3 ms·

Reducing variance is worth real money to people, since the marginal utility of the first dollar is much higher than the 1000th.

• incongruity 10y ago
  Adding to that - it's the flaw of averages at work in many of those examples. The listed "expected value" in some of those examples could never actually be a real outcome of the examples - it is completely rational to opt to take a 500 loss rather than risk a 1000 loss if you have, say, only 750 to your name and cannot be overdrawn (see debtors prison or whatever incentive one would like to imagine). The expected value of the $500 loss only happens in one option. In the $0 or $1000 loss, if you had repeated draws over time it would indeed trend towards an average of a $500 loss but in a single draw situation, you can never achieve the average so you absolutely must view it as different from the flat $500 loss and compare that to your imagined circumstances in this thought exercise.
• geomark 10y ago
  Yes. There are many cases where a lower expected value is a rational trade off for reduced variance.
• FullMtlAlcoholc 10y ago
  I completely agree. A bird in hand is worth 2 in the bush and all that. However, even today economic theories do not take this into account and assume rationality === choosing the highest expected payoff. I.E. we're all statistical savants
  • ericd 10y ago
    Yep, your main point that economics' rational man is a complete fiction is absolutely correct.
    • HarryHirsch 10y ago
      He's a convenient fiction though. Since to homo oeconomicus a certain expense of USD 500 and the 50 % chance of a USD 1000 expense are the same, expect him to be rolled out to sell you high-deductible health insurance and health savings accounts. H. oeconomicus is the model human, consequently he isn't cashflow constrained, and the fact that HSAs double as tax shelters for the wealthy doesn't concern the Ideal Man either. University economics is a religion, not a science, it offers up myths instead of testable predictions on observable reality. High finance is a different deal, but they surely don't build their models on university textbooks.
      • slededit 10y ago
        > University economics is a religion, not a science That's a bit harsh. Is every field that employs simplified approximations of complex systems a religion as well? With that definition even physics is out and that's one of the more concrete sciences.
        • FullMtlAlcoholc 10y ago
          At times it does appear to be a religion because of the cult of personality that proliferates (remember the Greenspan years?) and, most significantly, the blind faith and slavish devotion to free markets being a solution to everything. What I find most off-putting are those who advocate for wholesale changes in our society on little more than hypothesis and faith instead of running experiments starting at the smallest reasonable scale and then successively working your way up from that. It should at least pretend to try to applying basic scientific methodologies to its practice.Psychology has evolved to being a rigorous, evidence based social science. That psychology is not one of the foundations of studying economics is also problematic too.
        • mcguire 10y ago
          If it does not admit that it's simplified assumptions are simplified assumptions and seek to improve them, then yes. Which does not apply to physics. Economics has a weird tendency to invoke "outside shocks" to explain why their models fail, and seems to have difficulty identifying those shocks.
• kirrent 10y ago
  For a real world example, consider government bonds which have relatively low returns but are almost guaranteed.