There is a striking mismatch in contemporary society between the types of housing that elderly residents want and what the market delivers to them. Overwhelmingly, older Americans say they want to age in place. And if they must leave their current homes, most would like to stay in a home-like environment proximate enough to their present home to keep their existing social networks intact. In this chapter we examine legal rules that thwart the wishes of so many seniors to age in place and reside in homes that feel more residential than institutional. As recent history has shown, this desire to live in a scaled-down residential environment at the point in their lives when elderly people must lean on third parties to help with personal care is not merely a matter of personal preference. It can be a matter of life and death.

As the world approaches a half-decade since COVID-19 emerged, the horrific damage is plain to see. The elderly have been the hardest hit demographic group by far, with senior citizens representing more than three-quarters of lives lost to the disease in the US.Footnote ¹ But the elderly are far from homogenous, and the variation in contagion risk has been especially pronounced based on where seniors lived. The majority of older adults will need caregiving support during their lifetimes.Footnote ² Although 0.6 percent of the US population resides in assisted living or nursing home facilities, these residents account for up to 42 percent of COVID-19 deaths.Footnote ³ Some research suggests that nursing home residents were nearly twenty-seven times as likely to die of COVID-related causes as senior citizens who did not live in nursing homes.Footnote ⁴ Elderly residents of single-family homes and apartments fared so much better in comparison as the disease spread rapidly through congregate living communities. The average nursing home resident is in poorer health than the average senior citizen, but even accounting for that difference the disparity in mortality remains striking.

Nursing homes differed greatly in terms of their success deploying COVID-19 countermeasures. As we will show, there were nursing homes that managed to protect their residents against the ravages of the pandemic much more effectively than their peers that followed more traditional models of elder care. There is some research suggesting that nursing homes that provide their patients with the highest levels of care did successfully protect their residents from COVID-related deaths in comparison to lower-quality facilities. But even the research suggesting that higher-quality facilities reduced COVID-related deaths indicates that this admirable performance was accompanied by a marked increase in non–COVID-related excess mortality. So it’s possible that the higher-quality nursing homes saved residents from COVID-19 deaths while subjecting them to a higher risk of non-COVID deaths. This higher overall excess mortality may have resulted from the effects of isolation on the elderly.Footnote ⁵ Other research, including meta-analyses, finds no consistent relationship between nursing home quality metrics and COVID-19 mortality.Footnote ⁶

So if it wasn’t the highest-quality nursing homes that most successfully prevented deaths that resulted directly and indirectly from COVID-19, what kinds of nursing homes served their residents best? The answer is the Green House nursing homes, which do not necessarily have a higher quality of care than a typical nursing home but do offer a different benefit – they’re small, with a typical resident population of ten to twelve per facility. These Green Houses, scattered across the country, appear to have largely avoided the worst harms of the pandemic. During the pandemic’s first year, before vaccines became widely available, residents of Green House homes were about five times less likely than traditional nursing home residents to have contracted COVID-19, and about twenty times less likely to have died as a result of the disease.Footnote ⁷ This was true even though Green House residents were older and sicker than residents of other nursing homes, at least in some jurisdictions.Footnote ⁸ In Kansas, PEAK nursing homes that were not affiliated with the Green House movement but that practiced person-centered care and adopted some Green House policies, such as creating clusters of thirty or fewer residents who were consistently helped by the same nursing staff, saw COVID-19 infection rates 2.5 times lower than non-PEAK nursing homes.Footnote ⁹

The reasons why Green Houses and similar scaled-down nursing homes performed relatively well when stress-tested by COVID-19 are straightforward. As we explain below, Green Houses adhere to several organizing principles, including ten to twelve residents per home and private bedrooms. These attributes provided significant resiliency against the disease. Fewer residents and fewer staff rotating through resulted in fewer potential sources of infection entry.Footnote ¹⁰ Not surprisingly, smaller nursing homes saw fewer infections and deaths than their larger counterparts.Footnote ¹¹ Indeed, facility size and virus prevalence in the community were the two most powerful predictors of COVID-19 infection rates. In traditional nursing homes, private bedrooms and bathrooms are not the norm.Footnote ¹² The importance of private bedrooms in limiting the spread of infections was well understood long before the pandemic. According to one estimate, converting all shared bedrooms into private bedrooms would have decreased COVID-19 mortality by 30 percent in nursing homes.Footnote ¹³ Moreover, Green House residents were able to do what residents of other nursing homes could not do – minimize the risk of contracting COVID-19 while still maintaining social ties with their fellow residents.Footnote ¹⁴ Green House homes are more than twice as likely to offer protected outdoor spaces for socialization, such as screened-in porches, compared to traditional nursing homes.Footnote ¹⁵ In short, residents could reduce the risk of the virus entering the facility, lower the risk of the virus spreading if one resident did become infected, and continue to socialize within their protective bubble rather than being isolated.

It’s an open question whether the comparative data on COVID-19 mortality will spark a marked shift in the demand of well-resourced older Americans for Green House living arrangements. Rationally, elderly Americans should recognize that another pandemic is likely to hit the US in the decades to come. It is possible that memories of the pandemic will permanently alter consumer preferences and behaviors in ways that resemble the lifelong impact of the Great Depression on those who lived through it.Footnote ¹⁶ On the other hand, there is a possibility of survival bias. The present population weathered the last storm, and a focus on their own experiences may make them overconfident about their ability to survive the next deadly pandemic, even as changed health circumstances and age make them increasingly vulnerable over time. Assuming that salient experiences with COVID-19 do spark a shift in older Americans’ preferences, the key questions are whether legal and financial obstacles to their creation will frustrate that demand.

Seniors living independently were more likely to survive COVID-19 than the institutionalized elderly. Retirees with the income and ability to remain at home in lockdown conditions, getting groceries and other life necessities delivered, generally were able to avoid infection until the vaccines arrived.Footnote ¹⁷ Given a preference, seniors consistently tell survey researchers that their paramount desire is to age in place until doing so is no longer feasible.

Accessory dwelling units (ADUs or, parochially, “Granny Flats”) are a straightforward and easy way to satisfy that preference for aging in place as seniors’ housing needs change during the course of their life cycles. While some jurisdictions have moved to make ADUs easier to build in recent years, there is still a great deal of resistance to land-use liberalization. As the pandemic has shown, zoning restrictions that constrain the development of these affordable units have likely contributed to the loss of life by forcing seniors into higher-risk living environments.

Our claim in this chapter is straightforward. Older Americans overwhelmingly prefer scaled-down living. That preference exists when they can live independently, and it remains when independent living is no longer realistic. This preference for living small may receive a further boost as seniors and their loved ones come to grips with the data on infectious disease risk. Forms of housing that provided protection for their residents during the pandemic, especially Green House nursing homes and lower-cost, single-family residences such as ADUs, will become more desirable than they already were. Yet existing zoning laws in much of the US are a key impediment to developing the kinds of housing that seniors increasingly want at the scale that an aging society will need. We will discuss these impediments and identify promising reforms to satisfy this growing demand among the elderly for more intimate and protective housing arrangements.

22.1 The Green House Alternative

Geriatrician Bill Thomas pioneered the concept of the Green House long-term care facilities to address the shortcomings of traditional nursing homes. Green House homes are mostly nonprofit facilities.Footnote ¹⁸ Green House homes provide each of the ten to twelve residents with a private bedroom and en suite bathroom. The residents share an open kitchen, dining and living areas, and outdoor spaces.Footnote ¹⁹ Each house employs “Shahbazim,” typically nurse’s assistants who care for residents, do laundry, cook, clean, and order supplies. The available research suggests that Shahbazim are able to devote a higher percentage of their working hours to direct and indirect care of residents than is typical at traditional nursing homes, thanks in part to their ability to engage in tasks like cooking and cleaning while interacting with residents.Footnote ²⁰ The Green House’s founders deemed taboo traditional aspects of nursing homes – such as nursing stations, intercoms for paging healthcare professionals to rooms, and medication carts.Footnote ²¹ Residents of the Green Houses could set their own schedule for going to sleep and waking up, for receiving personal care, for eating, and for engaging in activities.Footnote ²² Where feasible, residents themselves would participate in cooking, laundry, gardening, and caring for household pets.Footnote ²³ Housing concepts that share features similar to Green Houses have emerged in other parts of the developed world, including the Clustered Domestic model in Australia, the Butterfly model in Canada and the UK, and Dementia Villages in the Netherlands.Footnote ²⁴

In 2003, the first Green House homes opened in Tupelo, Mississippi, embracing the Green House movement’s tripartite vision – providing a “real home,” a “meaningful life,” and an “empowered staff.”Footnote ²⁵ To facilitate evaluations of the model, efforts were made to ensure that the residents of the first Green Houses resembled those of comparable nursing homes insofar as possible, and all of the initial residents transferred from traditional nursing homes were managed by the same nonprofit.Footnote ²⁶ There are now some 300 Green House homes in thirty-two different states, mostly organized as nonprofits.Footnote ²⁷ At the same time, because of the small size of each home, the total population of all the Green Houses is approximately 3,200 people.Footnote ²⁸ At least in the original Green House homes in Tupelo, African American seniors were more prevalent than in Tupelo’s traditional nursing homes.Footnote ²⁹ The reverse is true of New York’s Green House homes, in which white residents are overrepresented.Footnote ³⁰

Though still relatively few in number, the Green Houses have been studied extensively by academics. The bottom line that emerges from this literature is that the quality of medical care delivered at the Green Houses is at least as good as what’s available at traditional nursing homes, but the quality of life for residents is much better. For example, a differences-in-differences study using matching to try to minimize the effects of selection bias found that Green House homes had lower hospital readmission rates and saw statistically significant reductions in the percentage of patients who were bedridden, catheterized, or suffering from pressure ulcers.Footnote ³¹ Other indicators of medical quality were not significantly superior among the Green House residents.

It’s with respect to measures of life satisfaction and patient autonomy that the Green House homes really shine.Footnote ³² Researchers have found marked differences in terms of satisfaction and overall emotional well-being, with residents’ and their relatives’ sense of their privacy, autonomy, and the physical environment at Green Houses rating significantly better than the traditional nursing home baseline.Footnote ³³ The literature also suggests that turnover and absenteeism among staff are noticeably reduced compared to traditional nursing facilities.Footnote ³⁴ Workers got to know residents better, and better continuity of care emerged, such that they had an easier time noticing changes in residents’ health status quickly.Footnote ³⁵ The private sleeping-quarters aspect of Green Houses is also a major selling point – surveys indicate that seniors prefer private bedrooms by a margin of twenty to one because these accommodations give seniors more privacy and control.Footnote ³⁶

The best evidence seems to indicate that Green Houses cost roughly the same to operate as traditional nursing homes, or perhaps a bit more. A 2016 study found that the average private-pay monthly cost of a Green House residence was 5 percent higher than the cost for a traditional nursing home.Footnote ³⁷ Another study found that the operating expenses of Green House homes were approximately 7.6 percent higher than the national average, putting the homes in the sixtieth percentile for spending among nationwide nursing homes.Footnote ³⁸ At the same time, there is evidence from an underpowered study indicating that Green Houses are associated with a decline in Medicare expenses of up to 30 percent, apparently because of decreased hospitalizations and transfers to skilled nursing facilities among Green House residents.Footnote ³⁹

Construction costs are likely to be significantly higher for Green House homes, though how much is difficult to estimate, with much of the differential driven by the fact that Green Houses allot more square footage to the average resident than traditional nursing homes with shared bedrooms.Footnote ⁴⁰ The available evidence suggests that occupancy rates in Green House homes are much higher than in traditional nursing homes, which may offset the higher upfront capital costs.Footnote ⁴¹ Taken together, then, the data make it uncertain whether the total monetary costs to society associated with Green Houses will be higher than those associated with traditional nursing homes. It is more plausible that they entail somewhat higher costs, but also conceivable that there is no meaningful difference.Footnote ⁴²

The Green House nursing home differs in one other respect from the traditional nursing home, and the difference looms large from a housing policy perspective. Green Houses are designed to have the profile of a large single-family home, and they can fit nicely into low-density residential neighborhoods. Their relatively small staff footprints also suggest that they could be situated in single-family neighborhoods without creating significant scarcity of street parking spots. As a result, there should be little principled opposition to efforts to zone them in any residential neighborhood where single-family homes predominate. To be sure, zoning opposition to projects in neighborhoods is often unprincipled, and some opposition to the siting of nursing homes and hospices in residential areas may reflect a perverse desire to keep aging and death out of sight, and therefore out of mind.Footnote ⁴³

Because the disabled are a protected class under the Fair Housing Act and many nursing home residents qualify as disabled, Green House developers should be in a strong position to challenge the denial of necessary zoning approvals under the Fair Housing Act.Footnote ⁴⁴ Though judicial interpretations of existing law make it rather clear that a Green House can be built in most areas that are zoned single-family residential only,Footnote ⁴⁵ Congress may wish to consider further amendments to strengthen the Fair Housing Act in a way that makes it crystal-clear that a municipality’s refusal to treat small-scale nursing homes on equal terms with similarly sized single-family homes is unlawful.

22.2 Accessory Dwelling Units

It turns out that the desire for continuity as people age is part of a broader phenomenon. At least in the US, older Americans strongly prefer to live in smaller versions of the single-family homes they lived in when they were younger. Thus, as the percentage of residents who are elderly in a community increases, demand for (and prices of) small, single-family homes tends to increase.Footnote ⁴⁶ Older empty nesters and young families wind up competing for the same “starter” homes.

The preference for American seniors to “age in place” as much as possible is equally powerful.Footnote ⁴⁷ Fully 83 percent of senior citizens agreed in a national survey that they wanted to remain in their current homes for as long as possible, and the overwhelming majority of these seniors agreed strongly with the sentiment.Footnote ⁴⁸ Most seniors find the prospect of moving to nursing homes distasteful.Footnote ⁴⁹ Senior citizens are especially likely to be tied to the community in which they live and the neighbors and relatives who surround them, making moves less desirable and less advantageous for the elderly than they are for younger citizens.Footnote ⁵⁰ For seniors who are attached to their homes, that attachment represents a substantial component of their psychological well-being.Footnote ⁵¹

While the community they’ve long resided in will have significant appeal, the house they’ve long called home might not be suited to the needs of elderly empty nesters. The home may have staircases that are difficult or dangerous to navigate for elderly residents whose mobility is impaired and who are at a greater risk of broken hips and other debilitating injuries. The home may be too large and expensive to maintain. New construction that is built for someone with impaired mobility and is right-sized for an elderly individual or couple will be ideal, but those sorts of dwellings are in rather short supply, especially in neighborhoods occupied largely by single-family homes.

Enter the ADU. ADUs are small houses built within the footprint of an existing parcel that already contains a single-family home. Some ADUs are apartments built over existing garages (i.e., coach houses). Others are new “tiny houses” built from scratch in backyards that resemble ground-floor studio apartments. Either kind of unit might provide great utility to a senior citizen who wishes to remain in place. An apartment built on top of a detached garage might be an ideal space for a caregiver to reside if an elderly resident wishes to remain in an existing single-family residence. And a ground floor unit might permit an elderly resident to stay in a beloved neighborhood while vacating the main house in favor of either a relative who can provide care and companionship, or a tenant who can supplement the senior’s income.

Embracing ADUs is an important pillar of affordable and inclusive housing policy. Surveys of owners who have built ADUs in the Seattle area find that Black and Latino households are more likely than nonminority households to construct ADUs, and that these units are particularly likely to be built in middle-income communities.Footnote ⁵² The same study found that ADUs were especially common in those neighborhoods with a higher concentration of senior citizens. A separate study of ADUs in three high-cost cities on the west coast (Portland, Seattle, and Vancouver, BC) found that the average construction cost of ADUs was $156,000, putting them well within the reach of many seniors looking to downsize or supplement their income.Footnote ⁵³ Fifty-seven percent of households living in ADUs consisted of one person, and another 36 percent comprised two people.Footnote ⁵⁴ Approximately 28 percent of all American senior citizens currently live alone.Footnote ⁵⁵

There is circumstantial evidence consistent with the idea that those seniors who live near their children but not under the same roof benefit from that proximity. As Margaret Brinig has pointed out, senior citizens who live within eight minutes of the home of one of their children are among the seniors with the lowest mortality risks and the most satisfied mental states, but senior citizens who live in the same household as their adult children face high mortality risks and suffer from diminished mental states.Footnote ⁵⁶ Obviously selection effects necessitate caveats – the elderly may need to cohabit with their children when their health declines or when a child’s financial resources disappear, and both dynamics along with the inherent stress of adults merging households and losing privacy will correspond with declines in the quality of life for seniors.

Zoning laws are the main obstacle to the desirable arrangement whereby seniors live near, but not with, their children and grandchildren. The US generally zones for homogeneity, at least within neighborhoods. Not only are many residential neighborhoods uniformly so, but the quality and cost of housing is often uniform as well. For an elderly person with adult children and minor grandchildren living in a neighborhood of single-family homes, residential units that are right-sized for a single elderly person may be hard to find, or even nonexistent. And because of their lack of density, these neighborhoods are often not well served by transit options. So when elderly seniors who can no longer drive safely need to get around, they may be entirely dependent on relatives, or taxis and ride-sharing services if they are available. High-density neighborhoods also saw lower COVID-19 mortality rates during the pandemic, in part because they tend to have a higher-quality and more accessible medical infrastructure.Footnote ⁵⁷

If ADUs are desirable for seniors who wish to age in place but find a lack of other options in their beloved neighborhoods as their needs change, what is stopping them from being built? The answer is generally local land-use laws, though financing restrictions also play a role.Footnote ⁵⁸ When researchers surveyed homeowners who had built ADUs, problems with the permitting process were the commonly identified causes of unanticipated delays and cost increases, and about one in every five survey respondents reported initially having their permit applications rejected.Footnote ⁵⁹ The perceived hassle of obtaining the necessary permits emerged in interviews with landowners who had not constructed ADUs in other research as well.Footnote ⁶⁰

The legal scholarship bears out the difficulties for governments that wish to promote ADUs. In 2002, California enacted statewide legislation to require municipalities to permit the construction of ADUs in neighborhoods that were zoned single-family residential, but many local governments responded with regimes that de facto made ADUs extremely difficult to build.Footnote ⁶¹ The political economy behind these efforts and the resistance is interesting. The politically potent AARP strongly supported the statewide legislation and used its considerable clout in Sacramento to get the legislation enacted, forming a coalition with progressive advocates of affordable housing.Footnote ⁶² But at the local level, real estate developers, construction unions, and homeowners interested in maximizing their property values and maintaining the character of their communities became powerful obstacles blocking reform.

Even when statewide laws make it easier to construct ADUs, cities can thwart those policies by refusing to relax parking space requirements, insisting that ADU builders obtain conditional use permits, or requiring that ADUs be owner-occupied, though California’s legislation thwarts several of these strategies.Footnote ⁶³ Other communities, notably including Los Angeles, enacted ordinances that have facilitated the growth of ADUs and prompted their construction across wider swaths of the city, including in more affluent neighborhoods.Footnote ⁶⁴ Nonetheless, ADUs remain more common in parts of Los Angeles that are close to light-rail stops, schools, and commercial districts, as well as neighborhoods with large Latino populations. They are observed less frequently in parts of Los Angeles where large numbers of high-income residents live as well as in parts of the city where most of the residents are Black.Footnote ⁶⁵ Policy decisions thus play a large role in determining whether ADUs become a viable tool for increasing housing supply and affordability, promoting dense development, and satisfying seniors’ preferences to age in place.

To overcome local intransigence in cities and towns that have tried to keep ADUs out, more aggressive reforms will be necessary at the state level. Christopher Elmendorf has identified approaches to housing affordability in his other work that may be well adapted to the problem of ADUs.Footnote ⁶⁶ Elmendorf envisions a series of steps whereby states create targets for localities to increase the housing supply; real estate developers and third-party interest groups gain new rights to sue localities that block applications for permits that would increase supply; and the state adds in financial penalties for localities that fail to increase the availability of housing over a predetermined number of years. These kinds of reforms could work well to promote the growth of ADUs. They have the advantage of altering the incentive structure for local governments from “how do we say no to these developments?” to “how do we get more approvals for these developments?” To comply with state mandates, local officials might even need to take steps to encourage single-family homeowners to consider construction of ADUs, which could help to overcome a major obstacle to their construction – which is residents’ ignorance about the opportunity to build them.Footnote ⁶⁷ Policymakers could also incentivize homeowners to build ADUs to help care for elders in their communities. For example, with proper supervision and regulation, single-family homeowners might become on-site caregivers for elderly residents of adjacent ADUs. In that way the homeowner could receive a stable source of income as both a landlord and a part-time caregiver, and the elderly resident would have assistance and companionship close at hand in the event of any emergencies.

To be sure, housing policy is only one important part of the necessary societal changes. Prior to the pandemic, advocates for the elderly pushed to increase the funding and availability for home-care services through Medicare, Medicaid, and perhaps nationalized long-term care insurance. Ways to improve further the prospects of seniors who want to age in place include increasing the availability of high-quality in-home caregivers and investing in highly skilled home-based primary care programs. Increasing the appropriate housing supply through ADUs is a necessary but not sufficient measure to facilitate that goal.

22.3 Conclusion

People change as they age, but the stability of their housing preferences is rather remarkable. The elderly generally wish to remain in their long-time homes for as long as they possibly can, and when that option is no longer feasible, many of them desire a nursing care facility that will preserve the normality and privacy that characterized their earlier years. Despite the consumerist culture of the US, the market provides fewer options for late-in-life residential living than for any other life stage, and the options that the market does provide are mostly institutionalized and unappealing to large segments of the elderly population. Consider the heterogeneity of housing options available to a forty-five-year-old American. It is not obvious that the option set available to that person’s parent, who is thirty years older, should be so much more restricted. Yet because the elderly are often paying indirectly for long-term care, through insurance or government entitlements, and because local laws substantially constrain the supply of housing, there is a mismatch between what seniors want (and arguably need) and what they get.

The COVID-19 pandemic may hasten a trend that was already underway, which is that many Americans wish to age in small housing units located within the neighborhoods they’ve grown attached to, or nursing homes that feel at least a little bit like home. To date, the nation has done a poor job of making these kinds of options broadly available to the elderly, and misguided land-use policies seem to be one major reason why. The pandemic underscores that some of the selling points of these housing arrangements are not merely matters of consumer preference but can be issues of life and death as well. COVID-19 may be the worst infectious disease of our lifetimes, but it will not be the last. As Americans begin living into their nineties and hundreds with increasing frequency, much of the existing resistance to the changes we anticipate here eventually will be overcome. Building the right housing now is an essential strategy to minimize the dislocation and anxiety that will occur when tens of millions of aging people discover that so many of their choices are unpalatable.