(THIS POST CONTAINS ERRORS AND HAS BEEN SUPERSEDED. See the corrected estimate here.)
A recent blog post at the Niskanen Center website argued that our estimate of the possible number of birth tourists was far too high. While that is certainly possible, the Niskanen analysis, authored by Jeremy Neufeld, demonstrates no such thing. Its error seems to be the result of two problems: a misunderstanding of the data combined with an implausible assumption.
In our original analysis from December 11, 2019, we first subtracted the number of births to foreign-born women living in the United States as shown in the American Community Survey (ACS) from births to foreign women as reported by the CDC on birth certificates. There were 51,327 more births to foreign-born women in the CDC data. This number represents our baseline estimate of the number of foreign women who might have had a child and then left the country. Second, we then subtracted from this number the possible number of births that were to foreign-born naturalized citizens who left, leaving our estimate of 33,000 possible birth tourists.
Niskanen's Neufeld countered by claiming that there were more likely 1,800 birth tourists. As far as we can tell, there were two primary mistakes in Neufeld's analysis, which we discuss below.
The Niskanen Analysis Seems to Have Used the Wrong ACS Question. The first criticism Neufeld makes of our work is that we underestimated the number of births to foreign women in the ACS. He claims that the number is really 870,970, while we find 845,896.1 Subtracting Neufeld's ACS number from births to foreign born women in the CDC data produces an estimate of only 26,253, which he reports. This is much less than the 51,327 we report. Neufeld's figure is not correct. We are not 100 percent certain, but it seems he mistakenly used the year-of-arrival question in the ACS to identify new foreign-born mothers, not realizing that many U.S.-born women also give a year of arrival in the ACS. This is one reason he ends up with a much higher number from the ACS and a much lower number of potential birth tourists. The citizenship question should be used to calculate births to foreign born women, not the year-of-arrival question.
In the 2017 ACS, there were 871,304 births to mothers who gave a year of arrival in the ACS, and if we exclude the 334 who gave "other" as their place of birth then the total matches the 870,970 figure Niskanen reports. This appears to be what Neufeld did to arrive at the figure he reports for the number of births to foreign-born women. Subtracting this much larger number from the 897,223 births reported to foreign-born mothers in the CDC data leaves 26,253 potential birth tourists, which Neufeld reports.
First, excluding the 334 mothers who reported "other" as a place of birth, if that is what he did, makes no sense because all of these new mothers indicated on the citizenship question that they are not American citizens, which by definition means they are foreign-born and should be included. On the other hand, 25,408 new mothers in the ACS were actually born in the United States and should not have been included by Neufeld.2 His inclusion of these native-born Americans results in a much higher number of births to foreign-born mothers in the ACS and is the primary reason he arrives at a much lower number of potential birth tourists when he subtracts the number of births reported by the CDC. If you include the 334, but exclude the 25,408, you get 845,896, which is the correct number of foreign-born mothers from the ACS citizenship question.3
If Neufeld did use the year-of-arrival question in this way, which seems to be the case, it is clearly a mistake. In any event, we have re-created the 2017 public-use ACS data and re-run the numbers and 845,896 is the correct number of births to foreign-born mothers in the ACS. This is the number that should be compared to CDC birth records and the difference is 51,327, as we had reported.
The Niskanen Analysis Mistakenly Excluded Potential Birth Tourists. The second part of the Neufeld's blog post reduces his much lower initial estimate further by assuming that 93 percent of the 26,253 potential birth tourists are either citizens or long-term residents.4 He states that "new mothers who CIS identifies as potential birth tourists had already been in the United States for 13 years, on average." He also states in reference to new foreign-born mothers that, "the ACS also reveals that over 90 percent had already been in the United States for at least two years."
It is not clear why he makes these observations. New foreign mothers in the ACS have to be living in the United States to show up in the survey and cannot therefore be birth tourists. As discussed above, the birth tourist estimate is based on people not found in the ACS, but who do show up in CDC data. The CDC only reports very limited information about new mothers' places of birth, which allows researchers to identify those born outside of the United States, but little else. CDC data does not include any information about citizenship or length of stay in the United States of new foreign-born mothers, as this information is generally not collected on birth certificates. Looking at the average length of stay of new foreign-born mothers in the ACS tells us nothing about those who are missing from the ACS.
Apparently Neufeld is assuming that new foreign-born mothers not in the ACS, but found in the CDC data, have the same characteristics as those who are included in ACS data. He does not provide any reason why he thinks this is the case. There are good reasons to think that those who left the country soon after giving birth, and thus show up in CDC data but not ACS data, are quite different from those in the ACS data. We do know from the Census Bureau's Current Population Survey that few young children with immigrant mothers leave each year.5 This should come as little surprise, as one of the primary reasons people come to America is so that their children will have the benefit of growing up in the United States. Leaving shortly after their births would defeat this purpose. Thus it seems extremely unlikely that many longtime immigrant residents have a child and then leave immediately. The notion that the difference between ACS and CDC data is almost entirely (93 percent) explained by well-established immigrants having a child and then leaving the country shortly after the birth, as the Niskanen analysis assumes, is not reasonable.
In sum, Niskanen's Neufeld first mistakenly reports the number of foreign-born women from the ACS, and then excludes many potential birth tourists from his estimate without good reason. We stand by our prior estimate of 33,000 possible birth tourists in the second half of 2016 and the first half of 2017. Of course, as we stated in our original estimate, our number is a "rough approximation, based on limited data". All of the caveats about our estimates mentioned in our original analysis still apply.
1 One figure in our blog post was transposed. We stated there were 854,896 births to foreign-born women in the 2017 ACS, the actual figure is 845,896. However, when we subtracted this number from what the CDC birth records show (897,223) to obtain our estimate, we did so from the correct figure, so the difference we reported of 51,327 in the December post was correct.
2 Almost all of these new mothers were born in Puerto Rico, Guam, American Samoa, and the U.S. Virgin Islands and are U.S. citizens at birth and should have been excluded.
3 As we made clear in our analysis, the Census Bureau defines the foreign-born slightly differently than the CDC. The CDC figures are based solely on the mother's country of birth, while the ACS has a specific citizenship question that records whether the mother was born abroad of American parents. We use the CDC definition of the foreign-born in this comparison and code the public-use ACS data accordingly so that it can be compared to the CDC data.
4 In our analysis, we assume that about a third of foreign-born mothers not included in ACS data, but who show up in CDC data, could be naturalized citizens. For this reason, our estimate of potential birth tourists is 33,000, not the entire 51,327 difference between CDC and ACS numbers. So our estimate does allow for the possibility that some long-time residents did leave shortly after the birth of a child, but not the implausibly high 93 percent that the Niskanen Center used.
5 The Current Population Survey Annual Social and Economic Supplement, unlike the ACS, asks individuals about their mother's place of birth. The number of four-year-olds in 2018 who were U.S.-born children with a foreign-born mother was just 2.2 percent less than the number of one-year-olds in 2015. The difference was less than 17,000 and well within the margin of error of the Current Population Survey, which is a much smaller survey than the ACS. This indicates that immigrants with young, U.S.-born children seldom take their children and leave the country.