Planning for active aging: exploring housing preferences of elderly populations in the United States

Increases in the elderly population (defined as people aged 65 years old and above by the United States Census Bureau) in the United States (U.S.) have reached an inflection point. This demographic shift fundamentally reshapes the magnitude of housing needs in U.S. communities. Planning housing markets for active aging, our study (1) examines how the overall housing market in the U.S. responds to increases in the elderly population, (2) identifies housing preferences of elderly populations by exploring the heterogeneous effects of increasing elderly populations on different types/sizes of houses. Our results suggest that increases in the elderly population are associated with a slight rise in prices in the overall housing market. It points towards the role of increasing elderly populations in shaping the housing market, bringing both opportunities and challenges in reframing urban and housing policy. The study also shows that increases in the elderly population have significant and positive effects on single-family homes and small homes, but no impact on condominiums and large homes. The heterogeneous effects on different types/sizes of housing represent the preferences of elderly populations to single-family homes and small homes. Given the increasing size and share of elderly populations and severe shortages in affordable elderly housing, our study suggests that a strong commitment to reframing urban policy and housing programs targeting assistance and support toward elderly populations such as the Sect. 202 Supportive Housing for the Elderly Program, is in great need of federal and local governments’ interventions.


Introduction
As baby boomers, born between 1946 and 1964, approach retirement age, increases in the elderly population have become striking in the United States (U.S.). The population size of those who are aged 65 years old and above has continued to increase in the last three decades, with more dramatic growth after 2010 (U.S. Census, 2017). Data from the U.S. Census (2017) shows that the population aged 65 and above in the U.S. increased by 19.7% between 2010 and 2016, compared with a 4.8% increase of the total population. This unprecedented growth in the size and share of the elderly population has a significant impact on social systems, including urban and housing policy.
Scholarly literature suggests that increases in the elderly population have significant effects on housing markets. Building off a primary assumption that people tend to buy homes during their working ages and are likely to sell their homes after they ascend into their retirement ages (Nishimura & Takats, 2012;Takats, 2012;Hiller & Lerbs 2016;Saita et al., 2016), previous literature has shown that increases in the population aged 65 years old and above decreases housing prices, though to varying degrees in different countries (Levin et al., 2009;Hiller & Lerbs, 2016). Similarly, relevant studies focusing on the U.S. housing market suggest a negative relationship between increasing elderly populations and housing prices (Mankiw & Weil, 1988;Takats, 2012;Saita et al., 2016).
The existing literature has made notable contributions in examining the impact of increases in the elderly population on the housing market. However, many prior studies focus on the impact of demographic changes on the overall housing market, in which one fundamental assumption is that increasing elderly populations affect housing submarkets to the same extent regardless of the types/sizes of houses. However, we find that this generalization of the impact of demographic shifts on the housing submarkets is questionable. One reason is that the long-running debate over downsizing and aging in place among elderly populations indicates that elderly populations prefer to stay in certain types/sizes of housing, for instance small homes and detached single-family homes (Banks et al., 2011;Bian, 2015). However, it is unclear whether increases in the elderly population contributes to heterogeneous patterns of housing price changes in different submarkets. Indeed, relevant research examining the heterogeneous effects of demographic shifts on different housing submarkets is fairly limited. To our knowledge, the works by Hiller & Lerbs (2016) in Germany and Simo-Kengne (2019) in South Africa are the only studies to do so. Therefore, our study aims to fill the research gaps in existing literature.
In our current study, we examine the impacts of increasing sizes/shares of elderly populations on the housing market in two related research questions. First, whether or not and to what extent the local housing market in the U.S. responds to increases in the elderly population. Second, whether or not elderly populations have preferences for certain types/sizes of houses to live/stay during their retirement age, and if so to what extent, do their preferences impact housing submarkets (i.e., houses with different types/sizes).
To answer these research questions, the remainder of our study is comprised of four sections. The next section summarizes empirical findings in existing scholarly literature. We then describe the data sources and analytical methodology used in this study, before discussing the results of fixed effects regression models. The last section discusses the findings and implications, makes suggestions for future research, and concludes the study.

Elderly population and housing markets
Demographic shifts such as increases in the elderly population create new challenges to urban and housing policy. A large number of studies have shown that this sharp growth in the population aged 65 years old and above has significant impacts on the housing market. In general, increases in the elderly population likely depress the overall housing market. A study by Martin (2005), through an international comparison in the United States, Japan, the United Kingdom, and Ireland, suggests a negative relationship between increases in the elderly population and housing prices. Similar results were found in another international study conducted by Chiuri & Jappelli (2008) in over 15 Organization for Economic Co-operation and Development (OECD) countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Sweden, the UK, and the U.S. Recently, a study using national level data for over 22 OECD countries, suggests that a one-percent increase in the elderly-dependency ratio (i.e., the ratio of elderly populations to working populations) decreases housing prices by a 0.6818% (Takats, 2012). In addition, Hiller & Lerbs (2016) found that in Germany, a 1% increase in the old-age dependency ratio leads to a 0.7856% decrease for real condominium prices, a 0.5155% decrease for real single-family house prices, and a 0.2218% decrease for real apartment rents.
Turning to the United States, Mankiw & Weil (1988), in a groundbreaking study, predicted that the increasing size and share of the elderly population would decrease housing prices by 47% from 1987 to 2007. Later, a national-level study by Takats (2012) suggested that a 0.6818% decrease in housing prices was associated with a 1% increase in the elderly-dependency ratio. In a state-level study, Saita et al., (2016) found that a 1% increase in the elderly-dependency ratio is associated with a 0.9067% decrease throughout the 50 states across the country. As stated above, the relationship between demographic shifts and changes in housing markets is important, although, it is still unclear to what degree/extent, increases in the elderly population impact the housing market. This forms the first research objective of our current study: examining how the housing market in the U.S. responds to increases in the elderly population.
Particularly and unlike previous studies, our study uses the county level as the unit of analysis. We do so for the following reasons: First, urban policies such as zoning regulations, which play a role in shaping the demand and provision of housing, differ widely by local municipalities. Prior worldwide and national-level studies lack consideration of the variation in the need of local communities. A county-level study will provide insights for reframing urban and housing policies in addressing the needs of local communities. Indeed, in order to incorporate community engagement in addressing various challenges of housing markets, an increasing number of studies have brought attention on housing affordability and accessibility to county governments (Ridings & Nakintu, 2018). Second, increases in the size and share of the elderly population vary considerably across the country due to local communities' resources, their capacity to provide public services, and weather conditions. These differences in factors such as community facilities and economic conditions, which likely contribute to elderly populations' housing preferences, cast doubt on whether the results of prior studies conducted at the national and regional levels can be generalized to the local level. A county-level study will help to capture variation in amenities, income levels, etc., among communities. Third, given that many local municipalities are cash-strapped, targeting affordable housing assistance toward the types/sizes of houses that suit elderly populations is an efficient approach and a pressingly needed commitment. To do so, identifying what types/sizes of housing are in great need among elderly populations, at a county-level, is one primary step in reframing local urban and housing policies. Taking all these factors into consideration, we conduct a county-level study to examine the effects of increasing elderly populations on the local housing markets and identify the types/sizes of housing that suit elderly populations' housing needs.

Elderly population's housing preferences
Even though a number of studies suggest that elderly populations show a strong desire to age in place, currently, this is not always realistic due to the lack of appropriate services and facilities (Lipman et al., 2011). As a result, we see a notable size and share of elderly populations expressing their expectation or anticipation of housing relocation. Research by Stoeckel & Porell (2009) suggests that health conditions and retirement status are salient and influential factors that raise housing relocation expectations among elderly populations. Interviews with 6,000 elderly individuals by Hansen & Gottschalk (2006) suggest 21% of elderly respondents relocate within five years. Among these elderly populations who moved houses, around 50% of them who were aged 62-years-old and above claimed that getting a smaller home is one primary consideration of housing relocation (Hansen & Gottschalk, 2006). In addition, research also shows that compared to their counterparts in other countries, elderly populations in the U.S. are more likely to relocate in their retirement age (Banks et al., 2011).
As people turn to their retirement age, increasing financial constraints and decreasing health conditions create new challenges for elderly populations looking for a home that suits their needs and maintains their quality of life. Indeed, in comparison with their younger counterparts, the size and share of elderly households who pay more than 50% of their income on housing are significantly larger, and these numbers increase as elderly populations transition from their 60 to 80 s (Lipman et al., 2011). Moreover, evolving needs for care services and utility costs as people age further heighten elderly populations' financial pressures. One approach for elderly populations to mitigate financial pressure is downsizing, that is, moving from larger homes to smaller homes with fewer rooms, moving from single-family units to multifamily units, or moving to less expensive units. A study by Bian (2015) suggests that the ratio of mortgage to house value, increases the likelihood to downsize among elderly populations. Banks et al., (2011) found that downsizing among elderly populations is associated with seeking available and accessible public facilities or reducing housing costs. Indeed, in a study by Huebner & Shipworth (2017), the authors found that downsizing by one bedroom for a single-person household with two or more bedrooms likely achieves 8% energy savings, indicating a fair amount of economic benefits for elderly populations.
To the extent that downsizing is a distinct and beneficial decision for elderly populations, we suspect that elderly populations' housing relocation expectations and decisions to small homes or less expensive homes likely have an impact on the housing market. Indeed, prior studies have shown that housing preferences among elderly populations contribute to heterogeneous effects of increases in the elderly population on housing submarkets for different types/sizes of housing. In Germany, a 1% increase in the elderly-dependency ratio is associated with a 0.7856% decrease in real condominium prices, a 0.5155% decrease in real single-family house prices, and a 0.2218% decrease in real apartment rents (Hiller & Lerbs, 2016). In addition, Simo-Kengne (2019) found that in South Africa, increases in the elderly population from 1995 to 2005 were associated with 0.0628% point decreases in the prices of large homes and 0.0752% point in the prices of medium homes, but show no impact on the prices of small homes. It suggests that the heterogeneity in the associations between increases in the elderly population and changes in the housing submarkets is distinctive and influential. However, it is unclear for policy makers and housing builders which types/sizes of housing are in most need of interventions of urban and housing policies. Relevant studies in the U.S. examining heterogeneity are rare. Inspired by Hiller and Lerbs (2016) and Simo-Kengne (2019), our study aims to investigate the heterogeneous effects of increases in the elderly population on the prices of different types/sizes of housing in the U.S. This forms the second research objective in our study: identifying elderly populations' housing preferences through exploring the heterogeneous effects of increasing elderly populations on different types/sizes of housing.

Methodology
This section describes the methodology used in our study. The unit of analysis here is at the county-level; the study period is from 2000 to 2010 over a ten-year interval. We start with an introduction to the measurement of each variable and the data sources incorporated in our regression models. We then describe the analytical method: fixed effects regression models. Lastly, we describe the study's limitations.

Dependent variables
Our dependent variable, housing prices, is measured by the Zillow Home Value Index (ZHVI) published by Zillow Research. 1 The ZHVI is "a smoothed, seasonally adjusted measure of the median estimated home value across a given region and housing type. It is a dollar-denominated alternative to repeat-sales indices." 2 Using the ZHVI to indicate changes in housing prices provides important benefits: the ZHVI (1) is derived from proprietary models that predict the estimated median value of homes in each area; (2) is available at the county-level; and (3) is provided in disaggregated form for housing units of different types and sizes. 3 In our analyses, to examine the impacts of increases in the elderly population on the overall housing market, we collected the ZHVI for all homes; to examine the heterogeneous effects on housing prices for units of different types, we collected the ZHVI for condominiums and single-family homes; to examine the heterogeneous effects on housing prices of different sizes, we collected the ZHVI for the single-family units with one, three, and five bedrooms or more. We then took the natural logarithm of the ZHVI to normalize its distribution.

Independent variables
Building upon previous studies on what factors contribute to the dynamics of housing prices (refer to Table 1), we measure changes in the population aged 65 years old and above using the elderly-dependency ratio. We calculated the elderly-dependency ratio as the population aged 65-years-old and above divided by the population aged 25 to 64 years old. We then took the natural logarithm of the ratio to normalize its distribution. This is the primary independent variable in our study.
As noted in prior studies, a variety of other factors can also contribute to housing price changes: one key factor is economic conditions (Takats, 2012;Saita et al., 2016;Hiller & Lerbs, 2016). Many other factors such as changes in the housing stock and housing type, are also mentioned in prior studies (Hiller & Lerbs, 2016;Mankiw & Weil, 1988;Paciorek, 2013;Saita et al., 2016). In the following analyses, we control for changes in population size, housing stock, and economic status. These variables are derived from data published by the American Community Survey (ACS), US Census. 4 The per capita income is used to capture the individual purchasing ability, and the unemployment rate is included to indicate how broader economic conditions impact the local housing market. Considering that the ability to maintain a property is another factor affecting people's housing choices (Luborsky et al., 2011), we calculated monthly owner costs as a percentage of household income to control for the impact of housing cost burdens on housing prices for units of different types/sizes. To ensure normality of the data, we transformed the total population, owner costs, and the per capita income to the natural logarithm type.
Overall, the dataset used here includes up to 3481 observations (the unit of analysis is the county-level) over two time periods (i.e., 2000 and 2010), which provides a sample of 1740 counties. The sample size accounts for more than 55% of counties (1740 counties in each decade divided by 3142 counties and county equivalents) across the country (U.S. Census, 2018).

Fixed effects model
Building on previous research (Takats, 2012;Saita et al., 2016;Nau & Bishai, 2018), we utilized fixed effects models to examine the association between increases in the elderly population and changes in the local housing market. Our fixed effects model controls for all time-invariant characteristics of counties over the study period (i.e., from 2000 to 2010), including geographic location, whether the county is urbanized or not, or existing land use regulations, provided that these factors do not change over time and do not have time-varying effects. To do so, we estimate the following equation: where the dependent variable is the housing price for each county i at time t; the elderlydependency ratio indicates the population aged 65 years old and above for each county i at time t; Time t indicates the effects of time on housing prices from 2000 to 2010. 0 is the intercept; 1 and 2 are the parameters of the main independent variable and the control variables, respectively; u i is a group-specific fixed effect used to control for the time-invariant characteristics in each county i; and e it is the error term for each county i at time t.
Admittedly, in panel analysis, both fixed effects models and random effects models can be used to examine heterogeneity (Simo-Kengne, 2019). In comparison with random effects, the fixed effects models adopted in our study bear its own limitations: (1) fixed effects models might be more subject to sampling variability by using only within-unit variation, compared to random effects models using within-and between-unit variation; (2) the coefficient in the random effects model stands for a weighted average of the generalized least square (GLS) estimates, which is not the case in the fixed effects model (Simo-Kengne, 2019). Taking these all into consideration, we conducted the Hausman test to examine whether the results of the fixed and random effects models differed significantly. In all cases, the tests produced a p-value of less than 0.05, indicating that the fixed effects models produced the preferred models, since they control for confounding due to timeinvariant characteristics of counties.

Limitations
There are also a number of limitations in our study that merit discussion. First, while the use of county fixed effects controls for time-invariant differences between counties, it does not control for temporal variation within counties over time. Thus, if the economic or demographic characteristics of counties change over time, which is likely to be the case, these factors need to be included as control variables in the regression models in order to control for their potential to confound estimates of the effects of increases in the elderly population on housing prices. Second, this study only addresses the heterogeneity of the relationship between increases in the elderly population and housing prices for units of different housing types categorized as single-family homes and condominiums. However, the heterogeneous effects of the increasing size and share of the elderly population may also exist over other housing submarkets, for example, multi-family units and apartments, as the population in different age groups have differing housing preferences. Third, our study faces the challenges of data availability, which limits the study period in our analyses that is currently from 2000 to 2010. The data on housing prices in 1990 was very limited and could not provide enough samples for the analyses, and at the time when this study was conducted, the housing data in the next ten-year interval (i.e., 2020) was not available in Zillow Research. This merits further exploration in future studies.

Results
To investigate the statistical association between increases in the elderly population and the local housing market, we first begin with the regression model that uses the ZHVI for all homes to measure changes in prices for the overall housing market. We then estimate a series of alternative regression models that examine the potential heterogeneous effects of increases in the elderly population on housing submarkets for units of different types and sizes. Descriptive statistics for all variables are shown in Table 2; the regression results for a series of models are shown in Table 3. Table 3 shows the regression results of various models that include the elderly-dependency ratio and a series of physical and socio-economic control variables. We start with an interpretation of the regression model with the ZHVI for all homes, as shown in Model 1. The results suggest that changes in the elderly-dependency ratio have statistically significant association with housing prices measured for the overall housing market (0.2048; p < .001). In terms of the control variables, a series of physical and socio-economic factors show different associations with housing price dynamics as indicated by the ZHVI. Increases in total population are associated with increases in the ZHVI (0.4850; p < .01). The size of the housing stock also shows a significant association with housing prices. Increases in the total number of housing units are associated with decreases in the ZHVI (− 0.4024; p < .05). Consistent with prior findings (Hiller & Lerbs, 2016;Mankiw & Weil, 1988;Takats, 2012;Saita et al., 2016;Simo-Kengne, 2019), economic conditions appear to be closely associated with changes in housing prices. That is, increases in the per capita income are associated with increases in housing prices (0.6497; p < .001), and increases in the unemployment rate are associated with decreases in housing prices (− 0.0068; p < .01). In addition, increases in the percentage of residents who are Black or African American show negative associations with housing prices (− 0.0101; p < .01). Our results also show that increases in monthly owner costs as a percentage of household income is associated with increases in housing prices for condominiums (0.0229, p < .01) and smaller homes (0.0314, p < .001). This points toward housing cost burdens as another incentive of downsizing, which potentially increases the demand, as well as housing prices, for condominiums and smaller homes.

Heterogenous effects on housing prices for units of different types
We now examine whether there is evidence that the association between increases in the elderly and housing prices differs by housing types. Model 2 and Model 3 show the association between increases in the elderly population and housing prices for condominiums and single-family homes.
Model 2 and Model 3 provide two different potential associations between increases in the elderly population and housing price changes for units of different types. In general, the results suggest that increases in the elderly population have a statistically significant relationship with the price of single-family homes but not the price of condominiums. That is, increases in the elderly-dependency ratio are associated with increases in housing prices for single-family homes (0.2068; p < .001). These findings are consistent with our hypothesis stated above that increases in the elderly population may have heterogenous effects on the prices of different housing types.
We also find that increases in the per capita income have heterogenous associations with the prices of different houses. Specifically, a 1% increase in per capita income is associated with a 1.4550% increase in condominium prices (1.4550; p < .001), compared to a 0.6269% increase in single-family house prices (0.6269; p < .001). This parallels the findings by Hiller & Lerbs (2016), suggesting that per capita income has a larger impact on condominium prices than on single-family house prices.

Heterogenous effects on housing prices for units of different sizes
Turning to housing prices for units of different sizes, Model 4, Model 5, and Model 6 show the association between increases in the elderly population and housing prices for singlefamily units with one, three, and five or more bedrooms, respectively. In general, Model 4, Model 5, and Model 6 suggest that increases in the elderly population have a statistically significant association with housing prices for smaller homes but not for larger homes. Specifically, increases in the elderly-dependency ratio are associated with increases in housing prices for one-bedroom units (0.2367; p < .05); however, coefficients of the elderly-dependency ratio for three-bedroom or five-or-more-bedroom units are not statistically significant. These again suggest that increases in the elderly population have heterogenous effects on housing prices of units of different numbers of bedrooms, and the elderly population may have preferences for one-bedroom single-family homes.

Discussion
Our study aimed to achieve two research objectives: first, examining how the local housing market in the U.S. responds to increases in the elderly population; second, identifying elderly populations' housing preferences by exploring the heterogeneous effects of increasing elderly populations on different types/sizes of houses. To do so, we used the elderlydependency ratio to measure increases in the population aged 65 years old and above from 2000 to 2010. The Zillow Home Value Indices (ZHVI) were used to capture fluctuations in housing prices. Our results highlight the heterogeneity in associations between increases in the elderly population and changes in prices of different types/sizes houses. Specifically, increases in the elderly population are associated with increases in the price of single-family homes and small homes, but not the price of condominiums and large homes. This is what might be expected if downsizing among elderly populations contributes to increased demand for smaller, owner-occupied single-family units. Compared to the study by Simo-Kengne (2019), which shows housing prices of small homes are less sensitive to increases in the elderly population, our study suggests an even stronger positive relationship between increases in the elderly population and housing prices of small homes. This may represent a higher possibility of housing relocation among America's elderly populations, compared to their counterparts in South Africa. Downsizing among elderly populations likely drives up the demand for and prices of smaller homes, mitigating the negative effects of increasing elderly populations on the other types of housing submarkets, for instance, larger homes.
Our study contributes to the existing scholarly literature in the following aspects. We first examine the relationship between increases in the elderly population and the overall housing market in U.S. counties. A county-level study allows us to focus more precisely on the variation in housing prices across communities and allows for a relatively large sample of 3481 observations. We also develop additional models to investigate the heterogeneity of the association between increases in the elderly population and changes in housing prices for units with differing types and numbers of bedrooms. To do so, we rely on the Zillow Home Value Indices (ZHVI) to measure housing price dynamics. We disaggregate housing prices/values by the type and size of the dwelling using the ZHVI as the indicators.
Using fixed effects regression models, we examine the relationship between changes in the elderly population and housing prices between 2000 and 2010. Our analyses suggest that increases in the elderly population have significant associations with changes in the price of the typical owner-occupied housing unit. For example, analyses using the ZHVI suggest that increases in the elderly population are associated with increases in the price of single-family homes and smaller homes and show no impact on the value of condominiums and larger homes.
We believe these results have potential implications for policy and planning practices. Understanding the relationship between demographic shifts and the local housing market at the county-level may help housing advocates, planners, and developers to evaluate the demand for and supply of new construction and to plan for impending demographic shifts. In addition, the heterogeneous relationship between increases in the elderly population and housing prices may help housing advocates and planners to identify the housing types that are suitable for and preferred by elderly populations. Specifically, our study suggests that elderly populations in the U.S. may prefer to purchase single-family homes and smaller homes and may wish to sell larger homes in their retirement age. However, the under-provision of appropriate housing options is identified as a primary barrier for elderly populations to do so (Huebner & Shipworth, 2017).

Implications
Our analysis points toward important implications for both policy and subsequent research on the potential relationship between changes in the size of the elderly population and the local housing market as well as the type and size of houses that elderly populations might prefer when they consider either downsizing or aging in place. For example, our analysis suggests that downward pressure on housing prices due to increases in the elderly population may primarily occur for larger homes while smaller homes may experience upward pressure on prices. These findings echo with a recent study by Landis & Reina (2018) indicating that housing downsizing among elderly populations potentially brings upward pressures on starter homes. Collectively, our study along with relevant literature suggest that future efforts on examining the relationship between demographic shifts and changes in housing prices for units of different sizes over longer time periods and/or at other geographical scales, such as the neighborhood level, are worthy of consideration.
Subsequent research could also explore the implications of the relationship between demographic shifts and housing shortages in greater details. For example, increasing demand for small homes and single-family homes likely further heightens the severity of shortage of affordable elderly housing in the U.S. The unprecedented sharp growth in the size and share of elderly populations may necessitate changes in national and local housing policy to support new forms of housing to address changing circumstances. For example, in a study examining how demographic and economic factors will shape to housing policy over the next few decades, Landis & Reina (2018) identify increases in the elderly population as the first consideration and suggest that housing preferences for downsizing will force elderly populations to encounter "heightened competition" from their counterparts, adding rising pressure on affordable elderly housing. To address the issue of housing affordability for elderly population, the Sect. 202 Supportive Housing for the Elderly Program, developed by the U.S. Department of Housing and Urban Development (HUD), aims to help "expand the supply of affordable housing with supportive services for the elderly, and provide very low-income elderly with options that allow them to live independently but in an environment that provides support activities such as cleaning, cooking, transportation, etc." However, in our knowledge, the Sect. 202 program shows preferences to multi-unit developments and rental housing and pose limitations on single-family homes and other types of homes. Given the research results in our study, we recommend that more flexible housing provisions/supportive programs to accommodate elderly residents who prefer small single-family homes, could greatly facilitate the achievement of aging in place and ensure the quality of live for elderly populations.
Moreover, although the housing programs including Low-Income Housing Tax Credit and the Sect. 202 provide affordable housing options for elderly populations in their retirement age, partially, our results suggest that additional national and local housing assistance programs are still in great need, which could be one research focus of future efforts on the housing affordability matter. According to the U.S. Department of Housing and Urban Development (2019), Sect. 202, as the only federal program addressing housing needs for very-low-income populations aged 62 years or older, is able to provide housing assistance for one in three eligible elderly households. This indicates that the current supply of affordable elderly housing falls far short of the need, as elderly populations in the U.S. increase rapidly. In furtherance of planning for active aging, additional housing policies and programs targeting housing assistance/interventions towards elderly populations are needed. Additionally, funding limitations for policy implementations and housing planning, as a consistent challenge faced by policy makers over the country, raise the necessity of prioritizing the provisions of housing that suits elderly populations' needs, furthermore lending credit to our study.
Furthermore, for states and localities, understanding how increases in the elderly population impact housing submarkets would help local officials to facilitate appropriate provisions/allocations of newly built housing units for the populations at different age groups. Particularly, a comprehensive review on zoning regulations, expanding the permits and provisions of detached single-family homes (DSFs) in U.S. communities, likely play a role in mitigating the shortage of affordable elderly housing. Another potential means for expanding the supply of smaller and lower-cost housing units in areas with an increase in the elderly population is accessory dwelling units (ADUs). ADUs are secondary dwelling units located on single-family lots; they are often referred to as "in-law" units, pointing directly to their potential role in providing housing for aging family members (Infranca, 2014). Subsequent research can example the ability of reforms to local land use regulations for addressing changes in the demand for smaller homes as a result of downsizing by the elderly population. Our study suggests that the heterogeneity between demographic shifts and changes in housing prices is worthy of future exploration. Further research on examining the provision and demand of houses with a variety of types/sizes, including DSFs and ADUs, in alignment with the various needs from different age groups, particularly disadvantaged populations for instance, elderly populations, is needed.
Funding There is no external funding support to acknowledge in this study.

Conflict of interest
The authors declare that there is no conflict of interest in this study.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.