THE CHANGES IN ECONOMIC INCENTIVES WHICH MOTIVATE DISCRETIONARY ACCOUNTING CHANGES: THE CASE OF THE SWITCH TO, AND THEN FROM, ACCELERATED DEPRECIATION (STRAIGHT-LINE, POSITIVE, AGENCY).

Abstract

This study extends previous work which attempted to describe why different managers use different accounting techniques (for financial reporting purposes only). The focus is accounting changes, specifically depreciation changes, both to and from accelerated methods. Seven hypotheses are proposed. Three of them predict a negative relationship between a manager's use of accelerated depreciation and (1) the firm's debt/equity ratio, (2) the existence of a bonus plan based on accounting income and (3) the potential impact of a depreciation change on earnings. These three hypotheses are referred to as H(D/E), H(BONUS) and H(IMPACT), respectively. The other four hypotheses predict a positive relationship between a manager's use of accelerated depreciation and (1) firm size, (2) the degree of voting control exercised by insiders, (3) the strength of current earnings relative to the preceding year's, and (4) industry-wide barriers to entry. These are referred to as H(SIZE), H(CTRL), H(EPS) and H(BTE), respectively. Designating accelerated as A and straight-line as S, the (tax/book) depreciation changes studied are the switch from S/S to A/A in 1954 and the switches from A/A to A/S in 1968 and 1969. All seven hypotheses were jointly tested with respect to the A/A to A/S changes in 1968 and 1969. The 1968 multivariate results indicate varying degrees of support for all the hypotheses except H(BTE), while the 1969 multivariate results support only H(D/E), H(BONUS) and H(SIZE). Only five hypotheses were jointly tested with respect to the change from S/S in 1954 (not H BONUS or H BTE ). Support is offered for H(SIZE), H(D/E) and H(CTRL), but not for H(EPS) or H(IMPACT). Further, the 1954 support is strongest when the 1968/1969 "switchback" firms are excluded from the analysis; that is, when the sample includes only firms which consistently used S as opposed to A for book purposes both before and after 1968-1969. Additional analyses used 1968 data for firms which switched from A/A to A/S in 1969. These "prior year" analyses reveal little new information except to suggest that the D/E ratio of the 1969 "switchers" rose significantly in 1969 from its 1968 level.