Okay, this piece is going to be a little different than the others on this blog (hence the different tab!) for a couple of reasons. First, it’s about New York City rather than Kirkwood or St. Louis, so if you’re not interested in that, feel free to stop reading now, (but I do think it’s fairly closely related to issues that plague St. Louis so you might as well keep reading just the same). The second thing to know about this piece is that it also doubles as an assignment for school. In fact, it is the final task I have left to complete for the NYU Masters program I’ve been engaged in for the past year. (Thank you for your grace as the gadfly content has fallen off a cliff as a result!) Now, without further ado, onto the project itself.
New York as Epicenter of Housing Crisis
The United States is in the midst of a housing crisis. Contrary to most of the online discourse, there’s a reasonable case to be made that New York is the epicenter of that crisis. In the period from 2010 to 2019, New York City has built fewer units of housing per capita (25 units per 1,000 residents) than Washington D.C. (71), Boston (46) and even perennial housing policy-goat (read: not G.O.A.T.) San Francisco which clocked in at 41 units per capita over the period. There is also no indication that New York’s position has changed over the intervening two years. While California has taken state-wide steps to address the housing shortage like legalizing ADU’s, Albany has taken nearly no action. And while housing policy is most effectively instituted at the state level, New York City bears some of the blame here as well.
By-Right Zoning
Since 2010, 80% of the housing built in New York has been by-right. No one had to say yes, developers simply submit their plans, the relevant city departments and agencies make sure the paperwork is in order, and boom, you’re allowed to build whatever the zoning code says you can. In most cases this means that there’s room to build taller and denser housing than currently exists, even if the zoning doesn’t allow for developments to be quite as tall and dense as demand for housing calls for them to be. Unfortunately, that limited built in elasticity is starting to be exhausted. According to the Citizens Budget Commission’s report on Strategies to Boost Housing Production in the New York City Metropolitan Area (the source of all the stats in the last two paragraphs and an incredibly useful resource overall), only 20% of New York’s lots zoned for housing still possess as-of-right potential to add housing via infill, the rest are built to capacity.
This leaves the city with two options for adding to its housing stock. First it could broadly up-zone entire neighborhoods and essentially raise the zoned capacity cap that it is rapidly approaching. This method is the one recently employed in neighborhoods such as SoHo and Gowanus. The second method is to keep the current zoning in place and demand that each proposal that goes outside the limits of that zoning receive special approval via the city’s ULURP process. The first is ideal. At a theoretical level, rules should be transparent and consistently applied to everyone. But also, because projects that abide by these rules are guaranteed to be approved, developers are free to buy property, assemble investors, and hire architects and so more housing gets built under this method too.
But the job of the politician is to gain power and to use that power to benefit their constituents. And so city councilors often prefer the second method, the method where every proposal requires discretionary approval and every project could theoretically require their okay, and via the miracle of representational democracy, the okay of the community.
Member Deference: Real or Scape Goat
Here’s where another important political institution, the practice of “member deference,” comes in. Because, New York’s city council, like Chicago’s and St. Louis’ Board of Alderman and those across the country, operates under informal rules in addition to the formal ones. Member deference (or aldermanic deference in St. Louis) is the idea that the entire council should defer to to preference of the representative of the district in which the project will be located. If the local representative doesn’t like the project, you should vote against it, even if you happen to think it’s a good idea, because you expect the same treatment down the line. If, say, a landfill is proposed for your district, you want to make sure the other council members’ votes will be there to block it, so if that means blocking a harmless new condo building on the Upper-west side, in the interim, so be it.
The problem is, the landfill has to go somewhere. And the condos have to go somewhere too and if everyone gets to block things in their own home district, it becomes really hard to build things anywhere. As the number of by-right development opportunities dwindles and more and more projects require council approval, the problem grows. It’s not the only reason why housing in New York is expensive, it’s not even one of the main reasons, but it is a reason of growing concern. And so this year, when AOC was trying to determine which city council candidates to endorse, one of the questions she asked in order to decide is the one below. Essentially: Are you willing to toss the practice of member deference?
I wanted to know if member deference was a substantial part of this problem or simply a maligned scape goat, symbolic of the sort of inside baseball people despise in politics, but of rather limited consequence.
Methodology & Considerations
Examining the influence of individual city councilors is difficult. New York is huge and the pressures its 51 council districts face in incredibly diverse. How do you compare housing that isn’t built in the Greenwich village because residents don’t want to hear a cement truck backing up at 6am to housing that isn’t built in Brownsville because developers don’t think the profit margins aren’t large enough there? How do you “hold all else equal,” when nothing is equal? I think we look to the boundaries.
New York City Council districts have borders. These borders are drawn along streets; one side of the street is one city council district and the other side another. Now, theoretically, there shouldn’t be all that much that is different from one side of a street than another. If someone wants to live on Wythe Ave in Williamsburg, it’s pretty unlikely that they’re going to be super picky about which side of Wythe they live on. Developers know that people want to live on Wythe, know that those people don’t really care which side of Wythe they live on, and so are free to pick which ever side of Wythe they think will be easier to get approval for if their project deviates from the existing zoning code. Because the West side of Wythe is represented by one city councilor, and the East side by another, differences between those councilors in their attitude towards development could lead to asymmetrical development with one side of the street getting tons of new housing and the other side getting very little.
In order to examine this potential relationship, we pull from the Department of Buildings database on housing permits filed between 2010 and 2020. That database includes all demolitions, major alterations, and new constructions involving housing units in any capacity. Because New York City Council district boundaries were redrawn in 2013 based on the 2010 census, we then throw out all the permits filed prior to 2014, the first year in which the new districts were in effect and the thus the first year in which a pattern along the current boundaries would theoretically begin to emerge. We then tie the permit data to a database compiled by yours truly of who was the council member of each district when. We then plot those permits on a map to examine any boundary effects that might be present.
Maps
Below is a comprehensive map of all the housing demolitions, major alterations, and new constructions for which permits were filed from 2014-2020. Each circle represents one project with the size of the circles corresponding to the number of units the project contributed to New York’s housing stock. If you click on a given circle, you will be given the name of the city councilor in the district at the time, the year, the permit was filed and the number of units the project contributed. On the left hand side of the map, if you click on the layer button, you can toggle on and off different criteria for which permits appear and which political boundaries you want to compare them to (right now I have school district boundaries and city council boundaries, more to come soon). I recommend toggling on All Housing Permits and Landmarked Housing Permits one at a time, as landmark permits are merely a sub-category of all permits.
Remember, we are looking here for a relationship between the boundaries themselves and the housing that is built. Please, explore the map and see what you can come up with (and throw any interesting boundaries you find in the comments below!)
Preliminary Findings
The complete permit map is hard to decipher. There are permits everywhere so it’s difficult to find places where development favors one side of a boundary more than another. But this, to some extent, is to be expected. If 80% of those permits are by-right, that means 80% of the data points are noise when it comes to the question of member deference. So then we just look at permits filed for landmarked sites. Why? Because landmarked sites, by definition, require discretionary approval. When it comes to landmarks, the councilor, if they want, can have a say. And when we narrow the data set to just projects that require discretionary approval, it does seem that small patterns begin to emerge. The most recognizable of these patterns are found in Manhattan, where the demand for housing is the greatest and the density of landmarked sites is enough that the naked eye can perceive what’s likely going on. The most obvious case is that of City Council District 2.
Future Research
While the map is fun, and I would argue useful in the exploratory sense, it can’t tell us all that much about the statistical relationship between intra city political boundaries and the development of housing. In order to make a definitive claim, we’d have to do some discontinuity analysis to determine if there was statistical significance along these boundaries. One method of achieving this analysis would be to measure the distance from housing developments to city council district boundaries using shape files. This would require analysis of each individual council district in comparison to each of its neighbors with the x-axis representing distance from the boundary and the y-axis the number of net housing units added. If a significant jump in the unit trend line could be detected at the median of the x-axis (the point representing the council boundary itself) a relationship could be inferred between district and housing production could be inferred.
Additionally, the difference in development between councilors within the same district is under explored. To answer that question would require a discontinuity analysis across time rather than across space, a proposition made more difficult by the relatively limited sample size.
The bad news is, most of this is likely above my pay-grade. At least without significant handholding from someone with more statistical expertise than myself. The good news is that this entire project is open source, (you can find everything you need at my git-hub page here), and that person might be you. So use it, I would love for you to. And when you do, let me know what you discover.