The disagreement between George Borjas and me on the desirability of guest workers is an interesting one to expand upon, not because this will change either one of our minds, but because it may help others evaluate their own views. So here are a few points.
1. First, George and I appear to be more or less in agreement about the basic economics. Our best point estimate has to be that guest workers will create a small net gain for the U.S. economy, with a large margin of uncertainty on both sides. They would also likely exert downward pressure on U.S. wages, especially in the short run. On the other hand, the gains to guest workers themselves is quite large. If you are in gross disagreement with any of these points, do not look to either one of us for support for your position.
2. Now assume you care only about the welfare of Americans, and that you put zero weight on the welfare of foreigners. In view of the small net gain and the distributional impact, whether you think a guest worker program is a good idea or not depends on how you weigh these two against each other. This is not an economic question per se. But here too I think George and I agree. Leaving foreigners out of the picture, the net gains are not worth the price to pay for the distributional "cost."
3. But, and here is where we begin to disagree, I do not find putting zero weight on foreigners to be an acceptable moral position. I am willing to accept--and in fact would advocate--a weighting scheme that overweights our own citizens relative to other countries'. But a weight of zero is surely unacceptable.
4. This has bite precisely because the gains to foreign workers are so huge--given the big differences in labor productivity in sending countries. Given these large differences, you would have to put if not zero, a near-zero weight on foreigners to still think that a guest worker program is a lousy idea. See Notsneaky's illustrative computations.
5. If one difference with George seems to be the weight you put on foreigners, another is whether a guest worker program can work as advertised. George does not believe you can enforce return, through appropriate carrots and sticks. Note that this objection is largely immaterial unless once again you put near-zero weight on foreigners. The logic of the argument so far does not rely on the migration to be temporary. It relies on the gains to foreigners, relative to distributional effects at home--and the permanence or temporariness of the program does not affect this calculus much.
6. Now I do think that a guest worker program where most people eventually return to their home countries, their place taken by new workers, is better than a once-off program because this spreads the gains around and also enhances the benefits to the home countries of the workers--it is the icing on the cake. As I said, George does not think this can be made to work. I believe it is worth a try. The role of the policy analyst is to design the incentive schemes and institutional arrangements that will allow the $100 bills left lying on the pavement to be picked up--it is not to say, "yeah, good idea in theory, but we cannot make it work in practice." My guess is that George's take on this would be less categorical if he were to put a non-zero weight on foreigners.
To sum up, the key differences seems to be the weight we put on foreigners--near-zero versus something somewhat larger--which is a moral, not economic consideration. If you are willing to countenance a non-zero, but still small weight, then you will be in favor of the guest worker program and more likely to believe that it is worth experimenting with carrots and sticks to overcome the practical difficulties.