arlier this evening, while we were all in the Chat Room, Jon asked me what I thought would happen to the stock market if Social Security was privatized. That took me aback, actually, because, for all the debate I've seen about it, no one seems to have really looked hard at what the effects would be on the markets. A lot would depend on the actual amount that would be transferred into the markets, and how much would go into stocks, and how much into bonds. But we can certainly play with some numbers and come up with some tentative preliminary conclusions.

In 2003, out of a total payroll of $4.3 trillion, The Social Security system collected total payroll tax revenues of $534 billion. Additionally, since benefits are subject to income taxation, the system collected an additional $13 billion in tax revenues, for a total of $546 billion.  Under the president's current plan, workers would be allowed to put up to 4 percentage points of their payroll taxes, i.e., about 35% of their actual payroll tax payment, into private accounts that would invest in the financial markets.

So based on the figures above, and without trying to do a dynamic analysis about what revenues will be in future years, let's say that there will be an additional ballpark figure of $200 billion per year being diverted into the financial markets.  At the moment, the total stock of US financial assets is around $33.4 trillion.  Given that disparity between total assets, and the amount of new investment, I think there's adequate liquidity  in the market to absorb the new investment without too much stress.  We're only talking about a 0.6% increase per year.

Assuming that the stock of US assets grows at a faster rate than 0.6% a year, the markets can easily absorb that amount of new investment, without an immediate effect on prices or returns. But the possible effects aren't limited to the immediate future.  Over time, the investment of Social Security funds would become an increasingly larger part of the private markets.   By 2050, the Social Security system, that is to say the government, would not only be the largest investor in the equity markets, but would hold about 25% of all equities. That raises a crop of problems that will have to be dealt with.

ne of the central talking points in favor of privatization is that returns from private equity investments have been historically higher that anything Social Security's current structure can provide.  There are two problems with that talking point however. 

The first problem lies in the use of the term "historically".  Past returns, investment advisers are required by law to tell us, are no guarantee of future profits.  In the case of Social Security Privatization, this might very well be more true than privatization advocates would have us believe.  As Fed Chairman Alan Greenspan has noted, if we are going to transfer Social Security funds from government treasury bonds into the equity markets, then we have to induce private investors to make a net transfer of their equity investments into treasuries. 

Two implications arise from this (although this is largely going to be guess work, because apparently, even Fed researchers aren't sure about the following predictions).  The first implication is that treasury yields will need to rise in order to induce the required investment shift into treasuries by private investors.  Second, equity prices will have to rise, meaning that the expected future return on equity will decline.  If that's true, then the future return on equity investments will be lower than the historical return.  And, as Mr. Greenspan points out, on a risk-adjusted basis, the expected return on equity investments from Social Security funds would certainly be less than current predictions estimate, although they would probably be higher than the system's current returns.

Next, privatization doesn't do anything to solve the problem of the current liabilities of the Social Security system.  All the transfer of Social Security moneys from treasuries to equities will accomplish is an asset reallocation.  It won't change the total stock of assets, and it wont change the net return on investment from all assets.  It may change the return for some individuals, raising the returns for Social Security investors while lowering the returns for private investors, but overall, it's a wash.

To meet the current retirement claims outstanding, the only thing that will help us out with that is either a substantial increase in the national savings rate, which implies congruent decrease in consumption, or by substantial increases in productivity.  Of those two options, increasing the savings rate is the more painful choice.  It means a lower standard of living as people pare back on their consumption now, in return for secure retirement.  The most pleasant option is to increase productivity, allowing us to get more economic output for the same amount of money.  Assuming, of course, productivity can be increased enough to even make up for the current shortfall. 

But, one of the key reasons for America's productivity increases has been because we have very efficient financial markets.  They do what free markets are supposed to do, which is to transfer resources, such as capital, to its highest valued uses.  Large-scale investments of Social Security funds might put that in danger.

As Robert Samuelson has put it:

 The moneys flowing into personal accounts would not be invested according to the "free market." Individuals wouldn't have the freedom to invest in Microsoft, General Electric or eBay. Instead, the moneys would be invested according to rules made by Congress, influenced by politics. There would be unrelenting pressures from interest groups, "experts" and public opinion. Some types of investing—or some types of companies—would be deemed better than others.

Once the government has a say in where investment dollars go, then you begin to run into problems.  This has certainly been the case at the state level, where it has been impossible to prevent state legislatures from mandating all sorts of investment rules that have reduced returns, for purely political reasons.  As Alan Greenspan has told Congress:

I doubt that it is possible to secure and sustain institutional arrangements that would insulate, over the long run, the trust funds from political pressures. These pressures, whether direct or indirect, could result in suboptimal performance by our capital markets, diminished economic efficiency, and lower overall standards of living than would be achieved otherwise.

The experience of public pension funds seems to bear this out...For example, it has been shown that state pension plans that are required to direct a portion of their investments in-state and those that make "economically targeted investments" experience lower returns as a result.  Similarly, there is evidence suggesting that, the greater the proportion of trustees who are political appointees, the lower the rate of return.

In an environment where the government is an institutional investor that controls a quarter of all equity capital, it's almost impossible to construct a scenario in which an investment plan run by the government will not devolve into an exercise in de facto industrial policy.  The government will make decisions about what companies should be winners or losers.  Even worse, political pressure to put a stop to things like  "Benedict Arnold companies shipping American jobs overseas" will eventually be too attractive to ignore, if past experience is any guide.

Companies will also become more politically risk averse.  The specter of upsetting a powerful institutional investor who holds 25% of your market cap is enough to give most CEOs nightmares.

The counter argument to that has been put forth by former Treasury Secretary Robert Rubin.

[T]he scenario Greenspan fears can be avoided by erecting barriers between Congress and the management of the trust funds. Those barriers would include creating an independent board, much like the Federal Reserve itself, to oversee the trust funds. Its members would be appointed by the president and confirmed by the Senate, serving staggered 14-year terms and shielded from dismissal from office for political reasons. In addition, the power of the board could be limited to selecting fund managers who would be required to make only passive investments in securities that represent broad market averages—so-called “index mutual funds.”

Well, that sounds good, but that raises a problem in an of itself, which I will address below.

Still, if you're concerned about the government doing all that investing, then the answer is to create a program that takes the investment choice away from the government, and to allow individual investors to allocate their money as they wish.

But that has problems as well.  Individual investors just aren't very smart.  It may not be libertarian-PC to say it, but the more private investors are allowed to make allocation decisions, the worse they do.  That's not an opinion, that's a fact. Investing isn't something you can dabble in part time.  It takes a significant amount of knowledge to churn your accounts, to buy high and sell low.  In aggregate, individual investors do best when they simply buy an S&P Index fund, keep putting money in it, and otherwise forget they have it.

Moreover, stock market investing does have an element of risk.  Unless you are going to posit that the higher rate of return from equities is as Brad Delong puts it, "a market failure that private accounts can profit from, rather than merely compensation for risk," then you have to address what, if anything, you're going to do to compensate retirees who suffer a 40% drop in equity prices like occurred in 2000-2001, the year before they retire.  The risk element of equity investment is there, and has to be dealt with. Professor DeLong points out that, as long as the risk is there, the government is much better situated to mitigate that risk than the individual investor is.

So, even a system of private accounts must have some very strict rules, such as those proposed by Professor DeLong, which would stipulate that:

• Private accounts are set up so that they are not eaten away by high administrative costs.
• Private accounts are set up so that they are not decimated by improperly balanced and diversified portfolios.
• Private accounts are set up so that they cannot be pledged or emptied by imprudent and impatient beneficiaries.

That means very little freedom of choice, and very conservative asset allocation rules, which gives rise to problems of its own.

Robert Samuelson, again:

Personal accounts would be a strange hybrid: part "private" investment, part public entitlement. This is a hard straddle. There's an unavoidable dilemma: making personal accounts safer for individuals may make the stock market less useful—less dynamic—for society. The conflict has already surfaced. One criticism of personal accounts is that they might subject beneficiaries to huge losses, because stocks fluctuate erratically. The administration counters that it would allow accounts to be invested only in "index funds"—for example, funds representing the Standard & Poor's 500 stocks. The idea is to minimize the risk of big losses on individual or speculative stocks. Sounds sensible. But it would bias the market in favor of existing companies, industries and technologies. It would discriminate against the new, exciting and different.

The problem here is exactly the same problem as the one that comes from Sec. Rubin's suggestion above.

Investment rules that seek to reduce risk by rigorously mandating conservative allocations prevent capital from flowing to its highest-valued uses. In other words, we're back to a productivity problem again.  New companies and new technologies would have to compete for a much smaller share of available capital, because so much investment capital was tied up in Social Security investments, with constrictive investment rules.  A shortage of entrepreneurial capital might very well put a crimp in technological progress upon which so much of productivity increases depend.

lmost any way you approach it, any government-mandated program of equity investments raises as many questions as it does answers.  It might increase the returns from Social Security, but only at the expense of lowering them for private investors.  Putting the government in charge of investments runs the risk of creating an industrial policy, while letting investors make choices exposes them to risk that we might have to make up for at extra public expense.  And the solution to either of those problems runs the risk of lowering productivity, and hence living standards.

If we really want to privatize Social Security then the best thing to do would be to...well...privatize it.  Let investors keep their money, make their own decisions, and accept that those decisions will have risks.

We won't do it though.  We want our money, and we want high returns, and we want it risk free  We want, in short, a free lunch.

I doubt we'll get it. 

UPDATE: Jon pointed pointed out that the amount of money I originally wrote would be going into private markets was wrong.  The correct amount would be $200 billion, not $20 billion.  I can only blame my post-midnight math for dropping the zero.