In this first of three related articles on unauthorized immigration, the authors describe what has been, to date, the primary method of estimating the unauthorized immigrant population in the United States: the residual estimation method. This method entails subtracting an estimated population of authorized immigrants from an estimated population of all foreign-born U.S. residents based on national survey data; the residual result is presumed to constitute the unauthorized immigrant population. Federal agencies and independent research institutes have used the residual method to provide measures of the unauthorized immigrant population that are useful in modeling factors affecting Social Security, other government programs, and the overall economy. The authors describe the residual estimation process, present historical estimates it has produced, and review selected recent alternative estimation methods that either refine and extend the residual method or take a markedly different approach.

In this second of three related articles on unauthorized immigration, the authors describe statistical techniques that researchers have used to estimate not only the size but also selected socioeconomic characteristics of the unauthorized immigrant population in the United States. Using data from national surveys, researchers impute the number of likely unauthorized immigrants as well as indicators of their earnings, education, occupation, and industry of employment. The authors describe the imputation methods those researchers have developed, review studies that have employed the methods and summarize selected findings, and discuss research on the correlation between U.S. residential permanence and human capital investment among likely unauthorized immigrants.

In this third of three related articles on unauthorized immigration, the authors introduce a new method of estimating the number of unauthorized immigrants in the United States by exploiting discrepancies between Current Population Survey (CPS) data and Social Security administrative data on Social Security numbers (SSNs). Potential unauthorized immigrant status is indicated when the SSNs reported by CPS respondents and the SSNs recorded in linked administrative data do not match. The authors use the nonmatching SSN data to identify likely unauthorized immigrants and apply a series of logical adjustments to refine the estimated population counts. Although the methodologies of the residual estimation method (described in the first of these three articles) and this new process are entirely different, the two approaches yield similar results. The consistency of estimates resulting from different methodologies suggests that both are likely accurate.