Why are Lowe’s and Home Depot so often neighbors? With help from TikTok sleuth JoseMonkey, we looked at the actual data and found dozens of store pairs close enough to share a parking lot. Turns out, the reason may lie in a 1929 model involving two ice cream carts and one very competitive beach.
If you really want your mind blown, do a search on Advanced Auto Parts and AutoZone. The story I remember hearing was that a couple owned one of those stores, and then divorced. To spite their ex, one of the partners started a new auto store chain and tried to locate any new branch as close to the existing competing store as possible.
This is such a fascinatin look at why Lowes and Home Depot cluster together! I never thught about it from the Hotelling model perspective before, but it makes so much sense. It's like they're both trying to capture the middle ground of the home improvment market rather than differentiating themselves geographically.
This was a thoroughly enjoyable post from top to bottom. Absolutely hooked me from the first paragraph. And I'm so glad you brought up the ice cream vendors on the beach; it's a classic example I've been using that for years to explain clustering to my friends.
I grew up in southcoast Massachusetts, about an hour from Boston. But even in my fairly rural area, there are two Dunkin Donuts a mere 235 feet apart. Unlike in Revere, though, it's not a median-barrier-divided road!
Reminds me of the time I was overhearing a Ph.D. student at U. Chicago talk about her dissertation at a Starbucks in southside Chicago in the early 2000s (find me Jose Monkey!). She found that wherever a Starbucks located, it increased the probability of another coffee shop opening or not going out of business nearby. The conclusion was that Starbucks brought more coffee consumers to the area than it could serve and thus did not take demand overall from the local competition.
One of my dissertation papers looked at the location of Major League Soccer teams and found that placing two clubs really close to each other increased attendance for both clubs. I was skeptical at first because I assumed they would just compete for the same fans.
But it made more sense through the lens of clustering: when teams are close together, more people are exposed to the product, which can increase overall demand in the region.
I'm pretty sure I've read that, actually! When it comes to multiple teams in a location, my casual observation is that human behavior often takes over: people segment themselves, sorting into one team fan base or the other based on demographics and the teams engage in monopsonistic competition. Intracity wins for those fanbases can proxy for culture wars as well (Celtic & Rangers), and increase the value of fandom, perhaps at a cost to society.
Historically, I think a lot of multi-team markets for baseball are divided into white-collar/blue-collar crowds. For MLS, LAFC clearly caters to Chivas fans and NY is split into the "we're the original NY team" crowd and the "we're ACTUALLY NYC and you're NJ" crowd.
But these are just my observations—I don't have any empirical analysis to back them up :D
My initial motivation was the establishment of NYCFC as a "natural rival" for the NY Red Bulls, but they weren't introducing second teams for other cities. I think your intuition is likely right.
It comes up a lot in Public Choice economics. One of the original sources for the theory is: Black, Duncan (1948-02-01). "On the Rationale of Group Decision-making". Journal of Political Economy. 56 (1): 23–34.
You should visit any large Asian city. There are entire neighborhoods dedicated to one type of store. Hardware stores, gold shops, dress shops, men's clothes, all right next door to each other.
I should add that the rise of the western style big box stores are devastating these neighborhoods. About 2 miles from me in downtown Bangkok, a Big C and a Lotus big box stores are right across the street from each other. The surrounding area is full of empty shop houses.
Those are great examples of clustering! I've always been fascinated by that topic, but I never had the skill set to really look into it myself.
That last point sounds very similar to what I see in many smaller towns with a Walmart nearby. It's like the downtown areas are ghost towns because everyone wants to do their shopping all in one transaction. It's a real shame, but I know I was guilty of this when I was pressed for time.
It was one of the main reasons I wanted to teach Urban Economics this past semester! It was so much fun to learn about the different clustering models.
Reminds me of the median, voter theorem.
If you really want your mind blown, do a search on Advanced Auto Parts and AutoZone. The story I remember hearing was that a couple owned one of those stores, and then divorced. To spite their ex, one of the partners started a new auto store chain and tried to locate any new branch as close to the existing competing store as possible.
I hadn't really thought about this in the context of voting, but that makes a lot of sense to look at it that way, too.
I'm going to need a source on that last part because that feels like a story I want to learn more about!
This is such a fascinatin look at why Lowes and Home Depot cluster together! I never thught about it from the Hotelling model perspective before, but it makes so much sense. It's like they're both trying to capture the middle ground of the home improvment market rather than differentiating themselves geographically.
This was a thoroughly enjoyable post from top to bottom. Absolutely hooked me from the first paragraph. And I'm so glad you brought up the ice cream vendors on the beach; it's a classic example I've been using that for years to explain clustering to my friends.
I grew up in southcoast Massachusetts, about an hour from Boston. But even in my fairly rural area, there are two Dunkin Donuts a mere 235 feet apart. Unlike in Revere, though, it's not a median-barrier-divided road!
Again, great post. Brings to mind Lewis Black's Starbucks bit: https://youtu.be/Sg-J2TS13GA?si=jG9itzLiahjyaF11
Reminds me of the time I was overhearing a Ph.D. student at U. Chicago talk about her dissertation at a Starbucks in southside Chicago in the early 2000s (find me Jose Monkey!). She found that wherever a Starbucks located, it increased the probability of another coffee shop opening or not going out of business nearby. The conclusion was that Starbucks brought more coffee consumers to the area than it could serve and thus did not take demand overall from the local competition.
One of my dissertation papers looked at the location of Major League Soccer teams and found that placing two clubs really close to each other increased attendance for both clubs. I was skeptical at first because I assumed they would just compete for the same fans.
But it made more sense through the lens of clustering: when teams are close together, more people are exposed to the product, which can increase overall demand in the region.
I'm pretty sure I've read that, actually! When it comes to multiple teams in a location, my casual observation is that human behavior often takes over: people segment themselves, sorting into one team fan base or the other based on demographics and the teams engage in monopsonistic competition. Intracity wins for those fanbases can proxy for culture wars as well (Celtic & Rangers), and increase the value of fandom, perhaps at a cost to society.
Historically, I think a lot of multi-team markets for baseball are divided into white-collar/blue-collar crowds. For MLS, LAFC clearly caters to Chivas fans and NY is split into the "we're the original NY team" crowd and the "we're ACTUALLY NYC and you're NJ" crowd.
But these are just my observations—I don't have any empirical analysis to back them up :D
My initial motivation was the establishment of NYCFC as a "natural rival" for the NY Red Bulls, but they weren't introducing second teams for other cities. I think your intuition is likely right.
It comes up a lot in Public Choice economics. One of the original sources for the theory is: Black, Duncan (1948-02-01). "On the Rationale of Group Decision-making". Journal of Political Economy. 56 (1): 23–34.
You should visit any large Asian city. There are entire neighborhoods dedicated to one type of store. Hardware stores, gold shops, dress shops, men's clothes, all right next door to each other.
I should add that the rise of the western style big box stores are devastating these neighborhoods. About 2 miles from me in downtown Bangkok, a Big C and a Lotus big box stores are right across the street from each other. The surrounding area is full of empty shop houses.
Those are great examples of clustering! I've always been fascinated by that topic, but I never had the skill set to really look into it myself.
That last point sounds very similar to what I see in many smaller towns with a Walmart nearby. It's like the downtown areas are ghost towns because everyone wants to do their shopping all in one transaction. It's a real shame, but I know I was guilty of this when I was pressed for time.
I have always found spatial science and economics crossing points interesting. It is the reason I wanted to learn gis
It was one of the main reasons I wanted to teach Urban Economics this past semester! It was so much fun to learn about the different clustering models.