April 17, 2016

In a 1996 article in Mother Jones magazine economist Paul Krugman (now a N.Y. Times editorial writer) explored the possible causes for growing income inequality.  Yes, twenty years ago income inequality was a “thing.”  A more recent study using hard granular data contradicts a widely held belief that income inequality has grown substantially in the past four decades.

I’ll look first at the older piece by Krugman.  I should note that Krugman was writing for a popular magazine, not an academic journal.  As with his weekly column for the N. Y. Times, Krugman’s goal is to sell his audience a story, to stir up the pot.  Critics who judge an editorial column with the same rigor as an academic paper should be forced to write “I will be kinder to my fellow human beings” one thousand times with a #2 Ticonderoga pencil till they get little pencil indentations in the pads of their fingertips.

Some, including a few contenders in the current Presidential campaign, sound the alarm.  “We’re shipping jobs overseas!”  Well, Krugman considered imports to be a possible cause for inequality but found that imports of goods from third world countries, where wages are markedly lower than the U.S., were only 2% of national income.  Stopping imports from those countries would have but a small effect on workers’ wages.

What about technology? Krugman asked.  As an example of technology, Krugman noted that the educational level of schoolteachers and corporate CEOs are approximately the same yet there is a wide chasm between the compensation of those in these two professions.  If Krugman were my brother-in-law at a Thanksgiving dinner, I’d have argued with him on this one.  “Hey, Paul, whaddayatawkin?!” I would have said, “CEOs are managing way more employees and capital than any teacher!”  Paul would have agreed with me, of course, and apologized for the error of his ways and I would have passed him the cranberry sauce.

 A 2013 study by Faleye, et. al., published in the Journal of Banking and Finance, used a systematic analysis of a salary and pay database to calculate the ratio of CEO pay to the average pay of an employee in the ranks of the company.  They identified and ranked a number of factors to explain the Pay Ratio, as I’ll call it. They found that three factors, ranked in effect, were most important, and by a wide margin:
1) size of the firm, or market cap;
2) tenure of the CEO;
3) return on assets, or profitability of the firm in a given year (Table 3).
Of those three factors, the first two factors, size and longevity, influenced the Pay Ratio far more – three times – than the third factor, the return on assets.

Let’s look at the first factor: the size of the firm.  We’ll use the stock market capitalization (CAP) as an indicator of size.  In the post war period from 1957 to 1980, 23 years, the CAP in the U.S. increased by 2.5 times.  In the subsequent 35 years since 1980, that CAP shot upwards over 13 times, even after two severe market downturns!!

So, CEOs are being paid to be responsible for the deployment of a lot of capital.  Remember, the return on assets (#3 factor) was much less important, so CEOs are being paid even if they don’t do particularly well with that capital in a given year.  Factor #2 was longevity so we can guess that CEOs who do perform well stay in the post.  Those who don’t get the proverbial boot.

Let’s turn to “Fee for service” financial advisors for a comparison.  Advisors typically charge 1 – 2% to manage money for their clients.  Naturally, an advisor – or CEO – who manages a combined $100 million will make more than one who manages $10 million.  The workers at each firm may earn approximately the same but the CEO of the larger firm should make more money and greatly increase the Pay Ratio.

What else can skew the ratio upward?  Using a mean, or average, rather than the median, the halfway number in a data set. The authors of this study found that average CEO pay was almost twice the median CEO pay, indicating that a relatively small number of very well paid CEOs skewed the average upward.  Krugman and other economists (Robert Reich, for example) touting inequality in the popular press use average, not median, CEO pay simply because the average shows a higher ratio than the median. “Whaddayadoin?!” I’d have to challenge Paul at the dinner table.   “In this case, an average gives a distorted view of the data,” I would protest.  Paul might smirk knowingly and grab the last helping of mashed potatoes while I was protesting his dirty, no good argumentative trick.

This study included 447 firms whose total revenues averaged almost $2 billion.  Apple had $600B in revenue last year so the study included both mega firms and large firms.  They found that the median Pay Ratio was 52, not 331 as the AFL-CIO claims, or 150 times, as Krugman claimed in this 1996 article.  Why are there so many different CEO ratios?  The mega companies like Apple, GE and Microsoft will naturally have the highest CEO ratios.  Organizations like the AFL-CIO who want to promote the idea of inequality might use only the pay data from the SP100, the top companies in the world.  In a popular magazine article, the writer doesn’t have to share the characteristics of the data set as one would do in an academic paper so it is relatively easy to convince readers of a particular point of view by careful selection of the data.

As companies have grown in size over the past three decades, the number of named executive officers (NEOs) in each of these large companies have grown.  As companies get bigger, the duties of these NEOs begins to approach that of the CEOs of yesteryear. To compare apples to apples, then, we would do better to compare the salaries of today’s NEOs with the CEOs of 1970. The 2013 study found that the NEO ratio was 23 times the average worker pay, much less than the Pay Ratio of 35 in 1970.  If we average the Pay Ratio (CEO) of 52 in this study and the NEO ratio of 23, we get an average of 37, just about the same as in 1970.

Regardless of the data, most of us are either convinced or not convinced that the Pay Ratio has increased dramatically in the past fifty years.  Our convictions are similar to our tastes – white meat or dark – in turkey.  We can only agree to disagree and know deep down in our hearts that we are, of course, right.