Where are the libraries? Where are the bookstores?
I haven't blogged about LibraryThing Local here, on Thingology. So, for the benefit of those who don't read the main blog, LibraryThing Local is a new sub-site devoted to finding, mapping and describing the world's bookstore, library, book fair and festival—as well as all the readings, signings, lectures and other events they host. Open to all for three days now, LibraryThing Local just hit 10,000 venues—all user-contributed.
As it grows, LibraryThing Local is geting more interesting. Below are some interesting visualizations of where the world's cities have bookstores (green dots), versus where they have libraries (blue dots).
Although none of the maps—with the possible exception of Cambridge—are complete, and not all the libraries are public, the pattern is clear: Bookstores cluster together in the high-traffic center; public library branches spread out into the outlying areas and are separated from each other evenly like identically-polarized magnets.
I don't think this basic fact will come as a surprise to many, but it's striking even so. It's worth thinking about why these two institutions—so different but also sharing much—are positioned so differently in space.
I think the easiest explanation is the difference between economics and politics. Economics favors businesses that can create the most amount of happiness—which is to say revenue— whether or not this makes access difficult for some people. Representative politics favors solutions that give all citizens good or equal access to the resource, even if the resultant distribution is inefficient in economic terms.
So, bookstores go where they're going to survive and grow. High-traffic areas are best for that, and competition isn't necessarily damaging and may even be good.** By contrast, library branches are never clustered together, which would seem inefficient. And towns position branches, either directly or through a process grounded in neighborhood representation, to ensure that no area is left out.
That's my take. There are, I'm sure, other good explanations. Here are some:
*Los Angeles is the anti-case. It's so spread-out that the bookstores have nowhere to cluster.
**Take Ann Arbor's Shaman Drum, an independent, the national flagship Borders, and the excellent used bookstore Dawn Treader are arrayed in a tidy row.
As it grows, LibraryThing Local is geting more interesting. Below are some interesting visualizations of where the world's cities have bookstores (green dots), versus where they have libraries (blue dots).
 |  |
| Cambridge, MA | Dublin, Ireland |
 |  |
| Sydney, Australia | Chicago, IL |
 |  |
| Toronto, Canada | Houston, TX |
 |  |
| Minneapolis, MN | Los Angeles, CA* |
I don't think this basic fact will come as a surprise to many, but it's striking even so. It's worth thinking about why these two institutions—so different but also sharing much—are positioned so differently in space.
I think the easiest explanation is the difference between economics and politics. Economics favors businesses that can create the most amount of happiness—which is to say revenue— whether or not this makes access difficult for some people. Representative politics favors solutions that give all citizens good or equal access to the resource, even if the resultant distribution is inefficient in economic terms.
So, bookstores go where they're going to survive and grow. High-traffic areas are best for that, and competition isn't necessarily damaging and may even be good.** By contrast, library branches are never clustered together, which would seem inefficient. And towns position branches, either directly or through a process grounded in neighborhood representation, to ensure that no area is left out.
That's my take. There are, I'm sure, other good explanations. Here are some:
- Google Maps dots are all the same size, but a city's main library is generally far larger than any branch library, and far larger, compared to a branch library, than city bookstores are to peripheral bookstores. If a city's main library were broken into bookstore-sized chunks, libraries would seem to cluster indeed!
- Libraries focus more on services to families, which naturally sends them where the families are.
- Libraries are often positioned near schools, which show a similar regional distribution.
- LibraryThing Local probably underestimates peripheral bookstores. Library branches are generally easy to find, but you need to know where a bookstore is to find it. You're more likely to know the big downtown bookstores.
*Los Angeles is the anti-case. It's so spread-out that the bookstores have nowhere to cluster.
**Take Ann Arbor's Shaman Drum, an independent, the national flagship Borders, and the excellent used bookstore Dawn Treader are arrayed in a tidy row.
Labels: city planning, economics, librarything local, politics, visualizations
posted by Tim at 7:51 PM
15 Comments:
That last is almost cheating, though, since Dawn Treader is just one block away from Borders! ;)
It's interesting to see how the distribution pattern is the same in just about every city.
(Small nitpick---the fifth map is mislabelled as Ontario, Canada; it's a map of Toronto.)
Howse about a KML file with all the locations, so they are viewable in e.g. Google Earth?
I was just about to mention the Ontario vs. Toronto error, but poodlerat got there first!
You left out St. Paul from the name of your Minneapolis map. St. Paul is where the library school is in the Twin Cities, and St. Paul will be hosting the Republican National Convention.
Interesting that you mention Ann Arbor. There used to be several other bookstores in that area, before Borders (and a few other factors, like Ann Arbor's obscene rents, but mostly Borders) drove them out. . .
What are libraires?
I think that specialist bookshops probably need to huddle together. A book junkie would get sore feet if he had a wide range of interests.
Don't forget that bookstores tend to cluster near colleges, for obvious reasons.
High-traffic isn't a guarantee of a bookstore presence, though. There's also socio-economic factors at work as well. Look at Brooklyn - almost all of the bookstores are in the Park Slope/Downtown Brooklyn area, with almost nothing in the outer areas except college bookstores and religious bookstores. It's not that people in these areas don't read - in fact, the libraries are always packed - but that books aren't something they feel the need to invest in (for pleasure, that is - school and religion are more by necessity).
That's what I found interesting, anyway. I wonder if it's really meant to be like this, or the presence of a decently-priced store would change people's buying habits.
Even in the case of LA, I think the general issue holds as long as you break it up into various suburban sub-clusters.
In Pasadena, which I'm most familiar with, the bookstores are mostly located on one particular street. (Until quite recently there were three good bookstores within mutual walking distance which was quite convenient). The public library branches follow the general pattern of near-uniform distribution.
The newer cluster (huddle) of Ann Arbor bookstores is Fourth Ave, not State St.
There's also this funny but persistent phenomenon of non-bookstores selling books (and often very interesting, very carefully selected books) amongst all the other merchandise. I'm not talking 2 titles at Starbucks, but rather the shelf or two of really cool books you can find at places like Vicki Honeyman's gift shop / studio / hair cutting salon / etc etc. Not Borders-sized, but the chance that any given book you will bump into will be interesting is way higher.
Neat! I blogged about this some myself -- I think if colleges were eliminated, the clustering effect would be even greater. University libraries cluster together (a bunch of the libraries on the Cambridge map are Harvard libraries), and college bookstores aren't located for economic gain so much as for convenience to their relatively small community.
So does anyone bother reading our comments? Leaving in a blatant error is bad.
Editing the error can take time too, but I'll fix it now.
Thanks.
It seemed like a 5-second fix (especially since you only changed the name, not the link), so I thought a week was enough time.
Incidentally, it wasn't a freebie originally. The free version came seven months later.
Post a Comment
<< Home