In many cultures, the term "guest worker" would be an oxymoron. Yet policy makers in both receiving and sending countries seem to like guest worker programs. The hope is that guest workers will fill labor shortage in the receiving countries, and at the end of an employment contract go back home with money and some new skills. There is also a belief that temporary migrants will remit more of their savings back home than migrants who plan to stay on in the destination country (often called the "host country", another oxymoron?).
The problem is that temporary migrants do tend to overstay. However as soon as migrants get residency status, they also tend to go back and forth between their home and host countries. Employers often prefer working with the same workers year after year to working with a new batch of workers every year. And of course, there are several high profile class action lawsuits still going on about guest workers from Mexico and elsewhere who were not paid their dues by the employers or the receiving or sometimes even the sending state. To ensure return, many guest worker programs require a part of wages to be paid back home. Many also propose that guest workers' compensation and benefits be less than those of native workers. There are also instances of schemes for forced remittances.
A key reason such temporary or circular migration schemes do not work is while they are politically tractable, they go against the fundamental forces of supply and demand. Migrant workers will not go back if their services are in demand, and they do go back when they lose employment. They hesitate to go home on losing their jobs, as they are doing now in many countries, if they know that tighter immigration controls will prevent their eventual return to migrant work.
Migration would rise or fall in tune with the economic cycle if labor markets were allowed to operate, without the complications of politics-induced distortions. But such first-best solutions are not easily acceptable in the context of the movement of people, mainly because of the fear that the national identity of the receiving countries may be undermined. Strangely, no one talks about what effect a large exodus of people would have on the national identity of the sending countries (especially when the latter are small nations).
That said, as a second-best solution, augmentation of the so-called guest worker programs can produce dramatic economic gains to the migrant workers and their countries of origin. Lant Pritchett said at a recent conference that allowing 3,000 additional workers from Bangladesh to the US would generate more financial gains in a year than the lifetime gains generated by some microfinance institutions. The Atlantic recently ran a nice story on this topic.