Last December the Congressional Budget Office published a comprehensive paper that describes how they approach analysis of health insurance reform proposals. It is a critically important (and somewhat technical) document for anyone who cares about health care legislation in the United States.
CBO is the referee for the budgetary costs of legislation. They estimate the effects on the federal budget of a bill or amendment, and those estimates are then inputs into legislative processes that determine what bills and amendments can and cannot be considered. In the Senate, for instance, there are many cases where an amendment can pass with a majority (51 votes) if any spending increase is offset by an equal-sized or greater spending cut, but for which you would need 60 votes if that spending increase is not fully offset.
Like a referee, CBO gets screamed at a lot by the coaches and players (Members of Congress and their staffs). Like a referee, their judgment calls (estimates) matter. And like a referee, what CBO says goes. It does not matter whether the player’s foot was or was not on the three-point line. What matters is what the referee says about whether or not his foot was on the line.
This study is a bit like a referee giving an interview before the big game, and explaining his philosophy toward refereeing certain aspects of the game. Smart coaches and players will adjust their strategies based on this information. At a minimum, it gives the spectators insight into what to expect as the big game approaches.
Here is one interesting thing that pops out from the Executive Summary of the report:
These problems [rising costs of health care and health insurance, and the number of uninsured] cannot be solved without making major changes in the financing or provision of health insurance and health care. In considering such changes, policymakers face difficult trade-offs between the objectives of expanding insurance coverage and controlling both federal and total costs for health care. (page ix)
They key phrase is “difficult trade-offs between.” CBO is clearly rejecting the argument made by some advocates of universal coverage, that covering more people will reduce federal spending. CBO is saying simply that if you want taxpayers to finance more health insurance coverage, then both federal and total health care spending will go up.
Interestingly, while I hear the counter-claim frequently from some on the Left (more coverage will reduce unreimbursed emergency room and clinic care, leading to a net savings to the taxpayer), I have never heard it from the Administration. The logic implicit in the Administration’s argument is not that “Universal coverage will pay for itself.”
It appears that the Administration’s logic is instead, “Yes, covering more people through taxpayer-financed program expansions will cost more, but we will more than offset those higher costs through other long-term reforms.” As I explained yesterday, they have yet to specify those cost-saving policy reforms.
Expanding taxpayer-financed health insurance coverage will cost taxpayers more, and will dramatically worsen our short-term and long-term budgetary problems. This is CBO’s “difficult trade-off.” Policymakers must either choose which problem they want to solve, or they must find so much savings through other reforms that they can more than offset the higher expenditures from the proposed new spending program.
This may seem like a trivial conclusion, but it is non-trivial in the legislative debate, and should have an important effect on legislation. The referee has spoken.
When I was growing up, I was taught that you change the oil in your car every 3,000 miles.
Suppose I take my three-year old car to Jiffy Lube for an oil change.
Jiffy Lube has all the latest information technology, as well as good data on both manufacturers’ recommendations and best practices.
After entering my license plate into their database and checking my odometer, the technician says, “Mr. Hennessey, it’s been only 3,000 miles since your last oil change. Your manufacturer recommends an oil change once every 12,000 miles. We have even better data based on comparing wear and tear on vehicles from all over the country, and we recommend once every 10,000 miles. Still, you have at least 7,000 miles to go before you need to change your oil.”
I argue, “But I thought you were supposed to change your oil every 3,000 miles?”
He replies, “Those were the old practices. We have better diagnostic technologies, better engines, and better oil. It’s now every 10,000 – 12,000 miles.”
I respond, “Thanks. How much does an oil change cost?”
Imagine if the technician were to say, “$50, but your insurance covers it. You only have to pay a $5 deductible.”
What would you do?
The President is absolutely right when he says, “We can’t allow the costs of health care to continue strangling our economy.”
The President’s budget director, Peter Orszag, is the lead Administration advocate for this policy. Director Orszag is right when he writes on his blog,
Now, many of you have heard me go on about how important it is to reform health care in order to bend the curve on long-term costs and get our nation on firmer fiscal footing … and this data shows how critical that effort is. When we say that health care is consuming too much of our GDP, we are not just citing an abstract statistic. These costs have real implications in sectors across our economy, limit our economic growth, reduce opportunities, and harden inequalities.
He then, however, argues,
This is why the Administration is making historic investments through the Recovery Act in efforts that will be crucial in bending the curve on the growth of health care costs while improving the health outcomes we can expect from our medical system. We are investing over $19 billion in health information technology to help computerize Americans’ health records, which will reduce medical errors and enhance the array of data that physicians and researchers have at their disposal. We are investing $1.1 billion in comparative effectiveness research, which will yield better understandings of which medical treatments work and which do not.
Additional information is good, but the example above shows why information by itself will not significantly slow the growth of medical care spending. Information must be combined with the incentive to purchase high-value medical care – a decision that involves both the medical benefit of the treatment and the financial cost. The government could use this information to reduce costs in health programs that it runs, like Medicare and Medicaid (I am certainly not endorsing that). But those of us with private health insurance are largely protected from the costs of the medical care we use because of the general prevalence of low deductibles and copayments. Even if we have better information, we may not care if the benefit of a particular medical treatment is small, as long as it seems really inexpensive. The Administration’s proposals on health information technology, electronic medical records, and medical outcomes research may improve health, but they will have little effect on slowing the growth of health care spending for those with low-deductible, low-copayment private health insurance.
I favor helping individuals get information so they can decide what is high-value for them. I imagine that those who favor a single-payor system would say those tradeoffs should be made for everyone by the government.
The Administration is giving an incomplete answer. They need to explain not just how much they will spend on health information technology, electronic medical records, and medical outcomes research, but how that information will be used to reduce cost growth, and by whom.
To be able to credibly claim that they will slow the growth of health spending, the Administration needs to answer the following questions:
- Who will be empowered to make decisions based on this improved information?
- Upon what basis will that decision-maker compare the costs and benefits of a particular medical treatment, good, or service?
- How will you change policy to create incentives for that decision-maker to choose high value medical care?
Until they provide answers, they cannot legitimately claim to be slowing the growth of health spending in the private sector.They are just increasing government spending on technology.
Jim Capretta has discussed this in greater detail on his excellent blog, Diagnosis.
Suppose I bought an iPhone yesterday for $500.
Suppose I argue that I will save $2000 this week, because I intend to refraining from buying an additional iPhone today, nor will I buy one this Wednesday, Thursday, or Friday.
Suppose I plan to buy a new flat screen TV tomorrow for $1500.
Can I claim I that have paid for my TV by cutting other spending, and that in addition I will be saving $500 this week?
This is what the Administration has done with war costs in their budget.
There is no debate about how much I will spend this week: $500 for the iPhone, plus $1500 for the TV, equals $2000 of total spending.
The question is instead whether I have increased or decreased my spending compared to what it otherwise would have been.
In this example, I argued that I will cut my total projected spending by $500, and I am also paying for the TV by cutting spending.
You argue that it is absurd to assume that I would buy an iPhone each day this week. The right baseline, you argue, is to treat the iPhone purchase as a one-time expenditure, and tp use a spending baseline of zero for the remainder of this week. Thus the $1500 TV purchase is a spending increase, not a spending cut.
The argument about whether the President’s budget increases or cuts the deficit is therefore a debate about the baseline – what would happen otherwise?
Rep. Paul Ryan (R-WI) did a good analysis of the war spending assumption in the President’s budget. He and his staff conclude that the President’s budget includes $1.5 trillion of phony savings (over 10 years) by inflating the war spending baseline the way I did with my mythical cancelled iPhone purchases. The President’s budget makes a similar $330 B assumption for Medicare payments to doctors.
Even more intriguing, the President’s budget (table S-5 in this document) argues that the $9 trillion of incremental baseline debt they argue they “inherited” should include $835 B of additional debt resulting from two laws President Obama signed: the stimulus law, and the omnibus appropriations law. Clearly that $835 B of additional debt was not inherited, and should be netted out against their claimed future deficit reduction.
Here is the math behind the Administration’s claim of fiscal responsibility, and CBO’s countervailing analysis. All figures are for the next ten years (2010-2019):
|Additional debt under the baseline||$9.0 trillion||$4.5 trillion|
|Additional debt under the President’s budget||$7.0 trillion||$9.3 trillion|
|Effect of the President’s budget on additional debt||-$2.0 trillion of debt||+$4.8 trillion of debt|
Let us walk through this step by step.
- The Administration has a radically different starting point than CBO. The President’s budget starts by assuming that $9.0 trillion of debt will be accumulated over the next ten years if the President’s budget is not enacted. The Congressional Budget Office assumes that $4.5 trillion of debt will be accumulated over the next ten years in the same scenario.
- The Administration assumes that its policies will result in $7 trillion of additional debt added over the next decade. That is $2.3 trillion less than CBO assumes. We saw why yesterday – the President’s budget assumes that the economy will grow faster than CBO assumes. This faster economic growth assumption would result in faster revenue growth for the government, and therefore smaller (but still huge) budget deficits.
- These two differences in assumptions result in two completely different views of the President’s budget. The President and his advisors argue they are being responsible by reducing the deficit by $2 trillion over the next decade, while someone relying on CBO’s numbers would say the President’s budget is horribly irresponsible and that it increases the debt by $4.8 trillion more than it would otherwise be.
This debate about whether the sign is a + or a – is politically significant. The -$2 trillion figure is the cornerstone of the Administration’s claim to fiscal responsibility. It allows them to justify big spending increases like the $600+ B new health entitlement.
At the same time, we should not let this important debate obscure that, even using the Administration’s more optimistic numbers, the President’s budget would mean that debt held by the public will increase by $7 trillion over the next decade, to a share of the economy not seen since the end of World War II.
Even if you believe the Administration’s deficit reduction claim (I do not), it is nowhere nearly enough deficit reduction. We need either to dramatically slow spending growth, or raise taxes, or some combination of the two. I support doing it all on the spending side while keeping taxes from increasing. Your view may differ. But we cannot accumulate $7 to $9.3 trillion more debt over the next decade and claim that we are being fiscally responsible.
Each year Congress enacts 12 annual appropriations (spending) bills. Those bills are the subject of vigorous and legitimate fights about spending priorities.
Included in these annual appropriations bills are spending for defense, veterans, military construction, highways, housing, education (except student loans), foreign aid and the foreign service, the FBI, CIA, and Department of Justice, most of the Departments of Commerce and Labor, Congress and the White House, the Department of Homeland Security, including Border Patrol and Customs, highways, airports, and ports, health and energy research, scientific grants to universities, the Interior Department and Environmental Protection Agency budgets, national parks, … you get the idea.
Much of what we think of as the federal government gets its funding annually through these 12 bills. As a result, these debates and tradeoffs occur each year. This is called “discretionary” spending, which goes through the “annual appropriations process.”
In contrast, several “mandatory spending programs,” aka “entitlements” are on autopilot. Spending occurs based on formulas written in law. Those formulas contain variables that change according to external factors (wages, inflation, food costs, health care costs). Most importantly, these programs are on autopilot. Spending continues from one year to the next according to these formulas unless the laws are changed.
If Congress does not enact an annual appropriations bill this year for the Department of Justice, then DOJ will have to shut down on October 1st.
If Congress does not enact a law this year affecting Social Security, Medicare, or Medicaid, those programs will continue spending money based on the automatic formulas within them.
There are many smaller entitlements other than the big three (Social Security, Medicare, and Medicaid), but in an aggregate budget sense, it’s the big three that matter most. Payments to federal retirees and the refundable elements of tax credits are the next biggest.
From an aggregate budgetary perspective, these big three programs are (i) huge, and (ii) growing faster than the economy. As a result, they are swallowing up the rest of the budget, and they are the principal source of future spending growth.
These automatic spending increases that are “built into the baseline” are not carefully reexamined each year, and are not forced to compete with other priorities. The national parks, scientific research, and defense budgets are at a tremendous disadvantage – each year they have to compete for the marginal spending increase dollar, while the Big 3 entitlements quietly grow without anyone really noticing too much.
Let’s look at some numbers.
If the President’s budget were to be enacted in full, four areas of spending would increase dramatically over the next ten years.
- The new health entitlement would go increase $100 B per year in 2019 (starting from nothing now).
- Non-defense discretionary spending would increase $74 B from this year to 2019.
- A collection of low-income cash support and food stamps programs would increase $61 B from this year to 2019.
- Federal spending on student loans would increase $29 B from this year to 2019.
Importantly, these are proposed policy changes from the default baseline. Now let’s look at a picture and see where the increased spending would go. This graph shows increases in spending, comparing 2019 under the President’s budget to spending this year. To be clear, this graph shows the level of projected spending in 2019, minus the level projected for this year (2009). The amounts on this graph are increases above where we are now, measured in billions of dollars.
The green bars show the President’s big spending increase proposals. You can see the big new health entitlement, the net effect of his new student loans proposal, and the spending-side effect of his “making work pay” credit and his expansion of the earned income and child credits.
You can also see that he would increase spending for non-defense discretionary. The red bar on defense, in contrast, is the amount that he would shrink defense spending. You can see he would make a similar change in the Medicare spending increase.
What about the yellow bars? They dominate the graph. They show the spending that will occur if we follow the baseline. For the top two bars, that’s a concept that we just do what we did last year, and increase everything by inflation. For everything else, that’s the effect of the autopilot effect of mandatory spending programs.
Look at those bottom three bars. While we’re fighting about the new health entitlement (I’m opposed), federal Medicaid spending will grow $171 B without any Congressional Debate. Federal Medicare spending will grow (net of premiums) $367 B with almost no debate. And Social Security spending will be $408 B higher in 2019 than in 2009, if Congress keeps burying their heads in the sand for the next decade.
By all means, let’s debate and even fight about the green and red bars. But if we ignore the spending increases in Social Security, Medicare, and Medicaid that occur without any changes to law, it will all be for naught.
p.s. This trend gets worse in the second decade.
There has been a lot of debate about whether the President’s budget improves or worsens the future deficit picture. This is a debate mostly about baselines – what do you assume would happen otherwise? Rather than engaging in that debate here, I am going to look at the results of what the President has proposed.
What would federal deficits and debt held by the public be if the President’s budget were to become law exactly as proposed?
We will look at it both from the Administration’s point of view, and from that of the Congressional Budget Office, which serves as the referee for Congressional legislation. While the two differ in some respects, the fundamental conclusions are the same.
We will begin with spending. You can see from this graph that CBO and OMB agree precisely on how much the President’s budget would spend over the next ten years.
Now we turn to revenues. The President’s budget assumes faster economic growth than does the CBO. A bigger economy leads to more revenues for government, so OMB assumes more revenues from the same set of policies as CBO.
Don’t be fooled by the scale of the graph. That’s a $500 billion difference in 2019.
It is this difference in revenue assumptions that leads OMB to have a more optimistic deficit forecast for the President’s budget than CBO. I am going to put the proposed deficits in historic perspective, and switch to % of GDP so we have a fair comparison over a long timeframe.
The light green line shows historic deficits (above the line). The light blue line shows the post-World War II average deficit of 1.7% of GDP. Red is the CBO’s estimate of deficits under the President’s budget, and yellow is the Administration’s estimate. The gap between the two is because of different assumptions about GDP growth leading to different estimates of future federal revenues.
I find it more interesting and worrisome that both estimates show annual budget deficits for the next decade that far exceed historic averages. The Administration’s estimate hovers around 3.0% of GDP, while CBO’s estimate climbs steadily to 5.7% of GDP by the end of the decade. Neither is anything to brag about.
Now let us look at the effects of the President’s budget on debt held by the public.
Again, red is CBO, and yellow is OMB. Orange is Budget Director Peter Orszag’s new net worth measure of debt held by the public, minus the value of financial assets. While interesting, we should not ignore the yellow and red lines. If you borrow money to buy stock, you still are in debt.
No matter which estimate you choose, the conclusion is inescapable: under the President’s budget, debt held by the public will climb to a level not seen since the aftermath of World War II. And in the early 1950s we were paying down debt. Now we will be increasing our indebtedness, just as the federal government begins to expend massive amounts to pay Social Security, Medicare, and Medicaid benefits for the Baby Boomers.
In case you’re interested, the last year of this graph is 2019. In that year:
- OMB’s estimate of debt net of financial assets is 60.5% of GDP.
- OMB’s estimate of debt held by the public is 67.2% of GDP.
- CBO’s estimate of debt held by the public is 82.4% of GDP.
Here are my conclusions:
- CBO and OMB differ on their economic assumptions.
- CBO and OMB differ on how they define the baseline (not covered in detail here). This affects the rhetorical debate about whether the President’s budget makes deficits bigger or smaller.
- CBO and OMB basically agree on the qualitative picture of the results of the President’s budget on future deficits and debt. They disagree more about the point of comparison than they do about the result.
- No matter whose economics or baseline you use, the result is terrible for federal deficits and debt over the next ten years, both of which are way above historic averages and not showing any positive trends.
Thanks to Steve McMillin for his help with this post.
This site was down for about 16 hours today, beginning late last night, due to a hardware problem at my hosting service. I am hopeful the problem will not return.
If you notice a few things moving around over the next week, I will be playing with some functionality and layout issues. They will be small enough that you may not notice them.
I am scheduled to be a guest on CNBC’s Squawk Box Monday morning, beginning around 7 AM EDT. Former Vermont Governor and DNC Chairman Howard Dean and I will be discussing health care, taxes, and spending.
I am a low-tax guy. I have worked on tax issues for 12 of my 15 years in Washington, helping elected officials lower taxes and prevent tax increases. You can see a list of the taxes President Bush cut here. I would like to cut taxes far below where they are today, and I will continue to make the case that America is better off with a bigger private sector and a smaller government. America’s long run fiscal debate, however, is instead principally about whether we will allow future spending increases to force taxes to increase dramatically above where they are today.
I believe that America’s greatest economic policy challenge is the projected long run growth of spending on three programs: Social Security, Medicare, and Medicaid.
I would like to show you why I believe this, and how it relates to taxes.
Now that we have reviewed how big a bite the government has taken out of the economy over time, let’s examine the competing tax proposals for the near future.
Revenues are only one element of a budget proposal. For a complete picture of the effect of a budget proposal on the rest of the economy, we should also look at deficits. At the same time, it’s useful to start by understanding how much each budget proposal would take from the economy in total taxes.
We are using the same graph format as before, in which we measure total federal revenues as a share of the economy. Please remember that even a flat line on this graph means the federal government is collecting more taxes in inflation-adjusted dollars each year.
Let’s compare the President’s proposed total federal revenues, with proposals from House Republicans and House Democrats. The House Republican proposal was authored by the ranking Republican on the House Budget Committeee, Rep. Paul Ryan (R-WI). The House Democrat proposal passed the House, and was authored by the House Budget Committee Chairman, Rep. John Spratt (D-SC).
From this graph you can see:
- Over the next few years, all three proposals show a steep dip in revenues, followed by a steep increase. This has nothing to do with policy and is entirely about the predicted recession and recovery.
- The President’s budget would have the government take 0.9 percentage points more of the economy than the Ryan proposal. That’s an extra 90 cents out of each $100 of income. This is a fairly steady gap over time.
- The budget passed by House Democrats has higher taxes than even the President’s budget. The House Democrat budget is only a 5-year proposal, so the long run gap could be bigger, but the House is collecting at least another 0.5 percentage points more than the President’s budget. The budget passed by House Democrats means that government will take an additional 50 cents out of each $100 of income above what the President proposed. Compared to the House Republican alternative, the House-passed budget would take an additional $1.40 out of each $100 of income for the federal government. That’s a lot.
Now let’s compare these proposals to the historic average. I’m a low-tax guy, so I would like to use the post-WWII average of 17.9%. In most discussions, however, the range ends up being between the 50-year average of 18.1%, and the 30-year average of 18.4%. I will include those bounds on this graph, even though my preferred comparison path is a bit lower than the lowest line.
By adding these historic averages, we can see that the House Republican proposal would bring revenues down to a bit below the 50-year historic average. Even under this proposal, the federal government would be getting more inflation-adjusted dollars each year, because a constant percentage still grows in real dollars as the economy grows.
The Obama tax proposal is above the top end of the historic range, and the Democrat House-passed budget is above even that.