Accounting for international joint ventures and the question of control: The case of Brazil, Colombia, and Mexico
Abstract
This study investigated international joint ventures in Brazil, Colombia, and Mexico to determine whether accounting methods employed by U.S.-based multinational venturers were appropriate for external reporting purposes, given underlying control relationships.
Generally accepted accounting principles prescribe the equity method for most joint venture investments. Two assumptions support this treatment: (1) majority percentage ownership is the only necessary indicator of control, and (2) joint ventures--characterized by shared control and minority interests--are noncontrolled investees. Official accounting bodies and others have expressed serious doubts about the adequacy of equity method disclosures. In addition, researchers have taken a more complex view of the ways in which one business entity controls another.
A questionnaire was mailed to U.S.-based multinational corporations with operations in one or more of the three countries. Ninety-seven (35% of all respondents) reported joint venture involvement, all but five as minority participants; they also provided data on accounting methods and control relationships. Two-thirds of minority participants reported that they controlled their joint venture investees. This perceived control rested on four independent control relationships identified through principal component analysis: managerial control, production control, ownership control, and control of intellectual property.
Principal component control scores for joint venture investments were used to test four reported accounting methods for within-method integrity and between-method differentiation. No significant differences were expected within each method, but each method was expected to be clearly differentiated from all others.
Pairwise t-tests found internal inconsistencies within each accounting method. Controlled equity method investments were significantly different from noncontrolled equity investments. Significant differences were also found within cost, proportionate, and full consolidation investments. Pairwise t-tests found poor differentiation between accounting methods. Similar investments were accounted for using any one of three methods: cost, equity, and full consolidation. Only proportionate consolidation investments were clearly differentiated.
Thus, the study concluded that accounting methods are inappropriate because they fail to capture underlying control relationships even as they exact a high informational cost for misclassification.