I found the issues in the paper fascinating, and I too was intrigued by what damage the CGI industry is doing compared to the status quo without CGI. If the ideal is a marketplace of CGI firms that use metrics perfectly matched to both investor preferences and market-tested performance norms, then is what we have a second-best solution or no solution? Without groups doing the work of CGI, we have scattered investors with differing access to information and analytical tools, as well as differing incentives to collect and analyze information. In steps the CGI firms to reduce the collective action problem, but we suspect that there are some principal-agent problems both in terms of ultimate goals and specific conflicts. But, is this world better than the alternative? And, given that investors have heterogenous expectations, can a CGI firm adequately represent all investors? Perhaps a thousand CGI flowers should bloom (the "Green Metric," the "Market-Tested Metric," the "Investor Power Metrick" etc.)

I think it's interesting to analogize CGI firms both to other organizers of individuals with rights/claims such as unions and plaintiffs' attorneys. Because the individuals have so little power, the organizers have quite a bit, but may have their own agendas. Which world is better? Also, these types of private rankings systems make a nice analogy to USNWR. Do you rail against USNWR as a manager? Do you try to negotiate your ranking? Are the rankings helpful to the investor/consumer as to the underlying fundamentals or just as to the market reaction to the ranking?

Mercer raises some very interesting points. I appreciate his references to the SEC’s other recent efforts to regulate corporate governance, and I agree that the paper would benefit by a more comprehensive discussion of these efforts. He mentioned that a discussion of the required disclosure of proxy voting policies could inform the paper, and the sad irony is that original thoughts for this paper were generated from my consideration of that debate, and whether the disclosure of those policies resulted in any change to the policies (one of several empirical projects that I have not been able to get off the ground).

There is a very interesting debate in the literature on how much shareholders should be involved in governance. I generally take a similar position as two of my law school professors, Steve Bainbridge http://papers.ssrn.com/sol3/papers.cfm?abstract_id=808584 and Iman Anabtawi http://papers.ssrn.com/sol3/papers.cfm?abstract_id=783044, who have both expressed skepticism of efforts to increase shareholder power. The rise of the CGI works hand in hand with the trend of increased shareholder activism, and although I do not want my paper to focus on that debate, I think Mercer is right that more needs to be said on the issue. Much of the increased shareholder activism is enabled by the CGI’s ability to collect votes on a particular issue—in the past, shareholders could not speak with a unified voice; the CGI helps give them this voice. But I wonder whether investors are really thinking about the implications of this power. As I mentioned in response to Larry’s comments, I don’t think that investors think much (or enough) about the recommendations and their implications—many ISS clients, for example, simply allow ISS to vote their proxies for them, without any client review, simply according to ISS’s policies.

Mercer’s other major point is to ask whether the CGI is not simply responding to a market demand—they are giving investors exactly what they want. I do make the point that many of the CGI’s recommendations are based on “myths”, but I do not believe that, in the case of the general fund manager (“social investing” funds excluded), those unexpressed preferences have any design other than profit. I do not think, for example, that funds have an unexpressed preference for more stakeholder-oriented companies, rather than companies that are more narrowly interested in providing long-term shareholder value. They are focused on financial performance, and the corporate governance industry claims to be giving them the tolls to achieve greater profits. On the other hand, I am concerned that the CGI may have more political motives than the funds (despite its efforts to couch it recommendations in terms of financial performance), and I am troubled by the fact that the CGI instead promotes its product as a means of increasing corporate performance, when in fact there is little evidence that the measures they promote do, in fact, promote corporate performance.

Mercer also notes that if the SEC put in place a king of governance coding system that will help identify relevant governance information (will XBRL get us there someday?), there would probably always be a market for someone who can provide context and guidance with respect to the information. If some information is not collected because the SEC finds by its reckoning that it is not relevant to performance, I suppose that some investors will continue to care about it, simply because it says something about the company that may not be related to performance but to other concerns—say, whether the company is a good corporate citizen, regardless of whether this helps the bottom line. Understanding these preferences, he suggests, is a more interesting problem. As Mercer noted (to paraphrase), I think that the market does not do a good job at providing investors with the product they desire, but that I believe they will get better at it as time goes on. Mercer believes that the market may already be giving the investors what they want—the market is reflecting investor preferences already. I am not sure that it is. First, the CGI market is not a very competitive market—it is dominated by ISS, and ISS has not been under much pressure to scrutinize the value of its ratings methodology (although it did undertake a significant review last year, perhaps in response to increased competition from GMI, Glass Lewis and the Corporate Library). ISS’s market dominance leads me to think that any unexpressed preferences are more likely to be ISS’s preferences than investor preferences, since they do not face significant pressure to cater to investors’ needs. However, ISS packages its governance metrics in performance language: the governance metrics, if implemented, will improve financial performance. To me this suggests that if investors are driving the metrics, they want the metrics to lead to increased performance. The problem, however, is that the metrics often do not lead appear to lead to better performance. I do not think we are at the point where we can ask Mercer’s question of why investors want what they want, because it seems that they just want improved performance. I don’t think that most investors recognize the disconnect between many of the CGI recommendations and improved financial performance (often they do not have the time to focus on the issue more deeply, which is why they outsource governance rating in the first place). To read ISS' discussion of its evolving governance methodology, you would think that every metric in their methodology matters to performance, and that ISS has proved this through extensive study. However, when you look at the studies that have been conducted (such as Brown & Caylor’s study, which was commissioned by ISS), you see that only some of the metrics have a connection with performance, and it is unclear whether other factors might explain why those metrics affected performance. That said, I will certainly think about this issue more deeply, because if the CGI is really just responding to investor preferences beyond financial performance, you can’t really fault the CGI—instead, as Mercer suggests, you must consider why investors want what they want, and further, whether they want things that might conflict with each other (e.g., a certain set of socially-conscious but costly governance standards might limit returns).

Thanks to Mercer for his insightful comments.