Examining Suffolk County’s Multifamily Housing Stock
In New York City, the high costs of rental housing are well documented. Low- and moderate- income renters have struggled to pay New York City’s high housing costs long before the COVID-19 pandemic hit. Median gross rent has increased steadily over the years, while median renter household income has not kept pace.
However it’s also New York City’s suburbs that play an important role in the economic future of the region, particularly as the region begins to recover from the economic fallout of the pandemic. In a recent report from the NYC Department of City Planning, analysis of employment, labor, and housing production trends illustrated that 973,000 regional residents — residents of Connecticut, the Hudson Valley, New Jersey, and Long Island — commute to jobs in the city. Of those in-commuters, 289,000 live on Long Island.
As the NYC DCP report shows, the southern and western regions of the country have seen faster job growth than the NYC metro region. With lower costs of living and strong job markets in these other areas, there has been little reason for younger workers to stay local. Indeed, while the labor force change has overall been positive on Long Island, adding 41,000 jobs since the Great Recession, there has been a loss felt solely in the age cohort of 25–54 year old workers. Further, according to an analysis performed by the Nassau County Comptroller, in 2017 nearly 34% of the college educated population was underemployed.
This is all to say that the housing stock on Long Island has been increasingly unable to meet the needs of its residents, and is key to the future prosperity of the region. Over the last 10 years housing production has largely been concentrated in New York City. This blogpost begins to explore some of the multifamily housing stock in Suffolk County, New York City’s most eastern suburb.
Based on available data of residential construction in Suffolk County, properties are concentrated in the south western and central parts of the county. There are fewer properties out east, most of which were built before the year 2000 (Image 2).
The Long Island Rail Road (LIRR) is an important commuter rail line connecting Long Island’s many towns and villages to New York City’s Penn Station. Suffolk County residents are likely less reliant, on average, to use the LIRR to commute to work than their Nassau County neighbors. This is due in part to the length of the commute, as well as the frequency of trains.
Many of the multifamily properties in my data appear to be clustered around transit, though there are some important outliers. For one, there’s a cluster of properties in the most south western part of the county in the towns of Babylon, Lindenhurst and Copiague. Directly north of this clustering we can see that there’s very little development around transit on the northern line. In addition, in the center of the county we can see that there’s quite a bit of development that’s not especially close to LIRR stations.
Of course, this doesn’t take into consideration the zoning restrictions that may also prohibit construction of multifamily properties in certain parts of this map. Zoning restrictions and political opposition may indeed be a prohibitive force against development in many of the sparse areas on the map.
Many factors go into the location decision making of builders and developers, as discussed — zoning, politics, accessibility to transit. However, beyond the existing stock, a look at construction shows a handful of planned or current construction taking place in primarily low- and moderate- income areas. The exception is a single senior housing facility being built in a wealthier area in the south (Image 4). High income census tracts are concentrated mostly on the north and south parts of Long Island, oftentimes neighboring tracts with very different income levels.
As these maps have shown, the construction of multifamily properties has been concentrated to some pockets of Long Island. Very little construction appears to be occurring in the eastern most towns and in areas of higher income.
An even more stark illustration of differing neighborhoods in the county is the map below, which looks at school segregation by public school district. Mapping school districts by the share of the student population that is white highlights a few important highs and lows. Outlined in red are the neighboring districts of Shoreham- Wading River (SWR) and Riverhead. As you can note from earlier, SWR generally consists of higher income census tracts, with no multifamily properties or LIRR stations nearby. In contrast, Riverhead is exactly the opposite — seeing lower incomes and more development.
There are a range of tools that might be utilized by New York State to solve the region’s pressing economic issues. Changing the zoning to allow for more homes and more transit-oriented development is one. Increased construction of all rental units, including affordable units, need be an important strategy of the region’s long term recovery. Above all, a sharp attunement to the legacies of systemic discrimination in housing and credit markets that created these current circumstances is imperative.
Data and Methodology
This data used in this post come from the following sources.
Suffolk County Department of Planning and Environment: The Research and Statistics unit of the Suffolk County Division of Planning and Environment analyzes a variety of statistical data about the region. The unit makes residential permit data available by town, largely compiled from the U.S. Census. I was interested in a more granular level of geography and instead utilised the county’s PDF files on residential and commercial development. For the purposes of this blog post I focused only on apartment complexes with 10 or more units, which notably leaves out condo and coop complexes. I first converted the PDF to a .xlsx before cleaning.
Using geocod.io I was able to geocode the list of apartment complex addresses. After reading into QGIS I noticed a few mistakes — some properties were misread as being located somewhere in Nassau County, and in one case in Vermont. After resolving these issues, adjusting the symbology to produce the point data in the maps above was pretty straightforward. I used rule based for Image 2, and single-symbol for Image 3. For the Image 4 I selected features using the expression function to create a layer of only those that contained Under Construction or Pre Construction statuses.
U.S. Census: census tract boundaries were downloaded from NHGIS and selected based on the COUNTYFP code (103). This shapefile was joined (by GISJOIN) to a .csv of 2018 5 year estimates of median income, also from NHGIS. Categorisation was based on HUD estimates of area median income.
New York State Education Department, BEDS Day Enrollment: enrollment , demographic, and expenditure data are available from the state as Microsoft Access Database files. The data used in this analysis are from 2019. I cleaned .csv exports from Microsoft Access and merged these to a shapefile of school districts, also available from the state.
After reading into QGIS, I adjusted the symbology to create the scale of share white. I selected the two school districts of interest and exported those as a separate layer to create the red outline.
Long Island Rail Station Stops: these data are available from CUNY Baruch, which I accessed through NYU. I first clipped by the layer of Suffolk County census tracts, before then creating a one mile buffer around each station to illustrate relative ease of access at a wider boundary than the point.