There was a comment below on Indian American ethnicity in terms of proportion. By “ethnicity,” I mean the dominant language-based groups which serve as the organizing unit of many Indian states. The usual figures I see quoted are that 50% of Indian Americans are Gujarati, 25% Punjabi, with the balance a host of other groups (e.g., Bengalis, Tamils, Assamese, etc.). Digging around, I found these data:
In 2006, 26.3 percent of Indian immigrants age 5 and older reported speaking Hindi at home. Gujarathi (14.1 percent) was the next most popular language, followed by English (10.1 percent), Panjabi (10.0 percent), Telugu (9.7 percent), Tamil (6.7 percent), Malayalam (6.1 percent), Urdu (3.4 percent), Marathi (3.1 percent), Bengali (2.2 percent), and Kannada (1.7 percent).
These data are from the United States Census’ American Community Survey. Here are the results where I simply removed the English speakers from the sample:
|Language||%||Ratio with India %|
From these data, it looks like I was wrong, South Indians may very well be more than 25% of the Asian Indian population. The idea that 3/4 of American Asian Indians are Punjabi or Gujarati seems refuted by these data. Many Gujaratis and Hindu Punjabis raise their children speaking Hindi, which is in any case very closely related to these languages, but even if you assume that the whole of the 29% who speak Hindi at home are Gujarati or Punjabi you only get 56% for these two groups. As it is, though immigrants from the core Cow Belt states where Hindi is the official language are almost certainly underrepresented among Asian Indians, some of the 29% are going to be from Bihar, Uttar Pradesh, and so forth. Urdu is sister language of Hindi which is spoken by South Asian Muslims, so that allows us to peg the low bound for the proportion of Asian Indians who are Muslims (many Muslims do not speak Urdu, but speak the language of their region).
The main caution I would offer about these data is that they are of Indian immigrants, and so skew toward more recent migration waves. I am willing to bet that Punjabis and Gujaratis were far less well represented among the migrants of the 1965-1980 period. Many of their children are not immigrants by definition, as they were born in this country, and a great number of these would speak English at home. Since about 25% of American Asian Indians are native-born adding these would change the results above if their proportions were different, though I think the snapshot of the first-generation immigrant community serves as a good first approximation. I believe that adding the native-born Asian Indians would reduce the proportion of Punjabis and Gujaratis further, so I simply don’t believe the .quoted statistics anymore.
Addendum: 7.4% of Asian Indians do not speak any of the languages listed above at home (including English). So in this 7.4% you will find Assamese, Oriya, Tulu, and so on.