Internal Audit Deficiencies as Predictors of Perceived Fraud Risk: Survey Evidence from Small Business Owners in the United States

Small enterprises in the United States may face heightened exposure to financial misconduct due to limited resources, lean staffing, and informal operational structures that can constrain the effectiveness of internal oversight mechanisms. Despite the recognized importance of internal audit functions in deterring and detecting fraud, empirical evidence on how specific audit deficiencies relate to fraud vulnerability in U.S. small enterprises remains limited. This study examined the relationship between internal audit weaknesses and perceived fraud risk among small enterprise owners and managers in the United States, with particular focus on deficiencies in audit independence, monitoring frequency, segregation of duties, and documentation practices. A survey research design was adopted, and data were collected from 250 respondents selected through snowball sampling using an electronic questionnaire with a five-point Likert scale. Simple regression analysis was employed to test the hypothesis that internal audit weaknesses are positively associated with perceived fraud risk. The findings revealed a statistically significant positive association between internal audit weaknesses and perceived fraud risk (β = 0.850, p < 0.001), with the model explaining approximately 73.1% of the variance in perceived fraud risk. Inadequate monitoring, poor segregation of duties, and insufficient documentation were identified as the most commonly perceived deficiencies linked to fraud vulnerability. The study concludes that deficiencies in internal audit structures are associated with heightened perceived fraud vulnerability in U.S. small enterprises. These findings have practical implications for small business owners, policymakers, and accounting professionals seeking to strengthen governance and financial accountability in resource-constrained organizational settings. Limitations relating to sample generalizability and the perceptual nature of the data are acknowledged.