In a strong libertarian state there would be no place for Social Security (Old Age payments). People would provide for their retirement completely out of their own savings; indigent elders would have to rely on their family or charity.
So many will see Social Security as a good argument against the strong libertarian position. Social Security seems to be a success: it is extremely popular, and poverty among the elderly has indeed plummeted.
In my note on strategy, I say that even when liberals conceed that a particular government program is deeply flawed, they can respond that we should mend and not end it. But in this case, the burden on the libertarian is even greater, since there is no consensus that Social Security is deeply flawed.
So what is the basic argument against Social Security?
The basic argument is that Social Security serves no positive purpose: true, it gives people money when they're old; but in order to pay for itself it has to take money from them when they're young. If Social Security didn't exist, working people could save the money they currently pay to finance SS, increasing their retirement income. So at first glance it seems that Social Security has no effect at all on retirement income.
So the basic argument against Social Security is that it's useless.
For the purposes of this apology, however, "useless" is not good enough. My strategy is to show that government programs make us worse off, not that they make no difference.
The above argument ignores the REDISTRIBUTIONIST aspect of Social Security. Perhaps the purpose of Social Security is to transfer money from the rich to the poor.
The higher your average working-life income, the higher your eventual Social Security payments. However, while SS taxes increase proportionately with income, SS payments increase less than proportionately, which suggests that SS is progressive: it redistributes money from the poor to the rich.
A counterargument is that poor people begin paying SS taxes earlier than rich people (since they begin working / end their education earlier); furthermore, poor people stop receiving SS payments earlier than rich people (since they die earlier than rich people): these two effects suggest that Social Security is regressive. A recent GAO study (PDF), however, indicates that the progressive aspects of SS outweigh the regressive, and hence that SS really is progressive. (But less so, of course, than would appear to someone thinking only of the progressive effect and ignoring the regressive effects.)
Suppose for the moment that the first effect is dominant, and that SS is progressive. In that case SS is not "useless"--it has a real effect on people's incomes, one which many people would think is desirable.
But if SS is no longer useless, it's also no longer harmless: since your net benefit from SS depends on your behavior (income), SS distorts incentives, producing a deadweight cost.
As a matter of fact, SS does appear to be redistributionist, but not so much in the manner described above. Rather than redistributing from rich to poor, SS redistributes from later generations to earlier generations: early participants in SS received greater returns than they could have by investing their SS taxes; later participants will receive lower returns.
In this case SS has not been "useless", and the above "deadweight" cost does not apply, since people don't choose which generation they're born in. Whether SS has brought a positive benefit depends on whether you like this particular redistribution (which, it could be argued, is progressive, since earlier generations are poorer than later generations).
A caveat: It's possible that this seeming redistribution is less significant than it seems. Earlier generations typically leave bequests to later generations. Therefore, earlier generations could offset the intergenerational redistribution aspect of SS by increasing their bequests. Indeed, it's natural to suppose that the reduction in bequests COMPLETELY offsets the SS redistribution: If, absent Social Security, you'd want to leave $50,000 to your son; and if Social Security raises your income and lowers your son's income by $10,000, it's reasonable to suppose you'll leave a bequest of $60,000 under SS.
A counter argument is that some people do leave 0 bequests, and would be glad to pocket the gains from SS redistribution at the expense of their heirs.
We could, more tenuously, make a similar argument against the previous redistributionist claim, that SS has an intragenerational redistributionist effect, from rich to poor. Since rich people as a group make a positive voluntary payment to charities for poor people, we might suppose that introducing an involuntary transfer will simply reduce the level of the voluntary transfers: redistribution through taxation reduces redistribution through charity. As in the previous case, it's natural to suppose that the reduction in charitable contributions COMPLETELY offsets the tax-financed redistribution. Clearly this argument can be used, not only against SS, but against any program for involuntarily redistributing income from group A to group B where voluntary redistribution from group A to group B is still positive after the involuntary redistribution is implemented.
Another possible "use" for SS is redistribute income from an individual when young to that SAME individual when old. E.g., suppose Joe would only save 6% of his income for retirement. The government can take another 5% as SS, and give it back to him when he's old, increasing Joe's total retirement saving to 11%. The flaw in this scheme is that if Joe's rational, he'll simply lower his voluntary saving to 1%, maintaining his total saving at 6%.
But what if, in the above example, SS takes 8% of
Joe's income. Then, even if Joe reduces his voluntary saving to 0,
he's effectively saving 8%, 2% more than without SS. Right?
Wrong: Joe can BORROW 2% of his income (on the strength of his future
SS payments), maintaining his total saving at 6%.
Return to main Social
Security page.
Return to main Apology
page.
This page maintained by Steven Blatt. Suggestions, comments, questions, and corrections are welcome at sjblatt@ocf.berkeley.edu.