Results in Brief

GAO found that quantitative data relevant to a variety of impacts of immigration are available but are frequently not adequate for the purposes of the S.2104 reporting requirements. Expanding information in some available data bases and developing additional data sources could fill many data gaps. However, to produce the strongest evidence concerning the impact of immigration generally would require longitudinal data sources or other similar methodologies that could suitably address cause-and-effect questions. One way to obtain such evidence efficiently regarding at least some impacts would be to graft new data collection onto selected large existing longitudinal studies.

GAO's Analysis

Data Availability and
Adequacy

To assess whether the data needed to answer questions about the six categories of outcomes would be (1) available and (2) of adequate quality, we examined prior reviews of the relevant data bases and drew on our prior studies of immigration. We consulted experts and interviewed officials at the agencies named in the bill.

There are three different methods of examining the "impact of immigra-
tion" in a quantitative way: current status, trends over time, and cause-
and-effect. These methods require different data and study designs. For
example, the size of the waiting lists for admission under the preference
system could be determined at two points in time under a "current sta-
tus" model. If many more data points were available, a "trend over
time" analysis could be done. A "causal" analysis would require addi-
tional data to estimate what the waiting lists would have been in the
absence of the new legislation and to determine reasons for patterns of
change.

When GAO considered the data requirements for the least demanding method of analysis-current status-the data available for addressing 24 of the 26 proposed outcomes were limited. Credible information on consequences that could be clearly attributed to immigration, utilizing a causal model, would be even more difficult to obtain.

GAO identified promising federal agency data bases and determined their
adequacy in measuring immigration impacts using the current-status
method. The data elements most often lacking concern alien status—
both the legality of the alien's presence and the class of admission
(immediate relative, refugee, or whatever)—and date of naturalization
or adjustment to legal immigrant status. The second type of commonly

Executive Summary

missing information is those educational and occupational characteristics which, as criteria, will be weighted heavily in the selection of the new independent immigrants.

Further, information on the impact of immigrants on resources and the environment and on foreign policy is limited. This is in part due to measurement difficulties in those areas. For example, the Environmental Protection Agency informed GAO that the agency was unable to provide information on the impact of immigration on the environment and was not planning to initiate such studies.

GAO finds that if the proposed bill were enacted, it would be possible to obtain some of these data through the augmentation of existing data bases or through additional data collection. The augmentation strategies with regard to current status and trend data include

more integration of different data sources within the Immigration and Naturalization Service;

additional national data collection about immigrant status on vital statistics records, including records of births, deaths, marriages, and divorces;

expansion of data collection by the Immigration and Naturalization Service, especially on education, occupational history, and language skills of immigrants applying for permanent residence or naturalization, and on the potential for future petitioning for relatives (so-called "chain migration") among new immigrants;

possible efforts to encourage states with large immigrant populations to develop compatible data systems; and

improved data on emigration (since the impact of immigration on the United States depends in a significant way on the degree of emigration, or out-migration, of immigrants).

Cause-and-effect data could be produced on some of the outcomes by adding some new data collection to existing longitudinal studies such as the Census Bureau's Survey of Income and Program Participation. It could also be useful to initiate other longitudinal studies aimed specifically at selected immigration issues.

GAO finds that available data are frequently not adequate for the purposes of the reporting requirements under S.2104. Therefore, GAO recommends that the linkage between impact measures and the process of

Agency Comments

periodic review of the numerical limits be removed from proposed legislation. However, if the Congress views this linkage as critical, then a variety of steps (such as those previously cited) should be undertaken to strengthen immigration data. A number of federal agencies could be involved in such a data improvement effort.

In the interest of time, the requester asked that GAO not seek agency comments on this report. However, GAO did discuss these findings and general conclusions with agency officials and took their observations into account.