How can AGE be less than YRSUSA1?
What am I missing>
After 1900, there is no direct question asking for the number of years since the person’s arrival in the United States (see YRSUSA1 comparability tab); accordingly, YRSUSA1 is derived from the year of immigration (YRIMMIG). The bulk of the cases where AGE is less than YRSUSA1 are a difference of 1 year; this is a logical outcome of deriving YRSUSA1 from the year of immigration (e.g., someone responding to the ACS in May with a December birthday who came to the US the same calendar year they were born would report an age younger than YRSUSA1).
The handful of other cases appear to be related to aggregation in YRIMMIG (it is also possible that the top-code in YRSUSA1/bottom-code in YRIMMIG might affect this in some years, but that was not an issue in the 2018-2020 data which I referenced when looking into this). In addition to being bottom-coded, YRIMMIG bins the oldest years together (but assigns the earliest year as the code for these groups). The categories change with each ACS sample; these changes aren’t clearly documented on the YRSUSA1 page (or the YRIMMIG page for that matter). I am linking the original version of the year of entry variable for the 2018 1-year ACS (US2018A_YOEP); this more clearly shows that the 1930 code is associated with 1930-1934 arrivals, etc. I will ask the IPUMS USA team to revise the documentation for these variables to more clearly flag how aggregation in the earliest years reported in YRIMMIG might affect YRSUSA1.
Thank you . I’m using ACS 1-year 2018-2020.
I was interested in the age at immigration to the US for foreign-born Hispanics. According to YRIMMIG, when subtracting YEAR OF BIRTH (YEAR - AGE) to YRIMMIG, 25% come into the country before age 12 – which sounds very extreme , so I was wondering if I was doing something wrong.
Thank you, hope you can help.
P PInheiro
Paulo S Pinheiro
MD MSc PhD
Research Associate Professor
University of Miami Miller School of Medicine
Sylvester Comprehensive Cancer Center
t305 243 2390
f305 243 6061
Without seeing your code, I cannot say for certain if anything is wrong.
However, I am wondering what about the share of children under 12 seems extreme to you; this is definitely not my area of expertise, but I assume the bulk of migrants would be working age persons (who are also of an age where they are more likely to have children that may migrate with them). A cursory search didn’t turn up any summary statistics about age distribution of the Latino/Hispanic foreign born population (though you may be interested in this book chapter, Pew Research Center report.), or Migration Policy Institute report for related summary information).
I also pulled the 2018-2020 ACS data and did a quick weighted tab of age categories for persons with HISPAN values of 1-4; my results are below (and similar, though not identical to yours).
I aggregated all perwts ACS 2018, 2019, 2020 by age among foreign-born Hispanics.