How Is This Still a Thing? "Generational Accounting" (Part III)

-- Posted by Neil H. Buchanan

In my Dorf on Law posts last Tuesday and Thursday, I commented on an op-ed in the NYT by Laurence Kotlikoff, an economist at BU who has been peddling something called "generational accounting" for approximately a generation now.  As I noted, my first law review article (after I transitioned from being an economist into legal academia) included an extended analysis of the many flaws of Kotlikoff's approach, flaws that are obviously designed to bias the results toward extreme right-wing policy preferences.  I was thus shocked to discover that generational accounting is "still a thing" -- not just in the fevered dreams of one right-wing ideologue, but with bipartisan support in the U.S. Senate.

For those readers who do not have time to read both of my previous posts (or my law review article), I offer this quick summary of Kotlikoff's approach.  He argues that the U.S. government is currently bankrupt, based on forecasts of possible deficits from today until the end of time (literally).  This "infinite-horizon" forecast, we are told, shows that all future federal borrowing has a present value of $210 trillion, which he calls the "fiscal gap."  Of course, two thirds of that amount is based on the borrowing that will supposedly take place after the next 75 years (seriously), and over 90% of the fiscal gap over any time horizon is based on the assumption that health-care cost increases will continue to outstrip the growth of the economy, until the health care sector simply swallows up the rest of the economy.

In short, "We are bankrupt right now" really means, "If medical care costs don't moderate, then Medicare will grow and grow, even as the rest of the federal budget -- Social Security very much included -- shows no sign of being a long-term problem."  Even so, Kotlikoff is trying to convince people that it is Social Security that needs to be cut, and cut NOW!

As I noted, the original claim behind Kotlikoff's approach was that fiscal gap measurements are not really predictions so much as warnings.  Unless we change our current laws, the argument goes, behold where we are headed!  Yet as Kotlikoff blithely admits, his forecasts are not actually based on "current law" but on an "alternative scenario" in which Congress does not rein in medical care costs.  And guess what?  If medical care costs are never brought under control, then there will be a problem!

I also noted that Kotlikoff's recent op-ed included a further, rather startling claim about "current law": "True, Social Security benefits could be cut by Congress and the president. But so can official debt, as Argentina’s likely default reminds us."  I allowed that Kotlikoff quickly changed the subject, seeming not to rely on that jaw-dropping logical error.  Over the weekend, however, I went back over some of his earlier writings on the subject, and I came across an op-ed that he published in the Harvard Business Review in 1993.  (The article appears only to be available for a fee, or through a university library's website.)  Here is what he wrote there: "[T]he government's obligation to make benefit payments to current and near-term Social Security recipients is certainly no less real than its obligation to pay interest on its Treasury bonds."

There are two striking things about this argument.  First, it is clearly inconsistent with the arguments from Republicans who want to use the debt ceiling to force the Democrats to agree to even deeper spending cuts.  The claim there is that the government is allowed to "prioritize," using incoming revenue first to cover payments on official federal debt, with everyone else who is owed money simply left hoping that they eventually receive payment.  If the legal requirement to pay even current Social Security recipients is contingent on the debt ceiling, then those obligations would apparently be much "less real" than obligations to pay principal and interest on federal debt instruments, on time and in full.

Second, and much more interestingly, Kotlikoff's claim would seem to negate the very idea that Social Security -- or, indeed, any other program -- can ever be cut.  In his 1993 piece, he says that not just current Social Security payments, but payments to "near-term" recipients, are as irreversible as federal debt obligations.  In his 2014 piece, he no longer even limits himself to the near-term, saying that cutting Social Security's future benefits is logically equivalent to defaulting on debt.

If that were true, however, then Kotlikoff's argument should be that we must raise taxes in order to prevent the fiscal doom that he predicts.  If formal default is bad, and if (as he further argues) informal default through inflation is bad, and if cutting Social Security and other promised payments is logically equivalent to defaulting, then the only path forward is to increase taxes.  We could stop passing laws that would increase future obligations, but we could not pass laws that reverse scheduled future benefits under current law.

Instead, however, Kotlikoff's entire shtick is to scare people with his fiscal gap numbers, in order to get them to start cutting Social Security and other future payments right away.  Apparently, therefore, cutting spending in the future is not the same thing as defaulting on federal debt, after all.

Which brings us, finally, to the generational part of generational accounting.  The fundamental analytical flaw in Kotlikoff's long-horizon accounting is that we could change our assumptions about future deficits in a way that can make the entire fiscal gap disappear.  For example, I could say that, starting in 2025, Congress will gradually introduce reduced spending and increase taxes, in amounts that are sufficient to make the infinite-horizon fiscal gap go to zero (or even to create a $210 trillion surplus, or any other number).

Surely, however, one could object that my saying so does not make it true.  Fair enough.  But what if Congress, led by the misguided-but-sincere efforts of people like Senators Coons and Kaine, actually passed a law that was scored in exactly the way that my whimsical supposition described, with long-term increases in taxes and reductions in spending sufficient to make the fiscal gap go to any number we like?

The initial retort to such a law is, of course, would be that a later Congress could undo the Coons-Kaine Act, so we cannot count on the law actually to be carried out.  In that case, however, we have merely exposed how utterly ridiculous is the exercise of predicting long-term fiscal gaps.  It is true that some laws will never take effect, but that logic applies to all laws, not just the ones that undermine Kotlikoff's argument.

The further retort is that "waiting makes it more expensive."  See, for example, yesterday's press release from one of the Kotlikoff-besotted deficit scold groups, which purports to show that we need to cut benefits for today's seniors in order to reduce the burden on future generations.  Of course, neither Kotlikoff nor any of his political enablers would put themselves on the line for the proposition that current retirees' benefits must be cut.  But even if they did, where is the logic in the idea that today's seniors need to pay a price to allow future generations to pay absurdly inflated prices for health care?

Kotlikoff has long claimed that future generations will pay a higher "lifetime net tax rate" than current generations are paying, but that is based on (among many, MANY other logical flaws) the idea that the proper measure of government's impact on people's lives is limited to the difference between the dollars collected in taxes and the dollars paid out in direct, personal benefits.  When government spends money to improve technology, or to clean the air or water, that is measured by Kotlikoff as a burden on future generations, not a boon.

There are, as one might expect, many more problems with "generational accounting" than can be summarized in a few blog posts.  The goal here has been at least to offer a taste of the absurdity of it all.  I will close with one further point.  In 2006 or 2007, I was sitting in on a semester-long seminar in which one of Kotlikoff's primary co-authors was a co-convener.  Each week, no matter the topic, the other co-convener would take a few minutes to endorse generational accounting and all of its scary implications.  Finally, in the middle of the semester, I could not keep quiet any longer.  I raised my hand and, as calmly as I could, laid out a litany of the most well-known arguments against Kotlikoff's approach.  We were early in that day's session, and there was more than enough time for anyone to engage with my comments.

Kotlikoff's co-author then spoke up, saying that he had been listening to such arguments for years, and that he finally had to have his say.  This was exciting for me, because there had certainly been no serious effort by Team Kotlikoff to engage with such arguments in any other forum.  Still, I admit to being momentarily worried: "What if he has a great argument that I -- and everyone else -- never thought of?  Yikes."  So I leaned forward, waiting for him to say what he had for so long wanted to say, anticipating a long, detailed reply at least to the most important points that have dogged Kotlikoff and his co-authors over the years.

I remember the entirety of his comment quite clearly: "Generational accounting is better than not having any way to measure long-term fiscal outcomes."  That was it!  The entire argument was that, hey, it's better than nuthin'.  But, of course, even that punt is most definitely not true, because generational accounting is not merely inaccurate, it is biased by design.  It is thus actually worse than nothing.  Even so, the person who uttered that non-defense is a frequent economic advisor to many high-level Democrats.  No wonder this snake oil is still on the market.