Employment and Wages of Hmong and Other Southeast Asian Refugees in the United States

Abstract

Hmong refugees fled from Laos to relocation camps in Thailand after the end of the Vietnam War, and thousands later resettled in the United States. We use Census microdata to explore measures of economic progress of working-age male Hmong refugees. To provide perspective, their progress is compared with that of other Southeast Asian refugees and other immigrant groups. Hmong refugees arrived in the United States with lower levels of human capital, so their economic progress was slow. Labor force participation rates were lower, and many earned low incomes in low-status occupations. Hmong men’s wages grew relatively slowly.

Keywords:

• Immigration
• immigrant assimilation
• refugees
• Hmong Americans
• Southeast Asian immigrants

JEL CODES:

• J15
• J71

Notes

1 Dunnigan et al. (1996) provides an excellent overview of the origins and resettlement of Hmong refugees. See also Downing et al. (1984).

2 For example, Faderman and Xiong (1998) include personal accounts from several Hmong refugees.

3 Here, as elsewhere in this study, we use data from IPUMS (Ruggles et al., 2017).

4 Cortes (2004) draws a sharp distinction between refugees—who come to the United States involuntarily and expect not to return to their country of origin—and economic immigrants, who choose to immigrate in search of better economic opportunities and retain strong ties to their home country (to which they might plan to return). In our study, we follow Cortes’s distinction for expositional convenience.

5 Since the microdata samples represent different percentages of the population in different years, the individual weights in the data are not directly comparable across years. Before pooling the data, we normalize the weights so the average weight is the same across years. This approach retains the correct relative weights between individuals within each year in the pooled sample.

6 Our algorithm for distinguishing between refugees and non-refugees is described in the online Appendix. The algorithm uses data from the annual Yearbook of Immigration Statistics from 1996 to 2016 (Department of Homeland Security, 1996-2016). Bollinger and Hagstrom (2008) used a more sophisticated method to estimate the probability that a person arriving from a particular country in a particular year was a refugee.

7 This is calculated as $100·(\hspace{0.17em}\text{exp}\hspace{0.17em}\left({\widehat{\beta }}_{21+\mathit{years}})-1\right)=100 ·(\hspace{0.17em}\text{exp}\hspace{0.17em}\left(0.175\right)-1)$ = 19.1%.