EDU + β2AGE + β3LOC + β4EMPL + β5GEND + β6REL + β7RELSITY + β8RACE + β9ETH + ɛ

Where BRIBE is bribery, EDU is education level of respondents, LOC is the urban/rural location, EMPL is status of employment, GEND is gender, REL is religion in which a respondent belongs, RELSITY is the religiosity of individuals, RACE is the race of the respondent and ETH is the ethnicity or tribe of respondent.

4. Findings

4.1. Descriptive Results

First, I carried out a descriptive analysis on the data. This was meant to give a descriptive pictorial of the main variables in the study. The variables analysed are bribery (corruption), education, employment status, gender, religion, religiosity, location, employment, and ethnicity. The descriptive results are presented in Table 1. The results showed that 63 percent of the respondents had paid a bribe, 40.37 percent had secondary education and 44.84 percent of the respondents were employed either part-time or full-time. The mean age of the respondents was 35.9 years with a range of 18 – 93 years (Fig. 1). The results also showed that 89.9 percent of the respondents were Christians and 90.53 percent of the respondents felt that religion was very important to them. The results further showed that 61.95 percent of the respondents were from the rural areas, 99.32 percent were of Black/African race and 66.91% belonged to the five largest tribes.

Table 1. Descriptive Statistics.

Variables Percent
Paid bribe 63.40
Not paid bribe 36.60
No formal schooling 6.6
Primary schooling 35.23
Secondary schooling 40.37
College/University schooling 17.79
Employment status  
Unemployed 55.16
Employed 44.84
Male 49.93
Female 50.07
Christians 89.90
Muslims 9.77
Others 0.33
Not at all important 0.46
Not very important 1.92
Somewhat important 7.10
Very important 90.53
Urban 38.05
Rural 61.95
Black/African 99.32
White/European 0.14
Coloured/Mixed race 0.14
Arab/Lebanese/North African 0.41
Kikuyu 20.07
Luo 12.56
Luhya 14.97
Kamba 10.83
Kalenjin 8.47
Others 33.09

4.2. Probit Regression Results

Table 2. Individual determinants of corruption in Kenya.

Variables Probit Model
Gender 0.190*** (-0.0569)
Race 0.694** (-0.347)
Location -0.00205 (-0.0602)
Age -0.00119 (-0.00223)
Ethnicity -0.215*** (-0.0685)
Employment 0.275*** (-0.0593)
Christian 0.452 (-0.503)
Muslim 0.162 * (-0.513)
Religiosity -0.668*** (-0.223)
Informal schooling only 0.867** (-0.381)
Some primary schooling 0.294** (-0.136)
Primary school completed 0.318** (-0.134)
Some secondary school 0.422*** (-0.143)
Secondary school completed 0.361*** (-0.135)
Post-secondary qualifications 0.600*** (-0.149)
Some university 0.267 (-0.243)
University completed 0.154 (-0.214)
Post-graduate 0.215 (-0.468)
Constant -0.49 (-0.625)
LR χ2 102.99***
Pseudo R2 0.0357
Log likelihood -1391.4203
Observations 2,197

Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

Table 2 presents the results of the probit regression. The results showed that gender had a positive and significant effect on corruption. The coefficient results show that the z-score of the likelihood to pay a bribe increased by 0.19 when the respondents were male. This suggests that male respondents were more likely than women to pay a bribe. The results also showed that race was positively related with corruption. The z-score of the likelihood to pay a bribe increased by 0.694 when the respondents were of Black/African race than when they were any other race. The results were significant at 0.01 level. This suggests that being Black was associated with higher likelihood of paying a bribe.

Related to race, the influence of ethnicity on the likelihood to pay a bribe was also assessed. As the results show, the study found a negative and significant association between ethnicity and corruption. Belonging to one of the big five tribes reduced the z-score of likelihood to pay bribe by 0.215. This suggests that the likelihood of a person paying bribe decreases if one belongs to one of the largest five tribes in Kenya. In other words, those that do not belong to the largest five tribes are more likely to pay a bribe.

One of the components of religion – religiosity - was found to have a negative effect on corruption. The z-score of the probability to pay a bribe increased by 0.668 for the respondents who agreed that religion was important to them. The results suggest that people who perceived religion as important to them were less likely to pay a bribe as compared to those that did not find religion important to them. However, being a Muslim or a Christian was insignificant in explaining the likelihood of individuals to pay a bribe.

The influence of employment status on corruption in Kenya was also assessed. The results showed a positive relationship between employment status and corruption. Being employed raised the z-score of likelihood to pay a bribe by 0.275. Thus, the chances of paying a bribe increased among the individuals that were employed as opposed to those who were unemployed.

On education, the study showed that all levels of schooling up to post-secondary education (but below the degree) were associated with an increase in the likelihood to pay a bribe. Those with some university education up to post-graduate education had non-significant effects on corruption. These results suggest that those with less than degree level of education were more likely to pay a bribe for some services. The results showed that age and location were insignificant in explaining the likelihood of the respondents to pay a bribe.

Figure 1. Age of the respondents.

5. Discussion and Conclusion

This study investigated the individual determinants of corruption among individuals in Kenya. The sample of 2,197 respondents was drawn from the Round 5 of Afrobarometer survey that was conducted in 2012. Since the dependent variable was binary, probit model was employed as the primary method of analysis as has been used by other scholars who have studied determinants of corruption.

The study found that gender influenced corruption as male were more likely than women to pay a bribe. These results are consistent with Pázmándy (2011) who found that men perceive less corruption that women. Since they perceive less corruption that women, they are therefore more likely to be corrupt than the women.

The study also showed that corruption was influenced by race as Black/African race was associated with more likelihood to pay a bribe than being a member of any other race. This is consistent with Wantz & McNally (2015) who found that Blacks support corrupt politicians of their own race at a higher rate than do whites or Latinos. However, this may only be true for non-educated or less educated Blacks as the study also showed that education had a negative effect on the likelihood of Black respondents supporting a corrupt politician. The influence of education on the relationship between race and corruption was not tested in this study and therefore no conclusions can be drawn to that effect.

Ethnicity also influenced corruption. It was noted that the likelihood to pay a bribe was associated with tribes that were not among the Big 5 tribes in Kenya. This may suggest that persons belonging to the smaller tribes in Kenya are disadvantaged in getting government services and must therefore bribe officials, who are mostly members of the largest tribes, in order to get the services. While no study had examined before the influence of dominant tribes on corruption, these results mirror those of Treisman (2000) who revealed that ethnic division had a positive and significant effect on corruption when the regression did not control for development but turned and insignificant when controlled for development.

The results showed that belonging to a specific religious grouping did not influence corruption but religiosity (the importance of religion to an individual) had a significant influence on corruption. Thus, people who perceived religion as important to them were less likely to pay a bribe. The findings are consistent with Shadabi (2013) who found that belonging to a specific religion did not influence corruption. The results on religiosity are also consistent with Peiffer & Rose (2014) who noted that corruption was influenced by religious contacts.

The study showed that employment status influenced corruption as the employed were more likely to pay a bribe than the unemployed individuals. This is consistent with Melgar & Rossi, 2008) who revealed that the unemployed were less likely to pay a bribe. This can be explained by the fact that the employed have the means (income) to pay up bribes in order to access some services they might require from the government such as acquiring passports and other government documents.

Education also influenced corruption as the results showed that the respondents with some education but no degree were more likely to pay a bribe. Thus, less educated individuals are more likely to be corrupt as opposed to other groups. This is consistent with Peiffer & Rose (2014) who found that education influenced corruption. The results can be explained in the context that those with less education require most favours in getting employment and are therefore more likely to bribe their ways in order to get the jobs. This has indeed in witnessed during police and army recruitment exercises where job applicants bribe to get employed in the forces.

Age did not influence corruption. This is consistent with Pázmándy (2011) who revealed that age did not influence corruption perception. This can be explained by the nonlinear effects of age as different scholars such as Melgar & Rossi (2008) have found that age influences corruption on different age levels. Peiffer & Rose (2014) also found that agwas only significant at older age. Urban location was found to have a non-significant effect on corruption suggesting that urban location does not influence corruption in Kenya. This is also consistent with Pázmándy (2011) who found that there were no significant differences in the way individuals who lived in various locations perceived corruption.

This study concludes that individual-level factors are significant determinants of corruption in Kenya. Specifically, the likelihood to pay a bribe increases for the men, Black/African descent, employed, and schooling of less than degree. Further, the likelihood to pay a bribe decreases when an individual is from the Big5 tribes and feels that religion is important. This shows how the fight against corruption can be hard to win given that some of the issues that influence corruption are inborn and cannot therefore be changed. The study recommends that policy makers should integrate individual-level policies with the macro-level policies for fighting corruption in Kenya. As such, the level of schooling should be improved by ensuring higher transition rates from high school to colleges and universities. While other factors such as race, ethnicity and gender cannot be changed, those that can be addressed such as schooling and religion can be targeted by policy makers.

Table 3. Empirical Studies on Individual Determinants of Corruption.

Author(s) Data and Model Key Findings
Treisman (2000) Transparency International’s (TI’s) indices for 1996,1997 & 1998; Business International; International Country Risk Guide (ICRG); Gallup International; OLS; cross-national survey Countries with protestant traditions were less corrupt
Svensson (2005) Control of Corruption (CC) Index; TI’s CPI; ICRG; International Crime Victims Survey (ICVS); OLS; cross-country survey Corrupt countries have lower levels of human capital stock (years of schooling for those aged over 25). Wage incentives can reduce corruption under certain conditions.
Barr & Serra (2006) Primary survey data; Probit; cross-country survey Corruption is a cultural phenomena
Frechette (2006) ICRG, TI, and World Competitiveness Report (WCR); FE; cross-country survey Income and schooling influence corruption
Kanold (2007) Afrobarometer Survey Data; OLS; national survey Corruption perception is influenced by urban residence. Other individual factors were non-significant.
Melgar & Rossi (2008) International Social Survey Program (ISSP); ordered probit; national survey Corruption is influenced by age (18-39), education level (high school, university), work in private enterprise, self-employment, unemployment, religion, and contacts.
Shaw (2009) A Partnership for Transparent Society survey data; probit; national survey Bribery is influenced by gender (women), past behaviour, corruption perception, and occupation of the father.
Morris & Klesner (2010) Americas Barometer Survey; 3SLS, SEM; national survey Perception of corruption is influenced by institutional trust
Melgar, Rossi, & Smith (2010) ISSP; ordered probit; cross-country survey Corruption is influenced by being a woman, education level, marital status, attendance to religious services, self-employment, opinions towards political system, sector of employment, and employment.
Pazmandy (2011) Eurobarometer 2009; OLS; cross-country survey Corruption is influenced by average years of education at country level and social status and unemployment at individual level.
Churchill, Agbodohu, & Arhenful (2013) World Bank’s control of corruption index; FE; cross-country survey Determinants of control of corruption are ethnic diversity, political stability, economic freedom, press freedom, and urban population.
Lee & Guven (2013) European Social Survey; SUR probit; cross-country survey Past experiences with corruption affects actual act of bribery.
Shadabi (2013) CPI; OLS; cross-country survey Religion does not influence corruption
Peiffer & Rose, (2014) Afrobarometer data from 18 countries; logit model; survey Individual influences of corruption were lived poverty, education, urban location, female, older age, and religious contacts.


  1. Pázmándy, M. (2011). Socio-Economic Influences on Corruption Perception – Empirical Evidence from 27 European Countries. Hamburg Review of Social Sciences, 5(3), 52-80.
  2. Bardhan, P. (1997). Corruption and Development: A Review of Issues. Journal of Economic Literature, XXXV(September), 1320-1346.
  3. Barr, A., & Serra, D. (2006). Culture and Corruption. Oxford: Global Poverty Research Group.
  4. Churchill, R. Q., Agbodohu, W., & Arhenful, P. (2013). Determining Factors Affecting Corruption: A Cross Country Analysis. International Journal of Economics, Business and Finance, 1(10), 275-285.
  5. De Maria, W. (2008). Measurements and markets: deconstructing the corruption perception index. International Journal of Public Sector Management, 21(7), 777-797.
  6. Frechette, G. R. (2006). A Panel Data Analysis of the Time-Varying Determinants of Corruption. New York: New York University.
  7. Kanold, C. (2007). Perceived corruption, public opinion, and social influence in Senegal. Cape Town: Afrobarometer.
  8. Kenya National Bureau of Statistics. (2014). Information on the Revised National Accounts. Nairobi: Kenya National Bureau of Statistics.
  9. Kimuyu, P. (2007). Corruption, firm growth and export propensity in Kenya. International Journal of Social Economics, 34(3), 197-217.
  10. Konold, C. (2007). Perceived Corruption, Public Opinion, and Social Influence in Segenal. Afrobarometer.
  11. Lee, W.-S., & Guven, C. (2013). The Influence of Cultural Values and Contagion Effects at the Micro Level. Bonn: IZA.
  12. Leite, C., & Weidmann, J. (1999). Does Mother Nature Corrupt? Natural Resources, Corruption and Economic Growth. IMF Working Paper, 1-34.
  13. McAdam, P., & Rummel, O. (2004). Corruption: A Non-Parametric Analysis. Journal of Economic Studies, 31(6), 509-523.
  14. Melgar, N., & Rossi, M. (2008). Perception of Corruption in Uruguay: the Effects of the Sector of Employment, Life-Course Adjustments and Education. Revista Economia Y Administracion, 70, 41-52.
  15. Melgar, N., Rossi, M., & Smith, T. W. (2010). Perceptions of Corruption in a Cross-Country Perspective: Why are Some Individuals More Corruption Than Others? Economia Aplicada, 14(2), 183-198.
  16. Morris, S. D., & Klesner, J. L. (2010). Corruption and Trust: Theoretical Considerations and Evidence from Mexico. Comparative Political Studies, 4(10), 1258-1285.
  17. Njonjo, K. S. (2013). Exploring Kenya's Inequality: Pulling Apart of Pooling Together? Nairobi: Kenya National Bureau of Statistics & Society for International Development.
  18. Peiffer, C., & Rose, R. (2014). Why do some Africans pay bribes while other Africans don't? Nairobi: Afrobarometer.
  19. Rabl, T., & Kuhlmann, T. M. (2009). Why or Why Not? Rationalisation Corruption in Organisations. Cross Cultural Management: An Introduction, 16(3), 268-286.
  20. Richmond, S., & Alpin, C. (2013). Governments Faulter in Fight to Curb Corruption: the People Give Most a Failing State. Nairobi: Afrobarometer.
  21. Salifu, A. (2008). Can Corruption and Economic Crime be Controlled in Developing Countries - and if so, it the Cost Worth It? Journal of Money Laundering Control, 11(3), 273-283.
  22. Shadabi, L. (2013). The Impact of Religion on Corruption. The Journal of Business Inquiry, 12, 102-117.
  23. Shaw, P. (2009). The Determinants of Educational Corruption: The Case of Ukraine. Fairfield University.
  24. Shehu, A. Y. (2005). Combating Corruption in Nigeria - Bliss or Bluster? Journal of Financial Crime, 12(1), 69-87.
  25. Stapenhurst, F., & Langseth, P. (1997). The Role of the Public Administration in Fighting Corruption. International Journal of Public Sector Management, 10(5), 311-330.
  26. Strategic Public Relations and Research Limited. (2005). National Survey on the State of Corruption in Kenya. Nairobi: The National Anti-Corruption Campaign Steering Committee.
  27. Svensson, J. (2005). Eight Questions About Corruption. Journal of Economic Perspectives, 19(3), 19-42.
  28. Transparency International. (2013). Global Corruption Barometer 2013: Kenya. TI-Kenya: Nairobi.
  29. Transparency International. (2014a). Clarification: Transparency International has not Ranked Kenya Fourth Most Corrupt Country in the World. TI-Kenya: Nairobi.
  30. Transparency International. (2014b). Kenya's Performance in Corruption Perception Casts Doubt on Reforms. TI-Kenya: Nairobi.
  31. Treisman, D. (2000). The Causes of Corruption: A Cross-National Study. Journal of Public Economics, 76(3), 399-457.
  32. Wantz, J., & McNally, D. (2015). Race, Accountability, and Corruption: A Survey Experiment. Western Political Science Association Annual Meeting, (pp. 1-37).
  33. Zaman, A., & Ur-Rahim, F. (2009). Corruption: measuring the unmeasurable. Humanomics, 25(2), 117-126.

Article Tools
Follow on us
Science Publishing Group
NEW YORK, NY 10018
Tel: (001)347-688-8931