Recently, the interest in wood products and bioenergy applications of transgenic forest trees is increasing worldwide, though plantations have been established to this purposes only in China. Information on the anticipated attitudes of consumers towards products from genetically-modified forest trees would therefore be of a particular interest both for developers and policy makers. This study investigated the purchasing behavior of potential Greek consumers towards the products from transgenic forest trees. In 2011, a survey was conducted based on randomly selected interviews of 418 potential consumers from all over Greece. Principal Components Analysis (PCA) was performed to identify the main factors affecting the potential purchasing behavior of consumers towards products from transgenic forest trees. Hierarchical and non- hierarchical cluster analysis was applied to PCA scores to identify homogeneous groups of consumers sharing a similar purchasing behavior. Discriminant analysis was used to cross-validate cluster membership of consumers based on PCA factors. Four groups of consumers showing similar potential purchasing behavior towards the products of transgenic forest trees were identified: (a) those interested in the quality of products; (b) those oriented towards lower prices; (c) those influenced by curiosity and labeling issues; and (d) consumers mainly interested in health safety issues and environmental impacts. Finally, a most frequent profile for each group of consumers was outlined according to their demographic characteristics and their opinions on the use of transgenic-tree derived products. Although it is unlikely that products from GM forest trees will be marketed in the next 10 to 15 years, information on the anticipated attitudes of consumers has to be taken into consideration by the developers and policy makers.
The global consumption of forest products, such as wood, paper and woody biomass energy, has been rapidly expanding between 1965 and 2007. For example, the annual average rate for consumption of paper products increased approximately by 3% in the above period. With the global economic crisis started in 2008, the consumption of paper products declined by 2-3% in countries like the USA, Canada and the EU (
Forest wood biomass is an important renewable resource that addresses multiple energy needs in the form of firewood, chips, charcoal, briquettes and pellets, as well as feedstock for the biofuels industry (
Transgenic forest trees are Genetically Modified (GM) through the insertion/deletion of specific exogenous DNA sequences in order to manifest specific traits (http://www.forestguild.org), such higher growth rates, lignin reduction and increased resistance to herbicides or forest pests. Despite the extensive research carried out on GM trees in the past 20 years, there is no currently market input for transgenic trees in Europe, the USA or other parts of the world, with the only exception of China (
Public acceptance and the behavior of consumers towards forest products from GM trees mainly reflect people’s concerns on the establishment of plantations of transgenic forest trees. Some concerns involve the potential spread of antibiotic and/or herbicide resistance genes from GM trees to other non-target species, the potential for long-distance pollen spread from long-lived trees, as well as the potential adverse, unexpected and unpredicted effects on biodiversity (
Although numerous studies have been carried out on public attitudes towards GMOs, no investigations have been carried out in Europe on the potential behavior of consumers towards products from transgenic forest plantations. Previous studies conducted in the USA indicated that consumers are concerned about purchasing final wood products from GM trees, though they generally show a positive attitude towards transgenic forest trees (
Although it is unlikely that products from GM forest trees will be marketed in the next 10 to 15 years (
This paper aimed at surveying the attitude and the potential purchasing behavior of Greek consumers towards products derived from transgenic forest tree plantations. The profile of potential consumers according to their demographic characteristics and their response towards the establishment of such plantations has been also investigated.
A conceptual model was developed to place key concepts into an identifiable framework (
The starting null hypotheses for this study were:
H01: consumers cannot be classified into groups according to their potential purchasing behavior towards products from transgenic forest trees;
H02: opinions of consumers towards the establishment of transgenic forest plantations are not related to a particular potential purchasing behavior (
H03: demographic characteristics of consumers are not significantly related to their potential purchasing behavior.
A survey with individual face-to face interviews have been carried out throughout Greece. Cluster sampling was used to form a representative sample of the whole Greek population. To obtain a geographically-balanced representative sample, nine out of the thirteen existing regions of Greece were selected for sampling, according to the methodology proposed by
A pilot survey was also conducted preliminarily to test if the questionnaire was properly designed to match the research objectives (
The main survey was conducted on November throughout December 2011 by interviewers from each selected prefecture. The sample size was adjusted based on the indications obtained from the pilot study, following the methodology proposed by
To extract the key information from the dataset made up with the responses of 418 consumers, multivariate analysis techniques were applied. Principal Component Analysis (PCA) was used to identify those variables accounting for the largest amount of variance within the dataset. To search for the smallest set of such variables, redundancy between variables was checked by applying the Bartlett’s test of sphericity and the Measure of Sampling Adequacy (MSA) to the correlation matrix. Variables showing a high proportion of large absolute correlation values and an MSA index < 0.5 were discarded from further analyses.
An orthogonal rotation (varimax method) was conducted and standard criteria (eigenvalue = 1, scree test and percentage of variance) were used in order to determine the factors in the first rotation (
PCA scores obtained from the above analysis were subjected to cluster analysis to classify the consumers with similar purchasing behavior into homogeneous groups. Both hierarchical and non-hierarchical methods were applied (
Quadratic Discriminant Analysis (QDA) was performed to assess whether the key factors identified through the factor analysis could accurately predict and discriminate cluster membership. Furthermore, the Friedman’s one-way test was performed to identify the relationship between the consumers’ opinion regarding the establishment of transgenic forest plantations and their particular purchasing behavior. Finally, the chi-square analysis and the Fisher’s exact test were carried out on contingency tables using the software package SUAUJCUJDA® (StatSoft Inc., Tulsa, OK, USA) to test for differences in demographic characteristics among the groups of consumers identified. Finally, the most frequent profile of each group of consumers based on their demographic characteristics was obtained.
Multivariate analysis was used to identify the key information from the 418 responses to the questionnaire used in the survey. The list of variables affecting the potential purchasing behavior of Greek consumers towards products from GM forest trees are reported in
Based on the cluster analysis carried out on PCA scores, consumers appeared to split into four groups according to their purchasing behavior towards products from GM trees (
The consumers interested in the quality of products from GM trees (group a) comprised 12 % of the respondents and were mainly influenced by the quality of the products and their advertisement. They were also interested in the potential negative environmental impacts of transgenic forest plantations, as well as in labeling issues regarding the indication of the transgenic origin of products. The consumers orientated towards lower prices (group b) comprised 30% of the sample. The potential low prices of such products, their advertisement and brand name, as well as health safety issues were the main factors affecting their potential behavior in purchasing the products of GM forest trees. Consumers influenced by labeling and curiosity issues (group c) were about 35% of the respondents. Most of these consumers declared that they were willing to purchase the products from GM forest trees mainly for curiosity reasons. However, they would like the transgenic origin of products to be indicated on labels. The potential low prices and the brand name of the transgenic products have a significant impact on the purchasing decision of these consumers. Finally, 23% of the respondents could be grouped as consumers who are interested in health safety issues and the possible negative environmental impacts (group d). No other factors were found to affect their potential purchasing behavior towards the products of transgenic forest trees.
Discriminant analysis was performed on the scores from the factor analysis carried out to assess the reliability of the groups obtained by the cluster analysis. A summary of such cross-validation analysis is presented in
The Friedman’s non-parametric test was used to explore the opinion of consumers on the establishment of transgenic plantations (
Overall, the first three groups of consumers (a, b, c) paid more attention to the economic impacts of the products from GM forest trees, and were moderately concerned of the possible environmental impacts associated with their production. That is, these potential consumers are willing to buy transgenic forest products, because their purchasing behavior is mainly driven by economic issues, such as price, quality, labeling and branding. These findings coincide with the results of previous studies dealing with the purchasing behavior of consumers towards non-transgenic forest products (
In order to outline the most frequent profile of the respondents based on their demographic characteristics, chi-square and Fisher’s exact tests were also applied on each consumer group previously identified, with the aim of testing for possible differences among groups for each characteristic.
Thus, our starting hypothesis H03: “Consumers’ demographic characteristics are not significantly related to a potential purchasing behavior” could be rejected.
This study contribute to better understand the potential purchasing behavior of consumers towards products from plantations of transgenic forest trees. In particular, our results indicated that there might be a potential market in Greece for products originated from transgenic forest plantations.
Four groups of potential consumers of products from GM trees with similar purchasing behavior were distinguished: (a) those interested in the quality of products; (b) those oriented towards lower prices; (c) those influenced by curiosity and labeling issues; and (d) consumers interested in health safety issues and environmental impacts.
In general, most potential consumers of transgenic forest tree products in Greece showed a purchasing behavior driven by economic issues (price, quality, labeling and branding). Therefore, there is the potential for market development of such products, that are not directly linked with human health impacts.
Although the products from plantations of transgenic forest trees are not expected to be commercialized in the nearby future, the scientifically-based information collected in this study on the anticipated attitudes of consumers may help future decision of both developers and policy makers dealing with transgenic forest tree products.
This study was initiated within the frame of the EU COST Action FP0905 (http://www.cost-action-fp0905.eu).
The Conceptual Model used for the classification of the potential purchasing behavior of Greek consumers towards products derived from transgenic forest trees.
Results of the principal component analysis carried out on the responses of 418 Greek consumers, with the variables affecting their potential purchasing behavior towards the products from transgenic forest trees.
PCAComponent | Eigenvalue | Variance(%) | CumulativeVariance (%) | Variable | Communalities |
---|---|---|---|---|---|
1 | 4.785 | 17.088 | 17.088 | V1 Low price for transgenic origin paper products | 0.583 |
2 | 4.168 | 14.887 | 31.976 | V2 Quality of transgenic origin paper products | 0.668 |
3 | 2.282 | 8.150 | 40.126 | V3 Brand name of transgenic origin paper products | 0.583 |
4 | 1.768 | 6.313 | 46.439 | V4 Certification regarding the place of origin of the transgenic origin of paper products | 0.541 |
5 | 1.526 | 5.450 | 51.888 | V5 Labeling regarding the transgenic origin of paper products | 0.594 |
6 | 1.355 | 4.838 | 56.726 | V6 Health safety issues regarding the transgenic origin paper products | 0.508 |
7 | 1.232 | 4.398 | 61.125 | V7 Possible negative environmental impact of the transgenic tree plantation used for paper production | 0.753 |
8 | 1.078 | 3.850 | 64.974 | V8 Advertisement of transgenic origin paper products | 0.740 |
9 | 0.961 | 3.432 | 68.407 | V9 Consumers’ curiosity regarding the transgenic origin paper products | 0.673 |
10 | 0.935 | 3.339 | 71.746 | V10 Low price for transgenic origin wood products | 0.710 |
11 | 0.803 | 2.869 | 74.615 | V11 Quality of transgenic origin wood products | 0.635 |
12 | 0.688 | 2.456 | 77.071 | V12 Brand name of transgenic origin wood products | 0.682 |
13 | 0.655 | 2.339 | 79.410 | V13 Certification of the place of origin of transgenic origin wood products | 0.608 |
14 | 0.617 | 2.203 | 81.613 | V14 Labeling regarding the transgenic origin of wood products | 0.686 |
15 | 0.588 | 2.100 | 83.712 | V15 Health safety issues regarding the transgenic origin wood products | 0.655 |
16 | 0.543 | 1.940 | 85.652 | V16 Possible negative environmental impact of the transgenic tree plantation used for wood products production | 0.782 |
17 | 0.487 | 1.739 | 87.392 | V17 Advertisement of transgenic origin wood products | 0.736 |
18 | 0.443 | 1.583 | 88.975 | V18 Consumers’ curiosity regarding the transgenic origin wood products | 0.730 |
19 | 0.431 | 1.540 | 90.515 | V19 Low price for transgenic origin woody biomass products | 0.647 |
20 | 0.409 | 1.460 | 91.975 | V20 Quality of transgenic origin woody biomass products | 0.602 |
21 | 0.353 | 1.262 | 93.237 | V21 Brand name of transgenic origin woody biomass products | 0.625 |
22 | 0.352 | 1.259 | 94.496 | V22 Special characteristics of transgenic origin woody biomass products | 0.580 |
23 | 0.327 | 1.168 | 95.664 | V23 Certification of the place of origin of transgenic origin woody biomass products | 0.550 |
24 | 0.305 | 1.089 | 96.752 | V24 Labeling regarding the transgenic origin of woody biomass products | 0.625 |
25 | 0.274 | 0.980 | 97.732 | V25 Health safety issues regarding the transgenic origin woody biomass products | 0.630 |
26 | 0.234 | 0.835 | 98.567 | V26 Possible negative environmental impact of the transgenic tree plantation used for woody biomass products | 0.674 |
27 | 0.218 | 0.779 | 99.346 | V27 Advertisement of transgenic origin woody biomass products | 0.715 |
28 | 0.183 | 0.654 | 100.00 | V28 Consumers’ curiosity regarding the transgenic origin woody biomass products | 0.679 |
Key factors and dimensions affecting the potential purchasing behavior of Greek consumers towards the products from transgenic forest trees.
Factor | Key attitude dimensions | Factor Loadings |
---|---|---|
Labeling | Labeling of wood products | 0.788 |
Labeling of paper products | 0.716 | |
Labeling of woody biomass products | 0.682 | |
Certification of the place of origin of wood products | 0.674 | |
Certification of the place of origin of paper products | 0.610 | |
Certification of the place of origin of woody biomass products | 0.564 | |
Health Safety Issues | Health safety issues regarding the woody biomass products | 0.732 |
Health safety issues regarding the wood products | 0.657 | |
Health safety issues regarding the paper products | 0.635 | |
Brand Name | Brand name of woody biomass products | 0.736 |
Brand name of wood products | 0.544 | |
Brand name of paper products | 0.534 | |
Special characteristics of transgenic paper products | 0.490 | |
Possible Negative Environmental Impacts | Possible negative environmental impact of the transgenic tree plantation used for wood products | 0.864 |
Possible negative environmental impact of the transgenic tree plantation used for paper products | 0.842 | |
Possible negative environmental impact of the transgenic tree plantation used for woody biomass products | 0.701 | |
Low Price | Low price for wood products | 0.802 |
Low price for woody biomass products | 0.758 | |
Low price for paper products | 0.707 | |
Advertisement | Advertisement of paper products | 0.806 |
Advertisement of wood products | 0.753 | |
Advertisement of woody biomass products | 0.609 | |
Quality | Quality of paper products | 0.724 |
Quality of wood products | 0.629 | |
Quality of woody biomass products | 0.569 | |
Curiosity | Consumers’ curiosity regarding the wood products | 0.758 |
Consumers’ curiosity regarding the woody biomass products | 0.731 | |
Consumers’ curiosity regarding the paper products | 0.653 |
Classification of consumers according to their purchasing behavior towards products from transgenic forest trees. (Prod. Quality): Consumers interested in the product’s quality; (Lower Prices): Consumers orientated towards lower prices; (Curiosity/Labeling): Consumers influenced by curiosity and labeling issues; (Health/Environ.): Consumers interested in health safety issues and environmental impact.
Factor | Group of consumers(Cluster centers) | Prob | |||
---|---|---|---|---|---|
Prod.Quality | LowerPrices | Curiosity/Labeling | Health/Environ. | ||
Labeling | 0.09730 | -0.46515 | 0.40216 | -0.04300 | 0.001 |
Health safety issues | -1.15498 | 0.48819 | -0.40129 | 0.57287 | 0.001 |
Brand name | -0.78262 | 0.24192 | 0.10984 | -0.08375 | 0.001 |
Potential environmental impact | 0.19923 | -0.12281 | -0.19710 | 0.36740 | 0.001 |
Low price | -0.36624 | 0.60421 | 0.29574 | -1.08728 | 0.001 |
Advertisement | 0.35850 | 0.34692 | -0.17698 | -0.38821 | 0.001 |
Quality | 1.00944 | 0.33586 | -0.58081 | -0.09216 | 0.001 |
Curiosity | -0.10825 | -0.42322 | 0.59287 | -0.28697 | 0.001 |
|
|
|
|
|
|
Summary of the discriminant analysis carried out on PCA scores for cross-validation purposes (total N = 418, N correct = 409, proportion correct = 97.8%). For the labels of the groups of consumers, see
Actual classification | Predicted classification of consumers | |||
---|---|---|---|---|
Prod.Quality | LowerPrices | Curiosity/Labeling | Health/Environ. | |
Prod. Quality | 48 | 0 | 0 | 2 |
Lower Prices | 0 | 125 | 1 | 1 |
Curiosity/Labeling | 0 | 1 | 146 | 0 |
Health/Environ. | 2 | 2 | 0 | 91 |
Total N | 50 | 128 | 146 | 94 |
N Correct | 48 | 125 | 145 | 91 |
Prop. of Correct Classification | 96.0% | 97.6% | 99.2% | 96.8% |
Consumers’ opinions towards the establishment of transgenic forest plantations. Numbers are the average ranks on the Likert scale adopted. For the labels of the groups of consumers, see
Opinions | Prod.Quality | LowerPrices | Curiosity/Labeling | Health/Environ. |
---|---|---|---|---|
It will contribute to an increase of job opportunities | 6.61 | 6.51 | 7.24 | 5.97 |
It will contribute to a production cost reduction | 6.42 | 7.13 | 6.51 | 5.97 |
It will contribute to an increase of farmers’ income | 6.33 | 6.41 | 6.07 | 6.11 |
It will contribute to reduction of production losses | 6.25 | 6.61 | 5.35 | 5.47 |
It might have negative impacts on the environment | 5.30 | 5.63 | 5.67 | 6.34 |
It might harm biodiversity and ecosystems | 6.02 | 6.07 | 5.78 | 6.27 |
It might have negative impact on wild native plants | 6.49 | 5.66 | 6.35 | 6.40 |
It might have negative impacts on human health | 5.41 | 5.67 | 6.03 | 7.22 |
It is not necessary | 5.41 | 4.96 | 5.70 | 5.07 |
It might contribute to climate change mitigation | 5.93 | 5.01 | 6.08 | 5.29 |
It is important for biomass production | 5.86 | 6.34 | 5.22 | 5.88 |
Profile of each consumer group regarding their demographic characteristics. Data reported are the percentages of the total respondents. For the labels of the groups of consumers, see
Group | Demographiccharacteristics | Prod.Quality | LowerPrices | Curiosity/Labeling | Health/Environ. |
---|---|---|---|---|---|
Education χ2[12] = 30.019, p<0.01; Fisher’s exact test 30.254, p<0.01 | Primary school | 10 | 13.3 | 6.2 | 7.4 |
Secondary school | 24 | 19.5 | 34.2 | 10.6 | |
High school | 34 | 36.7 | 39.7 | 45.7 | |
Bachelor degree | 32 | 27.3 | 16.4 | 33 | |
Postgraduate degree | 0 | 3.1 | 3.4 | 3.2 | |
Number of children χ2[6] = 14.150, p<0.05; Fisher’s exact test 13.597, p<0.05 | No children | 46 | 33.6 | 48.6 | 33 |
1-2 children | 34 | 37.5 | 32.2 | 48.9 | |
3+ children | 20 | 28.9 | 19.2 | 18.1 | |
Age χ2[12] = 41.037, p<0.001; Fisher’s exact test 38.131, p<0.001 | 20-29 | 10 | 25.8 | 8.2 | 18.1 |
30-44 | 18 | 28.1 | 34.9 | 33 | |
45-64 | 26 | 19.5 | 28.1 | 34 | |
65+ | 46 | 26.6 | 28.8 | 14.9 | |
Occupation χ2[18] = 52.699, p<0.001; Fisher’s exact test 50.077, p<0.001 | Private employee | 12 | 18.8 | 14.4 | 18.1 |
Civil servant | 30 | 20.3 | 43.8 | 20.2 | |
Free License | 10 | 11.7 | 4.8 | 21.3 | |
Retiree | 34 | 25.8 | 27.4 | 16 | |
Student | 8 | 12.5 | 6.2 | 11.7 | |
Unemployed | 6 | 7.8 | 3.4 | 11.7 | |
Other | 0 | 3.1 | 0 | 1.1 |