Changes in CBO projections of federal health spending: aging v. inflation

Brad Flansbaum was reading Aaron’s Carroll’s post and asking me via email about changes over time in CBO’s attribution of aging v. cost inflation as the driver of federal health care spending. It is an interesting question, but the answer doesn’t really change the fact that health care costs are the main spending side driver of federal budget unsustainability. Further, I don’t believe the relative impact of these two drivers of future health care costs necessarily implies a set of policy answers over others. Below I briefly compare the 2010 and 2012 CBO long run budget outlooks (you could do much, much more).

CBO’s long term budget outlook from June 2012 contains the following breakdown of the role of population aging v. cost inflation in explaining the growth in non-interest federal spending through 2037 (p. 15):

Focusing on health care programs (Social Security faces a purely demographic problem, while Medicare and Medicaid via the dual eligibles join the same problem with cost inflation), CBO says that population aging is responsible for between 52%-60% of the growth in federal health care spending through 2037, depending upon which budget scenario you use (extended baseline is the most optimistic scenario with federal debt-to-GDP stabilizing; the extended alternative scenario is the one where federal debt-to-GDP looks like Alpe d’huez).

Back track to the 2010 CBO long term outlook

Read more of this post

A tale of two (CBO) scenarios

If you can pick one thing in a policy argument, pick the baseline scenario against which to compare a series of policies.

The CBO’s baseline budgetary projections released this week show that the federal budget deficit will shrink greatly over the next few years. This is the so-called baseline scenario, that simply means that current law is followed. If you have a particular set of public policies you would like to enact, and you compare it to this scenario, it will be very hard for your policy tool kit to reduce the federal deficit in the next 10 years as compared to this baseline scenario because it contains some very large tax increases, and also some large spending reductions (Cuts from the Budget Control Act of 2011, Medicare payments to doctors).

However, I think there is almost no one who thinks all of the assumptions in the baseline projection will come to pass because it contains politically difficult things to enact.

So, the CBO also constructs an alternative fiscal scenario, that contains a mix of policies that might be viewed as more likely to come to pass. The alternative scenario assumes the following differences from the baseline shown above.

  • all expiring tax provisions other than the payroll tax cut are extended
  • AMT is indexed for inflation after 2011
  • Medicare payments to physicians in Part B are held constant at current levels
  • the automatic spending cuts from the Budget Control Act of 2011 do not take effect

The alternative fiscal scenario looks like this:

Deficit problem not fixed anymore, as deficits drop from the very large ones during the economic downturn, but begin to rise steadily again in the latter part of the decade as the tax receipts from our current tax code are unable to finance the default spending of our federal budget, particularly health care (Aaron points out we can’t get to sustainability via cutting discretionary spending).

I am not sure when, or even if, we will get a Grand Bargain to address the long run deficit issue. However, I am sure that if you can pick just one thing in the debate around these issues, pick the baseline against which to compare the policy options.


update: Here are the slides CBO Director Elmendorf used in testimony to the House (figures above taken from there).

Look back at CBOs 2000 long term budget outlook

Len Burman reminds us that the reality of the merging of the retirement of the baby boomers with the overall health care cost problem in the U.S. is not exactly breaking news. In the 2000 CBO long term budget outlook, CBO was urging caution about what to do with burdget surpluses (remember those!):

The aging of the large baby-boom generation and growth in the cost of health care will dramatically increase spending for federal health and retirement programs under current law. If policymakers act to ensure that the budget remains in surplus over the near term, the resulting drop in debt held by the public and the lower interest costs that follow will help offset some portion of that increase. Preserving the full amount of the projected surpluses could substantially delay the onset of fiscal problems and help boost GDP, providing a larger base of resources from which to meet the increased demand for spending. But even if policymakers preserved all projected surpluses, spending and revenues would be unlikely to balance over the next 75 years. {emphasis mine}

Even if all of the projected surpluses in year 2000 were reserved to offset these long term trends, an increase in taxes, decrease in benefits, or both would have been needed to totally pay for the retirement and health care needs of the baby boomers. That warning today, seems quaint.

Taxes and the Ryan Budget

Andrew Sullivan has provided qualified support for the Ryan budget, but doesn’t like the tax side, of which he says:

…the biggest flaw is the refusal to add new taxation to the proposal. Worse, it actually wants to reduce tax revenues.

I don’t think Andrew has this correct. In fact, the very large deficit reduction claims that Rep. Ryan makes in his proposal hinge on both spending cuts and a substantial increase in the total taxes collected as a percentage of GDP as compared to current levels.  The portion of Table 1 from the CBO letter to Rep. Ryan makes this clear (the full table is at the bottom of the post).

In 2010, total taxes collected (income tax, payroll, corporate, everything) equaled 15% of GDP. Ryan’s deficit reduction figures are dependent upon this increasing to 18.5% of GDP by 2022, and rising to 19% of GDP by 2030 and remaining at that level until 2050.  In fact, if taxes as a percent of GDP remained flat for the next 3 decades, we would have a deficit of around 3.75% of GDP in 2040 even with the Ryan budget cuts fully enacted!  Thus, Ryan’s budget assumes very large increases in the aggregate federal taxes collected as compared to the status quo, putting us back in line with historical averages over the past 40 years.

Ryan proposes decreasing the top marginal tax rate and reducing the number of rate bands, while claiming to substantially increase revenue collected.  There are several possibilities:

  • The tax side of the plan will not stand up to scrutiny when CBO scores it
  • Reductions of tax expenditures (home mortgage deduction, tax preference of employer paid health insurance, etc.) make it possible to raise this revenue
  • Income taxes will rise substantially for lower wage persons (it should be noted that ending tax expenditures is generally a progressive policy since the benefit of most such expenditures disproportionately helps higher income persons, so could balance higher income taxes paid by middle/lower income earners)
  • Other taxes could be raised

I take this plan as an acknowledgment that our current tax code cannot produce a balanced budget given any plausible level of spending.  Increasing tax receipts to 19% of GDP would put us near the historical average in terms of tax receipts over the past 40 years as the figure from the 2011 CBO budget outlook shows.

Of course, this is still well below average expenditures as percent of GDP from 1971 to 2010, but then that is why we have only had 4 balanced budgets since I was born. To have another balanced budget before I die, I suspect it will take something on the order of at least 20-21% of GDP collected in taxes (and spending reductions from current levels) given the movement of the baby boomers into Social Security and Medicare. I am somewhat agnostic about the overall structure of the tax code, and would like to hear a vigorous tax reform discussion focused around how to best raise a target percentage of GDP.  Moving from 15% of GDP to 19% of GDP collected in taxes is a step in the right direction, and it is notable to me that it has been embraced by the Republican Budget Committee chair. Democrats in the House are apparently going to release an alternate budget, and hopefully a useful debate will ensue, both about the level of taxation needed over the long run, as well as the makeup of the tax code.

Update 2: The House Progressive caucus with the outline of a budget that provides quite a contrast to Ryan’s, aiming for balance at tax receipts at 22.3% of GDP in 2021 (Ryan’s would get balance at 19% of GDP). This would include increases in top marginal income tax rates, raising the wage limit to which the Social Security payroll tax would apply to the 90th percentile of wages, and eliminate the taxable max on the employer side.  More on this later, but the Deficit Commission with balance point at 21% of GDP for balance around 2035, is where you would expect it to be; in the middle.  It will likely look better and better on the policy merits over time. Not sure about the politics.

Update: Austin sent me a longer time series of taxes collected as percent of GDP.

Long Live the CBO

I vividly remember being called out on strikes in a Little League Baseball game when I was 10.  As I plead my case that the last pitch was clearly a ball, the umpire said something I remember like it was yesterday.

Son. It is nothing until I call it. Go sit down.

And so it is with the Congressional Budget Office (CBO), the official umpire of Congressional legislation. They render judgments on what plans and eventually bills written into legislative language will do. They focus on the financial ramifications of legislation, but also assess the effect of laws on the number of uninsured persons, for example. The only certainty is that everyone is mad at the CBO at some point, and people are inconsistent in how they view CBO: sometimes provider of truth, while at other times, hopelessly misguided.

One of the favorite tactics of opponents of the ACA was to cast aspersions on the forecasts put forward by CBO. They didn’t like that their score of the ACA allowed proponents to argue the law could be implemented and cover 32 Million more people while reducing the federal budget deficit over 10 years. Some attacked the credibility of the CBO, while others said that politicians wouldn’t follow through with hard decisions in the out years, rendering the judgment of CBO misleading.

Yesterday on twitter, several noted that the CBO March 2011 Medicare baseline projections forecast a growth rate in Medicare over the period 2015-21 that is lower than that required by the ACA to trigger the the Independent Payment Advisory Board (IPAB) to make recommendations to slow the growth of the limited parts of the Medicare program to which IPAB applies prior to 2021 (won’t apply to hospitals, for example, until 2022). This was captured in a footnote (footnote d on page 2 of 5 of above link); the point being advanced by those opposed to IPAB was that since CBO was not forecasting  a spending reduction due to the work of the IPAB, that panel created by the ACA could be repealed without increasing the deficit over the 10 year window that the CBO uses for scoring.  For opponents of the IPAB, the CBO was misguided in March 2010 when they forecast savings due to the work of the panel, while the most recent judgment of CBO was evidence that the IPAB could be repealed without having to find a revenue offset.  It was the same CBO, using the same methods that rendered their best judgment given the data they had at the time.  The only thing that changed was who liked the answer they gave. The CBO gives, and the CBO takes away.

You can rest assured that some advocates of the ACA (who can be expected to mostly be reflexive opponents of the health policy components of Rep. Ryan’s budget) will have similarly situation-specific views of the work of CBO depending upon the argument they wish to forward. For example, it has already been said about Rep. Ryan’s Medicare voucher plan that future Congresses will not follow through on the amount of the voucher as specified by Rep. Ryan, because it will mean that people born in 1956 will have lots better Medicare benefits than those born in 1957.  This is the same argument made by opponents of the ACA in a variety of contexts: politicians in the future won’t make the tough decisions, and so any CBO judgment on the ACA is misleading. It is nearly an iron law of health policy that the hard things in my plan are easy to implement, while the hard parts of your plan are impossible. In the midst of this inconsistency, sits the CBO.

The Heritage Foundation provided some analyses of Rep. Ryan’s budget that sound quite impressive. They use so-called dynamic scoring methods to simulate the effect of tax reductions on economic growth, employment, tax receipts and therefore the revenue side of deficit reduction. Their assessments have been noted as being a bit rosy.  For example, the Heritage Analysis of Ryan’s budget projects that unemployment will be as follows (GOP budget) as compared to the No GOP budget default (chart via Dave Weigel):

Thus, Heritage is forecasting unemployment rates that seem implausibly low, at least by historical standards.  But, those are the judgments of an interested organization.  That doesn’t make them worthless, it just makes me want to know ‘what does the CBO think?’  It is important to point out that CBO has not scored the Ryan Budget proposal, because it has not been written into legislative language.  As CBO Director Doug Elmendorf’s letter of April 5, 2011 to Rep. Ryan says:

CBO has not reviewed legislative language for your proposal, so this analysis does not represent a cost estimate for legislation that might implement the proposal. Rather, it is an assessment of the broad, long-term budgetary impacts of the proposal, with results spanning several decades and measured as a share of GDP. It is therefore quite different from a cost estimate for legislation, which would require much more detailed analysis, focus on the first 10 years, and be based on more recent baseline projections.

The most speculative aspect of the CBO assessment of the Ryan proposal is the assumption of the amount of revenue that the tax code will bring in.  CBO has assumed, on the direction of Rep. Ryan’s staff, to assume that in 2022 the tax code will produce 18.5% of GDP, rising to 19% of GDP by 2030 and remaining static at 19% of GDP through 2050. Keep in mind that last year the tax code produced 15% of GDP in tax receipts. So, the deficit reduction claims about Rep. Ryan’s budget hinge on both the projection of spending reductions which have uncertainty but are likely fairly solid given the equivocations in the CBO letter, and an assumption that the proportion of the economy that will be collected in taxes will rise quite a bit in the next decade.  This claim is highly speculative, and is an assumption in the CBO letter.  Rep. Ryan has proposed a policy in his budget that alters the tax code to reduce the top marginal tax rates, and reduce the number of rate bands. As of yet, CBO has rendered no judgment about how much revenue that change in tax policy will actually produce. Table 1 from the CBO letter of April 5, 2011 to Rep. Ryan lays this out:

You can see under the actual column, that the current tax code brought in 15% of GDP in taxes last year, and that both CBO scenarios and Rep. Ryan’s proposal would increase the taxes received substantially a decade from now.  Ryan’s claim is that a reduction in top marginal rates will greatly increase the total receipt of taxes from 15% of GDP in 2010 to 19% of GDP by 2030.  CBO has assumed that to be the case for the purpose of providing this assessment of his policy, but eventually CBO will have to produce their own analysis.

For now, great credit to Rep. Ryan for laying out a budget, and especially identifying a set of spending reductions that can be debated and discussed (indeed, this budget is the blogger full employment act of 2011!).  However, the next steps are the most important.  The House Budget Committee must mark a bill and commit to specific legislative language on both the tax and spending side.  Then the CBO and the Joint Committee on Taxation must score the actual legislative language of the proposal, all the gory details (I wonder how long it will be?).  Will the tax code specified by the budget committee bring in 19% of GDP according to the CBO? Only time will tell.  And only after CBO has spoken can we fully evaluate Rep. Ryan’s budget, and the health policy proposals contained therein.  To paraphrase the Little League umpire I cannot get out of my mind, it is nothing until the CBO calls it.

Update: fixed a typo/wrong word use.