William Watson: Here’s evidence on why trying to limit CEO pay is absurd

Executive pay regulation in the United States leads to a litany of unintended and mainly harmful consequences, according to a new paper by Kevin J. Murphy of University of Southern California’s Marshall School of Business and Michael C. Jensen of the Harvard Business School.

That pay rules and laws don’t work out for the best shouldn’t be a surprise. They’re designed, as Murphy and Jensen write, by “uninvited guests to the bargaining table who have had no real stake in the companies being managed and no real interest in creating wealth for company shareholders.” (In fact, shareholders are overwhelmingly satisfied with existing arrangements. A “Say on Pay” advisory for 2016–17 shows 71 per cent of 2,444 firms on the Russell 3000 Index received more than 90-per-cent approval rates for executive pay among shareholders and only 1.8 per cent of firms’ plans didn’t get majority support among shareholders.)

Murphy and Jensen provide many examples of how laws and rules on CEO pay have had perverse effects. For instance: In 1993, Congress and the Clinton administration declared as unreasonable all compensation in excess of US$1 million to the CEO and four highest-paid executives of any public firm unless it was in some way related to the firm’s performance. If it wasn’t, the firm couldn’t count it as an expense.

What happened? Stock options being deemed performance-based, payment in stock options boomed, while base salaries fell from 41 per cent of executive compensation in 1992 to just 18 per cent in 2001. And because under then-existing accounting rules stock options had no cost to the firm, compensation ballooned. Hello, one per cent!

What else happened? Base pay of US$1 million became a corporate marker. Many companies did lower the salary component to US$1 million, which is not surprising, since anything higher was no longer deductible as an expense. But many other companies raised their salary component to US$1 million — Congress and the president having given the Good Housekeeping Seal of Approval to that amount.

The law also had the effect, since compensation had to be performance based, of causing many corporate compensation committees to adopt formulas — in many cases, simple-minded and excessively generous formulas — in place of discretionary awards arrived at after detailed and sometimes unpleasant discussions of the previous year’s performance. “By failing to make (these) inherently subjective appraisals, directors are breaching one of their most important duties to the firm,” Murphy and Jensen write. Legislators may not have intended this effect. They got it anyway.

Fast forward to 2018. In a populist flourish that the mainstream media mainly missed, the new Trump tax law retains the US$1 million Clinton limit, even though inflation has eroded its value to US$553,000 in 1992 dollars, but it includes all forms of compensation. That’s anti-corporate in that it means compensation costs greater than the US$1 million will be almost impossible to write off: until genius lawyers create ingenious loopholes, that is. Murphy and Jensen estimate that where the Clinton rules disallowed about 11 per cent of total compensation expense in 2015, in the same year the Trump rules would have disallowed fully 79 per cent.

But the act’s silver lining is that by not exempting any form of income, the new rules should end tax arbitrage across different means of compensation: “Compensation committees are now able to make tradeoffs between fixed and variable compensation, or discretionary or non-discretionary pay, without worrying about the different tax implications.”

People who favour higher taxation often argue taxes don’t really matter in corporate decision-making. Reading Murphy and Jensen destroys that notion. CFOs are always paying attention. To them, taxes do matter. A lot. The stampede from straight salary to stock options after the 1993 tax changes is one example. Another is that in 2006 when new rules came in for evaluating options companies switched wholesale to providing shares in performance-based ways other than options.

Likewise, a 1984 law limiting the generosity of “golden parachutes” to the executives of companies taken over in mergers “led to a proliferation in change-in-control agreements, which had previously been fairly rare.” Moreover, the typical parachute was the very same “three times base salary” that the law established as the tax-deductible limit. Once again, an intended ceiling became a floor.

Even relatively anodyne rules about disclosing executive compensation have their downsides. In 1934, just after the Securities and Exchange Commission was formed, proxy statements typically ran three to five pages, with just one page dealing with executive compensation. By 2007, Jensen and Murphy report, they averaged 70 pages, virtually all devoted to compensation.

The considerable effort that goes into writing such reports and both gaming and complying with detailed regulation of executive compensation is brainpower and hours that don’t go into producing output. No wonder productivity growth is glacial.