This is a really fun exercise; a rare example of something that's "data-centric" without being soulless.
I think it's fascinating how it illustrates weirdness about how Americans think about and categorize "ethnic" food. For example, the author's analysis of Google data shows Glendale, CA ranks #1 for "Highest prevalence of Mediterranean Restaurants." But I am nearly certain the majority of these, given Glendale's demographics, are in fact Armenian or Persian restaurants. Both Iran and Armenia are of course quite far from the Mediterranean region, but for whatever reason (rice? flat breads? grilled things on a stick?) have gotten lumped in with some Americanized, genericized conception of "the Mediterranean" that's indistinguishable from "the Middle East." I would imagine you'd find the same thing happening on Yelp etc.
> some Americanized, genericized conception of "the Mediterranean" that's indistinguishable from "the Middle East."
The same happens with the food itself. I had a chat with a restaurateur in Switzerland, and he explained all the modifications he had to make in order to sell "Chinese" food. "They didn't have bean sprouts when I first came, and they will look like they are dying if there's any amount of spice in it."
The famous example of this is Chicken Tikka Masala, which is a British take on Indian food. You can't open an "Indian" restaurant in the UK and not put it on the menu, just as you must have the step-ladder of spice with Indian sounding names (Korma, Madras, Vindaloo). IIRC similar to General Tso's Chicken when it comes to ordering Chinese in the US, gotta be on the menu.
People simply come to expect certain things with certain foods, often disconnected with the the place that inspired it. When you open an ethnic restaurant, it's almost like joining a franchise. You aren't formally paying MacDonald's when you open a Chinese takeaway, but you do have to have things on the menu that people recognize, so the labels "Thai", "Ramen", "Japanese", etc function a bit like a franchise.
A bit like writing a fantasy novel. You aren’t paying any Tolkien Estate licensing fees, but people expect the established elves, trolls, dwarves, goblins, dragons, etc.
A more historical example of the same phenomenon may be commedia dell'arte.
BTW this happens all around the world. There are some staple dishes found at every Chinese restaurant in Korea, which are only tangentially related to Chinese food sold anywhere else.
I recently spoke with someone who traveled to Thailand. She didn’t like the Pad Thai in Thailand and instead preferred the one in NYC with yellow color on the noodles.
I'd say that the step-ladder of spiced Indian dishes are more associated with takeaways and cheaper restaurants. High quality Indian restaurants in the UK will tend to feature a specific region and only have a handful of dishes.
Same with higher end restaurants in the US for Chinese and Indian food.
Depending on where you are the cheaper restaurants in immigrant communities will be similar.
You get the regional food the chef's mother made. Occasionally, there are local substitutions (different mangos, peppers, meat cuts).
Higher end chinese or indian restaurants are pretty rare in the US. Both cuisines have been relegated to the fast, cheap delivery/takeout space and places doing higher-end (and higher priced) dishes find it difficult to get customers.
Plus lots of salad and olive oil. I believe the use of "Mediterranean" is to avoid strange expectations about Middle Eastern food, which many people seem to erroneously expect to be more like Indian.
Agreed, I am wondering if you could extract food truck data from the various licensing databases. That question arises because in some places food trucks have replaced the statistically improbable 'hole-in-the-wall' restaurant for some of the same reasons those restaurants existed, relatively low cost of entry.
We've got a remarkable number of West African restaurants in Laurel, MD. (Not East African, like they do in DC down the road.) I'm becoming quite the connoisseur of the differences between Ivoirien, Senegalese, Gambian, and several other types of Jollof.
The town is known for its African American and Central American populations, but there's clearly a large African immigrant population that I just don't otherwise see.
There seems to be some correlation with how NIMBY a city is, and its access to diversified food options. (And probably entrepreneurship in general?)
Similarly, I would expect that the greater the dependence a city has on cars, the less diverse their food options are (leaning heavily into fast food.)
Houston standing out makes sense though. Despite its insane car infrastructure, I believe there are comparatively few restrictions on property use.
> Similarly, I would expect that the greater the dependence a city has on cars
I believe, this is simply if not reversed. A city with a good car infrastructure is far more likely to have niche restaurants, because people can easily _get_ there.
Now that Google Maps has the AI summary for restaurants, I wonder if this can be queried in the API? It would probably have the keywords for Xinjiang food or whatever the OP wants to analyze. Checking two Central Asian restaurants I've been to in my area, one is tagged "Restaurant" and one is tagged "Pan-Asian Restaurant", so not very illuminating. However, in their AI summaries both have keyword "Central Asian" and one even says more specifically "Kyrgyz"
Carrollton has a number of what I call east Asian malls. All kinds of great Asian restaurants with food options from very Americanized to highly regional.
The Italian city where I live (~100k) has 4 McDonald's, no burger kings (there was one, it closed years ago) and that's quite it in terms of fast food.
Now that I think of it, there's a subway. It must be for tourists because I've never heard about any local eating there
How is this possible? There's about 25 restaurants that serve 100k people, so 4k people each? Can you really run a restaurant with a 4k catchment? What proportion of people are eating out each day?
You're right. The numbers seem way too high. Nine Starbucks?? Not to mention that these are only the fast food spots. In total there are 300+ restaurants. I'm guessing that high density cities skewed the numbers way too high.
I suspect they are a bit high, but not insanely so. The author is using city limits rather than metro area it appears, so there is going to be some draw in from the suburbs.
100k/305 residents= 325 people/restaurant. Average per capita spending on food away from home is $4500. That means that each restaurant has $1,475,000 of addressable market on average, which seems totally viable? (https://www.michiganfarmnews.com/boom-in-spending-at-restaur...)
Some of the best Thai I’ve ever had outside of Thailand and Malaysia was in Allentown PA. Not sure that would get a hit since there are so many Thai places (by design).
This is a really fun exercise; a rare example of something that's "data-centric" without being soulless.
I think it's fascinating how it illustrates weirdness about how Americans think about and categorize "ethnic" food. For example, the author's analysis of Google data shows Glendale, CA ranks #1 for "Highest prevalence of Mediterranean Restaurants." But I am nearly certain the majority of these, given Glendale's demographics, are in fact Armenian or Persian restaurants. Both Iran and Armenia are of course quite far from the Mediterranean region, but for whatever reason (rice? flat breads? grilled things on a stick?) have gotten lumped in with some Americanized, genericized conception of "the Mediterranean" that's indistinguishable from "the Middle East." I would imagine you'd find the same thing happening on Yelp etc.
> some Americanized, genericized conception of "the Mediterranean" that's indistinguishable from "the Middle East."
The same happens with the food itself. I had a chat with a restaurateur in Switzerland, and he explained all the modifications he had to make in order to sell "Chinese" food. "They didn't have bean sprouts when I first came, and they will look like they are dying if there's any amount of spice in it."
The famous example of this is Chicken Tikka Masala, which is a British take on Indian food. You can't open an "Indian" restaurant in the UK and not put it on the menu, just as you must have the step-ladder of spice with Indian sounding names (Korma, Madras, Vindaloo). IIRC similar to General Tso's Chicken when it comes to ordering Chinese in the US, gotta be on the menu.
People simply come to expect certain things with certain foods, often disconnected with the the place that inspired it. When you open an ethnic restaurant, it's almost like joining a franchise. You aren't formally paying MacDonald's when you open a Chinese takeaway, but you do have to have things on the menu that people recognize, so the labels "Thai", "Ramen", "Japanese", etc function a bit like a franchise.
A bit like writing a fantasy novel. You aren’t paying any Tolkien Estate licensing fees, but people expect the established elves, trolls, dwarves, goblins, dragons, etc.
A more historical example of the same phenomenon may be commedia dell'arte.
BTW this happens all around the world. There are some staple dishes found at every Chinese restaurant in Korea, which are only tangentially related to Chinese food sold anywhere else.
I recently spoke with someone who traveled to Thailand. She didn’t like the Pad Thai in Thailand and instead preferred the one in NYC with yellow color on the noodles.
I'd say that the step-ladder of spiced Indian dishes are more associated with takeaways and cheaper restaurants. High quality Indian restaurants in the UK will tend to feature a specific region and only have a handful of dishes.
Same with higher end restaurants in the US for Chinese and Indian food. Depending on where you are the cheaper restaurants in immigrant communities will be similar.
You get the regional food the chef's mother made. Occasionally, there are local substitutions (different mangos, peppers, meat cuts).
Higher end chinese or indian restaurants are pretty rare in the US. Both cuisines have been relegated to the fast, cheap delivery/takeout space and places doing higher-end (and higher priced) dishes find it difficult to get customers.
Plus lots of salad and olive oil. I believe the use of "Mediterranean" is to avoid strange expectations about Middle Eastern food, which many people seem to erroneously expect to be more like Indian.
I always assumed it was due to a lot of Americans thinking "middle eastern" has a negative connotation.
Agreed, I am wondering if you could extract food truck data from the various licensing databases. That question arises because in some places food trucks have replaced the statistically improbable 'hole-in-the-wall' restaurant for some of the same reasons those restaurants existed, relatively low cost of entry.
To me, this is less about how Americans incorrectly categorize food (not that they don't do that) and more about how lazy algorithms do so.
People lack nuance. Nuanced views are computationally expensive.
We've got a remarkable number of West African restaurants in Laurel, MD. (Not East African, like they do in DC down the road.) I'm becoming quite the connoisseur of the differences between Ivoirien, Senegalese, Gambian, and several other types of Jollof.
The town is known for its African American and Central American populations, but there's clearly a large African immigrant population that I just don't otherwise see.
There seems to be some correlation with how NIMBY a city is, and its access to diversified food options. (And probably entrepreneurship in general?)
Similarly, I would expect that the greater the dependence a city has on cars, the less diverse their food options are (leaning heavily into fast food.)
Houston standing out makes sense though. Despite its insane car infrastructure, I believe there are comparatively few restrictions on property use.
> There seems to be some correlation with how NIMBY a city is, and its access to diversified food options.
How are you measuring NIMBYness?
> Similarly, I would expect that the greater the dependence a city has on cars
I believe, this is simply if not reversed. A city with a good car infrastructure is far more likely to have niche restaurants, because people can easily _get_ there.
Now that Google Maps has the AI summary for restaurants, I wonder if this can be queried in the API? It would probably have the keywords for Xinjiang food or whatever the OP wants to analyze. Checking two Central Asian restaurants I've been to in my area, one is tagged "Restaurant" and one is tagged "Pan-Asian Restaurant", so not very illuminating. However, in their AI summaries both have keyword "Central Asian" and one even says more specifically "Kyrgyz"
Wow, what is Carrollton like?
Highest prevalence of Korean Restaurants:
Carrollton, Texas 14.67%
Federal Way, Washington 12.45%
Santa Clara, California 8.74%
Garden Grove, California 8.20%
Irvine, California 7.75%
Fullerton, California 7.46%
Ann Arbor, Michigan 5.14%
Honolulu, Hawaii 5.13%
Killeen, Texas 4.40%
Torrance, California 4.25%
It is a suburb of Dallas. Carrollton has an Hmart and a big spa across the street.
There are a few Korean chaebols with offices in DFW. Samsung Electronics and Hyosung are the largest.
Carrollton has a number of what I call east Asian malls. All kinds of great Asian restaurants with food options from very Americanized to highly regional.
>> what is Carrollton like
Carollton, Texas is like the deep south met H-1B jobs.
Apparently an American city of 100,000 people has:
9 Starbucks and 4 Dunkin’s 6 McDonalds, 3 Burger Kings and 3 Wendy’s 4 Taco Bells and 2 Chipotles 9 Subways 3 Dominos and 2.5 Chick-Fil-A’s
The Italian city where I live (~100k) has 4 McDonald's, no burger kings (there was one, it closed years ago) and that's quite it in terms of fast food.
Now that I think of it, there's a subway. It must be for tourists because I've never heard about any local eating there
How is this possible? There's about 25 restaurants that serve 100k people, so 4k people each? Can you really run a restaurant with a 4k catchment? What proportion of people are eating out each day?
You're right. The numbers seem way too high. Nine Starbucks?? Not to mention that these are only the fast food spots. In total there are 300+ restaurants. I'm guessing that high density cities skewed the numbers way too high.
Quite often you see the mini-starbucks in areas like malls or other large retail stores instead of stand alone buildings.
In the town of 125k I used to live in there was a SB inside a bookstore, inside the mail, and a standalone building within 1000 meters of each other.
I suspect they are a bit high, but not insanely so. The author is using city limits rather than metro area it appears, so there is going to be some draw in from the suburbs.
100k/305 residents= 325 people/restaurant. Average per capita spending on food away from home is $4500. That means that each restaurant has $1,475,000 of addressable market on average, which seems totally viable? (https://www.michiganfarmnews.com/boom-in-spending-at-restaur...)
[dead]
Some of the best Thai I’ve ever had outside of Thailand and Malaysia was in Allentown PA. Not sure that would get a hit since there are so many Thai places (by design).