Welfare for Lazy BumsPosted: July 3, 2011
Welfare is a complicated issue. On the one hand, most people feel really bad if there are some poor folks in the community who just can’t make ends meet on their own, and need a helping hand. On the other hand, if we give too much of a handout, then there won’t be so much of an incentive for everyone to get a job, and you’ll have people lining up for free money, overwhelming the system. Besides, why should we help people who could be helping themselves?
Often, the people who are on welfare lack the skills to get a decent job (this is even aside from those who can’t work due to illness or old age). Perhaps they played basketball all day when they should have been in high school. Let’s run with this, even if it might not be true (it could be that people are just born into circumstances beyond their control). We’ll use a model with two periods, representing stages of life. The first represents high school, and the second represents the rest of life. The choice of whether to take school seriously occurs in the first period. The choice of whether to provide welfare to the indigent occurs in the second period. The bums prefer getting welfare to getting an education (E) leading to a job that pays (S), but much prefer getting an education and not starving (S – E) to not getting welfare (normalized to 0). Society would not want people to starve, but any assistance that they provide comes at a significant cost (C), which makes assistance to people who don’t desperately need it (which gives them payoff S – E + aW) prohibitively costly.
Suppose the bums, in the first period, decide that school sucks and they’ll instead play guitar – that’s where all the money and girls are, anyway, right?
Except they are no good at guitar, even though they think they are. So they flunk out of high school, and end up with no job, no gigs, and no money (and certainly no girls). So in the second period, these guys are already bums – they’ve missed the boat to a large extent (they could go for something like a GED, but that requires a program that’s sorta like welfare in the meantime, anyway), whether or not there is welfare. Since society prefers that the bums do not starve, any subgame-perfect Nash equilibrium in this model requires that, if the bums have not graduated high school, then society must provide them with welfare. Of course, this will incentivize them not to work in the first period, but this is unavoidable.
Now, one might object that this does not take into account how future generations will see this system (which we model by repeating the two-stage game above indefinitely). After all, if they can get away with not working, they will not obtain education now either, creating problems in later periods. However, if they see that the bums are “punished” by not being provided with welfare, they will know that the system isn’t messing around – when they say to go to school, they mean business. Thus the benefits to society from later periods, when they won’t have to pay welfare and everyone works, outweigh the suboptimal stage game solution of not providing welfare to the bums.
This may be true, or it may not, depending on a bunch of factors. First, it’ll depend on how much we value the welfare of future generations. Obviously, there is some discounting, the exact value of which (δ) will help determine whether there is a subgame-perfect Nash equilibrium without welfare: if δ is large, then we care more about future generations, and so we are willing to forgo welfare in the current generation to teach the young’uns a lesson; if δ is small, then we care more about the starvation of the bums than the lessons we can teach.
Second, it’ll depend on how much disincentive there will be to work by providing welfare. This might depend on how much welfare gives – if it gives them gold crowns, then there will probably be more people who will punt on school. This gets into some rather complicated economic issues, such as the shape of utility curves, optimal taxation, etc. Needless to say, the issue can’t be resolved here. Yet by the simple example here of the extensive-form game model of welfare, we can provide a new lens by which to view the problem.