Estimating Acceptability of Financial Health Incentives

Abstract

A growing body of evidence suggests that financial incentives can influence health behavior change, but research on the public acceptability of these programs and factors that predict public support have been limited. A representative sample of U.S. adults (N = 526) were randomly assigned to receive an incentive program description in which the funding source of the program (public or private funding) and targeted health behavior (smoking cessation, weight loss, or colonoscopy) were manipulated. Outcome variables were attitude toward health incentives and allocation of hypothetical funding for incentive programs. Support was highest for privately funded programs. Support for incentives was also higher among ideologically liberal participants than among conservative participants. Demographics and health history differentially predicted attitude and hypothetical funding toward incentives. Incentive programs in the United States are more likely to be acceptable to the public if they are funded by private companies.

Keywords

cancer prevention and screening financial incentive program health policy health promotion obesity smoking and tobacco use

Financial health incentive programs (hereafter referred to as “incentives” or “incentive programs”) are attracting increasing attention as a motivational tool for health promotion (O’Donnell, 2012). These incentives may take various forms, including cash payments and lower insurance premiums, and are defined as “cash or cash-like rewards provided contingent on performance of healthy behaviours” (Giles, Robalino, Sniehotta, Adams, & McColl, 2015, p. 146). One explanation for the growing emphasis on incentives is their increasing prominence in the health policy arena. In the United States, for example, the Patient Protection and Affordable Care Act (2010) allows employers to provide financial rewards for employee participation in wellness plans. Given the interest in implementing incentives as a means of health promotion, it becomes increasingly important to understand the programmatic and individual factors related to acceptability of incentives, as acceptability of incentives is needed for widespread adoption.

Although there is evidence for the effectiveness of incentives, incentives are not likely to be broadly implemented if they are not acceptable to the public, as public opinion has a substantial impact on policy (Burstein, 2003). Prior work has found both programmatic features (e.g., types of incentives, health behaviors incentivized, effectiveness of incentives) and individual factors (e.g., personal benefit, health history) are associated with acceptability of incentives (Diepeveen, Ling, Suhrche, Roland, & Marteau, 2013; Giles et al., 2015; Promberger, Dolan, & Marteau, 2012). However, most (but not all) prior research has been limited by a lack of representative and generalizable samples, reliability and validity of measures, qualitative research, and consideration of demographic factors (Giles et al., 2015). In particular, we could not identify a study that included political ideology, which may play a role in determining acceptability of health-related policies (Brodie, Deane, & Cho, 2011). Nor could we identify a study that compared the acceptability of privately funded versus publically funded incentive programs, even though differences in funding have been debated as a potential influence (Giles et al., 2015). Therefore, we extend the literature by addressing three limitations: representative sample, demographic factors, and funding source. We employed a U.S. nationally representative sample to examine individual factors (demographics, health history, and political ideology) related to the acceptability of incentives differing in terms of funding (public vs. private funding) and targeted health behavior (smoking cessation, weight loss, or colonoscopy).

Incentives are thought to change health behaviors by providing immediate tangible rewards for behaviors that would otherwise take time to produce desirable effects (Marteau, Ashcroft, & Oliver, 2009). The advantage of immediate rewards in this context has been explained by learning theory principles and present bias (Marteau et al., 2009). Systematic reviews and original research suggest that incentives can be effective for promoting smoking cessation, weight loss, and cancer screenings (Giles, Robalino, McColl, Sniehotta, & Adams, 2014; Mantzari et al., 2015; Volpp et al., 2008). We examine acceptability of incentives for smoking cessation, weight loss, and colonoscopy. While prior work has focused on incentives for breast, cervical, and colon cancer screenings (Stone et al., 2002), we focus on colonoscopy because it is recommended for all adults aged 50 years and older (U.S. Preventive Services Task Force, 2015).

Method

Participants

Data were collected from participants (N = 526) as part of the Annenberg National Health Communication Survey, an ongoing monthly survey of a nationally representative sample of U.S. adults. The experiment was embedded in a noncore module; the completion rate for this module was 53.6%. Full survey methods are detailed in prior publications (e.g., Bleakley, Hennessy, & Fishbein, 2006).

Procedures

After answering questions about personal health history, participants were assigned to one of six conditions using simple randomization into the 2 (funding source: U.S. Government vs. a private company) × 3 (health behavior: quitting smoking vs. reaching a healthy weight vs. getting screened for colorectal cancer with colonoscopy) between-subjects design. The health issues were selected from prior research on the effectiveness of incentive programs (Giles et al., 2014; Mantzari et al., 2015; Stone et al., 2002; Volpp et al., 2008). Participants read a description of a health incentive program that was manipulated to reflect one of the two funding sources and one of the three health issues (supplemental materials, available online at heb.sagepub.com/content/by/supplemental-data). Participants then completed attitude and funding allocation measures before answering a question about political ideology.

Measures

Attitude toward the incentive program, an outcome measure, was measured by having participants assess the health incentive program that they read about using semantic differential word pairs: bad–good, foolish–wise, harmful–beneficial, and unnecessary–necessary. The items were averaged to create a scale (Cronbach’s α = .92, M = 3.81, SD = 0.93, range = 1-5, negative end of the scale = 1).

Acceptability of incentives was further measured using a modified constant-sum comparison technique, which has been used to capture societal preferences for health care resource allocation rather than individual preferences (Skedgel, Wailoo, & Akehurst, 2013). Participants were told to imagine that they had $3,000 to divide among three different health programs. They read descriptions of three programs that matched the behavior mentioned in their experimental condition but differed in program implementation; options included an Internet-based information program, an incentive program, and an in-person education program. After reading the three program descriptions, participants read the following: “If you had to divide $3,000 among the three programs, how would you do it? You have to use all of the money and give it to at least one of the three programs.” They then saw three boxes, one for each type of program (information, incentive, education) and typed in the amount of the $3,000 they wished to allocate to each type of program. On average, participants allocated more funding to the incentive program than the others. Mean amount across all conditions for the incentive program was $1291.77 (SD = $837.92); the full measure and descriptive information for the other programs can be found in the online supplemental materials.

Political ideology was assessed by the statement, “In general, do you think of yourself as”; response options were very liberal (1) to very conservative (7). To assess health history, participants were asked if they had “ever been overweight,” and “ever been told by a health professional that you have cancer”; response options for both items were yes (1) or no (0). Participants were also asked, “During the past 7 days, how many cigarettes did you smoke on a typical day?” (fill-in-the-blank response). The market research company that administered the survey provided demographic characteristics (race/ethnicity, gender, education, household income, household size, and age). Equivalized annual household income was calculated by dividing the annual household income by the square root of the number of people in the household (Blakely, Kennedy, Glass, & Kawachi, 2000); it was highly skewed (skewness = 1.32, SE of skewness = 0.11) and thus multiplied by the natural log.

Analysis

The relationships were examined by estimating ordinary least squares stepwise regression models. Participants who reported smoking at least one cigarette per day in the past week were recoded as current smokers (1; 0 = nonsmokers).

Results

Overall, this sample more accurately reflects the U.S. population than samples used in clinical studies of opinion on patient incentives (Long, Helweg-Larsen, & Volpp, 2008). Participant demographics are presented in Table 1 along with data from the U.S. Census Bureau (2010) for comparison purposes, which few prior studies on acceptability have provided (Giles et al., 2015). This sample is also comparable to Internet-based panels used in prior survey research (Promberger, Brown, Ashcroft, & Marteau, 2011). The median annual equivalized household income was $38,891 (M = $45,000, SD = $33,896) compared with a reported national median of $31,000 (Organisation for Economic Co-operation and Development [OECD], 2011). Participant political ideology was well-distributed (24.9% identified themselves as somewhat to very liberal and 39.1% identified as somewhat to very conservative). Finally, 64% of our sample reported ever being overweight, 14% currently smoked, and 9% had ever had cancer.

Table 1.

Sample Demographics Compared With U.S. Census 2010 Data.

Variable	Sample		U.S. Census 2010
Variable	%	Mdn	%	Mdn
Race/ethnicity
White, not Hispanic	76		72
Black, not Hispanic	10		13
Hispanic	9		16
Other	5		6
Male	42		49
Age (years)		50		37
Education
<High school	9		14
High school	26		29
Some college	32		29
Bachelor’s degree or higher	33		28

Note. Sample demographics compared with data from the U.S. Census Bureau (2010). Participant mean age was 49.22 (SD = 16.21) years.

As shown in Table 2, participants who read about privately funded programs had more positive attitudes toward incentives than those who read about publicly funded incentives. Similarly, private funding predicted higher funding allocation for health incentives; participants who read about privately funded incentives allocated $209.96 more than those who read about publicly funded incentives (Table 3).

Table 2.

Effects of Funding, Behavior, Ideology, Demographics, and Health History on Attitude Toward Incentive Programs, Results From Ordinary Least Squares Regression.

Variable	Model 1: Program characteristics only, b (SE)	Model 2: Program characteristics and ideology, b (SE)	Model 3: Program characteristics, ideology, demographic variables, and health history, b (SE)
Private funding	0.20 (0.08)*	0.22 (0.08)**	0.20 (0.08)*
Incentivized behavior^a
Weight loss^b	−0.19 (0.10)	−0.19 (0.10)	−0.20 (0.10)
Smoking cessation	−0.26 (0.10)*	−0.28 (0.10)	−0.28 (0.10)**
Political ideology		−0.08 (0.03)	−0.07 (0.03)*
Race/ethnicity^c
Black, not Hispanic			0.05 (0.14)
Hispanic			0.10 (0.16)
Other			−0.47 (0.18)**
Male			−0.11 (0.08)
Education^d
High school			0.05 (0.17)
Some college			0.06 (0.17)
Bachelor’s degree or higher			0.22 (0.18)
Equivalized income (in thousands)			−0.07 (0.12)
Age			−0.00 (0.00)
Ever been overweight			0.20 (0.09)*
Smoke every day			0.13 (0.13)
Ever had cancer			0.01 (0.15)
F	4.11**	5.01**	2.49**
Adjusted R²	0.02	0.03	0.05

Note. N = 494. Unstandardized regression coefficients are reported. “Private funding” is a dichotomous variable; the comparison is to publicly funded incentive programs. Ideology scale is coded from 1 (very liberal) to 7 (very conservative). The outcome, attitude toward incentives, is measured on a 5-point scale. F refers to a test of the statistical significance of the entire regression equation. The adjusted R² is a measure of the proportion of variance explained that is adjusted for the number of variables in the model.

Comparison condition is colonoscopy. ^bWhen the weight loss condition was contrasted with the smoking cessation condition, the difference was not significant in any model. ^cComparison group is “White, not Hispanic.” ^dComparison group is less than high school.

p < .05. **p < .01.

Table 3.

Effects of Funding, Behavior, Ideology, Demographics, and Health History on Funding Allocation to Incentive Programs, Results From Ordinary Least Squares Regression.

Variable	Model 1: Program characteristics only, b (SE)	Model 2: Program characteristics and ideology, b (SE)	Model 3: Program characteristics, ideology, demographic variables, and health history, b (SE)
Private funding	188.36 (73.75)*	203.76 (73.38)**	209.96 (71.62)**
Incentivized behavior^a
Weight loss^b	−184.16 (90.65)*	−188.66 (89.97)*	−159.37 (87.83)
Smoking cessation	−177.93 (91.59)	−191.61 (91.01)*	−162.52 (89.48)
Political ideology		−74.83 (25.21)**	−70.34 (25.05)**
Race/ethnicity^b
Black, not Hispanic			−364.02 (125.90)**
Hispanic			−69.29 (132.78)
Other			−297.40 (153.18)
Male			−16.27 (72.43)
Education^d
High school			97.57 (144.01)
Some college			183.15 (144.76)
Bachelor’s degree or higher			250.35 (153.36)
Equivalized income (in thousands)			255.14 (104.06)*
Age			−5.95 (2.38)*
Ever been overweight			134.78 (74.92)
Smoke every day			248.30 (108.19)*
Ever had cancer			108.50 (125.89)
F	3.79**	5.09**	4.19**
Adjusted R²	0.02	0.03	0.09

Note. N = 511. Unstandardized regression coefficients are reported. “Private funding” is a dichotomous variable; the comparison is to publicly funded incentive programs. Ideology scale is coded from 1 (very liberal) to 7 (very conservative). Funding allocated to incentive programs could range from $0 to $3,000, thus coefficients can be interpreted as dollar amounts. F refers to a test of the statistical significance of the entire regression equation. The adjusted R² is a measure of the proportion of variance explained that is adjusted for the number of variables in the model.

p < .05. **p < .01.

Table 2 shows some evidence of differing levels of support for incentives by health behavior. Those in the smoking cessation condition had significantly more negative attitudes toward incentive programs than those who read about colonoscopy incentives. However, there were no significant differences between weight loss and colonoscopy incentive programs in terms of attitudes or funding allocation (Tables 2 and 3).

Individual characteristics also predicted support for incentive programs. Increasingly liberal ideology was associated with more positive attitudes (Table 2) and increased funding (Table 3) toward incentives. Participants reporting their race as “Other” had less favorable attitudes toward incentives than non-Hispanic White participants (Table 2). Non-Hispanic Black, lower income, and increasing age were all associated with a decrease in incentive funding (Table 3). Individuals who reported ever being overweight had more positive attitudes toward all incentive programs than those who had not ever been overweight, while current smokers allocated $248.30 more to incentive programs than nonsmokers. Interactions between personal health history and behavioral condition were not significant.

Discussion

This study examined how programmatic and individual factors are associated with support for different types of health incentive programs. Participants were more willing to support incentive programs that were privately funded rather than publicly funded. Participants also had less favorable attitudes toward smoking cessation incentive programs but did not allocate less funding to such programs when compared to colonoscopy incentive programs. In addition, liberals were more likely than conservatives to have positive attitudes toward and hypothetically fund health incentives.

Prior research has not explored the effects of incentive funding source or political ideology on support for incentives. The effect of funding source on support for incentives is consistent with the idea of self-interest (Sears, Lau, Tyler, & Allen, 1980); respondents are more supportive of incentives when public funds (presumably their tax dollars) are not involved. Some privately funded incentives, such as those funded by health insurance companies, private employers, or not-for-profit organizations, may be more acceptable than others. Future research should examine public acceptability of different types of privately funded incentive programs. The association between political ideology and hypothetical funding is consistent with perceptions of responsibility. Conservatives are less likely to support financial assistance than liberals, particularly if those in need of assistance are perceived as responsible for their situation (Farwell & Weiner, 2000). Thus, one strategy to increase the acceptability of incentives among the general public is to minimize personal responsibility over health. For example, there is a general perception that smokers are responsible for their smoking behavior (Promberger et al., 2011), but tobacco marketing also plays a role in smoking initiation (e.g., Biener & Siegel, 2000).

Prior research has also not examined whether acceptability is differentially associated with demographic factors (Giles et al., 2015). However, several demographic differences seem to exist, particularly in the hypothetical funding allocation measure. These differences suggest a non-Hispanic Black racial identity, lower equivalized income, and increasing age are associated with less support for incentives on a societal level, even though many of these demographic characteristics were not significantly associated with different personal attitudes toward incentives.

Consistent with prior research, our results showed some differences between behavioral condition and support for health incentive programs. Similar to Promberger et al. (2012), participants had less favorable attitudes toward smoking cessation incentives than colonoscopy incentives. However, this result contradicts Diepeveen et al. (2013), which found incentives targeting smoking were supported more frequently than other behaviors. One possible explanation for participants’ preference for programs that encourage colonoscopy is that individuals who need smoking cessation programs are more highly stigmatized than those needing colorectal cancer screening programs (Stuber, Galea, & Link, 2009). Additionally, this finding could be explained through alternate mechanisms. For example, most older adults are encouraged to have colorectal cancer screenings, while smoking cessation only applies to a segment of the population; thus, it is possible that more people felt the colonoscopy incentive would be personally beneficial.

Personal health history also had an influence on support for incentive programs. As in prior research (Long et al., 2008), current smokers allocated more money to all incentive programs than nonsmokers. One explanation might be control; smoking cessation is generally accepted as a difficult behavior to change, and those who have firsthand knowledge of this challenge may be more eager to accept a variety of cessation methods (Long et al., 2008).

This research poses several potential limitations. First, the selection of behavioral conditions may limit researchers’ ability to generalize the findings to other health behaviors. Second, this experiment used only a “reward” framework, so findings cannot be extended to programs that employ penalties as health incentives. Additionally, because these findings were produced in an experimental setting, they may not generalize to nonexperimental settings in which health incentive policies are often publicized and discussed. The between-subjects design may also be considered a limitation because comparisons cannot be made within individuals. We chose a between-subjects design to limit carryover effects and participant fatigue—disadvantages of within-subjects designs (Charness, Gneezy, & Kuhn, 2012)—and because it concealed our intent to compare multiple behaviors. Finally, the total amount of variance explained across the models is low, suggesting that there are other unmeasured variables that influence acceptability of incentives, such as program effectiveness (Giles, et al., 2015; Promberger et al., 2012).

Conclusion

The extent to which incentives are employed as interventions for health behavior change will depend on the actual effectiveness and public acceptability of incentive programs. The aim of this study was not to determine whether or not incentives should be used, but examine individual and programmatic factors that influence acceptability. Overall, this study found that acceptability of incentives varied by demographic characteristics, political ideology, incentivized behavior, and funding source in a U.S. nationally representative sample. Privately funded wellness programs are more likely to be well-received than publicly funded incentives, though acceptance could vary based on the health behavior incentivized and individual characteristics of participants. Framing of incentive programs should consider deemphasizing individual responsibility for health behaviors when possible and appropriate. Additional research examining the influence of various private funding sources and wording that deemphasizes individual responsibility is needed before these results should be used in real-world decisions.

Supplemental Material

sj-docx-1-heb-10.1177_1090198116664072 – Supplemental material for Estimating Acceptability of Financial Health Incentives

Supplemental material, sj-docx-1-heb-10.1177_1090198116664072 for Estimating Acceptability of Financial Health Incentives by Elisabeth Bigsby, Holli H. Seitz, Scott D. Halpern, Kevin Volpp and Joseph N. Cappella in Health Education & Behavior

Footnotes

Acknowledgements

The authors wish to acknowledge that survey support was provided by the Annenberg National Health Communication Survey, which is supported by the Annenberg Trust at Sunnylands, the Annenberg School for Communication and Journalism at the University of Southern California, and the Annenberg School for Communication at the University of Pennsylvania.

Authors’ Note

Holli H. Seitz is now at Mississippi State University. An earlier version of this article was presented at the 99th Annual Convention of the National Communication Association in Washington, DC. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors wish to acknowledge the funding support of the National Cancer Institute’s Center of Excellence in Cancer Communication located at the Annenberg School for Communication, University of Pennsylvania (P20-CA095856-06).

Supplemental Material

The online supplementary materials are available at .

References

Biener

Siegel

(2000). Tobacco marketing and adolescent smoking: More support for a causal inference. American Journal of Public Health, 90, 407-411.

Blakely

T. A.

Kennedy

B. P.

Glass

Kawachi

(2000). What is the lag time between income inequality and health status? Journal of Epidemiology & Community Health, 54, 318-319.

Bleakley

Hennessy

Fishbein

(2006). Public opinion on sex education in US schools. Archives of Pediatrics and Adolescents Medicine, 160, 1151-1156. doi:10.1001/archpedi.160.11.1151

Brodie

Deane

Cho

(2011). Regional variations in public opinion on the affordable care act. Journal of Health Politics, Policy and Law, 36, 1097-1103. doi:10.1215/03616878-1460587

Burstein

(2003). The impact of public opinion on public policy: A review and an agenda. Political Research Quarterly, 56, 29-40.

Charness

Gneezy

Kuhn

M. A.

(2012). Experimental methods: Between-subject and within-subject design. Journal of Economic Behavior & Organization, 81, 1-8. doi:doi:10.1016/j.jebo.2011.08.009

Diepeveen

Ling

Suhrche

Roland

Marteau

T. M.

(2013). Public acceptability of government intervention to change health-related behaviours: A systematic review and narrative synthesis. BMC Public Health, 13, 756.

Farwell

Weiner

(2000). Bleeding hearts and the heartless: Popular perceptions of liberal and conservative ideologies. Personality and Social Psychology Bulletin, 26, 845-852. doi:10.1177/0146167200269009

Giles

E. L.

Robalino

McColl

Sniehotta

F. F.

Adams

(2014). The effectiveness of financial incentives for health behavior change: Systematic review and meta-analysis. PLoS ONE, 9, e90347.

10.

Giles

E. L.

Robalino

Sniehotta

F. F.

Adams

McColl

(2015). Acceptability of financial incentives for encouraging uptake of healthy behaviours: A critical review using systematic methods. Preventative Medicine, 73, 145-158. doi:10.1016/j.ypmed.2014.12.029

11.

Long

J. A.

Helweg-Larsen

Volpp

K. G.

(2008). Patient opinions regarding ‘pay for performance for patients’. Journal of General Internal Medicine, 23, 1647-1652.

12.

Mantzari

Vogt

Shemilt

Wei

Higgins

J. P. T.

Marteau

T. M.

(2015). Personal financial incentives for changing habitual health-related behaviors: A systematic review and meta-analysis. Preventive Medicine, 75, 75-85. doi:10.1016/j.ypmed.2015.03.001

13.

Marteau

T. M.

Ashcroft

R. E.

Oliver

(2009). Using financial incentives to achieve healthy behaviour. BMJ, 338, b1415.

14.

O’Donnell

M. P.

(2012). We are beginning to speak with one voice: Consensus statement on wellness incentives. American Journal of Health Promotion, 27(1), iv-vii. doi:10.4278/ajhp.27.1.iv

15.

Organisation for Economic Co-operation and Development. (2011). Household income. In Society at a glance 2011: OECD social indicators (pp. 42-44). Paris, France: Author. doi:10.1787/soc_glance-2011-6-en

16.

Patient Protection and Affordable Care Act, Pub. L. No.111-148, 124 Stat. 119 (2010).

17.

Promberger

Brown

R. C. H.

Ashcroft

R. E.

Marteau

T. M.

(2011). Acceptability of financial incentives to improve health outcomes in UK and US samples. Journal of Medical Ethics, 37, 682-687. doi:10.1136/jme.2010.039347

18.

Promberger

Dolan

Marteau

T. M.

(2012). “Pay them if it works”: Discrete choice experiments on the acceptability of financial incentives to change health related behavior. Social Science & Medicine, 75, 2509-2514. doi:10.1016/j.socscimed.2012.09.033

19.

Sears

D. O.

Lau

R. R.

Tyler

T. R.

Allen

H. M.

(1980). Self-interest vs. symbolic politics in policy attitudes and presidential voting. American Political Science Review, 74, 670-684.

20.

Skedgel

C. D.

Wailoo

A. J.

Akehurst

R. L.

(2013). Choosing vs. allocating: Discrete choice experiments and constant-sum paired comparisons for the elicitation of societal preferences. Health Expectations, 18, 1227-1240. doi:10.1111/hex.12098

21.

Stone

E. G.

Morton

S. C.

Hulscher

M. E.

Maglione

M. A.

Roth

E. A.

Grimshaw

J. M.

. . . Shekele

P. G.

(2002). Interventions that increase use of adult immunization and cancer screening services: A meta-analysis. Annals of Internal Medicine, 136, 641-651.

22.

Stuber

Galea

Link

B. G.

(2009). Stigma and smoking: The consequences of our good intentions. Social Service Review, 83, 585-609. doi:10.1086/650349

23.

U.S. Census Bureau. (2010). American factfinder. Retrieved from http://factfinder2.census.gov

24.

U.S. Preventive Services Task Force. (2015). Colorectal cancer: Screening. Retrieved from http://www.uspreventiveservicestaskforce.org/Page/Document/UpdateSummaryFinal/colorectal-cancer-screening

25.

Volpp

K. G.

John

L. K.

Troxel

A. B.

Norton

Fassbender

Loewenstein

(2008). Financial incentive-based approaches for weight loss: A randomized trial. JAMA Journal of the American Medical Association, 300, 2631-2637.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB