Greenwashed Cigarette Ad Text and Imagery Produce Inaccurate Harm, Addictiveness, and Nicotine Content Perceptions: Results From a Randomized Online Experiment

Abstract

Introduction

The tobacco industry has a long history of circumventing regulations to present their products, inaccurately, as less harmful. Greenwashing (portraying a product as natural/eco-friendly) is increasingly used by tobacco companies and may mislead consumers to believe that certain cigarettes are less harmful than others. This study assesses the effect of some common greenwashing tactics on consumer product perceptions.

Aims and Methods

We conducted an online experiment with 1504 participants ages 18–29, randomized to view a cigarette ad manipulated for presence–absence of a combination of four different greenwashing techniques: greenwashed ad text, greenwashed ad imagery, recycled paper ad background, and image of greenwashed cigarette pack. Participants rated perceived absolute harm, relative harm to other cigarettes, absolute addictiveness, relative addictiveness, and relative nicotine content.

Results

Participants who viewed ads containing greenwashed text were more likely to have inaccurate perceptions about absolute harm (AOR = 1.72), relative harm (AOR = 3.92), relative addictiveness (AOR = 2.93), and nicotine content (AOR = 2.08). Participants who viewed ads containing greenwashed imagery were more likely to have inaccurate perceptions of relative harm (AOR = 1.55), absolute addictiveness (AOR = 1.72), relative addictiveness (AOR = 1.60), and nicotine content (AOR = 1.48). Forty-two percent of those who saw an ad with all greenwashed features believed the product was less harmful than other cigarettes versus 2% of those who saw an ad without greenwashed features.

Conclusions

We found that greenwashed text and imagery produced inaccurate risk perceptions. More active U.S. Food & Drug Administration (FDA) enforcement against such greenwashing and new FDA rulemaking to prohibit unnecessary imagery in tobacco advertising and establish plain packaging requirements would help protect consumers and public health.

Implications

These findings provide evidence that greenwashing tactics used by the tobacco industry increase inaccurate product risk perceptions. These tactics could be a way for the industry to make implicit modified risk claims, despite applicable U.S. Family Smoking Prevention and Tobacco Control Act prohibitions. Findings from this study support the need for prohibitions on these tactics, and the potential for such prohibitions to help protect public health.

Introduction

The term greenwashing refers to the “practice of falsely promoting an organization’s environmental efforts...,”¹ often in a manner that is deceptive.^2,3 This encompasses both “claim greenwashing,”⁴ which involves advertising copy that misleads consumers into believing that a company’s products or services are more environmentally friendly than they actually are, attempts to highlight environmentally friendly practices while ignoring or downplaying practices that may be environmentally harmful,⁴ and “executional greenwashing.” Executional greenwashing does not explicitly claim that a product or company is engaging in eco-friendly practices, but rather relies on imagery and indirect text to “artificially enhance a brand’s ecological image.”⁴ Evidence suggests that perceiving a brand or product as eco-friendly positively affects consumers’ brand perception and product purchase intentions.^5–8 Accumulating evidence indicates the tobacco industry engages in this practice.^9–13

Specifically, recent studies have documented the use of both claim and execution greenwashing in tobacco advertisements, documented both in the text (eg, natural descriptors, claims about sustainable farming) and imagery (eg, plants) in tobacco advertisements. For example, the cigarette brand Natural American Spirit commonly advertises that they use eco-friendly or sustainable farming and business practices, and other studies have documented the use of plant and farm imagery in Natural American Spirit marketing materials.^9–13 Other greenwashing tactics, such as imagery that symbolizes recycled materials (eg, materials that appear to look recycled, such as unbleached paper) or greenwashed pack imagery have similarly been identified as concerning for their potential to mislead or appeal to consumers.^14,15

The use of greenwashing in tobacco advertisements is concerning as studies suggest that the public misperceives tobacco products that use greenwashing tactics to be safer and less addictive than other products.^16–26 This is problematic, as no brands of conventional cigarettes have been deemed to present modified risk.²⁷ One of the first studies in this area found that the use of the term “natural” to describe cigarettes resulted in consumers misperceiving the cigarettes as less harmful.²¹ Several subsequent studies further explored this relationship, finding, similarly, that greenwashed terms such as “natural,” “additive-free,” and “organic” inaccurately lower perceived product risk.^{16,20,22,23,25}

Such inaccurate reduced risk perceptions can have important behavioral impacts. Previous work found that reduced risk perceptions may contribute to increased use intentions²⁴ and brand switching (as opposed to making a quit attempt) in an attempt to reduce risk.^16,28,29 Such reduced risk perceptions can also affect the way people smoke cigarettes. People who smoke have been shown to increase consumption when using cigarettes labeled “light” or “low-tar,”³⁰ and a recent study found that women inhaled and took more puffs from a cigarette branded as organic compared to a conventional cigarette without such branding.³¹

To try to ensure that tobacco product labeling and advertising does not change users’ or nonusers’ behavior in such harmful ways, the U.S. Family Smoking Prevention and Tobacco Control Act (Tobacco Control Act)³² makes it illegal to market any cigarettes or other tobacco product with false or misleading labeling or advertising and prohibits companies from including explicit or implicit reduced-risk or reduced-exposure claims in their labeling or advertising unless they first obtain an order from the U.S. Food and Drug Administration (FDA) allowing such modified-risk claims as “appropriate for the protection of the public health.” While FDA does not provide an exact definition of what constitutes a reduced-risk or reduced-exposure claim, the agency has described modified-risk claims as conveying that products are “lower risk,” “less harmful,” or “do not contain or are free of a substance.”³³ Accordingly, the FDA released a warning letter in August of 2015 to three companies (ie, ITG Brands LLC, manufacturer of Winston cigarettes with the MRTP claim “additive-free”: Santa Fe Natural Tobacco Company Inc., a subsidiary of Reynolds American, Inc. and manufacturer of Natural American Spirit cigarettes with the MRTP claims “natural” and “additive-free: and Sherman’s 1400 Broadway N.Y.C. Ltd., manufacturer of Nat Sherman cigarettes with the MRTP claim “natural”) who had engaged in such practices, informing them that their use of the terms “additive-free” and “natural” were unauthorized modified risk claims for explicitly or implicitly conveying that the products were “lower risk,” “less harmful,” or “do not contain or are free of a substance.”³³

As this action pertained only to these specific companies’ use of the terms “additive-free” and “natural” in tobacco advertising, broader FDA action is needed to protect consumers from a wider range of misleading greenwashing tactics by any tobacco company.¹⁸ Otherwise, when one specific term or phrase or practice is explicitly prohibited, the tobacco industry can quickly pivot to using others that are not explicitly prohibited, yet are still greenwashing tactics that make potentially illegal claims and mislead consumers and youth in harmful ways. For example, following FDA’s warning, Natural American Spirit began advertising their cigarettes as containing only “Tobacco and Water,” a term which at face value conveys information similar to the term additive-free, and which was found to still mislead consumers to perceive NAS products as less risky than other cigarettes.^12,34

Beyond this example, tobacco companies are already using a wide range of greenwashing tactics beyond the now explicitly restricted greenwashing terms “natural” and “additive-free” to portray their products and brands as eco-friendly.^{11–13,35–40} This includes both textual (eg, the use of language to depict a company or product’s eco-friendliness) and graphic (eg, the use of environmental imagery such as plants and outdoor settings) tactics. These tactics also range from the more explicit (eg, describing how a product contains only tobacco and water; imagery of green, plant-filled tobacco farms) to the more implicit (eg, encouraging consumers not to litter; recycling symbols to depict a company’s commitment to sustainability).^{12,13,35–40}

One study of Natural American Spirit advertising found that the use of such tactics increased after the use of the terms additive-free, natural, and organic was restricted,¹² indicating that the tobacco industry may be using newer, not-explicitly-prohibited greenwashing tactics which could potentially have similar harmful effects on consumer risk perceptions. Although suspected, it has not yet been explicitly established through research that these newer greenwashing tactics are false or misleading or serve as unauthorized implicit reduced-harm or reduced-addiction claims, thereby violating the Tobacco Control Act. To explore whether a broader range of greenwashing tactics currently in use misleads consumers in harmful ways, we conducted a 2 × 2 × 2 × 2 randomized experiment to test the consumer-perception effects of one textual and three image-based greenwashing tactics: (1) Claims about the product’s eco-friendliness, (2) Imagery of a tobacco-leaf, (3) Imagery of a greenwashed pack (ie, pack image featuring leaves), and (4) Imagery depicting that the advertisement was printed on recycled paper. The selection of these tactics was based on our analysis of greenwashing tactics currently used by cigarette companies⁴¹ and on prior research documenting greenwashing tactics.^{10–15,21,38} Specifically, we hypothesized:

H1: Participants who view an ad featuring greenwashed text will have lower perceptions of the product (1) absolute harm, (2) relative harm, (3) absolute addictiveness, (4) relative addictiveness, and (5) relative nicotine content.

H2: Participants who view an ad featuring greenwashed tobacco-leaf imagery will have lower perceptions of the product (1) absolute harm, (2) relative harm, (3) absolute addictiveness, (4) relative addictiveness, and (5) relative nicotine content.

H3: Participants who view an ad featuring imagery of a greenwashed pack will have lower perceptions of the product (1) absolute harm, (2) relative harm, (3) absolute addictiveness, (4) relative addictiveness, and (5) relative nicotine content.

H4: Participants who view an ad featuring recycled paper background imagery will have lower perceptions of the product (1) absolute harm, (2) relative harm, (3) absolute addictiveness, (4) relative addictiveness, and (5) relative nicotine content.

Methods

Sample

We aimed to recruit 1500 participants using a Qualtrics panel. This sample size exceeded Kim and Cappella’s⁴² recommendation for a minimum of 25 ratings per ad while allowing for adequate representation of participants with varying smoking status. Inclusion criteria were that participants be between the ages of 18–29; quotas were established for having never smoked (n = 500), having experimented with smoking (defined as having smoked <100-lifetime cigarettes and now smoking all or some days; n = 500), and having established smoking (defined as having smoked ≥100-lifetime cigarettes and now smoking some or all days; n = 500). We attempted to achieve equal age distribution between 18–20, 21–25, and 26–29 year-olds within each category. This resulted in 1560 responses. Quality control measures included attention control questions within the survey (incorrect answers were exited from the survey), and post hoc review of responses for validation of smoking status (participants answered in the screener, and again in the survey) and age (participants reported age in years in the screener, and date-of-birth in the survey), quality of open-ended responses, and repeated entries. This resulted in an additional 56 cases being excluded, from a final analytic sample of 1504 participants.

Procedures

We conducted a 2 × 2 × 2 × 2 between-persons experiment (Table S1) in which presence–absence of four greenwashing features were randomized across advertisements for two different fictional brands. Participants first provided informed consent, and then answered sociodemographic questions and questions about their tobacco use and perceptions. Participants were then randomized to 1 of the 32 conditions in which they viewed an ad for one of two fictional cigarette brands (Triumph or Victor) in which presence–absence of each condition was manipulated. The advertising text and imagery were adapted from existing cigarette advertisements. Figure S1 displays examples of the experimental ads. Participants could view the ad for as long as they wished and then answered questions assessing their perception of the advertised product. Participants had the ability to return to view the ad when answering questions if they wished. Upon completion of the survey, all participants were debriefed regarding the study and the harms of cigarettes. Study procedures were approved by the Johns Hopkins Bloomberg School of Public Health Institutional Review Board IRB00021941.

Measures

The five outcome measures were (1) perceived absolute harm, (2) perceived relative harm, (3) perceived absolute addictiveness, (4) perceived relative addictiveness, and (5) perceived relative nicotine content. To assess absolute harm and addictiveness, participants were asked how harmful/addictive the product in the ad was, with the following response options: Not at all, slightly, somewhat, very, extremely, and don’t know. Responses were dichotomized to reflect low (not at all or slightly) perceptions of harm and addictiveness vs. all other responses. To assess relative harm and addictiveness perceptions, participants were asked how harmful/addictive the product was compared to other cigarettes, with the following response options: a lot less, a little less, same, a little more, a lot more, and don’t know. Responses were dichotomized to reflect low (a lot or a little less) relative harm and addictiveness perceptions vs. all other response options. Perceived relative nicotine content was assessed by asking participants how much nicotine they thought the product contained, compared to other cigarettes, with the following response options: a lot less, a little less, the same, a little more, a lot more, and don’t know. Responses were similarly dichotomized to reflect low nicotine content perceptions (a lot or a little less) vs. all other response options. The dichotomization of responses was done to better align with the Tobacco Control Act’s language relating to an inaccurate portrayal of a product as modified risk (eg, that a product is presented as being “lower risk,” “less harmful,” “contains a reduced level or a substance,” “is free of a substance”) which does not distinguish between gradients of risk (eg, a lot less harmful vs. a little less harmful).³²

Sociodemographic measures were age, gender, race/ethnicity, and socioeconomic status (current income and income growing up). We also assessed smoking status (nonsmoker: Smoked <100-lifetime cigarettes and does not currently smoke at all; former smoker: Smoked ≥100-lifetime cigarettes and does not currently smoke at all; current experimental smoker: Smoked <100-lifetime cigarettes and currently smokes some or all days; current smoker: Smoked ≥100-lifetime cigarettes and currently smokes some or all days).

Statistical Analysis

Analyses were conducted using Stata 15.⁴³ We first calculated descriptive statistics to report participant characteristics. We used chi-square tests to assess where these characteristics differed significantly across experimental conditions. We then used chi-square tests to examine the bivariate relationship between exposure to each greenwashing feature and the outcomes of interest. Logistic regression analyses were used to test our hypotheses regarding the association between exposure to each greenwashing feature and the outcomes of interest. We first ran models for each outcome containing variables representing presence–absence of each of the four features and the brand of the ad (Victor/Triumph). We then ran models that additionally adjusted for participant gender, race/ethnicity, education, current income, past income, and cigarette smoking status. We also conducted sensitivity analyses in which our outcomes were dichotomized so that “somewhat” harmful/addictive responses were included with “not at all” and “slightly” harmful/addictive, and “don’t know” responses for relative harm, addictiveness and nicotine content were included with “a little less” and “a lot less” harmful/addictive/nicotine.

Results

Table 1 presents participant characteristics. No significant differences in participant characteristics between experimental conditions were found (Table S2).

Table 1.

Sample Characteristics (N = 1504)

	n	%
Age
18–20	222	14.76
21–24	425	28.26
25–29	857	56.98
Gender
Woman	1092	72.61
Man	367	24.40
Non-binary, non-gender-conforming, or another gender identity	44	2.93
Missing	1	0.07
Race/Ethnicity
Non-Hispanic White	881	58.58
Non-Hispanic Black	230	15.29
Non-Hispanic Asian	53	3.52
Non-Hispanic American Indian or Alaska Native	13	0.86
Hispanic or Latino/a/x	142	9.44
Another Non-Hispanic Race/Ethnicity	16	1.06
>1 Race/Ethnicity	164	10.90
Missing	5	0.33
Education
Less than or Some high school	125	8.31
GED (high school equivalency)	68	4.52
High school grad	555	36.90
Some college or associate’s degree	516	34.31
Bachelor’s degree or more	238	15.82
Missing	2	0.13
Current income
Doesn’t meet basic expenses	240	15.96
Just meets basic expenses	511	33.98
Meet needs with a little left or live comfortably	731	48.60
Missing	22	1.46
Income growing up
Poor	447	29.72
It varied	144	9.57
Above average or Pretty well off	907	60.31
Smoking status	6	0.40
Nonsmoker (including never smoker)	340	22.61
Former smoker	31	2.06
Current experimental smoker	444	29.52
Current smoker	687	45.68
Missing	2	0.13
Perceived absolute harm
Somewhat, very, extremely harmful, or don’t know	1361	90.49
Not at all or slightly harmful	142	9.44
Missing	1	0.07
Perceived absolute addictiveness
Somewhat, very, extremely addictive, or don’t know	1396	92.82
Not at all or slightly addictive	107	7.11
Missing	1	0.07
Perceived relative harm
Same, a little more, a lot more harmful, or don’t know	1188	78.99
A little or a lot less harmful	315	20.94
Missing	1	0.07
Perceived relative addictiveness
Same, a little more, a lot more addictive, or don’t know	1322	87.90
A little or a lot less addictive	179	11.90
Missing	3	0.20
Perceived relative nicotine
Same, a little more, a lot more nicotine, or don’t know	1265	84.11
A little or a lot less nicotine	238	15.82
Missing	1	0.07

No significant differences in participant characteristics by experimental condition were found.

Bivariate analyses found several differences in outcomes by presence–absence of ad features (Table 2). Among those who saw an ad with greenwashed text, 11.68% perceived the product to be not at all or slightly harmful (vs. 7.26%), 31.01% perceived the product to be less harmful than other cigarettes (vs. 11.08%), 8.74% viewed the product to be not at all or slightly addictive (vs. 5.53%), 17.25% perceived the product to be less addictive than other cigarettes (vs. 6.72%), and 20.70% perceived the product to contain less nicotine than other cigarettes (vs. 11.07%). Among participants who saw an ad with tobacco-leaf imagery, 23.94% perceived the product as less harmful than other cigarettes (vs. 17.94% of those who did not view an ad with tobacco-leaf imagery), 8.74% perceived the product to be not at all or slightly addictive (vs. 4.58%), 14.32% perceived the product to be less addictive than other cigarettes (vs. 9.50%), and 18.15% perceived the product to have less nicotine than other cigarettes (vs. 13.50%). Bivariate analyses did not find any significant differences in outcomes between those who did/did not view an ad featuring either a recycled background or greenwashed pack.

Table 2.

Perceived Harm, Perceived Addictiveness, and Perceived Nicotine Content by Ad Feature Presence–Absence

		Imagery of recycled background			Tobacco-leaf imagery			Greenwashed text			Greenwashed pack imagery
		No	Yes	sig.	No	Yes	sig.	No	Yes	sig.	No	Yes	sig.
Absolute harm
Somewhat/very/extremely harmful or DK	%	91.5	89.7	0.236	91.7	89.4	0.13	92.7	88.3	0.003	90.9	90.2	0.619
Not at all/slightly harmful	%	8.5	10.3		8.3	10.6		7.3	11.7		9.1	9.8
Relative harm
Same/A little more/A lot more harmful or DK	%	80.0	78.1	0.383	82.1	76.1	0.004	88.9	69.0	<.001	79.9	78.2	0.44
A little/A lot less harmful	%	20.0	21.9		17.9	23.9		11.1	31.0		20.1	21.8
Absolute addictiveness
Somewhat/Very/Extremely addictive or DK	%	94.2	91.6	0.056	94.5	91.3	0.014	94.5	91.3	0.016	91.9	93.8	0.142
Not at all/Slightly addictive	%	5.8	8.4		5.5	8.7		5.5	8.7		8.1	6.2
Relative addictiveness
Same/A little more/A lot more addictive or DK	%	88.3	87.8	0.749	90.5	85.7	0.004	93.3	82.8	<.001	87.7	88.5	0.634
A little/A lot less addictive	%	11.7	12.2		9.5	14.3		6.7	17.3		12.3	11.5
Relative nicotine
Same/A little more/A lot more nicotine or DK	%	83.3	85.0	0.386	86.5	81.9	0.014	88.9	79.3	<.001	84.3	84.0	0.885
A little/A lot less nicotine	%	16.7	15.0		13.5	18.2		11.1	20.7		15.7	16.0

Values significant at p < .05 are bolded.

Logistic regression analyses in which presence–absence of all ad features were included in the same model were then conducted. Table 3 reports the results of unadjusted and adjusted logistic regression analyses; results from the adjusted models are reported in text. These analyses found that the presence of tobacco-leaf imagery and the presence of greenwashed text were associated with lower product risk perceptions. Participants who viewed an ad with a tobacco-leaf image had 1.55(95% CI[1.19, 2.04]) greater odds of perceiving the product as a little or a lot less harmful than other cigarettes, 1.72(95% CI[1.13, 2.63]) greater odds of perceiving the product as not at all or slightly addictive, 1.60(95% CI[1.14, 2.23]) greater odds of perceiving the product as a little or a lot less addictive than other cigarettes, and 1.48(95% CI[1.10, 1.97]) greater odds of perceiving the product as containing a little or a lot less nicotine than other cigarettes. Participants who viewed an ad with greenwashed text had 1.72(95% CI[1.19, 2.49]) greater odds of perceiving the product as not at all or slightly harmful, 3.92(95% CI[2.94, 5.22]) times greater odds of perceiving the product as less harmful than other cigarettes, 2.93(95% CI[2.06, 4.16]) times greater odds of perceiving the product as a little or a lot less addictive than other cigarettes, and 2.08(95% CI[1.55, 2.79]) greater odds of perceiving the product as having less nicotine than other products. The presence of recycled background and greenwashed pack were not significantly associated with any outcome.

Table 3.

Odds of Perceiving Lower Harm, Lower Addictiveness and Lower Nicotine Content, by Ad Feature

	Absolute harm—not at all/slightly harmful			Relative harm—a little/a lot less harmful			Absolute addictiveness—not at all/slightly addictive			Relative addictiveness—a little/a lot less addictive			Relative nicotine—a little/a lot less nicotine
	AOR	95% CI	p	AOR	95% CI	p	AOR	95% CI	p	AOR	95% CI	p	AOR	95% CI	p
Model 1 (a)
Recycled background (ref:No recycled background)	1.23	[0.87 to 1.75]	.237	1.12	[0.87 to 1.45]	.384	1.46	[0.97 to 2.18]	.068	1.04	[0.76 to 1.43]	.806	0.87	[0.66 to 1.16]	.349
Tobacco-leaf image (ref:No imagery)	1.31	[0.92 to 1.85]	.133	1.48	[1.14 to 1.92]	.003	1.64	[1.09 to 2.45]	.017	1.62	[1.17 to 2.23]	.004	1.44	[1.09 to 1.91]	.011
Greenwashed text (ref:Non-greenwashed text)	1.70	[1.19 to 2.42]	.003	3.67	[2.78 to 4.84]	<.001	1.64	[1.09 to 2.45]	.017	2.93	[2.08 to 4.13]	<.001	2.11	[1.58 to 2.82]	<.001
Greenwashed pack (ref:Non-greenwashed pack)	1.11	[0.78 to 1.57]	.568	1.14	[0.88 to 1.47]	.333	0.76	[0.51 to 1.13]	.168	0.94	[0.69 to 1.30]	.720	1.03	[0.78 to 1.36]	.831
Ad brand (ref:Triumph)	1.05	[0.75 to 1.49]	.763	0.78	[0.60 to 1.01]	.057	0.98	[0.66 to 1.45]	.913	0.71	[0.52 to 0.98]	.036	0.82	[0.62 to 1.09]	.168
Model 2 (b)
Recycled background (ref:No recycled background)	1.25	[0.87 to 1.79]	.231	1.16	[0.89 to 1.51]	.273	1.49	[0.98 to 2.26]	.059	1.07	[0.77 to 1.48]	.699	0.88	[0.66 to 1.17]	.376
Tobacco-leaf image (ref:No imagery)	1.31	[0.91 to 1.88]	.144	1.55	[1.19 to 2.04]	.001	1.72	[1.13 to 2.63]	.011	1.60	[1.14 to 2.23]	.006	1.48	[1.10 to 1.97]	.009
Greenwashed text (ref:Non-greenwashed text)	1.72	[1.19 to 2.49]	.004	3.92	[2.94 to 5.22]	<.001	1.44	[0.95 to 2.18]	.087	2.93	[2.06 to 4.16]	<.001	2.08	[1.55 to 2.79]	<.001
Greenwashed pack (ref:Non-greenwashed pack)	1.05	[0.73 to 1.50]	.803	1.13	[0.87 to 1.48]	.363	0.67	[0.45 to 1.02]	.061	0.92	[0.67 to 1.28]	.628	0.99	[0.74 to 1.32]	.960
Ad brand (ref:Triumph)	1.10	[0.77 to 1.58]	.599	0.81	[0.62 to 1.06]	.123	1.01	[0.67 to 1.52]	.964	0.70	[0.51 to 0.98]	.036	0.79	[0.60 to 1.06]	.118

Values significant at p < .05 are bolded.

(a) Model 1 contained brand of ad and ad features (entered together).

(b) Model 2 contained brand of ad, ad features (entered together), and participant gender, race/ethnicity, education, current income, past income, and cigarette smoking status.

We conducted sensitivity checks (Table S3) in which our outcomes were dichotomized so that “somewhat” harmful/addictive responses were included with “not at all” and “slightly” harmful/addictive, and “don’t know” responses for relative harm, addictiveness, and nicotine content were included with “a little less” and “a lot less” harmful/addictive/nicotine. Results were largely the same, with the following differences: Greenwashed imagery was associated with absolute harm, and greenwashed text was associated with absolute addictiveness.

To further illustrate the contribution of greenwashing to shifts in consumer risk perception, we directly compared participants who saw an ad with no greenwashing (for either brand) to those who saw an ad with all greenwashed features. Figure 1 illustrates differences in risk perceptions between these two ads. Chi-square tests (Table S4) revealed that those who viewed an ad with all greenwashing features had significantly different patterns of all risk perceptions than those who viewed an ad with no greenwashing features.

Figure 1.

Differences in risk perceptions between ad with all greenwashing features and no greenwashing features.

Discussion

These findings illustrate the contribution of greenwashing—namely, tobacco-leaf imagery and eco-friendly language—to increasing inaccurate consumer misperceptions of reduced cigarette harm, addictiveness, and nicotine content, which can dampen cessation and increase initiation. Specifically, participants who viewed an ad featuring tobacco-leaf imagery had lower ratings of relative harm, relative addictiveness, absolute addictiveness, and relative nicotine content, while those who viewed an ad featuring greenwashed text had lower ratings of absolute harm, relative harm, relative addictiveness, and relative nicotine content. These findings are particularly concerning given the current prevalence of these tactics: recent analyses have found that claims about a product being eco-friendly (found in 27% of greenwashed ads) and imagery of flora (found in 32% of greenwashed ads) are commonly used greenwashing tactics.⁴¹

Prior research^{16–18,20,22–24} has already demonstrated the effects of other specific forms of greenwashing (eg, the terms “additive-free,” “natural”) on risk perceptions. While the FDA took action against some of those specific tactics, the tobacco industry continues to use a broader spectrum of other strategies to promote their products by portraying them as natural and eco-friendly. Moreover, most related research has focused exclusively on the labeling and advertising of Natural American Spirit cigarettes. While this brand commonly uses misleading greenwashing in their marketing,^11–13 other brands, including Marlboro, Winston, L&M, Nat Sherman, and Manitou, also engage in such greenwashing,^{9,13,36,40,41} indicating that the problem extends well beyond a single brand.

Added to prior research, the findings here indicate that restricting the tobacco industry from using greenwashed text and imagery in their advertising could help prevent and reduce inaccurate consumer risk perceptions that can increase tobacco use and harm. By increasing inaccurate risk perceptions, these tactics appear to violate the TCA’s ban on false or misleading tobacco product labeling or advertising and its requirement that tobacco product labeling not include any explicit or implicit reduced-risk claims without prior FDA authorization. Moreover, the industry is known to deploy questionable new advertising tactics when their earlier tactics prompt enforcement action or penalties or have otherwise been explicitly prohibited ^12,44,45 and these new tactics can continue to mislead consumers.^25,34 Accordingly, trying to stop misleading greenwashing tactics one at a time, as they appear, would be an inefficient, piecemeal approach, resulting in a perpetual game of catch-up to an ever-evolving industry. There are multitudes of words, phrases, and imagery that could convey the same meaning as a restricted tactic. For instance, research found that the use of the term “tobacco & water” as a replacement for “additive-free” still conveyed inaccurate reduced harm perceptions.³⁴ It might therefore not be possible for FDA to provide a comprehensive list of such words, phrases, and imagery, just as it would not be possible to provide a comprehensive list of such items that constitute defamation or threats of harm in legal cases. The FDA may therefore need to rely on regulatory and judicial precedent to make such determinations. Even with increased enforcement, new, more explicit restrictions on greenwashing tactics might not adequately protect consumers and public health against the industry’s use of misleading labels, advertising, and unauthorized and inaccurate implicit reduced-risk claims. More comprehensive and effective approaches to protect consumers and public health could include also requiring plain packaging (which would also limit the use of color on packs), prohibiting the use of imagery in tobacco advertising unless necessary for the industry to convey relevant, accurate product information to consumers (along with first Amendment compatible graphic warning labels to provide consumers with accurate harm and risk information⁴⁶), and additional measures to reduce youth exposure to tobacco products and their advertising.

Strengths of this study include its randomized design across ads for two novel brands, as well as the incorporation of advertising imagery and language used in real cigarette ads. The observation that risk perceptions did not depend on fictional brands moreover speaks to the robustness of the findings. This study additionally focused on young adults, a population at increased risk for smoking initiation and escalation.^47,48 Limitations of the study include the use of an online sample that was heavily female. The gender composition of the sample was unexpected, and future work should employ quotas to ensure balance by gender. Additionally, the tightly controlled design may not reflect real-world exposure to tobacco advertisements, in which individuals may be exposed to advertisements multiple times, and in different contexts. Future work may wish to replicate findings with different samples and in different settings. It is possible that the greenwashing manipulations used in this study were not clear to participants. For instance, it is possible the recycled paper background condition was not clearly discernable to participants (particularly compared to other work (eg, Lee et al.,¹⁵)) in which the recycled background was darker and clearly distinct from standard white paper. It is also possible that the greenwashing on the cigarette pack image may have been too subtle to be notable to participants. Interpretation of these findings should be balanced against other work which demonstrated the ability of greenwashing on cigarette packs to induce inaccurate reduced risk perceptions.^{17,23,24,49–51} Finally, we did not assess the mechanisms through which greenwashing led to inaccurate risk perceptions. For example, it is possible that the term “Tobacco and Water” led participants to believe the product was less risky because it did not contain any additives or was not artificial (as opposed to believing that the product was eco-friendly).

Greenwashing continues to be an insidious way for tobacco companies to make inaccurate and misleading modified risk claims, despite applicable U.S. Family Smoking Prevention and Tobacco Control Act prohibitions. More active FDA enforcement against such greenwashing and new FDA rulemaking to prohibit unnecessary imagery in tobacco advertising and establish plain packaging requirements would help protect consumers and public health.

Supplementary Material

Supplementary material is available at Nicotine and Tobacco Research online.

Funding

Research reported in this publication was supported by the National Institute on Drug Abuse at the National Institutes of Health (NIH) and the U.S. Food and Drug Administration Center for Tobacco Products (CTP; R01DA049814). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration.

Declaration of Interest

Dr. Moran served as a paid expert witness in litigation sponsored by the Public Health Advocacy Institute against RJ Reynolds. This arrangement has been reviewed and approved by Johns Hopkins University in accordance with its conflict of interest policies. Dr. Pearson receives grant funding from NIH/FDA. Dr. Pearson is also a consultant to Westat on the Population Assessment of Tobacco and Health (PATH) Study. Dr. Pearson is an expert witness for the Plaintiffs in a Multi-District Litigation invoking Natural American Spirit cigarettes. Eric Lindblom has served and is serving as an expert witness in litigation against cigarette companies. However, none of the lawsuits have pertained to American Spirit cigarettes or greenwashing issues.

Author Contributions

Meghan Moran (Conceptualization [lead], Formal analysis [lead], Funding acquisition [lead], Investigation [equal], Methodology [equal], Project administration [equal], Software [equal], Supervision [lead], Writing—original draft [lead], Writing—review & editing [equal]), Maryam Ibrahim (Methodology [supporting], Project administration [equal], Software [equal], Writing—original draft [equal], Writing—review & editing [equal]), Lauren Czaplicki (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Jennifer Pearson (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Johannes Thrul (Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Eric Lindblom (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [equal]), Shae Robinson-Mosley (Visualization [lead], Writing—review & editing [supporting]), Ryan David Kennedy (Conceptualization [equal], Investigation [equal], Methodology [equal], Writing—review & editing [supporting]), Ariel Balaban (Resources [lead], Writing—review & editing [supporting]), and Matthew Johnson (Conceptualization [lead], Funding acquisition [lead], Investigation [equal], Methodology [equal], Project administration [equal], Writing—review & editing [equal])

Data Availability

Data are available upon reasonable request to the authors.