Tuesday, July 22, 2008

What's really going on with Social Security.

This started out as a comment on Andrew Biggs' blog on Social Security, but grew too long to fit in the comment section, I felt, so here it is (where it grew longer yet!)

I believe it may have some general value in clarifying the basic issues confronting Social Security. Or not, that's for you to decide.

It follows Andrew replying to Bruce Webb (replying to me)...

Bruce, I think you're not right regarding the backward funding (meaning net transfers to early cohorts), and the table you cite tracking trust fund balances wouldn't really give a yes/no answer in any case.

If you look at this table you'll see that past/present participants have received $17.4 trillion more in benefits than they paid in taxes.

Future participants, even under scheduled benefits, will pay $1.5 trillion more in taxes than benefits. Add in the current trust fund balance and net things out and you have the infinite horizon shortfall of $13.6 trillion.

The upshot is that the infinite horizon shortfall that people deride (based on the 'infinite' part) isn't a function of over-payments to future generations, but to past ones. We already know that future cohorts must get out a lot less than they pay in, because past ones got out a lot more...


Yes, we do know that -- or should.

Again, with a paygo system it is a matter of arithmetic -- under paygo, total benefits = taxes in total, so if one group within the system receives benefits greater than the taxes they pay by $X, then everyone else must receive benefits less than the taxes they pay by $X. The Trustees' value for $X is above.

I've pointed out these facts and numbers (2007 version) to Bruce before, but he's been extremely reluctant to admit them. While I don't engage in the kind of name calling that others have elsewhere when I've brought these facts up, I think it is fair to state objectively that anybody who doesn't understand the existence of this "backward transfer" doesn't understand the first thing, literally, about the problems of Social Security reform.

Because it is entirely the rather large "backward transfer" that causes the ...

[] Intergenerational equity problem, due to the real rate of return on contributions falling from ...

36% annually for the first "Ida May Fuller"retiree cohorts, to

12% for those retiring in the 1960s (when Paul Samuelson praised Social Security for being an actuarially unsound Ponzi scheme that works), to

5% for the retirees of the 1980s, when Congress protected benefits for then-seniors by cutting benefits for, and increasing taxes on, the then-young, to

Under 1.9% and falling (less than the projected "risk free" rate on Treasury bonds of 2.9%, and thus a real economic loss) for today's young workers born in the 1970s and later.

[] Forced future tax increases/benefit cuts for the young, due to the $13.6 trillion remaining funding gap -- that must further reduce net scheduled benefits (and returns) for younger workers.

[] Loss-sharing problem, regarding how the inevitable losses to be suffered by the young should be spread between their generations.

[] Political support problem, regarding how to keep the public supporting Social Security in the future when they know they are losing trillions of dollars to it.

Consider: Prior generations received the above-noted by Andrew $17 trillion more than they put into Social Security, and thought very highly of it. Surprise! But if they had received $34 trillion less than that -- if it had made them poorer by $17 trillion -- would they still have thought so highly of it? In fact, would they have stood for that at all? That's the political future of Social Security.

[] Many false rhetorical claims about reform proposals, such as that any reform creates the dreaded "transition cost" that can't be financed.

But as Milton Friedman pointed out, it is the backward transfer itself that is the transition cost -- the need to fund past workers' benefits from workers' future taxes -- and this is a cost of the status quo. Reforms merely inherit it from the status quo. And it is completely bogus rhetoric to claim a cost of the status quo is a cost of reform. The status quo is going to have to fund this very same cost the same way as will any reform, with tax collections and/or benefit cuts, so it is hardly a cost of reform!

People who don't understand the "backwards transfer" can't possibly understand these issues, and how it drives them.

That said, darn few people understand it, and the Bush Administration did a dang poor job of explaining it (and these resulting issues) to the masses during its drive for Social Security reform. Karl Rove was no genius on that score. With predictable results.

But enlightenment is out there. For instance, one can look at what one might call the "Friedman scheme" to understand the fundamental situation of Social Security today, and the possible benefits of reform in the most extreme -- and thus most clearly visible -- case. Following Friedman (as per here and elsewhere):

Say you simply end Social Security cold-turkey while committing to pay all benefits earned to date, as of maybe 1/1/09. From then on everybody saves for their own retirement as they wish. The US government then must still pay the "backwards transfer" of $X trillion of benefits already earned but not paid for -- the "transition cost" to a world of "no more Social Security".

How would it pay for this? By imposing a payroll tax only on lower-level wages? Highly unlikely! Much more likely would be to finance the cost with the income tax -- using a progressive tax on a base about twice as large.

The point: The "backward transfer" is a sunk cost and going forward from here it must be paid for one way or another. But there is no logic in paying it with payroll tax because it is not a return on past-paid payroll taxes.

It was a general cash transfer, above the return on past-paid payroll taxes, from the nation and its politicians to the seniors of the past, made for better or worse (that doesn't matter now) and represented today by a remaining liability of $X. All other such general obligations the nation has ever incurred (for wars, space programs, Great Society programs) have been paid off with general revenue -- income taxes. This one should be and would be too, in a "no more Social Security world".

Now consider real-life Social Security as it is going forward today. It makes no more sense to pay off the "backward transfer" with payroll taxes on workers in our world than it does in the situation above. Yet it is what we are doing, and that is what's causing all the problems for Social Security today.

So now imagine a Social Security reform in which the cost of the backward transfer is removed from payroll tax and added to income tax. This frees all the contributions of workers to be saved for their own retirements in a diversified portfolio of investments (Treasury bonds, corporate bonds, equities, etc.) and increases their real rate of return from 1.9% or less to about 4.0% to 5.5% (private accounts or not, doesn't matter). Consequences:

* The "transition cost" of reform is $0. The government uses the exact same amount in tax revenue to pay down the liability of the "backward transfer". No change.

* Shifting the tax cost of funding the "backward transfer" from payroll tax to income tax is progressive -- let's call that good. (Also, being that the income tax base is much larger than the payroll tax base, fewer points of tax are needed, which is more tax efficient, and good.)

* Workers now receiving double to triple the return on their contributions will have much more retirement wealth per dollar contributed. This is good. They can even receive more at retirement than under today's system while contributing less than today -- giving them more disposable income over their working lives. This too is good. For workers this is in fact all very good.

* The accounting for Social Security becomes transparent. The "backwards transfer" becomes seen for what it is, not a fair return to participants of the past on their contributions, but a general transfer from the nation to them, above the fair return amount. (Whether the voters from 2009 on would wish to continue making such a transfer to Warren Buffett and all the other "rich" would be up to them.) And workers' future savings from their pay for their retirement in fact really become savings for their own retirement. Transparency is good.

(* Also, if a scheme like this had been implemented in the 1990s, when such ideas were first proposed, Social Security would be largely "out of the Treasury" for real by the 2030s, making the funding crisis for Medicare easier to deal with by points of GDP. Which would have been good. But too late for that now.)

Summing up the results we have "no net cost", "progressive", "good", "good", "good", "good", "good", ("good"). And all the problems resulting from the backwards transfer listed earlier are resolved going forward to the extent humanly possible.

So why does only a Milton Friedman consider such a thing? Why does everyone else think the backward transfer must be paid off with payroll taxes? When this produces a worse result by every measure? Even though "there is nothing in the nature of things that requires a particular tax to be linked to a particular expenditure."

Because, as Friedman also stated, FDR was right that the political psychology of the payroll tax is very powerful -- as LBJ's Great Society guy Wilbur Cohen said, peoples' perception and belief that that they are paying for their own benefits through the payroll tax (even when clearly not true) practically welds them psychologically to the payroll tax -- even to their own cost. It's politically mesmerizing.

Of course the "Friedman scheme" is a theoretical simplification -- there are practical problems with it, like maybe pushing income tax rates too high when the future cost of Medicare is added. But some tax must be used to collect the same amount, so if you don't like income tax you must find a better tax. (Some are proposing a new VAT.)

Yet I think it is very useful for clarifying one's view of what is really going on with Social Security, amid all the heated rhetoric and political noise.

A postscript about the "infinite" thing -- "infinite" isn't as long as you may think:

Many people hand-wave away the problems of Social Security thinking that since they are measured over an "infinite" term they will land on generations too distant in the future to matter to us, and probably can't be measured accurately anyway. This is a mistake.

When the finances of Social Security are measured over any limited, finite period (25, 75, 100 years, whatever) the net liability to the people living in that period is significantly understated.

This is because all the taxes paid by all the workers alive in the period are counted -- but the cost of benefits earned by and payable to millions of the same workers is not counted, due to benefit payment dates falling after the end of the period. The Treasury explains [.pdf] ...
It is important to understand that the magnitude of the infinite-horizon actuarial deficit is not driven by the use of distant or speculative long-range projections. Rather, the smaller size of the 75 year (or any finite period) deficit results from its use of a truncated time horizon.

The Trustees Report indicates that Social Security’s unfunded obligation for only past and current workers equals $14.4 trillion, which is actually slightly greater than the infinite-horizon shortfall. [Emphaisis in the orginal.]
Those "past and present" workers are you and me -- and the unfunded obligation for us is greater than that counted unto infinity.