VARIABLES AFFECTING THE CHOICE OF ACCOUNTING METHODS: AN INVESTIGATION BASED ON FIRMS ADOPTING ACCOUNTING CHANGES (INCOME, COVENANTS, POLITICAL)

This study examines the empirical validity of two of the variables suggested in the literature as relevant to the choice between income increasing and income decreasing accounting methods. \u27Political costs\u27 and \u27debt covenant related costs\u27 are the variables examined on the basis of a sample of seventy two firms that adopted voluntary accounting changes in 1981. For testing purposes, political costs are represented by four measures of firm size and a measure of profitability. Five financial ratios are used to represent debt covenant related costs. After specifying the type of accounting change implied by each measure probit models are estimated for hypothesis testing. Separate models are estimated on the basis of data for the year of accounting changes as well as for the preceding year. Except for a few univariate models with a fairly low level of significance, none of the models are found to be significant. Based on the premise that the variables are relevant to only certain subsets of firms the sample is partitioned into high political cost and low political cost subsets and also into high debt and low debt subsets. Several models estimated for subsets of firms with either a lower financial leverage or a higher level of profitability are significant. In both cases the most significant models are based on the data for the year of adoption of accounting changes rather than for the preceding year. While the results based on the total sample do not support the theory, the significant models estimated on the basis of subsets indicate a potential area for further investigation