Research Methodology, Security market model, Prediction errors, Daily returns data

Research Methodology, Market model prediction errors, Omitted variables

Complexity has been linked to the quality of an income tax system (Dean, Keenan, and Kenney [1980]), including its possible influence on the system's ability to generate revenues (New York State Bar Association [1972]). Given the IRS' recent estimate that $81 billion in annual revenue is lost through noncompliance (IRS [1983, p. 21]), the question of whether tax complexity has a significant effect on taxpayers' reporting positions is a potentially important issue. Complexity represents but one strand in a web of interrelated factors and propositions influencing compliant tax reporting in a democratic society. Nevertheless, it has been singled out as a factor affecting compliance and a study of its effect thereon can be viewed as one step in an ongoing program of study of noncompliance. This study involved two distinct phases. The first phase was devoted to obtaining operational definitions of tax complexity, using multidimensional scaling. These definitions of tax complexity were then used in the second phase of the study to test for potential effects of complexity on reporting position selections in four different tax situations. While phase 1 is critical to phase 2 of my study, this paper highlights the results of the latter. Details of phase 1 can be obtained in Milliron [1984]. In section 2 I discuss previous literature involving tax complexity. Section 3 provides an overview of the methodology and the research issues studied. Sections 4 and 5 describe the data collection procedures

Corporate failure, Reporting delays, Information dissemination, Annual reports

Asset valuation, Replacement costs, Fixed asset lives, Depreciation

Information evaluation, Information systems, Group decision making, Group performance

Information evaluation problem, Knowledge, Decision making

Information systems, Decision making, Lab experiment

Historical cost, Valuation

A major purpose of the Estes study was to determine, by using questionnaires, the expected usefulness of current cost information for various classes of assets, both current and long term.2 Two assumptions were made in conducting the study: (1) current costs were objectively measurable, and (2) the current cost information would be of a supplementary nature only. The sample was selected from three organizations: the Institute of Chartered Financial Analysts, the National Association of Bank Loan Officers and Credit Men (Robert Morris Associates), and the Financial Executives Institute. These groups were chosen because Estes believed they closely paralleled two major financial statement user groups: (1) investors, both current and potential, and (2) lenders. Questionnaires were sent to 300 members from each group. A total of 338 or 37.8% responded. The results of the study were that 81 percent of the item responses