Open Thread

By Robin Hanson · February 1, 2011 7:00 am · Comments (8) · ·

This is our monthly place to discuss related topics that have not appeared in recent posts.

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • http://liveatthewitchtrials.blogspot.com/ davidc

    What is your opinion on the heritage health prize? http://www.fastcompany.com/1722576/netflix-prize-style-competition-predicts-hospitalizations

    it is a $3 million prize to predict who will need hospitalisation. Do you think it will produce a good predictor? If it does will insurance companies use it not to prevent hospitalisation but to prevent themselves insuring those that will need to go to hospital?

    If the competition will allow insurance companies to not cover those that will need to use insurance does that make it unethical?

    • Chris

      David, failing to insure people who might need to go to the hospital would only make sense in individual insurance markets, and perhaps in the market for insuring small businesses.

      If you are selling insurance to a large organization such as JP Morgan, Ambrose Employer Group (outsourced HR for small business) or the Freelancers Union (purchases insurance for freelancers), you only have the option to offer insurance to the group.

      You might use the predictor to reject the *group* if the group has a much higher than average probability of hospitalization, but this is statistically unlikely except in cases where group membership dramatically increases hospitalization risk (e.g., a group of miners or commercial fishermen).

      • http://liveatthewitchtrials.blogspot.com/ davidc

        Good point Chris. The company may be selling to individuals.

    • anon

      There is no reason to expect that people at higher risk of hospitalization will be denied insurance outright; at most, they will be charged a differential premium depeding on their underlying risk. This is a reasonable outcome, since the point of insurance is to insure random events, not systemic risks. The algorithm allows higher-risk patients to address preemptable hospitalizations by improving preventive care, which is more efficent than spreading the cost to third parties via an insurance market.

      Yes, some people will always face higher costs due to other factors such as their lines of work. But ultimately, this newly-internalized cost will be borne by consumers of the resulting products such as fish and raw materials.

  • Adam

    What happened to the “Why Tax the Rich?” post?

    Has anybody addressed the absurdity of a country with a fiat currency going broke when all their debt is in said currency?

    The most pernicious bias in America is people believing the government’s finances are just like a household’s finances.

    The Federal government does not need money from rich people. They put the money out there to begin with and they make money ex nihilo.

  • Buck Farmer

    Who is doing the best research nowadays on what happens when our various biases, faculties, and cognitive habits and modules interact i.e. what’s the “equilibrium” solution when you stick these all together?

    Similarly, who’s doing the best research nowadays on modeling how all these biases together interact in a multi-person setting i.e. who is taking behavioral economics beyond a series of one-off cases and coming up with testable general theories?

  • Wonks Anonymous

    Alex J seems to be (unintentionally) suggesting a health care study for you to do.

  • Anon

    From Marion Nestle’s Food Politics:

    Michael Taylor tells food industry: FDA intends to enforce new food safety mandate. But it won’t accomplish much.