The Census Bureau set off reports of a 'multiracial boom' when it announced that, according to the 2020 census, multiracial people accounted for 10.2 percent of the U.S. population. Only the year before, the bureau's American Community Survey had estimated their share as 3.4 percent. We provide evidence that the multiracial boom was largely a statistical illusion resulting from methodological changes that confounded ancestry with identity and mistakenly equated national origin with race. Under a new algorithm, respondents were auto-recoded as multiracial if, after marking a single race, they listed an 'origin' that the algorithm did not recognize as falling within that race. However, origins and identity are not the same; confounding the two did not improve racial statistics. The fictitious multiracial boom highlights the power of official statistics in framing public and social-science understanding and the need to keep ancestry and identity distinct in both theory and empirical practice.

中文翻译：

---

多种族复杂性：2020 年人口普查和虚构的多种族热潮


人口普查局宣布，根据 2020 年人口普查，多种族人口占美国人口的 10.2%，引发了“多种族繁荣”的报告。就在前一年，该局的美国社区调查（American Community Survey）估计他们的份额为3.4%。我们提供的证据表明，多种族繁荣在很大程度上是方法论变化造成的统计错觉，这些变化将祖先与身份混为一谈，并错误地将民族血统等同于种族。根据一种新算法，如果受访者在标记了一个种族后，列出了算法无法识别为属于该种族的 “出身”，则他们会被自动重新编码为多种族。然而，起源和身份并不相同;混淆两者并没有改善种族统计数据。虚构的多种族繁荣凸显了官方统计数据在构建公共和社会科学理解方面的力量，以及在理论和实证实践中保持血统和身份差异的必要性。