Arnold Kling  

More on Privatization

PRINT
Government Mergers... Friedman on the Battle of Idea...

Paul Krugman makes a generous concession to privatization supporters.


For the record, I don't think giving financial corporations a huge windfall is the main motive for privatization; it's mostly an ideological thing.

I don't recall MIT having a course in the economics graduate program on discerning the true motives of one's opponents. But most of the information that Paul imparts in his columns seems to be based on that skill set.

The question that I have for Krugman and other defenders of the status quo is this: how much of an increase in economic growth must privatization stimulate in order to make it worthwhile? What I thought that Kinsley was saying was that unless economic growth increases by enough to eliminate all of the unfunded liability of Social Security, then privatization does not "work." To me, that sets the bar awfully high--in effect, you're saying that unless privatization provides umpteen trillion in higher real GDP, you'd rather get nothing. But perhaps I misread Kinsley.

Some other posts on the topic are by Nouriel Roubini, by Flit, and by Tim Worstall (and another one here)

If I were given my choice of one policy change in Social Security, I would rather fall on my sword for a higher retirement age than for privatization. But I do believe that privatization would raise economic growth. Moreover, privatization likely would make Social Security less regressive. I would really like to know what is so precious about the status quo.

UPDATE: Business Week's Michael Mandel thinks the way I do about the progressivity of privatization (and he cites me to boot). He also is the author of the book Rational Exuberance, which happens to be the title of a chapter in my own book.

For Discussion. How much economic growth would a reasonable person be willing to give up in order to maintain Social Security as it is without privatization?


Comments and Sharing


CATEGORIES: Social Security



COMMENTS (24 to date)
Boonton writes:

Arnold, I noticed you neglected to address any of the substance of Krugmans article. For example;

Privatizers who laud the Chilean system never mention that it has yet to deliver on its promise to reduce government spending. More than 20 years after the system was created, the government is still pouring in money. Why? Because, as a Federal Reserve study puts it, the Chilean government must "provide subsidies for workers failing to accumulate enough capital to provide a minimum pension." In other words, privatization would have condemned many retirees to dire poverty, and the government stepped back in to save them.

The same thing is happening in Britain. Its Pensions Commission warns that those who think Mrs. Thatcher's privatization solved the pension problem are living in a "fool's paradise." A lot of additional government spending will be required to avoid the return of widespread poverty among the elderly - a problem that Britain, like the U.S., thought it had solved.

Boonton writes:
For Discussion. How much economic growth would a reasonable person be willing to give up in order to maintain Social Security as it is without privatization?

That question is pretty stunning in its arrogance. You basically assume your proposal will generate economic growth and ask how much of that growth are we willing to pay to avoid implementing your proposal?!

How about looking at the question of why privatization will generate economic growth in the first place?

The question should be, compared to what? Social Security isn't a particularly efficient system for what it does. The CBO finds that money market mutual funds have fees as low as 0.36% of assets managed. The federal employees' Thrift Benefit Plan costs about $ 25 per year per participant, and it has a variety of investments available.

The system that today Krugman claims "works" was described in Fuzzy Math (2001) as $10 trillion in debt, and a staggering burden. He derided the "trust fund" as the government taking money out its left pocket to put into its right pocket.

Hunter McDaniel writes:

I would really like to know what is so precious about the status quo.

The status quo makes retirees depened on the good graces of politicians and government to receive their benefits and "protect" them from cuts. That is a position of power that the politicians have difficulty letting go of.

Boonton writes:
The question should be, compared to what? Social Security isn't a particularly efficient system for what it does. The CBO finds that money market mutual funds have fees as low as 0.36% of assets managed. The federal employees' Thrift Benefit Plan costs about $ 25 per year per participant, and it has a variety of investments available.

You underestimate the financial industries ability to charge fees. There are a few mutual funds with fees below a half a percent but there's more than a few with fees far above. A 1% management fee eats up returns like a cancer. Consider a $100 investment that would have earned 4%.

Year 1 Investment $100 - $1 fee = $99
Year 2 $99 + 4% = 102.96 - $1.0296 fee = $101.93.

A 4% return has turned into barely a 2% return. I highly doubt the Thrift plan's total fees are only $25. Perhaps that is just for the plan itself but it certainly doesn't cover the fees charged by the 'variety of investments available' which almost certainly come with a variety of fees poorly disclosed and largly hidden from any but the most observant.

Krugman's article hit the nail on the head. If you're going to cut down on the massive cost of fees charged by the finance industry then you have to opt to put all the money in a low cost index fund. Considering how large this amount of money is it would be more sensible for the gov't to simply create its own index and pool all the money together.

But then why would that be any different than simply decreasing its debt burden ala the Trust Fund surplus?

Jim Glass writes:
"You underestimate the financial industries ability to charge fees."

Gee, you should try to imagine the fees *I'd* charge -- if I didn't work in a competitive market. Now, as to the real world ...

Krugman forgot to mention the actual working example of the US The Federal Thrift Savings Plan, which manages private accounts holding real investments with a 0.1% expense rate -- hey, it's good enough for government employees!

He also forgot to mention that the actual average real-world administrative cost for private-sector, multi-employer defined contribution plans is 0.8% -- hey, that's less for running real retirement plans than Social Security charges for running a checkbook!

He also forgot to mention the experiences of the other 20-odd countries that now have personal accounts -- including how Sweden is very happy with its 2.5% personal accounts.

Wait -- that would make it apparent that Swedish social policy is too right-wing for US Democrats. So maybe he didn't mention that on purpose.

As to one of the two countries with plans he did mention as an example, a resident of one right now is asking, "Paul Krugman, ignorant or liar?"

And this is the dangdest thing: While Krugman is so concerned that with private accounts expenses might reduce returns from 4% to 3%, he overlooked entirely the fact that under the status quo returns are guaranteed negative.

For a guy who's so concerned about returns on contributions to SS participants, how curious is that?

He forgot to mention some other things too.

Boonton writes:

Jim,

Thanks for reminding us that the fees just to manage the private accounts can tip close to 1%, see my example of how a 1% fee can turn a 4% 'newspaper return' into a less than 2% realized return. Those are the fees before you actually invest your private account in anything. Once you get beyond those fees mutual funds add in their own fees that can be quite low for an index fund (still close to 0.50%) but are often higher than 1% or even beyond.

I suggest you review John Boogle's "Common Sense on Mutual Funds" for a more in depth review.

Boonton writes:
And this is the dangdest thing: While Krugman is so concerned that with private accounts expenses might reduce returns from 4% to 3%, he overlooked entirely the fact that under the status quo returns are guaranteed negative.

Jim,

We had a long discussion over this on http://www.janegalt.net/blog/archives/005063.html You never addressed the issue of calculating the cost basis of the SS tax. If, as there is good reason to believe, employees do not really pay 100% of the SS tax then your estimate for SS benefits returns are most likely way off. In particular the posts right around http://www.janegalt.net/blog/archives/005063.html#40187 demonstrate how sensitive return analysis can be if your costs are off even slightly.

Paul Zrimsek writes:

The only thing you'll learn from looking at Boonton's example is that a "1%" fee works out to 2% if you do as he did and charge it twice in the space of one year. I trust it's obvious to everyone else that a 1% fee would reduce an original 4% return to 3%, no worse.

Boonton writes:

Good point Paul. So a 1% fee would only be like a 25% tax on a 4% return, not a 50% tax. Of course we are just getting started with that 1% fee. Once your money is in the private account you have to buy a mutual fund which will have a management fee of anywhere from 0.30% to more than 1.5%. Add to that churning, the cost of the actual buying and selling the fund engages in.

If all goes well you looking at maybe 1.3% if you mandate low cost index funds as acceptable investments for the SS accounts (BTW, this will almost certainly be fought by the financial industry as 'price controls'). A lot has been written about mutual funds but oddly competition doesn't seem to have done the magic you would expect it to do. In fact, its a bit of a mystery why mutual funds are able to charge such fees at all considering the more 'managed' a fund is the more likely its return is expected to be inferior to the market return.

Edge writes:

To get growth with reasonable risk, we'd need to increase savings.

Having the government borrow the transition costs - $1 or $2 Trillion for 10 years, maybe $5 Trillion for 30 years - it's questionable whether we'd get growth, and it's certain that we'd increase risk, at least to the US government.

Going with a model like the federal thrift savings plans would probably be necessary to keep fees manageable.

But the thrift plans are run by the government, and that's why the costs are so low. They have very low overhead, limited product selection, limited marketing costs, and limits on access to the money.

For these tiny accounts, about the only way to keep fees low enough as a fraction of account value to match treasury bond yields is with a plan like the thrift plan. But it's much more limited and restricted than any private mutual fund company has ever offered, or is ever likely to offer.

As for the repeated claim that SS doesn't even match treasury bond yields - it certainly matches treasury bond yields on the value that is in the trust. The problem is, the amount in the trust isn't enough.

Moving funds out of the trust and into private accounts could give the giftees a "market return", but they' do so by shifting the burden onto everyone else to make up the difference. At $2 Trillion in 10 years, each family of four would have to pay $1000 a year just in interest costs to support the $2 Trillion put into private accounts.

That's hardly a recipe for increasing the savings rate. Well, maybe it wouldn't be. Taking $1000 a year from every family would certainly damp out consumption a bit.

The other question is whether we really want to load up the treasury with publicly held debt. Treating the US treasury like junk debt would almost certainly raise interest rates long into the future, and that would depress lots of things we don't want to depress, including equity prices, home prices, and employment, while increasing costs for other things, like mortgages and car loans.

It might work if we could keep running an external deficit around 6% of GDP for another 30 years. But I think eventually the drag of having our economy paying something like 15% of GDP annually to support interest payments to overseas central banks wouldn't be sustainable.

You write: "I don't recall MIT having a course in the economics graduate program on discerning the true motives of one's opponents. But most of the information that Paul imparts in his columns seems to be based on that skill set." Nice line! It's beyond me how Krugman imagines he's going to win more than a few people over using the kind of aggression, contempt, ego, name-calling that he does. Most people who aren't already on his side, it seems to me, are going to look at his displays, think "What a jerk," and rejoin the other team.

But maybe he's just trying to fire up the loyalists.

Nigel Kearney writes:

If the government didn't take so much of my money I would use it to repay my mortgage. Since I can't do that, a large financial corporation (i.e. my bank) makes lots of money by charging me interest.

But I don't think giving financial corporations a huge windfall is the main motive for taxpayer-funded superannuation; it's mostly an ideological thing.

Lawrance George Lux writes:

How much economic growth would a reasonable person be willing to give up in order to maintain Social Security as it is without privatization?

Fine partisan question. I pose another:

How much added taxation would Privateers accept to fund Privatization without unfunded additional liability? Would They accept another 3.25% increase in the FICA taxation?

There are a number of other questions:

How much Capital construction will be lost because of lack of funds due to funding the increased Debt?

How much will SS monthly benefits be reduced due to paying the Interest on the increased Debt?

How many projected Private Accounts will have to be subsidized by the SS Fund, due to the failure to maintain the minimum funding to provide the Elderly with an adequate monthly income? (There have been loose projection of up to 64%.)

These present some of the problems(stylistic or real) which Private SS Accounts can provide. lgl

Jim Glass writes:
Jim,

We had a long discussion over this on http://www.janegalt.net/blog/archives/005063.html You never addressed the issue of calculating the cost basis of the SS tax. If, as there is good reason to believe, employees do not really pay 100% of the SS tax ...

And as I pointed out to you then, if you want to insist on saying the Social Security Adminstration's and Advisory Boards' economists and actuaries are all wrong about this, without bothering to explain their error, that's fine, it is a free country.

But then, as I also said before, you still have to admit that somebody has paid and is paying the full 100% cost of the benefits promised after 2000.

And with the SSA actuaries saying full return on this cost will be negative for all annual cohorts retiring after 2000, as low as 50%-- compared to a pre-2000 record when total return was very high, above market -- then one simply can't deny that the status quo itself is engineered to deliver plunging, negative returns on the total, 100% cost basis of contributions.

And being that in this column Krugman poses as being so concerned about the rate of return on contributions, it's kind of odd he didn't mention that.

No? ;-)

Jim Glass writes:
It's beyond me how Krugman imagines he's going to win more than a few people over using the kind of aggression, contempt, ego, name-calling that he does. Most people who aren't already on his side, it seems to me, are going to look at his displays, think "What a jerk," and rejoin the other team.

But maybe he's just trying to fire up the loyalists.

Well ... it certainly worked for Howard Dean!

Boonton writes:
It's beyond me how Krugman imagines he's going to win more than a few people over using the kind of aggression, contempt, ego, name-calling that he does..

Interesting, the substance of Krugman's columns are consistently ignored here and on Jane Galt's site yet both feel the need to constantly quote him and use him as a focus of discussion. Krugman made two very potent points in his column; one about the massive amount of money the finance industry stands to gain in 'management fees' and the other about how countries that tried privitization of SS have ended up with increasing poverty among the eldery AND continued gov't subsidy.

I'm sorry you get so hurt by Krugman noting what the finance industry stands to gain. Do you really think the industry is ignorant of the money they can make? It's interesting how no one got upset in the pre-NYT days when Krugman attacked protectionism by noting how it was a policy that helped special interests at the expense of everyone else.

Boonton writes:
And as I pointed out to you then, if you want to insist on saying the Social Security Adminstration's and Advisory Boards' economists and actuaries are all wrong about this, without bothering to explain their error, that's fine, it is a free country.

Wrong about an academic point that they have no direct stake in. As we explored over on Jane's site, figuring out how much an employer is paying versus an employee is anything but easy. If you missed that discussion I've already provided you with the link to update yourself.

But then, as I also said before, you still have to admit that somebody has paid and is paying the full 100% cost of the benefits promised after 2000.

Which has nothing to do with returns. When I buy a bond someone has to make the interest and principal payments. If you require me to account for their cost in estimating my return then all investments will have zero return by definition.

And being that in this column Krugman poses as being so concerned about the rate of return on contributions,

More to the point he is noting that projections of returns being presented are deeply flawed. Connected to that point is the additional costs imposed on taxpayers in countries that have tried this sort of thing that he cited. That, naturally, should be deducted from any estimate of private returns.

Brad Hutchings writes:

For those of you over 50 today. It comes down to one thing... If, during your retirement when you are drawing Social Security, the payroll tax hits 20%, you're in trouble, because we younger folks will scrap the system. I'm with Arnold on which fixes to shoot an arrow through my foot for... (1) raise the dependency age steeply and quickly, then index it to life expectancy. (2) phase in means testing.

On the politics of this... Arnold's question gets precisely to that point -- that the opposition to reform is unreasonable. Not only that, but (scroll up to arguments here) they are quick to demagogue privatization as a payoff to Wall Street (you know Bush's friends like Bob Ruben). I think we could call and raise the bet... "This is Miguel Fernandez. He is 35 years old and started a home remodelling business in 2001. He effectively pays 15.3% of his own earnings to Social Security and Medicare. This is John and Elizabeth McPeak. They are retired in a 3000 sq foot home and are having Miguel remodel the kitchen this month. Elizabeth is driving her Mercedes to the salon to have her nails done today, while John plays a round of golf. Next week, they are taking their motorhome to Colorado for a one month vacation. The McPeaks collect over $3000/month from Social Security. It's time to means test Social Security so Miguel can share more of what he earns with his own family."

"Krugman made two very potent points in his column; one about the massive amount of money the finance industry stands to gain in 'management fees'..."

And, he is wrong about this. Especially if the privatization is held down to a measly 2% of payroll and invested in index funds.

" and the other about how countries that tried privitization of SS have ended up with increasing poverty among the eldery AND continued gov't subsidy."

And, he is also wrong about this. Krugman is either the laziest economist publishing, or the most dishonest.

Again, you fail to face the real question; compared to what? The current system has raised taxes more than sixfold over six decades, and still can't pay promised benefits.

Boonton writes:
For those of you over 50 today. It comes down to one thing... If, during your retirement when you are drawing Social Security, the payroll tax hits 20%, you're in trouble, because we younger folks will scrap the system. I'm with Arnold on which fixes to shoot an arrow through my foot for... (1) raise the dependency age steeply and quickly, then index it to life expectancy. (2) phase in means testing.

Over on Jane's site, I did up a little spreadsheet. It assumed that a single retired person was getting $25K per year and computed how much this would be per worker. I computed a series starting at 40 workers to 1 retired all the way down to 1 worker to 1 retired person.

Here is the bottom of the table:
Wkr.....Cost per wkr....Increase........% increase
6.......$4,166.67.......$595.24.........16.67%
5.......$5,000.00.......$833.33.........20.00%
4.......$6,250.00.......$1,250.00.......25.00%
3.......$8,333.33.......$2,083.33.......33.33%
2.......$12,500.00......$4,166.67.......50.00%
1.......$25,000.00......$12,500.00......100.00%

Clearly if you have one worker per retired person and a retired person gets $25K per year all of that will be funded by the one worker. What's interesting is the most painful increases come near the bottom of the scale. Going from 3 workers per retired to 2 workers retuires a 50% tax increase, the next step a 100% tax increase.

But what if you reverse the scale. What if you moved in the opposite direction? Going from 2 workers to 4 allows you to cut the tax per worker from $4,166.67 to $1,250. A decrease of 70%. However if you go to 6 workers the cut from 2 is $3,571.43. Adding two more only lowered your cost by another $654.76. The total % decrease is 86%.

The lesson is that you can achieve a huge amount of savings by simply modestly increasing the retirement age and you do not have to turn the system back to the days of 40 workers to 1 retired person.

Boonton writes:
And, he is wrong about this. Especially if the privatization is held down to a measly 2% of payroll and invested in index funds.

He addressed that too, he said that may indeed keep the fees somewhat managable but that runs right into what the privitization advocates are pitching as about 'choice'. If you think about it this is a sticky problem. Even a broad stock market index fund has the risk that stocks have as an asset. It's not a good idea to have all your retirement account in stocks if you are getting very close to retirement age. Many will tell you to slowly increase your bond holdings as you get closer to the period where you will need the money.

So what's the answer? Have a few gov't approved funds of low cost stock indexes, bonds indexes, fixed income etc.? Or have all the funds flow into one account which the gov't will buy a 'world stock index' like is advocated in _The Coming Storm_?

And, he is also wrong about this. Krugman is either the laziest economist publishing, or the most dishonest.

Care to show us how or shall we just take your word for it?

Again, you fail to face the real question; compared to what? The current system has raised taxes more than sixfold over six decades, and still can't pay promised benefits.

Really? I'm unaware of anyone who hasn't received their promised benefits.

Perhaps a picture will help?

Edge writes:

On the question: would borrowing and putting the liabilty on general revenues rather than FICA taxes be "progressive".

Given that the income tax base has been steadily shrinking, so that the same people pay income taxes as pay FICA taxes. Pretty soon, the same two people who support each one retiree's Social Security benefits may also be paying for all of government obligations.

"Tax reform" is as likely as not to further flatten tax rates, so whatever progressivity there is in the income tax code will be further eroded. Or, maybe we switch to a VAT tax to fund everything. That would at least broaden the tax base so someone other than the two workers would be paying it, but it would be similarly regressive as payroll taxes.

Comments for this entry have been closed
Return to top