A Study Of The Economic Consequences Of The Allocation Of Decision-making Power Within An Enterprise

Based on Chinese listed business groups,this dissertation investigates the impact of intra-firm decision rights allocation on internal resource allocation and external supervision.Because the distinguishing mark of the firm is the supersession of the price system by the authority of the entrepreneur-coordinator,the allocation of decision rights is crucial to the firm.There are two important research issues regarding intrafirm decision rights allocation.First,for the top managers,how to allocate decision rights to promote the efficiency of the internal resource allocation? Prior theoretical studies have formed a complete framework on this issue,but related empirical evidence is scarce and mixed due to the difficulty in measuring decision rights allocation.Second,how does the decision rights allocation affect the external supervision? The literature often ignores the top managers’ agent role and assumes that they act benevolently in the interest of the shareholders.This paper aims at comprehensively analyzing the impact of decision rights allocation on both internal resource allocation and external supervision,and providing empirical evidences using the unique data of Chinese listed business groups.By decision rights allocation,I refer to the centralization and delegation of the decision rights between the top managers and the divisional managers.When the decision rights are delegated,the divisional managers have great autonomy in the decision-making process,and the top managers restrain the divisional managers through performance appraisal.When the decision rights are centralized,the top managers play an important role in the decision-making process and issue orders to the divisional managers.In internal resource allocation,delegation and centralization both have comparative advantages and disadvantages.When decision rights are delegated,the divisional managers spontaneously utilize the information they own.However,delegation has costs,including the cost caused by the divisional managers’ self-serving behaviors,the cost of inter-divisional externality,and the cost of inter-divisional price generation when different divisions cooperate.Centralization makes up for the shortcomings of delegation,but entails information cost and renders the divisional managers undermotivated.Therefore,how decision rights allocation affects internal resource allocation is an empirical question.Furthermore,if the assumption that the top managers aim at maximizing the firm value is relaxed,decision rights allocation may affect the top managers’ information concealing behaviors as well as the outsiders’ supervision.When the decision rights are centralized,for the purpose of maintaining the private benefits extracted with decision rights,the top managers may deceive the outsiders.In the meantime,the top managers’ behaviors are firmly related to firm performance.This relation strengthens the top managers’ motivation to conceal information,but reduces the outsiders’ cost of supervising the top managers.Since it is difficult to measure decision rights allocation,most existing empirical studies rely on survey results or proprietary data.Taking advantage of the data from Chinese listed business groups,I innovatively measure intra-firm decision rights allocation with the abnormal concentration of the group’s cash flows in the parent company.This measure assumes that,the distribution of cash flows reflects the distribution of decision rights,and the allocation of decision rights between the parent company and the subsidiaries reflects the overall decision rights allocation of the group.Specifically,I calculate the proportion of the group’s employee compensation,gross profit and investment outflow that incur in the parent company,take the first principal component of these ratios,adjust it for the proportion of total assets in the parent company,and take the residual as the proxy for the centralization of decision rights.Compared to the measures by prior studies,my measure features repeatability and comprehensiveness.It allows me to base empirical tests on a large-scale panel dataset,and make full use of the abundant data on listed business groups.I validate this measure with other measures of centralization as well as tests on consistency,influencing factors and direct consequences.I study the impact of decision rights allocation on internal resource allocation from the perspectives of cash holding and investment.Cash is the lifeblood of the firm,while the investment determines the firm’s growth.Empirical results on the cash holding show that,centralization of decision rights is negatively related to cash holdings,but positively related to the value of cash holdings,indicating that centralization enhances the efficiency of cash holding policies.I examine how decision rights allocation affects both fixed asset investments and innovation investments.Empirical tests show that,centralization is positively related to fixed asset investment,especially when divisions are geographically dispersed.Centralization is also positively related to innovation investments.This relation strengthens when the firm has specialized R＆D centers,when the divisions are dispersed geographically,and when the top managers have technical background or display a willingness towards innovation through information disclosure.This relation is weaker for firms with higher financial leverage and firms in more competitive industries.I also find that centralization is positively related to the value of fixed asset investment and the proportion of exploratory innovation to exploitative innovation.Collectively,these findings suggest that centralization facilitates efficient investments by decreasing self-serving behaviors,internalizing externality,and fostering inter-divisional cooperation.To study the impact of decision rights allocation on the top managers’ information concealing behaviors,I test how centralization affects management earnings forecasts and earnings management.I find that,centralization is negatively related to the proportion of bad news in earnings forecasts,and leads to less precise disclosure and more external attributions of bad news.The more centralized the decision rights are,the weaker the stock market reactions to management earnings forecasts containing bad news,and the more asymmetric the market reactions to good news and bad news.Centralization is positively related to upward earnings management,especially when the top managers have accounting or finance background,and when the firm’s performance is worse than expected.Consequently,centralization is negatively related to the market reactions to annual earnings.These evidences indicate that,when decision rights are centralized,the top managers conceal unfavorable information through strategic disclosure.Lastly,to study how decision rights allocation affects the outsiders’ supervision,I examine the accuracy of the analysts’ earnings forecasts.I find a negative relation between centralization and analyst earnings forecast error.This relation is weaker when the top managers are newly appointed,but strengthens with the disclosure of management earnings forecasts.Further,the more centralized the decision rights are,the stronger the market reactions to analyst earnings forecasts.These evidences show that analyst can accurately predict firm earnings based on the relation between the top managers’ behaviors and the firm’s performance.To alleviate the potential endogeneity issue,I adopt the power distance attribute of regional culture,measured by the number of official schools in Ming and Qing dynasties within a radius around the firm,as an instrumental variable.The two-stage regressions,to some extent,confirm the causal relationship between decision rights allocation and the firms’ investment,information disclosure,and analyst behavior.The main contribution of this dissertation is to interpret decision rights allocation based on the transaction cost theory,and to study the comparative advantages and disadvantages of centralization and delegation in internal resource allocation.Further,I relax the assumption on the top managers’ incentive,and examine how decision rights allocation affects the top managers’ information concealing behaviors and the outsiders’ supervision.More importantly,based on the theoretical framework,I analyze in detail the economic consequences of decision rights allocation in cash holding,investment,information disclosure and analyst earnings forecast,and propose falsifiable hypotheses.This dissertation not only contributes to the literature on decision rights allocation,but also fills in the gaps in the literature of cash holding,information disclosure,analyst,etc.In the empirical research,I propose and validate a novel measure for decision rights allocation using data publicly disclosed by Chinese listed business groups.This approach,which ensures a high level of repeatability,is applicable to related studies.This dissertation is of practical significance.My findings indicate that,when the divisional managers’ self-serving behaviors and the inter-divisional externality obviously harm firm value,or when there is a need for inter-divisional cooperation,the firm should centralize decision rights.However,when the decision rights are centralized,shareholders should pay attention to the top managers’ strategic disclosure behaviors,and utilize information from other sources to form a thorough understanding of the top managers’ behaviors and the firm’s performance.Last but not least,the conclusions of this dissertation can be extended to other types of organizations.