Heavy Overthinking Permeates Flawed Academic Analysis of RomneyCare

For whatever reason, the author of the blog "A Certain Enthusiasm" has spent a lot of time trying to prove RomneyCare a success. I can't figure out whether he's a MittManiac or a lefty but he keeps trying to squeeze the old five pounds of you know what into the one pound bag. His current premise -- like others trying to rationalize the failure that is RomneyCare -- is that it is a success because 'private healthcare premiums' in Massachusetts are growing more slowly than elsewhere in the country.  It's the we're not as bad as them version of good.

Given its length and apparent independent thinking, his theory deserves a thorough review.  I have reviewed it. He's wrong. See below.  But I applaud his effort although it illustrates the overuse of statistical devices and overthinking of the derived "estimated results" by academics. (There is no "about us" link on the blog that I can see so I don't know if the author is an academic. But he makes my point about the academy nevertheless.)

• Why use statistical projections as derived estimated results when you have almost 100% complete census data available? 
• Why regress the flimsy data points the academics tend to depend upon when you can look back at complete data, and regression is basically for looking forward? 

In particular, I suggest his analysis has a problem with both source data and some of its analytical premises and devices.  It's hard to tell what he's using as his data soruce from his post because the link leads to a title page with about a billion spreadsheets branching off it.  But if he is using the same data set he used to float these concepts a month ago, he's using some survey-based stuff from the Census bureau.

And the survey -- if that's what he is still using -- counts nursing home spending.  RomneyCare had nothing to do with nursing home spending. He told me when I asked him about that he just meant to give a feel for what is going on in Massachusetts, not to specifically review RomneyCare.  But he seems to have changed his tune in the last month.

A better source of information on private premiums would be the Massachusetts Department of Healthcare Finance and Policy (DHCFP) web site which essentially depends on a census of policies rather than a survey of them. In particular, the author should have looked at the May 2011 Premium Trends report for better analytical premises related to

• Massachusetts' "merged market" ratings system, particularly relative to age
• Very important adjustments for the reduced benefits we have suffered here in Massachusetts since 2006
• The effect of the introduction of the Young Adult Plan if any 
• The introduction of exchanges as a means by which small employers can facilitate insuring their employees
• The effects of the "merged market" itself (which happened coincidentally with RomneyCare)

In addition, if the Expenditure Panel he mentions as a source is literally an estimate of 'private' spending (as I said, his source data is not clearly linked) rather than premiums, that alone could explain the difference between the author's analysis and conclusions and the state's own (and everyone else's) conclusions that premiums have skyrocketed in Massachusetts since 2006 when adjusted as described above for what is actually happening here.  A private spending estimate -- if that's what he used -- would not capture the billions of dollars shifted from private to public by federal and state subsidies of Commonwealth Care (the literal RomneyCare). 

 

And to my point about academics overthinking things, why back into an analysis of premium trends by starting with a partial view of spending when multiple sources publish the actual premiums?

 

Unfortunately DHCFP has not published 2010 data yet and the author would probably not be able to correct for all the variables that the state has appropriately corrected for in its own analysis.  But the state itself has what it calls private premiums going up over 7% a year since RomneyCare became effective or 50% faster than the author's estimate. Remember, this report is done by Democratic politicians who -- like the author apparently -- desperately want to 'prove' that RomneyCare, its then latest attempt at 'reform,' worked.

 

It is also most important to note that the 2006 changes in Massachusetts law were the fourth major 'reform' in Massachusetts since 1988, that there have been two major 'reforms' since 2006 and that another 'reform' is in the works for 2012.  Wouldn't you think that if the 2006 reform was the success the author apparently believes it is, that the Democrats would stop re-reforming it?

-- Dennis Byron