Regarding the memo from the CBO saying that raising the required medical loss ratio for health care plans would result in the CBO counting the insurers as branches of the federal government, Cato's Michael Cannon writes:

It's about how congressional Democrats have very carefully hidden more than half the cost of their health care bills.  
 
The Clinton health plan would have mandated that individuals and employers make payments to non-governmental "health alliances," which would pass the money through to private insurers.  In its 1994 score (http://www.cbo.gov/ftpdocs/48xx/doc4882/doc07.pdf)  of the Clinton health plan, Bob Reischauer's CBO included those mandated "private" payments in the federal budget - i.e., as federal revenues and federal expenditures.  
 
Yet you'll notice that the CBO's scores of this year's bills do not include the costs of similar individual/employer mandates as federal revenues or federal spending.  
 
My read of the CBO's score of the Clinton health plan is that the private-sector mandates accounted for around 60 percent of the Clinton health plan's total cost, the remainder being (traditional) government spending.  So how is it that the CBO made the full cost of universal coverage apparent to the public in 1994, but may now be revealing only 40 percent of the cost?
 
I think the answer is that congressional Democrats have very carefully tailored their individual and employer mandates to avoid CBO's definition of what shall be counted in the federal budget.  While at CBO, Peter Orszag altered the agency's orientation to make it more open and collaborative.  One of the things about which CBO has been much more open is the criteria it uses to determine whether to include mandated private-sector spending in the federal budget.  CBO even published a paper (http://cbo.gov/ftpdocs/102xx/doc10243/05-27-HealthInsuranceProposals.pdf)  on the topic.  If you read this profile (http://www.prospect.org/cs/articles?article=numbercruncherinchief)  of Peter Orszag by Ezra Klein (now of the Washington Post), I think you'll agree that this was likely an area of collaboration.  
 
The CBO's definition (http://cbo.gov/ftpdocs/102xx/doc10243/05-27-HealthInsuranceProposals.pdf) of which mandates shall count as federal revenues/expenditures is fundamentally subjective and arbitrary.  (I can get into that, if you want.)  But the problem is not so much the definition CBO has adopted.  Nor do I have a problem with CBO collaborating with legislators during the drafting stage.
 
The problem is that crafting the private-sector mandates such that they fall just a hair short of CBO's definition does not reduce those mandates' cost, nor does it make those mandates any less binding. But it dramatically reduces the apparent cost of the legislation.  It's the reason we're all talking about an $848 billion Reid bill, rather than a $2.1 trillion Reid bill.
 
The MLR memo is the smoking gun: it shows that this is what they've been doing with CBO all along.  Proposals that would result in a complete cost estimate are dropped.  Because we can't let the public see how much this thing really costs.

Update:  Cannon has turned this email into a blog post here.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.