What Is The Best New York Book Of All Time?

On June 22nd, New Yorkers will go to the polls to vote in a new type of primary, featuring ranked-choice voting. To help prepare for this new style of election, which asks the voter to order their favorite candidates from most preferred to least, we've been conducting simulations. Last week, we asked our readers to vote on their favorite New York City books from a list of candidates selected by the New York Public Library; the librarians chose thirteen books that they felt "represent the diversity, excitement, and beauty of the cultural center of the world."

Today, we are announcing the result of our Big Apple Book Ballot: after 3423 votes, and 12 rounds of ranked-choice elimination, The Catcher in the Rye by J.D. Salinger wins with 1389 votes, or 58.7% of the final round votes, beating Just Kids by Patti Smith, which received 976 votes, or 41.3%, of the final round. Below is a summary of all 12 rounds of voting—you can page through each round to see which book was eliminated, and how the votes for that book were redistributed:

Something interesting that we can note right off the bat is that throughout the 12 rounds of voting, the order of the books hardly changes. Sometimes, in a ranked-choice election, as votes are redistributed from less popular candidates to more popular candidates, the order of the frontrunners will change, indicating that a candidate might not have led in overall first choice votes, but they have more support overall than a candidate initially ranked higher than them. This might happen because voters often assemble into blocks, like liberal or conservative, and one block might have more candidates than another splitting the vote, but as the ranked-choice elimination proceeds, the blocks coalesce behind a single front-runner, and the larger block then takes the lead.

In this case, except for a very close jockeying between House of Mirth and Bonfire of the Vanities in the first three rounds for 4th place, and between Just Kids and The Amazing Adventures of Kavalier & Clay in the first four rounds for 2nd place, the order of the books didn't change from the first round, indicating that the first round choices largely reflected the overall preferences of our electorate.

Here's another way of viewing the results, as a Sankey diagram, where you can see that as each round progressed, the votes were largely redistributed to the top candidates in proportion to where they stood in round 1 (the version below shows the final 6 rounds, for the whole diagram, click here to see a larger version):

We know that all of these books were critically acclaimed, with strong connections to New York City, as they were selected on that basis by the knowledgeable librarians at the New York Public Library, so what other factors could explain our result? Many are possible—I've chosen three obvious ones and summarized them in a table below: length, year published, and, as a proxy for popularity among the reading public, the total number of reviews on Goodreads:

I think we can dismiss length, as we see both long and short books in the top and bottom five. It's true that the average for the top 5 books is 438 pages, and for the bottom 5, only 300, but that's largely because of the long length of Kavalier & Clay and Bonfire of the Vanities, which are outliers in our list because of their unusual length. Eliminating the top two top and bottom outliers, I don't see a huge preference for short or long books among our voting electorate.

Year of publication seems a little more promising, where the top five books have an average publication year of 1970, and the bottom five average 1992. It stands to reason that older books might have had more time to gain popularity, and that newer books simply might not have had time to be read by as many people. But once again, eliminating the top and bottom two outliers in the list leaves the averages for the top and the bottom five books quite close together, with the top five actually averaging in younger than the bottom five at 1997 compared to 1989.

The third column of our table is more promising- Goodreads is a site that allows book enthusiasts to keeps lists of what they read. Looking at the number of ratings gives us a good idea of the popularity of a book in their community, which might be a rough proxy for book popularity among the general public. That's a lot of supposition, but clearly the top ranked books in our list are much more popular than the bottom ranked books. The Catcher in the Rye has about 2.9 million ratings, compared to Native Speaker, our last ranked book, with only 6761. The top five books average almost 700,000 ratings, while the bottom five average only 22,000. Even with the outliers removed, the top books are four time more rated than the bottom ones.

In some ways, this is a good illustration of what happens when candidates with high name recognition or popularity enter a race against other candidates who are less well-known; they tend to rank highly from the outset, because more people are likely to have heard of them, and even if they are not the voter's first choice, they're more likely to be in their top selection set. This has implications for Andrew Yang's candidacy in the mayor's race.

Coincidentally, when we asked candidates their own favorite New York books, Yang chose The Catcher in the Rye.

• Eric Adams: The Power Broker by Robert Caro
• Shaun Donovan: Invisible Man by Ralph Ellison
• Kathryn Garcia: Let The Great World Spin by Colum McCann
• Dianne Morales: Down These Mean Streets by Piri Thomas
• Maya Wiley: Go Tell It On The Mountain by James Baldwin
• Andrew Yang: The Catcher in the Rye by J.D. Salinger

(Scott Stringer and Ray McGuire did not answer)

There is another point to be made about popular candidates. In our race, by the 12th round, 30.9% of votes were "inactive", meaning all of the books on that voter's ballot were knocked out before the final round, and their vote did not figure into the final decision. This would have been avoided if someone ranked all thirteen books, but in many ranked choice elections, only a limited number of selections are allowed. In New York's primary, that number is five, no matter how many candidates there are. So, if your ballot does not include a popular candidate in at least one of the five positions, there is a chance your vote may not be counted- it's as if you didn't vote at all.

If you want to explore our book rankings in more depth, visit our partner, RCVis, which provides a few more types of diagrams. And if you want to introduce a theory on why the vote turned out the way it did, add it in the comments below.