December 31, 2004
Long-Run Fiscal Priorities
Andrew Samwick writes:
Vox Baby: Answering Some Comments: Brad DeLong also makes some useful points about how things have changed to the point where he is out of the crisis mode (i.e., his view that the uptick in productivity since 1995 looks to be here to stay and this is not reflected in the Trustees' projections). However, I think these comments are incomplete. In particular, leading demographers think the Trustees' assumptions about longevity understate the decline in mortality and thus program costs. (See this paper for a recent example.) I discussed these issues in more detail in an earlier post.
So I still think we are in crisis mode, or, more precisely, that we are in "impending crisis" mode. If we do nothing, we hand a growing stream of annual Social Security shortfalls to future generations of workers with little policy flexibility to deal with them. I am trying to avoid that outcome. In the whole set of posts that I have done on Social Security, I haven't asked the Democrats to do anything that is more difficult than what I have asked the Republicans to do. I acknowledge that the problems facing Medicare are larger than those facing Social Security. That doesn't mean that we shouldn't solve Social Security's problems....
Touche on the life expectancy point. If I were running Social Security projections, I would up productivity growth, up immigration, leave natural increase alone, and up life expectancy. The net effect would be, with reasonable uncertainties, a 40% chance that the system would be in 75-year deficit. (As opposed to what happens when you take current Social Security central estimates, which with my uncertainties give an 85% chance that the system would be in 75-year deficit.)
I wouldn't, however, say that I'm not in crisis mode. I'm in Medicare crisis mode. I'm in ex-Medicare General Fund crisis mode. I'm even in Social Security crisis mode--after all, the first big hurdle the Social Security system faces is not that the system lacks resources to pay promised benefits after 2050, but that the General Fund--according to the Bushies' current policies--lacks resources to repay the Social Security trust fund after 2018.
As to Democratic proposals... Diamond-Orszag looks fine, but it doesn't address the systematic problems of governance that the Republican Party has inflicted upon America since the start of the 1980s. Whether you think that the Reagan deficits slowed American economic growth by only 0.3% per year in the 1980s and early 1990s or by somewhat more, it was without a doubt a domestic governance disaster of a magnitude that has only recently been matched by the renewed derangement of America's fiscal situation in the 2000s. This is why I'm attracted to the Federal Reserve model. Get the cabinet secretaries out as Trustees of Social Security. Replace them with guys appointed for fourteen-year terms. Mandate that they keep the system in actuarial balance--just like the Federal Reserve has a mandate to preserve price stability. And have them raise and lower the retirement age a bit at a time as the fortunes of the system wane and wax.
What I don't understand is why there aren't more right-of-center economists saying, "Stop this process now." They must realize that--as one Republican former CEA member says--you have to evaluate a policy initiative neither on the basis of whether your favorite set of reforms will be a net plus nor on the basis of whether the White House's ultimate proposal will be a net plus but on the basis of whether it will still be a net plus after Congress has done its worst to it. And I think that for this particular Bush initiative--as for almost all Bush initiatives to date, whether Medicare drugs or the steel tariff or the bra quotas or the corporate tax bill or the deficit--the answer is very clear.
Posted by DeLong at December 31, 2004 08:24 AM
TrackBack URL for this entry:
Listed below are links to weblogs that reference Long-Run Fiscal Priorities:
» Retirement Ages from Bradford Plumer
Brad DeLong wants to keep Social Security away from Congress' grubby little paws: "...I'm attracted to the Federal Reserve model. Get the cabinet secretaries out as Trustees of Social Security. Replace them with guys appointed for fourteen-year term... [Read More]
Tracked on December 31, 2004 12:37 PM
» Latest Social Security Scare Tactic: Americans Are from Pudentilla's Perspective
We believe Brad DeLong has answered the question of life expectancy satisfactorily. We note that by the time Il Ducetto is done with health care "reform," we'll probably have to revise life expectancy downwards anyway. Finally, we take the SCLM cover... [Read More]
Tracked on January 1, 2005 11:46 AM
What I don't understand is why there aren't more right-of-center economists saying, "Stop this process now."
Actually, I believe that quite a few have. Unfortunately, any criticism of Bushonomics is immediately labelled "left of center" regardless of their actual economic theories and assumptions.
The Presidency, and the party holding the presidency, now defines the terms of public discourse. A true moderate like Clinton wound up defining "liberalism" in the 90's solely because he was a Democrat. Bush, as a Republican President, defines "conservative"---and anyone who supports his policies, regardless of how insane they are, is considered a legitimate "conservative" economist.
And since just about the only economists who support Bush's policies are those that employed by far-right wing think tanks (Cato, Heritage, etc) where scholarship and analytical abilities matter far less than coming to the "right" conclusion. In other words, Bush has hijacked what it means to be a "conservative" economist, redefining "conservative" to mean "mad as a hatter".
Posted by: paul_lukasiak at December 31, 2004 08:55 AM
There is possibly a minor economic problem with Social Security that will in time have to be solved. There is however no need now to design the system anew. The push for a new design for Social Security stems from a political wish to finally set aside the legacy of Franklin Roosevelt's New Deal. We have a Republican Administration and Congress that wish to set a social legacy that will last at least as long as the New Deal legacy. Medicare costs may need to be addressed in a few years, and could well be addressed now, but Medicare is too explosive a political issue. Drug benefits have just been added to Medicare, and the costs of coverage for retirees are steadily eroding gains in Social Security payments, so there is happily no chance of cutting benefits if Republicans wish to hold Congress. Having Medicare bargain with drug and medical equipment makers over prices is not to the liking of the Congressional leadership, so the program will be left alone for now. Then, the turn will be to Social Security. Hopefully, Congress will find the political cost too high and not take apart this precious legacy. But, please do not tell me of a crisis and need to reform what works wonderfully well.
Posted by: anne at December 31, 2004 09:05 AM
Everything that the Republicans want to do with Social Security has to do with their desire to eliminate it and every other social program run by the government. Their thinking goes that if they can take a chunk of it and privatize it and there are no major disasters then they can just junk the whole program using the "success" of their previous efforts to justify it. They will also attempt to use that as a justification to eliminate everything else they don't like. Correct me if I'm wrong but didn't the Cato Institute issue a paper just about 4 years ago where they revealed what they really wanted as opposed to what they're supporting now? And that was of course the elimination of Social Security, Medicaid and Medicare.
Anyone really interested in fixing the Social Security problem would simply increase the payroll tax a very small amount, increase the current cap on income that's taxed drastically and be willing to increase the retirement age according to the current demographics BUT put in a health exception. If someone is "only" 65 when the age for full benefits is 68 but they can prove that they have health problems that would make retirement the best thing they can do (which would often mean they aren't going to live that expected life span) then give them their full benefits.
Posted by: Jim S at December 31, 2004 09:53 AM
There is absolutely no need to raise the retirement age. There is no Social Security crisis, and possibly no problem if productivity continues to grow reasonably. But, when we intimate there is a crisis we are lost. An increase in the cap on payroll taxes will be a fine buffer.
Posted by: anne at December 31, 2004 10:29 AM
The longevity issue misses the fundamentals. The fundamental is not age, it's disability.
The average manual laborer is effectively disabled for their job of choice in their 50s. Their backs give out.
The average tradesman is probably good into their 60s to 70s, depending on the nature of the work.
The average politician is good into their 70s and 80s.
The average physician and executive probably lasts into their 50s and 60s.
The average mathematician is in decline in their 30s (though as a teacher they can continue to work).
There's no real point in artificially extending retirement ages if the worker is effectively disabled for the domain in which they are competitive.
Estimates of social security costs need to model the impacts of disability. I suspect there is a huge economic payoff to therapies that slow the natural deterioration of the human brain. (By age 45 most of us are clearly working with a fraction of the neural network we enjoyed in our 20s, but we're surviving on a vastly improved experience set.)
Posted by: John Faughnan at December 31, 2004 10:33 AM
I agree with Faughnan completely. One of my brothers was a construction worker. He is now a retail clerk at 47 at one third his previous pay. My father was a systems analyst, he is now a retail clerk at 73 at one fifth his previous pay. Your ability to get jobs is related to your ability to do them as assessed by an HR person.
Posted by: wkwillis at December 31, 2004 11:00 AM
John Faughnan has some very good points, even though some of them may have been taken to illogical extremes. Yes most breakthought in math are done by young people, but the overwheming majority of mathemations can be self supporting until age 70. And that is the point for SS.
My father-in-law quit working as a truck driver hauling construction equipment when he was about 60 and worked in retail for another decade. He had less income but was still self supporting.
But I look would like a quick look at the data presented on aging in todays NY Times. They say that life expectancy rose from 47.3 in 1900 to 68.2 in 1950 and 77.3 in 2002. The jump from 1900 to 1950 was 44%, or modestly under 1% annually. The rise from 1950 to 2002 or a 13% rise was about a 0.25% annual rate.
They project it will rise to 83 in 2075 -- my calculator says this is 7.3% even though the NYT says it is 6%. But either way this works out to under 0.1% annually.
What the data is saying that the massive increase in life expectancy occured in the early part of last century and has been slowing significantly since about 1950. Of course this would be in line with at least what I always believed that the bulk of the jump in life expectancy stemmed from such public health measures such as clean water rather than private health measures like heart disease.
But when I look at the growth rates behind this data it is very hard for me to get very concerned about what the headline in the NYT article imples is a massive problem.
Am I missing something?
Posted by: spencer at December 31, 2004 11:19 AM
"the General Fund--according to the Bushies' current policies--lacks resources to repay the Social Security trust fund after 2018."
THERE it is. Greenspan said it twice last year, once during the primaries and once during the campaign: given the budget defecit, we have to cut SS benefits. No that the tax cuts have created an unsustainalbe situation, but that because of them, we have raided the SS accounts to pay for current spending.
And Kerry did not let out a peep; not a word.
Posted by: pragmatic_realist at December 31, 2004 11:59 AM
Spencer asks "Am I missing something?"
the question of increased "at birth" life expectancy is practically irrelevant to the social security issue---the huge decrease in child/infant mortality accounts for most of the gain in life expectancy prior to the 1950s...
the real issue (mentioned, but not terribly prominently, by the Times) is life expectancy of the "average retiree." If for some reason infant mortality were to rise appreciably, life expectancy at birth would decrease at the same time that life expectancy of retirees could still be increasing---putting additional pressure on the Social Security system. And with the increased emphasis on geriatric medicine, it is certainly within the realm of possibility that there will be new breakthroughs in the prevention of strokes, heart attacks, etc that could result in a significant increase in the life expectancy of seniors.
(this is not to say that I think there is a crisis.... I don't think you can really predict anything about what conditions will be in 40 years.)
Posted by: p.lukasiak at December 31, 2004 12:43 PM
Interesting discussion on demographics, but we apparently lean to quite a moderate problem as a result of the longetivity of retirees. Agreed.
Posted by: anne at December 31, 2004 03:17 PM
I agree with the poster who said we may live longer but maybe as disabled persons. Hope the medical profession can make those extra years productive years.
Posted by: dilbert dogbert at December 31, 2004 06:34 PM
Being both a PhD economist and a fully qualified actuary (FSA), I probably don't delude myself when I believe that I have better-than-average insight into longevity projections, and I would gladly bet that life expectancy at the relevant ages (65 and up) will be shorter than the current SSA projections. Their projected mortality rates at ages 70 and up increase faster than historical trends, and don't seem to account for the fact that we are getting fatter, that smoking has already bottomed out, and that diabetes, autoimmune disorders, and antibiotic-resistant bacteria are on the offensive.
Posted by: enfant terrible at January 3, 2005 09:40 AM
The net effect would be, with reasonable uncertainties, a 40% chance that the system would be in 75-year deficit.
Do you have a source or any substantial justification for this assertion? It seems to run counter to the sensitivity analyses conducted by the SSA.
[The sensitivity analyses conducted by the SSA *way* understate the variance of future developments]
Posted by: Victor at January 4, 2005 07:22 AM
Enfant Terrible -- Historical death rates have declined more slowly for older ages than for the rest of the population. The age-sex-adjusted death rate for ages 65 and over declined at an average rate of 0.71 percent per year between 1900 and 2000.
Accordingly, age-sex-adjusted death rates for ages 65 and over are projected to decline at average annual rates of about 0.29 percent, 0.68 percent, and 1.18 percent between 2028 and 2078 for alternatives I, II, and III, respectively.
Both of those quotes came from the http://www.ssa.gov/OACT/TR/TR04/V_demographic.html#wp155199 Trustee Report. 0.68 seems less than 0.71 to me. Where did they go wrong?
BTW, increasing mortality rates decrease life expectancy, they don't increase them. The issue we are considering here is the rate of decrease in mortality and corresponding increase in life expectancy.
Posted by: Victor at January 4, 2005 07:30 AM