Date of Award


Document Type

Open Access

Degree Name

Bachelor of Arts



First Advisor

Younghwan Song




wage, sample, analysis, selection, bias


Due to the increasing flow of immigrants into the United States in recent years, numerous researchers have been examining the socioeconomic characteristics of immigrants including wage differential. However, the majority of such wage analysis raises a key issue of the sample selection problem. This problem occurs when one has a non-random sample by ignoring the decision process to be participants of the sample, and it has a potential danger of a biased and inconsistent estimation. In the view of this, it is important to estimate the decision factors of employment status – being a wage earner or self-employed – before the wage analysis. The regression analysis follows that of Lofstrom (2002). He estimates the earnings of wage earners and the self-employed by correcting selection bias using a method introduced by Heckman (1979). Using the data from the 2003 and 2013 American Community Survey PUMS, my study aims to analyze the economic performance of immigrant wage workers by country of origin while correcting for sample bias and updating the findings of Lofstrom. The estimates find that immigrant enclave and earning differential variables have significant effects on the probability of being a wage earner. The negative sign of the correction term suggests negative selection into the wage/salary sector.