Habitat for Humanity, Washington, D.C. & Northern Virginia
3/3/2023
To: The Members of the Arlington County Planning Commission and Arlington County Board
CC: Mark Schwartz, County Manager
Claude Williamson, Director, Community Planning, Housing and Development
Anne Venezia, Housing Director
Anthony Fusarelli, Planning Director
SUBJECT: Public Comments on ZOA-2023-02 An ordinance to amend, reenact, and recodify the Arlington County Zoning Ordinance (ACZO) for Expanded Housing Option Development
Habitat for Humanity of Washington, D.C. & Northern Virginia (Habitat DC-NOVA) appreciates the opportunity to provide comment on the proposed amendments to Arlington County’s zoning policies to support missing middle housing, hereafter referred to Expanded Housing Option (EHO) development. Consistent with our prior public comments dated July 7, 2022 and January 23, 2023 (attached for reference), we encourage the Planning Commission and County Board to approve the most expansive version of EHO development in March 2023. Habitat DC-NOVA believes that allowing modest increases in density and housing type diversity in Arlington’s single-family neighborhoods is a necessary first step in ending exclusionary zoning, expanding housing choice, and improving affordability. The changes also would better enable Habitat DC-NOVA and the Virginia Statewide Community Land Trust to produce and preserve committed affordable homeownership units. The comments that follow focus on four main themes:
• Improving affordability in Arlington’s R-Zoned Districts requires expanding housing supply and building forms beyond single-family detached. The current EHO proposal is a foundational step in this process.
• To achieve the intended goals of missing middle housing, the County should not place artificial limits on the location and number of EHO developments. We encourage the County to adopt options 2A, 10B, and 7B.
• Greater regulatory flexibility opens more pathways to success. As such, the County should not adopt strict regulations on size and design. We support options 4B, 12B, 8A, 5C, 5E, and 3A. We also support the adoption of tree planting requirements under Option 6A. We prefer option 11B. We strongly oppose option 5B.
• The County should take additional action to maximize the impact of any approved zoning changes. We particularly support coordination with the County’s ongoing homeownership study, a robust educational/outreach effort, and the adoption of additional incentives for committed affordable housing, open space preservation, and high-quality/energy-efficient design. Finally, we encourage the County to make a “down payment” on affordable homeownership in the County’s Fiscal Year 2024 Budget by committing resources for the acquisition of properties to be added to the committed affordable homeownership stock.
The following pages provide more detailed explanations and justifications for each theme.
1. Improving affordability in Arlington’s R-Zoned Districts requires expanding housing supply and building forms beyond single-family detached. The current EHO proposal is a foundational step in this process.
In the midst of a nationwide housing shortage, and with much of Arlington already built out, the County cannot afford to unnecessarily restrict which neighborhoods are open for growth. We believe that adopting the EHO proposal will incrementally improve the broader attainability of Arlington’s housing market while also dramatically increasing the opportunity to create committed affordable homeownership units. The inventory of existing lower-cost single-family homes is small (see Appendix A) and shrinking every year under status quo policy. The County can either tie itself to single-family detached homes and let the starter home eventually go extinct, or it can create a pathway to create new homes that moderate-income households can at least hope to afford (see Appendix B).
As an affordable homeownership developer, we believe in the importance of expanding opportunities for ownership to a wider range of people. Throughout the past year, there has been considerable debate about the impact of any missing middle policy on homeownership. This is a false choice. More broadly, expanding housing choice (particularly in “rental deserts,” see Appendix C) is important to achieve the wider goal of combatting socioeconomic segregation. Improving attainability and providing more housing options in more areas can improve the housing stability of renters, which in turn can improve readiness for homeownership in the longer-term.
Furthermore, expanding opportunities to build rental housing is unlikely to come at the cost of reducing homeownership choice. First, given the net increase in density, it is probable that the number of owner-occupied housing units in R-Zoned Districts will increase even if the proportion of homeownership units declines. If 10 owner-occupied homes are replaced with six-plexes and just two of those EHO developments were built as condominiums, it would represent a net increase in the owner-occupied stock. Contrary to many concerns, homeownership units in 4-6 unit buildings is achievable and scalable. Our affiliate has considerable experience developing homeownership units in a variety of configurations, including stacked flats, two-over-twos, and smaller multifamily structures. And we have successfully built 3+ bedroom units within the smaller envelopes that would be characteristic of an EHO multiplex.
Homeownership development in mixed-townhome and two-over-two configuration in Washington, DC
Unlike rental housing, there are few dedicated subsidies for the production of committed affordable homeownership units. This limits the scale of potential homeownership development, and makes it difficult to compete for larger sites and sites in multifamily zoning districts. Given these funding limitations, expanding EHO opportunities throughout all R-Zoned Districts dramatically improves the ability of Habitat DC-NOVA to compete for sites. In addition to reducing the “barrier to entry” of site scale and acquisition price, each additional unit that can be added to a site reduces the per-unit subsidy that is needed to produce committed affordable homeownership.
2. To achieve the intended goals of missing middle housing, the County should not place artificial limits on the location and number of EHO developments.
All neighborhoods can and should contribute to Arlington’s growth and efforts to improve attainability. While the latest proposal would allow EHO developments in all R-Zoned Districts, some of the options being proposed could significantly limit the viability of more affordable development types. The ability to increase the number of units on a parcel is the key to making EHO viable on the largest and most expensive lots. We were disappointed that the options for 7-8 unit EHO developments were eliminated. The County should not impose any further restrictions on 5-6 unit EHO development. Doing so would reduce development viability in Arlington’s costliest, most exclusive neighborhoods, undermining the goals of integration and geographic distribution. We therefore encourage the adoption of Option 2A, establishing the minimum site area for EHO development at the same minimum lot area for each R-Zoned District. Option 2D is also a reasonable approach (though clearly inferior to Option 2A), as County staff analysis shows that it would not overly restrict 5- and 6-unit EHO developments on mid-sized lots. Among the other options, 2B is counter-productive from a transit-centric approach (limiting 5-6 unit EHO developments on smaller lots near transit). Option 2E is the most problematic, as it uses an overly restrictive definition of transit accessibility that excludes Langston Boulevard and also includes onerous lot size requirements.1 We believe that Option 2E would reinforce patterns of inequity, as many of the areas where 5-6 unit EHO developments would be prohibited are disproportionately white2 and wealthy.
In the spirit of enabling EHO development at the broadest geographic scope possible, we support Option 10B, which would enable EHO development in R-Zoned District areas that overlap with established planning districts. We understand that the County’s preferred redevelopment approach in these areas is to encourage higher levels of density through site assembly consistent with broader neighborhood plans. That development path would still remain under option 10B. However, not every site is appropriate for assembly, and not every owner is capable of undertaking larger scale development. For these properties, the “lowest-hanging fruit” under the status quo and Option 10A would still be by-right, large-scale single-family detached redevelopments. Option 10A would prevent interior conversions and “addon” EHO developments that do not require demolition, undermining one of the key benefits of the current proposal – allowing existing owners to “evolve” their homes to accommodate changing life circumstances. If the County is worried that adding another development path would prevent site assembly, it should revisit the relevant plans to reduce barriers/increase incentives for doing so, rather than restricting housing choice.
Finally, we strongly oppose a cap on annual EHO development. A cap creates a race to be early to file for permits, and we fear this time-sensitivity could create unintended consequences, such as:
• Skewing EHO development choices toward projects that are at the top of the market and toward product types that already have precedent in the market (larger duplexes and three-unit townhomes).
• Putting 4-, 5-, and 6-unit EHO developments that do proceed at a competitive disadvantage for getting permits, as these typologies may take longer to design.
• Discouraging overall design creativity, as the process for refining and optimizing plans to better fit the site and neighborhood can be iterative and, at times, lengthy.
• Preventing the development of committed affordable housing, as such projects may take longer to conceptualize and assemble financing.
While a cap may seem like a reasonable nod toward cautious, data-informed, and iterative policymaking, there is a risk that the opposite may be true. The above distortions would limit the usefulness of any data that the County could collect on the impacts of EHO development, as the projects that move forward would be less representative of what could be achieved without a cap. Therefore, we support Option 7B. If a cap is necessary, it should be set as high as possible and sunset (Option 7C). We also believe that any cap should be at the R-Zoned District level (instead of a County-wide cap) to encourage greater geographic distribution. EHO developments that include committed affordable housing (for example, units targeted to households below 80% AMI, Community Land Trust units, etc.) should be exempt from the cap.
3. Greater regulatory flexibility opens more pathways to success. As such, the County should not adopt strict regulations on size and design
Generally speaking, restrictive regulations make housing less attainable and committed affordable housing more difficult to produce. While many regulations (such as those related to health and safety) are still worthwhile, others raise the cost of housing and/or inhibit development to serve ends that are, at best, at the eye of the beholder. To promote good design, it is generally better to provide incentives for certain practices and address actual harms through policy, rather than using restrictive zoning policy as a proxy. As a final rule of thumb, there should be parity between building types – regulations that apply to EHO development but not single-family detached development tip the scales further toward development of the latter.
Specific to the current EHO proposal, there are already aesthetically-focused provisions that could restrict design options (building entrance orientation requirements, prohibition on exterior stairs to upper story), lack parity with single-family detached housing (trash/HVAC screening rules), and/or could lead to unintended consequences (prohibitions on front/side tuck-under garages could lead to longer driveways & more impervious space).
Conversely, we support the tree planting requirements under Option 6A. Though it does impose an incremental cost on development, it is appropriately scaled and is an example of a provision that directly addresses a clearly defined, unambiguous community need.
We also support other potions that encourage flexibility:
• Option 4B (allowing 5% lot coverage allocation for detached garages to be used for housing).
• Option 12B (allowing accessory dwellings). This option comports with the goal of expanding housing choice, allowing existing homeowners to adapt their own homes (interior accessory unit) or overall property (detached accessory unit) to meet changing circumstances. This would also allow an “EHO addition” to an existing home when a detached accessory dwelling is already present.
• Option 8A (allowing condos/co-ops on non-conforming lots without a use permit).
Caps on gross floor area (GFA) for EHO developments should not be imposed if they are not also applied to single-family detached housing. Given the regrettable removal of Option 11C from the request-to-advertise, we encourage the adoption of the least restrictive option, which is 11B. We believe that options 11A and the staff proposal are too prescriptive. In particular, we share the concerns raised in the County staff report that these options could discourage interior renovations of existing large homes to include EHO development. While we understand concerns about massing, it is likely that other provisions (height, setbacks, lot coverage, parking) would be more of a binding constraint for redevelopment than GFA caps.
Parking requirements are likely to have among the most impact on the viability of EHO development. Parking minimums impose direct costs, increase impervious space, and can reduce the number of units that are viable in a development. Developers know the market they intend to serve, and will often build the amount of parking that they know their purchasers/renters need, regardless of minimums. Not all HFH DC-NOVA properties are built to the parking minimum. While concerns about street parking availability are common, County data shows that crowding is geographically limited. Crowding is also often the result of spillover from other nearby commercial uses, rather than the housing units themselves. It is better to directly address parking scarcity where it exists through metering and permitting systems rather than by requiring excessive off-street parking. As such, we prefer option 5C (using the definition of transit proximity utilized in options 5A/C, rather than 2E), though option 5A is also a reasonable approach. We support the curb cut flexibility provision (Option 5E). We strongly oppose Option 5B, as developers should have the option to make a data-informed case for further parking flexibility.
Finally, we generally support Option 3A, which requires sites of at least one-acre to proceed via Special Exception Development. If Habitat DC-NOVA were to have the opportunity to develop a one-acre site, having a streamlined method for approving a multiple building EHO development would be beneficial. However, we encourage the adoption of clear standards for what might be required of such development. The benefit of such a policy is dependent on the value of the flexibility that could be granted and the timeliness/predictability of the process. It should be noted that adopting Option 3A would have the unfortunate trade-off in which the “lowest-hanging fruit” from a regulatory (if not financial) perspective is a large single-family home redevelopment.
4. The County should take additional action to maximize the impact of any approved zoning changes
Adopting the EHO policy would represent a positive, foundational step for improving housing attainability and choice in Arlington County. It is important for the County to build on that foundation. Throughout the missing middle study and debate, there have been continuous calls for additional analysis and the study of potential impacts. We believe that the County’s study has been robust and extensive and conclude that the results support the EHO proposal. Given the complexity of the housing market, there are limits to the predictability of how a market will respond to any policy change. As such, the best approach for data-informed policymaking is to implement the EHO proposal in combination with a robust plan to track development, measure results, and incorporate policy changes as warranted by the results in practice. We believe the plan for measurement and analysis included in the staff report is strong and should be implemented and adequately funded.
We also want to highlight the importance of the staff report’s plans to provide Educational Materials for Homeowners, focusing on parts of the County designated in the Racial and Socioeconomic Equity Analysis. We strongly support this effort and encourage the County to fund it robustly. Predatory acquisition of lower cost properties (often from lowerincome and/or older adult property owners) is a problem nationwide. Given this change to zoning, it will be important for property owners – particularly those with lower incomes or from historically marginalized communities – to have accurate information on their options and rights.
We also believe that this zoning amendment, if properly implemented, could create the opportunity for other policy and funding innovations – including those that are within the scope of the ongoing Homeownership Study3 – to provide affordable homeownership opportunities more efficiently and effectively for the households that we typically serve (at or below 80 percent of area median income). The County should proactively focus on this issue during this study process. It should also make a “down payment” on affordable homeownership in the County’s Fiscal Year 2024 Budget by committing resources for the acquisition of properties to be added to the committed affordable homeownership stock through the Virginia Statewide Community Land Trust or other mechanisms.
Finally, as the policy is implemented, data is gathered, and lessons are learned, the County should explore opportunities to adjust, enhance, and/or expand the policy to achieve broader community goals. This can include minor changes that are outside the scope of this advertised policy. For example, the County may consider addressing the prescriptive design measures discussed above. We would also support an effort to allow for Board of Zoning Appeals approval for minor modifications to EHO development standards that allow other community goals to be achieved (such as a modification to setbacks that would allow for additional tree preservation).
In lieu of prescriptiveness in the zoning code, we believe the County can take proactive steps to encourage better design practices or other community benefits. This could include the creation of a “design/form book” and/or pre-approved designs that are specifically developed to fit various neighborhood contexts. Such design efforts could be for educational purposes, or could be a mechanism for providing incentives. For example, a developer could receive administrative approval for a development that conforms with the form-book (or other form-based policy). The form book would provide designs that include modest additions to height, GFA, and/or units in exchange for preserving open space/trees, incorporating stormwater management practice, and/or adopting energy-efficiency measures. The County should also explore incentives for committed affordable housing production, including density/height bonuses and/or streamlined approvals.
In conclusion, we commend the County for undertaking this initiative, and recognize the effort made by County staff and commission members throughout this lengthy process. Thank you for the opportunity to comment on this critical topic. Please contact us with any questions.
Sincerely,
Susanne Slater
President and CEO
Michael A. Spotts
Director of Real Estate Development
Appendix A:
A 2022 Neighborhood Fundamentals4 review of sales listings for a 12-month period (Zilllow listings, ending 9/29/2022) to identify the number of comparably lower cost single-family detached homes that could be “lost” as a result of missing middle redevelopment. Of the 1,395 2+ bedroom homes (excluding condos) that sold in the last year, only 12 (less than 1%):
• Were single-family detached
• Cost under $600,000 (which is affordable at 129% of AMI)
• Were not likely tear downs, gut rehabs, and/or non-arms length sales
The average cost of these 12 houses was $529,042 (in the ballpark of estimates of the price point for some 1-2 bedroom new construction units under the County’s originally proposed missing middle framework). Of these 12, only 4 had at least 3 bedrooms (0.3% of all 2br + house sales).
Appendix B:
Missing middle price points would expand housing opportunities for key segments of the workforce. In 2022, Neighborhood Fundamentals conducted a review of occupational affordability using median wage levels for more than 100 occupations in the Washington, DC metropolitan region. This study analyzed who could afford various missing middle typologies (based on the initial consultant report) in sample one- and two-income households. The following chart shows a sample of 18 occupations from across the income distribution (0-90th percentile). As this chart demonstrates, typical single-family detached home in Arlington (the status quo) is not affordable to even a household with two-incomes each in the 81-90th percentile range, whereas missing middle typologies extend affordability further down the income spectrum. Importantly, per the PES consultant analysis, allowing 6-unit missing middle housing is critical to expanding affordability further down the income spectrum.
Appendix C:
In addition to the well-established nexus between housing cost and neighborhood diversity, two recent studies demonstrate the connection between a lack of rentals and segregation (https://www.jchs.harvard.edu/blog/rentaldeserts-perpetuate-socioeconomic-and-racial-segregation) and the connection between zoning barriers to multifamily housing and segregation (https://www.mercatus.org/publications/urban-economics/new-research-shows-how-zoning-slows-racial-integration).
County data shows the same pattern - the residence base of multifamily housing in Arlington is the most diverse (the 5- through 8-plex portion of the original County staff proposal), followed by single-family attached housing (2-4 plex), with single-family detached housing the least diverse (click to enlarge image).
This map shows the “rental deserts” as defined by Harvard’s Joint Center for Housing Studies in the above-linked study as neighborhoods where less than 2 in 10 units are rentals (the level associated with greater racial and economic segregation):
