4 ms·

Does anyone here know about the economics of gambling? It seems like the argument Karelis is making would have strong similarities there. I'd be interested to

Does anyone here know about the economics of gambling? It seems like the argument Karelis is making would have strong similarities there. I'd be interested to see if this suggestion is at all valid. To explain, in gambling (think lottery tickets, not poker) you trade $1 for something that actually has less than $1 worth of economic value. There is obviously another component at work (behavioral, not economic), call it hope. If economists have found a way to quantify hope in this context it seems that they could try the same thing with poverty, the results would certainly be interesting.

yummyfajitas 19y ago
I know a little bit. The relationship isn't very close. Gamblers tend to derive utility from gambling, i.e. the act of gambling gives them a rush. The extreme case of this is video games. I'll never win any money by purchasing a Wii (like the lottery, I lose with probability close/equal to 1). But I will derive more happiness from playing Wii sports than I will from the money I spent on it. Similarly, gamblers enjoy playing the game, and this gives gambling positive utility (for them). However, the argument Karelis seems to be making is that the marginal utility of increased income is zero (for some income range which constitutes poverty). People don't tend to do actions with zero marginal utility, so they don't try to increase their income. Basically, earning $15k/year doesn't make you any happier than $10k/year, so why bother to increase your income? (Note: he actually claims that "economics doesn't apply to the poor", which is either stupidity or hyperbole. ) One interesting policy implication of his model (if it is true, which I doubt): taxing the poor is good policy. If a poor person is no happier at 15k/year than at 10k/year, then we should tax all poor people making $15k/year at 33%. It won't make them any more unhappy, whereas taxing a richer person will. It also suggests that any wealth transfers to the poor that dont solve all their problems are wasted (welfare is "all or nothing").
- DmitriLebedev 19y ago
  I'm wondering what kind of utility function is needed to make such a strange MU function... As far as dents on the car are concerned, let me write what conventional microeconomy says. Here is the function to estimate the choices. U = sum(F(x(t))/(1+r)^t) by t from 1 to infinity (or life expectation), where F(x) is a momentary utility function, r is "perceptual" discounting rate (sorry, don't know the English term), which means that F(x) next year is 1+r times less nice than having it now. If you don't repair the car, x(t) will be constant, let's say X. If you do, first period it is x(0) = Y < X, but since t=1 x(t) = Z > X. The problem is that F(Y) has more weight than F(Z)'s. The more r is, the greater the difference and more probable the first choice will win. The problem of poor people is that (1) their r is much more than that of middle class, and they don't value the future, (2) their F(x) (perception of what is good) is different.
  - yummyfajitas 19y ago
    I think the authors ideas are less about time discounting, and more about non-linear relations in quantity of problems. The authors example is a man who's car has many dents. Utility might behave like U(0 dents) = 1, U(1 dent) = 1/2, U(n >= 2 dents)=0. If a car has 5 dents, the marginal utility of repairing one dent is 0 (U(5)=U(4)), so the owner does nothing. Of course, in real life, most utility functions behave in the opposite way. Going from homeless to a studio apt? Major gains in utility. Going from a studio to 1 bdrm? Nice, but not as fantastic.
- mattmaroon 19y ago
  The diminishing marginal utility of money would imply that the difference between $10k and $15k is quite large. Much larger than the difference, say, between $100k and $150k.
  - yummyfajitas 19y ago
    You are quite right, utility is usually steep near 0, and flattens out toward the top. I was simply pointing out that if the author is right, the difference between 10k and 15k would be very small (indeed, quite possibly smaller than that between 100k and 150k). In fact, I'd suggest that diminishing returns explains American "poverty" much better than flat utility curves. 45% of the poor own their own home, 95% are not overcrowded, 75% own a car (30% own 2 cars), nearly all have a TV (25% a big screen). Virtually no one is starving, over 70% of those the USDA classifies as "hungry" are obese. The typical poor family is also supported by 16 hours/week of work. It's quite possible that poverty is explained simply by a low marginal gain from increased work (the poor already hit the point of diminishing returns). All numbers are taken from here (http://www.heritage.org/Research/Welfare/bg1713.cfm http://www.heritage.org/Research/Welfare/bg1713.cfm), who in turn stole them from the census. See also this chart http://www.heritage.org/Research/Family/images/Chart1.gif http://www.heritage.org/Research/Family/images/Chart1.gif of working hours for poor families.
    - mattmaroon 19y ago
      I've actually gone from $10k to $15k once (technically $12 to $17) and it was humongous.
danteembermage 19y ago
One way to explain lottery tickets is to hypothesize that people don't typically process small probabilities well. So Radon looks like terrorist attacks looks like lightening. What people seem to do in practice is to lump all small probabilities into either one quantity that means "small but worth paying paying attention to" and zero. This decision making process works pretty well for day-to-day life, but it can get frustrated when the outcomes are very large, usually dying and big money. So some lottery players approximate the chances of winning to be zero and don't buy; others approximate small probability and get a large expected value of a ticket and buy many.
- mattmaroon 19y ago
  This is actually proven to be true. On the topic of terrorism and small probabilities, over 40,000 people die in car crashes per year in this country. It's essentially as if 9/11 has been happening every month for decades on our roads. But which do people care more about?
- Tichy 19y ago
  I am sure that most people don't get the probabilities right, but maybe there is also something else: if you factor the utility into the equation, the probabilities might not be so bad? E(money) < 0, but E(utility) > 0? As the other reply sad, people get a rush out of gambling, among other things.

yummyfajitas 19y ago

I know a little bit. The relationship isn't very close. Gamblers tend to derive utility from gambling, i.e. the act of gambling gives them a rush. The extreme case of this is video games. I'll never win any money by purchasing a Wii (like the lottery, I lose with probability close/equal to 1). But I will derive more happiness from playing Wii sports than I will from the money I spent on it. Similarly, gamblers enjoy playing the game, and this gives gambling positive utility (for them). However, the argument Karelis seems to be making is that the marginal utility of increased income is zero (for some income range which constitutes poverty). People don't tend to do actions with zero marginal utility, so they don't try to increase their income. Basically, earning $15k/year doesn't make you any happier than $10k/year, so why bother to increase your income? (Note: he actually claims that "economics doesn't apply to the poor", which is either stupidity or hyperbole. ) One interesting policy implication of his model (if it is true, which I doubt): taxing the poor is good policy. If a poor person is no happier at 15k/year than at 10k/year, then we should tax all poor people making $15k/year at 33%. It won't make them any more unhappy, whereas taxing a richer person will. It also suggests that any wealth transfers to the poor that dont solve all their problems are wasted (welfare is "all or nothing").

DmitriLebedev 19y ago
I'm wondering what kind of utility function is needed to make such a strange MU function... As far as dents on the car are concerned, let me write what conventional microeconomy says. Here is the function to estimate the choices. U = sum(F(x(t))/(1+r)^t) by t from 1 to infinity (or life expectation), where F(x) is a momentary utility function, r is "perceptual" discounting rate (sorry, don't know the English term), which means that F(x) next year is 1+r times less nice than having it now. If you don't repair the car, x(t) will be constant, let's say X. If you do, first period it is x(0) = Y < X, but since t=1 x(t) = Z > X. The problem is that F(Y) has more weight than F(Z)'s. The more r is, the greater the difference and more probable the first choice will win. The problem of poor people is that (1) their r is much more than that of middle class, and they don't value the future, (2) their F(x) (perception of what is good) is different.
- yummyfajitas 19y ago
  I think the authors ideas are less about time discounting, and more about non-linear relations in quantity of problems. The authors example is a man who's car has many dents. Utility might behave like U(0 dents) = 1, U(1 dent) = 1/2, U(n >= 2 dents)=0. If a car has 5 dents, the marginal utility of repairing one dent is 0 (U(5)=U(4)), so the owner does nothing. Of course, in real life, most utility functions behave in the opposite way. Going from homeless to a studio apt? Major gains in utility. Going from a studio to 1 bdrm? Nice, but not as fantastic.
mattmaroon 19y ago
The diminishing marginal utility of money would imply that the difference between $10k and $15k is quite large. Much larger than the difference, say, between $100k and $150k.
- yummyfajitas 19y ago
  You are quite right, utility is usually steep near 0, and flattens out toward the top. I was simply pointing out that if the author is right, the difference between 10k and 15k would be very small (indeed, quite possibly smaller than that between 100k and 150k). In fact, I'd suggest that diminishing returns explains American "poverty" much better than flat utility curves. 45% of the poor own their own home, 95% are not overcrowded, 75% own a car (30% own 2 cars), nearly all have a TV (25% a big screen). Virtually no one is starving, over 70% of those the USDA classifies as "hungry" are obese. The typical poor family is also supported by 16 hours/week of work. It's quite possible that poverty is explained simply by a low marginal gain from increased work (the poor already hit the point of diminishing returns). All numbers are taken from here (http://www.heritage.org/Research/Welfare/bg1713.cfm http://www.heritage.org/Research/Welfare/bg1713.cfm), who in turn stole them from the census. See also this chart http://www.heritage.org/Research/Family/images/Chart1.gif http://www.heritage.org/Research/Family/images/Chart1.gif of working hours for poor families.
  - mattmaroon 19y ago
    I've actually gone from $10k to $15k once (technically $12 to $17) and it was humongous.

danteembermage 19y ago

One way to explain lottery tickets is to hypothesize that people don't typically process small probabilities well. So Radon looks like terrorist attacks looks like lightening. What people seem to do in practice is to lump all small probabilities into either one quantity that means "small but worth paying paying attention to" and zero. This decision making process works pretty well for day-to-day life, but it can get frustrated when the outcomes are very large, usually dying and big money. So some lottery players approximate the chances of winning to be zero and don't buy; others approximate small probability and get a large expected value of a ticket and buy many.

mattmaroon 19y ago
This is actually proven to be true. On the topic of terrorism and small probabilities, over 40,000 people die in car crashes per year in this country. It's essentially as if 9/11 has been happening every month for decades on our roads. But which do people care more about?
Tichy 19y ago
I am sure that most people don't get the probabilities right, but maybe there is also something else: if you factor the utility into the equation, the probabilities might not be so bad? E(money) < 0, but E(utility) > 0? As the other reply sad, people get a rush out of gambling, among other things.