Thursday, December 06, 2012

Some Notes on the Federal Budget

 
The debates about the Federal Budget get to be pretty confusing.  The absolute magnitudes of money mentioned have little context.  For instance, this year’s budget deficit is $1.089 trillion. It is good to know that this amounts to 7% of  the $15.6 trillion US GDP.  Obama’s revenue proposal will raise $1.6 trillion over 10 years, I have to remind myself that a $15.6 trillion/year economy would have generated $156 trillion over that period (not accounting for growth in the economy) and so Obama’s proposal amounts to about 1% of the economy.

Fortunately, the US government issues documents that lay out the financial situation in some detail.   I will use primarily the following three to try to gain a quantitative understanding of the issues.

[1] Monthly Budget Review (Nov 2012) issued by the Congressional Budget Office (look under “Topics” at http://www.cbo.gov)

[2] The 2012 Social Security Report (THE 2012 ANNUAL REPORT OF THE BOARD OF TRUSTEES OF THE FEDERAL OLD-AGE AND SURVIVORS INSURANCE AND FEDERAL DISABILITY INSURANCE TRUST FUNDS) found via Google.

[3] The 2012 Medicare Report (2012 ANNUAL REPORT OF
THE BOARDS OF TRUSTEES OF THE FEDERAL HOSPITAL INSURANCE AND FEDERAL SUPPLEMENTARY MEDICAL INSURANCE TRUST FUNDS) found via Google

What I want to understand are:
1.     What are the federal revenues, now and projected into the future?
2.     What are the required federal expenditures, now and projected into the future?  Social Security and Medicare, the two paid benefit programs are the chief required federal expenditures.

I think if we understand these, the choices we have will be clarified.

 
Federal Revenues:

For 2012, [1] tells us, the major sources of federal revenues are (I computed the percentages)

Major Source
Billion $
% of revenue
% of GDP
Individual Income Tax
1132
46.2%
7.3%
Corporate Income Tax
242
9.9%
1.6%
Social Insurance
845
34.5%
5.5%
Other
229
9.4%
1.5%
Total
2449
100%
15.8%

It turns out that federal revenues historically have been about 18% of GDP, as depicted in the chart below from The Heritage Foundation (I have no idea of whether their future projection is any good.)
Federal revenues as a percentage of GDP hit a sixty year low in 2009, when it hit 14.9%.  In 2010, it was 15.1%, 2011 – 15.4%, 2012 – 15.8%.

(Taxes overall take up somewhat over 30% of GDP when state and local taxes are included.)

Federal Expenditures:
For 2012, [1] gives us these figures  (I computed the percentages)


Major Category
Billion $
% of GDP
Defense-Military
651
4.2%
Social Security Benefits
762
4.9%
Medicare
469
3.0%
Medicaid
251
1.6%
Unemployment Benefits
96
0.6%
Other Activities
1022
6.6%
Net Interest on the Public Debt
258
1.7%
TARP
24
0.2%
Payment to GSEs
5
0.03%
Total
3538
22.8%

Federal expenditures in 2008 were 20.6% of GDP, 2009 - 25%,  2010 - 24.1%, 2011 – 24.1%,  2012 – 22.8%.

In round numbers, pre-fiscal crisis, revenues were 18% of GDP, expenditures were 21% of GDP, and the deficit was 3% of GDP.  With the crisis revenues fell to 15% of GDP and expenditures rose to 24-25% of GDP, and there is the huge budget hole of 9-10% of GDP.

The Budget Crisis

To my mind, there are really two problems – one is the short-term problem described above, where as a fraction of GDP government revenues fell and expenditures rose sharply compared to their long term averages.  There is a long- term problem also, in that Social Security Benefits and Medicare expenditures are currently at 7.9% of GDP and rising. Per [1], comparing 2012 and 2011, Social Security Benefits rose by 6% and Medicare by 3%.   In 2012, Social Security Benefits grew at their recent rate, but Medicare grew significantly slower than the 7% average it recorded during the last 5 years.

Our political system is trying to use the short-term budget problem (and associated things, like the debt ceiling, the expiring Bush taxes, the fiscal cliff and the recessionary effects of a sudden drop in Federal expenditures) as a lever to address the long term budget problem.    This is partly because long-term crises (like climate change) do not move our political system to take action, partly because ideology and not common sense dictate the range of choices.  

Examining the long term problem

Examining the long term federal budget problem relies on projections of the future, which is highly uncertain. The Social Security report [2] thus provides three projections for the future of social security,  Low-Cost, Intermediate, and High-Cost alternatives.  The Intermediate cost projection is their best estimate; the Low-Cost is optimistic and High-Cost is pessimistic in its assumptions. I shall use the Intermediate cost projection.

By the way, the Social Security [2] and Medicare [3] reports are long and dense, I do not claim to have read them fully.

Before going into the numbers, lets look at some of the factors that go into the projections.  They are described in section V of [2], and included historical conditions and trends, and expected future conditions including “fertility, mortality, immigration, marriage, divorce, productivity, inflation, average earnings, unemployment, real interest rate, retirement, disability incidence and termination. Other factors depend on these basic assumptions.  These other, often interdependent, factors include total population, life expectancy, labor force participation, gross domestic product and program-specific factors.”

Wow, can one really predict anything with such a complex set of factors? Well, Appendix D of the document provides a sensitivity analysis, of which I quote a paragraph:

During the 25-year period, the very slight increases in the working population resulting from increases in fertility are more than offset by decreases in the female labor force and increases in the number of child beneficiaries. Therefore, program cost increases slightly with higher fertility. For the 75-year long-range period, however, changes in fertility have a relatively greater effect on the labor force than on the beneficiary population. As a result, an increase in fertility significantly reduces the cost rate. Each increase of 0.1 in the ultimate total fertility rate increases the long-range actuarial balance by about 0.13 percent of taxable payroll.
So, at least we get a sense of how much the misprediction of each factor will affect the conclusions.

An issue I’m eliding over that concerns the trustees of Social Security and Medicare, is how their expenditures compare to the specific social security and medicare taxes.  My thought is that the federal budget can be rearranged as long it remains within control and does not occupy an ever-growing percentage of GDP.

With that, we go to Table VI.F4 in [2] “OASDI and HI Annual and Summarized Income, Cost and Balance as a Percentage of GDP, Calendar Years 2012-90” and extract the following numbers from the Intermediate cost projection.
(HI is Medicare Hospital Insurance – Medicare Part A.  OASDI is social security, the table extract below is OASDI only)


Calendar Year
GDP in billions $
SS as % of GDP
2012
15.757
5.01
2013
16.441
5.06
2014
17,300
5.09
2015
18,303
5.10
2016
19,340
5.11
2017
20,392
5.14
2018
21,458
5.19
2019
22,488
5.28
2020
23,525
5.38
2021
24,597
5.49



2025
29,392
5.89
2030
36,679
6.25
2035
45,940
6.36
2040
57,653
6.31
2045
72,302
6.21
2050
90,396
6.12
2055
112,810
6.08
2060
140,739
6.06
2065
175,704
6.04
2070
219,280
6.04
2075
273,504
6.04
2080
340,865
6.06
2085
424,327
6.09
2090
527,996
6.13

In this projection, the worst burden that social security presents is around 6.36% of GDP in 2035, I suppose the demographic bulge of the Baby Boomers fade away after that.  Anyway, these projections cannot by their nature include technological changes and catastrophes, both natural and man-made that will visit us during this long span of years.   Also, my mind boggles at a US GDP of $528 trillion in 2090!

From the Medicare report [3], I simply quote:

As can be seen in figure I.1, Medicare’s costs under the Trustees’ current-law assumptions rise from their current level of 3.7 percent of GDP to 6.0 percent in 2040 and 6.7 percent in 2085. If the SGR restraint were overridden, as described above, Medicare costs would rise to 6.5 percent of GDP in 2040 and 7.8 percent in 2085. Under the full scenario, in which adherence to the ACA cost-saving measures also erodes, costs would rise to 7.0 percent of GDP in 2040 and 10.3 percent in 2085.
Notes:
SGR = Sustainable Growth Rate formula for physician fee schedule payment levels
ACA = Affordable Care Act (Obamacare)

So the long-term budget problem is that Social Security and Medicare will grow to 12-13% of GDP from the current level of 8-9% (provided Obamacare and SGR hold, otherwise the situation becomes much worse.
Solving the long-term problem
The previous Congress refused to raise the debt ceiling until the conditions that led to our current “fiscal cliff” were set.  Now all the expenditures were authorized by that or previous Congresses – so it shows the peril of trying to bind future Congresses by current agreements.  In any case the future is uncertain.  So what we should do is make sure that the trajectory is in the right direction.

If we want to hold defence military spending at 4% of GDP , and discretionary spending at its current level of  6-7% of GDP, and Social Security and Medicare grow to 12-13% of GDP, that means that federal expenditures must be permanently at around 23-24% of GDP.  To have no problem of growing debt, that means federal revenues need to be raised from their current historical level of 18% to a new long term average of 23-24% of GDP.    This seems to be politically unacceptable right now (in the future, who knows?).  So we have to bend the social insurance curves.

Of the two major programs, Social Security presents the smaller problem (about 1.4% increased share of GDP) and Medicare the larger one (3% or more increased share of GDP). I also think Social Security is more definite and predictable, e.g., does anyone have any idea of when customized medical treatment informed by genetic profiles or stem cell therapies will kick in and whether they will improve or worse the cost situation?).  

On Social Security, therefore, I support some long term fix, improving its revenues by increasing the base of collection, and some cuts.

On Medicare, I want an year-by-year approach, with constant efforts to bend the cost curve downwards, instead of some grand bargain that gives us the illusion of security for the program fifty years out. So, for instance, Medicare is expected to grow from its current 3.7% of GDP to 4.0% of GDP a decade from now.  Can we make that 4.0% to 3.9% instead?