If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Danish milk consumers are critical of advanced breeding methods in dairy production, but only 1 in 5 is unwilling to drink milk from dairy cows bred with semen derived from such methods
Department of Food and Resource Economics, University of Copenhagen, Frederiksberg 1958, DenmarkDepartment of Veterinary and Animal Sciences, University of Copenhagen, Frederiksberg 1870, Denmark
Assisted reproductive technologies and genetic technologies can accelerate progress in breeding programs in dairy farming, but it is unclear how consumers will react to the use of these technologies. Using representative questionnaire data on Danish citizens (n = 2,036) this cross-sectional study examined consumer attitudes to the application of advanced technologies in dairy cattle breeding. Attitudes were examined in 2 ways. First, we prompted about general attitudes to assisted reproductive technologies and genetic technologies in dairy cow breeding. Here we found that most of the participants were critical of cow impregnation involving hormone therapy and the insertion of cloned fetuses. Second, we used a vignette experiment to study whether acceptance of and willingness to drink milk varies with the type of technique that farmers use for their breeding work, as well as the traits being bred for. We included 5 breeding methods with differing degrees of technological complexity. Participants were randomly assigned to receive tailored information about 1 of the 5 breeding methods. The information specified that dairy farmers' own use of advanced technologies is limited to using semen in artificial insemination on the farm. The potentially concerning technologies are here not applied at farm level but are represented in the semen used in artificial insemination because they were used by breeders on earlier generations of cows and bulls to develop semen with higher genetic merit. There was much less concern about this indirect use of the technologies. Only 1 in 5 participants thought the most advanced method we prompted about (use of semen from breeding methods involving genetic engineering and cloning) was unacceptable. Unwillingness to drink milk from cows produced through such a breeding method was also modest (18%) and not much higher than the unwillingness to drink milk from a cow produced by natural fertilization (10%). A likely reason for the unexpectedly low level of unwillingness to drink milk is that people regard the genetic engineering as distant from the final product. We also found that high-frequency organic milk consumers were more critical of advanced breeding methods. Thus, 28% within this group were unwilling to drink milk from cows impregnated with semen derived from earlier generations of cows and bulls bred using gene editing and cloning. Further, this share rose if the high-frequency organic consumers were very averse to the manipulation of nature. The organic sector may need to cater to this subgroup (e.g., by ensuring the traceability of the semen that organic farmers use to artificially inseminate their cows).
A growing number of assisted reproductive technologies (ART) and genetic technologies are available to facilitate rapid genetic progress in dairy cattle breeding. Artificial insemination, multiple ovulation embryo transfer, sexed semen, and genomic selection are already widely used (
). Other technologies are being developed and implemented, such as ovum pick up (OPU) combined with in vitro production (IVP) and genomic selection of embryos (
). In addition to their use in the development of economically valuable efficiency traits in dairy production, the technologies can be used to pursue breeding goals addressing wider public demands, including those for improved animal welfare (
The technologies, though promising, come with the caveat that their use could become controversial. Farmers and affiliated industries require public support to operate (
). Historically, GE has been seen by a large number of people as “going against nature,” sacrificing the integrity of animals and potentially risky to human health (
In dairy breeding and farming, GE could get caught up in these controversies. Studies from Europe, the United States, Canada, and Brazil have found that consumers are critical of milk produced using GE and cloning and have a lower willingness to drink (WTD) milk from cows produced with such techniques (
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
Acceptance of a food of animal origin obtained through genetic modification and cloning in South America: A comparative study among university students and working adults.
). Still, a recent study found that US and Canadian citizens prefer “natural” over technological methods (such as GE) for promoting dairy cow welfare (
). In one study where attitudes to ART were examined, most German consumers were critical of the use of sexed semen, embryo transfer, hormone treatment of a cow, and cloning (
How a mix of advanced breeding tools (such as GE, cloning, IVP, and genomic selection of embryos) would affect public acceptance and consumer WTD milk is not yet clear. The aim of the present study is therefore to investigate acceptance and WTD milk in a case study country: Denmark. Denmark is a small, affluent country in northern Europe that maintains a significant, export-oriented dairy sector with high vertical integration due to its strong cooperative owner structure. The market share of organic milk in the country is very high (
We began by assessing knowledge and understanding of dairy breeding. We then examined general attitudes to several technologies with current and potential applications in dairy cattle breeding. Last, we investigated acceptance and WTD milk when the dairy farmers' own use of advanced technologies is limited to using semen in AI on the farm. Here the aim was to investigate acceptance and WTD milk when the potentially concerning technologies are not applied at farm level but are represented in the semen used in AI because they were used by breeders on earlier generations of cows and bulls to develop semen with higher genetic gains.
We anticipated that more advanced breeding methods would be associated with lower levels of acceptance and WTD milk than less advanced methods, but that this difference would be reduced to some extent when societally relevant breeding goals (animal welfare and positive environmental effects) are being pursued.
We further anticipated that consumers with strong animal rights views, and those who feel uncomfortable about human tampering with the natural world, as well as organic milk consumers, would exhibit lower levels of both acceptance of and WTD milk produced from advanced breeding methods.
MATERIALS AND METHODS
Participants
This was an observational study with data from a cross-sectional questionnaire where the Danish public was the target population. To obtain a nationally representative sample of Danish citizens (older than 17 yr), we contracted a survey bureau, Norstat Danmark A/S, that hosts a panel of Danish citizens (panel size approximately 90,000). The bureau invited a randomly selected gross sample of individuals in the panel to participate in the study via email. The email contained a link to the online questionnaire. Individuals who agreed to participate gave informed consent to their responses being used in research. The study received ethical approval from the Research Ethics Committee of Science and Health at the University of Copenhagen (Ref: 504-0033/18-5000). The data were collected in May and June 2021, and 2,036 participants completed the questionnaire.
Before the main data were collected, a pilot data-collection procedure was carried out in May 2021 (n = 201) to check that the descriptions of breeding methods and goals would be understood by the intended participants. Participants in this pilot examination were also recruited from Norstat Danmark A/S. Data from the pilot test are not included in the analysis reported in this paper, and participants taking part in the pilot were not invited to the subsequent main data collection. The questionnaire ultimately used to collect the main study data can be viewed in Supplemental File S1 (https://figshare.com/articles/online_resource/Supplemental_File_SI_https_doi_org_10_3168_jds_2022-22249_docx/21787337;
The vignette experiment is a method commonly used in questionnaire surveys to elicit how attitudes, opinions, and behavior are a function of different conditions relevant to the area of inquiry. This is done by systematically varying the description of, for example, a situation or a person (the vignettes) (
). In this study, we varied 2 factors with 5 and 4 conditions, respectively. This gave a 5 × 4 factorial design, and a total of 20 vignettes.
The first factor, breeding method, varies with the type of technique that farmers use for their breeding work. We included 5 breeding methods with differing degrees of technological complexity. The first and least advanced technique was bull fertilization (BULL). This technique is uncommon today; recently less than 10% of dairy farmers in Denmark have been found to use it (
), and it has been superseded by AI. However, as a fertilization process it resembles “natural” reproduction most. Given this, we treated it as a baseline through which we could study how acceptance and WTD change with more advanced methods. The 4 remaining methods all rely on the farmers' use of artificially inseminated semen. In this type of application, farmers may make indirect use of GE and ART. They do so when the semen made available to farmers by the breeding industry is derived from breeding schemes where these techniques have been used in the past, in the developmental phase. The second method was AI, in which semen from a bull donor at a bull station is used. It was explained that the semen is selected according to its ability to pass on desired traits to future generations of dairy cows, but without further specification of techniques used to achieve this. The third technique (AI+GS) involves AI in combination with genomic selection. This breeding method is relatively widespread in Denmark. Approximately 50% of dairy farmers use the genetic index Nordic Total Merit (
). The 2 remaining methods are still in the developmental phase and are not yet marketed in Denmark. The fourth technique (AI+OPU/IVP) accelerates genetic progress in dairy cows by combining OPU and IVP with genomic selection of embryos and embryo transfer. Details of this method have been described elsewhere (
). The fifth technique (AI+CLONING/GE) is again a mix of methods. Here cells are genetically modified and converted to fetuses using cloning. Embryo transfer is then used to transfer them to foster mothers. The concept and the details of this technique are described by
. When bulls developed using the last 2 of the techniques show sufficient genetic progress, their semen is made available to dairy farmers.
The second factor, breeding goal, concerns the traits being bred for. Here we included 3 breeding goals that mainly reflect those currently being pursued in Danish dairy cattle breeding. We added 1 further goal, which is not yet used in practice because semen with genetic gains on this trait is not yet marketed. The first 2 goals are production oriented, whereas the other 2 cater to societal concerns. The first goal, production of more milk per cow (milk quantity), is a common breeding goal. It can increase the milk producer's productivity, and thus profits. The second, higher milk quality (milk quality), may also have an impact on profits because it allows the dairy farmer to produce milk with a composition that adds value to the milk. The specific example given to participants was milk with less fat and more protein. The third goal is lower climate impact (climate impact) by means of reduced methane emissions from the cows. This is a promising future breeding goal (
) but currently it is not being pursued in the Danish dairy breeding sector (personal communication, Søren Borchersen, VikingGenetics). The fourth goal is improved animal welfare (animal welfare). This can be achieved through various means. Examples include breeding for polledness (
) and breeding for disease resistance. The specific example given to participants was ability to avoid diseases such as mastitis.
In the questionnaire, the presentation of the vignettes started with an introductory text that was similarly worded and presented in all 20 vignettes:
Danish milk producers are continuously ensuring that their current dairy cows give birth to heifers (i.e. future dairy cows). To that end, producers must select the bulls whose semen is to fertilize their current dairy cows. You will now be presented with a particular breeding method and a breeding goal that can form the basis of the producers' breeding work.[Breeding method]Before the fertilization of current and future dairy cows on the farm, the milk producer, often in collaboration with a consultant, assesses the various bulls available.
Following that common introduction, tailored information matching the 5 breeding methods was presented. See Table 1 for a description of the text provided. After that, tailored information matching the 4 breeding goals was presented (see Table 2).
Table 1Overview of information provided to participants for each of 5 breeding methods included in the 5 × 4 between-subjects survey vignette
Breeding method (short name)
Information given to participants
Bull
The selected bull is placed in the cowshed together with current and future dairy cows, after which it covers them.
AI
[opening text]: The bulls are not on the farm. They are instead assembled centrally at what is called a bull station. The bulls are selected according to their ability to pass on desired traits to future generations of dairy cows. [closing text]: The semen from the bulls at the bull station is tapped and then transferred to current and future dairy cows on the dairy farm.
The bulls are conceived by in vitro fertilization. The eggs used are sucked from promising young females. The fetuses are assessed on the basis of a genetic test when they are one week old. The most suitable fetuses are selected and inserted into foster mothers who give birth to the bull calves to be used for breeding by the milk producers. [closing text]
The bulls are derived from the cells of a bull with particularly good characteristics. In addition, the cells from the bull will have been genetically engineered so that the future bull will have even better traits. Subsequently, the genetically engineered cells are transformed into fetuses by cloning. These fetuses are inserted into foster mothers who give birth to the bull calves to be used for breeding by the milk producers. [closing text]
Because there is no AI involved in the bull vignettes, the first part of the opening sentence, where the breeding goal was specified to the participant (i.e., “The milk producer can, for example, obtain semen from a bull at the bull station”) was formulated differently as, “The milk producer can, for example, acquire a bull...” for all participants who received a bull vignette.
Milk quantity
Breeding goal: production of more milk The milk producer can, for example, acquire semen from a bull at the bull station which passes on to future generations of cows the ability to produce a lot of milk. This can help the milk producer to be financially secure and improve the export opportunities for Danish milk producers.
Milk quality
Breeding goal: higher quality milk The milk producer can, for example, acquire semen from a bull at the bull station which passes on to future generations of cows the ability to produce milk with a better composition of nutrients. This can help produce milk with less fat and more protein.
Climate impact
Breeding goal: less climate impact The milk producer can, for example, acquire semen from a bull at the bull station which passes on to future generations of cows the tendency to emit less methane. This may mitigate the climate impact of milk production.
Animal welfare
Breeding goal: better animal welfare The milk producer can, for example, acquire semen from a bull at the bull station which passes on to future generations of cows better resistance to diseases such as mastitis. This can contribute to the dairy cows having a better animal welfare.
1 Because there is no AI involved in the bull vignettes, the first part of the opening sentence, where the breeding goal was specified to the participant (i.e., “The milk producer can, for example, obtain semen from a bull at the bull station”) was formulated differently as, “The milk producer can, for example, acquire a bull...” for all participants who received a bull vignette.
In the between-subjects design employed, participants were presented with just 1 of the 20 vignettes. The reason for this is that the text length and complexity of information involved in the vignettes were considerable. That raised concerns about respondent fatigue (
). A further advantage of the between-subjects design is that it is not susceptible to confounding issues known to affect within-subject survey experiments such as the carry-over effect and the experimenter demand effect (
). To check whether the 5 breeding methods described to the participants were comprehensible, we asked how well they understood the principles described to them immediately after they had read the vignette. In general, the descriptions were well understood by most participants, both in the pilot data collection (n = 201) and in the actual data collection. In the actual data used in this study, 91% reported that they understood the details of the techniques “very well” or “well,” or that they “understood” or had “an ok sense of the general principles.” Further, there were no statistically significant differences in comprehensibility between the 5 breeding methods (see Supplemental File S3; https://figshare.com/articles/online_resource/Supplemental_File_SIII_https_doi_org_10_3168_jds_2022-22249_docx/21788018;
). Thus, the relatively complex descriptions in AI+OPU/IVP and AI+CLONING/GE were as comprehensible as the less complex techniques.
Acceptance of Breeding Work (Acceptance) (Main Outcome Measure)
This question was given to participants after they had read the vignette text to which they had been randomly assigned: “In my view, this is an unacceptable way to breed dairy cows.” Response options ranged from 1 (strongly disagree), over 4 (neither agree nor disagree), to 7 (strongly agree). We collapsed the responses into 3 categories: 1 acceptance (responses 1, 2, and 3); 2 Undecided (response 4); and 3 Nonacceptance (responses 5, 6, and 7).
WTD Milk from Cows Conceived through the Breeding Work (Main Outcome Measure)
This question was also given to participants after they had read the vignette text: “I am happy to drink milk from cows that are created using this breeding work.” Response options ranged from 1 (strongly disagree), over 4 (neither agree nor disagree), to 7 (strongly agree). We collapsed the responses into a dichotomy: 1 Lack of WTD (responses 5–7) or other = 0 (responses 1–4).
General Attitudes to GE and ART (Secondary Outcome Measures)
Inspired by the approach to eliciting attitudes taken by
, we presented participants with several techniques and asked them to indicate which of them “should not be used to make an existing or future dairy cow pregnant with.” The techniques we prompted about ranged from AI to cloning.
Self-Reported Knowledge about Dairy Breeding Methods
Participants were asked to report their level of knowledge about breeding in dairy production. They were also asked about the extent to which 4 fertilization methods, namely, a bull mates a cow or heifer, AI, AI with the use of genetic information, and more advanced methods (embryo transfer) are used by Danish dairy farmers. Response options ranged from “not at all” to “to a large extent,” together with a “don't know” response.
Dairy Consumption (Grams Per Day)
Participants who consumed milk, yogurt, whipped cream, or sour milk products were asked to specify the amounts they consumed on a monthly basis. We rescaled their responses into grams consumed per day.
Organic Milk Consumption
Participants were asked to indicate the share of milk they drank that is organic. The response options included “0%” and 10 positive responses (“1–10%” ... “91–100%”). There was also a “don't know” option. The 137 participants who gave this “don't know” response were excluded from the analysis. We treated organic milk as a continuous predictor in the multivariate regression analysis. In the subsequent presentation of data patterns, we divided participants into 3 groups (referred to as organic milk consumer groups in Table 3): nonusers (organic share = 0%), medium-frequency users (organic share = 1%–60%), and high-frequency users (organic share ≥ 61%).
Table 3Descriptive statistics on sociodemographic variables, dairy consumption, and organic milk share
The following products are included: light/fat milk, yogurt, sour milk, and whipped cream. Cheese and dairy used in processed products are not included in this calculation.
The following products are included: light/fat milk, yogurt, sour milk, and whipped cream. Cheese and dairy used in processed products are not included in this calculation.
Population data on consumption of dairy is from the most recent dietary survey carried out in Denmark in the period from 2011 to 2013 (Nadelmann Pedersen, 2015).
Data are weighted (see Statistical Analysis section for details) so the result is representative of the Danish population.
(%)
Nonuser (0% organic milk share)
19
Not applicable
Medium frequency (1% to 60% organic milk share)
45
Not applicable
High frequency (>60% organic milk share)
36
Not applicable
1 Shares may sum to more than 100% because of rounding error.
2 Sociodemographic shares are from Statistics Denmark.
3 Data are weighted (see Statistical Analysis section for details) so the result is representative of the Danish population.
4 The following products are included: light/fat milk, yogurt, sour milk, and whipped cream. Cheese and dairy used in processed products are not included in this calculation.
5 Population data on consumption of dairy is from the most recent dietary survey carried out in Denmark in the period from 2011 to 2013 (
This is a measure of discomfort with human interference in the natural world resulting from the belief that nature should be left alone and that such interference invites risks to humans and to nature (
to Danish (using a back-translation procedure). The scale includes statements such as, “Altering nature will be our downfall as a species.” Response options ranged from 1 (strongly disagree) to 7 (strongly agree). The internal consistency of the scale is acceptable in the present data (Cronbach's α = 0.74).
Animal Rights Orientation
This instrument measures attitudes to the notion that animals should be accorded rights (
). It consists of 3 statements (e.g., one was, “In principle, the use of animals by humans is unacceptable because animals are sentient beings”). The response options were 1 (fully disagree) to 5 (fully agree). The internal consistency of the scale is very acceptable in the present data (Cronbach's α = 0.94).
Sociodemographic Measures
We probed participants' age, gender, highest completed level of education, and area of residence in Denmark (NUTS2 level).
Statistical Analysis
A series of regression models were run for each of the 2 main outcome measures (acceptance and WTD milk). In all of the models described below, multinomial logit regression was used to estimate acceptance and logit regression was used to estimate WTD (using the mlogit and logit commands in Stata version 17, respectively).
First, we tested whether advanced breeding methods and breeding goals affect acceptance and WTD. Here we inserted breeding method and breeding goal as main effects. We also inserted their mutual interaction, as this would enable us to examine whether acceptance and WTD for more advanced methods were higher if societally relevant breeding goals linked to animal welfare and positive environmental effects, rather than production, are being pursued. Sample sizes in these analyses were 2,036 in the analysis of acceptance and 1,924 in the analysis of WTD. The sample for WTD was smaller because participants who did not drink milk (including vegans) were removed.
In the second series (baseline model 1), we ran a regression model in which several control variables were included (age, gender, highest completed education, region of Denmark, and dairy consumption) along with the main effects of organic milk consumption, animal rights orientation, and ATN. We then (baseline model 2) inserted interaction effects between organic milk consumption and the ATN and animal rights orientations, respectively. Baseline models 1 and 2 were run with the aim of ruling out possible confounding factors. In the last series (interaction effect models), we added interaction effects between breeding method and the 3 constructs organic milk consumption, animal rights orientation, and ATN, to test whether acceptance and WTD milk are affected differentially by breeding methods at different levels of these 3 variables. We used hierarchical regression to determine significant interaction effects by comparing the likelihood ratio test statistic of the baseline model with models with interaction effects. Interaction effects were inserted and tested one by one using Stata's lrtest command. P-values below 0.05 were treated as statistically significant. We present the character and direction of results using predicted probabilities from Stata's margins and predict command. In this second series of regression, 208 participants were removed for the analysis of acceptance, as they either responded “don't know” when asked about the share of the milk they consume that is organic or could not report a share because they never consume dairy products. That gave a sample size of 1,828. For the same reason, 187 additional participants were removed from the analysis of WTD, giving a total sample of 1,737 in this part of the analysis.
To account for a slight sociodemographic misrepresentation in the data relative to the Danish census (see Table 3), a weight variable was calculated that adjusts the sample to the proportions in the Danish population against 4 variables: age, gender, region in Denmark, and education. Descriptive results (proportions and means) and the first series of regressions are reported with this weight variable activated using Stata's svy command so that the figures are representative of the Danish population.
RESULTS
Response Rate and Nonresponse Analysis
Invitations were issued to 9,421 potential participants. In all, 2,036 completed the questionnaire (response rate 22%). In Table 3 sociodemographic details of the sample are shown, as are differences between the sample and the Danish population census. The sample deviated from the population census at a statistically significant level in both age (P < 0.001) and education (P < 0.001). In general, though, the deviations were modest. Thus there was a 4% point difference for the age categories (the highest deviation observed was for the age bracket 60 yr or older: population share 35%, sample share 31%), and one educational category deviated relatively more than the others—namely, vocational education (population share 32%, sample share 45%). The reported mean amount of dairy products that were consumed by participants in this survey (M = 296 g/d; 95% CI = 0.281 to 0.311) is very close to the mean observed in the most recent Danish dietary survey (M = 304 g/d) (
Self-Reported Knowledge and Beliefs about Dairy Breeding
Most of the surveyed participants (60%) declared that they have “no knowledge at all” about how dairy breeding takes place in Denmark, and 27% said that they have “little knowledge.” Only a minority claimed to have “some” (9%), “a lot” (4%), or “much” knowledge (1%). When participants were asked about the frequency with which breeding techniques are used by dairy farmers (see Table 4), the shares of those who responded “don't know” were quite high (22% to 50%). Many participants assumed that the traditional method of bull mating is uncommon. Thus 48% responded that it is used “to a very small extent” or “not at all.” Most participants (59%) assumed that AI is the most common technique, responding that this method is used “to a large extent” or “very large extent.” Around 1 in 3 responded that AI with the use of genetic information is used “to a large extent” or “very large extent.” Very few believed that more advanced breeding methods such as embryo transfer are common, and a very large share (50%) responded “don't know” to this question.
Table 4Beliefs about how often Danish dairy farmers use 4 types of breeding methods (%; n = 2,036)
General Attitudes to GE and ART in Dairy Cow Breeding
A small proportion of participants (13%) were against the use of AI to fertilize a heifer or cow (see Table 5), whereas approximately 1 in 4 were against sexed semen and in vitro fertilization. Many participants (39%) were against the transfer of fertilized eggs from a cow to foster mother, and even more were against the use of hormone therapy to make a cow produce more eggs (52%). Finally, 58% were against the use of cloning.
Table 5Share of participants who reject breeding methods (n = 2,036)
The data are weighted (see Statistical Analysis section for details) so that the reported frequencies are representative of the Danish population.
Are there any of the breeding techniques that you think should not be used to make a future or current dairy cow pregnant with? (It is possible to check several boxes.)
%
AI: Semen drawn from bulls not present at the milk producer's farm is transferred to the current or future dairy cow.
13
(Sexed semen): AI where it is ensured that a heifer is born: Using genetic information, sperm cells are selected, which ensures that the dairy cow to be pregnant will most likely give birth to a heifer (and not a bull). Fertilization takes place in the same way as AI (point 1).
23
In vitro fertilization and embryo transfer: In a laboratory an egg from a dairy cow is fertilized with semen from a bull. The fetus is then transferred to a foster mother who gives birth to the offspring.
28
Transmission of fertilized eggs from a cow to foster mothers: A cow with particularly good traits is stimulated with hormones (so she produces many eggs) and fertilized with semen drawn from bulls. Next, the fertilized eggs are rinsed out and put into several other cows that then function as foster mothers.
39
Hormone therapy: a dairy cow is given hormone therapy to become more fertile.
52
Cloning: Cells from a dairy cow (or a bull) are cultivated. One of the cells is used to make a fetus that is genetically identical to the dairy cow (or bull). The fetus is then transferred to a foster mother who gives birth to the offspring.
58
No: all the technologies are okay in principle.
22
1 The data are weighted (see Statistical Analysis section for details) so that the reported frequencies are representative of the Danish population.
Effect of Breeding Methods and Goals on Acceptance and WTD
Acceptance varied as a function of the breeding methods and goals with which the participants were presented (P < 0.01).
In Figure 1, the shares of participants who expressed nonacceptance, acceptance, or neither are plotted across the 5 breeding methods. The share agreeing that the breeding method is unacceptable (“This is an unacceptable way to breed cows”) is highest for AI+CLONING/GE (agree = 23%) and lowest for BULL (agree = 15%). The other methods fall in between. Disagreeing that the breeding method is unacceptable was highest for AI+GS (agree = 49%) and lowest for AI+CLONING/GE (agree = 38%). The other methods fall in between. Generally speaking, differences in acceptance across the methods were relatively modest.
Figure 1Acceptance across 5 breeding methods. Reported as percentages (with 95% CI) that were calculated on the basis of predicted probabilities. BULL = bull fertilization; AI+GS = AI in combination with genomic selection; AI+OPU/IVP = combination of ovum pick up and in vitro production with genomic selection of embryos and embryo transfer; AI+CLONING&GE = cells are genetically modified and converted to fetuses using cloning, then transferred using embryo transfer.
In Figure 2, the share of participants who expressed nonacceptance, acceptance, or neither is plotted across the 4 breeding goals. It can be seen that participants were more likely to agree that it is “unacceptable to breed cows” if the goal is to increase milk quantity (agree = 20%) and milk quality (agree = 22%) than they were if the goal is to reduce climate impact or improve animal welfare (agree = 16% in both cases). Additionally, more disagreed that it is unacceptable to breed cows if the goal is to achieve better animal welfare (agree = 49%), whereas the lowest level of disagreement was found here when the breeding goal is to improve milk quality (agree = 39%).
Figure 2Acceptance across 5 breeding goals. Reported as percentages (with 95% CI) that were calculated on the basis of predicted probabilities.
Turning to WTD milk, it was found that this also varied as a function of the breeding methods and goals with which the participants were presented (P < 0.05).
As can be seen from Figure 3, the main difference was found between AI+CLONING/GE and the other methods. The method AI+CLONING/GE creates higher unwillingness: 18% “disagree” that they are happy to drink milk from cows created with semen from this technique, whereas the corresponding figure for the other methods is 10% to 12%. We observed very modest differences in WTD milk across the 4 breeding goals (see Figure 4).
Figure 3Lack of willingness to drink milk across 5 breeding methods. Reported as percentages (with 95% CI) that were calculated on the basis of predicted probabilities. BULL = bull fertilization; AI+GS = AI in combination with genomic selection; AI+OPU/IVP = combination of ovum pick up and in vitro production with genomic selection of embryos and embryo transfer; AI+CLONING&GE = cells are genetically modified and converted to fetuses using cloning, then transferred using embryo transfer.
Figure 4Lack of willingness to drink milk across 5 breeding goals. Reported as percentages (with 95% CI) that were calculated on the basis of predicted probabilities.
For both acceptance and WTD, we did not find a statistically significant interaction effect between breeding goals and breeding method. This means that the lower acceptance and WTD milk associated with advanced methods was not offset when societally relevant breeding goals linked to animal welfare and positive environmental effects were being pursued.
Impact of Animal Rights Orientations, ATN, and Organic Milk Consumption on Acceptance and WTD Milk
Taking acceptance first, it can be seen from the lower part of Table 6 (interaction effect models) that none of the 3 interaction effects that were run were statistically significant.
Table 6Summary of statistically significant effects from multivariable regression analyses
Similarly, with WTD milk there were no significant interaction effects between breeding method and animal rights, and ATN, respectively. However, a significant interaction effect (P < 0.05) was found between breeding method and organic milk consumption. The character of this effect is represented in Figure 5. It is predominantly the high-frequency organic milk consumers who reacted negatively to advanced breeding methods, particularly in the AI+CLONING/GE case, where disagreeing that “I am happy to drink milk...” was highest with an average predicted probability of 28%, compared with 8% to 18% for the other methods. In contrast, breeding method did not affect WTD among nonusers, and it affected WTD very modestly among medium-frequency organic milk consumers.
Figure 5Lack of willingness to drink milk across 5 breeding methods, divided into organic milk consumer groups. Reported as percentages (with 95% CI) that were calculated on the basis of predicted probabilities. BULL = bull fertilization; AI+GS = AI in combination with genomic selection; AI+OPU/IVP = combination of ovum pick up and in vitro production with genomic selection of embryos and embryo transfer; AI+CLONING&GE = cells are genetically modified and converted to fetuses using cloning, then transferred using embryo transfer.
There was also a significant interaction effect between organic milk consumption and ATN (Table 6, baseline model 2). In Figure 6, the pattern of the tripartite effect of breeding method, ATN, and organic milk consumer groups is portrayed. Across all breeding methods, higher aversion (i.e., higher ATN z-scores) affected WTD in the high-frequency organic milk consumers. But the impact was particularly strong with AI+CLONING/GE, where the predicted probability of being unwilling to drink was greater than 0.40 (AI+CLONING/GE) at ATN z-scores above +1.
Figure 6Scatter plot [with fitted lines (lowess)] of aversion to tampering with nature [ATN (z-score)] and unwillingness to drink milk (reported as predicted probabilities) divided into organic milk consumer groups, per breeding method. BULL = bull fertilization; AI+GS = AI in combination with genomic selection; AI+OPU/IVP = combination of ovum pick up and in vitro production with genomic selection of embryos and embryo transfer; AI+CLONING&GE = cells are genetically modified and converted to fetuses using cloning, then transferred using embryo transfer.
In a previous study it was found that, generally speaking, both organic and conventional dairy farmers in Denmark were ready to use advanced breeding technologies. Both routinely used AI, genomic information, and sexed semen, and despite some ethical concerns they were also likely to use semen derived from advanced breeding systems. Further, embryo transfer was used by many conventional farmers (
). However, because consumers may prefer not to buy dairy products from cows produced through technologies such as GE and ART, this study examined public attitudes about this. We focused on an important distinction between different ways in which these techniques are used in practice on dairy farms (namely, directly or indirectly) with the aim of studying whether and how this distinction shapes consumer demand and attitudes.
This perspective generated surprising results. When the study participants were presented with breeding methods without further specification of their use, the rejection rates were quite high for the most advanced techniques: cloning (58%), transfer of fertilized eggs (39%), and hormone therapy (52%). Hence, it would appear that many Danes are concerned about, and critical of, these techniques in general terms. In contrast, acceptance and WTD milk were considerably higher when participants were given details in the vignette material about the ways in which similar advanced techniques were used indirectly in the dairy farmers' breeding work (i.e., through the use of semen developed through the breeding industry's use of the advanced techniques on earlier generations of cows and bulls). We varied the vignettes so that farmers' indirect use of advanced breeding techniques covered the entire spectrum of current and (expected) future methods insofar as GE and reproductive techniques are concerned: from natural fertilization (bull), through AI, to the use of semen derived from breeding schemes where egg transfer to foster mothers, genomic information, cloning, and GE are used. Surprisingly, our data suggest that Danish consumers were not substantially more resistant to milk from AI+OPU/IVP breeding, in which IVP, genomic information, and transfer of fertilized eggs are involved (nonacceptance: 20%), than they were to the presumably most natural breeding option that is available nowadays: BULL (nonacceptance: 15%). Similarly, there were no differences between reactions to these 2 methods in terms of WTD milk. Further, although rejection of the AI+CLONING/GE breeding technique, where cloning and GE are used, was higher, it was not greatly so.
There are several possible explanations of the low rates of rejection we observed when participants were asked about dairy farmers' indirect use of advanced methods. First, Denmark could be an exception. Perhaps in other countries where milk products are consumed people are less ready to accept advanced breeding methods, including GE. This could be connected with the high levels of institutional trust that characterize citizens in Denmark and adjacent countries (
). In our view, this is not a plausible explanation. In the case of animal cloning, citizens in Sweden, a country where institutional trust also is very high, are more critical than Danes, whereas citizens in Eastern European countries, where trust is lower, are less critical (
European Commission. 2008. Europeans' Attitudes towards Animal Cloning. Conducted by the Gallup Organization, Hungary, upon the request of Directorate General Health and Consumers, European Commission. Food and Drug Administration. Flash EB Series No. 238, Eurobarometer.
). Earlier studies have shown that Denmark lies between the most and least critical countries as regards acceptance of GE in food and the cloning of animals in food production (
European Commission. 2008. Europeans' Attitudes towards Animal Cloning. Conducted by the Gallup Organization, Hungary, upon the request of Directorate General Health and Consumers, European Commission. Food and Drug Administration. Flash EB Series No. 238, Eurobarometer.
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
A second potential explanation is that with time people are simply becoming less apprehensive about and critical of ART and GE. This may be a side effect of the increased use of in vitro fertilization and egg transfer in humans (
Acceptance of a food of animal origin obtained through genetic modification and cloning in South America: A comparative study among university students and working adults.
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
)—and thus they may not reflect this development. Whether criticism in fact is decreasing has not been studied recently in comprehensive international survey programs such as the International Social Survey Programme or Eurobarometer. We will have to await future studies to check the general validity of this explanation. What we can conclude at this point is that the share of Danes in this 2021 study who were critical of GE and reproductive technologies was markedly smaller than the share of Germans observed in 2012 (
) (in both studies, the participants were asked about their general attitudes to the use of these techniques). The rejection rates were as follows: sexed semen (Denmark, 23%; Germany, 53%), in vitro fertilization and embryo transfer (Denmark, 28%; Germany, 58%), hormone therapy (Denmark, 52%; Germany, 65%), and cloning (Denmark, 58%; Germany, 81%;
). This difference could be ascribed to a general decline in public skepticism. Alternatively, it may reflect attitudinal differences between the 2 countries.
A third possible explanation of the low rate of rejection is that the description of breeding goals given in the vignettes enabled the participants to see the justification and perceived benefits of the advanced breeding methods. It has been shown that perceived consumer benefits increase acceptance of GE technologies, although less so for GE used in relation to animals (
A fourth likely explanation is the distance of GE from the consumer product. Grunert and colleagues found that perceptions and acceptance of GE food products vary along a distance dimension, and that there is higher acceptance when GE applications only have been used as an aid in the processing of products and are not directly present in the final product (
). In line with this, Danes could have only limited concerns about the specific way in which Danish dairy farmers will rely on GE and ART in their breeding work at the farm because the technologies are regarded as rather distant from the offspring that produce the milk destined for human consumption. Against this, it should be noted that 2 other studies that also asked participants whether they would drink milk from the offspring of cloned cows found significantly higher levels of unwillingness to drink the resulting milk than the present study found (
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
). However, data collection took place more than a decade ago in these 2 other studies.
We find it likely—but cannot demonstrate this further using our data—that the fourth explanation, referring to the distance dimension, is the most powerful contributor to the low rejection rates observed for the more advanced breeding methods in our study. Our main argument for this is that rejection dropped considerably in the vignettes when the at-farm breeding work was explained in detail as compared with the levels observed when the participants were asked about general attitudes to the use of GE and ART. We assume that the second explanation (referring to a historical trend toward public acceptance of advanced technologies) and the third explanation (referring to the reassurance given when beneficial breeding goals are described) contribute less. Future studies could be designed to disentangle effects from the third and fourth explanations.
We were unable to confirm our expectation that the lower acceptance and WTD milk produced using advanced methods would be moderated to some extent when consumers know that societally relevant breeding goals linked to animal welfare and positive environmental effects are being pursued. In general, the production-oriented goals (milk quantity and milk quality) were only accepted slightly less than the 2 other goals (climate impact and animal welfare). This contrasts with other studies of attitudes to GE in dairy farming, where rejection varies considerably as a function of the breeding goal (
). It is difficult to compare these studies directly, though, because they vary in experimental setup and in the detail and kinds of description provided to their participants.
Some people believe that GE and advanced technologies violate the basic rights and integrity of animals (
). We therefore anticipated that consumers with strong animal rights views would exhibit lower levels of acceptance and lower WTD milk from cows inseminated with the semen of bulls produced with advanced breeding methods. However, we did not confirm this. We did identify a main effect from animal rights, though. This suggests that stronger animal rights orientations lead to criticism of any kind of breeding strategy, presumably because those who believe in the rights of animals often take the view that the breeding of farm animals (including bull breeding) is an illegitimate use of them that violates their rights.
We also expected that people displaying high levels of discomfort with human tampering with the natural world would have lower levels of acceptance of, and lower WTD milk produced from, advanced breeding methods (
). Again, we failed to obtain direct support for this. However, this was because our expectation here was too general. Aversion to tampering with nature is important, but predominantly it resonates with high-frequency organic consumers, as discussed further below.
Organic consumers have been shown to prefer “pure” products (
Healthy food is nutritious, but organic food is healthy because it is pure: The negotiation of healthy food choices by Danish consumers of organic food.
). So we expected users of organic milk to have lower levels of acceptance of, and lower WTD milk produced with, advanced breeding methods. We found partial support for this, in that the expectation was confirmed for WTD but not acceptance. Specifically, high-frequency organic milk consumers are more likely not to drink milk where AI+CLONING/GE has been used, as compared with the other methods. In this group, the levels of rejection also increase in tandem with rising ATN, particularly when the milk is derived from cows bred using semen developed through the breeding industry's use of advanced technologies on earlier generations of cows and bulls. However, this pairing occurs predominantly at ATN z-scores above +1. Further analysis (data not shown) reveals that Danes with this profile (high-frequency organic consumers and ATN z-scores greater than +1) are relatively rare, making up approximately 6% of all milk consumers and approximately 16% of high-frequency organic consumers.
The low response rate (22%) is a potential limitation of the study. However, limited response rates are problematic only if there is nonresponse bias (
). Our data sample included statistically significant deviations from the target population in age and education. These deviations were accounted for in all of the descriptive results reported in the paper through the use of a weight variable that adjusted the sample to ensure that it reflected the Danish population in terms of age, education, gender, and place of residence. We found that average levels of dairy consumption in the study were quite similar to those in census data. So although we cannot rule out nonresponse bias, the sample appears to portray the target population relatively well as regards milk consumption. The cost of genetically engineered products may affect the likelihood of their being purchased (
). Hence, it is a further limitation that we did not investigate the effect of cost by, for example, including a willingness-to-pay (WTP) element. We decided against including costs because the price ranges to be included, and hence the derived estimations of WTP, would have been highly speculative, as there are so far no realistic product prices to take as a guide in constructing such an analysis. Nevertheless, WTP methodology could have provided a complementary approach to the evaluation of technology acceptance examined in this paper. We recommend that future studies include a WTP perspective if possible.
Restrictions imposed by data collection costs meant we were unable to include a more comprehensive set of potentially relevant consumer traits. For instance, food choice motives (
) could influence acceptance of breeding methods and WTD. Similarly, environmental concerns might do so. Many questionnaire-based constructs reflecting environmental concern include questions, or a subdimension, focusing on harmony with, and respectful treatment of, nature (e.g.,
). But the constructs typically include additional components as well. For instance, the New Environmental Paradigm measure includes subdimensions related to the balance of nature, limits to growth, and the right of human beings to rule over nature (
). Measures of environmental concern, therefore, often conflate the dimension that was of primary interest in this study (i.e., concern over human tampering with nature) with other dimensions. This is why we opted to use the newly developed ATN (
). It exclusively measures discomfort with (or aversion to) tampering with nature and thus more specifically taps into consumer preferences for products with different levels of “naturalness.” Finally, as the technological applications studied in this paper focus on the control and alteration of animals, we speculated that it would be more appropriate to include a measure of animal rights (
) rather than a more general construct that measures the right of human beings to rule over nature (such as the third subdimension in the New Environmental Paradigm).
Among the many products derived from cows' milk, we chose to focus on WTD fluid milk (as opposed to, for example, yogurt, sour milk, ice cream, butter, and cheese) in the vignette experiment. We did so because this product is particularly highly loaded with symbolic meaning. In many European countries, fluid milk developed historically as a product perceived as a necessary foodstuff (
). We therefore speculated that any critical perceptions of advanced methods would be most clearly discerned in the consumer's willingness to consume fluid milk. Of course, this means we cannot draw any conclusions about the way willingness to consume other dairy products varies as a function of breeding methods and goals, and certainly future studies could usefully focus on the acceptance of breeding technologies across a wider variety of dairy products.
Turning to the implications of the study, our findings suggest that the most common breeding strategy in Danish dairy farms, AI+GS (
), is only rejected by a small proportion of Danish citizens, and that there is likely to be limited public criticism and rejection of milk produced using semen developed through the breeding industry's use of advanced technologies on earlier generations of cows and bulls. As Danes are on the whole neither critical nor uncritical of GE and cloning (
European Commission. 2008. Europeans' Attitudes towards Animal Cloning. Conducted by the Gallup Organization, Hungary, upon the request of Directorate General Health and Consumers, European Commission. Food and Drug Administration. Flash EB Series No. 238, Eurobarometer.
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
), the results are likely to be transferable to other European countries, although the more surprising aspects of the study results should be confirmed in other countries.
Although our results will be considered good news by the dairy sector, that positive evaluation is only warranted if the context in which dairy farming currently operates persists. One important contextual factor here is the generally rather limited public understanding of dairy animal breeding methods observed in this study, with 60% of participants stating that they have no knowledge at all about dairy breeding. It is unclear how attitudes would develop if people were to become more engaged in and knowledgeable about GE and reproductive technologies, both in their general use and specifically in their applications in livestock production. However, it is surely likely that many of those who are undecided about the use of the breeding technologies (35% to 40% of our participants neither agreed nor disagreed that the methods they were asked to consider were an “unacceptable way to breed cows”) would develop a more positive or negative stance. It is not impossible, of course, that a subgroup, or even the majority, of these undecided citizens would become more critical of advanced breeding. Given that this is a possibility, to avoid potential moral outrage (
), the breeding industry is urged to ensure—and be able to demonstrate if called upon—that the use of GE and ART such as IVP and embryo transfer does not negatively affect animal welfare, or introduce phenotypical changes to animals that appear to alter their nature, or introduce changes posing a risk to consumers or the environment.
Organic dairy farmers and their associations should pay particular attention to high-frequency organic milk consumers (about a third of all milk consumers in Denmark today). Although outrage and comprehensive rejection of milk produced with advanced technologies in this high-frequency group are unlikely, it is important, when dealing with the organic community, to take into account those who are highly averse to human tampering with nature, as this group is more likely to reject milk from cows developed in breeding schemes where cloning and GE have been used on earlier generations of cows and bulls. Organic dairy breeding presently relies on the conventional breeding industry only, in part because inbreeding and the lack of genetic gain are critical concerns in a purely organic breeding system (
). Currently, there is little or no transparency about the technologies used in the production of the semen provided to Danish and European dairy farmers. Although both cloning and the GE of farm animals are prohibited in the EU (
European Union. 2007. Council Regulation (EC) No 834/2007 of 28 June 2007 on organic production and labelling of organic products and repealing Regulation (EEC) No 2092/91.
), there is no monitoring of, or legislation on, imports of semen from advanced breeding systems produced outside the EU. Thus organic dairy farmers could be importing and using semen produced with the aid of these technologies. For this reason, the organic community should consider insisting on more traceability for the semen that is used by organic dairy farmers.
ACKNOWLEDGMENTS
The study presented in this paper was conducted as part of a strategic research project (funded by Innovation Fund Denmark, Aarhus, Denmark: 7045-00013B; and Danish Milk Levy Fund, Aarhus, Denmark) focusing on OPU-IVP-GS technology. We thank the project partners for comments on our research design and on an early version of the questionnaire. We also thank Eliza Ruiz Izaguirre (Department of Food and Resource Economics, University of Copenhagen, Copenhagen, Denmark) for help with a literature search, and Paul Robinson (Verbor Editorial Services, Bath, United Kingdom) for language editing. Partners include the private companies VikingGenetics, Trans Embryo Genetics, and EmbryoTrans Biotech. These partners had no role in conducting the study or in interpreting its results. The authors have not stated any other conflicts of interest.
REFERENCES
Aguinis H.
Bradley K.J.
Best practice recommendations for designing and implementing experimental vignette methodology studies.
Healthy food is nutritious, but organic food is healthy because it is pure: The negotiation of healthy food choices by Danish consumers of organic food.
European Commission. 2008. Europeans' Attitudes towards Animal Cloning. Conducted by the Gallup Organization, Hungary, upon the request of Directorate General Health and Consumers, European Commission. Food and Drug Administration. Flash EB Series No. 238, Eurobarometer.
European Union. 2007. Council Regulation (EC) No 834/2007 of 28 June 2007 on organic production and labelling of organic products and repealing Regulation (EEC) No 2092/91.
Gaskell G. Europeans and Biotechnology in 2010: Winds of Change? A Report to the European Commission's Directorate-General for Research. European Commission,
Brussels2010
Acceptance of a food of animal origin obtained through genetic modification and cloning in South America: A comparative study among university students and working adults.
00011-5&id=gr1.jpg){kind=link}
00011-5&id=gr2.jpg){kind=link}
00011-5&id=gr3.jpg){kind=link}
00011-5&id=gr4.jpg){kind=link}
00011-5&id=gr5.jpg){kind=link}
00011-5&id=gr6.jpg){kind=link}