بررسی اثرات در مورد صورتهای مالی قانون هزینه شکایت های قانونی در یک اقدام مدنی بخاطر سهل انگاری در برابر حسابرس

The litigation cost rule relates to which of the two parties in a civil lawsuit has to pay the legal costs. In those countries where the American system applies, each has to pay their own legal costs. In most other countries, the British system applies and the loser pays all the legal costs. By means of a single person decision-theoretic model, I examine the effects of this on auditing and financial statements under certainty and uncertainty conditions. Generally, the American system has the effect of restraining shareholders from suing unless they are able to cover their legal costs, thereby providing scope for under- or over-statements, depending on management’s wishes. This scope is denied under the British system and, as a result, audited financial statements will not be similarly biased.

Although class actions by investors against auditors and management are common, opportunistic earnings management (and worse) is apparently widespread and most importantly, there is negligible deterrence. Almost all actions against auditors are settled out of Court with no admission of negligence.1Choi and Pritchard (2012) and Ramseyer and Rasmusen (2013) have argued that in the US, despite the Private Securities Litigation Reforms Act 1995, litigation rates are too high in the area of securities litigation, bringing an otherwise fair and efficient legal system into disrepute. What is more, the litigation costs fall largely on the defendant corporation,2 its shareholders bearing the cost. There are probably many reasons for this occurrence. Coffee (2001) hypothesises that as the costs arising from litigation risk diminish accounting irregularities should increase and cites empirical evidence by Heninger (2001) of a positive relationship between measures of earnings management and litigation against auditors. This paper examines two related aspects: (1) the effect of the litigation cost rule as to who pays the costs (the plaintiff or the defendant) and, therefore, the likelihood of a class action and (2) the effects of this on auditing and opportunistic accounting. The rules as to who incurs the legal costs in a legal action differ across countries. In some jurisdictions the loser pays not only its costs but those of the winner. This is known as the ‘loser pays all’, ‘British’ or ‘European’ system and applies in, for example, Canada, Australia, Hong Kong and most of Europe. Elsewhere, the American system operates in the US and Japan and, to a less extent, in China,3 where each party pays its own legal fees irrespective of who wins. As a result, the American system creates significant ‘transaction costs’ for the winner (which may or may not be significant), whereas the British system does not. The purpose of this paper is to investigate whether litigation costs and the litigation cost rule affect the financial statements and contribute to a bias in financial statements when the differences in opinion between management and auditor are sufficiently large as to affect the implied value of the corporation. In other words, Type II errors: when the auditor accepts financial statements that are materially misstated. An important and obvious constraint on the auditor’s decision is the possibility of the shareholders as a class (or another party adversely affected) pursuing an action for negligence against it for misleading financial statements. If they win the action, they will have to pay their own costs under the American system. Therefore, they will only consider action if the expected damages are likely to exceed the costs incurred in bringing the action. Under the British system, as long as they win, they will incur no costs. Therefore, they are free to sue whatever the value of the damages. The paper is arranged as follows. In Section 2, I review the literature on which the analysis is based. In Section 3, I examine the costs and benefits arising to the auditor relating to disagreements with management in the context of a single-person decision theoretic model. I show that, in the absence of transaction costs for the plaintiff under the British system (if it wins) the auditor will act independently and, as a result, audited financial statements are unlikely to be biased. In Section 4, I introduce legal costs arising to the plaintiff from a legal action against the auditor for negligence (the American system). I show that not only do these create scope but provide an incentive for bias, although it is unclear how large the size of the bias is. In Section 5, I discuss some of the implications of this analysis in practice. In Section 6, I relax the certainty assumption and examine how asymmetric information may affect the results. I show that whilst the impact on accounting bias is largely unchanged, the likelihood of an action is significantly different under the two systems. Finally, in Section 7, I discuss the welfare aspects and the implications of the results.

Class actions by investors against auditors and management are common (particularly in the US) yet opportunistic earnings management is apparently widespread. This suggests deterrence is negligible. I have examined how differences between the British and American litigation cost allocation systems contribute to this. These are summarised in Table 2. More specifically, I have examined the relationship between the auditor and management when management wishes its corporation’s financial statements to imply a different value to that which the auditor thinks they should. I have shown that, where effectively there are no transaction costs, there will be no scope for discretion by the auditor and, as a consequence, audited financial statements will be unbiased. However, as transaction costs in the form of litigation costs incurred by the auditor are different, the equilibrium will be different under the American and British systems litigation regimes. I have argued that effectively transaction costs exist in the form of legal and related costs under the American system but not under the British system as these are passed onto the auditor as defendant. This leads to bias in financial information under the American system to the extent that the auditor would agree to a mild under- or over-statement (VCA or VCB in my notation) depending on management’s objectives, rather than an unbiased or ‘true’ figure (VT in my notation). It should also be remembered that the readers of financial statements do not necessarily know the objective functions of management and are therefore unable to even know the likely direction of the bias. Table 2. Litigation cost effects of the American and British systems compared. System American British Accounting bias Without a rogue auditor Systematic bias No bias With a rogue auditor Systematic bias More systematic bias than under the American system Class actions More cases including frivolous actions for which there is a greater incentive More cases including frivolous actions although there is a smaller incentive Contingent fees More cases but legal costs are less certain, does contribute to accounting bias More cases but legal costs are less certain; does not contribute to accounting bias More likely to be offered by lawyers than under the British system Level of spending on legal costs High level Higher level Plaintiff Optimistic Many cases More cases Discourages small lawsuits Encourages small lawsuits Pessimistic Few cases Fewer cases Discourages small lawsuits Strongly discourages small lawsuits Out of Court secret settlement (saving reputational costs) Where there is agreement between the parties reprobability of success A strong incentive to settle unaffected by the probability of success A strong incentive to settle: higher for the party with a low probability of success Where there is disagreement between the parties reprobability of success A strong incentive to settle but a narrower settlement window, leading to a lower likelihood of settlement A strong incentive to settle but a wider settlement window, leading to a higher likelihood of settlement Table options I have also examined the two systems under uncertainty in various situations; notably, where there is asymmetric information and uncertainty for (1) the auditor in ascertaining the corporation’s true value (Section 6.1) (2) shareholders in knowing if they have been provided with misleading financial information (Section 6.2) and (3) shareholders in assessing the most likely outcome of a lawsuit (Sections 6.3 and 6.4). The general effect of this analysis shows that the differences between the effects of the two systems on the likelihood of bias identified under the certainty assumption are repeated, if not exaggerated, under these conditions. When the auditor’s difficulty in ascertaining the true value (VT) was considered, it was found that the size of the bias under the American system depended on whether plaintiff’s legal fees were fixed or contingent on winning. The probability of detecting misinformation similarly affected the size of the bias under the American system. Nevertheless, the principal reason for the difference between the two systems arose out of the effect of the threat of audit switch under the British system which was unaffected by the introduction of uncertainty with the exception of where a ‘rogue auditor’ is present, either as an incumbent or alternative auditor. Undoubtedly, there is a welfare effect deriving from the accuracy of accounting information (increasing the likelihood that VT is reported) and audit quality. The raising of these is beneficial to society if they result in better investment decisions, a lower cost of capital, improved resource allocation and, even reduced investment fraud. No doubt there are other benefits as well. On the other hand, there will be additional auditing and legal costs involved in achieving this and these may not be covered by those additional benefits. 19 It follows (and is axiomatic) that litigation in the area of opportunistic earnings management is worthwhile to society if it is to have a deterrent effect and a compensatory value and that these benefits exceed both the legal costs involved and the additional accompanying public administration expenses. In contrast, litigation will be worthwhile to a plaintiff if the expected judgement and settlement it receives exceed the expected legal costs it incurs (Shavell, 1982). Clearly, these two cost–benefit calculations are different. Also, as these are different under the British and American systems, even if all other things are equal, the levels of litigation, expenditure on litigation and deterrence under the two regimes will be different and these are unlikely to equate to the socially desirable level. It was mentioned in the Introduction that both the settlement payments and litigation costs fall largely on the defendant corporation, its shareholders bearing the cost. This is, of course, different under the two systems for similar reasons. Although under both, the shareholders are effectively paying the award, under the American system they are also paying the company’s litigation costs and under the British system they are paying both parties’ litigation costs. In which case, shareholders stand to lose more under the British system. However, that is where the defendant is the company or its directors. Although my main interest is where the defendant is the auditor, the deterrence effects of actions against management may influence their attempts to mislead.20 The important point is that a wide settlement window, particularly that arising from a secret out of Court settlement, enables the defendant to avoid reputational costs. There are two main effects of this: (1) investors are unable to distinguish between ‘good’ and ‘bad’ auditing firms preventing them from competing on quality and (2) reputational risk is not eliminated but reduced and failures (such as Arthur Andersen’s collapse as a result of Enron) may occur and destroy an audit firm’s reputational capital. Finally, the assumptions and limitations: these theoretical results may not necessarily be good predictors of reality. As Donohue (1991) argues, they are sensitive to the model’s assumptions which may be unreal.21 Here, there are a number of simplifying assumptions and limitations. It is assumed that the two regimes are pure in the extreme: all the winner’s legal costs are shifted to the loser (the British system) or none (the American system). However, in practice, this is unlikely to always be so; according to an authoritative review of policies and practices by Jackson (2010), invariably this is at the discretion of the Court. It has been assumed here that in all other respects, jurisdictions are similar. This is of course, not true, particularly in many of those countries where the British system applies, notably continental Europe, i.e. outside the UK, as their capital markets are very different (for an analysis, see, for example, Coffee, 2002a, Coffee, 2002b and Coffee, 2002c). Some are code law countries, providing weaker legal protection to investors than common law countries. Their judicial systems are also less efficient than many other European countries. The lack of legal protection and undeveloped capital markets explains why most have a higher level of corporate ownership concentration than, say, the US and the UK. Instead, the banking sector is more important, and capital markets play a secondary role. As economic agents are able to use alternative sources on which to base their decisions, they do not demand a high quality of accounting information (Burgstahler et al., 2006). Also, whilst the US has a severe audit quality regime, many other jurisdictions, including those in most continental European countries, have more flexible ones. Here, there are few pressures on the auditor from financial and other institutions, society or the stock market, there is no obligatory auditor rotation and negligible risk of litigation. The effect has been that in these jurisdictions, there have been hardly any lawsuits brought against auditors (Geiger et al., 2006) and less pressure to ‘cook the books’ than for, say, American managers (Coffee, 2006, p. 80). There are other jurisdictional differences. Some allow class actions, out of Court settlements and lawyers to charge contingent fees; others do not because of the public policy implications. The effects of these are summarised in Table 2. There are other limitations to the analysis. The first is, of course, that it is limited to an application of a single person decision-theoretic model which prevents an examination of the strategic interaction between management and the auditor. Second, I have limited the definition of legal costs to include explicit legal costs and ignored the possibility of opportunity costs involved in a legal action such as the time spent on the action that could have been employed in income generation and so on. This may well be significant. However, this is equally likely to exist under both the American and British systems and should, therefore, be ignored in a comparison of the two. Nevertheless, the extent to which it does exist, bias will also occur under the British system. Finally, whilst I have been able to specify the bias in the American system as the difference between the maximum under- or over-statement that may occur without the auditor being sued [(VT − VCA) or (VCB − VT)] I have not been able to directly quantify it or assess its importance in practice although inferential evidence, mainly from a UK review of litigation costs and policies throughout the world ( Hodges et al., 2009 and Jackson, 2010), suggests its effect in practice is likely to be significant.