‘If I pay rent, I’m gonna smoke’: Insights on the social contract of smokefree housing policy in affordable housing settings

Abstract

Smoke-free housing policy in multi-unit housing has emerged as a promising tobacco control initiative, yet full compliance remains elusive and is a critical impediment to policy effectiveness. There is a gap in existing research on potential factors preventing optimal smoke-free policy adoption and corresponding solutions. Using qualitative and quantitative data from resident surveys (N=115) as well as key informant interviews, a focus group, and observational fieldwork, this study 1) assesses smoking behaviors and experience with secondhand smoke in buildings after adoption of a smoke-free policy; 2) examines resident and property management perceptions of the policy; and 3) identifies socioecological factors that influence the policy’s effectiveness in affordable housing settings in New York City. Findings indicate that residents view smoking regulations within the context of broader relationships with the housing provider and other residents. We argue that these “social contracts” strongly influence the effectiveness of smoke-free housing policies and must be acknowledged to maximize compliance.

1. Background

Smoking tobacco continues to be a leading risk factor for death and disability globally, accounting for more than one in ten deaths in 2015 (Reitsma et al., 2017). Although impressive reductions in smoking prevalence have been achieved in high-income countries, it remains the leading contributor to disability-adjusted life years in many of them, including the United States (Forouzanfar et al., 2016). Furthermore, 80% of current smokers globally live in low and middle-income countries that have not benefited from these reductions (World Health Organization). There is no safe level of secondhand smoke exposure; around 890,000 deaths attributable to such exposure occur annually among non-smokers (World Health Organization). Significant disparities exist in smoking status and tobacco-related health disparities within the United States (Margerison-Zilko and Cubbin, 2013; Bacio et al., 2014; Henley, 2016; Jamal, 2016). For example, smoking rates among adults receiving federal housing assistance are almost twice the rate of the general population (Helms et al., 2017); secondhand smoke exposure is almost twice as high among Blacks as Whites and is closely associated with poverty (Gan et al., 2015; Homa et al., 2015). The continued prevalence of smoking and its ill health effects highlights the urgent need to identify effective and equitable tobacco control practices and disseminate them globally.

Over the past 50 years, growing evidence on the adverse health effects of smoking and secondhand smoke have led to the adoption of policies to reduce environmental tobacco smoke in developed countries. Primarily these have focused on smoking bans in the public domain – in spaces such as workplaces, airplanes, hospitals, and restaurants – which have become extremely common (Hyland et al., 2012; American Nonsmokers’ Rights Foundation). These bans have proven to provide highly effective protection from secondhand smoke (Albers et al., 2004; Haw and Gruer, 2007). In addition to these direct protective benefits, smoking bans contribute to denormalization of smoking, which reduce its broad acceptability. Indeed, bans have been found to reduce smoking overall among smokers (Hopkins et al., 2010) and increase adoption of voluntary smoke-free rules in private homes (Cheng et al., 2011). A large body of evidence demonstrates the related positive impact on health outcomes, with asthma alleviation as a particularly noteworthy example (Hahn, 2010).

In the United States, where bans in public places are already well-established, housing has emerged as the next frontier in the tobacco endgame. Housing is where Americans spend the majority of their time (Klepeis et al., 2001) and therefore the home environment represents a major potential source of secondhand smoke exposure, especially among children (United States Department of Health and Human Services, 2006). Multi-unit housing residents are particularly at risk for secondhand smoke exposure (Chambers et al., 2015), because of higher tobacco use (Nguyen et al., 2016) and air exchange between units (King et al., 2010; Kraev et al., 2009). Studies have found substantial proportions of multi-unit residents reporting secondhand smoke incursions, ranging from 26% to 64% (Snyder et al., 2015). In total an estimated 27.6–28.9 million such residents who have rules against smoking in their homes still experience secondhand smoke exposure yearly (King et al., 2013). Methods that attempt to limit secondhand smoke spread indoors – e.g., separation of smokers and nonsmokers, ventilation, or air cleaning – cannot completely eliminate the risks (Homa et al., 2015; Bohac et al., 2011). Therefore, building-wide smoking bans are the only way to fully prevent secondhand smoke exposure in multi-unit housing. Evidence demonstrates greater protection from tobacco hazards for smokers and non-smokers alike – through decreased indoor secondhand smoke exposure, reduced overall smoking, and increased quit attempts (Kennedy et al., 2015; Kingsbury and Reckinger, 2016; Pizacani et al., 2012; Russo et al., 2015; Young et al., 2016). Smoke-free housing policies also offer potential cost-savings to landlords, residents, and taxpayers, due to avoided physical damage to units and adverse health outcomes caused by smoking (King et al., 2014; Ong et al., 2012). Furthermore, literature has shown a consistently widespread preference for smoke-free buildings among existing and prospective residents (Snyder et al., 2015).

Despite the health benefits, smoke-free policies have raised concerns over undue paternalistic restrictions on individual liberties (Jarvie and Malone, 2008; Shickle, 2009). That is, there is a tension between individuals’ right to act freely (including to smoke) and other individuals’ right to be protected from health risks posed by those actions (i.e. secondhand smoke exposure). This concern is particularly pressing for individuals’ private homes relative to public spaces, explaining why in-home smoking has traditionally been considered outside the scope of appropriate tobacco control regulation (Winickoff et al., 2010). However, the harm principle, which holds that restrictions on liberty are acceptable to prevent injury to others, has been used as ethical justification for smoking bans; moreover, smoke-free policies do not force individuals to quit smoking or deem smokers ineligible for housing, which would be a far more intrusive option (Jarvie and Malone, 2008; Shickle, 2009; Levy et al., 2017). A second ethical consideration is that the denormalization of smoking inherent in smoking bans can contribute to stigmatization of smokers and adversely affect their wellbeing, which is particularly concerning given that smoking is an addiction and disproportionately prevalent among vulnerable populations (Bell et al., 2010). However, since groups with the highest smoking and secondhand smoke exposure prevalence also stand to benefit the most from smoking bans from a health perspective, it can be argued that smoking bans alleviate inequality rather than exacerbate it (Bayer, 2008).

Over the past decade, smoke-free housing policies have increasingly been adopted and garnered support from federal agencies and advocacy groups nationwide (American hung Association; U.S. Environmental Protection Agency; National Center for Chronic Disease Prevention and Health Promotion US Office on Smoking and Health, 2014). However, to date adoption remains limited and has occurred primarily at two extremes – higher-rent units and public housing. The Department of Housing and Urban Development (HUD) in 2016 mandated smoke-free policies in all public housing developments nationwide by July 2018 (U.S. Department of Housing and Urban Development). The majority of low-income households do not live in public housing but instead reside in privately-owned buildings (which we refer to as “affordable housing”). Those who receive housing subsidies mainly access private units, in which HUD’s authority to enact mandates is more limited, through housing choice vouchers, project-based rental assistance, and the Low-Income Housing Tax Credit program (Congressional Budget Office, 2015). Moreover, many low-income households receive no assistance and live in housing with low rent (including rent-stabilized units). Affordable housing has received little attention with respect to smoke-free policies. Moreover, given the uneven distribution of such policies, vulnerable groups disproportionately lack access to smoke-free living environments (Homa et al., 2015; Stein et al., 2015; Stapleton et al., 2011). Low-income populations are also more likely to live in multi-unit housing (Brown et al., 2015; Schoenmarklin, 2010), and have fewer resources to seek alternative smoke-free environments, increasing their likelihood of involuntary secondhand smoke exposure.

To date, research and advocacy efforts have focused primarily on making the case for smoke-free housing policies and calling for the adoption of such policies. Although as noted there is still a need for more even distribution of smoke-free policies, at this point a shift toward compliance and enforcement is also a timely and logical next step to better ensure the intended benefits of smoke-free living environments (Anthony et al., 2018; Hennrikus et al., 2017). Suboptimal compliance threatens policies’ effectiveness, coupled with the inherent increased difficulty of enforcement in private spaces relative to public ones. While adoption studies have found that smoke-free housing policies result in a substantial reduction in smoking behavior and/or secondhand smoke exposure, there is not, to our knowledge, a study finding that a policy has fully eliminated secondhand smoke exposure. For example, one study found that 17% of residents reported “frequent exposure” to secondhand smoke 16 months after policy adoption, which, while a significant decrease from 41% prior to adoption, remains substantial (Pizacani et al., 2012). Thus, barriers to compliance and/or a lack of enforcement appear to exist and limit the effectiveness of adopted policies. Although identifying such factors is therefore essential to ensuring optimal adoption, a systematic review concluded that “issues related to smoke-free multi-unit housing policy compliance and enforcement have not been fully explored.” (Snyder et al., 2015) Recent qualitative research has begun to uncover an intriguing range of social-ecological factors influencing resident compliance and staff enforcement, suggesting that a holistic analysis that takes into account the context of the policy adoption is indeed called for in such research (Anthony et al., 2018; Hennrikus et al., 2017).

The present paper addresses gaps in the literature related to tobacco disparities and smoke-free housing policy adoption by assessing smoking behaviors and secondhand smoke exposure in their context, and thus identifying potential barriers to policy compliance and enforcement, specifically in affordable housing settings and among disadvantaged groups. This study took place in a high-poverty, majority-minority community in New York City with disproportionately high rates of smoking (up to 22%, relative to a citywide average of 15%) and smoking-related health conditions such as asthma (Dragan et al., 2015). The buildings included in the study are affordable housing, subsidized through the Low-Income Housing Tax Credit program – i.e., they offer below-market rental rates to lower-income residents. Thus, our findings can support the design and adoption of effective interventions for populations with high burdens of tobacco-related health disparities, and thereby reduce those disparities. We collected mixed-method data across multiple stakeholders to gain a holistic understanding of quantitative levels of smoking and secondhand smoke exposure and perceptions of smoking and smoke-free housing policies in low-income housing settings. This paper applies the ‘social contract’ framework as a lens for understanding the effectiveness of smoke-free housing policies, particularly compliance and enforcement challenges. We observe that the ‘social contract’ in the political and sociological senses, shape management and resident attitudes and behaviors toward the smoke-free policy.

2. Methods

The purpose of this study is two-fold: first, to characterize smoking behaviors and secondhand smoke exposure in multi-unit affordable housing settings with an existing smoke-free policy; and second, to identify factors influencing compliance and enforcement. In this collaborative, mixed-methods study, we examined three buildings managed by an affordable housing provider in New York City (further details are withheld to protect anonymity). A community-academic partnership was formed to track adherence with the policy, assess resident perspectives and related health outcomes, and provide recommendations for improving the status of the policy based on study results. This study was approved by the Columbia University Institutional Review Board.

2.1. Data collection

Data collection for this pilot study was conducted from May-September 2015 in 3 residential buildings—with 23, 64 and 105 units. These buildings were selected based on their size to encompass small, mid-sized and larger buildings and to assess if buildings that were smoke-free from the start performed differently than those with a retrofitted policy. Of the three buildings included in the study, two were newly constructed and smoke-free from the outset, while the other’s policy was retrofitted post-construction and occupation. Each building’s lease included an addendum regarding the policy that the tenants were required to sign at initial lease signing and/or annual renewal. The addendum broadly explained the policy’s parameters indicating that smoking was not allowed in common areas or within units; however, key details related to enforcement were not included. The smoke-free policy was in place for at least two years prior to the commencement of the study, allowing for assessment of effectiveness of the policy in existing smoke-free buildings.

The following qualitative and quantitative data collection methods were used:

1. Key informant interviews (N = 7) with building management/maintenance and resident leaders were conducted to understand the logistics of the smoke-free policy and perspectives on the effectiveness of the policy to date.

2. One focus group with 9 residents (smokers and non-smokers) was facilitated to assess resident knowledge and their perspectives of smoking in their buildings, the smoke-free policy, and recommendations to achieve greater compliance.

3. Resident surveys (N = 115) were administered across the three buildings. Surveys included questions about smoking behavior, secondhand smoke exposure, knowledge and perceptions of smoking as a building issue, residential satisfaction and also included validated measures such as the perceived stress scale and a brief health assessment (SF-12) (Jenkinson et al., 1997; Perceived stress scale, 1994). Based on this self-reported information, we were able to calculate a smoking prevalence rate of tobacco, marijuana, hookah, and electronic-cigarettes. Surveys were administered by research staff and took 30 min to complete. They were also audio recorded and transcribed for quality assurance purposes and for context regarding closed-ended questions. Participant identities are protected by pseudonyms.

4. Sensory inspection in buildings (N = 9; 3 times per building) were conducted at 3 separate time points in each of the buildings to track “signs of smoking,” such as tobacco and marijuana odor, and the presence of cigarettes, cigars, ashtrays, ashes, matches or lighters within apartments and around the building premises. The observations were conducted by paired members of the research staff during scheduled building visits (i.e. to recruit or administer surveys). “Signs of smoking” were captured via photographs and post-observation field notes indicating where in the building they occurred. Other issues inside homes such as clogged vents and holes in the walls, floors or ceilings, or in/around the building such as loitering, loud music and other concerns were also documented during the home visits and incorporated in post-survey notes.

2.2. Participant recruitment

Building residents were directly recruited via a door-to-door recruitment strategy in which members of the research team knocked on doors and explained the nature of the research study and invited residents to participate in the survey. Recruitment ended when we reached the targeted 60% response rate at each building (N = 14; 38; and 63, respectively). Several participants in one building expressed interest in participating in a focus group. Key informants were recruited based on their role in the building and by referral from the housing provider. Residents were compensated $20 in cash for their participation in each component of the study; building operators were not compensated.

2.3. Data analysis

Descriptive/summary statistics were conducted for the resident survey data and presented as percentages where appropriate. Chi-squared and Fisher’s exact tests were used to test differences in the expected and observed frequencies of responses between groups-perception smoking restrictions, and t-tests were used to test mean differences between groups, using a level of significance of 5% (0.05).

As the surveys were audio-recorded and transcribed, the transcripts were treated as textual data and analyzed, along with the focus group and key informant interview transcripts, the visual data, and field notes. The observational and interview-based data were analyzed using ATLAS.ti (version 6.1). Two members of the research team (DH and CS) engaged a two-phase inductive analytical process informed by phenomenology and grounded theory approaches to qualitative research (Creswell, 2007). These qualitative approaches helped to inform the analytical aims which were: first, to describe the shared phenomenon of residing in a building with a smoke-free policy and second, to provide a conceptual model for understanding the contextual nature of this phenomenon. These authors met regularly in the data analysis phase to jointly develop a codebook and agree on codes and themes. They also used their field experience and team meetings to triangulate results.

The first phase of analysis involved “open coding” in which all transcripts and visual documents were coded based on a descriptive assessment of the information appearing in the data. A codebook was subsequently developed that outlined and defined each code; the data were subsequently reanalyzed to apply codes from the codebook and identify several emergent topics that served to characterize and explain the nature of smoking in the buildings. The second phase entailed “axial coding” in which the codes were thematically grouped and patterns in the data were assessed to reflect causal relationships, contextual factors, intervening conditions, strategies and consequences. This phase of analysis helped to inform the “social contract of smoke-free housing” schema (depicted in Fig. 1) and formed the basis of the study results.

Fig. 1.

Building playground – gap between the mats are littered with trash and cigarette butts.

2.4. Sample characteristics

Two-thirds of survey respondents were female, with an average age of 39.7 years. The vast majority of respondents were either Black/African American or Hispanic/Latino. Over half of respondents had no more than a high school education, and nearly 60% of respondents lived in a household that earned less than $40,000/year; representing less than 200% of the federal poverty line and meeting the criteria for area median income that govern affordable housing development. One quarter of the sample reported fair or poor health and 40% of respondents reported a respiratory-related illness (See Table 1). While most respondents answered most questions in the survey, only 96 of the total 115 answered the question “Should residents be able to smoke in their own apartments?” Their responses are reflected in the tables below.

Table 1.

Demographics of respondents and their households, stratified by opinions on whether residents should be able to smoke in their own apartments.

Demographic characteristics	Total sample (n = 115) N (%)	Should residents be able to smoke in their apartments? (n = 96)		P-value *Sig ≤ 0.05
		Agree (n = 57) (59%) N (%)	Disagree (n = 39) (41%) N (%)
Gender				0.65
Male	38 (33.3)	17 (54.8)	14 (45.2)
Female	76 (66.7)	40 (61.5)	25 (38.5)
Age (mean/median)	39.7/39	37.3/36.0	43.7/39.0	0.03*
Older residents				< 0.01*
Respondents 60 or more years old	11 (9.7)	1 (10.0)	9 (90.0)
Respondents less than 60 years old	102 (90.3)	55 (65.5)	29 (34.5)
Children living in household				0.84
Households with children	53 (46.1)	26 (57.8)	19 (42.2)
Households without children	62 (53.9)	31 (60.8)	20 (39.2)
Race/Ethnicity				0.38
Non-Hispanic Black/African American	55 (49.1)	30 (63.8)	17 (36.2)
Hispanic/Latino	46 (41.1)	21 (58.3)	15 (41.7)
Other	11 (9.8)	4 (40.0)	6 (60.0)
Education				0.52
High school or less	55 (50.9)	24 (53.3)	21 (46.7)
College or more	53 (49.1)	28 (62.2)	17 (37.8)
Annual Household Income				0.04*
Less than 40,000	63 (63.0)	34 (69.4)	15 (30.6)
More than 40,000	37 (37.0)	16 (47.1)	18 (52.9)
Self-reported health status				0.58
Excellent/very good/good	84 (75.0)	44 (61.1)	28 (38.9)
Fair/poor	28 (25.0)	12 (54.6)	10 (45.5)
Respiratory-related illnessa				0.84
Yes	60 (52.2)	31 (58.5)	22 (41.5)
No	55 (47.8)	26 (60.5)	17 (39.5)
Impairment for moderate activitiesb				0.29
Limited a lot/a moderate amount	18 (84.2)	7 (46.7)	8 (53.3)
Limited a little/not limited at all	96 (15.8)	49 (61.3)	31 (38.8)

^aSelf-reported asthma, other chronic respiratory, lung, breathing trouble, sinus or allergies.

^bResponse to the question: “Does your health now limit you in…moderate activities, such as moving a table, pushing a vacuum cleaner, bowling, or playing golf?”.

3. Results

3.1. Should residents be able to smoke in their own apartments?

Residents generally supported a smoke-free policy in their building, with 96% of residents agreeing or strongly agreeing with the statement, “people should not smoke in indoor places if non-smokers are present” (data not shown). However, perspectives shifted when the question was framed as a query about placing restrictions on residents’ ability to smoke in-unit. When asked to rate their agreement with the statement, “people should be allowed to smoke in their own apartments,” the majority of residents agreed or strongly agreed (see Table 1).

Belief that residents should be allowed to smoke in their apartments varied by demographic characteristics and exposures with less support among older respondents and those earning $40,000 a year or more. Gender, children in the home, race/ethnicity, education, and maintenance-related issues however, were not significant predictors (see Table 1). While presence of adverse physical health conditions was also not significant, the qualitative data did find support for the policy based on its impact on health, particularly for children (see Table 3). Weekly or monthly exposure to secondhand cigarette smoke was negatively associated with this belief. The results were not significant for exposure to secondhand marijuana smoke. Unsurprisingly, living in households with indoor cigarette and marijuana smoking was a significant positive predictor of supporting smoking in one’s own apartment (see Table 2).

Table 3.

Resident perspectives on smoking in the building and the smoke-free policy.

Smoking and the ‘commons’	“They say you can’t tell a person what to do in their home, but when it leaks in the hallway, it becomes everybody’s problem,” (Kathryn) “I think that’s rude. You have to be considerate of the non-smokers. If it’s indoors, then you shouldn’t smoke… I think it’s just about being polite.” (Camila) “That smell [of smoke] comes through my windows, you know, it just travels. They’re smoking weed, they’re smoking cigarettes and it smells like I’m doing it in my home.” (Esther) “Smoke comes through my vents really bad. You smell it in my house as if I were a smoker and I’m a nonsmoker.” (Gabrielle)
Impact on health and safety	“The benefits [of the policy] is a cleaner and healthier building for us all to live in happily.” (Sarah) “I think it [the smoke-free policy] is good…for people who have asthma, and like kids who can’t prevent it, kids don’t want to smoke and you’re forcing kids to inhale that So that’s not good… I guess I’m most worried about the children.” (Stephanie)
Impact on the building environment	“They’ll put their cigarette out there [in the hallway]. I’m like, really? First you smoke and then you litter.” (Sarah) “I don’t think people really realize how it affects them and how it affects the other people in the building… it’s just so unbearable. If someone’s smoking in their bathroom, and it just travels −1 mean, the first thing you get up in the morning and you smell like cigarette smoke. It’s disgusting. We don’t know exactly who it is, but it’s traveling all through the vents. The lady on the second floor smells it as well So there’s definitely a problem there.” (Sarah) “That’s what pisses me off, they smoke in the hallway. I don’t care who I’m telling on. Let me say it again, they smoke in the hallway, and it’s disgusting.” (Lizette)
Concealment, Containment and Consideration	“If you have a person in their apartment smoking, right, have a little fan that the smoke is going out the window, and they [dragging sound] how can you be exposed to it, sweetie?… Look at the bottom of my door. I mean [skeptically], smoke is going out through there?” (Anthony) “If you’re going to smoke in your apartment, do everything possible to stop the smoke from leaving your apartment If it’s your apartment, keep it in your apartment Everyone else doesn’t need to suffer because you want to smoke… If people was to like, figure out a way to just keep the smell in their apartment, we wouldn’t have an issue with the building smelling like weed,” (Raphael)
Respect and compassion for smokers	“People been smoking cigarettes for years. They can’t just like, turn off like, one day just because of, you know, the situation.” (Christian) “If you’re a serious smoker I also understand their needs. If you enjoy drinks and then somebody tells you its alcohol free, so for the same argument [regarding the smoke-free policy].” (Tina)

Table 2.

Smoking and indoor environmental exposures of respondents, stratified by opinions on whether residents should be able to smoke in their own apartments.

Exposure	Total sample (n = 115) N (%)	Should residents be able to smoke in their apartments? (n = 96)		P-value *Sig ≤ 0.05
		Agree (n = 57) (59%) N (%)	Disagree (n = 39) (41%) N (%)
Time spent at home during the week				0.93
All/Most of the day inside	40 (34.8)	18 (60.0)	12 (40.0)
Some/A little/None of the day inside	75 (65.2)	39 (59.1)	27 (40.9)
Time spent at home during the weekend				0.64
All/Most of the day inside	59 (51.8)	30 (61.2)	19 (38.8)
Some/A little/None of the day inside	55 (48.3)	26 (56.5)	20 (43.5)
Second-hand smoke exposure to tobacco				0.05*
Daily	18 (17.0)	8 (50.0)	8 (50.0)
Weekly	11 (10.4)	3 (33.3)	6 (66.7)
Monthly/A few times	21 (19.8)	5 (38.5)	8 (61.5)
Never	56 (52.8)	35 (70.0)	15 (30.0)
Second-hand smoke exposure to marijuana				0.33
Daily	21 (19.1)	9 (52.9)	8 (47.1)
Weekly	17 (15.5)	5 (38.5)	8 (61.5)
Monthly/A few times	21 (19.1)	9 (60.0)	6 (40.0)
Never	51 (46.4)	31 (66.0)	16 (34.0)
Household indoor cigarette smoking				< 0.01*
Yes	24 (20.9)	21 (91.3)	2 (8.7)
No	91 (79.1)	36 (49.3)	37 (50.7)
Household indoor marijuana smoking				< 0.01*
Yes	11 (9.6)	9 (100.0)	0 (0.0)
No	104 (90.4)	48 (55.2)	39 (44.8)
Building maintenance satisfaction				0.82
Completely satisfied with maintenance	42 (38.2)	23 (62.2)	14 (37.8)
Partially satisfied with maintenance	49 (44.6)	24 (55.8)	19 (44.2)
Unsatisfied with maintenance	19 (17.3)	10 (62.5)	6 (37.5)
Cockroaches/insects/mice/mildew/mold				0.34
Yes	57 (50.0)	26 (54.2)	22 (45.8)
No	57 (50.0)	30 (63.8)	17 (36.2)

3.2. Do residents support the smoke-free policy?

Qualitative findings indicated tensions between residents’ perceived entitlement to full freedom versus the right to the nuisance-free enjoyment of the home environment. Participants described secondhand smoke exposure as a deeply troubling nuisance, health hazard, and encumbrance to their quality of life. Health and cleanliness were most often cited as reasons to reduce smoking in the buildings, a perspective shared by the residents and property management alike. Nevertheless, some residents (mostly but not always smokers) felt that restrictions on smoking represented an unacceptable infringement on individual rights; some were adamant that paying rent allotted tenants the right to fully enjoy their apartments including exercising their prerogative to smoke indoors. As such, many residents, including non-smokers, resisted the policy.

Attempting to cope with and mitigate secondhand smoke was difficult for residents; some even contemplated moving to avoid it (see Table 3). Consistent with other studies in low-income housing (Anthony et al., 2018; Hennrikus et al., 2017), many believed, incorrectly, that it was possible to conceal and contain the smoke, thereby reducing exposure to other units. Residents were, at times, sympathetic toward smokers, noting that smoking is an addiction that is difficult to quit and expressing the need to accommodate smokers. However, many viewed smoking indoors as a problem of “the commons” (common residential living space) that violated norms of consideration for others, particularly non-smokers and ensuring an adequate living environment for children and homebound neighbors (see Table 3). Importantly, while it is easy to respect the right to smoke in one’s apartment in the abstract, smoking inside a shared building is rarely a self-regarding behavior in practice. Smoke leaks into hallways, travels under doors and through windows, and is spread through vents (see Table 3)- at this point, the rights of non-smoking residents are adversely affected. This social and spatial nexus adds to the complexity of realizing smoke-free living environments in multiple unit housing.

3.3. Is the smoke-free policy effective?

3.3.1. Compliance

Surveys indicated limited compliance; residents reported smoking various substances inside their apartments including cigarettes (21%), marijuana (10%), hookah (5%) and electronic cigarettes (2%) (See Table 2). Secondhand smoke exposure remained high despite the policy. Almost half of residents reported secondhand tobacco smoke entering their home in the last year, with 17% stating that it occurred daily, 10% weekly, and 20% monthly. The figures were slightly higher for marijuana: residents reported higher daily (19%) and weekly (16%) incursions and comparable monthly incursions (19%), and fewer reported never being exposed (46%) (See Table 2). Visual inspection in the buildings indicated that smoking occurred regularly in common areas including stairwells, hallways, laundry rooms, and outside of a locked patio area (see Figs. 1–5). The evidence of smoking in common areas and self-reports of smoking in units demonstrated the lapses in compliance and enforcement of the smoke-free policy. Furthermore, smokers in most buildings were not accommodated by designated smoking areas or reinforcement strategies such as delineating where smoking was (dis)allowed in and around the buildings. There were no major differences noted by building size, but in buildings that were smoke-free since inception, residents had greater awareness of the policy and expressed more frustration with the violations.

Fig. 5.

The social contract of smoke-free housing and levers to promote policy effectiveness.

3.4. Figs. 1–4: evidence of smoking occurring in “smoke-free” buildings

Fig. 4.

Padlocked terrace gate, closed due to vandalism and marijuana smoking. Tenants identified it as a potential designated smoking area.

3.4.1. Enforcement

Lack of enforcement of the policy was noted by residents and building operators alike. Although key informant interviews with property management indicated that operating smoke-free buildings was a source of pride, they admittedly exercised minimal effort in enforcing the policy. The buildings featured little signage, save for decals posted on the main entry, and many residents had limited knowledge about the policy. Residents were frustrated by the lack of enforcement, and believed the policy could not be effective without it. They also noted another example of poor enforcement – on the rule against pets – which eroded residents’ attitudes about the seriousness of building policies and the standards of living articulated in the lease. While the building operators knew of residents who commonly smoked in the building, they hesitated to enforce the policy inside apartments, noting the burden of proof and also evincing a similar notion as tenants about protecting common areas but having less restrictions within units.

Building operators and residents shared a range of ideas about how to make the policy more effective, pointing to significant missed opportunities to strengthen the policy (see Table 4). Some participants recommended having designated smoking areas, placing more signage, providing lease counseling to familiarize tenants with the policy, and instituting reminders of the smoke-free policy on the monthly rent slips. When a violation of the policy is found, participants proposed escalating penalties including warning letters, monetary fines, and for repeat offenders, non-renewal of leases and/or eviction. Residents and property management expressed mixed feelings regarding eviction, with some viewing it as unjust, and others feeling that eviction was an appropriate response to ongoing flouting of rules all the while noting the practical difficulty of eviction given the aforementioned poor enforcement. Many residents suggested that cameras and/or hired staff might resolve security issues and assist with enforcement efforts; while property management agreed, they cited the lack of financial resources to support these approaches.

Table 4.

Policy status and strategies to strengthen its effectiveness.

Policy status according to residents	Smoking in the Buildings “Why should you have to live like that, put plastic around your windows or doors? I mean, you know, that’s not what you’re paying for, for somebody else, strangers or people that live in the building, to smoke in the hallways.” (Maria)
	Lack of Information “I didn’t even read my lease. All I knew was that I was getting an apartment, and I initialed and signed everywhere they told me to initial and sign. I found out it was a smoke-free building when I moved in. ” (Melissa) “You have to constantly educate them. Because people are just signing their lease.” (Doug)
	Lack of Enforcement “[The management company] also has to stick to it, and they have to abide by it, and they have to enforce it, and they’re not They’re not” (Emily) “You could put a million signs… but if no one’s here to enforce it, it’s not gonna really happen. It’s not” (Maria) “You have to constantly be on top of it Like, if you’re not showing them that you want to fix it, it’s just like, oh somebody complained about it so that’s why you sent the letter. They’re not really interacting with it” (Cynthia) “You can’t prove it because you’re not allowed to go into my apartment without my permission. So then that makes [enforcement] difficult” (Rachel)
Policy status according to building staff and property management	“It’s easily detected in the hallways, because you see people, you know, they don’t do it discreetly. They turn a cigarette off on the wall, they leave the butts on the floor. So besides just the health part, it also hurts the building…the cleanliness, overall look of the building.” (Property Management) “It’s kind of hard because like you can implement the policy but you really can’t detect or tell people what to do inside their apartments. And it’s kind of hard to say T know what you were doing in your apartment’ unless you have evidence.” (Property Management) “I always tell them, whatever you do in your apartment, is what you do in your apartment but far as stairways and lobby, I’m like, ‘please don’t smoke.’” (Super, Building #3)
Strategies to strengthen the policy	Cameras and Security “That’s why [property management] needs to do it That’s why they need to provide security and that will cut down on the destruction on the building as well… Then, you know, someone would be patrolling and a lot of that ruckus wouldn’t be happening. I really think that’s something that they need to enforce. That’s a way that you can also enforce the hallways that are no-smoking.” (Gabrielle) “If you had cameras in the hallways, in the staircases, then you’d be able to actually pinpoint that person and get rid of the problem.” (Property Management)
	Escalating Warnings and Eviction “I think they should be given a warning, like a ticket Because everybody that’s here is responsible for their guests, their visitors, and all that stuff.” (Carla) “All of’em have to do something, give a warning, a written warning. Cause it’s not gonna just happen, that you can throw them out That’s not happening.” (Eva) “I think either evicting or denying the renewal lease. I guess people know there are consequences to, you know, breaking rules.” (Tina) “You don’t want to evict a family and put them on the street that’s been smoking. But in a way, it’s like, how many times are you gonna disrespect management, and how many times are you going to keep doing something you know, you physically sat here, read and signed, that you’re not supposed to do?” (Property Management)
	Designated Smoking Areas They got to have a place for [smokers], too. You can’t put up a building and say, OK, unless you know, you only interview people that don’t smoke.” (Erica) “To make [the policy] successful, in my opinion, I think you will have to designate a specific area outside of the building to say, ‘Hey listen, this is where you’re gonna smoke, there’s like a little tin box for your cigarettes, ashes, because then they’ll say ‘you are thinking of us, because we’re smokers, you’re considering our feelings and you designated an area.’ This way they don’t have to jeopardize the building, the overall look, or everyone else’s health.” (Property Management)

3.5. What additional factors undermine the smoke-free policy?

Beyond the lack of enforcement, dissatisfaction around maintenance issues and tenant relations with management and one another also threatened the viability of the smoke-free policy. Some residents were steadfast in their belief that paying rent was accompanied by certain rights and privileges, including the right to smoke in their apartment. Issues with building design, maintenance, and operators’ non-responsiveness to tenant requests therefore posed challenges for the policy. Residents cited safety concerns; building upkeep problems; housing quality issues such as inadequate heating, mold, and pests; and denied access to building amenities as consistent and ongoing issues. They often explicitly linked these infringements to the rental contract and pointed to these violations as undermining management’s authority to impose the smoking policy as a condition of the lease. Furthermore, residents mentioned how maintenance issues and building design magnified the impact of secondhand smoke incursions, e.g., due to poor ventilation, an issue also supported by the literature (Bohac et al., 2011). However, as noted in Table 2., maintenance dissatisfaction was not associated with support for the smoking ban.

Conversely, property management held negative opinions of residents’ ability to value their living environment, contributing to friction between the parties. The lack of faith in tenants’ ability to care for the building, disregard for their concerns about maintenance, and dismissal of the tenants as “ignorant” summarize the belief that residents were fundamentally unable to comply with smoke-free policy. Property management was not alone in their frustration with tenants; residents also expressed disgust with how other tenants behaved toward the building, citing irresponsibility in preserving high standards in the building, including compliance with the smoke-free policy (Table 5).

Table 5.

Threats to smoke-free housing.

Entitled to the full enjoyment of apartment including the right to smoke	“I feel like you should be able to do whatever you want in your own apartment I don’t think somebody else should come and tell you what to do… It’s your apartment, you know, you rent it, you’re working hard to pay your rent” (Luciana) “I think that’s stupid. I can’t smoke in my house? I’m entitled to… People can’t smoke after Bloomberg. Like, ‘You can’t smoke in this restaurant, you can’t smoke outside, you can’t smoke’ – So where the f–k am I smoking at? I can’t smoke in my house? Next you’re telling me I can’t have sex in my house. I’m sorry to say it like that, but it’s the truth. So where am I supposed to smoke? If you pay your rent here, you should be able to smoke here.” (Alexandra) “Whoa, whoa, whoa, people got rights. People got – especially when they paying rent and you ain’t doing jack. So now you gonna, you gonna take away now a little thing [smoking in their apartment] that eases them up?” (Anthony)
Building Design, Maintenance, and Management	Buildine Design “In general the air circulation, our ventilation is really poor, so if somebody’s, for example, smoking something you’ll smell it You’ll wake up with the smell saturated in the apartment” (Tina) “I can tell you what my major problem here is: the windows. Ill give anything to have a sliding window or a window that I can freaking put my head out, and I could put a box fan. We don’t even get air, let alone cool air…” (Carla)
	Poor Maintenance “I been having a mold issue in this apartment for like two years. I can show you the pictures… I’ve been living here and I’ve been having this same [mold] issue.” (Jordan) “Even when they turn on the heat, I know last winter it would just really get too cold. It was a really bad situation. It’s like, the heater isn’t working. It’d be freezing literally freezing. I have [the baby] in like, full blown onesie pants… Ill go to my mom’s house and stay there because I couldn’t stay here. It’s crazy. I’m hoping that’s not going to be the same story, it’s horrible. Because then if not, my only other option is getting a heater and I’m not working.” (Camila)
	Lack of Responsiveness “I have issues with my cupboards, my refrigerator. And I have called and left messages. And no one happened to return. Not one phone call” (Rosa) “We have really no one to voice our concerns to ‘cause you call the office and you get the recording!” (Wendy)
Accountability: Resident Perspective	Resident to Resident “It’s some people that just is disgusting. You know, it’s just, for real they think this is the projects. And there are times you come home and you be like, damn, you know.” (Henry) “I feel bad for the people who clean, I really do, because to walk in the garbage room, you have to think we’re like animals. The way people put their garbage in there. It’s disgusting.” (Gabrielle) “Some people don’t realize, this is where you live at If you destroy it, where do you go? Maybe I’m old-fashioned, but I believe in where you lay your head, you don’t shit That’s my motto… It’s up to us. It’s not even up to [the property managers], it’s up to us, we have to make this happen. With the smoking the weed, you know, the cleanliness of the building.” (Wendy)
	Resident to Property Management “We are not nasty people… The upkeep of the building whoever’s in charge, you have to take responsibility.” (Erica) “I understand people that wanna smoke and stuff like that, fine. But again, like, [the super] used to smoke weed, I’m gonna just put it out there. He [even] had weed patties on the roof.” (Erica)
Accountability: Property Management Perspective	“The tenants have to respect where they live, and they have to maintain where they live. You have to have some type of pride in where you stay. And that comes with the attitude of the tenants. You can mop that stairwell 50 times a day. But if you constantly gonna have someone come out and throw something on the floor, urinate on the floor, smoke and throw a cigarette down, you’re defeating the cause. Because you’re trying to maintain a building that the tenants are constantly beating up. So it’s like you’re fighting each other.” (Property Management) “You know, tenants exaggerate. And I’ve learned that in this field. If you listen to them, they will make up a whole story, like Pinocchio.” (Property Management) “It’s an ignorant population we serve. [These] tenants aren’t thinking about their health or the health of others and will just go back to doing what they do [meaning smoking indoors] because they’re ignorant” (Property Management)

4. Discussion

With the benefits of smoke-free housing policies increasingly recognized and adoption on the rise, our findings make clear the importance of understanding factors that hinder or facilitate effective adoption in affordable housing – an underexplored area of research. Consistent with other studies (Kennedy et al., 2015; Pizacani et al., 2012; Anthony et al., 2018), we found that although the policy had technically been adopted, compliance was far from complete. By employing a mixed-methods approach, we found that social-ecological factors, in particular, residents’ relationships with each other and with property management, inform residents’ perceptions and likelihood of compliance with the smoke-free policy.

A strong tension regarding autonomy emerged, as found in previous studies on smoke-free housing policy in low-income settings (Anthony et al., 2018; Hennrikus et al., 2017). Residents frequently articulated a belief that they had a right to privacy and freedom in their own home for which they pay rent. They viewed the smoke-free policy as a paternalistic and inappropriate restriction on their liberty, undermining buy-in and compliance. On the other hand, residents also frequently expressed their right to be free from involuntary, unhealthful exposures, noting significant impacts to their health and well-being from secondhand smoke exposure. The evident harm caused to the building’s non-smokers, including children, provides what we feel to be sufficient justification for restricting smoking within the building. That is, people’s right to live in an unpolluted environment supersedes other people’s right to smoke indoors. These norms of reciprocity and mutual responsibility form the basis of what we have conceptualized as the ‘social contract of smoke-free housing.’

4.1. The social contract of smoke-free housing

In rental housing, a legal contractual agreement (lease) stipulates the terms of the provision of an apartment unit in exchange for payment (rent). A lease covers conditions and conduct expected of tenants, while the municipal housing code regulates the provision of habitable housing. However, as demonstrated in our findings, the social dynamics between the housing operators and tenants, as well as amongst residents, ultimately determine the housing experience. In a political sense, a social contract is a social arrangement in which rights, responsibilities, and restrictions that introduce limitations on individual liberties for the common good are considered fair because they are agreed upon by both parties (Hobbes, 1651; Locke, 1689; Moses, 2009). In the sociological sense, it relates to social behavior – “systematic patterns of shared, normative understandings that shape individuals’ behavior across institutions” and “overall norms of reciprocity.” (Rubin, 2012) Drawing on these understandings of social contracts, the context of smoke-free policy and the relationships between property management and residents and amongst residents themselves are subject to a social contract. Consistent with this sociological view, tenant relationships with management and maintenance personnel have been found to be the most influential factor in tenant satisfaction (James et al., 2009).

While our survey results indicate that satisfaction with management was not quantitatively associated with agreement with the policy, the qualitative findings reveal that management and resident attitudes and behaviors toward the smoke-free policy were embedded in this “social contract.” Residents asserted their entitlement to certain rights under their rental agreement, and believed property management consistently violated the social contract on both political and sociological levels. Specifically, residents frequently juxtaposed their good standing as tenants who kept their units in decent repair with management’s lack of responsiveness to maintenance and repair issues. This breach in conduct in maintaining a high-quality building standard effectively violated that actual and implicit contract of providing housing, which is bound by legal contract and enforceable by the local housing code. This ‘quid pro quo’ mentality gave license, for some, to disregard the policy, particularly when combined with the view that the smoking ban represented an unjust restriction on their liberties. Further, the lack of enforcement of non-smoking related policies (e.g., pet policy) also undermined management’s authority more broadly. In particular, the inattention to enforcement diluted the top-down approach necessary for the efficacy of the smoking ban.

In effect, the smoking policy introduced new responsibilities for residents to mind building quality standards and protect the health of fellow tenants. Yet, this “lever” was compromised by lapses in the responsibility chain going down from the housing provider to the residents, weakening the norms of reciprocity between the residents and the housing provider (i.e. well-maintained housing in exchange for rent). With management’s failure to uphold the rental agreement in terms of maintenance, enforce all policies stated in the lease and provide adequate customer service, residents viewed the prohibition of smoking in their apartments as unfair, inappropriate, and overreaching, and thereby against the housing-for-rent exchange. Furthermore, resident requests to hire security guards or install cameras went unheeded despite their belief that greater surveillance would deter smoking and enable identification of and intervention with noncompliant residents.

In keeping with the notion that residents had a responsibility to care for each other in their shared living environment, building tenants espoused the decrees of “social connections” and “collective wellbeing,” which respectively pertain to co-existence and the surrendering of personal liberties for the common good. These concepts were conveyed by residents that referenced a collective responsibility to maintain the building in acceptable condition and refrain from actions that are detrimental to fellow residents, such as not cleaning up after parties in common spaces or not disposing of garbage properly. Smoking indoors is another of these actions, both because indoor smokers frequently left behind cigarette butts, and secondhand smoke reaches other residential units. Despite sympathies toward the addiction of smoking, exposure to secondhand smoke was considered harmful to children, the elderly, ill and homebound. Thus, the social contract between residents was central to the perceived acceptability of indoor smoking behavior. In fact, these communal relationships are arguably unappreciated, though they may be a more effective strategy to motivate compliance compared to the hierarchical, and inherently punitive, relationship between management and tenants. Future efforts to ensure smoke-free housing policy compliance should emphasize the protections and obligations toward fellow residents, in effect serving as the “neighbor’s keeper.”

4.2. Adoption recommendations

On a basic level, the smoke-free housing policy must be known and its justification understood, and all stakeholders must have the tools to support it. Our results indicate key lapses in information about the policy and a grave assumption that residents carefully read the lease when in fact, many stated that they lacked awareness of the policy. As such, the parameters of the policy need to be clearly delineated and residents need to be properly informed of areas covered by the policy. This information should be made available at lease signing as well as annual renewal, and reminders can be included on monthly rent statements.

There should be opportunities for easy anonymous reporting, such as a hotline or mobile application for residents and staff to identify areas of the building or individual apartments where violations occur. Effective education on the health risks of smoking and secondhand smoke exposure should also be presented. In particular, children are especially vulnerable to secondhand smoke and qualitative findings indicated that residents were concerned about children’s exposure and health. Messaging that stresses risks to children should therefore be emphasized. As residents were not always aware of the harms that smoking in their apartment posed to other residents’ health better understanding that smoking in apartments poses involuntary harms to others could help increase acceptance of the policy and reduce concerns about restrictions on liberties.

Additionally, from an ethical standpoint, the policy should be designed to maximize its positive health impact while minimizing intrusion on residents’ liberties and any adverse consequences. First, as smoking is an addiction and quitting often requires multiple attempts, referrals and information about cessation support should be provided to assist residents who wish to quit. This recommendation is in keeping with a recent study that showed that public housing residents had disproportionately higher rates of smoking and while the majority had received smoking cessation advice from a medical provider, most did not follow up. This led the authors to conclude that “efforts to eliminate differences in tobacco use should consider place-based smoking cessation interventions that extend cessation support beyond clinical settings.” (Feinberg et al., 2017) Cessation efforts can be further bolstered by other tobacco control measures to reduce the availability of tobacco products in residential communities and with the previously-discussed educational information about the harms of smoking and secondhand smoke.

Second, acceptable smoking areas should be identified and officially designated. There were several areas that functioned as de facto designated smoking areas; there was also consensus among residents that those areas should be reclassified as designated smoking areas so as to accommodate smokers. It is possible that smoking areas may detract from a full realization of a complete smoke-free goal, particularly with respect to denormalization of smoking. However, smoking areas represent an appropriate compromise that minimizes the extent to which the policy restricts residents’ choices and respects the dignity of residents who choose to smoke. Given the expense and difficulty of moving, particularly for low-income residents, lack of smoking areas may pose an undue burden on residents who are smokers and unable or unwilling to quit. Furthermore, given that smokers were smoking in their apartments in the absence of such spaces, smoking areas may improve compliance.

Third, policies must be designed to ensure they do not cause greater harm than benefit to residents’ health. They can be marketed as opportunities to save money, improve health and promote self-care, which may be especially appealing in affordable housing settings. Regulation of the policy should avoid police involvement or evicting residents, as these approaches would increase housing insecurity and contradict the purpose of affordable housing provision, and deepen hardships among low-income and racial/ethnic minority populations. Because marijuana is an illegal substance in many states, smoking it outdoors may place residents at risk of arrest. Furthermore, the health risks of secondhand exposure to marijuana smoke as well as to e-cigarettes are not well established, and e-cigarettes could serve a harm reduction purpose when substituted for traditional cigarettes (National Academies of Sciences, Engineering, and Medicine, 2018). Housing providers should weigh these considerations in determining appropriate rules around traditional cigarettes, marijuana, and e-cigarettes, and sanctions for violating them.

Fourth, many hazardous housing issues cited by residents went unaddressed by management, raising questions as to why policies that involve such restrictions should be the sole focus for healthy housing initiatives. Instead, a smoke-free policy should be part of a broader concern for healthy housing. Property management must properly and promptly tend to maintenance issues and resident concerns, and more generally communicate and treat residents with genuine respect. Doing so is essential to cultivating goodwill among residents and ensuring their ongoing cooperation. This model may be of particular importance for smoke-free policy adoption in affordable and public housing. Marginalized (e.g., lower-income or racial minority) groups are more likely to experience substandard housing conditions (Dragan et al., 2015; U.S. Census Bureau), indicating that this contract breakdown may occur more often in low-income housing. Furthermore, residents of low-income housing have fewer resources to seek alternative housing with a higher-quality provider.

Finally, on a policy level, smoke-free housing should be adopted as widely as possible across all income levels. Smoke-free housing policy can positively impact the health of residents of any kind of building. Lower-income people should have equal access to the protections afforded by such policy; at the same time, policy mandates targeted only at lower-income people, such as the current public housing mandate, raise concerns about undue surveillance of the behavior of marginalized populations. Widespread adoption across all demographics would address equity concerns allowing everyone to benefit from smoke-free environments.

4.3. Strengths and limitations

Our comprehensive methodological approach relied on several complementary data sources to capture beliefs, preferences, and actions regarding smoke-free housing policies and to understand their drivers and potential leverage points. By simultaneously collecting qualitative and quantitative information, we were able to elucidate the explanatory factors behind the persistence of smoking in buildings with active smoke-free policies. Our quantitative findings indicated support for the policy in a broad sense but less support for in-unit restrictions; the nuances of this perspective were explained by the qualitative data which shed light on the implicit boundaries of the smoke-free policy. The value of this approach is particularly evident in our findings regarding the social contract of smoke-free housing.

Our focus on three affordable housing buildings in one urban area provides our study with both strengths and limitations. Findings may not be generalizable to settings with different affordability statuses or demographic characteristics. Conversely, this contained sample population strengthens our conclusions as it allowed for the identification of population-specific differences, an imperative for addressing disparities. The divergence of our findings from the literature – namely that older residents are more likely to support the policy, while previous literature has found no association – may indicate that these characteristics are specific to affordable housing. Similarly, our focus on a single housing provider allowed us to examine the role of the relationship between landlord and residents in the context of smoke-free policy adoption. It is not clear if the challenges reported here are characteristic of affordable housing in particular, or if they speak to the challenges of adopting smoke-free rules in multi-unit rental housing more generally. Future research should explore factors influencing compliance and enforcement in buildings with varying demographic characteristics and affordability statuses, including higher-income settings.

5. Conclusion

Our findings provide new insights regarding tobacco health disparities with respect to the design and adoption of smoke-free housing policies targeting racial/ethnic minoritized groups and low- and moderate-income populations. Smoking is substantially more prevalent among individuals living in affordable/public housing than other types, and they also experience more difficulty quitting. Although barriers to smoke-free policy compliance in other types of multi-unit housing may be similar, focused attention on low-income housing would conceivably achieve meaningful improvements in smoking-related disparities. Our findings indicate, however, that adopting a policy is not in itself enough to sufficiently achieve smoke-free residential settings and thereby rectify persistent disparities. As such, housing providers must seek to understand and address factors that impede effectiveness, which may be population- or setting-specific. In particular, in low- and moderate-in-come housing such factors may include deferred maintenance and frayed management relationships that undermine the perceived legitimacy of the policy. Relatedly, providers must ensure that the policy is embedded in a broader concern for housing as a platform for residents’ health, and does not risk causing greater harm through undue detrimental effects on their dignity, liberties, or residential stability – concerns that are particularly pressing for low-income, marginalized populations. Further research is needed on how various factors affect smoke-free policy adoption in other contexts, and should incorporate both qualitative and quantitative methods.

Fig. 2.

Hallway window littered with cigarette butts.

Fig. 3.

Staircases with cigarette filter and packet.