We systematically evaluated smoking-related costs in multiunit housing. From 2008 to 2009, we surveyed California multiunit housing owners or managers on their past-year smoking-related costs and smoke-free policies. A total of 27.1% of respondents had incurred smoking-related costs (mean $4935), and 33.5% reported complete smoke-free policies, which lowered the likelihood of incurring smoking-related costs. Implementing statewide complete smoke-free policies may save multiunit housing property owners $18 094 254 annually.

Approximately 10.6 million Californians live in multiunit housing (MUH),1 where units with smoke-free policies can be affected by environmental tobacco smoke morbidity and mortality effects through shared air spaces and ventilation or drifting from outside.2,3 Lack of information on MUH smoking-related costs (e.g., cleaning, replacement) may contribute to MUH owners’ and managers’ reluctance to implement smoke-free policies.4,5 We surveyed California MUH owners and managers to determine (1) the smoking-related costs borne by MUH owners, (2) the smoking-related costs prevented in MUH as the result of smoke-free policies, and (3) the economic benefits of all MUH implementing complete smoke-free policies.

Between July 2008 and February 2009, we conducted a computer-assisted telephone interview survey among 343 California Apartment Association (CAA) members who owned or managed MUH, with an overall response rate of 22.4% and an overall cooperation rate of 40.5%.6 CAA members were randomly selected and were sent presurvey notification letters proportionate to sizes of the 20 regional CAA chapters and to the small and large properties within each chapter (we defined “large” as  ≥ 16 units, which requires an on-site property manager).

We used survey items and categories adapted from the Property Owners and Managers Survey7 to ask respondents to estimate smoking-related costs beyond standard operations that were incurred during the preceding 12 months for the entire property with the most recently vacated unit. Categories included cleaning, repairs and maintenance, painting and decorating, trash collection, fire damage, property insurance, fire insurance, other insurance, legal costs, administrative costs, and other operating costs. We asked respondents whether the property had a complete smoke-free policy, which was defined as no smoking permitted anywhere on the property, including both in private units and in public (common) places. We then asked those who responded “no” whether any buildings, public places, or units on the property were smoke-free. If yes, we designated the property as having a partial smoke-free policy. If all responses were negative, we designated the property as having no smoke-free policy. Other domains of the survey included property, building, and unit characteristics and personal characteristics and beliefs of the respondent. Poststratification weights for the final sample reflected the overall statewide CAA member sampling frame.

We used Stata version 10.0 (StataCorp LP, College Station, TX) to perform all statistical analyses, using 2-tailed significance levels. We analyzed a zero-inflated negative binomial model8,9 of property smoking-related costs predicted by

    smoke-free policy status,

    the number of units,

    an on-site owner or manager,

    rent regulation,

    shared ventilation,

    shared furnaces, and

    respondent smoking status.

We used recycled predictions10 to estimate the base case and smoke-free scenarios for all California MUH by multiplying the predicted prevalence and amount of smoking-related costs with the total units in structures with  ≥ 2 units in California from the American Community Survey from 2005 to 2007.1

One third of properties had a complete smoke-free policy, but nearly half had no smoke-free policy. Small properties had more than a threefold higher rate of having a complete smoke-free policy compared with large properties (Table 1). More than one quarter of properties (27.1%) experienced smoking-related costs; large properties had nearly a threefold higher rate of smoking-related costs compared with small properties.


TABLE 1— Predictors of Smoking-Related Costs of Multiunit Housing Property With Most Recently Vacated Unit: California, 2008–2009

TABLE 1— Predictors of Smoking-Related Costs of Multiunit Housing Property With Most Recently Vacated Unit: California, 2008–2009

Descriptive Analysis
Zero-Inflated Negative Binomial Model
CharacteristicsSmall (< 16 Units), No. (%)Large (≥ 16 Units), No. (%)Total, No. (%)Logistic Model, Coefficient (P)Negative Binomial, Coefficient (P)
Total196 (65.0)147 (35.0)343 (100.0)
Smoking status*
 Never smoker (Ref)62.755.
 Former smoker33.029.031.6−0.138 (.717)0.562 (.128)
 Current smoker4.316.08.4−0.353 (.522)0.754 (.232)
Number of units at property***0.004 (.181)0.004 (.001)
 Average (SD)5.8 (3.5)100.2 (153.5)38.8 (93.8)
Has units with rent regulation
 Yes36.234.635.60.212 (.479)−0.093 (.786)
 No (Ref)63.865.464.41.01.0
Building with last vacated unit has central ventilation**
 Yes20.537.826.60.291 (.584)0.724 (.021)
 No (Ref)79.562.
Building with last vacated unit has individual furnaces***
 Yes88.270.582.00.149 (.718)0.338 (.574)
 No (Ref)11.829.518.01.01.0
On-site owner or manager lives at the property***0.610 (.052)0.239 (.454)
 Owner only9.60.46.4
 On-site manager only10.280.434.8
 Both owner and manager5.12.34.1
 Neither (Ref)
Smoke-free policy***
 Complete smoke-free policy (Ref)44.413.433.51.01.0
 Partial smoke-free policy12.035.720.30.857 (.049)1.015 (.108)
 No smoke-free policy43.750.946.20.722 (.079)0.252 (.639)
Smoking-related costs***

Note. Weighted analyses only. Significance tests for descriptive analyses compare small and large categories. Reported rates exclude responses coded as missing, don't know, or refused. Regression analyses examined a binary variable for on-site owner or manager. Zero-inflated negative binomial model statistics: ln(α) = 0.454; P < .001; A = 1.575.

*P < .05; **P < .01; ***P < .001.

Among all properties experiencing smoking-related costs (Table 2), the mean cost was $4935. Even after accounting for withheld deposits, the mean cost was $4252. The mean per unit cost was $282, with small properties having higher per unit costs than large properties ($578 vs $87). Properties with complete smoke-free policies experienced smoking-related costs, but less frequently and with lower mean amounts (16.3%, $1866) than did those of properties with partial smoke-free policies (39.7%, $9573) or no smoke-free policies (29.5%, $3425).


TABLE 2— Past-Year Operation Costs for Multiunit Housing Property With Most Recently Vacated Unit, by Smoking Policy: California, 2008–2009

TABLE 2— Past-Year Operation Costs for Multiunit Housing Property With Most Recently Vacated Unit, by Smoking Policy: California, 2008–2009

Property by Smoking PolicyMedian, $Min, $Max, $Weighted, $, Mean (SD)
All properties
 Overall cost64 400902 262 500167 655 (320 836)
 Smoking-related cost20005084 0004935 (11 334)
 Withheld deposit200013 000683 (1508)
 Smoking-related cost minus withheld deposit1000−260083 2004252 (10 945)
Completely smoke-free
 Overall cost48 60038051 441 000182 159 (437 521)
 Smoking-related cost240010085001866 (2706)
 Withheld deposit002400244 (568)
 Smoking-related cost minus withheld deposit800080001623 (2560)
Partially smoke-free
 Overall cost147 333300765 000245 203 (233 226)
 Smoking-related cost340022584 0009573 (18 914)
 Withheld deposit288013 000914 (2206)
 Smoking-related cost minus withheld deposit1998083 2008659 (18 204)
Never smoke-free
 Overall cost54 051902 262 500116 129 (308 434)
 Smoking-related cost20005027 0003425 (5273)
 Withheld deposit20005400723 (1224)
 Smoking-related cost minus withheld deposit1000−260025 5002703 (5311)

Our multivariable analysis showed that the likelihood of incurring smoking-related costs at a MUH property with a complete smoke-free policy was less than half that of those with a partial smoke-free policy (odds ratio [OR] = 0.42) or without a smoke-free policy (OR = 0.48), although the latter finding was marginally significant at P = .08 (Table 1). Having an on-site owner or manager was also significantly associated with incurring smoking-related costs. We found that smoke-free policy status was not associated with the amount of smoking-related costs; property size and central ventilation were the only significant associations.

We estimate that there are 104 237 California MUH properties, on the basis of 4 044 387 California MUH units1 divided by our survey's mean MUH units per property (38.8). Eliminating all smoking-related costs from the 27.1% of MUH that experience them would save each of the 28 248 properties $1339, for a total amount of averted smoking-related costs in 1 year of $37 824 296. MUH with complete smoke-free policies incur smoking-related costs, but at a lower prevalence rate (19.1% vs 27.1%) and overall amount ($991 vs $1339) than do MUH properties without complete smoke-free policies. If all MUH properties had complete smoke-free policies, 8339 properties would not experience smoking-related costs, and 19 909 properties would each save $348, resulting in total averted smoking-related costs in 1 year of $18 094 254.

To our knowledge, our study is the first systematic estimate of MUH smoking-related costs that are not fully compensated for by withheld deposits. Our findings suggest that MUH owners should expect significant savings from implementing complete, but not partial, smoke-free policies. However, we cannot determine from this cross-sectional survey whether MUH incurs smoking-related costs despite complete or partial smoke-free policies owing to recent transitions to smoke-free policies or as the result of enforcement problems. As far as we know, this study provides the first representative perspective on MUH by evaluating both small MUH (overlooked in previous studies4,5) and large MUH. Small MUH has a higher prevalence of complete smoke-free policies, which may be a secondary response to their higher per unit smoking-related costs compared with those of large MUH.

Our response rate (22.4%) is similar to internal CAA survey response rates. This survey was also suspended while in the field by the governor's executive order S-09–08 as a result of the state's budget crisis, which may have affected response rates, although findings were similar for those who responded before and after the survey suspension. Our study's self-reported costs may be subject to recall bias, but respondents were notified before the survey that they would be asked about property costs, and they provided reasonable responses to the detailed financial questions. We focused on the costs generated by smoking MUH renters and not condominium MUH because many smoking-generated costs may not become apparent until turnover or vacancy of the unit.

MUH smoking-related cost savings, combined with averted health care utilization, morbidity, and mortality from reduced environmental tobacco smoke exposure, suggest substantial benefits from the implementation of complete smoke-free policies in MUH.


This work was supported by a contract with the California Department of Public Health, California Tobacco Control Program (07-65582).

This work was presented at the American Public Health Association Annual Meeting, November 8, 2010.

We would like to acknowledge the assistance of the California Apartment Association, and especially Eric Wiegers, for partnering with us to conduct this project. We would also like to acknowledge the assistance of the Public Research Institute staff, including Kevin Adcock, Nicholas Alvarado, Diane Godard, Rebecca Lee, John Rogers, and Holley Shafer, for conducting the survey and assisting with the development of the survey and survey weights.

Human Participant Protection

This study was approved by the institutional review board at the University of California, Los Angeles, and complied with the Principles of Ethical Practice of Public Health of the American Public Health Association.


1. US Census Bureau. American Community Survey. Washington, DC: US Government Printing Office; 2007. Google Scholar
2. Samet JM, Norman LA, Wilbanks C, Pinto A. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. Atlanta, GA: US Dept of Health and Human Services; 2006. Google Scholar
3. Substances identified as toxic air contaminants. 17 California Code of Regulations §93000 (2007). Google Scholar
4. Center for Tobacco Policy & Organizing. Smokefree housing: owner/manager poll. June 2005. Available at: http://www.center4tobaccopolicy.org/polls-ownermanager. Accessed July 8, 2011. Google Scholar
5. Hewett MJ, Sandell SD, Anderson J, Niebuhr M. Secondhand smoke in apartment buildings: renter and owner or manager perspectives. Nicotine Tob Res. 2007;9(suppl 1):S39S47. Crossref, MedlineGoogle Scholar
6. American Association of Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 7th ed. Lenexa, KS: American Association of Public Opinion Research; 2006. Google Scholar
7. US Census Bureau. Property Owners & Managers Survey. 1995. Available at: http://www.census.gov/hhes/www/housing/poms/poms.html. Accessed July 8, 2011. Google Scholar
8. Sheu ML, Hu TW, Keeler TE, Ong M, Sung HY. The effect of a major cigarette price change on smoking behavior in California: a zero-inflated negative binomial model. Health Econ. 2004;13(8):781791. Crossref, MedlineGoogle Scholar
9. Stata[computer program]. Version 10. College Station, TX: StataCorp LP; 2007. Google Scholar
10. Ong MK, Mangione CM, Romano PS. Looking forward, looking back: assessing variations in hospital resource use and outcomes for elderly patients with heart failure. Circ Cardiovasc Qual Outcomes. 2009;2(6):548557. Crossref, MedlineGoogle Scholar


No related items




Michael K. Ong, MD, PhD, Allison L. Diamant, MD, MSHS, Qiong Zhou, MA, Hye-Youn Park, PhD, and Robert M. Kaplan, PhDMichael K. Ong, Allison L. Diamant, and Qiong Zhou are with the Division of General Internal Medicine and Health Services Research, Department of Medicine, David Geffen School of Medicine, University of California, Los Angeles. Hye-Youn Park is with the California Tobacco Control Program, California Department of Public Health, Sacramento. Robert M. Kaplan is with the Department of Health Services, School of Public Health, University of California, Los Angeles. “Estimates of Smoking-Related Property Costs in California Multiunit Housing”, American Journal of Public Health 102, no. 3 (March 1, 2012): pp. 490-493.


PMID: 21852657