Lit Review

[AaronGullickson]

Studies that can help frame broader concept of pan-ethnicity

• Lopez and Espiritu, 1990. An excellent article that is considered canonical in the field (along with Espiritu's 1992 book). They break down the factors affecting panethnicity as cultural (language, religion) and structural (racialization, class, generation, and geography). They then compare these features across four cases of Asian Americans, Latinos, Indo-Americans (East Indians), and Native Americans. Their main concusion is that the structural factors matter more than the cultural factors, largely evidenced by the weaker pan-ethnicity among Latinos, who are the most culturally similar but the most structurally divided. There is lots of good stuff here that relates to my paper, particularly their consideration of the role of language, class, generation, and geography -- all of which are addressed in the modeling. Their empirical analysis is based on organizations, but they note the importance of studying marriage patterns to look more at how individuals act on the ground.
• Kibria, 1997. Qualitative work on pan-ethnicity among second-gen Chinese and Korean Americans centered around issues of intermarriage. She finds that racialization does affect people's sense of pan-ethnicity and that for most this is not obviously through connection to pan-ethnic organizations and institutions. She also finds that respondents identify a pan-ethnic Asian culture that basically is the "model minority" stereotype and centers around conceptions of the immigrant expierence. So respondents say they want to marry a Korean American not a Korean, for example. She also notes that while there is a sense of pan-ethnicity it is mostly tied to East Asians, Filipino and Vietnamese Americans are seen as further removed. Finally, Kibria notes that pan-ethnicity is still hampered by the fact that national identities take precedence for most respondents, but this could change if assimilation were to wipe away the significance of those national identities - then pan-ethnicity is what is left. Kibria also has a 2003 book on this same topic that is frequently cited, but going with the article since that is what I have actually read.
• Wimmer, 2008. Both his AJS and ERS articles as well as book will ge useful in framing pan-ethnicity as a "topographical" move of boundary expansion.
• Kim and White (2010). Interesting paper that looks at panethnicity in residential segregation, but I don't think there is much here that is directly relevant. They do find weaker panethnicity among Asian ethnic groups, and they find that it is stronger where a higher proportion of the groups are foreign-born - basically a panethnic enclave for Asians that does not exist for other groups. Interesing contrast to the intermarriage data here which suggests stonger panethnicity for Asians - might be interesting to note in the conclusions.
• Okamoto and Mora, 2014. This is an ARS article. Very little actual focus on intermarriage in terms of pan-ethnicity, but they do address some important conceptual issues. They define pan-ethnicity as: "the construction of a new categorical boundary through the consolidation of ethnic, tribal, religious, or national groups." They emphasize that pan-ethnicity is particularly defined relative to race, ethnicity, or nationality by the maintenance of sub-group identities, which is conistent with Okamoto's earlier arguments about its layered nature. I understand the general point, but think that this point makes race, ethnicity, and nationality sound too easy. I prefer Wimmer's approach that sees this all as a singular concept where the possibility of layering is possible in most cases but not necessarily realized. They also point out the debate in the literature over whether structural or cultural conditions are more responsible for the emergence of pan-ethnicity, with the consensus seeming to be on structural conditions. Relatedly, they talk about how the concept of pan-ethnicity is often subsumed into theories of racialization and assimilation, but they argue against this - I would agree here, but thats because I think all three can be subsumed into a boundary approach emphasizing the process of boundary expansion, shifting, and blurring. They also note that panethnic identification tends to be stronger in the US among the native-born, and point out the role that language might play here. The only place intermarriage is brought in is with a discussion of the importance of group size and diversity ala Blau 1977.
• Mora, 2014. A good article on the growth of pan-ethnicity among Latinos through cross-field effects. Not sure how directly relevant it will be.

Studies on South Asian Distinctiveness in the US:

• Kibria, 1996. This is the article that describes South Asians as "ambiguous non-whites." She also talks about how South Asians' reluctance to be racialized within the US system helps contribute to this ambiguity and the continuing sources of separation from other Asian Americans, particularly in organizations and movements.
• Morning 2001. Excellent review of the literature on South Asian racial ambiguity here. The results of her analysis aren't as directly important for my work, although she does address the American Indian measurement issue.
• Schachter, 2014. This article is excellent as a good recent citation for the argument that South Asians do not fit easily within the "Asian American" identity in the US, partly for phenotypic reasons and perhaps other political and cultural reasons. Some good citations and discussion of earlier literature. Not as impressed by the analysis itself, but she does find that experiencing discrimination makes Indians less likely to identify pan-ethnically, which is inconsistent with prior work.

Studies that specifically look at pan-ethnic intermarriage

• Shinagawa and Pang, 1996. The analysis here is very basic. 1980 and 1990 Census data mostly from California. They just look at percentages of outmarriage, no adjustment for group size. Almost no discussion of data issues like timing of marriage and the IMA issue. Their biggest finding is that panethnic intermarriage among Asian groups has grown from 1980 to 1990. Not much else here, although they have an nice snippet from Espiritu's book.
• Qian, Blair, and Ruf, 2001. 1990 Census data. They use a 20-34 age range to restrict marriages and also remove couples where one person migrated between 1985 and 1990 to reduce the effect of pre-immigration marriages. They use the groups of Chinese, Korean, Japanese, Filipino, Southeast Asian, and Asian Indian (include Pakistani, Sri Lankan, and Bangladeshi). The results generally show that interethnic marriage is more common than marriage with whites, but also varies significantly by ethnicity, as does intermarriage with whites. They find that ethnic exogamy with Japanese Americans is most common and since this group is the most "assimilated" they see this as evidence of pan-ethnic assimilation. They also split their results. They also interact their terms with nativity and find that nativity increases both interracial and interethnic marriage. Also some good references here to work on how racialization affects pan-ethnicity for Asian Americans (mostly Espiritu 1992).
• Rosenfeld 2001. One of the most cited and a Demography paper. Rosenfeld's solution to the problem of uneven settlement patterns by ethnicity is to look at specific metropolitan areas. His main focus is Lost Angeles where he can jointly examine Asian and Latino pan-ethnicity, but he also does this for a few other cities and finds that the general results are similar. The comparison in LA is Mexican vs. Other Hispanic and Japanese vs. Other Asian, which I think is a bit crude. The log-linear models are also a bit tortured. However, the results are largely consistent with what I find in many ways. He finds that among the US born, Asian pan-ethnicity is stronger than Latino pan-ethnicity which he points out highlights the role racialization plays as opposed to cultural similarities among Latinos. Related to this, his other big finding is that pan-ethnic tendencies go down among native-born Latinos, which he attributes to the loss of Spanish as a pan-ethnic unifier. I don't see this so much in my data, because birthplace and language endogamy don't do much to shift over pan-ethnicity among Latinos.
• Qian and Cobas, 2004. So this article is basically Qian, Blair, and Ruf's approach applied to Latinos rather than Asians. They use the Latino groups of Mexicans, Puerto Ricans, Cubans, and "Central Americans." However, it is complicated by the strange decision to categorize Latinos by their racial id, despite many concerns about the validity of this measure for Latinos. They generally tend to find that among US-born non-white Latinos, it is more likely to marry a Latino of a different national origin than a non-Latino white. This is less clear for "White" Latinos, whatever that means. Like the other article, the overall odds of ethnic exogamy here are much lower than what I am getting for 1980 data, and I also think much lower than Fu's estimates with the same data. This could reflect the lack of adjusting for geographic specificity. In 1990, this would have had an important effect because of big differences in the settlement patterns of Mexicans, Puerto Ricans, and Cubans.
• Fu 2007. He is using long-form 1990 Census data and finds substantial pan-ethnicity among Asians, but less so for Latinos. Like my results, he finds Latino/White intermarriage more likely than Latino ethnic exogamy. He accounts for Census division to deal with regional issues. Only includes small set of ethnicities.
• Fu and Hatfield, 2008. I am not terribly impressed by the analysis here. They exclude the foreign-born population and make some other questionable methodological calls. I would say that the only useful finding here is that there is some affinity between Asian groups when they marry exogamously rather than to non-Asian groups. However, given their odd coding its touch to benchmark it. They use 2000 PUMS data.
• Okamoto, 2007. The paper uses 2000 PUMS data. It has a different methodology, with individual data nested in metropolitan areas, with an interest in how metro-level characteristcs affect odds of interracial vs. interethnic vs. ethnically endogamous unions (using multinomial logit models). Thus empirically, I don't think there is anything to compare to, but the theoretical arguments about a boundary approach are useful here and help frame my study. She has also made these points about the nested structure of panethnic identity, in other works including her book and they are similar to Wimmer's argument.
• Min and Kim, 2009. Uses ACS 2011-2006 to look at ethnic endogamy, ethnic exogamy, and interracial marriage among six Asian ethnic groups. They use raw percentages of the Asian partner, with no adjustments for group size, geography, double-dipping, or anything. This is not a good study. They seem to acknowledge at one point that these percentages largely reflect group sizes, but do nothing about it. Like why does this shit get published?
• Qian, Glick, and Batson, 2012. Uses Census 2000 data to look at four groups of Chinese, Filipinos, Mexicans, and Puerto Ricans. For all four groups, they find that the odds of pan-ethnic marriage are higher than intermarriage with whites. They split each group into those arriving at ages 14-19, 6-13, 0-5, and native-born. They find in each case that more time in US (or native born) is associated with more exogamy. Some differences across groups, but they also tend to find higher pan-ethnic exogamy for Mexicans and Puerto Ricans than Chinese and Filipinos.
• Bohra-Mishra and Massey, 2014. Ok, so this study uses the New Immigrant Survey from 2003. Their interests are broader than pan-ethnicity (more around theories of assimilation) and they are specifically looking only at intermarriage of immigrants, although one of their model specifically examines pan-ethnic tendencies by looking at marriage within region vs another region. They consider panethnicity as a form of selective acculturation ala segmented assimilation theory, which may be an interesting way to frame things. I am not very confident that they effectively account for group size in their analysis even though they include it in their multinomial models. They find much higher endogamy among Asians than other groups - I find it hard to believe this doesn't just reflect group size. Update: I thought about this one a little more and I think the fundamental problem is that they show the predicted probabilities from the models in Table 2, but because they are moving from odds to probabilities, they are probably re-introducing the group size differences that are supposedly controlled for.
• Smith 2017. This study is not going to be helpful. Everything from the theory forward is poor. I disagree with the characterization of mixed results from prior work or the idea that these various theoretical perspectives are competing, but rather represent different dimensions of the same underlying boundary processes. The very notion that you can somehow parse out the salience of Hispanicity from the salience of race is nuts both theoretical and empirically. Theoretically, it confuses some of the theoretical arguments where the question is the extent to which Hispanicity is treated like a race. Emprically, it is completely uncritical of the race reporting of Hispanics which is next to useless. And how can one create this space without considering pan-ethnicity among other groups like Asians. There is also this wonderful zinger: "For example, upward of 40 percent of people who identify as Hispanic identify as Other racially, suggesting that Hispanic is their understood racial/ethnic identity." I mean WTF!? It suggests no such thing.
• Qian, Lichter, and Tumin, 2018. This study uses 2009-2014 ACS data to look at intermarriage among Latinos. They use multinomial models to examine co-ethnic vs. pan-ethnic vs. white vs. black intermarriage. They use city characteristics as a predictor to try to get at local marriage markets. Ultimately, I don't know how useful it is for the current study because most of their focus is on the individual-level predictors of these things. I don't feel like they can possibly account for all the group size issues with their MSA variables and so I am somewhat skeptical of some of their findings. Table 1 does break down outmarriage percentages, but again no adjustments for group size. Their big finding is that foreign-born Hispanics are more likely to be in a pan-ethnic union than native-born Hispanics. They interpret this to mean a stronger preference for pan-ethnicity, but I think it is more likely that greater boundaries with the native-born non-Hispanic population restrict the pool for this group. Again, group size would be helpful here.

Miscellaneous

• Kao and Joyner 2006. Looks at panethnicity in best friends using Add Health data. It uses a multinomial logit model where you have interracial different ethnic, and coethnic as the options. I don't feel like they adequately control for group size issues as they only control for proportion own race and ethnicity in the models. They also get some weird results like finding Vietnamese are the most likely to be panethnic and Japanese are the least likely -- very inconsistent with other work. Their ultimate conclusion is that there isn't much evidence of panethnicity in adolescent friend networks. As a side note, it would be interesting to apply the counterfactual marriage model to the Add Health data. It should be do-able. You just sample alternate best friends from the same school.
• Hwang and Saenz, 1990. This article is seminal in its statement of the problem of including migrants in estimates of intermarriage without knowledge of where the union was formed. They show it upwardly biases rates of endogamy with examples of several Asian American groups.
• Jimenez, Fields, and Schachter, 2015. An article that highlights the growing importance of intra-ethnoracial diversity. May be useful in helping frame the final conclusions.