Perhaps the first useful lesson of economics is to think about costs in terms of opportunity. The cost of some action is the cost of the most valuable opportunity you have forgone by taking that action as opposed to some other. When you shell out $100 to buy some novels or porn or whatnot, the cost to you is not the loss of the piece of paper which no longer burdens your wallet. The cost is all the other stuff you might have purchased with that $100 that you’ve just blown on your Thomas Pynchon habit. If, by some mischance, there is a general inflation, so that both your income and most prices rise, but for some reason the price of Pynchon remains unchanged, even though it will cost the same $100, your next hit of literacy and pretension will feel much cheaper than the first, because you’ll forgo a lot less food and rent for the purchase.

A lot of left-ish proposals these days, including high marginal tax rates at high incomes and bans on share buybacks, are about increasing the cost to firms of making payouts to rich shareholders, thereby reducing the opportunity cost of other uses of the money. Some of these proposals I think are solid. Some I think half-baked. [1] But the basic logic behind the proposals is missed I think by a lot of smart commentators.

Firm managers and shareholders face choices about what to do with each $1 of revenue. Some choices are easy. The first zillion dollars go to cover liabilities they incur over the course of operations, paying suppliers, employees, rent or interest to capital providers. But then they face discretionary choices. Should they invest a dollar in expansion or new business lines? Should they increase the cushion in their payrolls above the absolute minimum their labor force might accept, paying an “efficiency wage” in the lingo for a happier, more devoted, potentially more productive workforce? Should they “pay” that dollar by not receiving it at all, by reducing prices, purchasing the goodwill of customers and potentially a bit of market share? Should that dollar be paid into some low-risk “cash equivalent”, to purchase extra insurance against hard times or just put off the decision? Or should they make payouts to shareholders, and let shareholders decide the best use of the dollar?

Under the model of capitalism that Milton Friedman famously championed and that became the “shareholder value” revolution, the presumption was that the best thing a firm can do is maximize shareholder financial welfare. Anything else, anything that might benefit employees, customers, or any other stakeholder was deemed an “agency cost”, an inefficiency. And you can make a strong theoretical case for this: If you believe that capital markets are high quality information systems that govern economic production, that shareholders allocate resources to their best possible uses to the benefit of society as a whole, then the ideal policy would be to return every dollar of unencumbered revenue to shareholders, and let shareholders choose to reinvest (or not) in potential uses at new or existing firms. That may be impractical, but what became the conventional standard was that managers should retain in the firm only what shareholders would have reinvested themselves, and disgorge the rest to find some more efficient use elsewhere.

However, if you do not believe that shareholder interests and the public’s interest in aggregate welfare are well aligned, this case breaks down. Then you might prefer firm managers to do things other than make the level of payouts that shareholders would prefer. You might prefer that they accede more easily to labor demands, or that they lower prices to customers, or that they invest more in speculative research and development, or that they delever their balance sheets and even build up cash cushions to reduce the probability of a disruptive insolvency with its attendant unemployment and the possibility the state will need to bail out pensions or depositors.

It is reported ad nauseam, when people point out that the US did very well under the high top marginal tax rates that prevailed from World War II through the 1980s, that those high rates were rarely paid. People bring this up as though it was some kind of policy failure. No, it was then and would be again quite the point of the policy. The purpose of very high tax rates at very high incomes is not to generate revenue. It is to make costly the practice of making payments to people who are already very rich relative to other things the payers could do with their money, and so reduce the opportunity cost of doing other things. If paying $1 to shareholders costs $1 of potential goodwill derived from better work conditions, maybe shareholders take the $1. If paying $1 (after tax) to shareholders costs $10 in worker goodwill, maybe shareholders let their lackeys indulge the help. Maybe not! Firms have all kinds of choices, and shareholders may try to have firms smuggle wealth to them through Luxembourg or ZCash or whatever. But that is costly too, and can be made very costly with determined policy and aggressive enforcement. All controls leak, but often they are effective anyway, because their purpose is not to keep anybody dry but to alter the relative costs of things.

Very high top tax rates are a means of encouraging “predistribution” rather than the tax part of tax-and-transfer redistribution. Their purpose, their very point, is to create those “agency costs” that economists from the 1970s until now have derided and demanded be ruthlessly excised from corporate practice. But every “agency cost” to shareholders is income to someone else, whether that takes the form of luxury offices and stupid jet travel for firm managers or better work conditions at higher pay for more employees. The ideologically tendentious presumption of the economics profession post-1970s has been that agency costs yield no real benefits, that they look much more like luxury offices for the C-suite than predictable schedules for service workers. But that was always just presumption, and historical experience does not support it. It is, I will admit, not a slam dunk case, it is only suggestive, that the ruthless efficiencies of contemporary labor markets and the shattering of union power happened just after we, in relative-to-prior-period terms, dramatically subsidized payouts to shareholders over other uses of funds. But it is suggestive. And it is plausible that “Treaties of Detroit” and Bell Labses, that corporate practices generally which favor workers, customers, and other stakeholders, are easier for companies to “afford” when shareholders have to give up less to purchase them. Which is precisely the effect, in the most basic economic terms, of taxing payouts to shareholders heavily. [2]

You can believe, if you like, that in fact the neoliberal case for shareholder primacy is correct, that shareholders allocate capital well on behalf of society as a whole and we should celebrate payouts as a way of directing resources to their best uses. That was once my view, and it is perfectly coherent, at least in theory, though I think experience has refuted it. You can argue, if you like, that firms would do worse things with the money than they do now, if they didn’t make payouts to shareholders. But it’s hard to believe, I think, that dramatically increasing the opportunity cost of payouts to shareholders will not result in some substantial reduction of payouts in favor of other uses of funds. Maybe it’s time to see (and also to shape, via other policy) what those uses might be.

---

This entry was posted on Wednesday, February 6th, 2019 at 4:23 pm PST.
You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.