중앙조달기업의 조달의존도와 이익조정 행태에 관한 연구

This study empirically examines the impact of public procurement dependence on earnings management in firms engaged in central procurement markets. Accounting information plays a vital role in stakeholders’ decision-making, yet managers often strategically adjust reported earnings when facing external pressures such as contracts, regulations, or performance evaluations. Procurement contracts, in particular, are characterized by strict conditions regarding value, delivery deadlines, and unit pricing, which may shape both accounting choices and real operating activities. Against this backdrop, the study investigates whether firms highly dependent on public procurement rely more on accrual-based earnings management or on real activity-based earnings management. Earnings management was measured using discretionary accruals and their absolute values as accrual-based indicators, while REM was assessed with a composite index and three components: abnormal operating cash flows, abnormal production costs, and abnormal discretionary expenses. Public procurement dependence was defined as the ratio of procurement contract value to annual sales, with the top 33% classified as the high-dependence group each year. The sample consisted of firms participating in central procurement markets from the 2000s onward. Regression analyses included industry and year fixed effects, with firm-level clustered standard errors. The findings reveal several patterns. First, among the low-dependence group, larger procurement contract sizes were significantly associated with lower discretionary accruals, consistent with Watts and Zimmerman’s (1978, 1986) political cost hypothesis. However, for the high-dependence group, this conservative effect was not statistically significant, indicating that dependence itself does not fully explain variations in accrual-based management. Second, in terms of real activity-based management, the opposite trend was observed: highly dependent firms significantly increased REM as procurement contract sizes grew, whereas no significant effect was found in the low-dependence group. Within REM components, abnormal production costs rose while discretionary expenses declined, suggesting overproduction to reduce unit costs alongside cuts in advertising or promotion. Overall, the evidence suggests that procurement contracts shape financial reporting incentives. Firms engaged heavily in procurement markets appear more inclined to manage earnings through real activities, which are harder to detect under procurement monitoring, than through accrual manipulation. Policy implications follow: oversight frameworks should broaden their scope to include REM indicators, not only accrual-based measures. Managerial evaluations should also incorporate earnings quality signals to prevent excessive reliance on real activities. Finally, this study acknowledges limitations, as it does not fully capture detailed contract features such as advance payments, inspections, and payment timing. Future research should explore how these factors, along with liquidity constraints, shape firms’ choices between accrual and real activity-based earnings management.