Top Ancestry or Hispanic or Racial Subgroups for U.S. County Subs and Places

[cinyc]

As I mentioned in the past few weeks, Census recently released their 2012-16 American Community Survey results. One of the tables that release contained is Reported Ancestries and Hispanic/Racial Subgroup information for Asians, American Indians, Alaska Natives and Pacific Islanders.

First, the CONUS map:



As you can see, German Ancestry predominates in most of the North except the Northeast. American is tops in much of the South outside of Florida and Southern Louisiana, and the Mexican Hispanic subgroup dominates much of the Southwest. One issue with Census' ancestry data is that African-Americans tend to check other ancestry instead of one of the Subsaharan African ancestries, which makes African-American predominant cities like New Orleans appear to be something else, (in New Orleans' case, German).

Next, Alaska:



Alaska Natives are not as monolithic as you'd think. German tends to dominate the big city areas, like in the Northern U.S.

And Hawaii:



The Japanese are the most prominent group in the Urban Honolulu and some neighboring CDPs and CCDs. Native Hawaiians dominate in Windward and Leeward Oahu, plus many of the Neighbor Islands, especially Molokai. You can tell where some of the military bases are on Oahu by looking to areas with German top ancestry.

[cinyc]

I will be adding a some close-up maps of major metro areas/states over the next few days. This will also likely go up interactively on Carto at some point.

First, the NYC Metro:



Italian is the top ancestry on Staten Island and many of NYC's suburbs. Dominicans are tops in Manhattan and the Bronx; Chinese in Brooklyn and Queens - though African-Americans' lack of choosing listed ancestries probably hurt their numbers in Brooklyn. That certainly hurt in Newark, whose top A/H/R is Puerto Rican on this map.

[cinyc]

O Here is a Venn Diagram showing the people who claim "American" and the people who voted Trump.

Some African-Americans likely choose American, too, for lack of a better choice, so I doubt it's a full set. But yes, many of the "American" ancestry towns are likely pro-Trump.

[cinyc]

though African-Americans' lack of choosing listed ancestries probably hurt their numbers in Brooklyn.

NYC's Black population is heavily Caribbean.

About half of NYC Black household heads are foreign born (see Table 1).

http://www.metropolitiques.eu/The-Evolution-of-New-York-City-s.html?lang=fr

There are probably a fair number of 2nd generation Caribbean Blacks in the "native" head of household category.  It wouldn't surprise me if more NYC blacks can trace their roots to the Caribbean than to the South at this point.

Census reports the Caribbean groups as ancestries. In Mount Vernon, NY, for example, Jamaican was top ancestry.

When I have time, I’ll have to run a test with jimrtex’s back-ending the African-American percentage. I suspect that’s going to create the opposite issue - a map that shows too many plurality African-American areas with AA as top ancestry/Hispanic/Racial subgroup.

[cinyc]

As a test, here's what the NYC Metro map looks like if you define African American to be Non-Hispanic AAs - all Subsaharan African Ancestries - All West Indian Ancestries, as defined on the total ancesty table. Someone can be, say, Nigerian and Dominican, so I may be slightly undercounting by using that metric:



As expected, I think this is a bit over-inclusive. I think of places like Elizabeth, NJ as more Hispanic than African-American, but because the Hispanic groups are fractured, it comes out as plurailty African-American.

jimrtex - do you think this metric is better?

[cinyc]

Can someone make a map of the whole country with this methodology.

I'm not going to redo Alaska or Hawaii. I doubt they'll change much, anwyay.

By the way can someone find a way to lump American in the English category, from what I've heard most of the people that mark American are white southerners of English ancestry that have been here so long they call their ancestry American.

I could, but it takes my computer a very long time to recompile the data when I make a change. I have to go into 3 different spreadsheets to compute things. So it's not on my immediate to do list.

[cinyc]

I like this map much better than the first one because it doesn’t ignore the African Americans in the black belt and the major cities, However I believe the map would be improved evan more if American was joined with English so we could clearly see which areas of the Deep South are populated by white southerners and which parts are populated by African Americans. For example, Mississippi is shown on the map as almost entirely African American, including areas we know are populated by White southerners, Unfortunately these areas are shown as African-American because the American and English catagories are separated from each other inaccurately showing African Americans as the largest group, However I understand it’s probably really time consuming to join the two groups together.

Wouldn't a simple racial map do that?

BTW - I noticed an error in my calculations. I subtracted the wrong rows. I've replaced the maps in the previous post. It flipped some of the NYC boroughs and suburbs back to what they were, since many NYC-area African Americans are West Indian.

[cinyc]

By the way can someone find a way to lump American in the English category, from what I've heard most of the people that mark American are white southerners of English ancestry that have been here so long they call their ancestry American.

This doesn't change things in the South a heck of a lot, in my opinion. Maybe at the margins, but South Carolina still looks pretty brown on the map.

Note that there are still some pockets of American and English on the map in jurisdictions where there was one of the two but not the other, largely in rural areas. This is because I made the map quickly using a =MAX function and didn't do anything for ties.

[cinyc]

This map is by far the most accurate ancestry map I’ve seen of the United States, What is most striking is how once you combine American with English it becomes so obvious that English is by far the largest ancestry among White Americans with German a close second, Irish third, and Italian a distant fourth. Finaly their is a clear correlation between areas of English ancestry and areas with White Evangelicals in the South and bordering parts of the Midwest and with the Mormon population in the West. Basically these English Americans are the real republican base.

Yes, American+English is the best map so far. I'd love to see more of close-ups of major cities (Seattle, please!).

This is likely to go up interactively on Carto after I crunch more ACS numbers. I don't want to have to re-do the maps every time I update the database. (Free Carto accounts don't get much space, so I have to be economical).

Anyway, Metro Seattle:



Here's the map with English and American as separate categories. It's more German:

[cinyc]

SF Bay Area:





No, I don't know how the Puget Sound Salish made their way down to California. But they're allegedly tops in Cold Springs CDP, Toulume County (covered up by the legend). My guess is it's an estimation or data error. Probably the latter.

[cinyc]

I've finally gotten around to putting an interactive version of this map on Carto. The Carto version is by Census Tract, which adds granularity in cities but not necessarily rural areas, where some County Subs/Places may be smaller than a census tract.

That interactive map is here:
https://cinyc.carto.com/viz/bbe0ba2d-3709-4933-9817-c545c4ca78ca/public_map

Clicking on a Census Tract will pop up a window that gives you the name and percentage of the Top Ancestry or Hispanic or Racial Subgroup for that tract.

There are also a series of interactive percentage maps for various ancestries/subgroups in my Carto account.

[cinyc]

I've finally gotten around to putting an interactive version of this map on Carto. The Carto version is by Census Tract, which adds granularity in cities but not necessarily rural areas, where some County Subs/Places may be smaller than a census tract.

That interactive map is here:
https://cinyc.carto.com/viz/bbe0ba2d-3709-4933-9817-c545c4ca78ca/public_map

Clicking on a Census Tract will pop up a window that gives you the name and percentage of the Top Ancestry or Hispanic or Racial Subgroup for that tract.

There are also a series of interactive percentage maps for various ancestries/subgroups in my Carto account.

Interesting. I assume you will be producing single ethnic group maps for other groups (Korean, Vietnamese, Filipino, Asian Indian, etc.)?

Are Japanese in LA more business oriented. and wealthier, so they can live closer to the coast?

On the main map, I would group related groups together by color, with the largest groups less saturated in color.

e.g. Asian and maybe Middle Eastern groups in green. For example, currently Vietnamese is barely distinguishable from Mexican. Since it is easier to distinguish among various green tones, this would work in areas where there are small communities of various Asian ethnic groups (Houston, Southern California, New York).

AIAN in saturated pinks, there might need to be much need for color separation, since there is physical separation. This would permit the Sioux, Cherokee, Choctaw, Navajo, Hopi areas to be visually tied together.

Hispanic in brown, dropping the saturation of Mexican down, so other groups such as Guatemalan, etc. stand out.

Black, in Dark Brown with African American dropped in saturation, so any other groups such as Haitian might stand out.

European in blue, with the Germans in a pale blue, with English. American, Irish somewhat darker, and then Norwegian, Italian, Dutch, etc. in brighter blues.

What percentage of Guatemalans in Harris County reside in C.T. 4327.01?

I will be making maps for some additional groups - those who are tops in ancestry in 100 or more tracts. You named many of the Asian groups, but there are also a few Hispanic subgroups (Puerto Rican, Dominican, Salvadoran, etc), plus French (excluding Basque), Polish, Portuguese and a few others.

The problem with changing the color scheme so that similar groups are the same color is that I think it actually makes the map harder to read. Colors for the top 20 or so groups were chosen so that they were heavily saturated and didn’t come close with with each other. The next 200 or so were assigned less saturated colors based on their then-rank. Because of a lot of ties among smaller groups, a lot of the smaller AI, Asian and PI groups end up with the similar colors, anyway. But if, for example, Navajo and Hopi were both red, it would be harder to distinguish Navajo and Hopi areas in AZ.

I’ll get back to you on the Harris County C.T. 4327.01 numbers when I get to my computer later today.

[cinyc]

Harris County C.T. 4327.01 is 47.54% Guatemalan and 35.54% Mexican. The top (non-Hispanic) ancestry is German at 2.9%.

In theory, the Guatemalan percentage should pop up on the Top A/H/R map if you click on that tract, since it is top A/H/R.

[cinyc]

What about Jews?  They're a plurality in certain NYC suburbs, probably in Brooklyn and Manhattan too and various other places like Brookline and Newton outside Boston, in pockets in Broward and Palm Beach etc.   

Jewish is not an ancestry reported by Census. Look for the areas dominated by Russians, Eastern Europeans, and, in some cases in NYC, Americans or Germans for areas in which Jews are the plurality.

Here's a cleaned-up map of the NYC Top Ancestry or Hispanic or Racial Subgroup with AA Estimate by Census Tract (Note the projection is Albers Conic CONUS, which distorts the direction a little bit - sorry Jim, but I made the Print Composer before switching projections and didn't want to redo it):

[cinyc]

The clarity of the break between Domininicans and Puerto Ricans in the Bronx is dramatic. Is there a street that is considered the boundary?

It looks like the dividing line is the Grand Concourse (or a block or two to the east, to take in the apartment buildings on the east side of the boulevard).

You must be logged in to read this quote.

Yes - though the top percentages usually aren't that high (in the 10-20% range). The Upper West Side is historically more Jewish than the Upper East Side, and German tends to correlate with Jewish in NYC, I think.

You must be logged in to read this quote.

The groups in Central Brooklyn are African American (dark brown), Jamaican (brownish green), West Indian (olive-like color), African (middle green) and Haitian (yellowish green). I wouldn't doubt that some of the Africans there are actually from Africa, instead of African-Americans descended from slaves. Central Brooklyn has a large West Indian and first and second generation African population.

The whole area isn't considered Bed-Stuy. Bed-Stuy is in the northwest portion of the African-American belt in Central Brooklyn. It's nearer to Williamsburg, which is why it is gentrifying. Some of the other black neighborhoods are Crown Heights (southeast of Bed-Stuy), (further southeast) Brownsville and Canarsie (near Jamaica Bay). I think the West Indian/African area is mainly in East Flatbush.

You must be logged in to read this quote.

Yes - and in some tracts, by a clear majority. It's the new Chinatown, spiraling out from Sunset Park to SW Brooklyn.

You must be logged in to read this quote.

38 in the NY portion of the map, which includes a few towns in Westchester and Nassau. Iranian might only be tops in Nassau, so it might be 37 in NYC. There are 5 more different top ancestries in New Jersey, according to my count (Cuban, Portuguese, Peruvian, Brazilian and Egyptian).

[cinyc]

The clarity of the break between Domininicans and Puerto Ricans in the Bronx is dramatic. Is there a street that is considered the boundary?

It looks like the dividing line is the Grand Concourse (or a block or two to the east, to take in the apartment buildings on the east side of the boulevard).

It is kind of interesting that there is a sharp demarcation. Were the Dominican areas historically black? The black areas of the Bronx are now on the northern edge spilling into Westchester.

I think so. The AA neigborhoods in the north Bronx have been there for decades. Mount Vernon in Westchester (the spillover) has also been black for a long time. As you can see from the map, though, Jamaicans are moving into this area.

You must be logged in to read this quote.

No, I didn’t. That would be one heck of a project. It takes me about half a day to put together all the spreadsheets for this project, which kills my computer.

You must be logged in to read this quote.

The groups in Central Brooklyn are African American (dark brown), Jamaican (brownish green), West Indian (olive-like color), African (middle green) and Haitian (yellowish green). I wouldn't doubt that some of the Africans there are actually from Africa, instead of African-Americans descended from slaves. Central Brooklyn has a large West Indian and first and second generation African population.

The whole area isn't considered Bed-Stuy. Bed-Stuy is in the northwest portion of the African-American belt in Central Brooklyn. It's nearer to Williamsburg, which is why it is gentrifying. Some of the other black neighborhoods are Crown Heights (southeast of Bed-Stuy), (further southeast) Brownsville and Canarsie (near Jamaica Bay). I think the West Indian/African area is mainly in East Flatbush.
[/quote]
I suspect that it might show up more if you aggregated the specific Subsaharan African groups, thought there are also likely to be distinctly Senegalese, Ghanaian, Nigerian, etc. neighborhoods.[/quote]Probably.

You must be logged in to read this quote.

SE Queens near JFK has traditionally been a middle-class black suburban type area within the city. There are exceptions (the Italians of Howard Beach and in older days, Ozone Park).

I’ll have to check if the border areas are Asian Indian. They probably are.

You must be logged in to read this quote.

I wouldn’t doubt that. A lot of Asian Indians tend to live in the suburbs.

[cinyc]

I'm bumping this old thread with a link to the 2013-17 ACS version of the interactive maps:

https://www.thecinyc.com/2013-17-ahr-multimap


This multi-map has maps of the Top 6 AHRs in a census tract, plus percentage maps of some of the 31 of the most-predominant AHRs in the U.S.

You can also turn on and off the County and County Sub layers.

If you're using a computer, you can change the opacity (to see underlying geography) and color step (useful for highlighting where the highest percentage of the less-predominant AHRs live - the default is 10 points - you can change it to 0.1, 2, 4, 5 or any number less than 10). Cell phones are too small to include the interactive legend that has this feature - though if you click on a tract, you will get a pop-up with its stats.

[cinyc]

Thanks. What is the origin of settlement for Hmong in western North Carolina? All interesting to see the expansion of the Ecuadorian population in north(east) Minneapolis into the Anoka(n?) peninsula.

The Scottish map does not seem to be working. The whole map goes black.

I don't know the derivation of the Hmong in western NC. I do know they settled in some unexpected places after the Vietnam war, including some places in Wisconsin. It's sort of like the Somali settlement patterns of today, also in some of the weirdest places, like Maine.

I misspelled the variable for Scottish in my Javascript code. That should be fixed now.

You must be logged in to read this quote.

That's because it usually is points, not percent in my election maps (which served as the protype for these). I'll consider changing that.

You must be logged in to read this quote.

In the Asian/Pac Islander multimap, I assume? I've fixed that. If it's in the data, it's a bit more difficult to fix.

You must be logged in to read this quote.

I'd really have to think about how to do that using JS and Mapbox. I don't know how off the top of my head. The only thing I know for sure that would work is to put an optional layer with low-pop precincts to toggle on or off at the bottom of the menu, like the County Sub layer. Otherwise, I'd need to do some math in the colorscheme formula.

You must be logged in to read this quote.

Top Subsaharan African group or just each Subsaharan African group? Each Subsaharan African group is mapped on the Ancestry Su-Z map here:

https://www.thecinyc.com/2013-17-ancestry-su-z

Czech and Czechoslovak are here:
https://www.thecinyc.com/2013-17-ancestry-a-f

Slovak is here:
https://www.thecinyc.com/2013-17-ancestry-g-so

I haven't tried to combine the three.

You must be logged in to read this quote.

The color step isn't responsive on the Top AHR maps, only the percentage maps. As to why it's not instantly responsive, that's probably a Mapbox issue that, unfortunately, I don't have much control over. The only thing I might be able to do is exchange the slider for a text input box - if I can even figure out how to do that. I'm not a coder. Most of my code is begged, borrowed and stolen from various Internet sites.

You must be logged in to read this quote.

That's a Mapbox issue in rendering the map. Unfortunately, I can't do anything about it. QGIS renders better, but QGIS isn't online.

You must be logged in to read this quote.

Maybe. I'd have to look into it. It actually totally disappears if you look at the maps using a cell phone, in exchange for pop-ups due to the lack of screen space. The interactive legend is also a holdover from my election mapping.

You must be logged in to read this quote.

[cinyc]

Census released its 2014-18 ACS data on Thursday. My new interactive map with the updated Top AHR/Percentage Reporting a particular AHR data is on my newer website here:

https://cinycmaps.com/index.php/ancestry-2013-17/2014-18-top-ahr

Click on Menu to change layers. Otherwise, the maps are pretty much the same as last year. No, Jim, I haven't updated the code to allow you to white out low-population tracts, at least yet. Seems too complicated to code, but I might give it a shot in the coming weeks.

[cinyc]

Is Fort Bend 6731.01 (Cinco Ranch) the most populous census tract? 70271 persons?

Yes.