Network World, September 3, 2010, Friday

Network World

September 3, 2010, Friday

Network World

About that "new research" that says CEOs got rich by firing workers...

We need more excess, not less.
By John Cox

It was a made-for-viral-Internet-meme: a report issued just before Labor Day, with the title "CEO Pay and the Great Recession", from the Institute for Policy Studies (IPS), complete with an illustration of a tophat-and-tails CEO brandishing a $12 million check "for laying off >3,000 workers."

After that, you didn't really have to read it. And judging from the firestorm of media and blogger coverage, not to mention the reader comments, few did. Even fewer asked who wrote the report, and why, and whether the study and its conclusions were actually valid.

If you wanted a bit more from the report, you could just read the helpful "1. Key Findings," which condensed all those baffling complexities about corporate governance, social values, shareholders and boards, business decisions, government policies, the economy, and the web of interrelationships among them all down to sound-bite-ready nuggets.

Just add outrage.

And outrage was added. In full measure. Media of all stripes "reported" on the report. The reporting usually took the form of regurgitating the Key Findings. Thank heavens for copy-paste: "CEOs of the 50 firms that have laid off the most workers since the onset of the
economic crisis took home nearly $12 million on average in 2009, 42 percent more than the CEO pay average at S&P 500 firms as a whole."

Network World's Paul McNamara briefly commented in his Buzzblog that it was "worth noting new research that quantifies what we would all suspect to be true: Bosses at big companies bank bigger bucks the more vigorously they wield the payroll ax."

I have to confess that I never suspected this. Perhaps because I can't see the connection between the two. A CEO decides he wants "more" and thinks he can get it by firing X number of people? Is there a formula for figuring that out? Like, "If I fire 3,000, I'll increase my bucks by $1 million per year." Why stop at 3,000?

Lots of media sites gave the report a local or special-interest twist. At CIO.com, for example, Meridith Levinson concluded that the report, which is about CEO compensation and layoffs, had an important lesson for IT managers: "But there's a surefire way IT professionals with management responsibility can improve their pay: Fire people."

Apparently, Levinson read the report's section on HP CEO Mark Hurd -- who fired 6,400 employees, before being shoved out himself, and reaped a going-away-present of "$12.2 million in cash and stock worth about $16 million" -- and concluded that since HP makes computers, then people who manage computers for a living could increase their take-home pay by firing their staff.

Setting aside the several logical discontinuities here, even the IPS authors acknowledged, weirdly, that Hurd's millions are the result not of a conventional compensation package -- annual salary, bonuses, and various stock benefits -- but of a specific "severance package."

Let's start with who published the report: the Institute for Policy Studies. You could be forgiven for imagining a group of tweedy wonks delving into arcane areas of, like, policy. The IPS website, just a click away, is fairly clear on its perspective and goals. You know: the agenda. Here it is: "Washington’s first progressive multi-issue think tank, the Institute for Policy Studies (IPS) has served as a policy and research resource for visionary social justice movements for over four decades — from the anti-war and civil rights movements in the 1960s to the peace and global justice movements of the last decade."

In other words, it has a specific perspective and frame of reference when it comes to public policy. "Progressive" usually means "left-of-center" though more typically "way-way-left-of-center." This doesn't mean the new "research" is wrong, necessarily. But it does mean the research isn't value-free.

The IPS study implies that the alleged connection between increased layoffs and increased CEO compensation is directly tied to the ""worst economic crisis since the Great Depression" -- the mess we're in now. In fact, the relationship of executive compensation and layoffs is something that's been a fertile research field for well over a decade.

Take, for example, a 2005 paper, "Corporate Downsizing and CEO Compensation," by Alexandros Prezas, Murat Tarimicilar, and Gopala Vasudevan.

This study found a "large increase in CEO equity-based compensation in the year prior to and the year of the downsizing." That would seem to support the IPS study. But the results of the downsizing, and the impact on the CEO's stock-based incentives and hence his compensation, varies. The sample of downsizing firms showed "small improvements in operating performance." But these improvements were mainly in firms with relatively low equity-based compensation plans.

A separate measure, financial returns in the period following the downsizing, are somewhat higher for downsizing firms, but typically for the larger firms, for those that hire a new CEO in the year of the downsizing, that have higher leverage, and that have better operating performance.

What does it mean? That's what I asked one of the authors, Professor Vasudevan, at University of Massachusetts Dartmouth, Charleton College of Business. "In our study also we find that CEO's who lay-off [workers] do not make any extra compensation," he replied via email. "In fact, the stock price typically is negative around the layoff announcement and the long term stock price performance is also quite poor. This would imply that these CEO's do not make any extra compensation following the layoff."

That's also the broadly similar conclusion of an even earlier study, in 1998, by Kevin Hallock, Cornell University: "Layoffs, Top Executive Pay, and Firm Performance."

Here's how Hallock began his now-12-year-old paper: "The popular press and some policy groups [IPS?!?] are increasingly reporting stories of firms with highly paid CEOs that fire thousands of workers only to see large increases in the firm stock price (and their own wealth) and their pay in the following year." The great thing about short memories is you can always be outraged by the same thing, because you always think it's a new outrage.

Hallock's purpose is "to document whether there is empirical [emphasis in original] evidence for the notion that CEOs [who] are heading firms that let workers go are more likely to see increases in their own pay in the following year for making these decisions."

The conclusion: no.

"Once firm-specific fixed effects [for example, the size of the company] are controlled for, the CEO pay premium for laying off workers disappears." He adds, "In addition, there is a small negative share price reaction [ie, the stock price drops] to layoff announcements."

According to Vasudevan, "The study by Kevin Hallock...looks at this [issue] exactly like the IPS study does and finds that, after controlling for other factors such as firm size, the CEO's of layoff firms do not make any extra compensation."

"There is a very strong relationship between the size of the firm and top management pay," he says. "Larger firms (based on sales, assets) tend to pay their CEO's more money. Hence the Hallock [study], and most academic studies, control for firm size and other variables in a regression. The IPS study does not."

The result is a report that at best states the obvious -- CEOs earn a whole lot more than their janitor -- and at worst is misleading, to give momentum to a variety of federal economic and tax policy changes favored by IPS. On page 13 of its report, the IPS authors lay out their principles of "economic fairness" and then starting on page 15, a scorecard of how well a flock of provisions and proposal further these principles.

One, mandated by the new federal financial reform law, requires all corporations to report the median annual total compensation of all employees, and document the ratio beteween CEO and employee pay. "This provision could boost efforts...to limit [CEO] pay excess via tax and procurement policies."

But there could be other, far more harmful effects. A higher CEO-employee pay ratio implies that the CEO is getting paid far more than the lowest paid worker, so IPS would envision a lower ratio as being desireable. But Vasudevan points out there are other ways to achieve a lower ratio: simply cut the lowest paid U.S. workers by off-shoring and out-sourcing these jobs to, say, India. "The CEO might look like a good guy, with a lower ratio, when he is actually getting rid of the low-paid employees and sending those jobs away," he says.

One actually sound proposal, which seems to be missing from the IPS list, is one that ties a CEO's equity-based compensation (stock options and other stock-based schemes) to a multi-year time frame. "The top management have to hold on to their stock and option grants for at least three years," says Vasudevan. "The idea is that forcing CEO's to hold on to their shares will ensure that they are not able to 'flip' them and perhaps leave the company if things don't work out."

One example of a compensation program that seems to be working is the one for Ford CEO Alan Mulally. Here's how a Reuters story summed it up earlier this year:

"Ford Motor Co Chief Executive Alan Mulally's total compensation rose almost 6 percent to $17.9 million in 2009 as the automaker posted its first annual profit in four years despite a severe recession that drove domestic rivals into bankruptcy. Mulally, who is credited with spearheading a turnaround of Ford without seeking a U.S. government bailout, earned $1.4 million in salary after accepting a 30 percent cut from 2008, and received no cash bonus for the second consecutive year. He received most of his compensation in the form of company stock."

The Reuters story continues: "The only major U.S. automaker not to file for bankruptcy last year, Ford's stock massively outperformed the auto sector and the U.S. equity market in 2009, rising more than four-fold. By contrast, the S&P 500 was up about 23 percent."

"Massively outperformed." The excess of success.