It turned out to be a bad idea to look for research papers that explain why big companies routinely pay their top executives well into eight figures. That is, unless you had a few weeks to spend on it.

Studies on what is driving the growth in executive pay packages have been appearing regularly at least since the era of big paydays for CEOs commenced in the 1980s. By now there have to be dozens of them.

If people really bought the simplest explanation, that pay packages of $25 million or more at the top of corporate America come from a well-functioning labor market, then there would be far less for the professors to consider. But "CEOs have been getting paid fairly for their value" is less than fully persuasive.

So that's why there have been studies like the one you may remember from 2016 that found that the best-paid CEOs are more likely to preside over relatively poorly performing companies, at least based on stock performance. This conclusion came from the consulting and investment technology firm MSCI, but at least one university research team had gotten to a similar conclusion.

After finding another study related to the counterproductive effects of performance-based cash bonuses, that was enough material from this vein. Other search terms quickly turned up research from a few years ago that found another distinctly "nonmarket" explanation for higher CEO pay — how the CEOs feel about their social standing in their hometowns.

Just living near other CEOs turned out to be enough to get paid more. That's because after mixing with the other big dogs at the country club and fundraising balls, the CEOs demanded more pay than their buddies were making.

Yet another study concluded that the CEOs of "sin" companies like casino operators were paid a lot more than CEOs in other industries, as compensation for the social stigma from operating slot machines or selling cigarettes.

One paper stayed at the top of the pile because it had a simple and compelling headline, "Executive Pay: What Worked?" The authors of this one had looked back to when CEOs weren't paid all that much, from about 1940 into the 1970s, even as business was booming.

In 1950 average big-company CEO pay was about 20 times that of the average worker's. Now, depending on who's doing the counting, it may be a couple of hundred times the average worker's pay.

As CEO pay surged beginning in the 1980s, efforts to rein it in mostly backfired, the authors found. A great example comes from the early 1990s, when runaway pay for CEOs became a hot political issue. One response was a tax rule that disallowed a deduction for any annual pay that exceeded $1 million. The biggest effect it had was igniting a boom in CEO pay.

So what happened? Well, if a company paid its CEO only $800,000, the CEO likely demanded an immediate 25 percent raise. After all, anything less than $1 million had just been officially deemed too little.

There was also a provision in the tax law that kept the tax deduction if pay was reasonably tied to performance. In practice that rule caused board compensation committees to set up a lot of 6-inch high hurdles for an executive to stumble over and still get paid a lot. And further, the perfectly reasonable-sounding idea of paying for performance really fueled the boom.

The problem was the structure of these new pay packages: a classic heads the CEO wins, tails somebody else loses. Executives didn't want to risk a pay cut, of course, particularly for something like a recession or a bear market in stocks. The boards of directors eased that worry by creating packages that were almost all upside, rewarding the CEO richly if all went well, maybe through stock options.

So why weren't these problems evident in the 1940s and 1950s? The research team was made up of law professors, including at UCLA and Temple University, and they were not very impressed by some of the common explanations for what had kept pay in check.

One idea is that high marginal income tax rates had made big salaries and taxable stock grants not really worth seeking, but of course higher tax rates didn't mean the big bosses were indifferent to pretax income. Nor did the professors find much evidence that corporations' outside directors were better stewards of the corporate checkbook back then. If anything, the boards of directors were more clubby than today's.

They did see good evidence that the job of CEO seemed to change from the 1950s to the 1980s, from administrator on top of a bureaucracy to leader asked to chart a course through a far more volatile and competitive market environment, with increasing globalization and technological change.

But the best explanation is that some of the social norms that mattered a lot in American business after World War II simply disappeared by the end of the 1970s. What had limited executive pay until then was mostly self-restraint, maybe motivated in part by self-interest. That is, getting to be known as greedy and self-interested derailed careers.

A compensation guide from the early 1950s specifically warned that "the executive will damage his own cause if he insists on being given the ultimate dollar to which he believes himself entitled."

The executives had also lived through the Great Depression and World War II, likely serving in the military during the war years and maybe not in a safe staff job. What really mattered to them was team survival and success, not necessarily bigger personal bonuses and stock grants.

So this provocative paper, all about "what worked" to restrain mid-20th century CEO pay, has only one answer, and that's that there's nothing to be learned that realistically can be implemented now.

Asking for CEOs to turn down higher pay, bonuses and stock grants could be tried, maybe. But it seems just as practical to start work on a time machine.

lee.schafer@startribune.com 612-673-4302