Prepared remarks for the SEC Hearings on September 13, 2000.

Rajib Doogar

Thank you for giving me an opportunity to appear before the Commission

Audit markets, in general, and, in particular, in the United States, have some distinctive features that make regulating audit firms activities a tricky task. The three key features I'll focus on today are:

  1. Doing a credible audit produces information also useful for consulting. Society is better off if this knowledge, once created, is efficiently used. Why force another firm to reproduce the information?

  2. The Securities Acts mandate audits of all SEC registrants. When demand for audits is mandated, market prices (audit fees) need not reflect the value of the audit to investors.

  3. Under United States law, auditors accused of culpability in the failure of a client can settle lawsuits out of court prior to an adjudication of the plaint on merits. Thus after an auditor lawsuit has been settled, investors never find out if the quality of the audit should be questioned.

These three features are relevant for interpreting my research findings. Before I present my principal findings, let me make a couple of comments to get the context of my research right as well.

You, as regulators, are concerned that auditors, tempted by the profits to be made by selling other services to their audit clients, i.e. what economists call rents, have gone too far towards efficient use of information at the cost of impaired audit credibility. This is quite plausible and even possible. This is the possibility I investigate in my theoretical analysis, an executive summary of which you already have. In your recommendations, you propose restrictions on the scope of the firm's services as a fix for the perceived lack of auditor credibility. However, to credibly link specific restrictions on the scope of the firm's services to your goal of enhancing audit credibility, you must argue that rents earned from performing one task are more tainted than rents earned from performing any other task. Yet a simple economic analysis would suggest that there is no difference, in the power to impair audit credibility, between the desire for profits from the provision of additional services or even from future provision of audits. Since I do not have a good theory of why the source of the rents should make auditors behave any differently, in my research I do not distinguish between various sources of rents. That's the challenge. Now to the findings.

In response to your proposals you have been told, first, that the proper scope of a firm's services should be left to market forces to decide. Second, you have been told that investor perceptions about audit credibility are not an appropriate basis for action. Analysis of the second and third features I have just highlighted shows that when audits are mandated, i.e. the market is not free to discipline auditors, both these arguments are defective.

Under the current rules in the United States, there are good reasons to believe that the market, on its own, cannot fix the problem of how much and what services an auditor should provide an audit client. Markets work well when prices reflect buyers' utilities for goods and services. Mandating an audit weakens the link between investors' utility for the audit and the fees paid for auditing. This, in turn, creates a divergence between what investors would want auditors to do and what profit-maximizing auditors might do. In my model this shows up as auditors having incentives to produce a mix of more non-audit services than investors would prefer. In principle, regulators may be asked "to do something" to reconcile auditors' incentives with those of investors. But since this is a problem created by mandating the audit in the first place, short of un-mandating the audit, in my analysis, there is no real fix for this problem.

The second finding of my research concerns the implications of the current litigation system for the well-functioning of audit markets. When a client fails and auditors are sued, in US law, auditors have the right to, and often choose to, settle the case prior to a disposal on merits. After such a settlement, investors never learn if the aUditor did a good audit but chose to settle to avoid further costs or did in fact do a bad job and settled to avoid public identification. Again, because auditor guilt or innocence cannot be established publicly, markets cannot do a good job. It is important to note that in my analysis, this effect is independent of whether audits are mandatory or voluntary. Low audit credibility, in turn, will drive up costs of capital, affecting the well functioning of capital markets and indeed of the US economy as a whole

What can or should you do about it? In my view, academic research is better at explaining than at prescribing. However the frictions it identifies can help in the search for remedies. I'd like to take 90 seconds to identify some things to think about if I may.

I believe three lines of thought could be profitable:

First, to allow market forces to discipline auditors, un-mandate the audit. Let she who wants an audit pay what it is worth to her.

Second, even after audits are un-mandated, to promote economic efficiency, it will still be necessary to have standards and you can encourage private sector entities to develop disclosure standards. You may even encourage multiple levels of quality and multiple standard setters to provide healthy competition among standards. This way your actions could not be interpreted as signaling mistrust about audit firms - rather you would simply free the market to decide whether the extant standards were good enough.

Third, there will still be a need to vigorously monitor whether entities that raise capital from the public do in fact comply with the standards they claim to adopt. Here I believe that the SEC and private bodies such as those sponsored by the accounting profession, by users and by organizations like Consumers Union can jointly assist market participants and entities that will compete in setting standards to deal with the monitoring task. The entertainment industry uses a voluntary rating system to classify its products and because individual producers have incentives to mis-classify their products, a class of private information intermediaries, in this case, parenting magazines have sprung up to satisfy consumers' information needs. Underwriters' Laboratories and Consumers Union are two other examples of entities that provide information useful to businesses and consumers without having formal regulatory oversight of their activities. These examples could offer starting points for thinking about the most cost-effective means of ensuring that capital markets obtain high-quality, relevant, information.

This solution will promote new industries, harness the ability of markets to discipline private conduct and allow human capital to flow to its highest valued uses, reserving the machinery of the State for what it does best: upholding laws and contracts and for investigating and punishing crimes. Notice that because financial statement credibility still needs monitoring, even if you were to un-mandate the audit, a strong and vigorous SEC would still be required to prevent fraud. And, while this is not a clincher, this approach would also be more in line with your prior policy of disclosure regulation rather than merit regulation.

Doing anything else, and in particular, undertaking something as ambitious as presuming to adjudicate the boundaries of the firm, a task that markets themselves are hard put to determine, is something a wise regulator might approach with an angel's caution since such actions suggest implicitly that one knows better than the market.