False Figure Fuels Furor

The debate over providing public education to children of undocumented immigrants centers on the “$2-billion subsidy� borne by California taxpayers. Those who quote this expense assume that it is the true cost of educating these children and will be the net savings to the public education system from barring illegal immigrants. As an economist, my curiosity is piqued by the underlying assumptions used to construct this number.

The basis for the $2-billion estimate comes from projections by the Immigration and Naturalization Service and the Census Bureau. Using these data, a common estimate for the number of undocumented children enrolled in California public schools is about 300,000. This number is then multiplied by the per pupil average costs, which for large urban school districts have ranged between $6,000 and $7,000 a year.

The problem with this estimate, however, is not the number of undocumented children enrolled in public schools, but the use of average cost data rather than marginal cost data--that is, the additional costs of educating the undocumented students. This is a major methodological flaw. Average costs include fixed costs such as buildings, debt payment on bonds and in some instances, multiyear negotiated contracts for salaried employees and variable costs such as classroom materials and supplies. In the case of K-12 education, there is a considerable amount of fixed cost hidden in the average cost estimate and a limited amount of variable cost. Therefore, eliminating access to K-12 education to undocumented students will not result in a $2-billion cost savings. The fixed costs of these school districts will remain. So the closings and teacher layoffs required to save this amount of money will not happen, because these fixed costs are required to meet the needs of the remaining students.

So what does this supposed $2-billion savings really mean? We can at best use it as a political red flag and hope that the federal government responds by providing us with more resources. As for the incremental impact of immigrant students, researchers have begun to sort out some of the additional costs to targeted programs like bilingual education. A preliminary study under way by the Urban Institute suggests that the incremental cost to the core curriculum for bilingual education may be only $1,000 per pupil. Using this approach would drastically reduce our cost estimate from $2 billion to $300 million or less.

With meaningful marginal cost estimates, we could then shift the cost burden to the real culprits involved in the undocumented migration flow: employers. Rather than making educational access the issue in immigration control, we could use the marginal cost estimates to develop targeted taxes to employers who hire undocumented workers. This incremental tax would be based on the number of projected undocumented employees. The tax would be applied across the board in an industry such as farming or garment manufacture, with individual employers granted exemptions if they can certify the status of their employees. The new tax would go to the local school district.

The beauty of using the tax system instead of the educational system as one mechanism for controlling immigration is that it targets where the hidden subsidy is and transfers that subsidy to where the real cost is borne. It reduces the cost for enforcement of immigration law from the government to the private sector. In sum, this is not only a better strategy for controlling the flow of undocumented immigrants, but also a way to shift the responsibility squarely where it belongs: on employers who choose to hire undocumented employees.