Arnold Kling  

Et tu, Barro?

PRINT
Second Thoughts About Stats vs... Feelings vs. Consequences...

Robert Barro writes,


A serious analysis with asking why we have Social Security. If we were not so used to it, we would find it odd for the government to collect money from young workers and give it to the old (mostly workers' parents). One rationale is that the government should help people who lack discipline to save for old age. I have never embraced this paternalistic view.

What Barro is arguing is that we should get rid of Social Security as we now know it, rather than replacing part of it with private accounts. He would like to trim back Social Security by making it a program that offers only a minimum benefit that is the same for all, in order to prevent extreme poverty among the elderly; indexing the minimum benefit to prices rather than wages; and raising the retirement age. Thus, he supports all of the major methods of shrinking future benefits.

The Washington Post was excited to report this as a split among conservatives on the Social Security issue. In fact, I believe that conservatives are quite united in their belief that it is a fiendishly bad program (fiendishly bad because the pay-as-you-go structure is so difficult to unwind). There are tactical differences over how best to go about taming the monster.

For Discussion. What are the pros and cons of Barro's (and Tyler Cowen's) idea of trimming Social Security benefits to a fixed minimum benefit?


Comments and Sharing


CATEGORIES: Social Security



TRACKBACKS (6 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/224
The author at PRESTOPUNDIT in a related article titled ECONOMIST Robert Barro turns his guns on writes:
    Social Security "private accounts." (pdf file) And Barro isn't the only classicly liberal economist who's turned on the "private accounts" version of Social Security reform. The WaPost has the... [Tracked on March 29, 2005 4:17 PM]
The author at aWpTiMuS.com in a related article titled Is It Really a Split? writes:
    Conservatives Splitting on Social Security, by Jonathan Weisman, Washington Post Why Private Accounts Are Bad Public Policy, by Robert J. Barro, Business Week Robert Barro, economics professor at Harvard say he once used to support private ac... [Tracked on March 30, 2005 12:27 AM]
COMMENTS (43 to date)
Ian Lewis writes:

I think the idea is not bad. But has any serious economist (or politician) put forth the idea of having each state maintain it's own Retirement Program? And if someone were to move from one state to the next, they could then transfer those benefits, or just get a cash payout. This would add some competition to the programs and make those in charge, at least a little, more accessible.

Axel Kassel writes:

Prof. Barro's remarks have, unsurprisingly, a great deal of good sense embedded in them.

Starting fresh with non-uniform demographic assumptions, a rational planner would indded be hard-pressed to justify a scheme that combines forced retirement savings and intergenerational welfare payments, both financed by a regressive tax on payrolls. And Prof. Barro's reform basket would indeed be a sensible approach to patching the system.

Congress is not well known for embracing sensible approaches, however, so I shall continue to embrace the private-accounts options as a useful transition device toward:
(1) Encouraging more people to think about financial planning, investing, and the advantages of ownership as distinct from a poltical entitlement that can be extinguished without notice, and that carries no actual property rights with it; and
(2) Easing the economic fall-out from the financial epiphany awaiting citizens of my children's generation a few decades out, when they realize that their families' welfare is being laid as a sacrifice on the altar of the United Church of AARP, many of whose members will be more flush than they. When the rising generation realizes that they outnumber the superannuated leeches who are raiding their resources, there will be electoral hell to pay. Encouraging private accounts could mitigate those future raids somewhat, possibly tempering the appeal of Draconian cuts, and offer some protection to those who might otherwise be taken totally be surprise by a political uprising.

I would, however, raise a slight objection to the following Barrovian passage: "From the perspective of the trust fund, returns look low because the fund's government bonds have paid less than stocks. But the premium on stocks is compensation for risk, as gauged by financial markets. Although the ability to hold stocks is
a plus, there is no free lunch of assured higher returns."

Indeed, there is not. But the use of the word "risk" here plays into the hands of the shrieking demagogues running the AARP's ad campaign, inviting visions of wiped-out principal. Holding shares of bond or blended funds can mitigate risk.

But even if we speak only of equity risk, a diversified, indexed holding reduces the salient risk to that of market volatility, not apocalyptic loss in a few ill-chosen Enronnish plunges. After all, stocks outperform bonds--especially Treasuries, even after you have carried out the dead. The inflation-adjusted total return for the S&P 500 Index since 1926 is reported to stand above 7 percent, and I understand that holding period included some grim stretches...

Boonton writes:
In fact, I believe that conservatives are quite united in their belief that it is a fiendishly bad program (fiendishly bad because the pay-as-you-go structure is so difficult to unwind). There are tactical differences over how best to go about taming the monster.

True conservatives would respect a program that has been embedded in the economy for so long and would be highly skeptical of plans to rip it.

But even if we speak only of equity risk, a diversified, indexed holding reduces the salient risk to that of market volatility, not apocalyptic loss in a few ill-chosen Enronnish plunges.

But Social Security does not have to rely on the stock market but the entire economy to pay benefits since it derives its funding from the payroll tax. Unlike a portfolio planner, SS doesn't have to try to guess which companies will have employees in 2040 it has perhaps ultimate diversification. This makes Social Security remarkably stable but it does carry a cost. Returns have to be closer to the growth of the overall economy.

Axel Kassel writes:

Re another posting:

"True conservatives would respect a program that has been embedded in the economy for so long and would be highly skeptical of plans to rip it."

I fear this confuses hidebound-tradition conservatives such as England's 19th-Century Tories with conservatives who prize liberty and limited government, and respect for rights as goods in themselves, not simply because they are referenced in faded parchment.

---

"Unlike a portfolio planner, SS doesn't have to try to guess which companies will have employees in 2040 it has perhaps ultimate diversification."

I fear the commentator stopped reading before noting the point about index funds, whose merits include NOT expecting a portfolio manager to prophesy which companies will succeed. As for Social Security, its (temporary) cash-flow surpluses are by law invested in Treasury securities. And after 2040, give or take a couple of years, there will be nothing in the trust fund and the program will be pure pay-as-you-sink. In either case, it has no asset diversification whatsoever.

Boonton writes:

I fear the commentator stopped reading before noting the point about index funds, whose merits include NOT expecting a portfolio manager to prophesy which companies will succeed.

Index funds are indeed great but they still represent only limited diversification. Being able to tax all payrolls in the US with the exception of under the table work is even more diversified.

As for Social Security, its (temporary) cash-flow surpluses are by law invested in Treasury securities. And after 2040, give or take a couple of years, there will be nothing in the trust fund and the program will be pure pay-as-you-sink. In either case, it has no asset diversification whatsoever.

1. No business in the US can offer a reasonable guarantee it will still be on ongoing concern in 2040. Since the Treasury abolished 30 year bonds you can't even use them to get you to 2040. No mutual fund is able to guarantee its existence to 2040 either.

2. There's very good reason to think the projections are highly biased against social security. Brad de Long, for example, has noted how their projections of productivity increases are pulling things down despite the fact that we have been experiencing years of productivity far above the projection.

3. Even assuming the projections are accurate, revenue never falls below 80% of promised benefits. Deficits max out around 2% of GDP. Compare that to today when we are happily clocking 4% of GDP deficits yearly.

4. By definition a true pay as you go plan with no surpluses or deficits yields a return equal to economic growth. The mathematical proof is simple:

Take a simple SS model that applies a tax rate T to income (I) and simply pays that benefit to the retired generation. What would be the return on Generation X?

Tax for generation X = T * I

Benefit for generation X = T * (I * r)
where r is the rate of growth

Return = Benefits / Taxes = T(I*r)/T*I = r

5. This argument is a bit more difficult to make in a few words but I'll try. There's a big difference between stated returns and realized returns. A while ago someone posted a study that showed the S&P had a rate of return for some time period of 13%. Investors in the market during that period, though, had a realized rate of return of only 3.5%. Why the dramatic difference? Because most people did not invest in the S&P 500 for that time period. People were buying stocks in different proportions than the S&P 500 and were sometimes winning big but also sometimes losing big. The losses of those investors allowed the few S&P 500 investors to reap the pure 13% return (as well as those lucky to do even better).

5.1 Basically if we could turn back the clock and force everyone to have invested in the S&P 500 at the start of that time period the result would not have been a whole nation of lucky people with retirement funds that had yielded 13%. The result would have been the S&P 500 would have had lower returns for that time period.

5.2 Long story short, there's a boundary to the returns that can be given to everyone. It is impossible for everyone to beat 'the market' because 'the market' is everyone! A piece of the economy cannot give everyone returns larger than the returns of the entire economy. So it is very foolish to assume the stock market will behave as it historically did if you radically change the economy and turn Social Security into millions of stock portfolios.

KipEsquire writes:

Something like this has also been proposed at the other end of the age spectrum: that a young physically and mentally competent adult, when reaching, say, 18 or when graduating from high school, would receive a one-time lump sum "safety net" payment that would cover all the expenses necessary for a "fair" start in life (I think I heard $40,000 when I was in college, but that was 20 years ago). In exchange for this, essentially all welfare and redistributionist government programs would be dismantled. If a person were to squander their "fair start" money, then they would have no recourse to the welfare state.

Lyn Burke writes:

Being in the exempt category, I plan to take all SS benefits at 62 and invest them in Roth/IRA's in my children's accounts (now in mid 20's) so that they will eventually have some return from SS when they retire. When I actually need SS, they should have a fairly good sized retirement account to make up for the fact that they will be paying such enormous amounts of payroll taxes that nothing will be left for them to save out of their own paychecks. I'll include the AARP dues that I no longer pay to fund the lies propagated by that organization.

Boonton writes:

Considering that even a poorly educated but fit 18-yr old can usually support himself or at least has parents willing to help him why would we want such a policy?

nmg writes:

Barro's plan is one step short of abolishing SS and replacing it with a simple welfare program for destitute seniors. This is of course, EXACTLY what should be done.

I notice that he fails to include two other necessary ingredients:

1. The benefits should only be given to the retirees who are destitute. It should NOT be a universal benefit.

2. Eliminate FICA.

#2 is more important to our economy and would almost instantly lead to a demand for #1 anyway.


nmg

Mark Horn writes:
A while ago someone posted a study that showed the S&P had a rate of return for some time period of 13%. Investors in the market during that period, though, had a realized rate of return of only 3.5%.
This is not at all surprising. If during the course of the next 5 days, a given stock's price goes up $1 per day, it's certainly possible that one investor (who invested for only the first day) will only make $1, and the next investor (who invested only the 2nd day) will make $1, and so on. At the end of 5 days, the average growth per investor is only $1 but the growth of the stock is $5. In this simple example, no one need lose any money yet the increase over a longer period of time is greater than the average increase of each investor.

In other words, measuring average investor increase over a time period is not a good measure of the investment.

Mark Horn writes:
5.1 Basically if we could turn back the clock and force everyone to have invested in the S&P 500 at the start of that time period the result would not have been a whole nation of lucky people with retirement funds that had yielded 13%. The result would have been the S&P 500 would have had lower returns for that time period.
Of course! You have two choices here. Either you allow investors to get out when they want to in which case they will take short term profits - bringing down the average - or you force them to stay in, in which case that limits trading and the ability to change the prices.

But even if that's true, that doesn't mean that investing in the stock market is a losing proposition. It may very well be that the economy is running slower because too much money is tied up in SS instead of being available for investment.

Mike Linksvayer writes:

Relative to a transition to personal accounts the Barro/Cowen package is low risk/cost and high benefit.

It would be hard to deform reducing benefit growth and increasing retirement age into something that increases the overall impact of non-market mechanisms. That isn't at all clear for (mandatory) personal accounts.

The third part of the Barro proposal, the same benefit for all, could increase the size of SS if the benefit were set too high, as it may have to be in order to get past opposition. So I'd stick with the two straightforward pieces of the package in the short term.

In the long term turn SS into a welfare program only available to the poorest of the oldest, or eliminate the thing, as the winds allow.

Boonton writes:

But even if that's true, that doesn't mean that investing in the stock market is a losing proposition. It may very well be that the economy is running slower because too much money is tied up in SS instead of being available for investment.

Nonsense, the SS surplus is not 'tied up'. It's primary effect is to reduce gov't borrowing. If the gov't borrows less then by definition there are more resources available to invest in the stock market (or other types of places).

Fundamantally, though, the stock market is simply a part of the economy. It's ability to actually provide realized returns is limited by the overall size and growth of the economy.

If not then you've basically found a magical money tree. Government can simply borrow $100B every month at 6% or so and invest it in index funds and reap returns of 13% over the long term. If this really held then all taxes could be abolished since gov't could fund itself from this 'magical money tree'. Rule 1: If you think you found a magical money tree something is fundamentally wrong with your analysis.

Boonton writes:

This is not at all surprising. If during the course of the next 5 days, a given stock's price goes up $1 per day, it's certainly possible that one investor (who invested for only the first day) will only make $1, and the next investor (who invested only the 2nd day) will make $1, and so on. At the end of 5 days, the average growth per investor is only $1 but the growth of the stock is $5. In this simple example, no one need lose any money yet the increase over a longer period of time is greater than the average increase of each investor.

Let's say the stock on day 0 was $10. The returns each day would appear as follows:

Day 1 $11 10%
Day 2 $12 9%
Day 3 $13 9%
Day 4 $14 7.7%
Day 5 $15 7.1%

The realized returns to the investors who got in and out each day average 8.56%. Yet the 5 day rate of return was 50%.

But let's go a step deeper. What is happening here is each day a person is willing to buy the share for $1 more than the previous day. The incremental investment each day is only $1. If a single investor had purchased those shares he would have made a handsome 50% return at the end of 5 days. if 5 investors had been day trading those shares their average return would have been a much lower, but still respectful 8.56%. In either case, though, the realized income was $5 and only $5.

Mark Horn writes:
Nonsense, the SS surplus is not 'tied up'. It's primary effect is to reduce gov't borrowing. If the gov't borrows less then by definition there are more resources available to invest in the stock market (or other types of places).
If I assume that the gov't is an effecient producer, then I can buy the argument that money used in gov't spending is not tied up. But the gov't is not an effecient prodcuer. Consequently any money used by the gov't is less effeciently deployed than if deployed in the market. I agree that the stock market is part of the economy. But I disagree that continuing to invest in an incredibly ineffecient producer is beneficial to the economy.

As a side note, I find it interesting that you don't consider the gov't giving a bond to the SSA not to be borrowing.

nmg writes:

Rule 1: If you think you found a magical money tree something is fundamentally wrong with your analysis.

That is funny to read from someone defending Social Security.

nmg

Lawrance George Lux writes:

Barro wants the One-Rate monthly benefit which I do, but he like all Conservatives minimize the importance of Social Security benefits themselves. Seniors would not maintain the stable Spending pattern, one that has spent Us out of every Recession since the 1970s, with reintroduction of uncertainty of Income--they abscent of Income-earning potential. lgl

Boonton writes:
As a side note, I find it interesting that you don't consider the gov't giving a bond to the SSA not to be borrowing.

It's borrowing in only an accounting sense but it is not borrowing from the private sector.

That is funny to read from someone defending Social Security.

Why so? I never said Social Security benefits come out of thin air. They come from the revenue generated from the payroll tax.

If I assume that the gov't is an effecient producer, then I can buy the argument that money used in gov't spending is not tied up. But the gov't is not an effecient prodcuer. Consequently any money used by the gov't is less effeciently deployed than if deployed in the market. I agree that the stock market is part of the economy. But I disagree that continuing to invest in an incredibly ineffecient producer is beneficial to the economy

This is an odd way to think about things. When the gov't runs a surplus (or less of a deficit) then it is borrowing less. Borrowing less from who? Why the private sector (plus other institutions like foreign central banks and so on). If the gov't increases its deficit then it increases its borrowing from the private sector.

If, for some reason, I thought project A could be done by gov't more efficiently than the private sector then it makes sense for gov't to borrow the money to do project A. Why? Because the rate of return on that project would be greater than the return the private sector demands for loaning the funds. Of course SS is a transfer program so it isn't trying to produce anything like gov't built highways 'produce' transportation or NASA produces scientific probes.

So when do people 'invest in the gov't'? When the gov't borrows more or less from the private sector?

mcwop writes:

NMG comments on an approach that I have been interested in lately, which is where SS's retirement benefit more insurance-like. A means test if you will. NMG is right to point out that with a lower expected benefit that there should be a corresponding premium reduction (lower payroll taxes).

Bob writes:

Boonton,

First, if the stock market earns 13% in a given period, the average return is 13%. The suckers that earn 3.5% don't allow the sharpies to earn 13%, if they did the overall market return would have to be lower than 13%. If you have the sharpies making more than 13%, then the math can work.

UNLESS you are making a diminishing returns argument - that the suckers have most of their money in zero-interest accounts and if they put it into the market that would lower future returns, preventing the sharpies from getting their 13%. Okay. But this does not drop the stock market return to GDP growth (unless you eliminate ALL leverage).

But is your point that in your mind private accounts' meager returns would make them a bad deal for all/most? Or that the fact that private accounts would be a bad deal for the suckers is enough to justify screwing the sharpies?

Or is this a M-M argument that capital structure doesn't matter so whether SS taxes go into private investment or pseudo-gov bonds doesn't matter (or are never collected) - the economy will be the same size either way?

I am surprised to see you bashing Barro's position - I thought it was fairly close to your own. Certainly he makes the point about private accounts being a red herring nicely.

Finally, your claim that the Fed Gov is better diversified than the S&P500 is not self-evident. The S&P500 includes large international operations, so unless you consider Iraq the first of the Fed Gov's offshore subsidiaries...

Also, in an imperfect market it is an error to pursue diversification to its extremes because the costs quickly swamp the benefits. Last I checked, our political market looked pretty imperfect to me. ;-)

Boonton writes:
NMG comments on an approach that I have been interested in lately, which is where SS's retirement benefit more insurance-like. A means test if you will. NMG is right to point out that with a lower expected benefit that there should be a corresponding premium reduction (lower payroll taxes).

We aren't that far from it. Since the payroll tax is regressive there's already a 'premium reduction' if you assume those with higher incomes are less likely to become impoverished in old age. Benefits are also 'taxed' so seniors that are working see reduced benefits to a degree.

You could make the program more insurance like if you audited retirees and reduced their benefits based on their accumulated assets (401K's, real estate equity and so on). Then benefits would be concentrated on those who 'really need them' and you could use the savings to either reduce the payroll tax, other taxes or fund other programs. I don't think people would like this very much, though.

You all seem quite eager to mess with Social Security but come up quite short on reasons why. As a form of social insurance SS provides a base of risk free income for retirees. Since it is funded by the gov't's taxing power it achieves a level of economic diversification that even a perfect portfolio could never come close to achieving. I believe its popularity also is caused by the social stability it provides. Since there is a guarantteed stream of income there is less political termoil as various groups demand to be bailed out.

The fact is the stock market returns more because it has a higher risk factor. That means returns have a higher deviation from the mean. Is it really reasonable to assume that large numbers of people whose retirements end up being 20% less than their neighbors simply because of a single bad year will sit idly by and accept it as fate or will they demand that the gov't bail them out?

Even a small bailout has serious implications for a system of 'private accounts'. It will almost be impossible to resist the demand of a large number of market losers to be partially bailed out but the precedent will create a moral hazaard as people accept higher levels of risk to achieve lottery like returns. As we know from the S&L debacle such situations create real 'hidden liabilities' as the cost of the bailouts grow year by year off the books until they cannot be ignored anymore.

Mcwop writes:

Boonton,
I think we are very far from it. Most that pay in collect retirement benefits. Some is taxed, but by large most get benefits out. This is different than insuring a few unlucky people. In turn if I get no benefits, I would want a reduction in the payroll tax (you do mention this).

Frankly, I do not like any of the proposals that I have seen so far. My conclusion right now is to leave SS exactly as it is. No payroll tax hike, no benefits changes, no increase in the retirement age. Unfortunately, I have a bad feeling that in 10 or so years, we will see a hike in the payroll tax, benefits reductions, and an increase in the retirment age. I do not want the system dismantled, but I do not want to see negative changes either (hike in the payroll tax, benefits reductions, and an increase in the retirment age).

I undertsand what the SS program does, I am open to changes that make the sytem stronger and better, but all of the proposals now floating around come up short on this.

Boonton writes:

Finally, your claim that the Fed Gov is better diversified than the S&P500 is not self-evident. The S&P500 includes large international operations, so unless you consider Iraq the first of the Fed Gov's offshore subsidiaries...

Also, in an imperfect market it is an error to pursue diversification to its extremes because the costs quickly swamp the benefits. Last I checked, our political market looked pretty imperfect to me. ;-)

It's not the Fed gov't that is diversified it is its taxing power. No private fund has the power to tax something like 'all corporate income' or 'all payrolls'. This ability means the gov't doesn't have to guess which companies will be successful, they just have to set a tax rate. Whatever company is the Intel or Microsoft of 2030 will be paying a heap of payroll taxes and corporate taxes to the Fed. Gov't. Your portfolio, however, will pay you nothing from those companies unless you're lucky enough to pick them today (or you are so spread out in an index fund that you get a taste of them anyway). The story doesn't end there, though, the taxing power extends to private and very small companies that you cannot tap through the stock markets.

Since the taxing power is limited to the US jurisdiction, I agree it cannot diversify beyond US shores (although large international companies usually end up paying something to the US gov't). I don't think that is sufficient to make up for diversification of the taxing power. If it was, though, there is a clever way the gov't could match it. The SS surplus could simply buy bonds from foreign gov'ts. Then the US gov't could tap the taxing power of the EU and a host of other national gov'ts (they use their taxing power, after all, to pay the interest on their bonds).

This isn't a magical money machine. Since taxing is destructive there's a point where a marginal increase in tax rates will cause a marginal decrease in tax revenue. So while Social Securities funding is remarkably secure compared to private 'diversified' accounts, there's a limit to how much SS can be expanded without becoming destructive.

I think 5% of GDP going to our national 'pension program' is about right. I would be weary of 10% or more. Brad de Long made an argument that SS should be cut because Medicare/Medicaid is a form of 'hidden social security' that promises to explode in costs. A reasonable argument but quite frankly I don't trust this administration to do anything honestly or correctly so I would opt for doing nothing for the time being.

First, if the stock market earns 13% in a given period, the average return is 13%. The suckers that earn 3.5% don't allow the sharpies to earn 13%, if they did the overall market return would have to be lower than 13%. If you have the sharpies making more than 13%, then the math can work.

Could you apply this to the very simple stock market we depicted here? We looked at a single stock that returned 50% in a 5 day period, yet the actual return realized by investors was 8.56%.

But is your point that in your mind private accounts' meager returns would make them a bad deal for all/most? Or that the fact that private accounts would be a bad deal for the suckers is enough to justify screwing the sharpies?

I think my point is best said by asserting that SS nicely offsets the risks of private accounts. For example, a risk with a private account is that you reach retirement with a nice lump sum but if your lifespan is very long you will end up using up all that money before you die. The point when you realize you are in trouble is likely to be when you are in your late 70's when it is unlikely you can go back to work; not in your late 60's when you can take some steps to beef up your income. Annuities are a private sector answer but they are not developed as a market and are often expensive. Social Security, however, has its risk in the opposite direction. If you live very long your payoff is all the better, its only living very little past 65 that you lose.

So the best of both worlds would be to use Social Security as a base and build up a private account (which we already have in the form of 401K's/IRAs plus non-retirement accounts like real estate equity). The privitization plans, however, seem to take their first step in carving out Social Security which I think is a bad idea. Why not simply do a SS+ program. Mandate 2% of payroll go into a private account above and beyond SS contributions. If you're already putting money in a 401K or IRA that amount gets deducted from your mandate. You've got the best of all worlds:

1. Private savings automatically increases by 2% of payroll.
2. If SS projections prove to be true then it is that much easier to asorb benefit cuts, tax increases or gov't borrowing if people have already built up a nest egg.
3. If SS projections turn out to be false then people get SS plus their private accounts to enjoy.

DeadHorseBeater writes:

Boontoon:
You're treating things as if the taxes taken were invested in government bonds, earning interest rate r. But in a PAYGO plan, no assets are accumulated. The interest rate (on T-bills, corporates, or whatever) doesn't enter into it. However, even if it did, that interest rate r would be a a risky asset, so much of the following analysis would still apply.

The rate of return in a PAYGO plan financed by a payroll tax is n+g where
n is the rate that the working age population grows relative to the elderly population
g is the growth rate of labor productivity=the growth rate of real wages

There is no strong reason to think n+g=r, even on average over the long run.

An asset that offered to pay you n+g would be a risky asset, since its returns would vary from year to year.
Notice that you could lower your overall risk exposure by not 'investing' 100% of your retirement assets in SS.
Some degree of diversification almost always reduces risk. You can even reduce risk by diversifying from a slightly risky asset into a riskier asset if the correlation of RoRs isn't too high.
For instance, IIRC, given a choice between gov bonds and the SP500, the risk minimizing portfolio is something like 90% bonds, 10% stocks. Something like 80/20 gives the same risk level as 100/0. (Exact numbers may be off, but principle is dead right.)

There is a 'moral hazard' danger, whereby people would be tempted to take on high amounts of risk, figuring that they win on the upside, but the government won't let them starve in the streets if they gamble and lose.
Barro's relatively high minimal benefit with no means testing would offer that downside payment unconditionally, thus eliminating the moral hazard. (Though it would also mean higher tax rates to fund it, thus creating different negative incentive effects.)

Mark Horn writes:
This is an odd way to think about things. When the gov't runs a surplus (or less of a deficit) then it is borrowing less.
The difference between borrowing and taxing is irrelevant to me. They're both equally bad. I would much rather see the government spending less money, thereby decreasing one or the other of borrowing or taxation.
If, for some reason, I thought project A could be done by gov't more efficiently than the private sector then it makes sense for gov't to borrow the money to do project A.
Agreed. However, that's an awfully big "if".
Of course SS is a transfer program so it isn't trying to produce anything like gov't built highways 'produce' transportation or NASA produces scientific probes.
But aren't you trying to justify SS as a mechanism to reduce gov't borrowing? In other words the gov't gets to use the SS surplus to spend on stuff. That stuff is produced. Of course, it's incredibly high cost production. Which is why I think we should limit the amount that we give to the high cost producer and favor instead efficient production. Which is, of course, another reason why SS is part of the problem: it's money that's raided by the high cost producer. The goal should be to give as little as possible to such inefficient production.
So when do people 'invest in the gov't'? When the gov't borrows more or less from the private sector?
People invest in the government whenever their money is used to fund something. Taxation is forced investment in the government. It's forced investment in an ineffecient, high cost producer. There is economic value in investing the same amount of money in a more effecient producer: the market.
Patrick writes:

Boontoon: you are comparing apples and oranges. The 50% return is a cumulative 5-day return. The 8.56% is an average 1-day return. If you compound that over 5 periods (1.0856^5 = 1.508 --> 50% return), you get the same number.

So your simple stock market is just fine, but your math is severely flawed. Sorry.

Could you apply this to the very simple stock market we depicted here? We looked at a single stock that returned 50% in a 5 day period, yet the actual return realized by investors was 8.56%.

Randy writes:

Pro: A pure welfare system would transfer money from the wealthy to the poor. The current system transfers money from the future to the present, creating an incentive for the politicians of each current generation to use the program as a patronage scheme.

Con: A pure welfare system would have free riders. But then again, the current system already has free riders, i.e., people who do not save knowing that they can count on politicians to grant generous benefits. Sorry, this isn't a con after all. A pure welfare system would actually create an enduring safety net - not a retirement plan - a safety net.

Recommendation: Separate the retirement portion of Social Security from the Welfare portion. Make the retirement portion completely voluntary. Make the welfare portion provide only minimal benefits. Private accounts are a very good first step to accomplish this.

Boonton writes:

You're treating things as if the taxes taken were invested in government bonds, earning interest rate r. But in a PAYGO plan, no assets are accumulated. The interest rate (on T-bills, corporates, or whatever) doesn't enter into it. However, even if it did, that interest rate r would be a a risky asset, so much of the following analysis would still apply.

True but in a pay-go plan the revenue grows by the growth rate of the tax base. If, for example, Social Security taxes and benefits are roughly constant at a certain % of GDP then returns will be equal to the growth rate of GDP.

The reforms of Greenspans commission during the Reagan administration actually made Social Security temporarily into a hybred type of program. The program prepared for the baby boomers by running a surplus. The effect of this was to reduce gov't borrowing from the public which generates a return in the form of reduced interest expenses. That return is positive (since interest rates have always been greater than 0) but is risky. The risk, however, is very small since gov't interest rates have nearly always been positive. Quite frankly, I don't see a better way for a gov't to prepare for a large expense expected in the future beyond borrowing today.


I agree that for the individual SS is a risky asset like all others. After all your return would be negative if you died a day before getting your first check or your return will be fantastic if you live a very long time. Obviously an individual should diversify by accumulating private savings to supplement Social Security. The overall returns in a pay-go system have to equal the rate of growth of the systems revenue. Who gets those returns is a different matter. Social Security has a very complicated way of allocating those returns (for example, to widows with children) that reflect society's political values.


But aren't you trying to justify SS as a mechanism to reduce gov't borrowing? In other words the gov't gets to use the SS surplus to spend on stuff. That stuff is produced.

No I'm pointing out that the SS surplus reduces gov't borrowing, if you propose to eliminate that surplus you propose to increase borrowing. My position is that gov't spending on 'stuff' happens with or without the SS surplus. If that is the case then reducing borrowing today has the effect of cutting gov't spending on interest which you should consider a good thing whether or not your're able to convince the gov't to cut back on all the other types of spending it does.

The odd position of the SS privitization advocates appears to be something like 'The gov't plans to spend $1.1T in 2040-20XX (discounted to present value). Therefore it makes no difference if gov't borrows $1.1T today or in 2040.'.

Boonton writes:
Recommendation: Separate the retirement portion of Social Security from the Welfare portion. Make the retirement portion completely voluntary. Make the welfare portion provide only minimal benefits. Private accounts are a very good first step to accomplish this.

Which, you could argue, is what we have. If you're happy living on just SS then don't volunteer to save anything. You're not a free rider because your benefit from SS is no higher than a person who saved everything they could.

Obviously what makes for a minimal benefit will be open to debate. Since social security and its defenders have always stated that people should not rely upon it for retirement you may want to consider the possibility that society considers the status quo to be the minimial benefits or close to it.

Randy writes:

Boonton,

Good point. Perhaps the way forward is to simply define the minimum benefit, pay no one anything but the minimum benefit, and then cut payroll taxes to a level sufficient to provide that minimum. If some chose to voluntarily contribute more (maybe someone who trusts the government more than the stock market), they could do so in private accounts.

So what would the transition look like? Interesting.

Mark Horn writes:
My position is that gov't spending on 'stuff' happens with or without the SS surplus. If that is the case then reducing borrowing today has the effect of cutting gov't spending on interest which you should consider a good thing whether or not your're able to convince the gov't to cut back on all the other types of spending it does.
What I would like is for the gov't not to have a free and easy loan against SS. I'd like the gov't to have to justify extra spending against borrowing instead of getting a free pass on spending when it raids the SS trust fund.

As far as taxing today vs. paying interest tomorrow, I'd refer you to the chapter in Landsburg's Fair Play, but I don't think that an online copy of that chapter exists. I agree with Landsburg: those things are a net wash - both are equally bad. Good is getting the gov't to stop spending money since they're so incredibly inefficient at it. A step in the right direction is to at least get them to admit that they're incurring a future debt when they keep spending as it is. That means taking away the free debt from the SS surplus. Since that means borrowing, and that's no worse than taxing today, it's a step in the right direction. Better would be to cut spending.

nmg writes:

You all seem quite eager to mess with Social Security but come up quite short on reasons why

You have got to be joking. There are so many things wrong with Social Security, fiscally, ideologically, and practically, that I have to assume you are joking.

nmg

DeadHorseBeater writes:

Unsurprisingly, poverty and financial myopia tend to be correlated.
Arguably, among it's many many purposes, SS seeks to protect financially myopic people from their myopia.
These people tend to not accumulate significant private diversification because they don't tend to save outside SS.
Thus there is a strong argument for some kind of default diversification through some kind of personal account system.
(People who are already self-saving and well-diversified would gain less by a personal account system. They might very well choose to not take advantage of the private account at all.)

Boonton writes:

Good point. Perhaps the way forward is to simply define the minimum benefit, pay no one anything but the minimum benefit, and then cut payroll taxes to a level sufficient to provide that minimum. If some chose to voluntarily contribute more (maybe someone who trusts the government more than the stock market), they could do so in private accounts.

But the system is funded by a payroll tax. Since people have different working periods and different rates of pay you will have a situation where some people will have contributed more than others.

What I would like is for the gov't not to have a free and easy loan against SS. I'd like the gov't to have to justify extra spending against borrowing instead of getting a free pass on spending when it raids the SS trust fund.

The loan is obviously not free. Anytime you use a $1 for anything you immediately incur the opportunity cost of all the other options you had for it. I think it was a good idea to reduce borrowing today to prepare for the baby boom generation.

Thus there is a strong argument for some kind of default diversification through some kind of personal account system.

Indeed, but keep in mind even being 100% 'invested' in just social security is better than having zero.

(People who are already self-saving and well-diversified would gain less by a personal account system. They might very well choose to not take advantage of the private account at all.)

Which is a good argument for a Social Security + type reform where the 2% private accounts come in on top of social security. If you're already putting money into a 401K then participating is optional. Let's not kid ourselves, though, this is paternalism. If you're going to use paternalism as a charge against social security then it is equally valid against so-called 'private accounts' that are really mandated by the gov't.

Randy writes:

Boonton,

Re; But the system is funded by a payroll tax. Since people have different working periods and different rates of pay you will have a situation where some people will have contributed more than others.

The idea is to transition from the current system to a pure welfare system - to turn it into a safety net, and only a safety net. The benefit would be set along the same lines as current welfare standards. The qualifications would be that the individual is above the standard retirement age with income below the poverty line. It would be a totally progressive tax so some would indeed pay in more than others. But the gain would be that the total tax paid by most workers would be considerably less than under the current system, allowing them to invest (or spend) the difference as they see fit.

As for the transition, it seems to me that all we have to do is gradually reduce benefits from the top down. At some point the minimal goal would be reached and the system would start producing a surplus, which could be applied to tax relief.

Boonton writes:

I would support letting Social Security peak at about 5% of GDP and let it move (as it naturally will) to a point where benefits roughly equal revenue. This may involve partially de-indexing initial benefits to wages after 2040 or s or gradually increasing the retirement age to pace increasing lifespans.

What I would do is set up an independent board similar to the Federal Reserve. They would have the power to make minor tweaks as needed to the system with major changes requiring Congressional legislation.

The system is fine but the structure of the benefits requires some modernization. For example, after ten years of marriage you get your spouses benefits. What I would do is have married couples automatically split their accumulating benefits as long as they are married. Social Security provides the best benefits to non-working spouses (aka the stay at home mom). Social conservatives may want to rethink a reform that provides another disincentive to the traditional family.

Bob writes:

Boonton,

"Could you apply this to the very simple stock market we depicted here? We looked at a single stock that returned 50% in a 5 day period, yet the actual return realized by investors was 8.56%."

It's just the difference between single period returns and compounded returns: 1.0856^5=1.5 . An investor that holds for the five days actually earns an 8.56% daily return, just like the five investors that trade. Although it is common to just divide by five and talk about a 10% daily return, it is incorrect to do so (in this case the error is material because the example has such high returns).

But the simple market makes a useful point about market returns nonetheless. Too often historical returns are reported as arithmetic averages, which generally overstate the true (geometric) return. Suppose the market yields 20% (arithmetic) over two years. It matters whether that is 10% each year, zero one year and up 20% the next, down 20% one year and up 40% the next, etc. The more volitile the market, the lower the actual (compounded) return to investors, which plays into your concerns.

Bob writes:

Sorry Patrick, I missed your post on compounding returns.

Boonton writes:

Thanks Rob & Patrick for correcting my error on returns. The idea I wanted to look at involved the difference between a percentage return and the fact that the actual hard dollar return in the example was $5.

If the economy permitted this company to issue a $5 return in just 5 days (this could be done thru dividends or an increase in the expected present value of its future income) then that is what is in the pot. If the number of investors double on the first day they have to split a $5 return, in other words they won't both earn 8.56% per day.

For an individual it is perfectly valid to say "gee, if only I had brought an S&P index fund ten years ago I would have made 13% returns!" because a single individual investor has no real effect on market prices. It is not valid to say "gee if we had put 2% of GDP in the S&P 500 ten years ago we would have made 13%". Such large changes would alter the dynamics of the market and alter the returns....if we had a time machine and could rerun the tape.

Randy writes:

Boonton,

From the polling results I have seen, the only fix that a majority of voters is willing to accept is reducing benefits to, or increasing taxes on, upper income recipients. Assuming that the politicians will respect the wishes of their constituents, the idea of gradually turning the current system into a progressive welfare system is precisely where we are headed.

As for benefits to traditional non-working spouses, these have always fallen into the category of welfare recipients, as they have never paid a dime into the system and yet they recieve benefits. So a change to a pure welfare system would not change their status.

Again, I am not recommending the creation of a pure welfare system unless it is accompanied by a decrease in payroll taxes. The idea is to tax to a level necessary to provide a true "safety net", but allow people to make their own decisions beyond that point. A responsible husband with a non-working spouse would certainly want to see that his surviving spouse were taken care of. Tax relief from a change to the system could be used to buy insurance, an IRA, or invested in some form of government sponsored retirement program (private spousal account?). And a responsible non-working spouse would take care to see that something was done. But worst case, she would still fall under the safety net.

Boonton writes:
As for benefits to traditional non-working spouses, these have always fallen into the category of welfare recipients, as they have never paid a dime into the system and yet they recieve benefits. So a change to a pure welfare system would not change their status.

This is an example of how it is sometimes more tricky to get an accurate picture of the true return on Social Security contributions for an individual. For many people, the above would not be viewed as welfare. They would view it as a benefit that their wife would be at least partially 'taken care of' in retirement and so would their minor children.

Of course this is in the eye of the beholder. The man who was married to his wife for 20 years but then got divorced in a particularily bitter and contested trial might not take as charitable a view of her being able to collect benefits off of him.

John Thacker writes:

nmg--

The problem with a pure welfare approach is that it's a big disincentive for the elderly to work (especially relaxing, lower wage jobs) and get their benefits cut. Hence the "minimum benefit to all" plan.

Randy writes:

John,

Why do we want the elderly to work? A pure welfare program would not become available until age 67 or so, and the assumption would have to be that people this age are too old to work.

It is interesting that you bring this up though, because I happen to believe that a primary purpose of the current system is to keep people from retiring. If I had been able to invest my own money instead of giving it to the government, I could have easily had enough to retire in my 50's, as could nearly anyone in the top half of those paying into the system. By taking the investment money away from this group, the government is able to keep them in the labor force for an additional 10 years or so - paying taxes.

Comments for this entry have been closed
Return to top