Friday, January 9, 2009

Investor’s Business Daily Misunderstands Social Security reform

Investor's Business Daily – a paper I otherwise like – has an editorial on Social Security reform that reflects a general misunderstand on the right about the merits of personal accounts. After running through the traditional menu of changes to make Social Security solvent – increase taxes, lower benefits, raise the retirement age, etc. – IBD's editors say:

There is another way, however, at least with Social Security…. That way is privatization. Obama's party lashes out when even partial privatization is brought up. But taking retirement out of government's hands works. It has worked in Chile. It has worked in Britain.

It's worked, as well, in Texas, where in 1981, Galveston County employees voted 78% to 22% to opt out of the Social Security system and put their trust in a private retirement plan. Brazoria and Matagorda counties followed Galveston out of Social Security before Congress outlawed in 1983 the option of leaving Social Security.

Now more than 5,000 Texas participants are in a private plan where the rate of return far exceeds that of Social Security's average of 1.25%.

The Institute of Policy Innovation, a Texas think tank, reports that a low-income worker making $17,124 a year and retiring at 65 will get $782 per month from Social Security while someone with an identical salary retiring at the same age would receive $1,285 a month from the plan the Texas county workers invest in. A retiree who earned roughly $51,000 a year will get $1,540 monthly from Social Security, $3,846 from the Texas plan.

I'll stick to the Galveston example, since it's the cleanest illustration of why straight comparisons of Social Security to a personal accounts plan can be misleading.

For background, until the early 1980s local municipalities were allowed to opt out of Social Security and set up their own plans, much as many state and local workers had never been in Social Security in the first place. While most municipalities set up defined benefit plans, Galveston, Texas was an early adherent of defined contribution plans, aka, personal accounts. Under this plan, workers invest a portion of their salaries in accounts holding market investments, and draw down their accounts to provide income in retirement. And, it is true that in general these accounts pay a higher return and higher benefits than Social Security. So what's not to like?

The problem is that what can be done for a few people, like workers in Galveston, can't be done for everyone. Here's why: Social Security is a transfer system, meaning that today's taxes immediately pay for today's benefits. If people decide to invest their taxes in personal accounts, there will be a continuing "transition cost" to maintain benefits for current retirees.

Now, if a very small group of workers is allowed to opt out, the rest of the system can effectively carry their transition costs without anyone noticing. That's what happened with Galveston: the rest of us are paying to finance Social Security benefits for retiree in Galveston, which allows workers in Galveston to invest their payroll taxes privately and earn higher returns. For a small group, it's not a problem, even if it raises questions of fairness.

But if everyone did this the transition costs couldn't be swept under the rug. And those transition costs are very large: Social Security currently owes around $17 trillion for benefits that have been earned but not yet paid out. If we levied a surtax to finance those benefits, the total rate of return earned under a private account plan would be almost exactly the same as under Social Security.

The argument that personal accounts would pay a higher rate of return – that they would "make Social Security a better deal" – is intuitively attractive but ultimately counterproductive because it encourages people to avoid the hard choices involved with Social Security reform. There are a number of good reasons to support personal accounts, but a higher rate of return isn't one of them.

2 comments:

Jim Glass said...

For a small group, it's not a problem, even if it raises questions of fairness. But if everyone did this the transition costs couldn't be swept under the rug.

And those transition costs are very large: Social Security currently owes around $17 trillion for benefits that have been earned but not yet paid out. If we levied a surtax to finance those benefits, the total rate of return earned under a private account plan would be almost exactly the same as under Social Security.

The argument that personal accounts would pay a higher rate of return – that they would "make Social Security a better deal" – is intuitively attractive but ultimately counterproductive because it encourages people to avoid the hard choices involved with Social Security reform.

~~~
Well, Milton Friedman argued that the $17 trillion "backward transfer" should be considered a cost society chose to incur, for better or worse, and thus should be paid for from general revenue (income tax). As was World War II, the moon program, etc.

In that case workers could indeed get a much better return from their SS contributions, as they would be paying only for their own retirements, not those of their predecessors, and they could have private accounts or whatever. Their benefits would be funded.

That's not a free lunch because the $17 trillion still has to get paid by somebody. But coming from general revenue it would come from a much larger tax base (good), at a lower tax rate (good), progressively collected generally from higher-income taxpayers (good), and would leave lower-income workers getting much higher returns from their payroll tax-funded retirement contributions (good), which could then be smaller for comparable benefits (good).

Friedman said that the near unanimous popular idea that the $17 trillion has to come from payroll tax has some sort of mesmerizing political power (I forget his exact words) that stymies reform, but that it's illogical.

He also denied that the there is any "transition cost" in privatization plans, because the $17 trillion cost is in the status quo and must be funded by it -- privatization ideas only make this cost visible, while defenders of the status quo obscure it and kick it down the road. But something can't be a transition cost if it exists even if there is no transition.

Yes, moving the $17 trillion onto general revenue would be a big, hard choice, and people generally aren't thinking at all in that direction yet. Though Kotlikoff has proposed dropping that cost on a new national sales tax or VAT, I believe.

Andrew G. Biggs said...

Jim,
You're right, but I guess what I'd say it that what you describe is a bit different from what IBD is talking about. IDB describes basically a first-order efficiency change -- you were getting a 1.5% rate of return on your Social Security taxes, but with personal accounts you'll be getting a market rate of return (e.g., 3% with almost no risk, double that by investing in stocks). Compounded over a full working lifetime, that produces a HUGE difference in benefits at retirement.
What you've described is paying the transition costs with higher income taxes. Now, if you factor the higher income taxes into the rate of return equation, the return on a personal accounts based system is exactly the same as under current law. It looks higher because part of the costs are taken off the books, but in reality it's the same. You point out that there may be some efficiency gains because income taxes have a broader base than payroll taxes, which is true. But there may also be some efficiency losses, since top marginal income tax rates are three times higher than top payroll taxes rate (and in theory, the deadweight loss from a tax rises with the square of the tax rate). In any case, the efficiency gains will be second order.

I'm not arguing that personal accounts are a bad idea, and Kotlikoff's approach of financing the transition with a VAT tax may be the best possible approach to reform. The only point argued here was that the straight rate of return gains that IBD argues for are largely illusory.

Andrew