Tuesday, December 3, 2013

I would think if there's one single thing an economist should be able to do better than anyone else, it would be identifying a serious risk of endogeneity and calling it out: John Cochrane edition

Posted by dkuehn at 7:10 PM
John Cochrane thinks two New York Times articles are a sign of cognitive dissonance on health economics.

I do not.

13 comments:

  1. Bob MurphyDecember 3, 2013 at 7:20 PM

    Wait a second... I could see you arguing that there is no even apparent contradiction, or that we need to use counterfactuals. But I'm not seeing how risk of endogeneity is the solution here. Can you help me out?

    ReplyDelete
    Replies
    1. Daniel KuehnDecember 3, 2013 at 7:25 PM

      Countries that spend twice as much as other countries to get worse results tend to pass health reforms.

      People keyed into endogeneity bias when conducting policy evaluations ask themselves "is there some other factor that would make things look bad in a simple comparison of means that comes along with the policy in the first place".

      This is very much like Antony Davies observing that a scatter plot of growth rates and fiscal stimulus doesn't look too good for fiscal stimulus.

      Delete
    2. Daniel KuehnDecember 3, 2013 at 7:26 PM

      I certainly agree with you that we could criticize this without even invoking the more involved issue of endogeneity.

      Delete
    3. Ken PDecember 4, 2013 at 8:37 PM

      The "worse results" is a stretch.

      Delete
    4. JRothDecember 5, 2013 at 4:39 PM

      A stretch? Look at life expectancy, or even height for crying out loud, and most of the OECD countries are ahead of us. What better gauge of the results of a health care system than A. keeping people alive and B. facilitating their healthy growth before adulthood?

      Delete
    5. John SDecember 7, 2013 at 9:47 AM

      Re: Life expectancy--some have pointed out that if you remove violent deaths and traffic accidents, US life expectancy is tops at 76.9.

      http://www.forbes.com/sites/theapothecary/2011/11/23/the-myth-of-americans-poor-life-expectancy/

      http://www.aei.org/files/2006/10/17/20061017_OhsfeldtSchneiderPresentation.pdf
      p. 17: Homicide and vehicle fatality rates

      Americans are also notoriously more obese than other OECD countries. Just my guess, but obesity-related diseases (e.g. diabetes) may contribute to higher health care costs.

      US Male avg height is not noticeably lower than countries with more socialized healthcare systems: level with UK, higher than Canada and France. I'd guess nutrition and genetics are more important factors.

      http://en.wikipedia.org/wiki/Human_height
      http://www.cdc.gov/nchs/data/series/sr_11/sr11_252.pdf (pdf page 22/48)

      Delete
    6. John SDecember 7, 2013 at 9:52 AM

      Arnold Kling's "Crisis of Abundance" is an informative book on healthcare. His conclusion is that the 3rd party payer system (among other factors) incentivizes Americans to overuse "premium (i.e. high-tech) medicine" such as MRIs and other expensive procedures that don't lead to statistically better outcomes.

      http://www.cato-unbound.org/2007/01/07/arnold-kling/insulation-vs-insurance

      Delete
  2. Jonathan Finegold CatalánDecember 4, 2013 at 12:46 PM

    If those price rises occurred after health reform, how would your critique change?

    ReplyDelete
    Replies
    1. Daniel DDecember 4, 2013 at 4:59 PM

      They'd have to have increased at a greater rate than they had before health reform. I generally think it's not a good idea to say either way how health reform has impacted cost until we have at least five years of data post implementation. I'd say the same thing to Krugman as I would to John Cochrane.

      Delete
    2. Ken PDecember 4, 2013 at 9:32 PM

      Why shouldn't we compare them to the decreases currently taking place prior to health care? Do you honestly believe they would have continued rising like they were previously?

      Big pharma has shucked hundreds of thousands of jobs in recent years. Patents on major drugs have expired. Companies are drooling over the possibility of selling to billions in developing nations by manufacturing to a lower price point. Biotech is recovering since the recession crash and poised to take on the role of innovator that big phrama with it's centralized structure never did well. The Antibody Drug Conjugate platform is proving itself and is ideal for the disposeable factory approach to manufacturing and presents a model more appropriate for pharma. Walmart is showing interest in provider space. First, pharmacies and eyeglasses, now vaccines, later I expect nurse practitioner services and more. Companies are coming up with surgery factories in India to have high throughput solutions. ACA can interfere with some of this, particularly by trying to protect current hospital models, but I would not expect the previous cost trajectory to hold with or without ACA.

      Delete
    3. Daniel DDecember 5, 2013 at 11:49 AM

      So why then is the public sector growing more slowly over the past few years than the private sector? Is the public sector just better at incorporating these cost savings into their payment structure? This is why I say we need to have caution and look at longer trends. How exactly is the ACA "protecting current hospital models" by cutting payments to hospitals?

      http://www.whitehouse.gov/blog/2013/11/20/new-report-council-economic-advisers-recent-slowdown-health-care-cost-growth-and-rol

      Delete
    4. dr2chaseDecember 5, 2013 at 10:45 PM

      Define "after". Our former elementary school principal, a great guy who ran a great school, good scores, inclusive, everything you could want in a school, for some reason we asked him one day how long it would take for a new principal to "turn around" a school if it had problems.

      "Five years". For an elementary school. I think our health care system has just a wee bit more inertia than that.

      Delete
    5. Daniel DDecember 6, 2013 at 6:26 PM

      I agree with you, but we shouldn't even start to speculate until 5 years of data is out post the full law going into effect.

      Delete

All anonymous comments will be deleted. Consistent pseudonyms are fine.

Subscribe to: Post Comments (Atom)