View From a Height

Public Savings Accounts

Nathan Newman of New York University wants to know why his idea of having the government invest Social Security money in the stock market is such a bad idea. I don't think any of the arguments I present here will persuade him, but an honest effort demands a thoughtful response.

Efficiency...while current administrative costs for social security are just $15 per worker involved, private accounts could cost as much as $50 per year. Such a system would knock out up to thirty percent of the touted accumulated returns on investment.

Could knock out up to 30%, on average knocks out much, much less than that. In the first instance, I think he overstates the cost. Most of the reasons for increased costs that he cites are already either built in or would remain irrelevane. Even give an extra $35 a year, though, if an employee gets a 12% contribution on $30,000, that's $3600 per year. Thirty-five dollars is almost exactly one percentage point interest. If current after-inflation returns are close to zero, and they are, the extra costs come to 14 of the added return. Mutual fund companies with hundreds of thousands of customers routinely make do with much less than that. Even to give the government that much is to confirm conservative suspicions about giving them the cash in the first place.

Government investments could be directed to investments that strengthen wage growth: The idea that the government might actually use all this capital in the social security trust fund in a pro-active manner gives rightwing economists hives, but it makes a lot of sense. James Glassman from the American Enterprise Institute attacked Clinton's proposal to have the government invest in stocks with the S scare word:

I do not want to sound overly dramatic, but, by definition, the plan is a step toward the dictionary definition of socialism: government ownership of the means of production.

Yet even in Glassman's hyperbolic venting, he admits that the government would own only 3.7 percent of the total value of the shares on the stock market, hardly enough to implement five-year plans. The government's existing regulatory powers over business gives it far greater potential control of business operations than this tiny minority share of stock would. Even with this reality, many promoters of government investing in the stock market call for elaborate controls to require blind investments in stock indexes.

This is just nonsense. Anyone with experience in the financial markets knows that a 3.7% share in a company is gigantic. It guarantees broad say in policy decisions, great leverage in proxy fights, and frequently means a seat on the board. Outside of confiscatory taxation, I can't think of a better example of galloping socialism than to start giving the government board seats.

Add to that the fact that this money wouldn't be invested evenly across the entire universe of funds, and you realize that for many companies, government ownership would exceed 5%. At those levels, it becomes almost impossible to punish bad management by just dumping shares without crashing the stock. This has exactly been the justification that CalPers has used to start exercising extra-legal regulatory powers over a number of its biggest holdings.

To compound the problem, which Newman's suggested use of this power may or may not be benign, there's nothing inherent in the proposal that would insulate the government from making political decisions with the money. In the long run, I don't believe such safeguards are possible.

The only way to manage that problem is to, in effect, forbid the government from trading by forcing it to hold some large, fixed portfolio. This doesn't replace distributed decision-making with centralized decision-making, but with no decision-making. Those shares have to just sit there, free-riding on the market. You'll get a better return than T-bills, but I'm not sure you'll be contributing in any meaningful way to economic growth.

I don't want to get into the whole colossal off-shoring issue, which is where Newman takes this argument. However, it's worth noting that investing in inefficient domestic companies virtually guarantees the investors - you - lower returns. In effect, you're subsidizing inefficiency, a national version of making everyone pay more for groceries because you don't like WalMart.

Newman also argues that collective investment spreads the risk more. That's true. But it also highlights the difference in incentive that inevitably makes government returns less attractive. The government's job may be to spread risk, but an individual investor will tend to be motivated by a desire to maximize return.

Finally, there's this:

Privatizers want to talk about "returns on investment" from social security, while ignoring the fact that social insurance systems are not like regular investments. Individuals start out with a fixed amount of capital and the only returns that matter are from annual investment returns. But governments investing their capital have two sources of economic returns from those investments -- conventional returns from equity or bond markets PLUS the taxes derived from the domestic economic growth fueled by those government investments.

This is a bait-and-switch. In a fair comparison, returns are returns, no matter who's doing the investing. And domestic economic growth only happens when there are returns to investors. Governments don't get any more out of that than anyone else.