Thursday, January 12, 2006

Tails of the Compensation Distribution

by Tom Bozzo

Via Brayden at Pub Sociology (via Daniel Gross via the Wall Street Journal), executive compensation scholar Lucian Bebchuk has new data on the amazing growth of executive pay:
From 1999 to 2003, the five top dogs at each of the 1,500 largest publicly traded firms cumulatively took down $122 billion in salary, bonus and stock, compared with $68 billion from 1993 through 1997.
Untangling the presentation of the data, that's an average of slightly more than $3.25 million in annual compensation in the latter period.

It's particularly remarkable in that the denominator includes a lot of third and fourth lieutenants to unremarkable CEOs, and the 1999-2003 stock market rollercoaster was not exactly a banner period for investors. Indeed, Bebchuk's data show that in the 2001-2003 period, executive pay claimed a whopping 9.8% of the companies' net income, vs. 5% in 1993-1995.

I can't agree with Larry Ribstein that 'excessive pay' — to the tune of giving away some 5% of net income for no obvious return — is "one of the least important issues in corporate governance." Ribstein's claim that "[w]hat really matters to shareholders and society is what the executives are doing or not doing for their money" is less tendentious. But the argument strikes me as looping back upon itself, as it's likely to be resolved in a finding that corporate executives are wildly overpaid (relative to a sober assessment of the value of executives' marginal products) thanks to significant departures from the a competitive labor market model, and so excessive executive pay is what really matters to shareholders and other stakeholders after all!

It does follow more-or-less immediately that, even if the SEC isn't misguided in pursuing regulation of executive compensation per se, Gordon Smith of Conglomerate is exactly right in concluding that there's no reason to believe that merely shedding more light on executive compensation and letting the 'market' work the rest of it out on its own (without blowing up some of the dysfunctional institutional structures in the executive pay 'market') will do much if anything of social value. Those structures encourage executive pay ratchets as every firm must, as a matter of policy (or at least official narrative), employ above-average executives at above-average wages. Until the "market" can inject some sense into those arrangements, there's no reason to expect a magical solution.


Meanwhile, on the other end of the spectrum, my conservative pal Bryan Smith sends me a link to a Tech Central Station article by John Henke claiming that, contra those of us who focus on negative real (inflation-adjusted) wage growth, workers really never had it so good as total compensation and disposable personal income have been modestly outstripping inflation. (*) TCS, for the uninitiated, should be taken with a grain of salt in its roles as cheerleader for corporatism and finder of the sunny side of the present economic arrangements.

In comments, I had objected in part that the focus on the faster growth of nonwage compensation costs comes in large part out of health care expenditures, where prices happen to be increasing much faster than general inflation. Bryan responded asking whether those increased expenditures thus 'don't count.' Fair question.

The answer hinges on two meanings of "real" economic variables. The mechanistic sense is dividing the current-dollar value of something by a measure of price. The purpose of such a calculation is to obtain a "real" quantity (or, for advanced readers, quantity index number) that's free of money illusion. Clearly, if you're spending more money but not getting more actual stuff with it, you're not better off in the sense of having increased "real" consumption.

What Henke tries to do is to draw a connection from the mechanistic calculation (total compensation divided by a price index) and worker welfare. Now, it's true that if health care costs weren't exploding, the increase in real total compensation probably could be interpreted as the articles author would like to suggest — an increase in workers' "real" claim on resources. But in the real world, companies are spending more money while the "real" content of non-wage benefits is declining, and the mechanistic calculation overstates what workers are really getting. (The mechanism is a bit subtle. In some ways, the quality of health care is improving thanks to new pharmaceuticals, surgical techniques, etc. However, the pace of the quality improvement probably doesn't outstrip that of the cost shifts to employees without which the measured value of employer-paid benefits would be rising even faster.)

So, while employers are paying more in total compensation, employees are not actually getting more from either the cash wage component or the nonwage component, and thus in the more important sense of the term, "real" compensation is not rising.

Henke also misleadingly deploys statistics on disposable personal income (DPI), as he puts it "compensation that we the proletariat get to pocket." He notes that DPI has increased 3.1, 2.4, and 3.4 percent in 2002, '03, and '04, respectively, adjusted for inflation. However, the DPI statistic he quotes is the national aggregate, which is subject to scaling with population. Per capita DPI, three lines up on the same table, increased by more modest amounts of 2.1, 1.4, and 2.4 percent. And, since official statistics probably understate inflation in the recent past, "true" real DPI growth is likely lower.

Even that does not accurately reflect the well-being of the "proletariat," since per capita DPI averages the proles and the plutocrats; the income distribution in the U.S. is, highly skewed in favor of the well-to-do, and income inequality has not improved. A correct calculation would eliminate the effect of high earners on the means. Additionally, any goosing of DPI due to the Republicans' intertemporal tax shifts will eventually have to be paid back in the from of lower DPI growth for somebody at some future point.

All in all, Henke does not make his case that workers are seeing as gilded of an age as their corporate overlords.

(*) The article also deploys some silly statistical arguments in an attempt to minimize recent Census data showing stagnant median household incomes.
On the top part of the post, does stock include stock options that might have some restrictions?

I do think that the market will correct for executive overpayment over time, that is, if executives are actually paid too much. College and NFL football coaches get paid multi-millions per year for coaching. I don't think $10 million for a CEO of a company that employs tens of thousands of people is outrageous, but maybe you could explain why. I think a good CEO has an obvious return on investment. See: Steve Jobs.

As for the real wage decreases, it seems to me that your argument places the blame for stagnant wages on rising health care costs, not Bush's stewardship of the economy. Wages would be increasing if health care costs weren't increasing so much faster than inflation, right? I guess you could blame Bush for not fixing the health care system, but I'm actually surprised that given the huge increases in health care costs that wages have stayed as high as they have. Even when adjusted for inflation, wages are only slightly down. In any case, I think it's a weak argument to say that the economy is bad because increases in health care costs have stalled wage growth, even though the GDP is growing at 3+% per year. That's slicing the tomato pretty thin, which I believe was Henke's whole point, considering his example of the Dole campaign in 1996.

I am sympathetic to your argument that the deficit shifts decreases in DPI to the future. Is there some indisputable number that accounts for wages, benefits, taxes paid, and the future of the deficit and properly adjusts for income inequality into one simple number? You could call it the Bozzo number, but I think it might show that working in America isn't so bad.
Bryan: The original research is here (PDF). It notes that the compensation includes the values as of the grant date of restricted stock awards as well as a Black-Scholes value of stock option awards.

Gordon Smith notes that the last round of executive pay sunlight was a disaster. IMHO you have to do something about the 'every executive is above-average' problem as a prelude to letting the market self-correct.

None of that is to say that there aren't any CEO's worth their astronomical pay. But I don't see any reason to believe that the distribution of business ability is such that the *average* for the top *7500* executives should be north of $3M.

Even Steve Jobs was very lavishly paid (not insignificantly in the form of a G-V jet) before Apple had big profits to show for its well-regarded Jobs II-era products.

As for wages, they could (in theory) be better if health care cost growth could be fixed. They could be higher yet if employee compensation were stable as a share of GDP. You can blame Republican policies, I think, for tilting the playing field in favor of capital vs. labor.

Really, what I'm saying, along with some other left-leaning economists, is that the headline growth stats are much rosier than the actual improvement in worker well-being. If these stats were really in line wiht productivity growth, TCS flacks wouldn't have to try to break the champagne out over improvements in living standards that are close to zero even in overly optimistic readings. Some of it is due to factors where Bush is simply doing nothing; some is due to deliberate policy choices.

As for the last question, the nature of combining multiple dimensions of well-being is such that the answer is 'no.'
Jobs's package with AAPL is why I won't buy the stock: those 80MM optional shares would significantly dilute the value of any purchase if held for the long term, since the amount of the company owned would decrease as the stock price rose.

So I'll severely dispute Bryan's contention, because the hidden cost will more than outweigh the value of AAPL--unless the content distributors are really stupid enough to leave AAPL with a monopoly.
Sorry again to post on an old topic, but I read more about this in the WSJ online ($$$$, this free link might work). The author, Jesse Eisenger, argues that CEOs are paid way too much and that shareholders should take notice. He thinks hedge-fund managers should be paying attention, but since they make >$25 million maybe they aren't really paying attention.

However, maybe the hedge-fund managers like the current setup, though, because they realize that excessive CEO pay will hurt a company's bottom line and therefore they can short that stock or buy stock of companies who will outperform because their CEOs aren't overpaid. Maybe that's why hedge fund managers seem a bit indifferent to CEO pay...
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?