Arnold Kling  

Social Security Indexing Debate

PRINT
Futurology... Private Investment is Dangerou...

Bill Sterling argues that the debate is long overdue.


If there was a substantive debate on wage indexation in 1976 it seems to have been entirely an inside-the-beltway affair.

Fearing to step on the dreaded third rail, President Ford ignored recommendations of the Consultant Panel on Social Security that were unusually clear about the long-run inequity and unsustainable nature of the shift to wage indexation...

We are pretty certain that in an honestly conducted survey about whether most people favor wage indexing versus price indexing of initial benefits the main answer would be: Huh?

With President Bush having made Social Security reform a high priority, the good news is that a genuine debate appears to be starting.


For Discussion. If you were to start with a blank piece of paper, would you choose wage indexing or price indexing for Social Security?


Comments and Sharing


CATEGORIES: Social Security



COMMENTS (22 to date)
Bob writes:

Asking us to design a SS system but limiting us to a single dimension? With apologies, let me broaden the debate to include retirement age.

Index initial benefits AND subsequent benefits to the average of price and wage increases. I fail to see the logic of indexing initial and subsequent benefits differently.

Index the retirement age to life expectency.

But any system we design will be screwed up by a future Congress that takes advantage of inevitable demographic fluctuations to buy votes. There is merit to the proposal for an independent entitlement board ala the Fed simply because, just like monetary policy, politicians' incentives make a crisis inevitable. Or we could just do away with the whole thing...

snacknuts writes:

neither. you'd used a price index based on a weighted average basket of goods & services purchased by retirees. as a result the medical care component (rising at around 6% per annum) would factor more heavily for old folks. also what bob sez.

the strategy of congress has always been to beseige future generations (who of course cannot vote) with current liabilities. the sensible thing to do however would be to raise the retirement age a few years.

Edge writes:

Any private savings plan has a goal of replacing a certain portion of your pre-retirement income.

Thus, with a clean slate, you would base retirement benefit on pre-retirement income.

Thus, you wage index, or perhaps as an alternative, index to a minimum acceptable living standard.

Lawrance George Lux writes:

Most who advocate Price-indexing know not what they do. They advance the notion of curtailing the Income of an increasing segment of the Consuming population, something which will reduce Business profits at a greater pace, than Savings in unpaid Benefits. lgl

spencer writes:

How about stating the question in a way that
makes clear what you are talking about. The real question is do you think SS receipents should participate in the almost ever rising standard of living. If you think that the current ss payments should be set to provide 40% of the 1935 standard of living then you would favor price indexing. Price index assures that ss benefits decline as a share of the pie forever and is an almost guaranteed way of making sure ss becomes so unattractive that it loses support and dies.
If we get price indexing the battle is over and ss has been destroyed.

Eli writes:

Price indexing. I don't think it's fair for each generation to retire in more luxury than it provided for the previous generation. Wage indexing is tantamount to a Ponzi scheme.

That said, I think each generation should finance its own retirement, and retire in precisely the amount of luxury that it can afford.

Clearly, we wouldn't design anything like what we have today. And, btw, since sunk costs are sunk we ought to be thinking as though we do have a blank piece of paper before us.

I suggest we replace the current system with a negative income tax for those above 65. And expand and simplify the 401k-IRA tax favored accounts. That way people can decide for themselves when--or, if--to retire.

Tom Myron writes:

The current objection to private accounts seems to be the vaugeries associated with the stock market. Rather than limiting private account investment to this venue why not consider the large number of savings banks throughout the country. Granted the rate of return does not have the potential of the equity markets (best seems to be around 3.5%) but the FDIC insured feature of savings banks certainly reduces the risk. Additionally, perhaps after an account has been funded for a number of years (say 10 yrs.) the account owner would be able to direct savings bank management to allocate a portion of the accrued funds to increased risk investment areas. The other advantage might be that using savings banks, monies would be spread over the entire country.

Victor writes:

I believe that the price-indexing discussed in the Hsaio report differs from the price-indexing proposed by Bush. Bush wants to keep wage-indexing within the AIME calculation, but keep overall benefits indexed to the CPI by adjusting the PIA factors.

The '76 Panel appears to me to be considering price-indexing within the AIME calculation.

I'm having browser issues, so I can't load chapter 3 to verify. But the summary discussed the fact that benefits still rose faster than inflation. Elsewhere, they also seemed to be focused on the indexing within AIME.

As far as the bigger picture goes, this is irrelevant, I suppose. But if you are asking about blank pieces of paper, these issues are important.

Jim Erlandson writes:

Before asking HOW to index Social Security payments, we should ask IF they should be indexed. And before that, we need agree on what problem Social Security is supposed to solve.

If we don't know where we're going with Social Security, how can we plan to get there?

Bob writes:

Spencer, not at all. I guarantee Congress would raise benefits when it was possible to do so without immediately raising taxes. You *have* to design a system where spending lags your means in order to offset Congress' bias. Not that you could actually sucessfully redesign SS for the long hall unless you take the decision-making out of Congress' hands but in the context of the discussion your point is myopic. Those who favor changes that put SS in a stable equilibrium are saving, not destroying, the program. Those who advocate doing nothing plant the seed of its demise.

Boonton writes:
Price indexing. I don't think it's fair for each generation to retire in more luxury than it provided for the previous generation. Wage indexing is tantamount to a Ponzi scheme.

Actually right now wage indexing only applies to the starting benefits...after that benefits are price indexed. Assuming wages grow faster than inflation (wage growth - inflation = productivity) then you have no problem at all.

As for generation in time T having 'more luxory' than generation in time T-1...you're really objecting to the fact that things get better as time goes forward.

Eli writes:

Boonton,

I am not "objecting to the fact that things get better as time goes forward." I am objecting to the fact that Generation T-1 is retiring with a higher standard of living than they would have if they were forced to save for their own retirement. And Generation T is footing the bill. In turn, Generation T is having its even cushier retirement financed by Generation T+1. Isn't this situation characteristic of a Ponzi scheme?

And yes, I am aware that price indexing is used after initial benefit levels are set. However, I don't see how that diminishes my argument.

In any case, I'd like to see the system phased out. You want to retire? Save money in your productive years.

Jim Glass writes:

If one is starting a new system from scratch, the answer is to use neither price indexing nor wage indexing, just real economic returns on real savings. (Like the original SS Act of 1935 basically actually did!)

Then if you decide you also want to transfer something to older folk who won't be in the system long enough to save much before retiring, set up a separate program to do just that. Make it clear and transparent that that is what you are doing, and have everyone know just who is transferring and receiving how much. When that program has done its job for that one generation, wind it up and finish it. Close it. Done. Finis.

And you'll be left with a sound system based on real economic savings.

Now, to get to that situation from where we are today is possible, as both Friedman and Kotlikoff show:

1) Take the $10 trillion backwards transfer that's causing all the trouble and segregate it from the rest of the system like it should be. That's all benefits earned but not paid and unfinanced to date. Then pay that off with income taxes (Friedman) or consumption taxes (Kotlikoff).

2) Have workers then have real savings in prersonal IRA-type accounts, say 6% annually.

The result is completely progressive: workers get more spendable money up front, higher returns (positive ones!), and thus more money in the future too. And the cost of the $10 trillion backwards transfer is shifted from low-wage workers to the rich, either high income earners or big consumers. Win-win-win from a progressive point of view.

Shifting to inflation indexing OTOH is just a patch on an open sore that solves nothing -- it is regressive, hurting the poor the most, and so will be (and should be) fought by all who think SS's best role is as a safety net for the poor.

It also drives returns on SS contributions even more negative to the young, which is the real thing that will kill SS politically a generation from now. So it really hurts SS on the merits in two ways.

If one wants to cut benefits to close the funding gap while also (for some unknown reason) not wanting to improve the rest of the program, keeping it as close to its broken obsolete self as possible, then the route to go is wealth means testing, like Australia uses.

Why, in a system that is insolvent, should Warren Buffett be collecting transfers from the paycheks of his Dairy Queen employees, as his insurance against poverty?

Why, 20 years from now, should we have a massive income tax increase to pay off the SS trust fund bonds (on top of a twice as massive increase to finance Medicare) in order to pay SS benefits to Bill Gates???

People aren't asking this now, but they surely will be 20 years from now when facing those massive tax hikes. But then it will be too late to avoid the train wreck.

Boonton writes:
I am not "objecting to the fact that things get better as time goes forward." I am objecting to the fact that Generation T-1 is retiring with a higher standard of living than they would have if they were forced to save for their own retirement. And Generation T is footing the bill. In turn, Generation T is having its even cushier retirement financed by Generation T+1. Isn't this situation characteristic of a Ponzi scheme?Why, in a system that is insolvent, should Warren Buffett be collecting transfers from the paycheks of his Dairy Queen employees, as his insurance against poverty?

Why, 20 years from now, should we have a massive income tax increase to pay off the SS trust fund bonds (on top of a twice as massive increase to finance Medicare) in order to pay SS benefits to Bill Gates???

It's always sad to see people resort to class warfare.

Edge writes:

There's really no need for a massive tax increase 20 years from now.

25 years ago, our debt as a fraction of GDP was about 33%. Today it's about double that.

If we need a massive tax increase 20 years from now, it'll be mainly about the fact that we're currently spending almost $1.50 for every dollar of revenue brought in for general revenues.

In the mean time, we've got more to worry about with foreign central banks, because our average debt maturity is around 3 years. We'll have to find new buyers for all that debt, too. The trust fund debt is at least more predictable than the publicly held debt. All the SS trust fund is asking between now and 2028 is that we start paying some of the interest in fifteen years.

Jim Glass writes:

"It's always sad to see people resort to class warfare."

First, Boonton, it is very bad netiquette to smush two different persons' comments together as if they were one, so let me separate them for you.

I am objecting to the fact that Generation T-1 is retiring with a higher standard of living than they would have if they were forced to save for their own retirement. And Generation T is footing the bill. In turn, Generation T is having its even cushier retirement financed by Generation T+1.

Now, one may wonder why it isn't class warfare when T puts its hands into to T+1's pockets, and how class warfare breaks out only when T+1 objects. Curious, that.

In 1983 we had a group of retirees and near-retirees who started out paying 3% payroll tax and who were promised the greatest benefits that SS will ever give anyone. But there was not enough tax revenue to pay for them.

So to please those retirees by maintaining their record-size benefits, Congress made young workers start paying tax at a new high 12.4% rate and slashed their benefits.

As the direct and intentional result of this, the '83-era retirees get a huge positive gain from SS on a lifetime basis, while the younger workers take a big loss on a lifetime basis from SS.

But this is not warfare between the generations.

Warfare breaks out only when -- sadly -- the younger generation objects. ;-)

Why, in a system that is insolvent, should Warren Buffett be collecting transfers from the paycheks of his Dairy Queen employees, as his insurance against poverty?

Why, 20 years from now, should we have a massive income tax increase to pay off the SS trust fund bonds (on top of a twice as massive increase to finance Medicare) in order to pay SS benefits to Bill Gates???


"It's always sad to see people resort to class warfare."

Aw, gee, Boonton, are you against the graduated income income tax too?

It seems the "progressives" mantra these days is that letting 'the rich' keep a little more of their own income by dropping their tax rates a couple points is bad, bad, bad ... but making direct transfers to the rich is good!

Questioning transfer payments from the poor to the rich is "class warfare". That's liberally amusing! ;-)

Anyhow, you might try answering the question on the objective merits, rather than referring to your subjective sadness:

Why should we increase income taxes post- 2018 to supplement the retirement income and pay the medical costs of Bill Gates? As current law mandates we will.

What is the economic/social/moral benefit that will be obtained? Precisely?

Jim Glass writes:
"There's really no need for a massive tax increase 20 years from now."

Oh yes there is. If you want to pay off those SS trust fund bonds -- they aren't going to pay off themselves you know! -- and have the government pay for Medicare too.

"25 years ago, our debt as a fraction of GDP was about 33%. Today it's about double that."

A trivial change compared to what's coming. Due to SS and Medicare alone the annual deficit will reach 20% of GDP on current law before the SS trust fund is spent. (Assuming one can "spend" a debt owed to oneself.)

So the increase in the national debt that you are complaining about which happened over 25 years will be matched every one-and-a-half years. Good, eh?

"If we need a massive tax increase 20 years from now, it'll be mainly about the fact that we're currently spending almost $1.50 for every dollar of revenue brought in for general revenues."

$1.23 in 2004. Bad enough, but nothing compared to what's coming.

"All the SS trust fund is asking between now and 2028 is that we start paying some of the interest in fifteen years."

All the SS trust fund is asking is that we pay off $5 trillion (today's current value) and interest on it over the period from 2018 to 2042.

For perspective, the entire national debt held by the public accumulated from George Washington's time to today is $4.4 trillion. Did you ever hear anyone talk about how easy it would be to pay that off over 24 years with an little income tax increase? That little income tax increase is going to be >20%.

And that's on top of the little income tax increase for Medicare, which has accrued liabilities twice as large, giving us another 40% increase. Together we are talking about an income tax increase of 60%. Just during the period to 2042. More afterward.

GAO projects the combined cost of Social Security and Medicare increasing by 15% of GDP to reach 22% of GDP by 2050 -- compared to an entire federal government consuming 20% of GDP today, and total federal income tax revenue today less than 10% points of GDP.

So if you don't like debt but you do like these programs, there are some income tax increases coming, eh? You do the math.

For the record, the US Treasury's unfunded liability for Social Security over the next 75 years is $12.5 trillion, current value. (Rather more than the $3.5 trillion reported by the Social Security Adminsitration, because it doesn't count little government obligations like paying off the trust fund bonds).

And that SS liability when up by $810 billion in 2004 -- compared to total personal income tax collections of $811 billion.

Of course, Medicare's unfunded 75-year liability as of end 2004 was $24.6 trillion, current value. And growing a lot faster than SS's.

That's a $37 trillion liability combined for the two of them, current value, and growing much faster than the economy.
[Treasury, pdf, p.11]

So, if you are concerned about debt as you seem to be, it seems likes there's a lot more for you to be concerned about than you've been aware of.

Take your eye off that mole hill -- look up at and admire the mountain!

Boonton writes:
Now, one may wonder why it isn't class warfare when T puts its hands into to T+1's pockets, and how class warfare breaks out only when T+1 objects. Curious, that.

I thought it was Bill Gates and Warren Buffet we were getting upset about. T is responsible for T+1 just as parents are responsible for their children. SS nicely ties T's interests to T+1's and T+1 has an interest in T+2's welfare.

The same exists in a 401K system. How will Mr. Gates cash in his 401K? Why by selling his shares to people in the T+1 crowd. Or if his 401K generates income from dividends or bonds it comes from income forgone by T+1.

A trivial change compared to what's coming. Due to SS and Medicare alone the annual deficit will reach 20% of GDP on current law before the SS trust fund is spent. (Assuming one can "spend" a debt owed to oneself.)

Jim, Medicare is like the elephant in the room and SS a large dog...yet you always seem to insist the dog comes before anything...such as:

GAO projects the combined cost of Social Security and Medicare increasing by 15% of GDP to reach 22% of GDP by 2050 -- compared to an entire federal government consuming 20% of GDP today, and total federal income tax revenue today less than 10% points of GDP.

Notice how Medicare always gets hitched to social security...even though Medicare is open ended...there's really no way to predict medical costs in 2050 except variations on straight lining the recent past a half century into the future (note anyone doing that in 1950 to prepare for 2000 would have ended up way off).

If you cap SS at a certain % of GDP you are guaranteed positive returns as long as growth is positive. If growth isn't positive then market returns will do nothing for you.

Edge writes:

$1.23 in 2004. Bad enough, but nothing compared to what's coming.

I'm not talking about the unified budget. I'm talking about on-budget. OK, so we're only spending $1.43 for every dollar of revenue brought in , rather than $1.50. By practically any means of accounting, finances for the part of the federal budget funded by income taxes rates an F, while Social Security's finances rate about a B.

If we run that sort of deficit, you could eliminate Social Security and Medicare and you'd still need a massive tax increase 20 years from now, because our federal debt would grow exponentially. We'd need to pay about 4% of GDP indefinitely just to pay the interest on the debt you'd run up in 20 years . The logical, rational solution is to do something like we did between 1993 and 2000 - plan ahead, rather than running the rest of government at such a massive shortfall. We reduced federal debt by about 10% of GDP between 1993 and 2000. That gain has been wiped out in the past four years. That's irrational policy for a government projecting a future need to pay off a large portion of its debt starting 20 years from now.

You've lumped together Medicare and Social Security, and discounted not only the $1.5 Trillion currently in the SS trust, but also the additional $2 or so Trillion that's yet to go into the SS trust in the form of surplusses between now and 2018, to arrive at an unfunded liability for Social Security.

If you believe Medicare is a problem, then propose a way to fix it. But Medicare being a problem doesn't mean we have to phase out Social Security, or worse still, phase out Social Security with a privatization plan that requires a massive increase in publicly held debt for 45 years before the program starts to come back into balance with benefit cuts.

"T is responsible for T+1 just as parents are responsible for their children."

We're all just one big family, eh? I suspect that many will not agree when the time comes to have their taxes massively raised.

Boonton writes:

And when will that be? Remember if SS tax revenue never dips lower than 70% of benefits scheduled when will the massive tax increase come? How objectionable can it be when T spent quite a long time overpaying SS taxes in order to generate the trust fund surplus?

Comments for this entry have been closed
Return to top