Arnold Kling  

The Budget Debate, VII

The Budget Debate, VI... About EconLog / Econlib...

I want to follow up on the attempt by the Center for Budget and Policy Priorities to compare the future bulge in entitlement spending to the Bush tax cut--see this thread. The CBPP was able to shrink the gap between entitlement spending and tax revenues in the middle of the century from something in the range of 5-10 percent of GDP to something less than 2 percent of GDP.

With Paul Krugman's extensive use of the term "present value," I inferred that this was due primarily to using a nominal discount rate. Instead, I now think that footnote 5 in the CBPP report is the major explanation:

The reason is two-fold: First, the Administration's figures exclude the Social Security and Medicare Trust Funds. They thus focus on the future cash flow associated with the programs (which ignores the existence of the trust funds), rather than the actuarial deficits (which include those trust funds)... Second, the Administration's figures include Part B of Medicare (Medicare physicians' services). Our figures include the Hospital Insurance component (Part A) of Medicare, but not Medicare Part B. Under federal law, Medicare Part B, like most government programs, is supported primarily by general revenues rather than by a dedicated trust fund.
Basically, what the CBPP is saying is that entitlement expenditures should not be counted if they are funded by something other than current payroll taxes. The trust funds are accounting buckets for tax payments from previous years, plus accrued interest. General revenues also are a different accounting bucket from current payroll taxes.

Thus, simply by changing the accounting buckets, the CBPP was able to dramatically reduce future entitlement expenditures. According to this logic, we could eliminate the entitlement deficit entirely by accounting changes, such as switching the funding mechanism from payroll taxes to general revenues.

Compared with this shrunken Social Security deficit, the CPBB is correct that the Bush tax cut appears large. By the same token, I would appear big in comparison with Shaquille O'Neal--if I could put 70 percent of him in a different accounting bucket.

Comments and Sharing

CATEGORIES: Social Security

COMMENTS (10 to date)
Patrick R. Sullivan writes:

Krugman has gone to the well of the CBPP before:

This is one of my favorites, as it it recommends the action facetiously suggested by Paul Zrimsek:


Patrick R. Sullivan writes:

Let's try that quote from CBPP again:

" To restore Social Security's long-term financial balance, three options

[snip the first two]

" . Raise the returns earned on the Social Security Trust Fund."

Bernard Yomtov writes:

Come on Arnold. Criticize the methodology if you like, but be fair in your presentation.

First, the study does not, as you imply, try to hide the fact that it deals only with the Hospital Insurance part of Medicare in an obscure footnote. This is explicitly stated throughout the text of the study.

Second, you might quote the footnote in its entirety, and not leave out the authors' rationale. After the part you quote, the footnote goes on:

" Calculating an actuarial deficit in Part B thus is akin to computing a "deficit" in the Defense Department or in other parts of government that are supported by general revenues; such a "deficit" has little meaning unless it is calculated for all federal programs taken together, relative to all projected general revenues.  Nonetheless, even relative to the Administration's figures for Social Security and Medicare, the tax cuts are substantial, and it is inappropriate to exclude them from a discussion of the long-term fiscal outlook."

I attributed your previous unfounded criticism to honest error. Now I'm not so sure.

Arnold Kling writes:

It was an honest mistake before. My guess is that you did not see the impact of their methodology, either, or you would have explained it in one of your comments.

Let's not make personal attacks here. The question is, does it make sense to disregard the cost of entitlements if they are paid for by something other than current tax revenues? Do you agree with that approach? Do you think that it contributes to a constructive discussion of public policy?

Scott writes:

Help me understand what you folks are arguing about. Is Arnold’s statement “Basically, what the CBPP is saying is that entitlement expenditures should not be counted if they are funded by something other than current payroll taxes,” correct?

If not, why not?

I agree with Arnold that this is misleading. If, as Jim Glass said in a previous post, the GAO estimates that expenditures on Medicare plus SS will be as high as 14% by 2050, then CPBB's report is *highly* misleading.

Bernard, you are technically correct that the report refers only to the 'Medicare Hospital Insurance' program throughout. This does *not* excuse them. If they wanted to be clear that they were not taking all of Medicare into account, they should have clearly differentiated between Medicare Hospital Insurance and Part B at the beginning of the report, in the actual text (as opposed to in a footnote). The average reader will probably miss the footnote and will not know the difference between the various parts of Medicare. CPBB should know this.

In addition, the nature of Medicare (as opposed to defense) is that there is an implicit govt obligation even if 'Part B' comes out of the general fund. Medicare benefits everyone over 65. Younger people have a reasonable expectation that, just as it is there for 65+ folks now, so it will be there for them when they reach that age.

Following from this implicit obligation, we must look more closely at projected future expenditures for Medicare than for other areas of the budget, which could be more easily adjusted at a later date, should society wish to extend or curtail certain programs or types of spending. 'Actuarial Deficit' does not seem to be the best measure here; perhaps we should just stick with 'expenditures' or 'net govt expenditures' to better understand what amount will not be covered by the 'pay as you go', program-specific revenues.

Lastly, I'm on the fence as to how the so-called 'trust funds' should be treated. I don't actually know what the current status is of these trust funds...are they still considered to exist? When we calculate the overall federal budget surplus/deficit, does it still exclude the funds in the 'trust funds'? If so, I think there's a decent argument for counting them when calculating 'net expenditures'.

Jim Glass writes:

"If, as Jim Glass said in a previous post, the GAO estimates that expenditures on Medicare plus SS will be as high as 14% by 2050, then CPBB's report is *highly* misleading."

Increase by 14% of GDP, to reach 22% of GDP. (Compared to total federal government expenditures today of 20% of GDP.) With post-2050 expenditures still rising indefinitely into the future. But GAO cuts off its analysis there because it takes the position to be so fiscally unsustainable as to be implausible.

And this is without counting any new entitlements such as prescription drug benefits.

Bernard Yomtov writes:


You raise a legitimate issue with respect to the two parts of Medicare. I do, however, still think you should have quoted the entire footnote.

As the last sentence points out, even including all of Medicare, the tax cuts are "substantial" with respect to entitlement costs, and I agree that

"it is inappropriate to exclude them [tax cuts] from a discussion of the long-term fiscal outlook."

That, after all, is CBPP's main point. Do you think they are wrong?

Jane Galt writes:

Bernard, it might be instructive to look at the history of tax cuts and increases. Reagan decreased rates; Bush raised them; Clinton raised them further; and now George W. has trimmed them back to roughly pre-Clinton levels. So in fact, ten years seems empirically too long to count the tax cuts; recent history suggests they won't last beyond the Bush Administration.

The larger question is whether we're concerned about the size of the shortfall or it's value to us, today. Most of the people alive today will feel the pinch of the shortfall starting early in the next decade, when the deficit-masking SS surplus begins to drain out to the Boomers and taxes or spending have to be raised. However, by the time it reaches truly crisis proportions, a majority of voters will probably be out of the net tax-paying group. In that sense, a high discount rate is appropriate. It's also thoroughly immoral. If you want to compare the impact of the deficits on the US citizens alive in any given year, there's no contest between a deficit of 14% vs. the comparitively trivial deficits attributable to these lost revenues.

Jim Glass writes:

I mentioned GAO analysis of the SS/Medicare funding deficits previously.

To get a clear picture of the *scope* of the accounting sleight-of-hand that the CBPP engaged in (as explained by our host) to make most of the cost of Medicare "go away" and let Krugman reach his "startling" conclusion, let's compare GAO's deficit projections to CBPP's and Krugman's:

The CBPP claim, as per Krugman,

"The new study, carried out by the Center on Budget and Policy Priorities, estimates the present value of the revenue that will be lost because of the Bush tax cuts .... comes to more than the Social Security and Medicare shortfalls combined."

Wow. So the Bush tax actually *created* the future entitlement funding gap? Even though so many people thought it was there all along? Yes, says Krugman, with these flat claims....

"The Bush tax cuts, not the retirement programs, are the main reason why our fiscal future suddenly looks so bleak.

"... the revenue that will be sacrificed because of those tax cuts ... would have been more than enough to 'top up' Social Security and Medicare, allowing them to operate without benefit cuts for the next 75 years"
Let's remember that flat claim: The revenue lost to the Bush tax cuts would not only have been enough, but actually "more than enough" to allow SS and Medicare "to operate without benefit cuts for the next 75 years"

OK, now lets look at GAO fiscal projections made on the fiscal baselines of 2000 and 2001, that is, *before* the Bush cuts were enacted, and while a decade of surpluses was still expected.

A summary of these projections is at They cover various scenarios, but just take a quick look at the charts and graphs.

E.g, Figure 2 shows the annual fiscal deficit by itself reaching 20% of GDP -- equal to the size of the entire federal government today -- between around 2045 and 2060, depending on the scenario. At that point GAO stops its analysis because it deems a deficit of 20+% of GDP-and-rocketing-upward too implausible to consider.

As Brad DeLong would say, the government is then bankrupt. And that is *without* the Bush cuts!

Now that is one heck of a difference from Krugman's claim that without the Bush tax cuts "revenue would have been *more* than enough to 'top up' Social Security and Medicare, allowing them to operate without benefit cuts for the next 75 years", eh?

*Without* the Bush tax cuts, deficits reach 20% of GDP and the gov't goes bust maybe 30 years before Krugman's 75 years are up -- yet, "The Bush tax cuts, not the retirement programs, are the main reason why our fiscal future suddenly looks so bleak". Eh?

Well, perhaps, if your methodology is to exclude the larger part of Medicare from your definition of "Medicare".

And after Krugman does just that, he of course accuses others of using deceitful accounting for devious political purposes. ;-)

Comments for this entry have been closed
Return to top