The K Street Perversion of Cost-Benefit Analysis

[twitter-user link=”yes”]

Business efforts to stifle a simple, congressionally mandated rule requiring public companies to disclose the ratio of CEO pay to the median-paid employee provides an insightful lens on the perversion of cost-benefit analysis.

Approved as part of the 2010 Dodd-Frank Wall Street Reform Act, Section 953b counts as the simplest of the 400 statutes that regulatory agencies must translate into rules with which business must comply. The U.S. Securities and Exchange Commission (SEC) proposed a rule in October, setting December 2 as the deadline for submitting comments. Public Citizen filed a formal comment with the SEC. In addition, more than 40,000 Public Citizen members and activists filed comments in support of this rule. To date, more than 105,000 commenters have filed letters in support of this rule.

This support reflects investor anger that CEO and senior management pay has escalated beyond economic justification. Pay for top-level execs now drains 10 percent of corporate profits where in 1990 senior management pay took only 5 percent. When finalized, the new rule will allow investors to “unit-price” CEOs. An investor weighing, say, a purchase of Oracle stock versus Hewlett Packard can now look to one more variable: Is the CEO relatively inexpensive, or costly? In this case, the Oracle multiple for CEO Larry Ellison would be more than 1,200 (expensive!) while Hewlett Packard’s Margaret Whitman is about 350 (certainly cheaper).

But K Street attorney Eugene Scalia (son of the Supreme Court justice) pioneered a roadblock for all such rules: cost-benefit analysis. In the case of “Business Roundtable v. the Securities and Exchange Commission,” he convinced the court that the SEC’s cost-benefit analysis of another mandated rule regarding nominations for corporate boards didn’t fully appreciate all the costs. In that specific case, the Business Roundtable argued that unions would cause pension fund managers to violate their fiduciary duty, which let them use the nomination of candidates to corporate boards as a bargaining tactic. It wasn’t that the SEC failed to respond to this absurdity; the court agreed with Scalia that the SEC didn’t respond enough.

Come now the K Street hyperbolists who claim that the junior high school statistics required for 953b to find the median-paid employee may cost millions of dollars at each company. We’re not making this up. According to the Federal Register (Oct. 1, P. 60587), here are the estimates proffered by industry representatives who met with SEC staff (it is quoted verbatim lest it be questioned):

• “Approximately 201 to 500 hours per year, plus significant costs;
• $3 to $6.5 million for a multinational manufacturing company with 90 separate payrolls;
• $4.725 million for a multinational consumer products company (including an estimated 50 hours per country for employees located in 80 countries);
• $100 million dollars for a multinational company; and
• $350,000 to implement plus $100,000 a year for ongoing compliance for a global technology company.”

On its face, the numbers are contradictory. How can one company do it for $100,000, but another can’t figure it out for less than $100 million?

The SEC doesn’t name names in this proposal, but it does post the name of the six companies with whom it met on a separate website. By process of elimination, it can be assumed that IBM, with 400,000 employees, qualifies as a “global technology” firm estimating compliance cost of $100,000. And it is Exxon-Mobil that claims that compliance would cost $100,000,000.00. Exxon employs 79,000. Exxon is claiming, in effect, that it would cost more than $1,000 per employee to calculate which one is the median-paid employee.

Public Citizen does not find this figure to be credible. In fact, children usually advance more believable excuses to escape doing the dishes (“Homework, Dad”).

Even if one counted every employee at Exxon, it would not cost $1,000+ per employee. But one need NOT find each employee’s compensation. If a firm with 100,000 workers pays 10,000 of them a minimum wage and employs them part-time, then this simply means the bottom 10 percent is established. The specific compensation of each of these bottom 10 percent employees need not be identified. If 30,000 are full-time and paid minimum wage, then the bottom 40 percent is established. If the next-best-paid 20,000 employees earn more than the minimum wage, then only about 5-10 percent of them need be examined to identify median employee pay.

The extensive industry lobbying effort to claim that the calculation of the ratio is expensively complicated may be motivated by a wish to protect high senior management pay. Phil Angelides, who led the Financial Crisis Inquiry Commission that investigated the economic collapse of 2008, observed, “The fact that corporate executives wouldn’t want to display the number speaks volumes.”

Public Citizen has documented the lobbying effort. A Public Citizen report in 2011 found that industry lobbyists have spent more than $4.5 million trying to avoid the 953b rule completion. This figure has undoubtedly grown in the two years since publication. Given that IBM contends it will cost $100,000 a year to comply, which we believe is exaggerated, that means the lobbying effort for one year could pay for the compliance costs that 45 companies of IBM’s size would supposedly bear. IBM, of note, is the nation’s second largest employer.

The sad fact remains, however, that if a company claims a cost, the SEC seems to feel duty-bound to record it. Cost benefit analysis, which may sound benign, becomes a dangerous playground for policy-making.

Bartlett Naylor is the financial policy reform advocate for Public Citizen’s Congress Watch division. Follow him on Twitter at @BartNaylor.

Sign up to receive a weekly email highlighting the best from Public Citizen’s blogs.