Is cohousing good for democracy? Comparing political participation among residents of cohousing communities and traditional condominium developments

ABSTRACT

This study assesses the argument that cohousing constitutes a politically fertile form of civil-society association, which hones political skills and cultivates political efficacy among residents, by comparing self-reported political participation levels of residents of cohousing and residents of condominiums. This design allows for more causal leverage than found in past research. However, the findings are still exploratory. Data from a nationwide survey of residents of cohousing and condominium developments in the U.S. were used to test the hypothesis that cohousing residents participate in politics at a higher level than do residents of condominiums. While the sample was small (n = 311), the results suggest both a strong self-selection effect among residents of cohousing and independent effects for involvement in housing-community practices and activism. The results also suggest that residents of condominiums increase their political participation due to cohousing-like practices and activism. Overall, the findings constitute a tentative step toward establishing causation, thus augmenting earlier research on cohousing as a form of civil society that fosters political engagement, and potentially bolster a civil-society case for repurposing condominiums into cohousing communities.

KEYWORDS:

• cohousing
• collaborative housing
• intentional communities
• civil society
• political participation
• political engagement
• cohousing effects

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. CohoUS is a nonprofit organization that facilitates and promotes cohousing.

2. CohoUS maintains a directory at this site of cohousing communities that elect to be included. The directory is organized by state and includes basic information, such as community size and community age/stage of development. Published academic research on cohousing in the U.S. has utilized this directory.

3. This is a 2012 cross-sectional nation-wide survey of individuals living in cohousing communities, led by Angela Sanguinetti with assistance from the Cohousing Research Network (CRN): (http://www.cohousingresearchnetwork.org/). See Berggren (2017), pp. 60, footnote #3, Sanguinetti (2014), p. 88 for details on this survey.

CRN carries out research on cohousing, assists others with research, and serves as a network for cohousing researchers working within and outside of the U.S. CRN is affiliated with CohoUS, but conducts research independently.

4. According to the Foundation for Community Association Research, homeowners associations accounted for 54–60% of these figures, cooperatives for 2–4%, and condominium communities for 38–42%.

The Foundation for Community Association Research represents the interests of the advocacy group, Community Associations Institute (CAI); yet it appears to be as reliable of a source as currently available for the purpose of attempting to estimate the number of condominium units in the US. A 2015 report by the Center for Administrative Records Research and Applications (CARRA), which is part of the US Census Bureau, argued for the need to improve the condominium question in the American Community Survey, since there is “… no dataset of every condominium unit in the United States available” (Flanagan-Doyle, 2015, p. 9). The report states that CAI fact books on community associations are compiled from American Community Survey data and are supplemented by CAI member data in order to produce estimates of condominium units specifically (Flanagan-Doyle, 2015, p. 3).

5. It is important to note that others have found that the rise of housing – including condominium – associations has had negative consequences for individuals as citizens vis a vis their elected representatives and the government more generally. In some areas, housing associations have effectively displaced local municipalities as primary systems of community governance and service provision (Barton & Silverman, 1994; McKenzie, 1994; Nelson, 2011; The Harvard Law Review Association, 1985). Furthermore, housing associations exercise a high level of control over common areas and behavior in them (Nelson, 2011, p. 547), thus creating the potential for violations of individual rights (The Harvard Law Review Association, 1985) and for socially problematic pressures to conform (Nelson, 2011, p. 547). I focus on the effect of positive forms of interactions within housing communities on political participation.

6. As indicated, the large majority of cohousing communities are legally organized as condominium associations. The most recent data on traditional condominiums indicates that there are 138,800 condominium community associations distributed across all 50 states and DC (Foundation for Community Association Research, 2020); whereas the CohoUS directory shows that there are only 165 cohousing communities (old CohoUS site) – which if included in the Foundation for Community Association Research figure would represent a tiny portion – in 39 states, including DC (CohoUS).

7. DataTree is a comprehensive searchable database of U.S. real estate and property ownership information, including hundreds of variables relating to individual properties, homeownership, and mortgages. It is used primarily by banks, credit unions, and mortgage lenders; but is useful for research such as that undertaken here. There are search filters for condominium and owner-occupied designations.

8. The Cohousing Research Network (CRN) and CohoUS maintain and update cohousing community unit addresses.

9. I attempted to find the median-market-valued cohousing unit in each cohousing development. In a limited number of cases, I had to approximate this value based on market values of proximate single-family homes.

10. DataTree can only search and find condominium unit addresses, not the larger condominium associations of which they are a part.

11. Most condominium units were valued within 30,000.00 of the corresponding cohousing community unit medians, but when the searches turned up nothing in this range, I adjusted the parameters.

In addition, since the legal form taken by most cohousing developments is condominium, I examined the spreadsheets for any units from cohousing developments that were picked up in the searches and then deleted these units.

12. In some instances, searches produced spreadsheets with zero or very few traditional-condominium units, or alternatively hundreds of units distributed across different developments. Since cohousing developments vary in size between 10 and 50 units, I set the minimum size for a potentially comparable condominium development at 10 units. When a search turned up only 0–9 sufficiently geographically proximate or similarly valued traditional-condominium units, I excluded the cohousing community around which the search had been conducted and entered the next cohousing address in the list into the search engine.

13. As noted, DataTree can only search and find condominium unit addresses, not the condominium associations or developments where they are located.

In the event of multiple condominium developments found to be potentially comparable to the corresponding cohousing community, I selected those developments that were closest in size to the cohousing communities.

14. I already had contact information for cohousing communities, as provided by CRN and CohoUS.

15. As discussed earlier, one of the selection criteria for housing-community pairs was similarity, not exact sameness in number of units. The total numbers of units in each type of community therefore is not identical.

In addition, the average cohousing community size was approximately 23 units, and the average traditional-condominium development size was approximately 19 units.

16. Sending $2.00 bills was intended less as a material incentive and more as a novelty that might attract sufficient attention to spur survey participation. While it is common practice in survey research to include small cash incentives as part of recruitment, in this case, sending $2.00 bills only to the traditional-condominium residents likely introduced bias into the sample selection process. However, I thought it necessary to attempt to counter-balance possible bias stemming from the larger number of cohousing community contacts, relative to condominium-association contacts, that I was able to reach and persuade to help me with survey recruitment.

Ten residents who declined to take the survey returned the $2.00 bills along with a note stating that they declined to fill out the questionnaire.

17. The survey included adaptions of items from the National Cohousing Survey, Phase III (Sanguinetti, 2012). Survey items were further adapted, with help from the Cohousing Research Network, to be applicable to residents of both types of housing development – cohousing and condominium.

18. While it is well known in polling research that respondents tend to over-report pro-social behaviors, including participation in politics, the measures of political participation used here, as indicated, are based on those of the long-running scientific poll, the National Election Studies series, which regularly tests measures for validity and reliability (The American National Election Studies).

19. Including non-electoral forms of political participation.

20. The original intention was to construct a variable based on the respondent’s participation in these activities. However, there was too much data missing from these variables to do so.

21. Some of the variables used in the analyses below are ordinal and dichotomous and it cannot be assumed that they are normally distributed; therefore, I used a non-parametric version of the partial correlation test. Results cannot be generalized beyond the sample.

22. Due to the data limitations indicated, including a small N (n = 311), a nonrandom sample, and the reliance on self-reports of survey respondents, any results stemming from tests of the hypothesis can only be tentative.

23. Community practices measured whether the range of practices occur in residents’ communities, not residents’ direct involvement in these practices.

24. To maintain anonymity of residents, in accordance with IRB requirements and because in many states there was only one pair of communities, I did not identify cohousing communities and condominium developments by name.

Additional information

Notes on contributors

Heidi M. Berggren

Dr. Heidi M. Berggren is an Associate Professor of Political Science and Chair of the Women’s and Gender Studies Department at the University of Massachusetts-Dartmouth. She serves as the Assistant Director of the Cohousing Research Network. Dr. Berggren’s research, focusing on the relationship between cohousing and political engagement, civic engagement, and democratic citizenship, has been published in numerous journals and supported by the American Political Science Association, the University of Massachusetts President’s Office—Creative Economy Initiatives Fund, and the University of Massachusetts-Dartmouth. She is currently in the process of setting up a panel study of residents of cohousing communities throughout the United States in order to further advance general knowledge about the social effects of cohousing.