The Congressional Budget Office has released a preliminary estimate (“score”) of the draft Kennedy-Dodd health care bill.Some in the press are reporting that this is a “$1 trillion bill.” This poorly explains the true budgetary impact of the bill for several reasons. The bill would increase spending by more than $1.3 trillion (that’s “thirteen hundred billion dollars”) over ten years, and even that understates the impact because the bill phases in over the first five years.

I explained last week that the bill is incomplete. I expect a later version of this text will be marked up by the Senate Health, Education, Labor, and Pensions (HELP) committee, and then combined on the Senate floor with a companion bill marked up by the Senate Finance Committee. Even in its final form the HELP Committee text will be only half a bill. I further expect the HELP Committee bill will end up increasing the budget deficit, while the “pay-fors” (offsets) will come from the Finance Committee bill, which it appears will raise taxes and cut Medicare and Medicaid spending.

New health care entitlement spending

CBO has estimated the effect on the federal budget of the new subsidy component, which I described last week:

People from 150% of poverty up to 500% (!!) would get their health insurance subsidized (on a sliding scale). If this were in effect in 2009, a family of four with income of $110,000 would get a small subsidy. The bill does not indicate the source of funds to finance these subsidies.

There are at least six other major components of the bill that have not yet been scored by CBO. Part of this is due to the time crunch, but most of it is because the Kennedy-Dodd staff have not yet given CBO specific enough legislative language for CBO to do their thing.

  1. The budgetary effects of neither the individual mandate nor the employer mandate are included in this score. I think CBO will find these provisions would raise revenues for the government and reduce the deficit. While the leaked draft of Kennedy-Dodd was specific about the employer mandate, the official version has just the placeholder language, “Policy under discussion.” Both mandates leave wide discretion for the Secretaries of Treasury and HHS to create a level and structure of taxation “to accomplish the goal of enhancing participation in qualifying coverage.” It is extremely difficult for CBO and their tax counterparts, the Joint Committee on Taxation (JCT) staff, to estimate something like this.
  2. The estimate does not include the budgetary cost of expanding Medicaid to childless adults with income below 150% of the poverty line. I expect that this will add hundreds of billions of dollars to the cost over the next decade.
  3. It does not include the requirement that health plans define “children” as dependents up to age 27. I expect this will raise costs.
  4. It does not include the effects of the Medical Advisory Council’s ability to define benefits, or the requirements that plans rebate premiums to the insured. I think this too will raise costs.
  5. It does not include the budget effect of having a “public plan option.”
  6. There are a bunch of other programs in the bill, including a new disability program and lots of new public health programs.

What we have is a score of two provisions — the new subsidies for people between 150% and 500% of poverty to buy health insurance, and subsidies to small businesses with low-wage employees. We can learn a lot from the table on page 10 of CBO’s estimate. Press coverage is focusing on the “one trillion” number for the net deficit impact, making the common mistake of losing important information by ignoring the gross components of that net number. I am more concerned with the size of the new health care entitlement spending, which is $1.3 trillion over the next ten years.

2010-2019 total
Exchange subsidies 1,279
Small business subsidies 60
Payments by uninsured individuals -2
Medicaid/SCHIP outlays -38
Tax effects on deficit -257
Net deficit impact 1,042

95% of the spending, $1,279 billion over the next ten years, comes from the new subsidies for individuals. This spending does not begin until year 3 (2012), and it’s not fully effective until year 6 (2015). This is a normal effect of implementing such a huge and complex policy — it takes several years to phase in. While the $1,279 billion represents the actual effect on the federal budget of this bill, we can see that the phase-in reduces the cost quite significantly. The green area is 71% of the area under the blue line, which is my estimate of the hypothetical cost of a bill that were it fully effective on day 1. I am not arguing that the “real” cost is the area under the blue line, but instead that focusing only on the 10-year total disguises the true long-term cost of this new entitlement spending. The Kennedy-Dodd draft creates new health spending entitlements that would grow 6.7% per year, faster than our economy, which CBO projects to grow about 4% per year (nominal) in the long run. This means the new health entitlement spending would eat up a larger share of the economy over time.


Effect on private sector spending

CBO has not estimated the effect of this bill on private sector health care spending. That’s not their core mission, but I hope they do so when they have a more fully specified bill. It is a critical metric identified by the President as one of his key tests for an acceptable bill, and it’s a core element of my four-part test for measuring health care cost control.

Effect on health insurance coverage

I will let the pictures tell the story. Here is a before and after of the effects of implementing the Kennedy-Dodd bill. I have chosen 2015, the first year in which the subsidies would have their full effect. Pictures for succeeding years would look similar. I have written before that I think the 51 million figure (now 46 million) overstates the problem to be solved.


And here is the effect of the Kennedy-Dodd bill:


You can see that Kennedy-Dodd would mean that 16 million otherwise uninsured people would get health insurance through the new exchanges, as a result of the subsidies.

In addition, another 22 million people who will otherwise have health insurance will take those subsidies. Three million are a shift from one taxpayer-financed program (Medicaid or SCHIP) to another (the new exchanges). But CBO estimates that 19 million people who are now using their own resources will take advantage of the new subsidies and get health insurance through the exchanges.

This is the problem with creating a new subsidy for something that people are already doing. Analysts say that the new taxpayer subsidies “crowd out” private spending. These people are better off, in that they now have funds available to spend on other things. But if the goal is to reduce the number of uninsured, it is inefficient because half the people benefiting from the new spending are substituting public dollars for private ones.

This makes the cost per newly insured numbers look bad. If you look at just the effects on spending, it’s more than $11,400 per newly insured person. Even if you take into account the higher tax revenues that result from the movement out of employer-based health insurance, the net cost to the taxpayer is still more than $9,100 per newly insured person.

I will soon apply my four-part test to this preliminary Kennedy-Dodd draft, and will update these estimates as the policy develops.

(photo credit: Referees in the Tunnel by Scott Abelman)