Thursday, February 10, 2011

Hsiao and the Chocolate Factory By Charlie Bucknam 2.8.11

Last May, the legislature passed Act 128 which provided for the expenditure of $300,000 to engage a consultant to develop three options for State-run health care including a “single payer” system.  The Act’s stated primary focus was to reduce health care costs in Vermont.  It assumed that health care costs would be reduced with a government run health care program.  The Commission on Health Care Reform, also created by Act 128, hired William Hsiao, a Harvard economist, in June of last year to develop the three options.   

The long awaited Hsiao Report was delivered to the legislature on January 19, 2011.  Unfortunately, the report was never written to be read by the public.  Indeed, it was not even read by its three authors.  In its 124 pages there are more than 66 typos, several contradictory statements about facts, and footnotes to dubious and outdated sources.  It is replete with cherry picked data,   poorly documented claims, assumptions upon assumptions, and tedious modeling involving so many input variables that the conclusions are completely unreliable. More importantly, the report never analyzes how health care costs are actually reduced, but assumes—and then concludes—there will be  cost containment through government run health care.  If the proposed government run health care system is anything like the Hsiao report, it will be opaque, confusing, sloppy, ineffective and vastly overpriced.   
The report can be divided into two parts:  the first 74 pages of the report are devoted to restating the history of health care in Vermont, reviewing “constraints” on developing a completely new system, and outlining the methodology used to estimate “savings” in health care costs.  The last 50 pages are Hsiao’s three proposals, as mandated by Act 128, for government health care “reform”.  Here I focus on the first 74 pages.     
Hsiao’s history recounts how the legislature and regulators have worked for years to increase government involvement in health care for Vermonters through community rating, Catamount Health, and VHAP, among other programs.  Hsiao reports that these efforts have failed to slow the increase in health care costs in Vermont.   In fact, Hsiao admits that Vermont’s per capita health care costs grew an average of 38% more than the national average every year for the last 20 years, resulting in Vermont going from 42nd in per capita health care spending to 9th  in the nation.  Yet Hsiao never analyzes why our health care costs have increased so dramatically. If he had, he would have concluded that government regulation was the primary reason for increased costs.  His failure to analyze this vital fact of Vermont’s health care history demonstrates one of the major flaws in this report: inconvenient facts which do not support pre-ordained conclusions are ignored.
The report then lists what Hsiao calls the “constraints” on Vermont government run health care—and they are formidable even through Hsiao’s rose colored glasses.  Hsiao notes that not the least of the many constraints to the imposition of State-run government health care is the recently passed Patient Protection and Affordable Care Act (PPACA) pushed through Congress last year.  Any plan that the legislature may adopt will require numerous waivers to avoid having to comply with the many provisions in PPACA.  Nowhere in this report is there an analysis of the administrative and legal costs associated with obtaining all the waivers necessary to implement the Hsiao recommendations.  In addition, no one knows the effects of PPACA.  Robin Lunge, attorney for Vermont’s Banking, Insurance, Securities and Health Care Administration, recently told Vermont legislators that PPACA is “ridiculously complicated”.    These two facts by themselves, suggest that our legislature’s efforts at government run health care are premature at best.  

According to Hsiao, however, there are other daunting obstacles, including the federally mandated ERISA law and regulations, the requirement to obtain waivers under Medicare and Medicaid, the “fiscal” problems of paying for the increased costs; and resistance from “stakeholders” ---doctors who would be receiving lower payments, unions who would lose their gold plated plans, and hospitals who would lose to out-of-state competition.  Why would Vermont embark on costly legal and administrative challenges to PPACA, the crown jewel of the Obama administration’s first two years, without knowing the impact of PPACA?  And why would Vermont spend taxpayers’ funds on legal challenges to ERISA, administrative and legal costs for obtaining waivers from other federal programs, and legal and political challenges from “stakeholders”?   Reading about the “constraints”, one is astonished to find out that Hsiao decides to plow ahead anyway with proposals for government run health care.  Of course, he has no choice.  That was what he was paid $300,000 to do. With Vermont’s budget in substantial deficit, the effort, like the Hsiao report, is an unnecessary and costly waste of taxpayer’s money.
Finally, Hsiao employs “modeling” to determine the cost “savings” from government run health care.  Hsiao employs the Gruber Microsimulation Model (GMSIM), a fancy sounding model which upon further analysis is a mirage.  The model employs innumerable undisclosed “inputs” (data and assumptions), including  assumptions about people’s behavior which are never specified; combined with a formula, the details of which are not revealed, and comes to a---guess what?—conclusion that the author seeks.  It’s like Charlie and the Chocolate Factory:  Hsiao controls the knobs for the data and assumptions and the magic formula to reach the result he wants.

These 74 pages are the prelude to the three options Hsiao presents.  Unfortunately, Hsiao’s options are set forth with the same sloppiness, inattention to detail, and lack of scientific or real world support that tarnishes the first part of the report.  The result is proposals that have no predictive value whatsoever.  I will address the three proposals in another article. 

1 comment:

  1. There are two "single payer" bills before our legislature already...introduced on February 8 barely two weeks after delivery of the Hsiao Report and before any serious critique of the report could be heard. The second instalment on my article will take longer than originally planned because I am attempting to connect the Hsiao Report to the bills in the legislature.