1. Permitting Process
- Accessory Dwelling Unit (ADU) Permitting (Special Permit or Allowed/Conditional)
2. Occupancy Requirements
ADU Employees or Family Only (Yes or No)
ADU Prohibition on Rental (Yes or No)
ADU Owner Occupancy Required (Yes or No)
ADU Elderly Only (Yes or No)
3. Physical Requirements
ADU Restricted to ONLY Primary Structure (i.e., No Outbuildings like Garages) (Yes or No)
ADU Max. Size (% of MAIN UNIT) (Percentages ranging from 25* to 50%)
ADU Max. Size (SF) (Floor area ranging from 600 to 1500 sq ft)
Individual districts in Fairfield County have combinations of these variables. In general terms, the level of ADU zoning restrictiveness will be based on the cumulative number of requirements imposed by each zoning district. We translated these attributes into a flagging system based on boolean values (0 and 1). The flags (or 1 values) indicate greater restriction. For example, requiring a special permit or a public hearing would be considered more restrictive than an as-of-right process, and as such is marked with a positive flag or 1. In general terms, the greater the number of flags the more restrictive the zoning for ADUs.
All the field attributes indicating Yes or No, as described by Desegregate CT’s data compilation, were replaced with boolean values (1 for Yes and 0 for No). To flag dimensional criteria as boolean values, we defined the thresholds for when parameters are considered restrictive. In general terms, requirements that limit the size or the physical configuration of an ADU are more likely to increase restrictive conditions. We defined these as anything restricting the size of the ADU to 30 percent or less than the principal unit (consistent with approved legislation) and restrictions that limit the maximum area of an ADU to 800 square feet or less.
We also weighted the individual flags (attributes) based on the themes: permitting, occupancy, and physical requirements. These were weighted in a way that assumes that all three themes have an equal level of influence in restricting ADUs. Weighting the attributes in this way also corrects the variability in the number of attributes compiled per theme (e.g. the permitting theme only has one attribute, while occupancy has four).
We then joined the ADU zoning attribute information with land use and housing stock data available at census block level. We used the spatial join feature in ArcMap and tested multiple matching options and spatial rules. The matching option that most accurately aligned boundary discrepancies between zoning districts and census blocks was “closest” with a specification of a quarter mile distance in the search radius parameter.
Finally, we estimated the average ADU restrictiveness score per municipality by calculating the mean based on all zoning districts within the towns boundaries.
Estimating ADUs and Single-Family conversions in Fairfield County
RPA staff conducted a geospatial analysis to estimate the potential for creating Accessory Dwelling Units and single-family conversions in Fairfield County, Connecticut (e.g conversion of single-family homes into duplexes). The result is not a forecast, rather an aspirational vision that would maximize the region’s housing potential over the long term.
For comparative purposes and to project the potential impact from policy and regulatory changes, we factored the ADU zoning restrictions score. In this alternate projection, the share of incremental units is constrained and much lower to the aspirational vision. This alternate scenario also follows irregular growth patterns as a result of a patchwork of zoning regulations.
The mix of variables on a given geography affects the likelihood of ADU creation or conversions of single-family homes in Fairfield County. The greater the number of favorable factors (large lots, proximity to transit, assessed values that suggest underdevelopment, and flexible zoning rules) the higher the probability of ADU creation and single-family conversions. To estimate the potential of ADUs and conversions, we applied a geographically weighted methodology that projects housing stock growth based on the mix of variables and their interrelations at the block level.
The variables included in this analysis are the following:
Number of Dwelling Units per housing type per block - Estimate Total
Single Family Attached - Estimate
Single Family Detached (small) - Estimate
Single Family Detached (large) - Estimate
Proximity to transit - Boolean Value (0 or 1)
One mile to rail transit station - Yes/No
Quarter mile to a bus stop - Yes/No
Market Value of Single Family Residences - Mean values per Locality
Mean Improvement to Land Value Ratio -
Max Improvement to Land Value Ratio
Mean Total Property Value
Mean Improvement Value
Mean Land Value
Zoning Restrictions on ADUs - Score/Index (range between 0 - 10)
Permitting process
Occupancy Requirements
Dimensional Requirements
After compiling all the variables listed above, we estimated an incremental baseline for the year 2040 consistent with the assumptions and housing needs assessment used by the Fourth Regional Plan. Demand was based on the housing needs assessment of the Fourth Regional Plan aspirational scenario. Under such a scenario the tri-state area would add 2.6 million units (from 8.8 to 11.5 million). Fairfield in particular would go from 362,000 in 2015 to 475,000 in 2040 (an increment of 113,000 units).
Under this scenario, Fairfield County would see an additional 40,000 housing units (or an 11 percent increment over current conditions) in the form of ADUs and conversions. The majority of these would be located in neighborhoods served by public transit, and localities that have improvement ratios below the county average.
This projection assumes that the number of units per block and derived from single-family residences would increase by the following percentages:
5% of Single-family attached
5% Single-family detached (small)
10% Single-family detached (large)
Additional increments would depend on the following attributes:
5% more when located in proximity to a bus stop
10% more when located in proximity to a transit station
5% more when the municipality has a mean improvement ratio below the county’s average
This potential future scenario outlines a net increment of housing units that ranges between 5% and 30% above the existing stock of single-family units per block. In this aspirational unconstrained scenario, it is also assumed that zoning reforms and programs that support ADUs and single-family conversions are in place.
For comparative purposes and to project the potential impact from policy and regulatory changes, we factored the zoning restrictiveness score. In this alternative constrained scenario, the increment of housing units uses the unconstrained potential, but assumes an inverse proportion depending on the level of restrictions as ranked by the zoning index.
In this alternate projection, the share of incremental units is less than half when compared to the aspirational vision. The alternative scenario would follow irregular growth patterns along political boundaries and a patchwork of zoning regulations. Without zoning reforms, the net increment would be almost a third when compared to the projections estimated in the aspirational unconstrained vision (approximately 16,000 additional units or only 4% over the existing housing stock).