In the last month or so, I’ve seen some remarkably dubious studies flogged around what Lambert calls the Innertubes, all ringing changes on the same themes: outsized pay for those at the top is a reflection of a state of nature. Power laws in pay are to be expected and therefore the winners are deserving. Cathy O’Neil shredded one example of this type of
propaganda research. A more recent one, that I could not even bring myself to shred because I thought calling attention to it would serve to dignify it, tried claiming that the rising pay disparity between CEOs and average worker wages was due to the pay levels of CEOs at….hold your breath…”super companies.” A professor who does heavy-duty statistical work who looked at the study confirmed my reading: “It’s basically junk.”
Lawrence Mishel and Alyssa Davis of the Economic Policy Institute have not only gone after that junk paper, but with it, the general issue of whether ever-excalating CEO pay is justified. A new report focuses on the fact that rising executive pay has been a significant cause of the doubling of the income share of both the top 1% as well as the top0.1%. Let us not forget that CEO pay drives a host of other pay levels: other C level execs, board members, and advisors to CEOs like top consultants and top law firm partners (it’s unseemly for them to be seen to make more than their clients, so rising executive pay gives them an umbrella for charging more).
The study is clear and forcefully argued. Key points from its summary:
Over the last three decades, compensation for CEOs grew far faster than that of other highly paid workers, i.e., those earning more than 99.9 percent of wage earners. CEO compensation in 2013 (the latest year for data on top wage earners) was 5.84 times greater than wages of the top 0.1 percent of wage earners, a ratio 2.66 points higher than the 3.18 ratio that prevailed over the 1947–1979 period. This wage gain alone is equivalent to the wages of 2.66 very-high-wage earners….
That CEO pay grew far faster than pay of the top 0.1 percent of wage earners indicates that CEO compensation growth does not simply reflect the increased value of highly paid professionals in a competitive race for skills (the “market for talent”), but rather reflects the presence of substantial “rents” embedded in executive pay (meaning CEO pay does not reflect greater productivity of executives but rather the power of CEOs to extract concessions). Consequently, if CEOs earned less or were taxed more, there would be no adverse impact on output or employment.
Critics of examining these trends suggest looking at the pay of the average CEO, not CEOs of the largest firms. However, the average firm is very small, employing just 20 workers, and does not represent a useful comparison to the pay of a typical worker who works in a firm with roughly 1,000 workers. Half (52 percent) of employment and 58 percent of total payroll are in firms with more than 500 or more employees. Firms with at least 10,000 workers provide 27.9 percent of all employment and 31.4 percent of all payroll.
The entire paper is very much worth reading. It has a lot of solid lower-level arguments and evidence. For instance:
The alignment of CEO compensation to the ups and downs of the stock market casts doubt on any explanation of high and rising CEO pay that relies on the rising individual productivity of executives, either because they head larger firms, have adopted new technology, or other reasons. CEO compensation often grows strongly simply when the overall stock market rises and individual firms’ stock values rise along with it (Figure A). This is a marketwide phenomenon and not one of improved performance of individual firms: most CEO pay packages allow pay to rise whenever the firm’s stock value rises and permit CEOs to cash out stock options regardless of whether or not the rise in the firm’s stock value was exceptional relative to comparable firms.
I have one small quibble with this otherwise solid and important paper. Mishel and Davis correctly describe excessive CEO pay as rents but they fail to describe the mechanism by which these rents are extracted. The main one among public companies is breathtaking simple and is guaranteed to make sure that top executive keeps rising at a rapid clip. From a 2008 post:
While most commentators on CEO pay correctly focus on the role of options-based rewards in goosing pay from generous to stratospheric, the role of compensation consultants seldom gets the attention it merits.
One practice that I have seen get perilous little mention is where the pay targets are set. Based on their belief of what constitutes good modern practice (influenced in no small degree by the pay consultants) most boards set general target ranges for how they would like the CEO to be paid relative to peers. The comp consultant then helps define and survey the peer group’s pay ranges, setting a benchmark for how the CEO in question is to be paid.
That all sounds fine, right? Well, except just as all the children at Lake Woebegone are above average, no board likes setting a target below peer group norms. I have heard of numerous examples of targets being set somewhere in the top half (66th percentile, top quarter, top 20%), hardly any at the mean, and none I know of below average (although GE’s Jeff Immelt set his pay at a remarkably modest level, saying it was bad for morale and inappropriate for the CEO to be paid vastly more than other C-level executives). If readers know of any examples of companies (other than those with substantially owned by insiders) where the target for CEO pay is below the median of comparable companies, please let me know.
So with this mechanism in place, any CEO who has fallen below median pay who is targeted to be in a higher group will have his pay ratcheted up, independent of performance, merely to keep up with his peers, This increase raises the average and creates new laggards. The comp consultants have institutionalized a leapfrogging process that keeps them busy surveying competitor reward levels and keeps top-level pay rising relentlessly.
And there seems to be a creep in cultural values that accepts, nay endorses, the opposite process at work further down the food chain.
And as we pointed out in that post, we have a double standard as far as other workers are concerned:
Consider the way in which views that are contrary to most wage earners’ interests have been internalized (or at least are promulgated in the media). One meme I have noticed surfacing in the debate over the automaker bailout is that UAW employees are paid more than average workers.
Now in and of itself, that statement is meaningless. You need to have an idea of worker productivity to see whether that it out of whack (and for some odd reason, the bloated and highly paid management cohort almost never gets mentioned in these discussions, nor do the massive state level subsidies to the foreign transplants). Perhaps I missed it, but I do not recall seeing any longitudinal work on labor costs (that sort of analysis would help bring some badly needed facts to the table).
But why is framing the discussion around averages alone dangerous? Let’s say we collectively want to bring car worker pay down to some sort of average. That has the effect of lowering the average. You will have groups that were formerly at the average that are now above it. And if you accept the implicit logic “above average pay is bad” (fill in the blank as to why), you have a race to the bottom due to pressure on the relatively better paid to take less which puts pressure on aggregate pay.
And once you understand how that logic is put into practice, you can see how it is put into practice on all sorts of other fronts: attacks on teacher pay and benefits as “too high” now that wages and job stability for private sector workers have been badly eroded. Or as we pointed out, Greek pensions are assailed as “too high” when it could just as well be argued that pensions in other Eurozone countries have been cut too far and should be increased instead. But playing on jealousy is remarkably effective, and workers will need to recognize how easy it is to set different groups who could make common cause off against each other to distract attention from the puppet-masters who make out even better than before.