I read somewhere that a physicist could predict the movements of a single ball bouncing around inside a small, frictionless box, say in outer space (removing gravity helps to imagine this). All he needs is the state of the system at one point in time, and he can tell you precisely (that is, within a small margin of error), where the ball will be a year from now. Put a second ball in the box, allowing for frequent collisions, and he can still do it. But this would take enormous computing power. Put a third ball in the box, and a computer the size of planet earth couldn't do the job. I don't know if this is true, but if it is it stands as a good illustration of chaos theory. Beyond a certain degree of complexity, even systems consisting of very simple parts become unpredictable.
This occurred to me as I was thinking about immigration reform. The economic consequences of immigration are usually described in this way: the supply of cheap labor lowers the costs of labor at the entry level. This is bad for the least competitive native workers. It is good for the immigrants, for the businesses that hire them, and probably for the American economy as a whole. This helps to explain why developed nations don't really try to control immigration: those who benefit from it are more numerous and more politically powerful than those who do not. However, say advocates of reform, reducing the flow of immigrants to a trickle would do wonders for the least advantaged native workers, and legalizing undocumented workers would mean a lot more people paying their full share of taxes, etc.
An article in the LATimes suggests that this model is too simple the capture the dynamics of immigration reform.
Legalizing the immigrants already here will move some of that competition up the labor ladder. Hanson gave the example of hotel workers. Many are illegal, which means they're stuck in the worst jobs. As their status changes, so will their positions. The result: more people will compete for the post of, say, assistant manager.
Suppose we shut off illegal immigration completely (a goal which seems unrealistic), and legalize the existing immigrant population. We then have only to calculate the effect of millions of newly documented workers in the work force. These workers will suddenly be able to compete for jobs that have hitherto been closed to them. This would have the effect of driving down wages in middle tier jobs. Unions are going to be vociferously opposed to this.
It looks to me as though this has to be good for entry level workers, at least at first. As many of those struggling with hotel mattresses or walking in fruit groves leave to take better jobs, the wages of those who remain will have to rise. But that means that the magnetic pull of jobs seeking workers will tug irresistibly against any legislative/security force barrier we try to erect across the border. Unless that barrier is very effective, we may end up reducing average wages of workers currently protected against competition from immigrants without in fact reducing labor competition for entry level workers. We will end up with the same problem we have now in short order.
The truth is that no one knows what the effect of any proposed reform will be. There are too many balls bouncing around in this box, and the box geometry is just too complex for prediction. But if we could establish some measure of control over the border, as well as enfranchising the existing population, then at least we would have our hands on the levers. Moreover we would be able to measure the effect of reforms over time. Maybe that's the best we can hope for.