The easy way out of a being ambushed by a pro-immigration argument is to point out that the mere existence of welfare will attract immigrants, thus creating havoc in our dear country. Of course, this is said by partisan anti-immigrant groups. As well, such an argument was made by well known economists of the 70s and 80s. Mostly because they lacked the data and their assumption was based on arguing that non-natives react differently to welfare. As if American’s reaction to different state-based benefits were never of concern. This is mostly a native bias that occurs in most people.
First, ignore the immigration problem and focus on for instance American interstate migration. America is a developed nation, a nation that has within most states a substantial welfare base besides existing federal benefits. States not only control their own benefits but also the allocation of some federally determined benefits. So, this makes the data perfect for analyzing and determining if welfare can induce poor people to move among states for more benefits in a developed country.
Using the National Longitudinal Survey of Youth (NLSY) from 1979-1992 Phillip B. Levine and David J. Zimmerman found little evidence that welfare benefit differences among states in the US caused migration among states. They specifically looked at:
The pattern of cross-state moves among poor single women with children, who are likely to be eligible for benefits is compared to the pattern among other poor households
What occurred is the little evidence can be seen of families who qualify for Aid for Dependent children moving versus other poor individuals. And they go further to note:
The are no more likely to move from low versus high benefit states than other poor individuals not eligible for AFDC. Neither are they more likely to move to a higher benefit state. This suggests that either other factors dominate their decision to move or that the perceived costs associated with moving exceed any potential gain from increased benefits.
Levine and Zimmerman produced this paper in 1999 as a answer to debates during this period concerning the failure of the “Great Society” programs. They were addressing the issues when states start retaking federal tasks in welfare and found that this didn’t create the issues predicted by some during this period, particularly the welfare magnet theory.
If one can conclude that welfare magnet theory is unfounded amongst states then I don’t think its hard to apply the same to international immigrants. Considering that for many immigrants the cost of moving is even greater and their living conditions may have become worse with lower incomes, coming to America or Europe for welfare is not exactly a good deal (almost always in the short run).
A survey of various research (section 5.1) on “welfare magnet theory” leads to the conclusion that for different countries at different times the balance of usage of welfare between natives and immigrants changes. Its not really significant and when significant differences occur they can be explained by the changes in welfare systems or changes in labor structure as a result of law or recessions. However, in the survey issues are noted that explain this discrepancy:
One issue with many of the welfare assimilation studies is the lack of separation between welfare eligibility and usage. In particular, most studies do not evaluate the extent to which various immigrant groups are eligible to work and/or to receive welfare benefits in the host country. Changes in work eligibility over time might offer interesting insights in the reasons for the greater reliance on welfare by the immigrants. Similar to the wage and employment assimilation studies, welfare assimilation studies also suffer from the issue of selective re-migration that generally has not been accounted for. In addition, none of the studies have estimated the extent to which welfare dependence is related to the “welfare magnet” effects versus employment obstacles such as discrimination, insufficient language skills, transferability of educational degrees and lack of work permits.
One of the papers, Borjas (1999), makes a case that since immigrants are clustered in high benefit states they are more sensitive to welfare compared to natives. However, this study falls for the criticism set out in the survey above. The issue is that arguing that a state like California, which has many immigrants and provides high levels of benefits attracts immigrants towards welfare is inconclusive. The research seems to fall short to mean anything. Too many variables could affect decisions. Various research came along and noted this, particularly Zacodny (1997) and Kaushal (2005).
What both papers address is the issues in finding proper experimentation parameters. This is because as mentioned before sometimes immigrants incentives and choices are large and expansive and reasoning from large but constricted samples can lead to inconclusive answers and sadly inconclusive answers have been applied to some policy in the US.
What is discovered from Zacodny and Kuashal is that immigrants location choices inside a particular country is related to work (or particular goals) and most importantly the number of immigrants in the city. This actually paints a clearer picture. And explains why New York and Los Angels are still attracting immigrants till today. Both were homes for immigrants of various generations for a long time. The issue with previous research is in my opinion the lacking grasp of determinants of immigration, as well as a the lack of depth in the variables that are overseen. The papers I mentioned that noted the lack of magnet theory both clearly looked at data in the US that looked at the changes following policy that limited benefits to immigrants in some states and while remaining in others.
I think going past this and addressing causes for changes in welfare uses (or more specifically the reaction to discrepancies in welfare benefits among nations) in Europe or Canada is even more helpful in understanding “magnet theory” but still the importance of understanding the various determinants, as well as outside macro reasons for use of welfare are important. Sadly some research overstated the existence of the magnet theory by not understanding this. Many overstated the possibility of stress by immigration on public finances mostly by using the magnet theory. Especially many politicians who till today still wield the magnet theory as the ultimate anti-immigrant tool.
PS I will follow up on Scandinavian/Canadian immigration, and possibly more in depth into pro magnet theory papers. I just had to get this out because this argument still gets used by politicians and even some of my peers.
I should note that some of econometrics in the research particularly Borjas and Kaushal/Zacodny I had trouble comparing just because my experience with econometrics is only beginning since I am only an undergraduate student. But for the most part I attempted to compare the research, even the quantitative parts. Overall, Borjas (1999) and others, in much of their work, seem weak and inconclusive.If you’re an econometrician or know more about labor economics I would appreciate your response.