Tuesday, May 24, 2005

Voluntary Accounts and Solvency

I have been thinking for a while about why the President is gaining such little momentum in his efforts to promote Social Security reform. One reason that is underappreciated, in my opinion, is the disconnect between how reform is being motivated and how it is being designed. Take a look at this website, which is the Treasury Department's guide to Social Security reform and matches up pretty well with what the President has been saying on the road.

The motivation for reform is the projected insolvency of the system--whether you think about it in terms of Trust Fund exhaustion (in 2041), a long-term actuarial deficit (1.92% of taxable payroll), or, as I do, a permanent actuarial deficit (3.5% of taxable payroll). The "Need for Action" page of the Treasury's website characterizes the financial challenge reasonably well and reiterates some of the President's principles for reform. So far, we're doing fine.

But then it finishes with a section titled, "Voluntary Personal Retirement Accounts Are An Important Part Of Comprehensive Reform." This is where it gets confusing. Voluntary accounts, as they are structured here, may in fact be an improvement to the system (more on this below), but they do not necessarily do anything for solvency. Okay, some of you sharp-eyed readers will note, this part of the page doesn't actually say that PRAs do anything for solvency, only that they are important for the reform to be comprehensive. But what in the world is it doing on a page devoted to the "Need for Action?"

This is the disconnect. The personal accounts--which seem to get the lion's share of the coverage and appear to be the only specific component of the reform that the President will discuss--are not structured to address the reason why we are being told reform is necessary--to restore solvency. That isn't helpful, and it is probably confusing a lot of people.

Because the accounts are voluntary, people who choose them have to give up something in return. What they give up is a portion of their benefits relative to those who do not opt for the accounts. As has become standard since the 2001 Commission, future benefits are reduced by the value of all of the diverted payroll taxes, compounded at some low interest rate. In the current discussions, that interest rate is 3% after inflation. This is the projected long-term rate on Treasury bonds held in the Trust Fund (subtract the CPI inflation rate of 2.8% from the nominal interest rate of 5.8%). It would be better specified as the actual Treasury bond rate (so that stocks only had to perform better than bonds, not some absolute level), but that's a minor point in this context, so let's just assume that the two are equivalent for most of what follows.

In order for the the accounts to help with solvency, in the sense of lowering the $11.1 trillion unfunded obligation (over the infinite horizon), the rate at which contributions to PRAs offset the traditional benefit has to exceed the projected Treasury rate credited on the Trust Fund's balance. People opting for a PRA would then allow the Social Security system to earn a higher rate of return on its loans to them than it does on its "line of credit" to the rest of the government.* No one is suggesting that the offset rate will exceed the Treasury bond rate (except indirectly, if the 3% real offset rate is higher than what the Treasury rate will turn out to be), and thus no one is proposing a voluntary account that helps directly to restore solvency.

What about the Administration's explicit claim that voluntary accounts are an "important part of reform," now modified to exclude any connection to solvency? As a start, an economist looking at the financial aspect of the program would say that they do improve efficiency. Currently, there are people who do not save for retirement beyond what they expect from Social Security. As a result, they have little financial net worth that they can invest in equities. Suppose that they would be in this position after the Social Security is reformed to restore solvency. (More would be in it if solvency were restored through higher payroll taxes, and fewer would be in it if solvency were restored through lower future benefits.) The availability of a loan to invest in the stock market that has an interest rate of the Treasury bond (or even a bit beyond) can be appealing to people who currently do not save enough to be in the stock market on their own and are unwilling to borrow at market interest rates to invest in equities.

So the potential improvement from voluntary accounts is that they allow more people to be in the market, thus improving efficiency of risk-bearing according to any neoclassical model of economic decision-making. This may not be a big effect, either for the market or any one individual. (I should investigate this in the Survey of Consumer Finances at some point.) And none of it derives from people like me, who are already saving enough on their own to have already decided how much equity and bonds they want in their portfolios.

To summarize my views:

  1. The Administration is motivating reform correctly, based on the need to restore solvency to the system.
  2. What we have seen of the Administration's plans for reform involves reductions in the growth future benefits that would appear to restore solvency.
  3. The Administration is trying to build a consensus for reform by selling the idea of voluntary personal accounts.
  4. That people cannot figure out how #3 is connected to #1 is one reason why the process appears to be stalled.
  5. Voluntary accounts can improve the efficiency of risk-bearing in the economy, but this is also not connected to #1 and is not a part of the Administration's rhetoric.
  6. The importance of personal accounts to Social Security reform remains their ability to protect current surpluses from being spent (as discussed in yesterday's post) and to allow investments in equities of a sufficient magnitude to meaningfully restore solvency (as discussed in this earlier post).

*For more on this issue, see Peter Orszag's Congressional testimony. Peter also notes that if the accounts are sufficiently large, not all loans will be repaid, further reducing the PRA's contribution to solvency. Subject to the caveat about the legal purpose of the Trust Fund I noted in this post, I am not sympathetic to Peter's other criticisms of PRAs.

Other blogs commenting on this post


PGL said...

I'm truly enjoying your recent posts so forgive a couple of quibbles. On #5, I still hold to the position of Barro and Becker that privatization will not change overall asset demands and hence will have no net impact on overall expected return and risk. And I've seen very little - beyond the premise that a few folks might be borrowing constrained - to suggest otherwise. My other quibble is more politics than economics - the Bush crowd has been trying to sell us on some massive free lunch that somehow the alleged (and perhaps nonexistent) gains from #5 are so massive - that no one will lose from the benefit cuts. Of course, this suggestion is simply put - a bald faced lie.

PGL said...

Max gives you a plug and links to the WaPo article that mentioned you. WaPo seems to think you were the chief economist for the Bush43 CEA as I read the article. Huh? Bush41 - yes? Bush43 - no? What am I missing here?