Arnold Kling  

A Must-Read on Social Security

PRINT
Three that Caught My Eye... Counterexample on Female Patho...

A Well-Tailored Safety Net should be read by every policy wonk or would-be wonk who is focused on the long term fiscal outlook in general and on Social Security in particular. There are two reasons to read it.

1. Author Jed Graham has one original idea that is quite interesting.

2. He offers clear, fair analysis of many of the alternatives for dealing with Social Security.

He calls his idea "old-age risk-sharing." I would call it "age-related retirement benefits." In math geek terms, I would call it a "continuous retirement-age function."

Currently, your retirement benefits are a crude step function of your age. You can get a benefit of $X if you wait until the Normal Retirement Age, or you can get some set fraction of $X if you retire early. Under Graham's approach, you would get $X when you hit, say, 85, but up until then your benefits would start below $X and increase gradually.

Politically, this means that Congress does not have to raise the retirement age, with all the psychological drama that involves. Instead, they cut benefits for younger retirees, so that if you want to maintain a better lifestyle after age 67 or whatever you have to either save more early or work longer. Graham would point out that working when you are 70 is an easier option than working when you are 85, so this pattern of age-related retirement benefits is more reasonable than an all-out benefit cut.

It also has some advantages over a drastic increase in the retirement age. For one thing, you could inflict some of the benefit cuts in the near term (say, 2013) without causing as much of a ruckus as raising the retirement age on such short notice. For another, you make the boundary between "not retired" and "retired" fuzzier, which I think is appropriate, given that for many people these days, sitting on a rocking chair in Florida is not their idea of how to spend their late 60's and early 70's.

More of my comments follow.

On p. 15:


Think of the Social Security Trust Fund as a cookie jar where a couple tries to save for retirement by tucking away a $100 bill every week. But every weekend, they give that $100 to their kids to spend on movies, pizza, new clothes, and video games. Still, the couple doesn't worry about their inability to save, because each time they remove a $100 bill from the cookie jar, they replace it with a piece of paper that says "You kids owe us $100 plus interest to be paid out in an annuity upon our retirement."

Of course, most of us have long understood this, but you would be amazed to see who writes as if the trust fund actually exists.

p. 57:


For those who claim Social Security at 62--as more than half of all workers now do--the monthly benefit checks they get for the rest of their lives will be 25 percent smaller than if they had waited until they turned 66--the so-called Normal Retirement Age...

Don't you think these people need to nudged into waiting until their Normal Retirement Age? Why doesn't a life insurance company offer a "bridge" loan of four years of income now, to be paid back out of your higher Social Security benefits later?

p. 78, he offers a succinct argument for age-related retirement benefits:


Only by scaling back Social Security's promise of an early retirement, can we avoid the need for benefit cuts in very old age, when people can least afford them.

The book also does a nice job of laying out the pros and cons of other well-known reform proposals, from Diamond-Orszag to Paul Ryan. Those chapters are what make this a book for policy wonks, not just an article on Graham's own proposal. Overall, Graham strikes me as fair to both sides, not out to please either the left or the right.

I wish him good luck in getting "old-age risk sharing" or what I am calling "age-related retirement benefits" into the policy discussion. In any case, we need a grown-up discussion of Social Security, and this book can contribute to that.



COMMENTS (9 to date)
Peter writes:
Don't you think these people need to nudged into waiting until their Normal Retirement Age? Why doesn't a life insurance company offer a "bridge" loan of four years of income now, to be paid back out of your higher Social Security benefits later?

The reason is that Social Security payments have a special legal status unlike anything else, and can never ever be attached by a creditor for almost any reason. The only ones I know of are child support and back taxes owed.

So you would be giving them an unsecured loan the size of a small mortgage. People taking Soc. Sec. at 62 probably don't qualify for a $75,000-$100,000 personal loan. A more reasonable course might be to take out a 2nd mortgage (or 1st if the house is paid off), but that market is in a shambles now.

Doc Merlin writes:

"Politically, this means that Congress does not have to raise the retirement age, with all the psychological drama that involves. Instead, they cut benefits for younger retirees, so that if you want to maintain a better lifestyle after age 67 or whatever you have to either save more early or work longer."

This was effectively already done when they started taxing portions of SS benefits.

rhhardin writes:

A social security trust fund can't exist. The government must return every penny it takes in to the economy at once, lest the money supply fall.

The bookkeeping can vary, but it always winds up as an IOU in the trust fund.

And of course money isn't wealth to the nation, but just tickets in line to say what the US economy does next, presumably something for you. It does no good to squirrel away tickets if there isn't going to be a future show.

azmyth writes:

Perhaps "increasing annuity" would be a good term. If payments increased enough, you could roll Medicare into Social Security and eliminate all the health care bureaucracy associated with Medicare, because payments would increase as health care costs increased.

Two Things writes:

Doc Merlin, are you sure your words fully express your intended meaning?

The income tax on [50% of] social security benefits penalizes both saving and working longer, because under our progressive income tax system it pushes retirees into higher tax brackets.

The Social Security Administration writes: "As a general rule, early or late retirement will give you about the same total Social Security benefits over your lifetime. If you retire early, the monthly benefit amounts will be smaller to take into account the longer period you will receive them. If you retire late, you will get benefits for a shorter period of time but the monthly amounts will be larger to make up for the months when you did not receive anything."

What SSA doesn't point out is that the higher your monthly benefits, the larger percentage tax you pay on them. That means the total you collect from the government (net of benefits and taxes) is less if you retire later (assuming your total post-retirement income, including savings interest or capital gains, is more than the minimum income tax threshold).

I understand that including (some or all of) social security benefits in income for tax purposes is a kind of back-door means-testing of social-security benefits. Those who continue to earn money, including interest on savings and/or capital gains, end up taking less from the government (net of social security and income tax).

The SS tax discourages post-retirement work as well as savings. Social security recipients pay higher marginal taxes on their wages than younger workers with similar jobs, making them particularly uncompetitive workers-- they're old and slow and their incentives (net wages) are lower.

I'm not trying to argue about whether that's "fair" under our system of income taxation, only to point out that any given job, even part time, offers less "take home pay" to a retiree than to a younger worker-- which is probably contrary to what you want if you would prefer to see old people supplement their savings + social-security with wage labor.

Two Things writes:
Why doesn't a life insurance company offer a "bridge" loan of four years of income now, to be paid back out of your higher Social Security benefits later?

Because "later" SS benefits aren't enough larger to pay the combined life-insurance premium and interest on that loan (well, maybe at today's Fed-induced artificially-low interest rate, but not at historical rates). Like a life annuity, SS payments stop when the beneficiary dies whether or not he still owes something to a lender, so the risk of default would be very high. (The loan, of course, would be unsecured-- you pretty much can't hypothecate Social Security benefits.)

rhhardin writes:

The SS recipient isn't trying to maximize money taken from the government, but to get the best insurance against living too long.

In that case, postponing benefits as long as possible is the best strategy. That maximizes the income if you live a long time.

If you don't live a long time, you don't need it.

Michael Jordan writes:

Here's a thought on how to deal with the Social Security issue. How about if we take the individuals who are receiving this benefit and offer them the opportunity to work? I realize that factory jobs and ditch digging would be out of the question for most of the 60 and over class. It wouldn't even be fair to expect it of them. But I know from personal experience that many retirees find out that, in fact, they don't enjoy retirement because they are, plain and simple, bored. So right off the bat I believe you would have a reasonable share who would be willing to do light work just for the benefit of having something to do and feeling useful once again. I am sure that there are tons of over-worked secretaries and executives who wouldn't mind having someone to shuffle papers and take care of some of the simpler, less physical aspects of office life.

This would also benefit everyone simply because you wouldn't lose the age and experience that comes with these workers. They have all learned a host of things from their experience in the real world. Putting them into some of the less physical government positions would make their experience available. That, in my opinion, would do a great deal to help this country out. As it is, we are looking at losing a huge amount of experience when the baby boomers swing into full retirement mode.

Better yet, make a portion of their income linked to working. Unless they have a medically valid reason for not working at all, make, say, 10% of their benefits contingent on them working 20 hrs. a week. Those figures are just an example, of course.

By doing this we would allow those that wish to continue to be productive members of society, and it would ease some of the financial issues plaguing the nation's economy. Even if the benefits were not direct as far as reducing benefits, I believe that a large amount of cumulative labor - and thus dollars - would be contributed to the general financial picture. I would see it as a win-win for everyone if it is done correctly.

richard writes:

I have never understood the trust fund claim because it is easily debunked. SSA has query tool where you can actually look up the composition of the "assets" in the trust fund. The SSA goes on to great pains to distinguish the trust funds holdings between Marketable Securities (T-Bills) and Un-Marketable Securities (IOUs). As one might imagine the amount of Marketable Securities in the Trust Fund is miniscule, like a quarter billion. During the crash I even saw Brad DeLong say that we should take all of those T-Bills from the Trust Fund and invest them in stocks. He actually worked at the Treasury (he should know better).

Comments for this entry have been closed
Return to top