McCain on Social Security

This is a kind of perfect storm of John McCain economic policymaking, a lethal combination of bad ideas and total lack of comprehension:

On Social Security, the Arizona senator says he still backs a system of private retirement accounts that President Bush pushed unsuccessfully, and disowned details of a Social Security proposal on his campaign Web site.

What's at issue here is whether McCain wants to just cut Social Security benefits or combine cuts with a stab at privatization. Neither is, in my view, a good idea at all. It's conceivable that at some future point reductions in the benefits growth rate will be necessary, but then again that might not happen. The responsible course of action for the short-term is to focus on the portions of the government balance sheet that face actual short-term budgetary problems rather than hypothetical long-term ones. That'll mean a tax reform aimed at increasing efficiency and simplicity but also revenue levels, and it'll mean rejecting McCain's costly imperial conception of American foreign policy.

Comments

"[McCain] disowned details of a Social Security proposal on his campaign Web site"? Straight talk!

Posted by Steve | March 4, 2008 11:50 AM

The responsible course of action for the short-term is to focus on the portions of the government balance sheet that face actual short-term budgetary problems rather than hypothetical long-term ones.

Yes, how irresponsible of candidates to think about long-term implications of near-term policy choices.

Come on, Matt.

Posted by right | March 4, 2008 11:59 AM

Because, "right", we all know that focusing on imaginary problems is so much more useful than focusing on real ones.

Our actual retirement problem is with Medicare (where it's inseparable from the disaster of the health care "system" in general), not SS.

Posted by Steve LaBonne | March 4, 2008 12:08 PM

Yes, how irresponsible of candidates to think about long-term implications of near-term policy choices.

There should be a FAQ for this.

SS has been running a surplus for decades. It will remain in the black for many years to come. Even when it runs into the red, it will be many more years before it runs into debt, if ever.

If we should be concerned about SS running deficits, but not debt, than obviously we do not consider SS savings to be 'serious'. But then why would we look to increse those savings today when SS is running a surplus?

If we are only worried about SS running into debt, that might happen 50 years from. It also might not. Doing something about it now will not be that much more painful than doing something about it farther down the road.

The only sense of the Republicans strategy is to use short term policy (cutting seniors benefits) is to avoid a long term effect of avoiding having to pay back the debt owed to the SS fund which could be achieved by raising taxes on the rich.

Posted by mpowell | March 4, 2008 12:16 PM

I agree overall, but otoh eliminating the income cap like Obama has suggested if combined with rate reductions could take the issue off the board for the foreseeable future and reduce taxes on lower and middle income Americans. But yeah this is not an issue that needs to be addressed with any urgency in the next administration. Spend the effort on global warming and health insurance reform instead.

Posted by Ron | March 4, 2008 12:24 PM

McCain's slogan is "Viva la muerta"-- if you're dead, you don't need Social Security.

Posted by MattF | March 4, 2008 12:25 PM

disowned details of a Social Security proposal on his campaign Web site

Hilarious. Why is there a SS proposal on his website that the candidate himself doesn't even support? Might it be because he knows so little about economics, and cares so little about domestic issues that one of his advisers wrote it and he never bothered to even look at it?

What a bumpkin.

Posted by urbino | March 4, 2008 12:31 PM

Remember that McCain has "Greenspan's book" when it comes to economics. Mind you, there is no such book...

Posted by SavageView | March 4, 2008 12:50 PM

There should be a faq on this

There are already many (pick your favorite think tank!. While the "surplus" can last until 2040+, it will have an impact on the budget and national debt much sooner because the assets held in the 'trust fund' will be liquidated and refinanced externally. It will exacerbate any other budgetary problems of the day starting around 2018 - and given the tendency of the bond markets to anticipate supply - perhaps well before then.

Medicare is orders of magnitude larger. SS is a fire next door while medicare is the one in your living room.

OT, but for the life of me, I don't know why proponents of the 'trust fund is real' crew don't know the difference between real obligations passed around a single balance sheet and equally real obligations held at arms length. The latter have to be *funded*, which is where we run into trouble.

Posted by "Mindles H. Dreck" | March 4, 2008 12:53 PM

I see no reason not to schedule future benefit cuts in something like a public pension program as far in advance as practical, because that gives people time to adjust their retirement planning.

For example, I personally think it makes ample sense to require people to work for a longer period before receiving their full pension benefits, at least when those benefits take the form of an annuity. The logic is simple (and very unlikely to change): people are living longer and healthier lives, so should have to work a bit longer in order to fund annuity-style pensions. But they should also be informed well in advance that they will need to work a bit longer in order to receive a full pension, which will give them a fair opportunity to make appropriate decisions about careers, homes, savings, and so on.

In that sense, I think waiting for the relatively predictable shortfalls in Social Security to actually show up before cutting benefits is kinda dumb. Rather, while you should probably not cut benefits now, you should schedule them for the appropriate time, and allow everyone affected to make their plans accordingly. And if by some chance it turns out those benefit cuts are not justified by these predictable demographic trends (although that is pretty unlikely), then you can take them off the schedule. But generally it is better to err on the side of giving people more pension benefits than they expected than the other way around.

Posted by DTM | March 4, 2008 1:22 PM

Notice that the same article, McCain's own chief economic aide, Douglas Holtz-Eakin, attributes negative consequences to the same Bush tax cuts that McCain is going around touting as absolutely necessary for Blah Blah Blah. McCain really is a walking waffle iron.

Posted by Martin | March 4, 2008 2:04 PM

Notice that in the same article, McCain's own chief economic aide, Douglas Holtz-Eakin, attributes negative consequences to the same Bush tax cuts that McCain is going around touting as absolutely necessary for Blah Blah Blah. McCain really is a walking waffle iron.

Posted by Martin | March 4, 2008 2:05 PM

The most amazing thing about John McCain isn't that he sometimes says things so contradictory or stupid that people believe that they must be honestly held, well reasoned thoughts. It's that he is so stupid it is amazing that anyone listens to him at all.

OTH, he may be a step in our path to the future

Posted by Peter Driscoll | March 4, 2008 2:17 PM

"Spend the effort on global warming and health insurance reform instead.

Posted by Ron | March 4, 2008 12:24 PM"

Maybe he's hoping that global warming will kill us all before there's a social security crisis. *Ahhh*, the benefits of a Republican nominee who believes in global warming, always thinking ahead.

Posted by Reality Man | March 4, 2008 2:23 PM

Oh boy.
"but otoh eliminating the income cap like Obama has suggested if combined with rate reductions could take the issue off the board for the foreseeable future and reduce taxes on lower and middle income Americans."
Can you point me to anything Obama has said that would indicate he would reduce FICA taxes on lower and middle income Americans? It is not on his website, only the cap increase.

A cap increase sounds progressive but really isn't. The really wealthy in this country don't earn most or even much of their wealth from wages which is the only category of earnings exposed to FICA, raising the cap may hit Shaq pretty hard, and your family doctor, but will leave your typical hedge fund manager laughing. If you want additional progressivity in the overall tax system do it on the income tax side not in a way that punishes the professional class while giving the investor class a free pass.

As for retirement planning it is not clear why phasing in benefit cuts to ease sticker shock several years down the road makes any sense at all. We can pay 100% of promised benefits until 2041 at which point if current projections hold the benefit will have to be cut from one 160% better in real terms than retirees get today to one that is only 120% better, that is Social Security transforms from a great deal to a really good deal. Given that over the last eleven years the date of Trust Fund depletion has been pushed back from 2029 to 2041, a rate of more than a year per year (mainly because of overly pessimistic assumptions about current year growth), and that the numbers suggest that is going to continue, this is really only something to keep an eye on. To suggest that a potential in context mild cut in benefits thirty four years out means that I should accept a couple of decades of lesser benefits starting in 2023 (when I retire) doesn't make sense.

If the economy grows over the next twenty years at anywhere near the average rate it did over the last twenty the problem vanishes completely. If the generation who will start taking Social Security starting in 2041, which is to say Gen Y and includes our host, is really worried about this maybe collectively you can get off the couch and grow the economy like your parents and grandparents did. I did my part, I'll be 84 in 2041 and 89 in 2046 (which is where CBO pegs depletion), asking me to take cuts because younger generations can't keep long term Real GDP above 2.0% doesn't really seem fair. And that pretty much is all it takes.

Posted by Bruce Webb | March 4, 2008 2:27 PM

Private social security accounts didn't go over so hot last time, I can only imagine how much more popular the idea will be now that the market is dropping. The best time to privatize social security would have been during the dot com bubble, when everyone thought that stocks only went up.

For this election people are going to want to hear government guarantees, and they'll have little interest in being given the chance to assume market risk.

Posted by daveNYC | March 4, 2008 2:45 PM

"The logic is simple (and very unlikely to change): people are living longer and healthier lives, so should have to work a bit longer in order to fund annuity-style pensions."

That logic is simple enough to have occurred to the demographers who work in the Office of the Chief Actuary in the Social Security Administration who in turn have worked it into their models. If you have a specific quarrel with the mortality projections in Table V.A3-Period Life Expectancy http://www.ssa.gov/OACT/TR/TR07/V_demographic.html#wp151333 or those in Table V.A4-Cohort Life Expectancy http://www.ssa.gov/OACT/TR/TR07/V_demographic.html#wp194031 then please share them.

There seems to be an odd belief that somehow nobody in charge actually sees the implications of Boomer retirement when in fact all of that data is built into the models already. We don't need to increase retirement age if we can grow the economy at trend. It really doesn't matter how simple the logic appears, does it survive encounters with the numbers? In my experience very, very few people have bothered to actually examine the tables and figures of the Social Security Reports, preferring instead to pick up the impressionistic version of 'crisis' as seen in the newspapers. Social Security does not need a fix until someone can show me with numbers that it does. It is not like the data is hidden, Reports for all years back to 1942 area available in a variety of formats at http://www.ssa.gov/OACT/TR/index.html . For recent years you have a choice of PDF or HTML or if you are old fashioned even get a paper copy sent with free First Class mailing.

Second hand knowledge about Social Security is worthless because with some fairly rare exceptions most reporting on this is limited to what reporters can glean out of the press release, and even the more diligent rarely seem to venture beyond the 15 page Summary (essentially a political document). If you want to speak knowledgeably about this you really need to engage the data tables, you can bet big, big money that print reporters and columnists whose names don't start K-r-u-g, are actually in a negative information position on this issue. Negative in that most of what they 'know' is factually incorrect.

Posted by Bruce Webb | March 4, 2008 2:48 PM

Bruce,

First, I don't see why it is an either/or proposition: we could make our taxation more progressive both by raising the payroll tax cap and by addressing how we tax investment income.

Second, you are papering over exactly what is likely to happen with Social Security. It is true that the Trust Fund will likely allow the currently-scheduled benefits to be maintained until around the middle of the century, but the program will start collecting less revenue than it pays out in benefits much sooner (and much more certainly) than that.

Now, in theory the Trust Fund is there to cover temporary shortfalls, but there is no necessary reason why it should all be used up in the next 30-40 or so years. In other words, there is no reason why the point at which the Trust Fund is 100% depleted is the first point at which we should start caring about the fact that the Trust Fund is being depleted.

And in fact, only a portion of the anticipated shortfall is essentially temporary (the part caused by the "Baby Boomer" retirements distorting the demographic balances). The other portion of the shortfall is essentially permanent (the part caused by people living longer and healthier lives), and the Trust Fund by its nature cannot be used to fix a permanent problem.

Third, along those lines, you are also papering over the fact that our grandparents tended to live shorter lives than our parents will live, and in turn we will live even longer, and so on. So the problem is really not a matter of people being less productive while they work (they are actually becoming steadily more productive, in fact). It is a matter of people living a lot longer once they stop working, and eventually it is just unrealistic to expect productivity gains to make up for ever-longer retirement periods.

Indeed, that growing imbalance will inevitably show up in real GDP per capita, and it would be decidedly UNfair to blame that effect on working people. Indeed, the real GDP per WORKING person will likely continue to grow at a very healthy pace. But as the ratio of non-working people to working people continues to rise, it creates an increasing drag on real GDP per capita that the working people alone simply cannot be expected to overcome.

In short, likely you (and I) are the ones asking for the "unfair" change in the basic deal presented by Social Security, at least if we expect to keep the same level of benefits over a much longer period of time without working any longer in advance to fund those annuities. So there is actually nothing unfair at all about the program asking us, right now, to work a little longer before we start collecting those benefits, since that is properly understood as actually just preserving the basic economic terms of the original deal.

Posted by DTM | March 4, 2008 2:52 PM

Bruce,

Just a quick followup: of course the program administrators are cognizant of the demographic trends. But what they lack is the authority to do things like increase the age at which people can retire with full benefits.

And again, we cannot expect to indefinitely "grow the economy at trend" without increasing retirement ages, if by that you mean increase real GDP per capita as opposed to real GDP per working person. That is because a decrease in real GDP per capita becomes unavoidable if the ratio of nonworking to working people increases beyond the rate at which productivity gains can keep up.

Posted by DTM | March 4, 2008 3:00 PM

"While the "surplus" can last until 2040+, it will have an impact on the budget and national debt much sooner because the assets held in the 'trust fund' will be liquidated and refinanced externally. It will exacerbate any other budgetary problems of the day starting around 2018 - and given the tendency of the bond markets to anticipate supply - perhaps well before then."

Have you quantified this effect? Well the Trustees have in both current dollars and in inflation adjusted constant dollars and the effects start out very small. http://www.ssa.gov/OACT/TR/TR07/trLOT.html
Current dollars can be seen in Table VI.F8 and constant dollars in VI.F7. Initially the impact will be around $30 billion inflation adjusted around 2017 and will not hit triple digits until 2023. (In fact the Trust Fund continues to increase until 2023, it is only then that any principal needs to be paid out.) As late as 2030 the gap between income and cost is only projected at $260 billion. The notion that these are huge sums that will paralyze the bond market is frankly nonsense. Get some control over the General Fund side and this is easily financeable. Which is why few opponents of Social Security actually deploy any numbers except the Infinite Future Horizon estimate. The numbers are not particularly scary. If you look them up.

And these are the ones drawn from the historically pessimistic Intermediate Cost assumption. Plug slightly better numbers in your model and the effect becomes even lighter and farther out. Under the perfectly plausible Low Cost model the Trust Fund never goes to depletion and only requires payment of about 25% of the interest owed to fully fund benefits. Why doesn't anyone ever talk about the Low Cost alternative? Why don't they highlight the actual payroll gap? Because they don't want you to know how little growth it would take to backfill the entire gap, or how small the actual tax gap is. Bush took tax increases off the table right at the outset because he knew that it was by far a cheaper solution. Per the Trustees, even given long term Real GDP mired at 2.0%, backfilling the entire projected gap would cost 1.95% of payroll. Do the math. If the economy does better than that then that gap shrinks. In 1997 the payroll gap was 2.23%, by doing nothing from 1997 to 2007 the net effect was the equivalent of a 2% annual tax cut. People who tell you 'we can't afford to wait' either don't know the numbers or are lying. 'Nothing' has been a numerically proven successful plan for Social Security for over a decade. Left alone the problem is getting smaller and more distant in time, that is hardly the definition of 'crisis'.

Posted by Bruce Webb | March 4, 2008 3:15 PM

If McCain decides to trot the Bush Social Security privatization plan back out for another go, I think Matt's epistle on the matter still works pretty well as the response.

Even though most Americans know objectively that SS is pay as you go, when they realize that the privatization plan defaults on bonds they paid for, they'll get really upset. The Feds have never welshed on their debt, and I can only imagine what'll happen to the value of Treasury Notes if they started... to the tune of ~1.5 trill.

I originally found the article in '05 during a Google search on the subject, my first exposure to Matt's writing.

Posted by cmholm | March 4, 2008 3:21 PM

Bruce,

I've looked at the various models the Trustees have commissioned, including the stochastic models. The upshot is that while there is indeed a considerable amount of uncertainty over exactly when various events will occur, the basic shape of the curve does not vary.

Posted by DTM | March 4, 2008 3:50 PM

DTM "And again, we cannot expect to indefinitely "grow the economy at trend" without increasing retirement ages, if by that you mean increase real GDP per capita as opposed to real GDP per working person. That is because a decrease in real GDP per capita becomes unavoidable if the ratio of nonworking to working people increases beyond the rate at which productivity gains can keep up."

I think you are proving my point here. Without quantification you have effectively built your entire case into an unsupported 'if'. Make the case that productivity gains can't keep up to that ratio, but using numbers this time. Instead you are just presenting an impressionistic case backed by pretty much nothing.

All of these factors are built into the current models and those models are available for detailed inspection. You can look for yourself at the various demographic assumptions, taking special care to examine immigration numbers (which seem to be artificially suppressed) as they show up in Table V.A1 and look at the results as they appear in Covered Worker Ratio (worker to retiree) and Total Dependency Ratio (worker to all dependents including children) both of which are found in Table V.A2.
Then you can take your conclusions from that and examine Productivity and CPI assumptions in Table V.B1 in conjunction with the Employment and Real GDP assumptions in Table V.B2. And we can go from there. Otherwise it is kind of a waste of time.

The logical case you are making is familiar to me, I really am not new to this debate. It simply does not hold up on encounter with the numbers. http://www.ssa.gov/OACT/TR/TR07/trLOT.html

Do you really think that immigration will decline in both absolute and relative terms from the 1.1-1.2 million rate typical of the last ten years to the 900,000 a year combined legal and illegal in the current Intermediate Cost model? This is rather an odd result of a crisis that is framed in terms of a worker shortage. Do you really think that productivity will decline to 1.7% by 2013 and stay at that level forever? Do you think that Real GDP will go to 2.2% by 2013 and then slowly slip to ultimate 2.0%? Is this really the median potential outcome? Are the odds really 50/50 that the actual outcome will be worse than that?

Accepting dates drawn from Intermediate Cost alternative (and all dates you see whether 2012, 2017, 2023 or 2041 for various bend points are taken from that model) implicitly means defending each and every assumption that was used to derive them. This debate has been endlessly discussed from the top down, the logical positions are fully staked out. What is not happening is an honest discussion from the bottom up using the actual numbers from the data tables. This is not an accident, when opponents of Social Security encounter the actual numbers their case dissolves.

If you like take a run at it, build the numeric case for Intermediate Cost, explain why Low Cost or results close to it are impossible. But fair warning. I never get into arguments with Civil War buffs, they generally have spent for too much time poring over those books and touring those battlefields, I wouldn't dream of taking Matt on about anything related to the NBA, he knows the players and I don't. On the other hand I am perhaps the oddest of odd ducks, the Social Security buff, and have been at this since 1997. I'll agree it is a really strange hobby, then again I think that memorizing the Order of Battle at Antietam is a little odd, to each his own.

Posted by Bruce Webb | March 4, 2008 3:51 PM

Bruce,

We seem to have overlapped posts. Like I said, I have seen the various models, including the stochastic models. I certainly wouldn't claim the low cost end of those projections are "impossible". But they are unlikely, at least according to the best efforts of those deriving the models. And they don't change the basic picture.

Just to address one point: again, I don't find those productivity and GDP projections at all implausible. Again, an increasing dependency ratio (to use the Trustee's terms) is not a bad bet at all, and that will show up in productivity and real GDP numbers calculated on a per capita basis.

Posted by DTM | March 4, 2008 4:03 PM

Oh, and to reiterate one policy point:

Because there are benefits to allowing people to make retirement plans well in advance, it seems quite sensible to schedule any projected benefit cuts (such as an increase in the retirement age) well in advance. But if we do that and it turns out the most favorable scenarios in the models come true, then we may well be able to cancel those scheduled cuts. Again, as a policy matter, it is better to err on the side of scheduling cuts and then cancelling them as opposed to not scheduling cuts and having to make them with little or no notice.

Posted by DTM | March 4, 2008 4:08 PM

Sorry, one last comment:

I did go to look at the linked tables, and confirmed something I remembered.

As far as the projections of the dependency ratio is concerned, the biggest difference between the models was not immigration, but rather relative fertility and mortality rates. Specifically, the Low Cost model would require a return to a fertility rate the U.S. has not seen since the early 1970s, and that would be contrary to the trend we have seen around the world. Moreover, the Low Cost model requires much higher mortality rates than the Intermediate model (about 40% higher by the end of the series).

So, the Low Cost model is not dependent on just happy things, like faster productivity gains. It is also dependent on unhappy things, like slower mortality gains.

Posted by DTM | March 4, 2008 4:21 PM

DTM I too have looked at the Stochastic projections (first introduced with the 2003 Report). They answer a different question than the one posed. Instead of testing any specific assumption they simply calculate whether random variation would move you from Intermediate Cost to High Cost on the low end or Low Cost. They do this by essentially holding Intermediate Cost steady and running 6000 different trials against that. I am not a statistician and the language seems designed to confuse the casual reader but these projections assume what needs to be tested, that Intermediate Cost is a good choice for a median projection. When actual results continually come in at or beyond the same side of your confidence interval, certainly the result from 1997 to 2004, it suggests something is wrong with your model, or worse your modelers. If the productivity' crash we have seen since Q4 2005 continues then perhaps Intermediate Cost gets some renewed support. But to say the least projections of near perma-recession are not exactly proof of the magic of tax cuts.

Posted by Bruce Webb | March 4, 2008 7:06 PM

DTM are you claiming the ability to casually inspect the relevant fertility and mortality tables. And conclude they outweigh a 20 to 30% underestimate of immigration over the next thirty years? The mortality models don't diverge much for decades an d the impact of fertility by definition is offset by the 20 years plus it would take for this new cohort to affect the workforce. Plus you still seem to be afflicted by an inability to put any of this in numeric form. Color me unimpressed.

Posted by Bruce Webb | March 4, 2008 7:25 PM

DTM exactly why should I privilege the retirement planning needs of Gen Y retirees faced with a modest benefit cut that may or may not happen, if it does will leave the retiree with a benefit in real terms 20% better than a similarly situated worker gets today, and is caused by the failure of people born since 1980 to grow the economy in the same way my demographic cohort did?

Your argument adds up to "Boomers need to take benefit cuts because Gen Y and Millenials are going to be huge slackers going forward"

Don't make me tell you how I had to walk to summer school uphill both ways in waist deep snow and 100• Heat. Get off the couch and give us 2.6% productivity and quit whining.

Posted by Bruce Webb | March 4, 2008 7:43 PM

I don't even need 2.6% productivity. 2.0% productivity and 2.6% Real GDP brings this one home or at least extends any problem for decades beyond 2041.

Posted by Bruce Webb | March 4, 2008 8:00 PM

I see now that you used some numbers. Can you explain with examples first why Low Cost mortality numbers represent a 40% worse result than IC numbers and in any event is less realistic?

I don't see ' Low cost is sad, let's look on the bright side' as really being a policy argument'

Posted by Bruce Webb | March 4, 2008 8:20 PM

McCain wants to live in Sparta.

Do they have Social Security in Sparta?

Posted by mike | March 5, 2008 1:00 AM