Tired, poor, on welfare: what is the effect of immigrants on the bottom line - and on something more important?

National Review, Dec 13, 1993 by George J. Borjas

What is the effect of immigrants on the bottom line--and on something more important?

MORE immigrants will enter the United States during the 1990s than in any other decade in the country's history. We are now admitting over 800,000 legal immigrants annually, and at least 200,000 illegal aliens manage to evade the Border Patrol and settle here permanently. It is not surprising, therefore, that immigration has become a charged political issue. Previous debates revolved around the questions of how well immigrants assimilate and whether they take jobs away from natives. The rapid growth of entitlement programs in the past three decades introduces an additional explosive question: Do they pay their way in the welfare state?

As with all accounting exercises, one's assumptions effectively determine the answer. Two recent assessments, by Donald Huddle and Julian Simon, reach the conclusions that 1) immigrants are a huge drain on the system (Huddle), or 2) "immigrants are a good investment" (Simon). Huddle, in a report prepared for the Carrying Capacity Network (an anti-population-growth group), concludes that the net costs of immigration exceeded $40 billion in 1992. Simon concludes that immigrants paid over $2 billion more in taxes than they took out of the welfare system.

The question is a complicated one. In brief, however, Simon is entirely wrong; Huddle is partly correct, though not always for the reasons he gives; and both misunderstand the real nature of the immigrant/welfare problem.

Huddle calculates that immigrants pay less in total taxes than they take out of the welfare system. Extrapolating the results of a study conducted in Los Angeles County to the country as a whole, he claims that immigrants increase expenditures in welfare and other social programs by about $62 billion and that they pay only $20 billion in taxes, imposing a tax burden on natives of $42 billion annually.

Huddle's estimate of net costs seems too high. For example, it assumes that immigrants pay only about 7 per cent of their income in total taxes (less than the California sales tax alone!). Perhaps that is the tax rate that the relatively unskilled immigrants residing in L.A. County face, but it is unlikely to be the tax rate faced by immigrants nationwide. Huddle also claims that for every six immigrants, one native is displaced from his job and joins the welfare rolls, adding $12 billion to the costs of immigration. The evidence that immigrants drive natives into welfare is weak. In fact, during the 1980s, a period of rapid immigration, the fraction of natives on welfare declined. If we reject Huddle's assumptions and assume that immigrants pay 20 per cent of their income in taxes and that they don't affect native welfare costs, we would then conclude that immigrants pay in $57 billion and take out $50 billion, for a net contribution of $7 billion.

The Other Side of the Coin

IN VIEW OF the dramatic changes that have occurred in U.S. immigration since 1970, it is obvious that Simon's calculations (based on a 1976 Census survey) give an erroneous account of the impact of today's immigration. In fact, it is doubtful that immigrants were such a good investment even in the 1970s. Simon reports, for example, that the typical immigrant household entering in 1974 received $1,400 in services and paid $2,700 in taxes, leading him to conclude that immigrants more than paid their way in the welfare state.

Though the calculation is quite simple, it is also quite wrong. A good hint that something is amiss occurs when Simon reports that natives also pay about $1,000 more in taxes than they receive in benefits. But how can this be? Welfare programs, after all, redistribute income from some groups to others. If both immigrants and natives pay more in taxes than they receive in benefits, who is being taxed and who is being subsidized?

The problem lies in comparing total taxes paid with expenditures in a particular set of programs. As we know all too well, our taxes fund thousands of programs. Shouldn't we, therefore, allocate taxes among programs before we determine if immigrants pay their way through the welfare system?

To see how this can lead to very different conclusions, let's update the calculations to 1990, when cash benefits received by immigrants totaled $4 billion. We do not know precisely the extent of immigrant participation in non-cash programs (such as Medicaid and food stamps). However, if immigrants use these other programs to the same extent that they receive cash benefits, they would account for 13 per cent of the cost of all means-tested entitlement programs, or about $24 billion. In 1990, immigrant households received a total income (net of welfare payments) of $285 billion. Recent estimates of the total tax burden faced by U.S. households at various income levels suggest that the average federal tax rate for the immigrant population is around 22 per cent. Including state and local taxes would bring it up to around 30 per cent or around $85 billion.