The Little Neighborhoods

No, not the small neighborhoods - the ones that start with the word "Little." Over on my other blog I make the claim that "there are no Koreans in Little Korea, there are no Brazilians in Little Brazil, and there are definitely no Italians in Little Italy." I decided I should crunch the numbers for a claim like that, so we're going to check it out. Nate Silver would be proud. I'm also adding in Little Italy in the Bronx, Little India in the East Village, and a couple others that NYC & Co (the city's marketing arm) has invented for it's 9 in 09 campaign - Little Sri Lanka in Staten Island, and Woodlawn in the Bronx, which NYC & Co. insist is "Little Ireland."

Now, every New Yorker knows that while "Little This, That, or The Other" might not have people of that particular persuasion, is almost always has restaurants. So as part of this, one question I'm going to be answering for each neighborhood is "more restaurants, or more people?" My quick guesses:

Little Korea: more restaurants.
Little Italy: more restaurants.
Little Brazil: this is a tough one. More people. Both in single digits though.
Little Italy in the Bronx: more people.
Little Sri Lanka: more people.
Little India: more restaurants.
Woodlawn (Little Ireland): more people.

Now, the numbers will not be that accurate. These are all micro neighborhoods, meaning the largest (Woodlawn) is only 4 census tracts. We have to go all the way back to the 2000 census to get tract-level data, and we all know the city has changed a great deal since then. There are ways you can developed trending numbers based on the American Community Survey from 2007 and other sources, but not for such detailed data and small geographies. So keep in mind the question is not going to be "how many Italians are in Little Italy?" it's is going to be "how many Italians were in Little Italy in 2000?" And I'm going to be comparing this against the restaurants I count there in 2009, so the "more people or more restaurants" question won't be an entirely accurate comparison.

Second, there's the question of what constitutes someone who is "Korean" or "Italian" or "Brazilian." Is it someone who was born in the respective country? Someone who speaks the language? Someone who self-identifies as being of that particular ancestry? These are all questions asked on the census, and all questions that can be used to determine someone's ethnicity. Does a kid from Jersey whose great-grandparents were from Italy count as "Italian" if he's currently living in a Mulberry Street tenement with two of his NYU roommates? For purposes of this exercise, I'm going to be as liberal as possible in determining who is "Korean" or "Italian" or "Brazilian."

Third, we will have better data for some neighborhoods than others. This is because the Census gathers racial and certain ethnicity information on their short form that everyone gets, but gathers place-of-birth, language, and ancestry information from only the 1 in 6 households who get the longer form. The census bureau specifically counts "Korean" and "Asian Indian" as "races," but Italian, Irish, and Brazilian as ethnicities, so as a result we have 100% data for Koreans and Asian Indians, but only sample data for the others. Sri Lankan is interesting, because while the short form lists Korean, Japanese, Chinese, Asian Indian, Vietnamese, and Filipino as races under the "Asian and Pacific Islander" category, after that it goes to "Other Asian." So we have potential 100% data for Little Sri Lanka, but it has to be looked at pretty carefully.

We'll start with the easiest one, Little Korea, next post.