استفاده از شاخص مالمکو برای اندازه گیری تغییرات در بهره وری و کارایی از شرکت های حسابداری در ایالات متحده قبل و بعد از قانون ساربانس اکسلی

There have been many criticisms of the Sarbanes–Oxley (SOX) Act passed in July of 2002 to correct business accountability and performance practices. The act has a major emphasis on accounting and its practices. This paper attempts a response to these criticisms by investigating changes in productive efficiency for 62 of the largest US public accounting firms between the periods (2000–2001) and (2003–2004)—the periods before and after enactment of SOX in July of 2002. DEA is used to calculate Malmquist indexes of productivity and efficiency changes. This index is used because it can distinguish between changes in technical efficiency, which limit the possibilities, and changes in the performance efficiencies for each firm. Contrary to many of the criticisms, results indicate that accounting firms have exhibited significant post SOX growth in productive efficiency which is better than pre-SOX performances.

There has been a large and very vocal outburst of criticisms of the Sarbanes–Oxley (SOX) act passed by Congress in July of 2002 and signed by President Bush. This act is aimed at correcting practices believed to be associated with the recent parade of business and accounting scandals. See Prentice and Spence [1] for a comprehensive survey of these criticisms and possible responses. See also Prentice [2]. Many of these criticisms are reminiscent of the similar criticisms evoked by the Securities and Securities Exchange acts of the 1930s which represented responses to similar deficiencies in business accounting and financial practices that were believed to have led to the 1929 stock market crash and the subsequent “great depression.” These 1930s criticisms have long since abated, however, and, as noted in Prentice [2], the regulatory activities—especially the accountability and full disclosure requirements associated with these laws—have now become models for other countries to follow. See also Ijiri [3] who describes the reaction to the legislation directed to accounting reform in the 1930s as well as to SOX—which Ijiri refers to as the “reform bill,” that was in process at the time of his writing. This is all preparatory to his discussion of the need to focus on “procedural fairness” instead of the “fairness” concepts that have been a guide to FASB and its predecessors going back to the 1930s. Comment 1. The prospect of prohibiting simultaneous provision of accounting-audit services and the more profitable MAS (management advisory services) was of particular concern to the accounting industry which mounted a campaign against this legislation. However, a veritable flood of scandals, some involving accounting firms, swept aside these objections. See Asare et al. [4] who used a designed experiment to test these conflicting claims and found that their results were consistent with the claim of the accounting industry that this practice was not associated with higher risks.Other papers also appearing in JAPP that deal with effects of SOX on accounting include Zhang et al. [5], which deals with internal control weaknesses and finds that recent auditor changes are more likely to be associated internal control weaknesses. See also Krishnan [6]. Gordon et al. [7] find strong evidence that corporation information security activities are receiving more focus since the passage of SOX, and Cullinan et al. [8] who find a significant association between executive loans and financial misstatements. Thus, these studies as well as others we examined, deal with propriety whereas we deal with performance efficiencies. In this paper we report on an inquiry directed to ascertaining performance effects of SOX (if any) on 62 of the largest US public accounting firms. We selected accounting firms partly because of data availability (at the early stage, 2005, when the data for this study were collected) but also because accounting firms and practices constitute a center piece of this legislation. For this study we divided the data into two groups: post-SOX (2003–2004) and pre-SOX (2000–2001). We then employed a relatively new version of the Malmquist Index to evaluate the changes, if any, in performance efficiencies of these 62 firms. We chose the “Malmquist index” to study these changes because this index makes it possible to distinguish between effects that reflect changes in efficiency that are due to the firm performances and the effects that are associated with changes in the productive possibilities associated with legislations, etc. (The latter effects can be regarded as externalities that affect the productive possibilities for accounting performances—e.g., effects on financial markets, where critics claim that expenses arising from the need for complying with the information requirements imposed by SOX result in diversions of IPOs from US to foreign financial markets.) See Prentice [2].

As has just been noted, the changes occurring after SOX showed that these 62 accounting firms (from among the 100 biggest firms) increased their productive efficiency with the main improvements occurring in non-big 4 firms. This, of course, refers only to technical efficiency and productivity. In part we focused on technical efficiency because of lack of data availability on unit prices and costs and in part because this efficiency is necessary to other types of efficiency—such as cost, profit, and returns-to-scale efficiency. This follows from the fact that technical inefficiency (=waste) can be removed without requiring the use of any resources. See Cooper et al. [23]. Remark 7. The measure, θ, that we use is sometimes referred to as a measure of only “weak” efficiency because it does not include the non-zero slacks that change the inequalities in (2) to equivalent equations. When a “complete” efficiency measure is wanted, recourse may be had to the “slacks based” measure used in the Malmquist indexes described in chapter 11 of Cooper et al. [23]. As previously noted, to incorporate returns to scale efficiency considerations in Malmquist type indexes would require recourse to different models from the ones used here. This would also require developing extensions or alternatives to Malmquist type indexes and require treatments that could differ considerably from the models and methods reported in this paper. A study of allocative efficiency and productivity could be of considerable interest. However it would require access to information on per unit prices and costs that would take the form of salary and other emoluments for partners and others in these firms, and we had no access to such information. Indeed, if this information were available such studies could be extended to “profit” and “overall efficiency,” etc. and thereby provide insight into these additional aspects of firm performance. See Cooper et al. [23]. Other models and other objectives could also be brought into play such as the chance constrained programming formulations of DEA used in Cooper et al. [39] to provide access to other objectives such as the “satisficing” behavior described in Simon [40]. This study shows that despite the criticisms noted in the opening to our introduction, the post-SOX period showed improved performances at rates that exceeded pre-SOX rates of improvement. This is desirable, of course, but government regulations could be carried still further to improve efficiency and productivity (as well as the “profits”) in firm performances. This could lead to changes in regulatory formulations and procedures that could also help avoid worsening performances as was done in this case for the big 4 firms. See, for instance, Banker et al. [11] for what was lost in interdicting the offerings of MAS services to audit clients of public accounting firms in SOX because of practices that could have been dealt with by other means—e.g., by using the PCAOB board to audit the auditors. “Quis custodiet custodies?” asks John Stuart Mill in his Contributions to Representative Government. See Mill [41, p. 233]. We adapt this to the present context by rephrasing it as “who will audit the auditors?” A potential for answering this question in a positive manner lies in the Public Accounting Oversight Board provided in SOX. Given sufficient resources and adequate authority, PCAOB could allow accounting firms to supply both MAS and audit services to the same client as in pre-SOX with advantages to both and without involving deterioration in A&A (or MAS) performances. Had this kind of use of auditing the auditors been available pre-SOX, by, say, vesting such authority and giving adequate funding to the Public Oversight Board (POB) previously maintained by the American Institute of CPAs, the propriety objective in SOX could have been achieved with possible further improvements in efficiency gains associated with MAS services. Thus, the large MAS associated efficiencies noted in Banker et al. [11] could have been retained without the accompanying improprieties that were manifested in performances like those of Arthur Anderson & Co (a member of the “big 5”).