3. Legal Migration and Illegal Migration: The Effectiveness of Qualitative and Quantitative Restriction Policies

There are, of course, examples that have reach the opposite conclusions. Among them, Galor (1986) is the most important study to conclude that, although migration is harmful for the people of the host country, it will not affect its economic welfare.

All are static analyses. The framework of the pioneers (Berry and Soligo 1969) is a two-factor one-good model. Rivera-Batiz (1982), Quibria (1989), and Kondoh (1999) adopted a two-factor two-good model, where one is nontradable. Wong (1995) extended the m-factor n-good model.

Some countries, like the United States, have adopted quantitative restrictions. Recently, however, these countries have started using “ability and future contribution to the host country” as the main criterion for granting immigration permits, rather than the traditional “having family ties” (see Chiswick 1988).

Ohta (1990) summarized these two papers and concluded that internal enforcement is better than border enforcement in attaining the two inconsistent targets of a larger national income and a more accurate disclosure of illegal migrants. Yoshida (1993) concluded that tighter restrictions of illegal migrants would reduce the economic welfare, not only of the home country, but of the rest of the world. Djajić (1987) analyzed the cost of border enforcement policy in equilibrium by adopting a dynamic Harris–Todaro model.

In this case, he is detected upon arrival and so cannot get the higher wage in the host country at all. He is fined and deported and has now lost the precious chance to improve his skill, which was available to him if he had not migrated.

Concerning the theoretical analysis, O’Connel (1997) proposed “speculative migration,” in which greater uncertainty about the destination region leads migrants to “try their luck” in foreign job markets.

Our postulate about internal enforcement differs from that of Djajić (1997), where it is the employer of the illegal workers who pays the fine.

In reality, many countries adopt a fixed penalty by fining. Of course, there are some countries that do not impose any monetary penalty; however, if we consider the period in prison before deportation, with the obvious loss of money-earning opportunity, we may regard that as equivalent to paying a fine. In countries where the fine depends on the period of one’s illegal stay, it is possible that some illegal workers would be induced thereby to return home freely, but this feature was not a part of our model.

See Atkinson (1973) and Bhagwati and Hamada (1982).

Generally speaking, the number of newborn individuals is endogenous, because it depends on how many workers have migrated to E. However, in our model, we assume that workers decide, in the first period of their lifetime, whether to migrate or not. They then also produce their children (two per couple) who, as newborn, must stay in their birth country regardless of whether their parents migrate. The children also will decide their own futures during the first period of their lifetime (which is the second period of their parents’ lifetime). See Kemp and Kondo (1989) for an example of an endogenous population growth formation.

See Grossman (1983) for an example of a similar approach.

The return travel cost is often shared by the government of the host country that enforces deportation; we may consider that this cost is included in the fine.

This assumption excludes the possibility of middle-aged illegal migration.

In the usual case with risk, there is an insurance system to hedge the worst losses. But in our model, we consider an illegal migrant who tries to get an informal job in the country E. In reality, the black-market brokers who help a worker to migrate illegally do not pay any interest on such insurance systems, because there are plenty of potential candidates who intend to migrate without insurance. Of course, the formal banks or insurance companies avoid dealing with illegal workers because of their illegality. Instead of being financed by the formal financial or insurance sector in the host country, most poor migrants are, in reality, financed by their immediate or extended families. Lucas and Stark (1985) showed that migration is a credit insurance between the migrant and his immediate family. Ilahi and Jafarey (1999) showed that 58 % of Pakistani migrants borrowed from their extended family and that 46 % of the migration costs are financed by the extended family in Pakistan (38 % in the Philippines). It is worth noting that those who are financed by their families are fortunate and that there are plenty who have no such connection. As mentioned in Sect. 3.1, the poorest people in developing countries cannot afford to pay a few hundred dollars for travel costs; they must accept their fate with resignation.

Y ₄ > 0 is the stability condition of i ^**.

If a border enforcement policy is adopted by E, illegal migrants must pay increased travel costs, because of the extra charge for disguise. Furthermore, we would have to introduce into our model the possibility that a stronger border enforcement would reduce the number of workers who is permitted to enter E.