Category: Housing

Allowing cottage courts in Chicago requires changing the zoning code

Cottage and bungalow courts (also called clusters and pocket neighborhoods) are a common multi-family design typology in cities outside of Illinois. You’ll find them in California (where they’re more often called bungalow courts) and Tennessee, for example. They are characterized by having multiple detached one or two-story houses on a single and compact development site. The houses commonly face a shared green space and have shared parking behind them.

A primary benefit of cottage courts – and smaller footprint housing, generally – is that it creates more for-sale detached houses (which are in very high demand in Chicago) while sharing land [1]. In other words, cottage courts create more detached houses using less land.

It is not possible to build a cottage court in Chicago because of how the zoning code is written. The purpose of this article is to discuss specific text amendments that would have to be collectively adopted in order to allow this housing type.

screenshot of a page from a guide about Missing Middle housing in Chattanooga, TN. The image has a site plan for a cottage court, an isometric view of the rendered cottage court, a narrative about the sample project, and statistics.
A sample cottage court intended for an audience in Chattanooga, TN, created by the Incremental Development Alliance.

Zoning code barriers

  1. The Chicago zoning code allows only one principal building per lot. A detached house constitutes a principal building and only one of them is allowed. The code would have to be changed to allow more than one principal buildings per lot. If the city council wants to limit this allowance to cottage courts then other code would have to be modified to define a cottage court (similarly to how it has a definition of a townhouse). (Section 17-1-1300 – this anti-housing provision is in the first chapter of the Chicago zoning code!)
  2. Cottage courts should be fee simple [2], to make them easier to mortgage and sell, which means subdividing a lot into one lot per house. The Chicago zoning code does not allow subdividing a lot below the minimum lot area for the particular zoning district governing a lot. For example, currently in RS-3, if more than one principal building per lot was allowed, each cottage court house would have to occupy 2,500 s.f., which is an untenable size for cottage court development.
  3. Minimum lot area per unit standards also severely restrict development of cottage courts. To build four houses in an RS-3 zoning district one would need a 10,000 s.f. lot!
  4. Rear setbacks would need to be reducible, preferably without the need for a variation from the Zoning Board of Appeals. Because the houses are oriented to face a common green space at the interior of the lot (not at the front or rear of the lot), the rear of the house may be close to the side property line, violating the rear setback standard of ~30 feet.
  5. Side setbacks would need to be combinable or eliminated as a requirement for cottage court development, so that the houses could be closer together or the designer could have more flexibility in their orientation.
  6. Parking requirements would need to be more flexible, both in quantity and in design. Parking is currently allowed only in the rear setback, but these houses may not have a rear setback because of their inward orientation. To maximize shared green space parking requirements should be reducible for this housing type. The Chicago TOD ordinance may be relevant here, as it now applies in RM-5, and higher, residential zoning districts, but the cottage court needs to be allowable in the RS-3 and RT-4 zoning districts (as these are for more common).

Attorneys, designers, and developers: Are there any other Chicago zoning code standards missing?

Examples of cottage courts

I know of a few cottage courts in Chicago (et. al. means there is an address range):

  • 7436 S Phillips Ave (et. al.). These were built prior to 1952, according to historic aerial images, and are individually parceled (see barrier #2 above).
  • 7433 S Euclid Pkwy (et. al.). The houses were built between 1938 and 1952, according to historic aerial images, and are individually parceled. Two of the parcels are vacant.
  • 1802 S Kildare Ave (et. al.). The houses were built after 1950.
  • 3020 N Waterloo Ct (et. al.). This townhouse court was built between 1970 and 1975. These houses are arranged like townhouses but they are not compliant with the present townhouse code (adopted in 2004) because they are too dense and have too little parking. The density is equivalent to 2.5 homes per standard lot (which, in Chicago, is 3,125 s.f.).
3020 N Waterloo Ct has 25 homes and one car parking space per home.

Burlington, VT, approved a zoning plan in March 2024 that permits “cottage courts” in places “where lots make it feasible”. Read their new Neighborhood Code, which created the graphics below showing before and after the code change.

Graphics showing the new housing types that are allowed in the updated Burlington, VT, zoning code. (I made the collage.)

All of the images below are from the Missing Middle Housing website, created by Opticos Design, an architecture firm that has popularized the term “missing middle”.

Notes

A tiny house village may also be considered a cottage court. The influential architecture firm, Landon Bone Baker, once designed a proposed tiny house village (knowing full well it was not legal) for Thresholds and Easter Seals in Chicago.

[1] It’s also assumed that sharing land means sharing land costs, and land costs are a significant part of purchasing a house. In the areas where demand for detached houses is the highest, land costs are also the highest. Cottage courts create more detached houses with less land.

[2] fee simple has a legal meaning, but here I mean “the house owner also owns the land beneath the house”, and the pair are collateral for the lender that, on paper, looks like any other detached house the lender has mortgaged.

Inclusionary zoning calculator will tell you how many units a developer can afford to make “affordable”

An “inclusionary zoning” calculator can help you determine how much affordable housing your town should require that developers build in their new construction residential buildings.

I learned about Grounded Solutions Network’s Inclusionary Housing Calculator at the second-ever YIMBYtown conference in Oakland, California, two weeks ago.

YIMBY (yes in my back yard) is a movement to reduce barriers to building more housing in order to be able to house everyone at a level they can afford. It’s a movement for other things, and it means a lot of different things to a lot of different people but the end result is that more housing needs to be built.

An interested person inputs a lot of values relevant to their local housing market into the IHC and it will calculate the cost of construction per unit and the rental income from those units, and then will figure the profit margin for the developer. What makes this “inclusionary” is that one also needs to enter the desired portion of units that are set aside as “affordable” (to people making a certain income) and subsidized by the developer’s rental income.

I put the IHC through a real world exercise by inputting as much data as I knew about a rejected proposal in Pilsen.

The first proposal from Property Markets Group had 500 units, and 16 percent of them were set aside (news on this and their subsequent proposals) [I cannot find the source of the “16 percent on-site” factor]. Chicago’s Affordable Requirements Ordinance, or ARO, requires that 10 percent of the units are affordable, and that 25 percent of those 10 percent must be built on site. The other 75 percent can be built on site, or the developer can pay an in-lieu fee per unit.

Needless to say, 16 percent on-site is much, much higher than 25 percent of 10 percent. A neighborhood organization, the Pilsen Land Use Committee, however, requires 21 percent in the area, and the city council member, Danny Solis, 25th Ward, adheres to.

PMG said they couldn’t go that high, and that’s what I wanted to test.

According to this Inclusionary Housing Calculator, could the developer make enough profit (considered as 10 percent) if the building had 21 percent of units as affordable?

In this exercise, the answer was “no, PMG could not make a profit if they had to set aside 21 percent of the units as affordable.”

But the calculator showed that they could earn a 12 percent profit if 16 percent of the units were affordable. 

Some of the inputs are actual, like the sale price of the land (found in the Illinois Department of Revenue’s transactions database), but I had to make up some inputs, including the apartments’ bedroom mix, and the future rental prices of those apartments.

Further reading

  • It’s tough for people to move into one of these set-aside apartments in Chicago (DNAinfo Chicago, July 28, 2017)
  • Inclusionary zoning cannot create enough affordable units (City Observatory, February 11, 2016)
  • Other housing cost calculators like this one (City Observatory, July 26, 2016)