An unsettling increase in economic inequality over the past few decades has become widely accepted as a fact. What's caused it, and where things go from here, are still hotly debated.

There seem to be three basic theories about what's going on:

One is that changes in the modern economy have concentrated income and wealth. Transformative technologies and the globalization of commerce have placed vast markets at the disposal of "superstars" in management, entertainment, finance, science, etc., leaving the rest behind.

Another idea is that ever greater inequality is the inexorable, historic tendency of capitalism, except when interrupted by great wars and revolutions or by bold government policies to redistribute wealth.

A third notion is that public policies were deliberately altered over the past three decades or so (whether malignantly or misguidedly) diverting wealth from the middle class to wealthy elites.

Many analyses incorporate elements from each of these stories.

But another theory that gives off the messy aroma of real life holds that inequality has worsened in part as the mortifying, accidental byproduct of efforts to get tough with the rich and crack down on their excesses.

On his always stimulating Conversable Economist blog (conversableeconomist.blogspot.com), Macalester College economist Timothy Taylor recently described his version of this saga — "which will forever exemplify the Law of Unintended Consequences."

Taylor points to federal legislation back in 1993 — President Bill Clinton's first budget measure. It was most controversial at the time for a deficit-cutting income tax hike on the rich. But the law also included a provision designed to put a lid on soaring executive pay.

It "plac[ed] a $1 million cap on salaries for … top corporate executives," Taylor writes. "However," he adds, "the cap was only about salaries, not about pay that was in some way linked to performance."

As intended, this tax incentive set off an unprecedented boom in "performance pay" for executives — chiefly stock options through which corporate chiefs cash in only when their companies' stock prices rise. The idea is that this kind of results-based compensation is more deserved than gigantic salaries.

But Taylor explained in an interview that few anticipated how this "large-scale shift to paying top executives with stock options" would coincide with a remarkable bull market in stocks — the dot-com boom of the late 1990s. "The stock market more or less tripled in value during the five years from late 1994 to late 1999," Taylor writes, "and so those who had stock options did very well indeed."

Taylor doesn't doubt that technological change, globalization and other forces also pushed executive pay higher. But he argues that the twin gold rushes of the 1990s — the stock options boom and the dot.com boom — turned the Clinton-era "attempt to pass a law that would limit CEO pay" into an inadvertent alchemy that "reset expectations about what was 'reasonable' to pay top employees …"

It's been widely noted that in America stratospheric pay for work, more than inherited wealth, has fueled growing inequality. Taylor shows that the timing of the 1990s stock option surge coincides suggestively with measures of the general rise in American inequality. He shows, for example, that the share of total income collected by the top 0.01 percent of families varied between 1 and 1.5 percent for a half century from the 1930s to the 1980s. It rose to between 2 and 3 percent in the '80s, then soared above 5 percent by the end of the '90s. It's bounced around that level ever since.

I asked Taylor whether the initial jump in inequality in the 1980s might confirm the argument of many that supersize incomes soared with the end of confiscatory top marginal income tax rates in the Reagan era (top rates were still 70 percent in 1981 and are just over half that now). No one, it's said, had much reason to seek lofty cash income when the government grabbed so much of it.

Taylor agrees that was a factor, and notes yet another. Saluting Taylor's theory from his perch at Forbes, British economist Tim Worstall adds the idea that our era's rise in executive compensation may be more apparent than real — in part, "we used to … underestimate executive compensation; now we measure it properly."

It seems that while cutting top tax rates, 1980s policymakers also imposed taxes on many fringe benefits that had long been exempt from taxation. In earlier decades, everything from club memberships to car allowances to dozens of free or discounted services had been lavishly provided to executives and not legally counted as "compensation." Some bosses, suffering a "corporate royalty complex," had trouble distinguishing what was the company's from what was theirs.

That changed to a degree in 1984, Worstall says, when fringe benefits were made taxable income. But one big result was simply that execs started asking for more cash instead of fringe goodies, since previously tax-free perks were tax-free no longer.

Taylor draws several lessons from these speculations. First, that "what happens at the top is strongly influenced by laws, taxes, rules about how you can pay people." And that it's easy for rules to misfire, or to be exploited by smart people with huge incentives to "dance around with them." Writing better rules would be "harder than it looks."

Second, if, as he suspects, our era's surge in inequality results in significant part from such policy missteps, Taylor says "it argues against the claim that all this higher pay is just a matter of [corporate leaders] being 'worth more' in a globalized economy." Yet it also argues against the fear that inequality must inevitably continue to increase, concentrating ever more wealth and continually deepening social divisions. It looks more like a onetime mishap.

But if he's less alarmist than many about today's swollen inequality, Taylor allows that it justifies reasonable redistributionist policies, and he's eager to see the media and disclosure rules expose executive compensation to more public scrutiny.

D.J. Tice is at Doug.Tice@startribune.com.