U.S. consumer support for genetically modified foods: Time trends and assessments of four GM attributes

ABSTRACT

There is a large literature about consumer acceptance of GM foods dating back almost three decades, but there are fewer studies that investigate how support for specific GM attributes contribute to general support for novel plant technologies. In addition, there is little information on how support has changed over time. Using survey data from 2018 to 2023 in a U.S. State (Vermont) (n = 3101), we analyze changes in support for a variety of GM attributes over time. There are three major findings. First, there is movement toward neutrality in support for various GM attributes, but opposition continues. Second, there is variability in support for different GM attributes. People are most supportive (least opposed) to GM attributes that improve flora (plant health or drought tolerance), and most opposed (least supportive) of attributes that impact fauna (specifically fish). Third, multivariate regression reveals that assessments of individual GM attributes contribute to levels of overall support of the use of GM technologies in agricultural production.

KEYWORDS:

• Attitude
• consumer behavior
• food labels
• genetic modification
• GM
• GMO
• novel technologies
• support

1. Introduction

Many consumers remain more opposed than supportive of the use of novel food production technologies, which is perplexing to some and telling to others.^{1,2 1} One thread of literature refers to people as not able or willing to understand science (see, for example).^3,4 This literature concludes hesitancy toward genetically modified (GM) foods is unfounded because these foods are safe. Another thread uses a bioethics approach asserting there are many reasons people are opposed to GM foods that are unrelated to the safety or health implications, including autonomy to make decisions (see, for example⁵). This literature concludes a more precautionary approach should be taken to introducing GM foods into the marketplace and points to the need for increased information for consumers. Lancaster’s⁶ theory that goods “are not just goods,” and it is the intrinsic and extrinsic attributes they possess that yield consumer satisfaction, is useful to employ when considering GM production method. Individual attributes may contribute heterogeneously to satisfying consumer preferences.

Knowledge of attributes alone may not be the key to rejection or acceptance of novel foods.^7–9 The study of consumer behavior recognizes the importance of affect on choice (see, for example).¹⁰Affect refers to feelings, including liking and disliking a product. Benefits, both perceived and real, are derived from these attributes).^11,12 The opposite can also be true. Risks, both perceived and real are also important¹³M. Costa-Font.^14–17 Benefits and risks include externalities that go beyond the product itself, such as environmental gains or losses from the use of GM technologies in food production.^18,19 People’s support for GM is one way to identify consumer preferences.

In this applied study using data from 2018 to 2023 and a sample of 3101 respondents from the U.S. state of Vermont, we analyze changes in support for a variety of GM attributes over time. We include data on overall support for the use of GM technologies in food production, and for four specific GM attributes: crop resistance to pests and/or weeds, resistance to drought and/or virus conditions, resistance to browning in produce (apples, potatoes), and superior growth in fish (Aquabounty salmon).

1.1 Consumer Support for GM Technologies

In general, people have expressed limited support for GM technologies since the introduction of the Flvr Savr and Zeneca tomatoes in 1994 and 1996, with levels of support varying by study.^2,20 Words associated with viewpoints include not only safety but also interference with the natural order of things, freedom of choice, environmental concern, trust, risk, cultural identity, equity, fairness, consent, autonomy, and food neophobia.^9,17,21–35 Limited support of early GM technologies has been associated with benefits not accruing to the final consumer (e.g., farmers not needing to apply pesticides), and it has been noted that or the successful introduction of such innovations, societal acceptance needs to be addressed at early stages of the technology’s development.^9,27

1.2 Applied Economic Approaches to Identifying Preferences for Novel Technologies

We focus our review on the U.S. context, recognizing the vastness of the GM literature that includes an international perspective (See, for example^14,36,37). It has been noted that the U.S. experience is different from that of other countries.^38,39 We also focus on the applied economic approaches to identifying preferences, recognizing there are many theories and methods used to explain acceptance of GM (See, for example^{8,15,18,29,32,40,41}).

Experimental methods that analyze stated preferences measured by willingness to pay (WTP) conclude that consumers prefer non-GM food alternatives, implying a general opposition to GM technologies. The experimental literature focused on GM labeling prior to 2008 has been summarized by others.^38,39 Lusk et al.,³⁹ conducted a meta-analysis of 25 studies of 57 valuations for GM food and included studies up to 2003. The results include the international context, but a majority of studies included U.S. respondents. Across studies, the authors calculated a 23% weighted average higher WTP for non-GM food compared to GM food. Dannenberg³⁸ conducted a meta-analysis of 51 studies between 1992 and 2007. U.S respondents were less WTP a non-GM premium compared to international counterparts, although a non-GM) premium was found in the majority of cases. Premiums were lower when the GM food provided consumer benefits (taste, nutrition) and when the GM technology was plant, not animal related. More recent studies conclude that research participants are WTP a premium between 1% and 21% for non-GMO foods over that of the GMO equivalent. Between 32% and 53% of consumers are not willing to pay any premium. When provided with information about benefits of GM, WTP for GM food increases.^42–47 Overall, WTP for non-GM foods has decreased over time, indicating a movement toward more support for GM foods.

Lancaster’s⁶ utility maximization model brought attention to product attributes as contributing to consumer preference. We can also include product benefits that the attributes convey as important components of selling brands and building brand loyalty.^11,48 Further, some researchers have considered consumer values and intuition about GM technologies and have shown that these concepts matter.^9,19,49–51 There has been advancement in the study of consumer support for GM technologies, including a movement from overall judgments to acceptance of particular GM techniques.^52–55 Some studies have included individual attributes of GM.^35,56,57 A limited number included changes over time, and the need for such studies has been mentioned in the literature since at least 2012.^58,59

The economics of information has a place in the examination of consumer preferences for GM technology benefits. On July 29, 2016 United States President Obama signed the National Bioengineered Food Disclosure Standard (NBFDS) into law (Public Law No. 114–216) which directed the United States Department of Agriculture (USDA) to establish a national standard to disclose certain food products that are “bioengineered,” making the United States the 65^th country requiring mandatory labeling of products containing genetically modified (GM) ingredients).^60–62 The NBFDS was implemented in 2020, and mandatory overall compliance went into effect (for both small and large manufacturers) on January 1, 2022.^63;64 The mandatory labels joined voluntary labeling of foods not derived from GM plants or GM salmon, for which Federal guidance was introduced in 2015 and updated in 2019 (Draft Guidance for Industry: Voluntary Labeling Indicating Whether Food Has or Has Not Been Derived From Genetically Engineered Atlantic Salmon).^65,65 For example, the Non-GMO project has more than 62,000 certified food products carrying its label.⁶⁶

While there was much discussion in the agricultural economics arena about labeling increasing opposition to the use of GM technology in food production, one study found that after labeling was in place, consumers were less opposed.^58,67–69 Others have found that increased familiarity over time can lead to positive consumer habituation and thus less opposition.^70–72 Our study examines changes in support over time for four GM attributes and overall and investigates how individual GM benefits are associated with overall support for the use of GM in food production.

2. Hypotheses

We test two hypotheses in this study of overall support for GM technologies and support for individual benefits provided by GM of food. The first hypothesis is based on several studies described above that show consumer WTP for non-GM foods has decreased relative to GM foods, indicating a shift in preferences toward GM and the assertion that habituation without dire consequences decreases risk aversion.^72,73 Since their introduction decades ago with the FlavrSavr tomato, there have been no widespread indications that GM food is dangerous to humans and limited evidence that the environment can be harmed directly by GM crops (see, for example,^{20 74,} .⁷⁵ There is evidence that herbicides sprayed on GM crops can be harmful to the environment and human health, but the modified food is not harmful.^76,77

H1:

Support for the use of GM technologies in food production has increased.

The second hypothesis is based on both Lancaster’s theory that product attributes and benefits lead to satisfaction and accounting for more autonomy in the marketplace with the diffusion of labeling may contribute to increased support.^{27,32,58,78–80} We investigate how consumers perceive benefits derived from GM technologies contribute to increases in overall preferences for the use of GM in food production.

H2:

Support for the benefits provided by GM technologies used in food production are associated with overall support for the use of GM technologies in food production.

3. Materials and Methods

We use a series of cross-sectional survey data for the years 2018–2023 from 3101 Vermont, U.S. respondents.² All surveys received institutional review board approval. Vermont is the one state that experienced mandatory labeling twice before the National Bioengineered Food Disclosure Standard (NBFDS) was signed into law in 2016 and can be considered a bellwether to investigate trends in consumer support/opposition for GM attributes. Vermont’s experience has been used or referred to in several other studies.^{58,67,68,68,79,81–84,84–86}

We utilize uni-, bi-, and multivariate empirical methods. First, we describe the sample, including the passage of time, and socio-economic characteristics of the sample, and the variables GM support, measured using four different GM attributes and overall. See Table 1 for details.

Table 1. Sample descriptive statistics.

Variable	Definition	2018	2019	2020	2021	2022	2023	Total
Kiddum	Presence of children under age 18	18.0%	20.3%	26.5%	10.9%	16.0%	16.2%	18.2%
Inc75	Household income > $75K	43.1%	40.3%	43.3%	58.9%	51.6%	50.5%	47.4%
Female	Identify as female	59.4%	64.7%	53.5%	50.8%	52.8%	50.1%	55.3%
College	Completed associates degree or higher	79.4%	80.5%	82.1%	91.9%	76.5%	75.0%	79.4%
Data year	observations	510	538	430	248	702	673	3101

Overall support for the use of GM in food production was measured on a five point, 0 to 4 scale, with 0 = strongly opposed, 1 = opposed, 2 = no opinion/neutral, 3 = supportive, and 4 = strongly supportive. The same ordinal scale was used for variables that measured support for four GM attributes: crop resistance to pests and/or weeds, crop resistance to drought and/or virus conditions, resistance to browning in produce (apples, potatoes), and superior growth in fish (Aquabounty salmon). See Fig. 1 for a graphical representation of support over time. Demographic variables include gender of the respondent (Female = 1 if identify as female and 0 otherwise), income of respondent household (Inc75 = 1 if income is greater than $75K and 0 otherwise), presence of children under age 18 in the household (Kids = 1 if children under the age of 18 are present and 0 otherwise), completing an associates degree or higher (Coll = 1 if completed and 0 otherwise), and the passage of time with 2018 as the base year and 2019, 2020, 2021, 2022 and 2023 (yes for particular year = 1 and 0 otherwise). Measures of demographics were chosen to ensure comparability over the five year period as data were collected as five cross sectional studies, with varying categorizations of the demographic information.

Figure 1. Unadjusted support for novel technologies.

3.1 Statistical Methods

MANCOVA is the appropriate statistical test to test H1, as it accounts for the correlations between overall support for GM and support for the four more specific attributes of GM technologies (See Table 2) and we use it as a robustness check on the use of the more basic ANOVA. Conducting five individual ANOVA analyses one at a time on each GM attribute over time and on overall support is flawed, resulting in inflation of Type I error (rejecting a null hypothesis when the null is true). The statistical problems related to multiple testing are well documented.⁸⁷ MANCOVA has several requirements and assumptions and thus limitations. While there is controversy about treating ordinal data as interval, researchers have analyzed GM support variables as interval measures.^58,88–92 We point out that ordinal probit regression, which we use to answer our second research question, relies on an underlying interval scale.⁹³ MANCOVA is robust with regard to the assumptions of violations of normality, and statistical packages readily address violations of homoskedasticity, identified by Box’s test, using Pillai’s test for significance.⁹⁴ MANCOVA also requires there be no significant interaction effects among the control demographic variables. Our choice of demographic characteristics is limited by this restriction, and the included variables do represent demographic characteristics included in other studies: income, gender, presence of children under age 18 in the household, and educational level. There were no significant differences in the unadjusted and predicted adjusted means using MANCOVA and ANOVA. A graphic representation of adjusted means is shown in Fig. 2

Figure 2. Adjusted support for novel technologies.

Table 2. Pearson correlation coefficients – GM attributes.

	To protect against drought/virus	To repel pests/weeds	To prevent browning	To support superior fish growth	Support for GM use in food production
To protect against drought/virus	1	.719**	.648**	.497**	.646**
To repel pests/weeds	.719**	1	.666**	.547**	.618**
To prevent browning	.648**	.666**	1	.641**	.616**
To support superior fish growth	.497**	.547**	.641**	1	.556**
Support for GM use in food production	.646**	.618**	.616**	.556**	1

**Sig. ≤ .01 N=3103.

Therefore, we present the more simple, but comparable ANOVA results as Table 3.^95,96

Table 3. Comparison of support for GM over time.

		Overall	2018	2019	2020	2021	2022	2023	F-test
GMO support	Overall support for the use of GM in food production	1.44 (1.23)	1.20^b,d,e,f (1.24)	1.19^a,d,e,f (1.18)	1.42 (1.27)	1.70^a,b (1.17)	1.55^a,b (1.17)	1.62^a,b (1.27)	14.969***
Virus	Support for technologies that increase resistance to viruses or drought	2.12 (1.36)	1.79^c,d,e,f (1.43)	1.79^c,d,e,f (1.39)	2.13^a,b,c (1.44)	2.45^a,b,c (1.23)	2.28^a,b,c (1.27)	2.32^a,b (1.28)	20.823***
Herbinsect	Support for technologies that repel weeds or insects	1.58 (1.36)	1.13^c,d,ef (1.19)	1.24^b,d,e,f (1.24)	1.77^a,b (1.50)	1.97^a,b (1.37)	1.74^a,b (1.34)	1.78^a,b (1.38)	29.531***
Browning	Support for technologies that reduce browning	1.36^d,e,f (1.26)	1.16^c,d,e,f (1.26)	1.13^b (1.20)	1.39^a,b (1.39)	1.62^a (1.21)	1.50^a,b (1.21)	1.43^a,b (1.24)	10.545***
Growth	Support for technologies that support superior animal growth	0.94 (1.12)	0.88 (1.16)	0.90 (1.12)	0.96 (1.20)	1.03 (1.09)	0.96 (1.05)	0.94 (1.13)	0.85
N		3101	510	538	430	248	702	673

a=Sig. different than 2018 p≤-01; b=Sig. different than 2019 p≤.01; c=Sig. different than 2020 p≤.01; d= Sig. different than 2021 p≤.01; e=Sig. different than 2022 p≤.01; f= Sig. different than 2023 p≤.01.

To answer H2, we compare ordinary least squares regression (OLS) and ordinal Probit. Overall support for the use of GM technologies is the dependent variable and support for individual GM attributes are the independent variables, controlling for the same demographic characteristics used in the ANOVA analysis described above. As with the bivariate analyses, the more simple OLS regression results are comparable to the more statistically complicated Probit analysis.

4.0 Results

Table 1 provides descriptive statistics of the sample, including our control demographic variables. Figure 1 shows graphically support over time for overall use of GM food production and for each of the four GM attributes over the six-year period. Figure 2 shows mean support after controlling for demographic characteristics. Table 3 shows significant differences over time in support. Table 4 presents results of the regression analyses showing the part worth contribution of support/opposition for each GM trait on overall support/opposition for the use of GM technologies in food production. Our results are robust to differences in specification as described above.

Table 4. Contribution of GM attributes to overall support for GM in food production.

GM Attribute (continuous)	OLS	GM Attribute (ordinal)	Ordinal Probit
To protect against drought/virus	.28 (.017)***	To protect against drought/virus
		Strongly oppose	reference
		Oppose	.94 (.14)***
		No opinion	1.01 (.13)***
		Support	1.51 (.13)***
		Strongly support	2.08 (.14)***
To repel pests/weeds	.15 (.018)***	To repel pests/weeds
		Strongly oppose	reference
		Oppose	.26 (.10)***
		No opinion	.47 (.11)***
		Support	.48 (.11)***
		Strongly support	.78 (.13)***
To prevent browning	.17 (.019)***	To prevent browning
		Strongly oppose	reference
		Oppose	.12(.09)
		No opinion	.29 (.09)***
		Support	.50 (.10)***
		Strongly support	.71 (.17)***
To support superior fish growth	.21 (.02)***	To support superior fish growth
		Strongly oppose	reference
		Oppose	39 (.06***
		No opinion	.64 (.09)***
		Support	.65 (.08)***
		Strongly support (reference)	1.10 (.16)***
year2019	−.02 (.05)		−.00 (.07)
year2020	−.04 (.06)		−.02 (.08)
year2021	.08 (.07)		.13 (.09)
year2022	.05 (.05)		.10 (.07)
year2023	.11 (.05)**		.17 (.07)**
Constant	0.21 (.05)***
Actual mean	1.44 (1.23)
Predicted Mean	1.52 (1.33)
Predicted Category (Mean)			1.33 (1.19)
Sig. p ≤.05 ** Sig. p ≤.01***	Adj. R^2 = .52	n=3101	McFadden R2 =.25

Control variables: education, gender, income, presence of children.

ANOVA analyses show a movement away from opposition for all technologies except “to support superior fish growth” between 2018 and 2023. All differences are significant at p ≤.01 (Table 3). Mean support overall (scale from 0 to 4) for “support for the use of GM technologies in food production” increased 35% increase between 2018 (1.20/4) to 2023 (1.62/4) (F = 14.69; p ≤ .01). Mean support for “to reduce drought/virus” increased by 29.6% between 2018 (1.79/4) and 2023 (2.32/4) F = 20.82; p ≤ .01). This specific GM characteristic exhibits highest overall level of support compared with other GM attributes and support for GM technologies overall. Mean support for “to repel pests/weeds” increased by 57.5% between 2018 (1.13/4) and 2023 (1.79/4) and represents the highest percent increase in support over time(F = 29.53; p = ≤ .01). Mean support for “to reduce browning” increased 23.2% between 2018 (1.16/4) and 2023 (1.43) (F = 10.54; p ≤ .01). Mean support for “to support superior fish growth” did not change significantly over the time periods and had the lowest support of all the GM attributes included (F = .89, p > .05). While support for three of the four included GM attributes and overall support “for the use of GM in food production” have improved, they remain below “neither support nor oppose” for “to repel pests/weeds,” “to prevent browning,” “to support superior fish growth,” and overall “support for the use of GM technologies in food production.” Support for “to prevent drought/virus” has positive support in 2023 (M = 2.32/4), between neutral and somewhat supportive categories.

The general answer to the question, has support for individual GM technologies and overall changed over time is yes. There has been movement toward being more supportive of GM, peaking in Covid year 2021 and moderating, but still higher than 2018, which had the lowest reported support for all measured years. Consumer support for the four GM attributes and overall has increased over time with the exception of the GM attribute “to support superior growth in fish.”

To answer our second research question, we estimate the contribution of each GM attribute to overall support for the use of GM in food production as put forth by Lancaster.⁶ Results are presented in Table 3. Regardless of whether the equation specified overall support for the use of GM in food production and each of the GM attributes as interval or ordinal variables, the individual GM attributes all significantly contribute to overall support. While both specifications had similar explanatory power, the OLS specification more closely predicted actual support compared with the ordinal specification. Therefore, we focus on the OLS results. GM modification to “resist drought/virus” had the largest contribution to overall support, .28 of a scale point for a one unit increase in support (p ≤ .01), while “to repel pests/weeds” had the lowest contribution, .15 of a scale point (p ≤ .01). The contributions of “to prevent browning” and “to support superior fish growth” were between the others, .17 and .21 of a scale point (p ≤ .01). Overall, we can say that a one unit change in support for any one of the benefits is associated with approximately a fifth of a point (.20) on a scale of 0 to 4.

5. Discussion

Overall support for the use of GM in food production and for each of the GM attributes included in this study increased by between 23% and 54% between 2018 and 2023, except for the GM attribute “to support superior growth in fish,” which did not change significantly and has the lowest level of support of all GM attributes. Findings that people are less supportive of GM used in fauna is documented.^{35,38,40,97,98} We found lower levels of opposition over time for GM attributes that describe crop health and have implications for increased yields and thus food production and food security. Others have found increased support for attributes that are beneficial for humans and the environment.⁹⁹ Support for GM that repels pests and weeds, increased by the highest percentage, to become the GM attribute with the second highest level of support, trailing behind the most supported “resist drought/virus.” These attributes have been available longest and are found for corn and soy, of which over 90% are GM, and sugar beets, of which close to100% are GM.¹⁰⁰ People have experienced products that contain ingredients made from these crops without individual harm. Increased familiarity can lead to positive consumer habituation and thus less opposition.^70–72 However, the significance of time occurred only in 2023, compared to 2018. Support was highest in 2021 and then moderated. This could have been due to the COVID pandemic. Shoppers increasingly used home or curbside delivery and did not have the opportunity to read labels while shopping.¹⁰¹ Support is lowest for a reduction in browning and superior growth in fish. It may be that people consider these as superfluous and perhaps unnecessary GM foods. On the other hand, Simplot potatoes are produced using newer gene editing methods (CRISPR-Cas9), and there is a growing discussion that consumers may be more accepting of these newer gene editing techniques, compared with first generation genetic modification.^102–104 Shew et al.¹⁰⁴ referred to gene editing as not GMO. Our finding that people remain opposed to the attribute of reduced browning does not support these other findings about the process of the modification. While it has been asserted reducing browning can reduce food waste, findings by Lusk et al.,⁵⁷ found this attribute at the middle of a list of benefits people seek in their food choices. People may not make the distinction between the semantics of genetic modification versus gene editing.⁵² Our study did not focus on GM method, but on the attribute resulting from the modification. It appears that regardless of the how a trait is introduced, people continue to be less accepting of fauna modification compared to flora modification.⁹⁷

What do these results mean in the marketplace? Support for GM attributes is heterogeneous, and although respondent support for these benefits contribute to overall levels of support for the use of novel food technologies in food production, support overall for GM remains low. As Helliwell et al.¹⁰⁵ point out, it may be unreasonable for producers to expect consumers to differentiate between types or methods of GM. This points to taking a step back in making any specific, targeted recommendations for information dissemination, as was suggested by Sleboda & Lagerkvist.¹⁰⁶ If messaging is created for different GM attributes, information overload may lead to increased confusion over risks, benefits, reasons to support, reasons to support, etc. For example, as new technologies are found to produce that people care about, including plant health and a secure food supply, specific communication about these may improve overall support. But, if communication about new technologies with animal applications increases, overall support may decline. The finding that there can be “leakage” from acceptance of a particular novel technology to another further supports this.¹⁸ Information overload is real.¹⁰⁷ One thing is certain, mandatory (NBFDS) labels are now required and voluntary certified labeling (Non-GMO Project) is increasing. Both labels provide simple information that a GM method was used on an ingredient contained in the food being purchased if it is a multi-ingredifent product or to modify an attribute in a single ingredient food, but no specifics on attributes nor method of the GM. No label can provide all information.¹⁰⁸ With both mandatory and voluntary labels in place people can find and use information that meets their preferences for GM food overall. There is emerging evidence that about 50% of consumers will see labels, one source of information, and for a third of those the label will not influence their purchase decision. For another third, the label will reveal preferences for non GM products.⁶⁸ Other studies have shown that signaling impacts of “contains” GM labels do not reduce support for these technologies.^68,86,109 Others found labels reinforce support.⁵⁸ As with the introduction of any new product or service, future research must continue to track consumer preferences and demand as the market decides which products produced with novel food technologies consumers are most willing to purchase. Placing our results in the GM literature space leads us to conclude that providing factual, simple information about GM use in product formation and letting the market decide over the course of time, without overburdening consumers with specific information about the way a benefit was introduced may be the best approach. Why? Our data show a trend toward acceptance, especially when there is a consumer or societal benefit. This supports the idea of habituation over time, albeit slowly.

Given the higher level oversight of Federally regulated or voluntary non GMO Project labels, it is unlikely that the informational landscape will result in sub-requirements that include type of modification nor product attribute. That said, it is likely that marketers will fill in the gaps if consumers ultimately make choices based on attribute, type of technology or both. It is likely regulators and certifiers will watch the landscape and determine whether regulations require adjustment.

This study has limitations. It was conducted in a single U.S. state and the sample was more highly educated compared to the general population. Thus, we cannot generalize to the U.S. population. This study focused on benefits of GM technologies and not specific types of gene modification. Future research might consider both the benefit and type of GM. Previous studies have taken an either/or approach. Either the resulting attribute was studied, or the technology used was studied.^{47,55,57,79,103} Strengths include data collection over a six-year period in a state which is familiar with GM through legislation of information availability. Future research should track revealed preferences (actual purchases) of consumers, include more studies about heterogeneity of consumer preferences, information search, investigate stated support trends over a longer period of time, and include a national sample. While overall support for GM foods is low, our research suggests there is variability in support for various GM attributes, as well as an overall transition toward attitude neutrality in recent years.

Control variables: education, gender, income, presence of children.

Acknowledgments

This research was supported USDA NIFA award numbers VT-H01404, VT-H01811, VT-H02113, and VT-H02706.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the National Institute of Food and Agriculture [VT-H01404, VT-H01811, VT-H02113, and VT-H02706].

Notes

1. Novel food production techniques go by a variety of names and the vocabulary is becoming more diverse. In the recent literature, authors are distinguishing newer techniques and distancing the research away from GM, by using terms such as gene editing (GE), bioengineering (BE), disruptive technologies (DT), and gene technology (GT). Because this paper includes not only genetic modification but also gene editing, because the average consumer continues to refer to the general process of genetic engineering as genetically modification, we choose GM to describe the range of plant engineering techniques.^52,110

2. The n varies by univariate, ANOVA, and MANCOVA analyses due to the number of respondents to specific questions and the number of variables used in each analysis.