1 Introduction

On December 31, 2020, the Andres Manuel Lopez Obrador administration published a decree calling for Mexico to phase out glyphosate and genetically modified (GM) maize for animal and human consumption by January 31, 2024.Footnote 1 A total ban on GM maize would have widespread implications across the Mexican economy, particularly impacting the production and consumption of livestock products and tortillas. Mexico has the highest per capita consumption of eggs globally, averaging 409 eggs per person in 2021 (WATTPoultry, 2022). Eggs are the second largest (17%) source of protein in the Mexican diet, only behind poultry meat, which accounts for 39% (Union Nacional de Avicultores, 2022). Tortillas provide around 13–20% of the daily caloric intake for Mexicans, who eat an average of six daily (World Perspectives, 2022). Per capita, tortilla consumption in Mexico is approximately 75 kg per year (USDA FAS, 2022).

The potential impacts of the decree on Mexican consumers are still being determined. A World Perspectives (2022) study estimated that implementing a ban on GM maize for livestock feed would lead to an increase of 66.7% in poultry meat prices in Mexico and reduce consumption by 78%, as maize represents 60–73% of the cost of production of broilers (Union Nacional de Avicultores, 2022). The World Perspectives (2022) study also estimated that tortilla prices would increase by 30% in the third year after the ban and by an average of 16% in the decade following the ban’s implementation. The increase in tortilla price is mainly driven by the response to the decreased availability of yellow maize via the ban, resulting in increased area planted to yellow maize in Mexico for livestock feed, reducing the supply and increasing the price of white maize (and tortillas). Further, given an average Mexican tortilla is estimated to consist of approximately 82% white maize flour and is blended with 18% yellow maize flour, there are direct (yellow maize) price impacts on tortilla prices as well (USDA FAS, 2023). Moreover, prices may rise for Mexico’s poorest populations to the point where eggs become a luxury item, even given that Mexico is the largest per capita consumer of eggs globally.

Approximately 45% of Mexicans live in poverty, and 23% are food insecure (Shamah-Levy et al., 2017). If the estimated price changes reported by World Perspectives (2022) are correct, the ban will likely exacerbate food insecurity. Engel’s Law asserts that low-income consumers spend a greater proportion of their disposable income on food than higher-income consumers, which holds in countries with lower poverty rates like the U.S. (Lusk and McFadden, 2021). For example, the lowest-income decile in Mexico spends 52% of its income on food, and lower-income Mexican consumers spend a larger proportion of their relative and absolute income on tortillas than the wealthiest decile (Otero et al., 2015).

Satter’s Hierarchy of Food Needs (Satter, 2007) posits that individual or household food needs can be described by different sequential levels (i.e., enough food, acceptable food, ongoing access to food, good-tasting food, novel food, and instrumental food) and the needs at lower levels must first be satisfied before progressing to the next level. It could be argued that banning the imports of GM maize into Mexico is attempting to provide “novel food” in the form of more expensive products derived from non-GM maize, which is at odds with sourcing “enough food” for the many consumers at the lowest level of Satter’s Hierarchy of Food Needs. Understanding the needs of lower-income consumers is also important because previous research has shown that the preferences concerning GM food and the associated risks and concerns perceived by consumers are diverse, so the “average consumer” analysis can be misleading (Hu et al., 2004).

Before imposing the decree, government officials in Mexico did not consult with consumer groups or livestock producers who rely on GM maize imports to survive economically. One of the primary reasons President Obrador gave for the decree was to “… use agroecological practices and inputs that are safe for human health” (USDA FAS, 2021). Juan Cortina, the president of the Mexican National Farm Council, expressed concern about the ban on GM maize for the livestock sector in Mexico: “Even if we wanted to import it [maize for feed] from somewhere else, it does not exist, transgenic or non-transgenic. Those volumes do not exist for importation from somewhere else, especially efficiently, as it is currently acquired from the United States” (Reuters, 2021). President Obrador also failed to meet with Mexican maize producers as he suggested: “…replace the use of glyphosate by workers who can clean up the crops with machetes” (Dominguez, 2021). As such, the decree can be seen as an expression of political will rather than a reflection of the needs and desires of farmers, producers, scientists, and consumers (Ventura, 2022).

Maize is Mexico’s most important crop in production and consumption, accounting for over 80% of all cereal production, 14.5% of the agricultural gross domestic product, and 20.9% of the household food expenditure (Secretary of Agriculture of Mexico, 2017). Mexico bought more than 20 million metric tons of maize from the United States in the 2021-22 marketing year, second to only China, with an estimated value of $4.92 billion (USDA FAS 2023). Maize in Mexico can be segregated into two main markets: yellow maize, which represents 39% of the maize market and is mainly used for animal feed, and white maize, which represents around 61% of the total maize demand and is destined for human consumption. Mexico imports over 75% of the yellow maize (which represents only 13% of the domestic production) and white maize (which represents 87% of the domestic production) (Secretary of Agriculture of Mexico, 2017). Mexico relies heavily on U.S. maize imports, mainly yellow maize, for livestock production. Over 90% of U.S. maize is GM, which would drastically impact bilateral trade should the GM ban be implemented. Over the last five marketing years (2017/2018–2021/2022), yellow maize accounted for an average of 95% of U.S. exports to Mexico, with white maize traditionally used for tortillas accounting for the remaining 5% (Williams et al., 2022). An average Mexican tortilla is estimated to consist of approximately 82% white maize flour and is blended with 18% yellow maize flour (USDA FAS, 2023).

On February 13, 2023, the Mexican government published a new decree clarifying that the ban applies to the sale of GM maize seed and GM maize for human consumption in Mexico and added an exemption for GM maize for animal feed (Mexico, 2023). U.S. maize exports to Mexico in 2022 were worth about $5 billion, and maize for human food use comprises about 21% of Mexico’s maize imports from the U.S. (Garrison, 2023). The newly revised decree states that Mexican authorities will carry out “the gradual substitution” of GM maize for human consumption (The Associated Press, 2023a). Still, it sets no date for doing so and says potential health issues will be the subject of study by Mexican experts “with health authorities from other countries” (The Associated Press, 2023b).

The revised decree clarifies that Mexico reserves the right to take precautionary measures it considers essential to protect public health and the environment, including the genetic integrity of its diversity of native maize (Mexico, 2023). U.S. industry representatives and government officials have repeatedly insisted that Mexico’s actions are based not on science but on protectionist policies and that the decree contradicts Mexico’s commitments under the Sanitary and Phytosanitary (SPS) Measures chapter of the United States-Mexico-Canada Agreement (Lawder, 2023). GM import bans have been used to protect local agriculture in other places. For example, Heilongjiang, China’s largest producer of non-GM soybeans, banned GM imports in 2016 despite the central government’s plan to commercialize GM varieties. The import ban was used to protect the Heilongjiang non-GM soybean sector after the importation of GM soybeans caused several non-GM crushing plants to close in the region (Zhang & Wu, 2024). Other countries (e.g., European Union member states) have enacted slow, stringent approval processes for importing GM varieties that can implicitly ban imports (Kerr, 2015).

Interestingly, one of the tenets of Article 6 of the Presidential Decree claims that it aims to contribute to food security. Still, it is difficult to see how increasing the price of yellow maize under the first version of the decree, estimated to increase by 81% in some studies, in Mexico will enhance food security (Macall et al., 2021). Food security risks associated with GM bans have played out in other countries. For example, in 2020 Zimbabwe lifted an import ban on GMs that had been in place for 12 years after the worst drought in decades that resulted in more than half of the population needing food aid (Ndlovu, 2020). Similarly, Kenya banned GM crops in 2012 and then lifted the ban in 2022 after experiencing its worst drought in 40 decades and soaring food prices (Oloo, 2022).

There is some literature on the impacts of the presidential decree on producer and consumer prices (Macall et al., 2021; World Perspectives, 2022). Previous literature has shown that marginalized Mexican maize producers refrain from discounting GM maize relative to non-GM varieties (Birol, Villabla, and Smale, 2009). Still, a gap exists on whether Mexican consumers support the ban and would be willing to pay (WTP) for the impacts of its implementation. To our knowledge, research has yet to be conducted on what consumers are WTP for non-GM maize products, specifically for the potential price increases associated with implementing the initial decree.

Given that the decree was not a referendum, this study sets out to survey Mexican consumers about their knowledge of the potential GM import ban, if they support the ban, and if they are WTP, the estimated price increases on three staples (chicken, eggs, and tortillas) which will likely be affected by the decree. Maize has cultural and dietary importance in Mexico, and this study fills this gap by first determining which consumers (by socioeconomic group) would be WTP for the ban and if that premium is large enough to offset the predicted price impacts of the ban. Importantly, this study focuses on whether more affluent consumers drive support for the ban and if their WTP differs from low-income Mexican consumers. One large contribution of this study is to estimate if the ban is being forced on low-income consumers who could likely not afford to pay the necessary premium for non-GM maize products, leaving the poor with even fewer options to try and alleviate food insecurity. The results of this research will provide policymakers with important information about whether Mexican consumers support the decree and if they are WTP the expected price increases by its implementation. More generally, this study’s findings can serve as a roadmap for future studies about the economic impact of agricultural policies and the importance of assessing consumer preferences to gauge the political feasibility and economic implications of policy proposals.

2 Materials and methods

2.1 Survey Overview

This study used a web-based survey designed in online software created by Qualtrics and the survey was distributed by Qualtrics to an online panel of 1,301 Mexican respondents in April of 2023. A coauthor whose native language is Spanish translated the survey questions into Spanish, and the survey responses were translated from Spanish to English. Respondents had to be 18 to qualify for the study and eat chicken, eggs, and tortillas. Approximately 5% of the sample did not consume all the food products, so 1,238 respondents completed the survey. A quota-based sampling approach was used to ensure that the sample of respondents matched the regional population proportions of Mexico (Central, Eastern, Northern, Southern, and Western). Several other demographic variables were measured at the end of the survey, including age, education, gender, grocery shopping responsibilities, income, race, and whether they live in a rural, suburban, or urban area. The wording for the education and race questions was retrieved from the Mexico census (Mexico, 2020). Summary statistics for the sample’s demographic characteristics are shown in Appendix Table 1.

2.1.1 Dichotomous choice contingent valuation questions

After the qualifying and quota questions, respondents were informed they would make hypothetical purchasing decisions for three food products (chicken, eggs, and tortillas). The hypothetical purchasing questions aimed to estimate the premium Mexican consumers were WTP for a non-GM version of the products over a GM version. For the GM versions, it was communicated that the chickens (meat) and hens (eggs) were fed GM maize, and the tortillas were made using GM maize as an ingredient.

Product selection between GM and non-GM versions was elicited using dichotomous choice contingent valuation (DCCV) questions, and premiums for the non-GM versions over the GM versions of products were estimated using a double-bounded approach. The double-bounded approach asks each respondent two DCCV questions, in which all respondents are presented with the same initial DCCV question. Respondents then answered another DCCV question with a higher or lower premium depending on the answer to the initial DCCV question. The double-bounded approach to DCCV provides more efficient WTP estimates than a single-bounded approach that asks only one contingent valuation question (Hanemann et al., 1991). Contingent valuation is widely used in economic studies (Adamowicz, 2004); however, WTP values estimated from contingent valuation can suffer from hypothetical bias, and discrete contingent valuation can produce high WTP values (Jacobsen & Hanley, 2009). Therefore, respondents were provided a brief cheap talk script because using a cheap talk script can reduce WTP estimates from DCCV questions (Lusk, 2003). The cheap talk script shown to respondents is presented in Appendix Fig. 1.

Fig. 1
figure 1

Responses to the dichotomous choice contingent valuation (DCCV) questions [Note: Percentages in parentheses next to prices are premiums for the non-GM versions, and percentages in parentheses next to the number of respondents are sample proportions.]

Full size image

Prices for the GM versions of the three products were obtained across the country from a major Mexican retailer, reflecting prices for the specific products at the time of data collection. The premiums for the non-GM versions of chicken and tortillas were selected from a World Perspectives (2022) report, which estimated that poultry prices would increase by 67% due to higher feed costs and that tortilla prices would increase by an average of 30% in the third year after the ban. The highest premium for the non-GM versions in the DCCV approximately matched these estimates; the premiums were approximately half for the first question, and the lowest premiums were approximately a fourth.

Prices, in Mexican pesos, used in the initial DCCV question for the products were: chicken (GM-$52/kg, non-GM-$69/kg), eggs (GM-$35/dozen, non-GM-$43/dozen), and tortillas (GM-$13/kg, non-GM-$15/kg). These price differences between GM and non-GM alternatives represent non-GM premiums of 33%, 23%, and 15% for chicken, eggs, and tortillas, estimated to take effect if the ban is implemented (World Perspectives, 2021). Those who selected the higher-priced non-GM version in an initial DCCV question were then asked a second DCCV question with a higher premium for the non-GM version; the premiums for the non-GM version were 65%, 46%, and 31% for chicken, eggs, and tortillas, respectively. Those who selected the lower-priced GM version in an initial DCCV question were then asked a second DCCV question with a lower premium for the non-GM version; the premiums for the non-GM version were 12%, 14%, and 8% for chicken, eggs, and tortillas, respectively. The initial DCCV questions shown to respondents for chicken, eggs, and tortillas are presented in Appendix Figs. 2, 3 and 4; the follow-up questions were the same as the initial question, except the non-GM version had a higher or lower premium depending on the choice made in the initial question.

2.1.2 GM Safety questions

After answering the DCCV questions, respondents completed several questions to determine opinions about the safety of food produced with GM inputs. The first set of GM safety questions provided respondents with a statement and asked for their level of agreement with the statement; the response options were on a 5-point Likert scale (i.e., strongly disagree, somewhat disagree, neither agree/disagree, somewhat agree, and strongly agree). Three questions were asked about the safety of products used in the contingent valuation questions (i.e., chicken, eggs, and tortillas). For example, the statement shown to respondents for eggs was “Eggs from chickens fed GMO maize as a feed ingredient are safe for me to consume.” Then, two questions were asked about the safety of GM maize, one for maize from Mexico and one from the U.S. For example, the statement shown to respondents for Mexico was “GMO maize grown in Mexico is safe for me to consume.” Data from these questions allow for testing of heterogeneity across stated safety concerns for products and countries.

Respondents then answered a second set of safety questions similar to the first set of safety questions as an additional measure of heterogeneity in safety concerns; however, the second set of questions asked respondents to rank the GM products and maize from Mexico and the U.S. by safety. The first question asked respondents to rank the safety of chicken, eggs, and tortillas; tamale husks were added to the ranking question to include a GM maize product that could be used for preparing food but not consumed. The second question asked for GM and non-GM maize safety rankings from Mexico and the U.S. Lastly, respondents were then asked whether they believed they consumed foods with GM ingredients.

2.1.3 Awareness of and support for the Import Ban and Support sensitivity questions

Respondents were then asked questions to determine awareness and support for the decree. The awareness question was asked first with response options yes, no, or I don’t know. Only those aware of the decree (i.e., a yes response to the awareness question) answered a subsequent question asking if they supported the import ban, with response options yes or no.

Then, follow-up questions examining reasons for support were asked to those who supported the import ban (i.e., a yes response to the support question). Supporters were first asked to select the reason(s) why they supported the decree from the four response options: human health concerns, environmental concerns, protecting cultural heritage, or protecting Mexican heritage. Respondents could select multiple reasons, and then the reasons selected were piped into a subsequent question asking them to rank the reasons selected.

Lastly, supporters of the decree were asked two questions to measure the sensitivity of support to potential jobs lost or increased maize prices associated with the import ban. World Perspectives (2022) estimated that the original decree would result in 56,958 jobs lost annually in Mexico and an increase in maize prices of 19% in the next ten years and up to 30% in the first few years after implementation. These values were used to ask supporters if they would still support the decree given a potential loss of jobs for 55,000 Mexicans and if they would still support the decree given the increase in maize price increases, with response options yes or no.

2.2 Data Analysis

Awareness, support, and reasons for supporting the decree banning GM maize imports were examined first. Differences across the reasons for supporting the decree were determined by estimating tests between the proportions of the reasons selected, and t-tests were estimated to determine differences in the subsequent ranking of chosen reasons. P-values from the proportion and t-tests were adjusted using the Bonferroni method to account for the multiple comparisons between reasons for support. Additionally, a test of proportions was used to determine if support of the decree was more sensitive to potential loss of jobs or increases in maize prices.

Consumer characteristics associated with awareness and support for the decree banning GM maize imports were determined by estimating a separate logistic regression model for each outcome. The dependent variables for these models were coded as one if a respondent was aware of or supported the decree and zero otherwise. The estimated logistic regression models for awareness and support can be specified by:

$$Pr\left(Y_i=1\right)=X'\beta,$$
(1)

where X is a matrix of consumer characteristics, including the region of Mexico a respondent resided in, age, education, gender, grocery shopping responsibilities, income, race, and whether they live in a rural, suburban, or urban area, and β is a vector of marginal effects to be estimated.

Heterogeneity in stated opinions about the safety of GM products was estimated using t-tests. Separate analyses were done for the GM products (i.e., chicken, eggs, and tortillas) and GM maize produced in Mexico versus the U.S. Then separate analyses were done for the safety ranking of products, including tamale husks and the safety ranking of GM and non-GM maize from Mexico and the U.S.

Responses to the DCCV questions were used to estimate the premiums consumers were WTP for the non-GM versions of chicken, eggs, and tortillas. WTP was estimated using a double-bounded model like that described in Hanemann et al. (1991). Price differences between the GM and non-GM versions were converted to percent differences so premiums could be directly compared to the price increases associated with the import ban estimated by World Perspectives (2022). Two specifications of the double-bounded model were estimated for each product. One specification, Model 1 (M1), only included a constant term to estimate the average premiums for the non-GM premium. The other specification, Model 2 (M2), included independent variables to determine the associations of awareness of the decree, support for the decree, and respondents’ income with the premium for non-GM versions. Coefficient estimates for these variables allow us to determine how much of the premium is associated with those who supported the decree and whether lower-income consumers have lower WTP values for the non-GM versions. All estimations were conducted using Stata® Standard Edition 17.

3 Results

Nearly 94% of respondents believed they currently consumed some foods containing GM ingredients. Thus, consumers have some understanding of the proliferation of GM crops or foods in their current diet. However, more than half of the respondents (54%) were unaware of the decree banning the import of GM maize. Of the 46% who were aware of the decree, 77% supported the ban (36% of the full sample).

Concern about human health was selected by 85% of those supporting the decree, protecting Mexican heritage was chosen by 48%, environmental concern was selected by 40%, and protecting cultural heritage was chosen by 26%. There were significant differences (P < 0.01) between the reasons for supporting the decree, except there was no significant difference in the proportions selecting to protect Mexican heritage and environmental concerns. The subsequent ranking order for the reasons supporting the decree reflected the selection proportions; thus, human health concern was the largest reason Mexican consumers supported the decree; there was no difference in the rankings of protecting Mexican heritage and environmental concerns or protecting cultural heritage had the lowest ranking.

The proportion of respondents who were aware and supported the decree decreased from 77 to 46% (21% of the full sample still supporting) when provided information about the potential jobs lost due to the decree and to 56% (26% of the full sample still supporting) when provided information about increases in maize prices. A test of proportions indicated that consumer support of the ban was more sensitive to potential job losses compared to increased maize prices (P < 0.01). Thus, the potential costs associated with the decree, particularly the negative impact on employment, are important considerations when considering support for the ban on GM maize imports.

The marginal effects estimated by the logistic regression models for awareness and support are presented in Table 1. There were few significant associations between consumer characteristics and awareness or support for the decree banning the import of GM maize. Respondents with a graduate degree or who bought at least half the groceries for a household were relatively more aware of the decree. On average, those residing in the Western region of Mexico and female respondents were less aware of the decree. However, the females who were aware of the decree were more supportive of the import ban on GM maize.

Table 1 Marginal effects estimated from the logistic regression models
Full size table

Responses to the questions about the safety of GM products and GM maize with five-point Likert-scale response options are shown in Table 2. The weighted average of responses was below three for all the questions (i.e., 1 = strongly disagree, …, 5 = strongly agree), indicating that consumers disagreed that GM products or maize were safe to eat. The mean response for tortillas was significantly (P < 0.05) higher than chicken; thus, consumers tended to believe that using GM maize as feed was slightly safer than poultry raised using GM maize feed. There were no other significant differences in opinions about safety across the products. GM maize grown in Mexico was considered (P < 0.01) safer than GM maize grown in the U.S.

Table 2 Responses to the questions about the safety of GM products and GM maize
Full size table

Responses to the safety rankings for GM products and GM and non-GM maize are shown in Table 3. There was no significant difference in the safety ranking of tortillas and tamale husks, which had the highest safety rankings. The safety rankings of both tortillas and tamale husks were significantly higher than eggs or chicken (P < 0.01), and the safety ranking for eggs was higher than chicken (P < 0.01). Results for the safety rankings of products align with research in the U.S., concluding that consumers are generally more averse to fresh products like meat from animals fed with GM maize than processed products using GM maize (Lusk et al., 2015). There were significant (P < 0.01) differences in the safety ranking of all non-GM and GM versions of maize from Mexico and the U.S. Non-GM maize from Mexico was ranked safest, non-GM maize from the U.S. was ranked the second safest, then GM maize from Mexico, and lastly, GM maize from the U.S. Thus, consumers appear to be more concerned about the GM status of maize compared to the country of production.

Table 3 Safety rankings for GM products and GM / non-GM maize
Full size table

Figure 1 illustrates the responses to the DCCV questions (prices in Mexican Pesos). The coefficients estimated by the double-bounded contingent valuation models are presented in Table 4. On average, the sample was WTP a premium of 73% for chicken, 50% for eggs, and 50% for tortillas produced with non-GM maize. These premium estimates are larger than the potential price increases of 67% for chicken and 30% for tortillas estimated by World Perspectives (2022). In the M2 estimates, the coefficients for Support were significant and positive for all products, and the coefficients for the lowest-income category of an annual income of less than 7,000 pesos were significant and negative for all products. Thus, respondents who supported the ban had relatively higher premiums for non-GM products, while those with lower incomes had relatively lower premiums. The coefficients for two of the other income categories (income between $7,000-$20,000 and between 20,000 and 60,000) were also significant and negative for eggs and tortillas, while the coefficient for the income category between $60,000-120,000 was not significantly different from the reference income category (income higher than $120,000). Taken together, the average WTP estimates from M1 are driven by consumers who support the ban and have income levels high enough to cover premiums for non-GM products.

Table 4 Coefficients estimated from the double-bounded contingent valuation models
Full size table

Respondents who supported the import ban were WTP non-GM premiums of 91% for chicken ($0.736 − 0.091 + 0.266), 71% for eggs ($0.578 − 0.154 + 0.285), and 66% for tortillas ($0.578 − 0.145 + 0.226). However, respondents in the lowest-income category, even those who supported the ban, had an average WTP below the sample average. Approximately 22% of the sample was in the lowest-income category, and 42% of those low-income respondents were aware of the decree (9% of the full sample). The low-income respondents unaware of the decree were WTP premiums of 55% for chicken ($0.736 − 0.187), 37% for eggs ($0.578 − 0.213), and 40% for tortillas ($0.578 − 0.179). On average, low-income consumers who supported the decree were WTP premiums of 72% for chicken ($0.736 − 0.091 + 0.266 − 0.187), 50% for eggs ($0.578 − 0.154 + 0.285 − 0.213), and 48% for tortillas ($0.578 − 0.145 + 0.226 − 0.179), while those who did not support the ban were WTP premiums of 46% for chicken ($0.736-0.091-0.187), 21% for eggs ($0.578-0.154-0.213), and 25% for tortillas ($0.578-0.145-0.179). Thus, the premiums stated by low-income consumers who do not support the ban are not enough to fully cover the potential increase in the price of chicken, eggs, and tortillas.

4 Discussion

In 2020, the government of Mexico enacted a decree seeking to ban glyphosate and GM maize imports by 2024. A new decree in February 2023 clarifies the extent of the measure and states that the ban will apply to GM maize used for human consumption and exempts GM maize for feed and industrial use. If implemented, the ban could significantly increase the price of chicken, eggs, tortillas, and other maize-based products, all of which are significant calories and protein sources in the Mexican diet. While high-level government discussions and technical consultations about the glyphosate and GM maize ban between the U.S. and Mexico are ongoing, our study shows several interesting findings. Regardless of the original motivation for the ban, our results show that it may be politically popular (77% of the attentive public, those aware of it, support it). This aligns with previous research that found Mexican consumers perceived GM food as risky (López et al., 2016). Information is essential for forming policy preferences, as highlighted by our findings that the popularity of the decree decreased significantly when consumers received information about potential job losses and maize price increases resulting from the ban. In this context, a quick look at the information related to the decree and available in Spanish shows that most of the focus has been on the health impact of glyphosate and a close association between glyphosate and GM maize.

Very few news articles have focused on the potential economic and food security consequences of the ban, while the potential benefits of other GM maize events beyond herbicide tolerance have been ignored. There are multiple potential benefits associated with adopting GM crops (McFadden et al., 2021). For example, a randomized controlled trial found that Bacillus thuringiensis (Bt) brinjal increased yield by 51% compared to non-GM brinjal while reducing pesticide use and exposure to toxicity (Ahmed et al., 2021). Similar results have been found for Bt maize in the Philippines (Mutuc et al., 2013) and South Africa (Gouse et al., 2005). Sanglestsawai et al., (2014) found that the yield benefits of Bt maize are larger among small, poorer farmers. Drought and heat-tolerant maize can help expand maize production areas currently considered marginal and, more importantly, can become an adaptation practice to climate change and the increasing occurrence of droughts and heat shocks (Tesfaye et al., 2018). GM crops, including GM maize, can also help combat undernourishment and malnutrition through biofortification (Bouis et al., 2020).

In a videoconference on March 29, the government of Mexico presented scientific evidence about the health risks of GM maize, but it was primarily associated with the presence of glyphosate, again ignoring other GM traits that could prove useful in combating food insecurity. The biased information may lead to biased public opinion on the subject, which may help explain the high level of support among the attentive public and the enactment of the decree. Diamond et al., (2020) state that when opinions between the public and scientific community differ, policymakers are pressured to respond to constituents’ attitudes rather than scientists. The results also show a country-of-origin bias (U.S. maize is perceived as less safe than Mexican maize), which is not supported by science but affects consumers’ perceptions.

To our knowledge, only one study (World Perspectives, 2022) assessed the potential impact of the GM maize ban and found that it could significantly impact the price Mexican consumers pay for maize-based foods. Our findings suggest that, on average, consumers are WTP price premiums equivalent to the estimated higher prices for chicken, eggs, and tortillas. These findings could be seen as encouraging signs for supporters of the GM maize ban, although, a caveat for all WTP estimates in this study is the hypothetical elicitation of preferences and possible overestimation of premiums. Nevertheless, low-income households unaware or unsupportive of the decree were not WTP similar premiums as other households for the non-GM versions and had a lower WTP for chicken than the potential price increase estimated by World Perspectives (2022). Further consequences on food security must be acknowledged as consumers, primarily from low-income households, are forced to spend more income and a disproportionately higher proportion of their income on maize-based staples.

Previous studies found that high food prices can reduce access to nutritious food and worsen the nutritional and health status of the population (Brinkman et al., 2010). Results from the 2020 National Survey on Health and Nutrition (Shamah-Levy et al., 2021) show that the increase in food insecurity between 2018 and 2020 was related to a decrease in diet quality due to the lack of resources to purchase food. Food insecurity affects 59.1% of Mexican households, particularly among indigenous populations in the southern states. The ban on GM maize for human consumption can affect particularly the southeastern region that relies on imported white maize, which is cheaper than sourcing it from the production areas located mainly in the central region of Mexico (Suarez, 2023). The potential negative impact of the ban on food security contradicts one of the main arguments presented in the December 31, 2020 Decree that introduced the GM-maize ban. Further, the effects of a GM ban on food security have been previously realized in other countries, and the effects are amplified by unforeseen occurrences like weather events (Oloo, 2022; Ndlovu, 2020).

The lack of information about Mexican consumers’ WTP for non-GM maize products highlights the need for feedback about consumer preferences in the GM ban policy process. While we do not know the motivations behind the President’s decision to enact the ban, the 2020 decree that introduced the ban on GM maize states the importance of research and collaboration between researchers, farmers, and indigenous communities that could be used for policy design, clearly ignoring the potential contribution of consumer studies. Producing timely scientific information on consumer preferences could help formulate better policies by broadening the information available for decision-making.

These results can be seen as encouraging for stakeholders interested in advancing the ban. Still, they should be further analyzed to ascertain impacts on food security, considering the importance of maize-based products in the diet and the high level of food insecurity in Mexico. History should teach us that increases in tortilla prices in Mexico can lead to political and civil unrest, as was the result of the 2007 tortilla crisis. When maize tortilla prices rose from eight pesos per kg to more than ten pesos in late 2006, hundreds of thousands of Mexicans took to the streets of Mexico City to protest. Within a few weeks of the initial protests, President Felipe Calderon managed to contain most of the price increases, and tortilla prices fell back to 8.5 pesos per kg, which was low enough to staunch a popular uprising (Financial Times, 2010). As such, our results may indicate that Mexican consumers are, on average, WTP the estimated premiums associated with implementing the decree; the real question is whether those premiums are sustainable for lower-income Mexican consumers in the long run. Moreover, the discussion about banning GM maize should be expanded beyond its association with glyphosate and include the potential benefits from other traits, such as Bt maize, that reduce the use of insecticides and drought tolerance and other traits such as biofortification and high protein maize.

This study has limitations. Contingent valuation was used to estimate WTP, and hypothetical bias is a concern when using contingent valuation. Future research could use revealed preference data (e.g., retail scanner data) to determine the effects of the import ban on prices and quantities demanded for the products in this study. Also, food security rates could be used to estimate the calories forgone due to the import ban. However, it will likely take several years for enough data to be available post-ban to examine these. The price increases used in this study for the non-GM products were pulled from estimates by World Perspectives (2022), and it is possible that the actual increases will be different from what was estimated. Future hypothetical research could use larger and smaller price increases to determine sensitivity in the premiums consumers were WTP for non-GM products relative to GM products.