government on their debt is 3 percent per year, and this rate is also used to calculate the net present value of an incremental immigrant. Later we will test our results with different budgetary assumptions and with interest rates of 2, 4, 6, and 8 percent.

We assume that these same age profiles, shifted upward as described, also describe the tax payments and benefits received by descendants of immigrants. The descendants are projected based on the fertility of immigrants, which is assumed to converge in two generations to that of the general population, following the assumptions described in Chapter 3. Over their life cycles, 30 percent of immigrants are assumed to return to their countries of origin (Duleep, 1994), taking with them their children under the age of 20 born in the United States. We take into account that these return migrants may or may not be entitled to Social Security benefits, depending on the length of time spent in the United States.21

Baseline Results

Definition of Baseline Scenario and Alternative Scenarios

We will carry out the basic calculations for a baseline scenario, and then examine how results change when we vary certain assumptions, one at a time. Here are the baseline assumptions, and the variants (in parentheses):

  • Starting in 2016, and thereafter, fiscal policy will hold the debt/GDP ratio constant at the level of 2016. (Alternatively, the ratio will be constant from now on, or it will never be constant, and taxes and benefits will simply rise with productivity at 1 percent per year.)
  • This budgetary adjustment will be achieved by a 50-50 combination of raising taxes and reducing benefits. (Alternatively, all the adjustment will be made to taxes, or to benefits.)
  • The real rate of interest is 3 percent. (Alternatively, we use 2 percent and


growth for the first few decades. Pension costs require special treatment. We assume that the current number of state/local pensioners was set by state/local employment 30 years ago, and we shift the age profiles of pension receipt upward each year between 1965 and 1994 by the ratio of state/local employees in each year to the number in 1965. (The difference between the mean age of Social Security beneficiaries and FICA taxpayers is 30 years.) The per capita number of state/local employees grows from 4.9 percent in 1970 to 6.1 percent in 1992, for example. After 1994, we allow the age profiles of state/local pensioners to shift upward at I percent per year along with all the other age profiles, assuming that growth of state/local government relative to the population size has run its course, and that further growth will just keep pace with population.


Unfortunately, we did not have time to incorporate into our analysis the legislated change in the normal retirement age for Social Security. We believe, however, that doing so would make very little difference.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement