CEO Pay Ratios Didnt Cause the Crisis, and Won't Fix Inequality
Section 953 of Dodd Frank requires the Securities and Exchange Commission to write a rule requiring public companies to report the ratio of the CEO's total compensation to median employee's total compensation. The SEC recently issued a final rule on a 3-2 vote. That's too bad, because pay ratios didn't cause the recent crisis and don't address the inequality that matters.
The crisis was about the spread of insolvency risk throughout the financial system. There was a crisis because banks and other financial institutions held highly rated tranches of asset-backed and mortgage-backed securities (ABS and MBS) and especially collateralized debt obligations (CDOs) that went bust.
A recent study by Professors Isil Erel, Taylor Nadauld and Rene Stulz showed that neither the level of compensation, bonuses, nor option-like features of executive pay factored into bank holdings of the very securities that went bust. What did explain those holdings were regulatory arbitrage opportunities created after the Federal Reserve, Federal Deposit Insurance Corporation and Office of the Comptroller of the Currency finalized the "Recourse Rule," which lowered their capital charges from 8 percent to 1.6 percent.
The study also reports evidence of a "securitization byproduct effect," which means that originating banks had incentives to hold on to the highest rated tranches, instead of the equity tranche, to signal that they stood by their product. Here again, this could well be due to the Recourse Rule. NYU Professor Jeffrey Friedman and Wladimir Kraus show in their book "Engineering the Crisis" that the Recourse Rule increased capital charges on equity tranche holdings from 8 percent to 100 percent, i.e., "dollar-for-dollar." No wonder banks didn't hold the equity tranches.
While Section 941 of Dodd Frank (the so-called skin-in-the-game provision) tries to force banks to hold equity risk, it really just undoes part of the Recourse Rule-to stem the risk of future banking crises, consider simplifying and raising capital requirements.
If pay ratios don't help us understand the crisis, you might still argue that pay ratios at least address growing concerns about inequality. But that's not right either, since most discussions of inequality, measuring variation across the population, examine only part of the story-typically wages, income or wealth, ignoring its multidimensional nature. A 2010 issue of the Review of Economic Dynamics, summarized results of a methodology that examines the link from variation across the population in wages and hours worked-hence income-to expenditures and lastly wealth for nine countries, including the United States, through 2006.
In short, since income equals the hourly wage times hours worked, you have to understand variation in wages earned and hours worked across the labor force to understand variation in earnings. For a given tax code, once people know their after-tax and after-transfer, disposable income, they decide how much spend. That means variation in expenditures across the population reflect many factors including peoples' decisions about spending and working, the tax code and transfers and labor market conditions that affect wages. Finally, whatever you don't spend goes toward accumulating assets, from which you can observe variation in wealth across the population.
Unlike other measures of inequality, variation in expenditures across the population ties directly to well-being and vulnerability; it's harder to go a day without food and water than without a paycheck or retirement account contribution. Moreover, if variation in expenditures lies below that for income, and it usually does, then that's a sign that people find ways-including transfers, savings, loans, gifts or even the sharing economy-to share their risks of lost income.
The study for the United States shows that in more recent times, while variation in wages for men and women each rose, variation in hours worked stayed flat for men, and although higher, fell for women. Since earnings depend on wages and hours worked, overall, variation in earnings rose for men, and to a lesser extent for women. Variation in expenditures rose slightly, but was much lower than that for income, and wealth. That means people here usually share their income shock risks.
Fabrizzio Perri, a Monetary Advisor at the Minneapolis Fed, extended some of those results to cover the crisis period. He found that during the crisis, while variation in pre-tax and pre-transfer "market" income rose, disposable income inequality remained flat, and this was primarily due to our tax code, rather than transfers. More importantly, he finds that during the crisis, peoples' vulnerability increased, as declines in income resulted in larger than usual declines in expenditures.
All told, expenditure inequality and vulnerability relate to risk-sharing, not pay ratios. Since pay ratios don't address the inequality that matters, and didn't bring on the crisis, why should investors, let alone the SEC, care?