Questions for conservatives on Social Security

Stephen Bainbridge raises three good questions on the Social Security debate. I thought I’d take a stab at answering them. I don’t know the exact figures that go into the actuarial analysis of Social Security, so I use illustrative examples below.

1. Would we achieve significant actuarial improvements in the health of the Social Security system by (a) changing the method by which the benefit is calculated from being based on wages to one based on prices (see Tyler Cowen's post for details) and (b) increasing the retirement age? Social security was designed for an era in which most folks would live to receive benefits for months rather than years. Why not deal with that problem directly? (Glenn Reynolds has a solution that goes somewhat in the other direction.)

Both of these would improve the health of Social Security, but both of these are benefit cuts. (The term “cut” is used colloquially, to indicate that the benefit would be less than that calculated under the current system.) There are only two reasonable approaches to saving Social Security, raising taxes or cutting benefits. (Unreasonable approaches would be to expect an increase in the working population supporting the retired population or to expect a decrease in life expectancy for retired individuals.) However, if returns are higher when invested in a private account, then we can have a similar nominal benefit for less cost.

Benefit cuts are more palatable when they don’t result in a loss of money. Let’s say $6,000 of annual Social Security taxes over a working life funds a $2,000 monthly Social Security benefit at retirement, but would fund a $2,600 monthly annuity if invested independently. Let’s say current Social Security taxes will only fund a benefit 70% of the current size in the future. We could reduce benefits to $1,400 a month, bringing the system into actuarial balance, or do the actuarial equivalent in changing the index or raising the retirement age. Now, if you take half of those Social Security taxes and invest them in the deferred annuity described above, you’d get half of the reduced Social Security benefit ($700), and half of the annuity benefit ($1300), resulting in the same monthly benefit.

2. If we can achieve significant savings and ensure the health of the system with the changes mentioned in # 1, is there a non-ideological reason for introducing private accounts? Even proponents of private accounts concede that the transition costs will require trillions of dollars of government borrowing. Do we conservatives really want revenge on FDR and the New Deal at that price? Personally, speaking as a small government fiscal conservative kind of guy, I'd give up personal accounts if any money thereby saved was spent on deficit reduction or, better yet, an income tax rate cut.

There are two reasons I can think of to introduce private accounts. One is the fact that you get more benefits from a private account if you don’t live to see retirement. I am unmarried and childless, so if I die now, my whole Social Security benefit is the $255 death benefit, less than what I pay into Social Security each month. My retirement savings, on the other hand, can be passed to my heirs. Social Security benefits can be at their current level thanks in part to the tontine nature of the program, where each death increases the benefits to the survivors.

The second reason is protection against future shortfalls. If in 2042 private accounts means both taxes collected and benefits paid are half the size, then the projected shortfall is half the size. Let’s say the actuarial estimates implicit in current projections are that the 67 year olds in 2042 will live to 84 (female) and 82 (male) on average. However, what if scientific discovery and new medical treatments mean that people live to 89 and 87 instead? Benefits would have to be cut a lot more than currently estimated to keep the system in balance. (In the case above, people will also realize a smaller monthly benefit from their annuity, due to increased life expectancy.) The smaller the future deficit is, the smaller the impact to the government will be if current estimates are wrong.

3. Why aren't conservatives talking about other entitlement programs, such as Medicare, which reportedly is scheduled to go broke long before Social Security does?

We should be. But I suspect Medicare would be even harder, politically, to fix. It may be that politicians will only be able to fix Medicare if it can first be demonstrated that Social Security can be fixed without the political world coming to an end.

Medicare is also more complicated, as the future shortfall depends on both the factors affecting Social Security (life expectancy and size of the workforce supporting retirees) and unknown future medical costs.