« Why Oh Why Can't We Have a Better Press Corps? (Yet Another Social Security Edition) | Main | Turning Up the Heat »

December 24, 2004

Jeff Madrick on Social Security

Jeff Madrick gives us his bottom line on Social Security:

The New York Times > Business > Economic Scene: The Bottom Line on Overhauling Social Security: Investing part of the payroll taxes we pay in stocks and bonds to produce a personal nest egg is alluring. But until there is a detailed plan to analyze, we cannot know the true consequences. At last week's economic summit meeting, the Bush administration began to give clues as to its plan. The administration appears to be leaning toward the Reform Model 2 proposed in December 2001 by its Commission to Strengthen Social Security, which was under the direction of former Senator Daniel Patrick Moynihan and Richard D. Parsons, the chief executive of Time Warner. The plan is disheartening....

The Bush privatization plan coming into view would work as follows. Workers now pay to Social Security 6.25 percent of their wages up to $87,900, matched equally by their employer. Workers would be allowed to divert four percentage points, up to a maximum of $1,000 a year, to private investments in stocks and bonds. The investment accounts would be limited to highly diversified mutual funds, or even index funds, and the transactions costs would be kept to a minimal 0.3 percent.... Social Security benefits guaranteed under the present system would be cut slightly for each dollar the individual worker diverts to his or her private accounts. More important, future benefits would also be cut by indexing them to the rise in consumer prices rather than, as is now done, to rising wages....

Will the magic of private investment accounts make up for the reductions? The answer is no. We can compute how retirees fare if they earned the historical average of 4.6 percent a year (after transaction costs) on a portfolio of stocks and bonds. William Dudley, chief economist of Goldman Sachs, has calculated that benefits for a typical retired one-earner family would come to about 93 percent of the projected benefits from the present Social Security system in 2022. In 2075, the benefits would fall to only 77 percent of present-system benefits....

[B]y no means is everyone going to earn the 4.6 percent average. The history of private investment unmistakably shows that there is a wide dispersion of results. Some will invest only in bonds, others may choose the wrong mutual fund, or switch from one to another at exactly the wrong times. And some will have the bad luck of retiring in the midst of a bear market. Many workers will therefore inevitably earn less, and some considerably less.... Mr. Dudley worked out some calculations under Reform Model 2. The results are stunning. If a worker earns just the respectable expected bond rate of 3 percent a year, or 2.7 percent after transactions costs, then the typical one-earner family will retire on only about 58 percent of the projected benefits under current law. If the investor earns zero over time, which may well occur for some investors, the projected retirement benefit is only a little more than 38 percent of the current benefit. These are considerably worse than the projected adjustments needed to bring the present system into balance....

Consider the potential impact from borrowing as much as $2 trillion... the projected transition costs.... [T]he ultimate decision Americans will have to make is over the purpose of the public pension system. Social Security was originally intended to guarantee a minimum retirement income for workers. For all the alarm over Social Security, the present guaranteed system can be fixed with only moderate tax increases and through less harsh benefit reductions, particularly if Americans are prepared to raise the official retirement age slightly.... The privatization plan the Bush administration is leaning toward, in contrast, will divide people into winners and losers. It may make some workers better off in retirement, and may well reduce costs somewhat to government, if all goes well. But a significant number of American retirees will do poorly.

Posted by DeLong at December 24, 2004 09:43 AM


Sadly, as proven by this past election, many many Americans just won't be interested in the details (the facts) concerning Bush's plan for SS. Bush has already announced his intention to conduct a media campaign and if nothing else he is a good snake oil salesman. I think it's time for the Dems to start asking the American people if they can trust Bush to do the right thing with their future. He is already creating considerable pain for those bearing the brunt of his incompetence in Iraq with no end in sight for that debacle. Why should anyone be any more confident about entrusting their retirement security to him. The Dems should be suggesting that even if this was a good idea, given Bush's track record, are these really the people we would want implementing it. Taking a page out of the Repugs strategy book the Dems should spare no effort in reminding everyone about Bush's close ties with Ken Lay and how things worked out for the retirement plans of Enron employees. One crucial element is going to be whether or not the media gets off their knees and starts doing some real reporting once the Bush sales campaign kicks off.

Posted by: Dubblblind at December 24, 2004 10:19 AM

Someone should point out to Mr. Madrick that a 0.3 percentage point management fee is hardly "minimal." That's about twice the current expense ratio of Vanguard's S&P 500 index fund, and a third higher than the fee on Vanguard's balanced (stocks and bonds) index fund. Also, that 0.3 points doesn't include trading commissions, bid-ask spreads, account maintainance fees, and all the other little ways that Wall Street nickels and dimes its customers. Add those, and costs could easy be a half a percentage point and probably more.

That still may not sound like much, but consider the compounding effect over a 40 or 50 year acccumulation period.

A simple example: Imagine $100,000 invested at the 4.6% "average" stock-bond return cited by Madrick. After 40 years, that would grow to about $604,000. Now imagine that same 100k invested at a 4.1% return. After 40 years, it would be $499,000 - almost a 20% reduction in total return.

Apply that math to $3 or $4 trillion (or whatever)in privatized assets, and you can easily see why Wall Street pumped so much money into Shrub's last campaign.

Is this a great country, or what?

Posted by: billmon at December 24, 2004 10:52 AM


Posted by: pragmatic_realist at December 24, 2004 11:31 AM

I think we should demand that any proposal they put forward also have a budget for providing for privatization's losers - the ones who lose money on stocks and don't have enough to live on.

The US government is not going to let them starve, and the still-guaranteed part of the SS benefits (if any) won't be enough. So the government WILL wind up paying something to them. (Perhaps it could show up as a budget for quelling the resulting riots if they don't.) This should be reflected in the budget projections of any privatization plan.

Posted by: Dave Johnson at December 24, 2004 11:39 AM

Ilia Dichev of the U of Michigan has done research that suggests that real live investors historically earn significantly less than the buy-and-hold returns cited in support of privatization. You can find an early version of his paper at:

www.usc.edu/schools/business/ FBE/FEA2004/FEApapers/J-SA_IDICHEV.pdf

Posted by: Bernard Yomtov at December 24, 2004 11:40 AM

The political problem you face is captured perfectly in the title of George Carlin's book: "When will Jesus bring the pork chops?" The American small-d democrat is always prey to the pathetic illusion of becoming a petty bourgeois giant. Something or somebody or Jesus will, they hope, transform their payday-to- payday condition into marvellous prosperity. Look at the lottery ticket buyers. Bush pretends to be both Jesus and Carnegie.

Posted by: g-lex at December 24, 2004 02:27 PM

More important indexing accounts Dudley considerably reduce all benefits.

Posted by: cloquet at December 24, 2004 07:33 PM

** Someone should point out to Mr. Madrick that a 0.3 percentage point management fee is hardly "minimal." ... Imagine $100,000 invested at the 4.6% "average" stock-bond return cited by Madrick. After 40 years, that would grow to about $604,000. Now imagine that same 100k invested at a 4.1% return. After 40 years, it would be $499,000 - almost a 20% reduction in total return.**

Which leaves a mere gain of $399,000, or 399% on investment.

Now *let's compare* to the status quo, by looking at real numbers from the Social Security Administration's Actuaries for the current value of taxes paid and benefits earned by average-wage workers who entered the work force in 1994, and who thus are in their 30s today.

average income male ...
tax: $86,720
benefits: $62,889
return: - 16%

average income female...
tax: $90,620
benefits: $76,185
return: -15%

(These figures also apply to married individuals where both spouses work and collect benefits on their own earnings records.)

And note that because these benefits are 30% underfunded, if SS remain paygo then by the Iron Laws of Arithmetic these return drop from -15%/-16% to about -40%.

Hey, how, ... um .. confused is it to condemn private accounts for having a 0.3% fee that drops return on investment to only + 399%, while endorsing instead a status quo that gives a *negative* return of -15% to -40% over 40 years? Hello??

Look -- No defender of the status quo can honestly criticize private accounts for having fees that might reduce positive returns, without mentioning the negative returns *guaranteed* by the status quo. It is totally just, well, "disingenuous" is a polite word.

The SS Actuaries say that every annual cohort retiring after 2000 will receive back from SS less than they paid in through taxes -- it's negative returns for all from now on, getting worse and worse for the young.

This is of course the dramatic, historic change in the nature of SS -- which always provided guaranteed above-market returns in the past -- that no Democrat defending the status quo ever dares mention.

Yes, markets have some risk -- but to *guarantee* a negative return over 40 years takes a government, no market can do that.

Posted by: Jim Glass at December 25, 2004 01:29 PM

Jim, a link or two to the sources of those numbers from the Actuaries would be handy. As well as some explicit recognition of the economic projection that underlies them, particularly the 30% gap. Because I am browsing through the tables and the only place I am finding that number is in
Table VI.F3.--Summarized OASDI and HI Income Rates, Cost Rates, 
and Balances  for 25-Year Subperiods,1 Calendar Years 2004-78

And that -29.90 derives from the High Cost estimate, and an examination of the economic numbers underlying that show numbers so dismal that the future of the country is at risk: 1.2% for 2004, 2.6 for 2005, 1.7% for 2006, .1% for 2007, and 1.3% in the out years after 2013.
Economic Assumptions under the Three Alternatives.

This topic has endured too much selective reporting over the years and now over the last few weeks. Vague references to the "Actuaries" or the "Trustees" are no longer pulling much weight.

Because I look at this graph, and then consider the very reachable numbers that produce outcome ( I ), I am not seeing any 30% gap.
http://www.ssa.gov/OACT/TR/TR04/II_project.html#wp106217 Trust Fund Ratios

Tell me you are not using the High Cost Alternative to produce your numbers. And while you are at it give your best estimate of the impact of 4.0% growth in 2004 on models that had it at 1.2% (High) 2.7% (Intermediate) and 2.8% (Low) Cost Alternatives.

Because we have a dynamic system here, one extremely sensitive to changes in the front end. Simply assuming we should be making policy choices relying on a static number that the real world has left behind is not serving us well here.

Posted by: Bruce Webb at December 25, 2004 02:10 PM

The president proposes borrowing money from the Chinese and giving it to the young to invest in the stock market. You don't need to run the numbers on this, just consider: if this would work, we would all be billionaires. The government could borrow, invest directly in the market, and give each of us a million dollars a day, tax-free.

But -- you don't need to wait for the government to get around to this. Just take out a HELOC, invest in the market now, and retire in a few years with millions.

In fact, it's odd that some corporation hasn't discovered this already. Issue corporate bonds, invest the proceeds in the market, and watch your profits grow without bound, to trillions of dollars a second.

Only a great mind like that of George Bush could grasp an idea this deep.

Posted by: Jim Smith at December 25, 2004 06:44 PM

"The president proposes borrowing money from the Chinese and giving it to the young to invest in the stock market. You don't need to run the numbers on this,"

I stipulate for this discussion, only, that such is Shrub's plan.

What's wrong with that? I mean, the Chinese have a growing economy with one major drawback -- a centrally planned and managed market.

If the Chinese central bank loans the money to INDIVIDUALS - Americans, in this case, since the Chinese planners don't trust their own citizens -- to make investment decisions, the decisions will be better than those of the central bankers. (If we don't believe this, why stay in a free-market society?)

Because the Chinese economy is growing, (if this were not true, how would they have excess to lend?) at least some of the money the Americans invest will be in Chinese enterprises. Most likely, the ones the Chinese central planners undervalue -- again, if the planners happened to value it fairly it would not appear as a bargain to American investors.

So, what's the net? The Chinese get extra investment for their most promising projects. The American investors profit beyond what they'd get in purely Treasury bonds -- or purely domestic investments. Win win!

And, as a social side effect, we liberalize China.

What's so "evil" about the Shrub's plan -- as asserted above -- from a liberal social democrat economic perspective?

Better yet, how does this compare to the "do nothing" plan?

Now, just for full disclosure -- I like the John Anderson proposal of the 1980 race: Raise gas taxes by 50 cents a gallon and apply the funds collected toward retirements. I point out that Jimmy Carter lost against a DIVIDED Republican SS-reform field -- Carter on the "do nothing" side and Reagan and Anderson both promising "do something".

If the Democratic party wants to continue advocating "do nothing" -- and making that a cornerstone of their platform during elections -- be my guest. Be sure to tell me how smart this is.

Posted by: Pouncer at December 26, 2004 06:35 AM

"More important, future benefits would also be cut by indexing them to the rise in consumer prices rather than, as is now done, to rising wages..."

I am wondering if the assumption here is correct, or if there are realistic scenarios where this change could turn out to have the opposite of the intended effect. Possibly: an overall inflationary environment with falling real wages that continues for a lengthy period of time?

More specificlly: what effect will oil prices and energy availability have on the above scenario? Large increases in oil prices could easily produce a situation where inflation outstrips wages.

Several well-known oil geologists contend that oil production worldwide cannot increase significantly above the current level (so-called "peak oil"), and will start to fall within the next few years, leading to astronomical (& inflationary) price increases. A lot of oil analysts seem to focus on about 2010-2012 as the time frame for the oil peak, and the rosiest informed prediction I've seen is for the peak to occur around 2037. Even 2037 is well within the time frame we are considering for Social Security "reform," so it seems to make sense to attempt to address this before assuming that switching to price indexing from wage indexing will save any money.

As an aside, what would be the economic effect of negative growth in the amount of energy actually available? We have never been in a situation where the actual amount of energy available to humanity as a whole has dropped, but that is the logical consequence of an early date to "peak oil." There aren't, at present, any realistic substitutes for oil. North American natural gas production is already falling rapidy, and the shipping fleet required to import significant amounts of it doean't exist and won't for at least a decade as I recall. Significant increases in wind, solar, and nuclear power are also decades away. Energy- in some form- is *required* to do _anything_. Is there any scenario where we can see significant economic growth during a period of real decline in available energy? And if not, is there any reason to assume economic growth when looking forwards at Social Security?

Posted by: JohnDL at December 26, 2004 09:55 AM