The Role of Fraud Management Policies in Mitigating the Severity and Reporting of Fraud Cases in the Public Sector: The Case of County Governments in Western Kenya Brian Omondi Onyango A Thesis Submitted in Partial Fulfilment of The Requirements For The Award of Degree of Master of Commerce at Strathmore University Strathmore Business School Strathmore University Nairobi, Kenya October 2022 This thesis is available for Library use on the understanding that it is copyright material and that no quotation from the thesis may be published without proper acknowledgment. i DECLARATION I declare that this work has not been previously submitted and approved for the award of a degree by this or any other University. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made in the thesis itself. © No part of this Thesis may be reproduced without the permission of the author and Strathmore University Brian Omondi Onyango Reg No. 100951 Signature: Date: 2nd September 2022 Approval The Thesis of Brian Omondi Onyango was reviewed and approved by the following: Supervisor: Dr. David Mathuva: Senior Lecturer Strathmore Business School Signature: Date: 2nd September 2022 ii ABSTRACT In the twenty-first century, fraud has become a problem that management and other stakeholders are working hard to resolve. In the Kenyan County Governments, fraud is on the rise, despite ongoing efforts by the office of the auditor general and other state investigative agencies to combat it. This study sough to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector in 11 selected county governments in western Kenya. More specifically, the study sought to investigate the role fraud control plans, mechanisms for reporting fraud and zero-tolerance to fraud policies influence the severity and reporting of fraud cases. The study borrowed from the fraud triangle and deterrence theory to address the general objective. Guided by a positivist research paradigm, primary data using a drop and pick questionnaire were collected from 316 respondents from the 11 county governments who included accountants, external and internal auditors and forensic investigative personnel. The data were analyzed using principal component-factor analyses, spearman correlations and ordered logistic regression approaches. The relevant descriptive and inferential statistics were also employed in the analyses. According to the results, two important fraud management aspects were established: deterrence effect of fraud management policies and fraud case reporting instances. The results illustrate that having an anti-fraud plan with zero tolerance to fraud, an independent fraud reporting system and deploying an integrated macro policy in the fight against fraud are useful mechanisms in deterring and combating the risk of fraud. The results also show that installing a fraud reporting system with real consequences to the offenders leads to an increase in the reporting of fraud cases. In line with the results, a strict anti-fraud plan with zero tolerance to fraud is useful in encouraging more fraud cases to be reported. Further, the deployment of an integrated macro policy on fraud might lead to reduced fraud case reporting. According to the study's findings, county governments in Kenya should adopt fraud management policies to assist in reducing the severity of fraud and help in timely reporting and handing of fraud cases. The findings are useful in informing fraud risk management policies in the counties as well as other public sector entities. iii TABLE OF CONTENTS DECLARATION .....................................................................................................................i ABSTRACT........................................................................................................................... ii LISTS OF FIGURES...........................................................................................................viii LIST OF TABLES ................................................................................................................ ix LIST OF ABBREVIATIONS AND ACRONYMS................................................................. x DEFINITION OF OPERATIONAL TERMS ....................................................................... xi ACKNOWLEDGMENTS ....................................................................................................xii DEDICATION.....................................................................................................................xiii CHAPTER ONE: INTRODUCTION .................................................................................... 1 1.1 Background of the Study ............................................................................................... 1 1.1.1 Fraud Management Policies and reporting of fraud cases in the public sector. ................. 4 1.1.2 Kenyan County Governments. ....................................................................................... 5 1.2 The Problem Statement ................................................................................................. 6 1.3 Objectives of the Study ................................................................................................. 8 1.3.1 General Objective ......................................................................................................... 8 1.3.2 Specific Objective of the Study ..................................................................................... 8 1.4 Research Question ........................................................................................................ 8 1.5 The Scope of the Study ................................................................................................. 9 1.6 The Significance of the Study ........................................................................................ 9 1.6.1 Policymakers and the Government................................................................................. 9 1.6.2 Researchers and Scholars. ............................................................................................. 9 1.6.3 County Governments. ................................................................................................. 10 1.7 Chapter summary ........................................................................................................ 10 iv CHAPTER TWO: LITERATURE REVIEW ...................................................................... 11 2.1 Introduction ................................................................................................................ 11 2.2 Theoretical Framework ............................................................................................... 11 2.2.1 The Fraud Triangle Theory .......................................................................................... 11 2.2.2 Deterrence Theory ...................................................................................................... 13 2.3 The Empirical Review ................................................................................................. 15 2.3.1 The role of a fraud control plan in mitigating the severity and improved reporting of fraud cases 15 2.3.2 Contribution of Mechanism for reporting fraud towards minimizing the severity and improved reporting of fraud cases ........................................................................................... 17 2.3.3 The role of Zero tolerance to fraud policies with regard to reducing the severity and improved reporting of fraud cases ........................................................................................... 18 2.4 The summary of Literature Review and Research Gap ................................................. 20 2.5 Conceptual Framework ............................................................................................... 24 2.6 Operationalization of Variables. .................................................................................. 25 2.7 Chapter Summary ....................................................................................................... 27 CHAPTER THREE: RESEARCH METHODOLOGY. ..................................................... 28 3.1 Introduction ................................................................................................................ 28 3.2 Research Philosophy. .................................................................................................. 28 3.3 Research Design ......................................................................................................... 28 3.4 Target Population. ....................................................................................................... 29 3.5 Sampling Procedure and Sample Size. ......................................................................... 30 3.5.1 The sample and the respondents. ................................................................................. 30 3.6 Data Collection Techniques and Procedure .................................................................. 31 3.7 Data collection procedures .......................................................................................... 31 v 3.8 Research Quality ......................................................................................................... 32 3.8.1 Reliability test. ............................................................................................................ 32 3.8.2 Validity test ................................................................................................................ 33 3.9 Data analysis and presentation ..................................................................................... 33 3.10 Ethical Considerations................................................................................................. 35 CHAPTER FOUR: PRESENTATION OF RESEARCH FINDINGS ................................. 36 4.1 Introduction ................................................................................................................ 36 4.2 Response rate ............................................................................................................. 36 4.2.1 Demographic profile of the respondents ....................................................................... 37 4.2.1.1 Respondents’ Organizations ........................................................................................ 38 4.2.1.2 Gender of respondents ................................................................................................. 38 4.2.1.3 Job occupation of the respondents ............................................................................... 39 4.2.1.4 Respondents’ qualifications ......................................................................................... 39 4.2.1.4 Respondents’ professional certification ........................................................................ 39 4.2.1.4 Respondents’ years of professional experience ............................................................. 40 4.3 Descriptive Statistical findings .................................................................................... 40 4.3.1 Descriptive results Plan for fraud control ..................................................................... 40 4.3.2 Descriptive results for Mechanism for reporting fraud .................................................. 41 4.3.3 Descriptive results for Policy of zero tolerance for fraud .............................................. 43 4.3.4 Descriptive results for fraud severity and reported cases ............................................... 44 4.4 Diagnostic Tests ......................................................................................................... 45 4.4.1 Normality test ............................................................................................................. 45 4.5 Principal Component-Factor Analysis (PCFA) ............................................................. 47 4.5.1. PCFA analysis for the independent variables................................................................ 47 vi 4.5.2 PCFA analysis for the dependent variable .................................................................... 52 4.4.3 Correlation analysis .................................................................................................... 55 4.4.3 Ordinal (Ordered) Logistic Regression......................................................................... 57 CHAPTER FIVE: DISCUSSION OF FINDINGS, CONCLUSION, AND RECOMMENDATIONS ...................................................................................................... 62 5.1 Introduction ................................................................................................................ 62 5.2 Discussion of findings ................................................................................................. 62 5.2.1 The role of a fraud control plan in mitigating the severity and improved reporting of fraud cases 62 5.2.2 Contribution of Mechanism for reporting fraud towards minimizing the severity and improved reporting of fraud cases ........................................................................................... 63 5.2.3 The role of Zero tolerance to fraud policies with regard to reducing the severity and improved reporting of fraud cases ........................................................................................... 65 5.2.4 Fraud severity and reported cases. ............................................................................... 66 5.3 Conclusion.................................................................................................................. 67 5.4 Recommendations ....................................................................................................... 69 5.4.1 Policy recommendations ............................................................................................. 69 5.4.2 Managerial recommendations ...................................................................................... 69 5.4.3 Recommendations for further research ......................................................................... 69 5.5 Limitation of the study ................................................................................................ 70 5.5.1 Timing of the study ..................................................................................................... 70 5.5.2 The scope of the study based on time ........................................................................... 70 5.5.3 Limits of research methodologies and tools used .......................................................... 70 REFERENCES ..................................................................................................................... 72 APPENDICES ...................................................................................................................... 79 vii APPENDIX 1: A LETTER OF INTRODUCTION ........................................................ 79 APPENDIX 2: QUESTIONNAIRE TO THE COUNTIES AND OAG ......................... 80 APPENDIX 3: EMAIL TO RESPONDENTS FOR ONLINE QUESTIONNAIRE ...... 87 APPENDIX 4: PARTICIPANTS INFORMED CONSENT ........................................... 88 APPENDIX 5: LIST OF COUNTY GOVERNMENTS IN THE SAMPLE .................. 91 APPENDIX 6: FULL RESULTS FOR REGRESSION MODEL 1 – FRAUD DETERRENCE MODEL ..................................................................................................... 92 APPENDIX 7: FULL RESULTS FOR REGRESSION MODEL 2 –FRAUD CASE REPORTING MODEL ........................................................................................................ 93 APPENDIX 8: LETTER OF ETHICAL APPROVAL TO CONDUCT RESEARCH .. 94 APPENDIX 9: PERMISSION TO CONDUCT RESEARCH FROM NACOSTI ......... 95 viii LISTS OF FIGURES Figure 2. 1: Conceptual Framework .................................................................................... 25 Figure 4. 1: Research response rate ..................................................................................... 36 Figure 4. 2: Normality test using box plots .......................................................................... 46 Figure 4. 3: Probability plots for the dependent variables ..................................................... 46 Figure 4. 4: Scree Plot of Eigenvalues ................................................................................. 51 Figure 4. 5: Scree plot for the eigenvalues – dependent variable .......................................... 54 ix LIST OF TABLES Table 2. 1: Key Studies and Research Gaps ....................................................................... 20 Table 2. 2: Operationalization of Variables. ....................................................................... 25 Table 3. 1: Target Population ............................................................................................ 29 Table 3. 2: Sample Size ..................................................................................................... 30 Table 3. 3: Reliability test results ....................................................................................... 32 Table 4. 1: Demographic characteristics............................................................................. 37 Table 4. 2: Plan for fraud control results ............................................................................ 40 Table 4. 3: Mechanism for reporting fraud results .............................................................. 42 Table 4. 4: Policy of zero tolerance for fraud results ........................................................... 43 Table 4. 5: Fraud severity and reported cases ..................................................................... 44 Table 4. 6: Variance inflation factors and tolerance ............................................................ 46 Table 4. 7: Kruskal–Wallis equality-of-populations rank test .............................................. 47 Table 4. 8: Eigen Values and Sampling Adequacy Tests..................................................... 47 Table 4. 9: Principal Component Analysis-Factor Analysis ................................................ 49 Table 4. 10: Factor loadings for the dependent variable ........................................................ 52 Table 4. 11: Eigen values– dependent variable ..................................................................... 53 Table 4. 12: Spearman’s Rank Correlation results ................................................................ 56 Table 4. 13: Ordered Logistic Regression results- Reported cases Model .............................. 58 Table 4. 14: Ordered Logistic Regression results- Fraud Severity model ............................... 60 x LIST OF ABBREVIATIONS AND ACRONYMS ACFE CPAs Association of Certified Fraud Examiners Certified Public Accountants EACC FCP NACOSTI NFA OAG SEM Ethics and Anti-Corruption Commission Fraud Control Plan National Commission for Science, Technology, and Innovation National Fraud Authority Office of the Auditor General Structural Equation Model TDF TFP ANOVA ICPAK INTOSAI KMO PCA PC PwC SESRC SPSS Techniques for Detecting Fraud Techniques for Fraud Prevention Analysis of Variance Institute of Certified Public Accountants of Kenya International Organization of Supreme Audit Institutions Kaiser-Meyer-Olkin Principal Component Analysis Percent Price Waterhouse Coopers Strathmore Institutional Ethics and Scientific Review Committee Statistical Package for the Social Science xi DEFINITION OF OPERATIONAL TERMS Fraud Management Policies Practices that raise awareness amongst employees that fraud response plans have been developed to deal with and lessen the damage caused by any fraudulent attack (ACFE, 2021). Fraud Prevention These are strategies implemented by institutions to reduce the likelihood of resource waste (Alamer, 2015) Fraud Detection These are control measures implemented by institutions to reduce the possibility of institutional resource waste (Mat et al., 2013) A plan for fraud control A Fraud Control Plan is a plan or method that is responsibly planned and implemented in order to avoid fraud in an organization (KPK, 2014). Mechanism for reporting fraud A Mechanism for reporting fraud is a technique used by fraud reporters to relay information regarding various perceived and actual fraudulent activities to the relevant authorities (NFA, 2010) Policy of zero tolerance for fraud According to Khadra and Delen (2020), the Policy of zero tolerance for fraud refers to a scenario whereby the relevant authorities have put in place mechanism to ensure that fraud is unacceptable at whatever level. xii ACKNOWLEDGMENTS I could not have finished this thesis without the help of many people who helped me in a variety of ways to make it a success. First and foremost, I thank my wife, Lauryn Akinyi who happens to be my classmate in this program, for her patience, prompting endurance, and love, which have helped me continue along the arduous path of writing toward a finished thesis. Thanks are also due to my supervisor Dr. David Mathuva for his constructive criticism and encouragement that helped me shape this thesis. Dr. Mathuva has taught me the value of being a dedicated supervisor to a research student. And last but not least, I thank my friends who read bits and pieces of this thesis and offered constructive criticism. xiii DEDICATION This study is dedicated to my late mother, Mama Priscah Onyango, who was a firm believer in education. She was always there for me when it came to matters of education. I dedicate this study to my father, Mzee Dan Onyango Mohol, for his support and unending prayers. I also dedicate this study to my sons Harvey and Chiwo, as well as my unborn children, whom I hope will be bright and enthusiastic about education. This research is dedicated to my members of staff for their support which has helped me continue along the arduous path of writing towards a finished thesis. Finally, I dedicate this research to anyone who’s trying to make the world a better place by fighting fraud. 1 CHAPTER ONE: INTRODUCTION 1.1 Background of the Study Fraud is a threat to all businesses (Khadra & Delen, 2020). The impact of fraud on states, organizations, and citizens is the same: the rule of law is undermined, rights are violated, the institution is not transparent, public resources are lost, and national integrity is weakened. As a result, fraud and its consequences have become a worldwide issue (Khadra & Delen, 2020). Employees, individuals outside the business, or another organization benefit directly or indirectly from fraud to the disadvantage of the organization (Erbuga, 2020a). Some examples of fraud are: accepting bribes or gifts; diverting a potentially profitable transaction to an employee or person outside the organization that would otherwise generate profit for the organization; fraud, which is usually the misappropriation of money or property, and the falsification of financial information for the purpose of concealing the act, thus making it difficult to detect; intentionally concealing or misrepresenting events, transactions, or data; unauthorized or illegal use of confidential or proprietary information; unauthorized or illegal handling of computer networks or operating systems, theft, and so on. The entire cost of fraud is difficult to assess objectively, and calculating the indirect damages caused by fraud was even more challenging. Some scams will go undetected for a long time. As a result, the study can only discuss in terms of fraud loss estimates. The results of the survey, which were based on national fraud reports, were published by the Association of Certified Fraud Examiners (ACFE, 2021). Fraud cost firms throughout the world an average of 5% of their annual income. Effective fraud resolution necessitates coordinated and strategic effort from international, national, and local partners who are aware of the problem and its importance. States should adopt laws and regulations at the national and local levels to deter fraudsters from participating in the process, and those who do should be suitably sanctioned. The problems of business fraud, which result in a loss of profit and assets for the firm as well as a loss of reputation, can be solved by implementing forensic accounting procedures and building an anti-fraud strategy. 2 The current operating environment is ripe for fraud. Large companies across the world such as Enron, WorldCom, Satyam, and Lucent are darned for their ending because of money-related fraud. For that reason, mechanisms for detecting and assessing financial fraud should be created (Yue et al., 2017). Financial fraud is a threat to the accounting and auditing professions, necessitating the development of countermeasures. Even though fraud can be detected, the effectiveness of fraud detection is a difficult task in the accounting profession. Historically, auditors have been unable to save fraud from happening because of a loss of required facts mining skills, inexperience due to its low frequency, and the use of non-computerized computer systems (Ngai et al., 2020). According to a study, 25 percent of business firms in India are affected by financial misconduct, demonstrating the importance and depth of corporate fraud. Fraud in corporate organizations involves large sums of money and causes hostility in a variety of industries. According to a study conducted in the year 2021, 0.5 p.c of fraud perpetrators were non – management staff (KPMG Malaysia, 2021). In contrast, this category of fraud perpetrators represented about 0.34 p.c in the year 2009, a daunting detail to the government and the management since there was a 0.16 p.c increase the subsequent year. The customers were the second commonest variety of fraud perpetrators, who were followed by those in positions of management, who represented 0.18 p.c respectively. Others discovered included perpetrators who charge 0.08 for services and suppliers who charge 0.06. The data also revealed that the rate of stealing outgoing funds was highest in 2013, at 0.67 percent, compared to 0.57 percent in 2019. The theft of physical assets was ranked at 0.58, followed by the theft of incoming resources at 0.34. Individual assessments revealed that theft of money and cash receipts (0.26 p.c) was the foremost common kind of fraud followed by inaccurate billing (0.16 p.c), and stock larceny (0.13 p.c) (KPMG Malaysia, 2021). According to Ernst and Young (2020), governments and firms agree that corruption and felony are prejudicial to businesses and that vital steps ought to be taken to scale them back. The augmented variety of fraud cases in Nigeria for example has been blamed on the government's dilapidation in providing social amenities. According to Modugu and Anyaduba (2018), there is a negative perception of state auditors for failing to prevent embezzlement or misuse of public funds within the state departments. For this reason, the use of fraud management policies in mitigating the severity and reporting of fraud cases is highly recommended. It is dire because only 20% of Nigeria's population bears the brunt of the country's revenue (Abdullahi & Mansor, 2018). This results in skewed resource distribution. 3 Even though Nigeria is Africa's largest oil producer, at least 70% of its population lives in poverty because of corruption and resource mismanagement (Adebisi & Gbegi, 2015). Since the country's independence in 1963, fraud has been a major issue in Kenya (Nyamu, 2012). The introduction of devolved county government systems in 2013 didn’t help the situation, although several legislations and laws, such as the 2016 Bribery Law, the 2016 Public Procurement & Disposals Act, and the law on Proceeds of Crime & Anti-Money Laundering of 2009, have been enacted to combat fraud. Individuals pay bribes to reap authorities’ offerings (EACC, 2017). Wajir County recorded the very best percent of these searching for public offerings paying bribes (ninety percent), observed with the aid of using Meru County (88. five percent), Trans Nzoia County (83. three percent), and Kajiado County (81. five percent) (EACC, 2017). According to the 2019-2020 Auditor General's report, a few counties, inclusive of West Pokot, incurred excess expenses in violation of the Public Finance Management Regulations 2015. According to a similar report, the Wajir County government had cases of unreconciled inventory, whereas Uasin Gishu County government reported cases of inaccuracies within the financial reports and unaccounted for revenue (Auditor-General, 2019). According to an equivalent Auditor General’s report, most Kenyan Counties lack an internal audit section and a functional audit committee to observe the governance process, responsibility process, and management systems (Auditor-General, 2019). This provides the impression that most counties are having issues implementing fraud prevention and detection techniques and have many fraud cases reported; it additionally implies that the Counties either have weak fraud detection and prevention techniques or haven’t enforced the antecedently counseled techniques. Previous studies have focused on the role of fraud management policies in mitigating the severity and reporting of fraud cases in the private sectors (Abdulrasheed et al., 2018; Fadipe- Joseph & Titiloye, 2016; Inaya & Isito, 2016; Kanu & Okorafor, 2018; Kawugana & Faruna, 2018; Odi, 2020; Owolabi, 2019;Uchenna & Agbo, 2018). Most of these studies were done in Nigeria, and they didn't really pay attention to the relationship between fraud management policies and the severity and reporting of fraud incidents. Kiragu et al. (2015) pointed out that there was a noteworthy absence of study in Kenya on the connection between preventative, detective, and responsive fraud management policies (deposit money bank fraud management policies) and non-financial bank performance (efficiency and operational performance). Based on this identified problem and gaps, this study sought to identify the role of fraud management 4 policies in mitigating the severity and reporting of fraud cases in the public sector, a case study of Western Kenya’s County Governments. 1.1.1 Fraud Management Policies and reporting of fraud cases in the public sector. Fraud management policies are anti-fraud policies that informs employees that action plans have been developed to deal with and lessen the harm caused by any fraudulent assault. For the purpose of this study, a plan for fraud control, a mechanism for reporting fraud and a policy of zero tolerance for fraud will be used as the fraud management policies. Public sector entities as non-profit organizations aim to contribute to the public good. The management's capacity to institutionalize effective and efficient fraud management policies will determine whether these objectives are met. According to Fatoki (2020), high and lower-level managers of public sector organizations were vulnerable to internal fraud threats, which had a negative impact on their operational performance globally. All public sector organizations face fraud threats on a global scale. Numerous fraud-related actions within public sector organizations have resulted in the collapse, enormous losses on investments, high legal fees, the incarceration of important individuals, and a loss of trust in the public sector organizations. In the USA, among other financial fraudulent activities affecting publicly traded companies, investors lost $180 billion in the World Com fraud scandal of 2002, $150 million in the Tyco scandal of 2002, $1.4 billion in the Heath South scandal of 2003 (the largest publicly traded company), and $3.9 billion in the America International Group (AIG) scandal of 2005. Relatively speaking, the majority of public sector companies have faced significant obstacles to achieving their desired performance, including fraud activities and inadequate internal control systems. The majority of public sector entities, particularly in developing nations like Kenya, Ghana, Nigeria, and South Africa among other African countries, experienced a decline in performance and collapse of public sector entities, as noted by (Uwaoma & Ordu, 2019; Gitau & Samson, 2016; Kinyua, 2019) due to public sector organizations' failure to implement high-tech controls that correspond to the fraud innovation methods used by those organizations. The erosion of public trust and top public officials' fraud activities, which have led to financial hardship, have been the two main issues facing Kenya's public sector over the past 20 years. Poor management of public resources is the result of the rise in fraud activities in Kenya's 5 public sector. Because of its slowness, Kenya's legal system is unable to curb fraud-related actions. The delayed Kenyan legal system's response to fraud in the public sector has encouraged ongoing deception among top officials, directors, and executives, which has resulted in a steady fall in performance and the demise of the majority of these organizations. 1.1.2 Kenyan County Governments. CoK (2010) established 47 counties based on the previous administrative districts established by Kenyan Provinces and Districts Act that was enacted in the year 1992. The devolved units as they are commonly referred to CoK (2010), oversee the implementation of the development agenda within the counties. This is done through consultations with the Kenyan National Government. The implementation of the development agenda within the devolved units is usually undertaken annually through budget allocation to various counties. Despite the budget allocation to various counties, it is common knowledge that the county governments continue to encounter various financial challenges because of corruption, misappropriation of funds, and ineffective financial systems (Adari, 2007). Fraud-related cases have been witnessed in various state agencies as per the auditor general reports on the performance of those agencies. It is against this backdrop therefore that fraud prevention necessitates a larger embrace of moral practice as well as a working relationship between the devolved units and the Kenyan national governments with a clear and formidable approach to fighting fraud. Devolved units in Kenya encounter fraud risks because of poor implementation of fraud management policies and internal control weaknesses, which frequently result in loss of monies (Njoroge, 2003). Njui (2012) conducted a study on the role of forensic investigations and the strength of internal control systems in the enhancement of better management in public institutions in Kenya. In his findings, he found out that the strength of the internal control system has a significant positive influence on better management within the public institutions in Kenya. He also found that risk management also has a positive influence on corporate governance within public institutions. Finally, the study showed that compliance and consulting have the least influence on the better management of public institutions. In view of this study, it is imperative for county governments to use a unit of purpose to combat and fight fraud. The responsibility of the national government in combating fraud within the devolved units, therefore, is to facilitate the creation of the enabling spaces for the county governments to undertake anti-fraud initiatives effectively and efficiently. 6 Due to an increase in instances of fraud and misappropriation of public funds by top management, county governments in Kenya have been having issues with a persistent delay in operational performance and ineffective service delivery. Ineffective preventive, investigative, and response fraud management techniques, as well as rising executive and top manager fraud activity, may contribute to Kenya's public sector organizations' subpar performance. Information asymmetry, immoral behavior, ineffective governance frameworks, and moral hazards are additional potential causes of fraud activities that have been found Gitau and Samson (2016), and Kinyua (2016) 1.2 The Problem Statement Profit and non-profit organizations are both vulnerable to a slew of scams ACFE (2021). In a government setting, fraud can be perpetuated through the denial of government services (Bierstaker et al., 2006). It is against this backdrop therefore that there’s a consensus among various stakeholders within the government that institutional procedures within the law should be followed, failing which massive government entities would face losses. This would result in the use of public offices for personal gain. Despite the presence of fraud management policies in Kenyan County Governments, Fraud is still occurring. A strict application of fraud management policies in mitigating the severity and reporting of fraud cases is required to reduce the chances of fraudulent activities occurring (Munir et. al., 2016) The inability of these policies to deter fraud should be beefed up by instituting identifiers for fraud and other relevant fraud prevention strategies. The County government's reliance on external audit reports may fail to produce the desired results even after they have been implemented (OAG, 2020). This could lead to a loss of organizational sustainability, which would bring the service delivery system to a halt. Furthermore, the County government may respond reactively rather than pro- actively because of this discovery (OAG, 2020). Local studies on the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public and private sector include Sang (2017) and Kiragu et al. (2015) who studied bank size, occupational fraud risk, and investigated the factors that influence the use of fraud management policies in Kenyan commercial banks. Both studies used the survey research design as well as the correlation and regression analysis techniques. They discovered 7 a strong and positive correlation between bank size and the risk of occupational fraud. The investigations also showed that the banking system has seen a rise in fraud activities due to a lack of comprehensive anti-fraud measures and non-compliance with fraud management policies. This study has a geographical gap because it looked at businesses in the private sector, whose fraud mitigation procedures might be better than those of county governments. Globally, Njanike et al. (2009) conducted a study on the efficacy of forensic auditing in identifying, investigating, and preventing bank frauds. The study used both descriptive and inferential statistics to analyze the data. According to the findings, forensic auditing departments face a number of difficulties, including a lack of material resources, technical expertise, interference from management, and ambiguous recognition of the profession. As a result, the top bank employees' involvement in bank fraud has increased. In Nigeria, Amanze and Onukwugha (2017) and Olay and Dada (2017) looked at the roles of fraud management policies in the banking industry's loan fraud detection and prevention system. These studies, which employed a survey research design, showed that there was little fraud control in Nigerian banks; the results also showed that fraud management policies including, the risk assessment management; system audit; and financial report verification procedures adopted by the banking sector were unable to stop fraudulent activity among Nigerian banks. The establishment of a multi-agent framework by regulatory authority is a stand-alone system that can detect fraud in the banking industry in Nigeria, according to their studies, which also revealed that fraud management policies captured by risk assessment, system audit, and financial report verification were not statistically significant in mitigating the fraudulent act. Due to their concentration on the banking industry, whose fraud management policies differ from that of the county governments, these studies present a geographical gap. While the empirical studies that have been conducted on this topic (Sang, 2017; Kiragu et al., 2015; Njanike et al. 2009; Amanze & Onukwugha, 2017; Olay & Dada, 2017) and others are crucial and have made a significant contribution to the literature, it is imperative to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector in the context of Kenyan County Governments. This is because the empirical studies on this area of study have demonstrated both location contextual gaps and conceptual methodological gaps as has been demonstrated from the local and global studies above. 8 1.3 Objectives of the Study 1.3.1 General Objective The overall aim of this study is to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector in 11 selected county governments in western Kenya. 1.3.2 Specific Objective of the Study The specific objective of this study include: i. To establish the role of fraud control plans on mitigating the severity and improved reporting of fraud cases among the Western Kenya County Governments. ii. To determine the contribution of mechanisms for reporting fraud toward minimizing the severity and improved reporting of fraud cases among the Western Kenya County Governments. iii. To assess the role zero-tolerance to fraud policies play with regard to reducing the severity and improved reporting of fraud cases among the Western Kenya County Governments. 1.4 Research Question The study is intended to answer the following research questions: i. What is the effect of fraud control plans on mitigating the severity and improved reporting of fraud cases among the Western Kenya County Governments? ii. What is the effect of the mechanism for reporting fraud toward minimizing the severity and improved reporting of fraud cases among the Western Kenya County Governments? iii. What is the effect of the zero tolerance to fraud policies with regard to reducing the severity and improved reporting of fraud cases among the Western Kenya County Governments? 9 1.5 The Scope of the Study This study sought to establish the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector: The case of County Governments in Western Kenya. The conceptual scope covered fraud management practices in these counties. This study was therefore valuable to the county governments in Kenya and would thus allow them to mitigate the severity and reporting of fraud cases in the public sector. In terms of the geographical scope that was covered, the study specialized in 11 county governments in the Western Kenya region. The Accountants, Internal and External Auditors, and Forensic Investigators were the participants in the study due to their understanding of fraud management policies and since they action public policies touching on monetary problems within the counties. Respondents were drawn at random from the Auditor General's Office and included forensic investigators attached to the county governments under study. The study was conducted between May 2022 and June 2022 and was limited to a quantitative research approach. The study was underpinned by the fraud triangle theory and deterrence theory. 1.6 The Significance of the Study The application of fraud management policies in mitigating the severity and reporting of fraud cases raises various areas for future research, however, this study centered on assessing the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector: The case of County Governments in Western Kenya. Policymakers and the government, researchers, scholars, and the county governments among other stakeholders were identified as the main beneficiaries of this study. 1.6.1 Policymakers and the Government The study’s findings will aid the Kenyan government and policymakers in developing a legal framework to combat fraud based on the recommendations established by the study. 1.6.2 Researchers and Scholars. The study’s findings are valuable to future researchers, scholars, and academicians who want to conduct research on the effectiveness of fraud management policies on fraud hindrance and 10 detection within the devolved units or Kenyan county governments. Moreover, the study contributes to future literature and fills the current research gap. 1.6.3 County Governments. The study’s findings may aid in identifying gaps for fraud mitigation within Kenya’s county governments. The findings of the study would also benefit the Management team and those in charge of internal forensic audits in County Governments by allowing them to implement and streamline the best fraud management policies. 1.7 Chapter summary This chapter presented the introduction of the study, which is divided into six sections. Section 1.1 discussed the background of the study which included a critical examination of the independent and dependent variables and their selection is justified; Section 1.2 conducted a critical examination of the statement of the problem, which stated clearly why this study should be carried out; Section 1.3 presented the objectives of the study, which included both the general objective and specific objectives; Section 1.4 presented the research questions; Section 1.5 presented the scope of the study and finally, Section 1.6 presented the significance of the study, which included how the study will help various stakeholders. 11 CHAPTER TWO: LITERATURE REVIEW 2.1 Introduction The chapter includes relevant theoretical models used for the study as well as the empirical literature on fraud management policies in mitigating the severity and reporting of fraud cases. Further, it captures some of the existing literature on fraud management policies in mitigating the severity and reporting of fraud cases, as well as why it is necessary for organizations to implement fraud management policies based on past studies. The chapter compares the similarities and differences in the findings of assorted studies carried out in numerous sectors and geographical locations to work out the role of fraud management policies in mitigating the severity and reporting of fraud cases. Additionally, it includes narrowing the information gap. Lastly, the chapter uses the existing literature to give a summary of how fraud management policies ought to be implemented. 2.2 Theoretical Framework This section provides a theoretical overview of fraud management policies in mitigating the severity and reporting of fraud cases. In this study, a multi-theoretical approach was used because it provides a more comprehensive understanding of a phenomenon. This argument is supported by empirical papers that have been published in leading repositories. The fraud triangle and deterrence theories influence the research. 2.2.1 The Fraud Triangle Theory This theory was created by (Cressey, 1953). In the year 1950, a criminologist undertook research with an aim to identify the drive within individuals who engage in fraudulent activities. He examined 250 criminals in 5 months (Abdullahi & Mansor, 2018). According to Cressey’s (1953) findings, fraud is committed for 3 major reasons: pressure, opportunity, and rationalization. As a result, the fraud triangle theory emerged. 12 Based on the pressure as an element for committing fraud, Albrecht et al. (2018) emphasized the importance of using the phrase "perceived" when discussing pressure or opportunity to commit fraud because the pressure or opportunity may not be real and is solely dependent on the perpetrator's perceptions. According to Lister (2021), the essential aspect of committing fraud is pressure or incentive. He talked about three types of pressure: personal, work-related stress, and external pressures. Vona (2019) investigates various personal and corporate pressure as key motivational proxies for fraud. Greed, living beyond one's means, excessive expenses or personal debt, family financial problems, and drug addiction are all examples of pressure. When pressure turns into an obsession with achieving corporate goals regardless of the repercussions, the situation becomes unstable and eventually leads to catastrophe. Individuals are more prone to participate in dubious behaviors in this situation, which may lead to fraud (Hooper & Pornelli, 2010) The opportunity, according to Rae et al. (2017), is a flaw in the system that the employee has the opportunity, power, and capacity to exploit and possibly conduct fraud. The weaker an organization's internal control system is, the more likely it is that the fraud was hidden. Financial fraud may not occur even when an employee is under a lot of strain, according to Hooper and Pornelli (2010), unless there is an opportunity. Individuals are heavily influenced to commit fraud by organizational internal control flaws, a bad auditing system, a lack of accounting records, and a lack of segregation of duties. Tunerj et al. (2019) suggest that even if a person is under pressure or has a motive, he will not be able to commit fraud unless the opportunities are presented. Related party transactions, according to Moyes et al. (2005), would be the second factor among the multiple opportunity risk variables. According to Wilks and Zimbelman (2020), related party transactions can only be ranked third among the several elements that indicate the presence of fraud opportunities. The third aspect of the fraud triangle hypothesis, as well as the fraud diamond theory, is rationalization. This concept states that when conducting fraud, a fraudster must express a variety of ethically acceptable behaviors that was used to justify his plan before breaking the trust. The perpetrator's opinion that the dishonest and unethical action is something other than illegal activity is referred to as rationalization. If the offender is unable to rationalize his unethical behavior, he is unlikely to commit fraud. "I was only borrowing the money," "I was entitled to the money," "I had to steal to provide for my family," "I was underpaid/my employer had cheated me," are some of the moral justifications used by fraudsters (Cressey, 1953). It's 13 also worth noting that observing rationalization is tough; as a result, it's hard to read the minds of con artists (Cressey, 1953; Wells, 2005). Fraudsters frequently have specific attitudes that allow them to justify their dishonest behavior (Hooper & Pornelli, 2010). Incentives/pressure, opportunity, rationalization, and capability are all interconnected. According to Howe and Malgwi (2019), the established gap between incentive/pressure and opportunity is closed when the fraudster can justify his unethical behavior. The fraud triangle theory has been chosen over fraud diamond or pentagon since it clearly informs the study and its variables since the respondents in the study especially accountants can commit fraud and at times may be under financial pressure and can rationalize processes. Since the respondents in this study are the executors and policy implementers within the county governments, they may in their own interests choose to implement the internal control systems or even ignore the policies laid down. The theory, therefore, supports the usefulness of examining fraud management policies in mitigating the severity and reporting of fraud cases. 2.2.2 Deterrence Theory The origins of deterrence theory can be traced back to the writings of the 18th-century philosophers Jeremy Bentham (1948) and Beccaria Cesare (1963), who argued that humans were self-interested and rational thinkers, driven in their actions by an economic 'hedonistic calculus' in which they act to maxim (Paternoster & Raymond, 2010; Tombs et al., 2013). This theoretically opens a Pandora's Box of 'deterrence,' whereby raising the costs of behavior through sanctions reduces the willingness to pursue a bad course of action, but fraud remains an enemy of the public service. Individuals' fear of punishment, according to this theory, has a negative impact on their intent to commit fraud (Nagin et al., 2011). Deterrence is useful in public entities for controlling employee behavior, particularly in the fight against fraud. Deterrence is widely regarded as the most effective method of combating fraud. As a result, there is an urgent need to develop strategies to prevent individuals from acting fraudulently. It is self-evident that prevention is preferable to cure, and that hiring a forensic accountant is the only way to combat fraud (Nagin et al., 2011). As governments around the world recognize the importance of counteracting employees' fraudulent actions and the potential risks they may pose, deterrence theory has become increasingly important, and forensic accountants can play an important role in the deterrence of fraud (D’Arcy et al., 2011). D’Arcy et al. (2011) discovered that detecting and punishing 14 fraud perpetrators, as well as raising awareness and providing clear information on repercussions if one is a perpetrator, reduced the amount of fraud. According to this theory, the harsher the punishment and the harsher the sanctions, the better the employees behave, which is the key remedy in fraud control (Herath et al., 2009). Deterrence techniques force people to consider the possibility of being caught as well as the consequences of their actions before deciding whether to break the rules (D’Arcy et al., 2009). One of the most difficult challenges in determining whether deterrence methods have been effective in fraud control is that the results are difficult to evaluate and monitor (Herath et al., 2009). Even with the help of forensic accountants, monitoring employees can be very expensive and complicated. Continuous checks by forensic accountants, for example, may instill fear in employees, especially if they are aware that their actions are being monitored, owing to the complexity and profound differences in their organizations and the methods of control put in place (D’Arcy et al., 2009). Regardless of the issues with monitoring the outcomes, deterrence theory is widely used in organizations to prevent and detect fraud (Herath et al., 2009). Overall, deterrence is still considered the most powerful theory in fraud detection and prevention. Deterrence theory is relevant to the study since it was used to predict employee behaviors in various situations because behavior is either supportive or disruptive of fraud control (D’Arcy et al., 2011). The reason for this is that fraud occurs because of employees failing to follow financial control policies by using their designations for personal gain (Herath et al., 2009). As a result, deterrence theory is one of the most applied theories in fraud management policy (Siponen et al., 2010). Fraud deterrence theory informs the study and its variables because fraud does not happen at random; it happens when there are right factors for its occurrence. The enablers of fraud are the root causes and enablers of fraud deterrence outbreaks. It was critical to identify potential fraud opportunities and how fraud management policies could eliminate factors that cause fraud. Whereas preventing fraud encompasses both short- and long- term actions, it is possible that it is not the same as fraud detection, which has frequently been a source of confusion. Fraud detection entails examining old transactions for signs of unusual transactions. To Deter fraud, an assessment of the conditions and behaviors that fosters fraud, as well as what is likely to happen when anti-fraud measures are not put in place such as the use of fraud management policies, could be implemented. 15 2.3 The Empirical Review This section of the study provides a critical examination of the empirical literature on fraud management policies in mitigating the severity and reporting of fraud cases. Further, it goes ahead to demonstrate the identified research gaps discovered during the critical review of the literature. To reduce the probability of fraudulent activities from taking place and mitigation of financial losses, businesses should adopt fraud detection and prevention strategies that are sufficient to combat fraud (Bierstaker et al., 2006). When it comes to fraud prevention, the use of regulations and policies is pivotal (Bolton & Hand, 2002). Since detecting fraud follows when preventing fraud has failed, it entails identifying a previously committed action (Bolton & Hand, 2002). Research methods that will assist investigators in preventing and detecting potential fraud should therefore be adopted. Varied perception-based research studies are conducted to work out the usefulness of fraud management policies within the private and public sectors (Sang, 2017; Kiragu et al., 2015; Njanike et al. 2009; Amanze & Onukwugha, 2017; Olay & Dada, 2017). The studies have resulted in policy recommendation frameworks outlining the best policies for mitigating the severity and reporting of fraud cases in the public and private sectors. Because of the sensitivity of the issue and the fear of reputational damage, obtaining secondary data for the purpose of assessing an entity’s efficiency of anti-fraud strategies is a challenge to the researcher. Furthermore, because of the significant differences in both economic and geographical environments where most of these studies were conducted, making the generalization of those findings in Kenya would be misleading. As a result of this, knowledge and literature gaps existed. As a result, a study is needed to determine the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector. 2.3.1 The role of a fraud control plan in mitigating the severity and improved reporting of fraud cases Alam (2010) asserts that there are three ways to prevent crime: pre-emptive, preventive, and repressive. When repression has been used to commit a crime, repressive measures are imposed. whereas pre-emptive and preventive measures focus on preventing crimes from 16 happening in the first place. The best approaches to take when combating fraud are ones that are either preventive or preemptive rather than repressive. Preventive actions should be prioritized above repressive ones for a number of reasons (BPKP, 2015): when corruption occurs, there is a significant financial loss; the amount of stolen state funds that are recovered is very small; the reputations of both individuals and institutions are harmed; the legal system and potential suspects must invest time and money in the litigation process; and the longer the corruption case is kept secret, the greater the chance that the corruptors will use other fraud to conceal their actions. The construction of a program that is effective at preventing and detecting fraud, one of which is through the implementation of the Fraud Control Plan (FCP), is one of the preventive measures that public agencies can use to prevent fraud. A responsible strategy or process that is created with the goal of effectively preventing fraud in an organization is known as a "fraud control plan" (KPK, 2014). Studies like Kiragu et al. (2015) and Sang (2017) investigated the factors that affect fraud management practices in Kenyan commercial banks as well as bank size and occupational fraud risk. Both investigations used correlation and regression analytic techniques, as well as a survey research design. They discovered a strong and positive correlation between bank size and fraud control plans. Their findings also showed that the number of fraudulent operations in the banking sector had increased due to a lack of comprehensive fraud management policies and non-compliance with fraud mitigation techniques. The study focused on the factors that affect fraud management policies in Kenyan commercial banks in contrast to the current study, which sought to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector. This, therefore, presents a conceptual gap. Khanna and Arora (2019) examined the reasons for bank fraud and the implementation of preventive security controls in the Indian banking industry. In particular, the study evaluated the effectiveness of fraud control plans, ratio analysis, trend analysis, and commercial anti- fraud strategies in mitigating the severity and reporting of fraud in the banking industry. The implementation of fraud management policies considerably improved the severity and reporting of fraud cases in the banking industry, as determined by descriptive statistics and the Ordinary Least Square (OLS) model. Further, the study empirically discovered that the major causes of bank frauds were a lack of training, overworked staff, and low compliance levels. The study concentrated on firms listed in the banking sector, as opposed to the current study which sought to investigate the role of fraud management policies in mitigating the severity 17 and reporting of fraud cases in the public sector. This, therefore, presents a location contextual gap. 2.3.2 Contribution of Mechanism for reporting fraud towards minimizing the severity and improved reporting of fraud cases Fraud can take numerous forms and occur in a variety of circumstances. There is no one-size- fits-all approach towards minimizing the severity of fraud, and the key to success is the establishment and improvement of a fraud reporting mechanism. The National Fraud Authority (NFA) estimates that the public sector in the United Kingdom has lost £17.6 billion in fraud, with a total loss of £30 billion a year for the country as a whole, comprising both public and private sectors (NFA, 2010). According to research conducted by the London Boroughs Fraud Investigators Group (LBFIG) in 2009, local government fraud totaled £621 million (Wesley Lane, 2010). According to this study by Wesley Lane, the public sector is responsible for more than half of all estimated losses in the United Kingdom. Fraud prevention is the responsibility of everyone in the company, from the CEO to the lowest- ranking executive. Strong controls are required to ensure that the perceived risk of fraud is reduced. As a result, proactive measures to detect fraud, such as a clear mechanism for reporting fraud, questioning, fraud assessment, and an anonymous hotline, give employees the responsibility and opportunity to help stop fraud. These measures will also boost the perception of detection, lowering the risk of fraud and preventing new frauds. It will contribute to the creation of a positive work environment that decreases the pressure or motivation to commit fraud while also improving employee morale and ethics (Peterson & Zikmund, 2004). In Nigeria, Amanze and Onukwugha (2017) and Olay and Dada (2017) looked at the roles of auditors in the banking industry's loan fraud detection and prevention system. The results of these studies, which employed a survey research design, showed that the level of fraud control in Nigerian banks was low; they also showed that the mechanisms for reporting fraud, system audits, and financial report verifications adopted by the banking industry were unable to stop fraudulent activities among Nigerian banks. The establishment of a multi-agent framework from regulatory authority is a stand-alone system that can be integrated by banks to combat loan fraud, according to their studies, which also showed that audit roles captured by mechanism for reporting fraud, system audit, and financial report verification were not 18 statistically significant in determining the fraudulent act in the Nigerian banking industry. The scope of the study is limited to Nigeria, as opposed to this study which is limited to Kenyan County governments. This, therefore, presents a location contextual gap. Adetoso and Akinselure (2016) studied fraud prevention and control in the banking sector of Nigeria. Internal audit and whistleblowing were utilized as stand-ins for fraud prevention, while management control was used as a stand-in for fraud control. Because its proxies, such as the mechanism for reporting fraud, internal audit, and whistleblowing, were taken into consideration, the study demonstrates that there was a substantial association between fraud control and fraud prevention. The study also showed that, based on the sample study, fraud management strategies must be able to increase fraud control and fraud prevention of commercial banks since the proxies for both variables significantly influence one another. The study focused only fraud prevention and control in the banking sector, as opposed to the current study which sought to determine how the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector; the case of county governments in Western Kenya thus presenting a conceptual gap. 2.3.3 The role of Zero tolerance to fraud policies with regard to reducing the severity and improved reporting of fraud cases According to Khadra and Delen (2020), all organizations, public or private, are susceptible to fraud. The impact of fraud on states, organizations, and citizens is the same: it undermines the rule of law, violates rights, lacks transparency, wastes public resources, and undermines national integrity. As a result, fraud and its consequences have become a global problem (Khadra & Delen, 2020). Fraud against the organization is typically committed for the direct or indirect benefit of employees, individuals outside the organization, or another organization (Erbuga, 2020b). Accepting bribes or gifts; diverting a potentially lucrative transaction to an employee or person outside the organization that would otherwise generate profit for the organization, and so on are examples of fraud. Fraud is not an individual problem; rather we face systematic fraud, particularly in the public sector, because of the establishment of a private monopoly over the public sector. Therefore, the first step that needs to be implemented in an organization to develop an efficient fraud control system is the development of an anti-fraud strategy. The anti-fraud strategy is based on 19 a policy of integrity and zero tolerance for fraud. Corruption, embezzlement of assets, and misrepresentation are examples of fraud. There are scammers inside and outside the organization. The following factors affect fraud performance: Pressure, Opportunity, and Rationalization. The cost of fraud can be difficult to estimate. In the context of the Kenyan County Governments, a political will, effective legislation, anti-fraud culture, strategy, and a policy of zero tolerance for fraud all play a role in fighting fraud. Enofe et al. (2017) conducted an empirical analysis of bank fraud in Nigeria and preventive strategies. The study's purpose was met by taking into account the zero-tolerance fraud policy, the Internal Control System (ICS), Corporate Governance (CG), and Compliance with Banking Ethic (CBE). Regression and correlation analysis were used to assess the strength of the relationship between bank fraud and preventive measures. According to the study, a zero tolerance for fraud policy, a robust internal control system, sound corporate governance, and adherence to banking ethics all significantly and favorably affect the prevention of fraud in the banking sector. The study only investigated how compliance policies, management overrides for controls and segregation of duties affected fraud mitigation. These techniques are a limited indicator of fraud mitigation. The study therefore presents a contextual gap. Karuti (2020) examined forensic accounting and fraud control in county governments within Mt. Kenya region. The aim of the study was to investigate and produce definitive results on how fraud management policies, fraud management policies, degree of forensic accounting knowledge, and application of forensic accounting skills affected fraud control in the County Government of Kenya. Inferential statistics from regression analysis include tests for normality, multicollinearity, model fitness, model specification, and a test of a hypothesis with a 95% confidence level. The study found that county government personnel did not sufficiently apply fraud management policies to counter the potential of fraud. The study also found a favorable and significant correlation between fraud management policies and fraud control. Since the study used parametric analysis on non-parametric data, it presents a methodological gap. Alfian et al. (2017) examined the effect of antifraud strategy on fraud prevention in the banking industry. This study aimed to investigate the impact of an anti-fraud strategy including a policy of zero tolerance to fraud-on-fraud prevention in the banking sector. Quantitative methods, particularly explanatory research approaches, were used to assess the research data. The results of this study demonstrated that the three pillars of prevention, detection, and investigation all 20 have an impact on preventing fraud in the banking industry. This study has a geographic gap because it focused on private sector companies in developed economies, whose anti-fraud tactics may be superior to those used locally. The study did not demonstrate how fraud management policies as the antifraud strategy can aid in mitigating severity in fraud cases thus presenting a conceptual gap. 2.4 The summary of Literature Review and Research Gap In this chapter, a discussion of the fraud triangle theory and the deterrence theory, both of which were relevant to this study was done. It presented the studies that were carried out that were related to the objectives of this study. The similarities and differences based on previous research outcomes in the same subject area were ably presented through the elaborate empirical review. The techniques used in the previous studies were also presented. The chapter established that a restricted study was conducted within the context of Kenyan County governments to see the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector. At its best, fraud in the county governments has been regarded by the office of the auditor general as a serious threat to the counties. Generally, fraud has been associated with the rise of economic crime and alarming regulatory flaws. This problem has been on a sharp rise in developing countries and economies, Kenya included (PwC Global Economic Crime Survey 2016). Most of the studies reviewed (Sang, 2017; Kiragu et al., 2015; Njanike et al. 2009; Amanze & Onukwugha, 2017; Olay & Dada, 2017) used a descriptive research design and adopted the survey methodology. The Primary data for the studies were gathered from various participants and were then quantified to assess their effectiveness in the study by use of the Likert scale. To measure and classify the level of fraud hindrance and detection strategies, the studies used descriptive statistics such as mean, mode, median, standard deviation, and ranking techniques. Sengur (2012), on the other hand, used Friedman's test model to determine whether there was a significant variation in auditors' perceptions of the effectiveness of fraud management policies in mitigating corruption, asset misappropriation, and financial statement fraud. Table 2. 1: Key Studies and Research Gaps 21 The Key studies, the Research Gaps, and how the Research Gaps were filled by the study are summarized in the table below: Author Title Findings Research Gap and how the study will fill them. Kiragu et al. (2015) Bank size and occupational fraud risk: Empirical evidence from commercial banks in Kenya. They discovered a strong and positive correlation between bank size and fraud control plans. Their findings also showed that the number of fraudulent operations in the banking sector had increased due to a lack of comprehensive fraud management policies and non- compliance with fraud mitigation techniques. The study focused on the factors that affect fraud management policies in Kenyan commercial banks in contrast to the current study, which sought to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector. This, therefore, presents a conceptual gap. Khanna and Arora (2019) Reasons for bank fraud and the implementation of preventive security controls in the Indian banking industry The implementation of fraud management policies considerably improved the severity and reporting of fraud cases in the banking industry, as determined by descriptive statistics and the Ordinary Least Square (OLS) model The study concentrated on firms listed in the banking sector, as opposed to the current study which sought to investigate the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector. This, therefore, presents a location contextual gap. 22 Author Title Findings Research Gap and how the study will fill them. Amanze and Onukwugha (2017) Loan fraud detection system for banking industries in Nigeria. Using data mining and intelligent agents: The The establishment of a multi-agent framework from regulatory authority is a stand-alone system that can be integrated by banks to combat loan fraud, according to their studies, which also showed that audit roles captured by mechanism for reporting fraud, system audit, and financial report verification were not statistically significant in determining the fraudulent act in the Nigerian banking industry The scope of the study is limited to Nigeria, as opposed to this study which is limited to Kenyan County governments. This, therefore, presents a location contextual gap. Adetoso and Akinselure (2016) Fraud prevention and control in the banking sector of Nigeria The study demonstrates that there was a substantial association between fraud control and fraud prevention. The study also showed that, based on the sample study, fraud management strategies must be able to increase fraud control and fraud prevention of commercial banks since the proxies for both variables The study focused only fraud prevention and control in the banking sector, as opposed to the current study which sought to determine how the role of fraud management policies in mitigating the severity and reporting of fraud cases in the public sector; the case of county governments in Western Kenya thus 23 Author Title Findings Research Gap and how the study will fill them. significantly influence one another presenting a conceptual gap. Enofe et al. (2017) Empirical analysis of bank fraud in Nigeria and preventive strategies According to the study, a zero tolerance for fraud policy, a robust internal control system, sound corporate governance, and adherence to banking ethics all significantly and favorably affect the prevention of fraud in the banking sector The study only investigated how compliance policies, management overrides for controls and segregation of duties affected fraud mitigation. These techniques are a limited indicator of fraud mitigation. The study therefore presents a contextual gap. Opiyo (2017) Role of forensic accounting services in fraud mitigation among parastatals in Kenya According to the study's findings, parastatals in Kenya include compliance policies, management overrides for controls, and segregation of duties, all of which significantly aid in reducing fraud. The study only investigated how compliance policies, management overrides for controls and segregation of duties affected fraud mitigation. These techniques are a limited indicator of fraud mitigation. The study therefore presents a contextual gap. Karuti (2020) Forensic accounting and fraud control in county The study found that county government personnel did not sufficiently apply forensic Since the study used parametric analysis on non-parametric data, it 24 Author Title Findings Research Gap and how the study will fill them. governments within Mt. Kenya region accounting rules to counter the potential of fraud. The study also found a favorable and significant correlation between forensic accounting skills and fraud control. presents a methodological gap. Alfian et al. (2017) Effect of antifraud strategy on fraud prevention in the banking industry The results of this study demonstrated that the three pillars of prevention, detection, and investigation all have an impact on preventing fraud in the banking industry The study did not demonstrate how fraud management policies as the antifraud strategy can aid in mitigating severity in fraud cases thus presenting a conceptual gap. Source: Researcher (2022) 2.5 Conceptual Framework A conceptual framework helps a study by identifying the research variables and clarifying their relationship (McGaghie et al., 2001). The fraud triangle and the deterrence theories were used to develop the conceptual framework for the study. These theories championed the case for the use of the dimensions for the independent and dependent variables to measure the relationship between fraud management policies and fraud severity and reported cases. As a result, fraud management policies become the study’s independent variable. The effect of the independent variable dimensions was evaluated. Fraud severity and reported cases, therefore, become the dependent variable. This relationship is depicted in figure 2.1 25 Independent Variables Dependent Variable Figure 2. 1: Conceptual Framework 2.6 Operationalization of Variables. Operationalization facilitates the transformation of abstract concepts into observable characteristics that can be measured using indicators. Both the dependent and independent variables were measured using a rating scale ranging from 1 (strongly disagree) to 5 (strongly agree). The indicators used in the studies are summarized in table 2.2. Table 2. 2: Operationalization of Variables. Variable Constructs Definition of Operation Measures for rating Data Analysis Supporting Literature Independent Variable: Fraud Management Policies Fraud Control Plans The ability of the organization to institute and implement the construct. A five-point Likert scale on the implementation level and the believed usefulness Descriptive analysis, inferential analysis Kiragu et al. (2015) Fraud Management Policies  Plan for fraud control: Integrated macro anti-fraud policy; fraud risk assessment plan; community awareness program; fraud reporting systems; conduct and disciplinary standard  Mechanism for reporting fraud: effective fraud reporting; independent reporting lines; clear fraud reporting channels; follow-up on fraud reports; fraud reporter immunity  Policy of zero tolerance for fraud: anti-fraud strategy; existence of zero-tolerance policy by top managers; anti-fraud plans in place; code of business ethics and governance aligned with anti-fraud policy; existence of contemporary model anti-fraud policy  Reported Cases  Impossible to account for fraud cases; increase in referrals in post zero tolerance era.  Fraud Severity  Fraud decreases with anti- fraud policy; severity of policy deters fraud occurrence; Controls  Organisation type  Gender 26 Variable Constructs Definition of Operation Measures for rating Data Analysis Supporting Literature Mechanism for reporting fraud The degree to which this construct is used in the Organization A five-point Likert scale on the implementation level and the believed usefulness Descriptive analysis and inferential analysis Amanze and Onukwugha (2017) Policy of Zero Tolerance for fraud The extent to which this construct is applied in the organization A five-point Likert scale on the implementation level and the believed usefulness Descriptive analysis and inferential analysis Alfian et al. (2017) Dependent Variable: Fraud Severity and reported cases Number of reported cases of simple and complex fraud The degree to which the organizations are reporting low cases of simple and complex fraud A five-point Likert scale on the implementation level and the believed usefulness Descriptive analysis and inferential analysis Ocansey (2017) Control variables Organization type Type of the organization the respondent works for =1 if it is a county government, 2 if it is the office of the auditor general Descriptive analysis , Regression analysis Gender Gender of the respondent =1 if it is male, 2 if it is female Descriptive analysis , Regression analysis Source: Researcher (2022) 27 2.7 Chapter Summary The study’s literature review was described in this chapter. From the onset, the chapter began by describing the two theories that underpinned the study: the fraud triangle theory, and the deterrence theory. The empirical studies on the relationship between fraud management policies and fraud severity and reported cases were then discussed in detail followed by the presentation of the study’s conceptual framework illustrating the nature of the relationship between fraud management policies and fraud severity and reported cases as per the literature review. The chapter ends with the operationalization of both independent and dependent variables for measurement in the questionnaire. 28 CHAPTER THREE: RESEARCH METHODOLOGY. 3.1 Introduction The study’s third chapter was utilized to chronologically outline the methodology that was followed throughout the study. A summary of the research philosophy, design, population, sampling method, data collection tools, data analysis, and presentation were also included in this section. 3.2 Research Philosophy. A system of rules and beliefs regarding the development of knowledge and its fundamentals is known as philosophy (Saunders et al., 2014). This study's foundation is congruent with empirical positivism, an objective epistemological strategy that enables the researcher to objectively examine theories and theoretical notions (Bell et al., 2018). This philosophy was chosen for the study because it is grounded in empirical positivism and is done in an impartial manner without the influence of the researcher or the subjects. Additionally, the philosophy was appropriate for this study because its main objective was to estimate the relationship between the study variables using quantitative methods. As a result, the paradigm was determined by the study to be the most successful in conducting the survey. 3.3 Research Design Research design is the model that enables the researcher to find solutions to problems and guidelines at different stages of research (McLaughlin, 2012). The research design provides a framework for the research to be carried out. A quantitative research design was utilized in the study. An appropriate research design is necessary to empirically test the structural linkages predicted by the conceptual model of the study. By using data from questionnaires, the researcher hoped to improve the answer to the research questions (Creswell, 2013). To answer the study's research questions, the researcher collected quantitative data. A descriptive design was used for the study. Descriptive designs enable the collection, summarization, presentation, and interpretation of information for the purpose of clarification. It also involves many people and describes population characteristics using an unbiased sample. Because it depicts accurate current facts through data collection for testing hypotheses or answering questions to conclude 29 the study, the descriptive research design is one of the best methods for conducting research in human contexts (Creswell, 2014). The study used a field survey and a descriptive research design. With the help of this research design, the researcher was able to collect empirical data on both independent and dependent variables simultaneously. The researcher was also able to describe the relationships between the variables thanks to this research design. 3.4 Target Population. The term "target population" refers to a universal set of all members of a real or hypothetical set of people, events, or objects to which an investigator wishes to generalize the results (Bryman, 2012). A target population is a collection of people, things, or occurrences that share observable characteristics that the study is interested in (Gibson, 2017). Kenya has a total of 47 counties, according to the County Governments Act (2012). The study analyzed 11 County governments in the Western Kenya region as well as the office of the Auditor-General. Western Kenya County Governments were chosen because they are leading in Fraud occurrence according to the 2020 annual corporate report by the Office of the Auditor-General (OAG, 2020). This formed the study’s unit of analysis. A target of an investigation is referred to as a unit of analysis and might include, among others, an individual, a group of individuals, an organization, or a country (Bryman, 2012). The unit of analysis for the current study was the accountants, internal auditors, external auditors and forensic investigators. This is because accountants and auditors, both internal and external, oversee the implementation of the accounting and the fraud management policies used by the counties to fight fraud. The study's findings, while establishing the body in which the respondents were deployed, are shown in Table 3.1. Table 3. 1: Target Population Frequency Percent 11 Western Kenya Region County Governments 405 90.00 The Office of Auditor General 46 10.00 Total 451 100 Source: Kenya Census results (2019) 30 The study also discovered that a delicate balance of respondents was achieved, with approximately 405 respondents representing approximately 90 percent drawn from Kenya's County Governments and approximately 46 respondents representing approximately 10 percent drawn from the Auditor General's office at the National Audit Office. The sample respondents for the study were 405 internal and external auditors and certified public accountants attached to the counties and 46 forensic investigators from the office of the OAG. 3.5 Sampling Procedure and Sample Size. Sampling is a process, procedure, or technique for selecting a subset of a population to participate in a study. It is the process of selecting a group of people for a study so that the people selected to represent the large group from which they were selected (Kothari, 2008). Bernard (2013) for his part, defines a sample as a small proportion of the total population, a selection from the population. The sample was selected using a stratified random sampling method. The stratified proportional random sample provides more accurate estimates of general population parameters and ensures that a more representative sample is drawn from a relatively homogeneous population. Stratification attempts to reduce the standard error by allowing some control over the variance (Latham, 2007). Divided into three groups in the study: Internal Audit, Accounting and External Audit staff from the County governments and Accounting, and Forensic Investigation staff from the Office of the Auditor-General respectively. The study took a sample of at least 30% from each shift apart from the Forensic Investigation staff. According to Mugenda (2008), 10% of the population must be sampled for a population of 1000 or more, while 30% of the population must be sampled for a population of less than 1000. As a result, the sample size was 451 staff members. 3.5.1 The sample and the respondents. The study established that the sample was chosen, and the respondents were as shown in table 3.2. Table 3. 2: Sample Size Job Occupation Frequency Percent The Internal Auditors 135 30 The Forensic Investigators 46 10 31 Job Occupation Frequency Percent The External Auditors 135 30 The Counties CPAs 135 30 Total 451 100 Source: Kenya Census results (2019) 3.6 Data Collection Techniques and Procedure For the study survey, a questionnaire instrument was constructed. In descriptive and explanatory investigations, questionnaires are frequently used to collect information from respondents about a certain topic (Saunders et al., 2014). The process of creating a tool (research questionnaire) with a set of prepared question items that aim to elicit replies in a systematic way is known as questionnaire development (Christensen et al., 2011). With the use of closed-ended questions and 5-point Likert scale systems, the research instrument acquired a structured format. The construction of the study questionnaire was influenced by earlier research and the constructs used to measure the study variables. As a result, the statements were in line with the study's objectives. The questionnaire was divided into four sections: Section A gathered information on the demographic profile of the respondents in the study. Sections B and C were structured to address the study’s independent and dependent variables dimensions, that’s a plan for fraud control, a mechanism for reporting fraud, a policy of zero tolerance for fraud, and the reduction in the reported cases of simple and complex fraud. A five-point Likert scale was used to rate the questions for their efficacy (strongly agree - 5, agree - 4, neutral - 3, disagree - 2, and strongly disagree - 1). 3.7 Data collection procedures All respondents received the questionnaires using a drop and pick later procedure (Fowler Jr, 2013). As per the study's sample, each respondent was contacted separately. The distribution of the questionnaires that were given to the respondents was tracked using a register that was kept. The researcher saw to it that the required permissions were obtained so that data collecting could take place. Additionally, the researcher used Google forms to gather data electronically 32 when the drop and pick approach could not be used, which improved the convenience of the data collecting process. 3.8 Research Quality This is a reference to a small-scale replication and trial of a significant study that aids in evaluating the reliability and validity of the research instruments that will be used, as well as practical issues related to the administration of questionnaires (Bhattacherjee, 2012). A pilot test is required to identify various flaws that are more likely to occur, deficiencies in the study, as well as various difficulties that are most likely to arise during the research process (Creswell, 2014). A pilot test was conducted with several randomly chosen county governments. In this study, 10% of the sample made up the pilot group. 3.8.1 Reliability test. The degree to which a research instrument delivers consistent results after repeated trials is considered a reliability test (Cooper et al., 2014). Cronbach’s alpha was employed in this study to determine the questionnaire’s internal consistency. Legitimacy and consistency of findings on instruments contribute to important data interpretations (Creswell, 2013). A greater value indicates that the data items examined have a strong association, whilst a lower value indicates that the items tested have a weaker relationship. Cronbach’s alpha was employed in this study to determine consistency in the range of 0 to 1. Closer values to 1 indicate that the variables under investigation are easily quantifiable. If the alpha is between 0.7 and 0.99, the reliability item can be accepted (Fraenkel & Wallen, 1996). The study conducted pilot tests among the 11 county governments and the office of the auditor general, which allowed for reliability statistics to be conducted. Table 3.3 shows the reliability test results. Table 3. 3: Reliability test results S.No. Construct N Cronbach’s Alpha Items Verdict 1 Fraud severity and reported cases 12 .934 4 Accepted 2 Plan for fraud control 12 .778 5 Accepted 3 Mechanism for reporting fraud 12 .825 5 Accepted 4 Policy of zero tolerance for fraud 12 .720 5 Accepted Source: Research data, (2022) 33 The tests yielded Cronbach alpha scores of above 0.7, which led to the acceptance of the study questionnaire. 3.8.2 Validity test The content validity index for each instrument was calculated using two sets of data from each instrument. Validity was determined by giving two experts to evaluate the relevance of each item in the instrument to the objectives and rate each item on the scale of very relevant - 4, quite relevant - 3, somewhat relevant – 2, and not relevant – 1. Validity was determined using the content validity index (C.V.I), C.V.I = items rated 3 or 4 by both experts divided by the total number of items in the questionnaire. This is symbolized as n¾ ⁄ N. If the CVI computed for each tool was greater than 0.6, the instrument was considered valid as advised by (Jeffry, 2011) 3.9 Data analysis and presentation The researcher made sure that all of the quantitative data were edited, coded, entered, and cleaned before the quantitative data could be evaluated. This guarantees accuracy and coherence. However, the researcher also used descriptive statistics and inferential statistics to evaluate quantitative data. SPSS version 22 was used to accomplish this. Descriptive statistics were employed to compile a summary of the demographic profile data. The descriptive statistics used in this study were percentages, means, standard deviations, and frequencies. The presentation of the study's findings included tables and figures. Bar and pie charts were among them. In order to evaluate the relationship between the categorical variables, the study also performed Spearman correlation analysis. In addition to component analysis, which was used to limit the number of variables and concentrate on just those that have the most influence, ordinal regression analysis was also performed. Ordinal logistic regression was relevant for the study since it was used to determine the strength of the relationship between independent variables (fraud management policies) and dependent variables (reported cases and fraud severity). The dependent variable in this study was assessed using a Likert scale with five possible values. Based on the factor analysis, two factors for the dependent variable were extracted. These included Factor 1 - Deterrence effect of fraud management policies (Deterrence) and Factor 2: 34 Increased case reporting of fraud activities due to the existence of policy (Increased Case Reports). This, therefore, led to the development of two ordered logistic regression models for the study. With regard to the independent variables, the factor analysis results led to the extraction of four factors namely: Factor 1: Fraud reporting and disciplinary measures (FraudReport_Discipline), Factor 2: Anti-fraud plan and zero tolerance to fraud (Anti- FraudPlan_ZeroTolerance), Factor 3: Independent fraud reporting system (IndependentFraudReportSystem) and Factor 4: Integrated macro-policy with clear responsibilities for fraud (IntegratedMacroPolicy). These formed the independent (test) variables for the study. Two control variables were also incorporated: gender of