The impact of environmental certification on preferences for wood furniture: a conjoint analysis approach.
The objectives of this exploratory study were to: 1) determine the relative importance of five wood CD rack product attributes; and 2) identify demographic and psychographic variables associated with the respondents who rated environmental certification as the most important attribute in forming their product preferences. To meet these objectives, we surveyed a convenience sample of 265 Oregon State University undergraduate students enrolled in an international business class during the 2001/2002 academic year. The results indicate that respondents viewed environmental certification as a favorable product attribute. However, for the typical respondent, the importance of other product attributes outweighed that of environmental certification. Despite environmental certification's limited importance for the typical respondent, it was the most important attribute for 20.8 percent of the respondents. Backwards-stepwise binary logistic regression was used to identify four variables useful for predicting which respondents would place the highest importance on the attribute environmental certification. A key finding was that willingness to pay more for certified forest products (CFPs) was highest among those who placed the greatest importance on environmental certification. Although our sample was not representative of the general population, these results provide insight about possible marketing implications. First, since environmental certification was a relatively unimportant purchase decision criterion for the average respondent, CFPs marketed in mainstream distribution channels, e.g., big-box retailers, are not likely to realize price premiums. However, since willingness to pay was greatest among those who placed the highest importance on environmental certification, CFP price premiums may be possible through market segmentation.
The Green Gauge Report, published annually by the market research firm Roper ASW, documents the public's attitude toward the state of the natural environment. Roper ASW (2003) results show that increasing numbers of American consumers demonstrate their environmental concern through environmentally conscious purchase behavior (ECPB), i.e., the practice of buying products on the basis of their environmental attributes. Similarly, in the mid-1990s, The Hartman Group, another market research firm, conducted a nationwide study of U.S. food consumers (Hartman Group 1996, 1997, 1999) and the results showed price, quality, and availability/convenience were the most important purchase decision criteria. However, they also found that: 1) about 50 percent of food consumers were influenced by environmental considerations; 2) about half of food consumers viewed ecolabels (on-product symbols that differentiate a product from similar products based on environmental impacts associated with production, distribution, use, or disposal) as a key source of information about a product's environmental attributes and they would like to see more detailed information on ecolabels; and 3) 63 percent reported they would be willing to pay more for a product that has a positive environmental impact (Hartman 1996, 1997, 1999). Results from the Hartman and Roper reports as well as a variety of other sources (e.g., Schwepker and Cornwell 1991, Ottman 1993, Bhat 1996, Nimon and Beghin 1999) caused many companies to believe that consumers would buy a product based on its environmental attributes. In reality, most environmentally friendly products face apathy and price resistance from the majority of consumers. For example, 20 years of opinion polling by the National Renewable Energy Lab concluded that 56 to 80 percent of American voters were willing to pay more for environmental protection or renewable electricity. However, as of May 1998, only 1 to 2 percent or 45,000 electric customers across the country had signed up for "green" electricity (Holt and Wiser 1999). Apparently, Fowler (2002) correctly asserts, "after a decade of attempting to appeal to shopper's environmental sensibilities, many companies have concluded that shoppers seem far more willing to pay for convenience than for ideology."
What causes the disconnect between expressed environmental concern and subsequent ECPB? The Green Gauge Report suggests that some consumers perceive environmentally friendly products as inferior (Roper ASW 2003). For example, 41 percent of respondents reported not buying green products because they feared the products wouldn't work as well. Thus, the disconnect may be an artifact of green marketing's early days when inferior, overpriced earth-friendly products sought to replace readily available, trusted, high-quality brands (Ottman 1993).
Dunlap (1989) offers several explanations for the discrepancy between environmental concern and ECPB. First, increasing government attention to environmental problems causes the media and public to assume the problem is being addressed, thus, they turn their attention to other issues. Second, the public perceives institutions rather than individuals as the culprit. This common perception causes individuals to believe environmental responsibility lies with organizations. Third, people may be willing to change some, but not all, aspects of their lives for the sake of the environment, i.e., willing to recycle but not carpool. Fourth, the public may be uneducated in methods of environmentally responsible behavior. Finally, the absence of strong leadership regarding environmental protection may cause some to feel lifestyle changes are unimportant.
In this context of heightened environmental concern, but lagging ECPB, the concept of forest certification (third-party verification that forest management practices are consistent with predetermined criteria) has emerged. It was introduced in the early-to-mid 1990s and it's based on the notion that consumers are concerned about the state of the world's forests and would therefore prefer purchasing forest products that are certified to originate from well-managed forests. Thus, forest certification is conceived as a market-based approach for sustainable forest management. Some certified forest products (CFPs) display ecolabels that indicate an independent third-party inspected and approved the forest management practices in the forest from which the product originated. The movement by the global forest products industry toward forest certification is clear. The number of forest acres under certified management has grown to 306 million acres worldwide (Rametsteiner 2002). Additionally, major lumber products retailers in the United States and Europe including The Home Depot, IKEA, BandQ, and The Body Shop have made commitments to source their lumber products from environmentally certified vendors (Anderson et al. 2002). Despite these trends, consumer demand for certified, ecolabeled forest products has been slow to develop.
Previous studies of consumer demand for CFPs
Because the process of forest certification can be a costly investment, many forest landowners and forest products manufacturers want evidence confirming the assumption that consumers will discriminate purchases in favor of CFPs. Numerous studies have addressed the issue, although none have produced empirical evidence supporting the notion of consumer preference for CFPs. Teisl et al. (2002) found that environmental labeling of forest products could influence consumer decision making, but price and quality would likely continue to be the most important considerations. Their study was unique because they used focus groups. Smith (1999) conducted a review of published studies about consumer attitudes relative to forest products marketed with emphasis on social, environmental, or sustainable attributes. He identified 184 publications and classified 25 of them as major research efforts. Results from those studies are mixed. Several indicated that marketing a forest product's environmental attributes would have a positive effect on consumer behavior (Winterhalter and Cassens 1994, Ozanne and Vlosky 1997, Bigsby et al. 1997). Others indicated that a forest product's environmental attributes were relatively unimportant in consumer purchase decisions (Ozanne and Smith 1996, Groonros and Bowyer 1999, Forsyth et al. 1999, Rametsteiner 1999). Several studies identified a consumer segment that placed high value on environmental certification (Ozanne and Smith 1996, Ozanne and Vlosky 1997, Ozanne and Smith 1998, Forsyth et al. 1999, Groonros and Bowyer 1999).
Smith's review concluded that the dominant methodology used was mail questionnaires and personal interviews. The format typically involved direct questions about attitudes and intended behavior with respect to purchasing CFPs. There are two drawbacks associated with this methodology. First, the respondent considers the importance of any given attribute individually rather than in the context of other product attributes. Therefore, the researcher can determine whether or not environmental certification, for example, is an important attribute, but he or she cannot determine its importance relative to other product attributes. A second drawback is an apparent weak link between expressed attitude and actual behavior. Attitude is defined as a learned predisposition to consistently respond positively or negatively toward an object. It is an extremely popular concept because it is generally believed that attitudes are predictors of behavior. However, there are many examples of attitude/behavior inconsistency (Wicker 1969, Oskamp et al. 1991). In the case of CFPs, we believe that respondents tend to answer questions in a "socially desirable" manner, thereby overstating the importance placed on a product's environmental attributes and their willingness to pay a premium for an environmentally certified product. Because of these drawbacks, Smith (1999) urged further research that would observe actual consumer purchases of CFPs.
[FIGURE 1 OMITTED]
The observation of actual purchase behavior is a logical next step for research on this topic since it bypasses the weak link between expressed attitude and actual behavior. However, such a methodology has limitations as well. From a practical perspective, our experience has shown that a study of actual consumer behavior is exceedingly difficult. Gaining full cooperation from a CFP supplier and CFP retailer while maintaining an adequate level of academic and scientific rigor is challenging at best. From an external validity perspective, we submit that the researcher must use great care in designing such a study because it is questionable whether the observation of purchase behavior can be interpreted as consumer preferences. For example, given our busy lives, it can be argued that a consumer chooses the best available product, rather than taking the time to find the product that perfectly satisfies his or her needs. In other words, a consumer with low involvement in a purchase may buy a product simply because it is convenient and not because it is their preferred product. If such a scenario were true, a researcher would incorrectly interpret purchases as preferences for a product, rather than purchases made because they were most convenient.
Conjoint analysis is a multivariate statistical technique first introduced in 1964, which has evolved into a family of techniques that has received extensive use since the mid-1970s (Green and Srinivasan 1990). Briefly described, it is a method by which researchers can realistically model which product attributes matter most to consumers. We do not offer a full explanation of the technique; for more information the interested reader should consult other publications (Green and Srinivasan 1990, Hair et al. 1995, Reddy et. al. 1996, McCullough 2002). We chose to use conjoint analysis in this study as a tool for modeling consumer response to CFPs because it allows us to determine the importance of environmental certification relative to other product attributes. Despite its widespread use in other disciplines, a review of the forest products literature revealed that few studies used the technique. Reddy and Bush (1998) and Reddy et al. (1996) used it to evaluate consumer perceptions of softwood lumber quality. Bigsby and Ozanne (2002) and Cooper et al. (1996) used it to model consumer preferences for CFPs. In both of those studies, which used wood furniture as a hypothetical product, environmental certification was relatively unimportant.
The two main objectives of this study were to: 1) assess respondent preferences among selected wood furniture attributes with special attention on the relative importance placed on environmental certification; and 2) identify a set of explanatory variables useful for predicting which respondents rated environmental certification as the most important product attribute. Such a model is useful since we believe that respondents who place the highest importance on environmental certification constitute a target market for CFPs.
A fundamental concept underlying this research is that all humans possess needs and desires, some of which can be satisfied through the purchase of goods and services. Consumer behavior is the study of individuals, groups, or organizations and the processes they use to select, secure, use, and dispose of products and services to satisfy their needs and the impacts these processes have on the consumer and society (Hawkins et al. 2001). To frame our study theoretically, we adapted Hawkins' et al. (2001) model of consumer behavior (Fig. 1). According to the model, needs and desires drive an individual's purchase behavior. Needs and desires are influenced by each person's self-concept, which in turn, is formed by influences in that person's life, i.e., values, beliefs, and attitudes. As we gain life experiences our self-concept and lifestyle change based on what we have learned.
To operationalize the theoretical constructs depicted in Figure 1, we created a questionnaire that was divided into two sections: a dependent measure and several explanatory measures.
[FIGURE 2 OMITTED]
Dependent measure. -- In the first section of the survey, respondents were asked to imagine they were purchasing a wood CD rack while considering the following attributes: price, wood type, environmental certification, adjustability of shelves, and storage capacity. Within each attribute, several levels were defined, e.g., price = $75, $100, or $125. Each level of each attribute was clearly defined on the survey instrument. Eight versions of a wood CD rack were created using a fractional, orthogonal conjoint design. The conjoint design was presented to the respondents as shown in Figure 2. The task of each respondent was to rank the eight versions of a wood CD rack in order from most desirable to least desirable. This scenario of comparing different versions of similar products is thought to realistically recreate the purchase process a consumer encounters in real life (McCullough 2002). Statistical analysis of each respondent's rankings yields the relative importance for each of the five product attributes. We used each respondent's importance ranking for the attribute environmental certification as a criterion for classifying respondents into one of two categories: 1) those who rated environmental certification as the most important product attribute; and 2) all others, thereby, creating a binary dependent variable.
Selection of the product attributes. -- The first step in creating a conjoint analysis study is deciding what product attributes need to be included in the model. The environmental certification attribute was chosen on the basis of its relevance to the objectives of the study. The 4 remaining attributes (price, wood type, adjustability of shelves, and storage capacity) were selected from an original list of 16 wood CD rack attributes consistently mentioned in promotional material. The basis of attribute selection was their importance rankings in a survey of 15 students enrolled in a wood science class at Oregon State University. During this process, size and color of the rack also emerged as very important considerations. We felt that those two attributes are indeed important, but that they were not determinant. Therefore, we stated on the questionnaire that the respondent should assume the size and color of the wood CD rack perfectly matched their needs. The levels for the attribute price ($75, $100, and $125) were selected because they are within the range of published prices for wood CD racks.
Explanatory measures. -- The second part of the questionnaire was exploratory in nature in that we identified 14 constructs that we hypothesized to be correlated to ECPB:
1. Demographic. -- Respondents reported their age, gender, personal income, and parental income.
2. Willingness to pay. -- We asked each respondent to assume that they were going to purchase a wood CD rack that cost $100. We then asked them how much more they would be willing to pay for the rack if it were environmentally certified. Their response choices were: 0, 5, 10, 25, 50, and more than 50 percent. Adapted from Ozanne and Vlosky (1997).
3. Awareness of certification. -- Respondents reported whether they had heard of forest certification prior to completing the questionnaire. The scale used was 1 = know nothing about and 7 = know a great deal about. Single item scale, developed for this study.
4. Knowledge of forest certification. -- Respondents reported their personal level of knowledge about forest certification on a scale from 1 = know nothing about and 7 = know a great deal about. Single item scale, developed for this study.
5. Understanding of environmental certification. -- Respondents reported whether or not they understood why products are environmentally certified. The scale used for this item was 1 = agree and 7 = disagree. Single item scale, from Ozanne and Vlosky (1997).
6. Importance of environmental packaging. -- Respondents reported their attitude about the importance of displaying environmental information on product packaging. The scale used for this item was 1 = agree and 7 = disagree. Single item scale, developed for this study.
7. Credibility of environmental packaging. -- Respondents reported their attitude about the credibility of environmental claims made on product packaging. The scale used for this item was 1 = agree and 7 = disagree. A composite score for this variable was computed by summing the items in the construct. Developed for this study. Cronbach's Alpha = .781.
8. Knowledge of environmental issues. -- Respondents reported their knowledge of five general environmental issues: acid rain, world population growth, global warming, pollution from pesticides, and destruction of the rainforest. The scale used for each item was 1 = know nothing about and 7 = know a great deal about. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .915. Developed for this study.
9. Engagement in environmentally friendly behavior. -- Respondents reported their engagement in three environmentally friendly behaviors, including recycling, joining/supporting an environmental organization, and boycotting purchases from specific companies whose products damage the environment. The scale used for each item was 1 = never and 7 = always. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .686. Developed for this study.
10. Engagement in environmentally conscious purchase behavior. -- Respondents reported how often they intentionally purchased organic fruits and vegetables, recycled paper products, environmentally friendly detergents, and products not tested on animals. The scale used for each item was 1 = never and 7 = always. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .822. Developed for this study.
11. Altruism. -- Respondents reported how likely they would be to engage in six specific behaviors that can generally be described as helping someone in need. The scale used for each item was 1 = I wouldn't do this and 7 = I would do this. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .723. Adapted from Rushton, Chrisjohn, and Fekken (1981).
12. List of values. -- Respondents reported how important nine values are in their daily life and then selected one value from the list of nine that was most important. Cronbach's Alpha = .900. From Kahle and Kennedy (1989).
13. Environmental concern. -- Respondents answered four questions regarding their level of concern for the state of the natural environment. The scale used for each item was 1 = agree and 7 = disagree. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .867. Adapted Dunlap and Van Liere (1978).
14. Perceived consumer effectiveness. -- Respondents reported the degree to which they believe their individual actions as consumers affect the state of the natural environment. The scale used for each item was 1 = agree and 7 = disagree. A composite score for this variable was computed by summing the items in the construct. Cronbach's Alpha = .552. From Straughan and Roberts (1999).
The Cronbach's Alpha scores indicate the reliability of a scale. The Cronbach's Alpha values for constructs 9, 11, and 14 suggest that these scales are less reliable than optimal. However, these scales were adapted from other studies, in which they were shown to be reliable. Therefore, we elected to retain them as potential explanatory variables.
We sampled undergraduate students enrolled in an international business class at Oregon State University during the winter and spring terms of the 2001/2002 academic year. A self-administered questionnaire was used to collect data. Respondents were given approximately 20 minutes to complete the questionnaire at the beginning of a regularly scheduled class session. Respondents were clearly informed that participation was voluntary and not related to their grade in any way.
A total of 293 questionnaires were received, of which 265 were used in the analysis. There were 28 (9.5%) that were unusable due to incomplete answers (10 questionnaires) or response patterns that didn't meet the assumptions required by conjoint analysis (18 questionnaires).
The majority of respondents (78.7%) reported that they had never heard of forest certification prior to completing the questionnaire. Of those that were familiar with forest certification, they generally learned of it in other classes, rather than having bought CFPs in the past. The mean self-reported knowledge level about the forest certification issue was 2.02 (SD 1.19) on a scale where 1 = know nothing about and 7 = know a great deal about.
Nearly 55 percent of respondents were female and the average respondent age was 22 years. The median personal income range was less than $20,000. The median parental income range was $60,001 to $80,000.
Table 1 provides a complete summary of the demographic characteristics of the respondents.
Method of analysis
Conjoint analysis. -- First we analyzed the conjoint results for each individual. We assumed that the wood CD rack evaluations were an additive function of the five product attributes we defined and that these evaluations are interval level data (Louviere 1988). These are common assumptions in conjoint analysis and they allow use of ordinary least squares regression to estimate the parameters of our wood CD rack attribute importance model. Additivity implies that a respondent's utility for a whole product is simply the sum of the utilities for each attribute. Several questionnaires (6.1%) were eliminated from the data set because the conjoint responses were non-additive. In other words, their ranking scores showed either no consistent pattern of preference for any product attribute, or an interaction effect between two or more attributes. An [r.sup.2] value less than .90 was used to eliminate non-additive responses.
With the data "cleaned" of respondents whose rankings were not consistent with the assumptions of conjoint, we next determined the preferred levels for each attribute for each respondent. This was accomplished by computing the part-worth for each level of each attribute. Then the part-worths were used to determine the relative importance of each attribute for each respondent. We also used each respondent's importance scores to divide the respondents into five subgroups based on which attribute they rated most important. Up to this point, all analysis was done at the individual respondent level. In the next step, we adjusted a regression that yielded the part-worths for respondents at the aggregate level. The aggregate part-worth values were then used to calculate the relative importance of the attributes for the "typical" respondent. Finally, we performed the same aggregate part-worth calculations within each of the five subgroups to determine the relative importance of the product attributes for the typical respondent in each subgroup.
Binary logistic regression analysis. -- In this phase of the analysis, our objective was to identify a parsimonious set of variables useful for 1) predicting which respondents placed the highest importance on environmental certification; and 2) examining the effect of the selected variables on respondent preferences for environmental certification. Therefore, we used backwards-stepwise (likelihood ratio) binary logistic regression to identify which explanatory variables were useful in predicting membership in the group of interest. Binary logistic regression is a multivariate statistical procedure used to predict a dichotomous dependent variable from a set of dichotomous or polytomous independent variables (Hair et al. 1995). An alpha-to-enter value of .05 and an alpha-to-exit value of 0.10 were used. These procedures resulted in a "best" model containing four explanatory variables. Note we use "best" in quotes because according to Ramsey and Schafer (1997), the variable selection process should be sensitive to the objectives of the study, and the particular subset chosen is relatively unimportant.
Thus, we arrived at the final model through a combination of stepwise selection and subjective judgment using three criteria: parsimony, ease of variable acquisition, and ease of interpretation. SPSS statistical software version 11.0 was used for the analysis.
The importance of environmental certification
Table 2 displays part-worth estimates and relative importance estimates from our sample. We performed these calculations in aggregate (all respondents as one group) and for each of the five subgroups. Each subgroup was comprised of the respondents that rated a particular product attribute as their most important. For example, the environmental certification subgroup contained 55 respondents who rated environmental certification as their most important product attribute. Note that the sum of the number of respondents in the subgroups is 256 rather than 265; this is because 9 respondents had equal importance scores for two or more attributes, and thus we could not group those 9 by most important attribute.
The part-worth scores indicate the preferred levels of each attribute; the greater the positive value of the part-worth estimate the higher the preference for that level. Table 2 shows that the aggregate and subgroups all preferred the following level of each wood attribute: the lowest price, solid wood, environmentally certified, adjustable shelves, and greater storage capacity. These results indicate that environmentally certified is a preferred product attribute compared to non-environmentally certified.
Given this result, the question becomes, in the context of other product attributes how important is environmentally certified? The importance scores indicate this. The importance scores are calculated by: 1) determining the range of part worths for each attribute; 2) summing the ranges; and 3) dividing the range for each attribute by the sum of ranges. Relative importance is an indicator of the weight a respondent places on an attribute relative to the others as he or she forms preferences. One can infer from a comparison of two attributes' relative importance scores how much more important one is than the other. For the aggregate, price was clearly the most important attribute, about 1.5 times as important as wood type, wood origin, and adjustability and about 2.5 times as important as storage capacity. The importance scores change quite dramatically between subgroups. For example, note the reversal of the aggregate results in the environmental certification subgroup where the importance of the attribute wood origin is about 2.5 times that of price.
Note the consistently high importance of price across all subgroups. It is either the first or second most important attribute. Our aggregate results are consistent with those of Ozanne and Smith (1996) who found that consumers consider the following factors highly important in the purchase decision: quality (in terms of both construction and materials used) and price. Environmental impact, on the other hand, was not a primary purchase decision criterion in their study. On the other hand, our results are not consistent with Ozanne and Bigsby (2002) who found that price was relatively unimportant in New Zealand consumers purchase decisions for outdoor wood furniture.
These results show that the respondents do not place equal weight on product attributes when forming their preferences. For example, we know from the part-worth values reported in Table 2 that the typical respondent prefers an environmentally certified wood CD rack to a non-certified rack. However, we also know that compared to price, wood origin is relatively unimportant. Thus, when forming preferences the typical respondent is willing to sacrifice environmental certification for the sake of a lower price. We also know from Table 2 that the importance of a given attribute varies greatly from individual to individual. In other words, at the aggregate level, wood origin appears relatively unimportant, but when looking at the importance ratings among the five subgroups, 55 respondents or 20.8 percent rated it their most important attribute.
Predicting respondents' utility for wood origin
This exploratory part of the analysis attempted to identify which explanatory variables were significant in predicting respondents who placed the highest importance on environmental certification. Backwards-stepwise logistic regression was used to reduce the number of explanatory variables to the four displayed in Table 3. Note from Table 4 that the model correctly predicted which group a respondent belonged to 82.3 percent of the time. Breaking down predictive accuracy further, the model correctly predicts those who do not place highest importance on environmental certification 97.1 percent of the time. But it was only able to correctly predict the respondents that were most likely to place the highest importance on environmental certification 26.1 percent of the time.
It was also of interest to interpret the regression coefficients in terms of the effect they have on group membership. The parameter estimates are displayed in Table 3. These results show that after accounting for the effects of the other variables in the model: 1) females are 1.67 (95% CI = .82 to 3.40) times as likely to rate environmental certification as the most important attribute; 2) those respondents willing to pay at least a 5 percent premium for a CFP are 1.56 (95% CI = 1.18 to 2.06) times as likely to rate environmental certification as the most important attribute; 3) those respondents who express increased levels of environmental concern are 1.16 (95% CI = 1.05 to 1.28) times as likely to rate environmental certification as the most important attribute; and 4) surprisingly, those respondents who expressed lower levels of altruism are 1.08 (95% CI = 1.01 to 1.15) times as likely to rate environmental certification as the most important attribute.
Marketing and managerial implications
Although these data are from a convenience sample of undergraduate students and are therefore not generalizable, the data do provide insight about possible marketing and managerial implications. Further research is required to verify the following points.
These results suggest that for the typical respondent, environmental certification has a positive effect on their preferences for a wood CD rack. However, more influential competing effects often outweigh environmental certification's positive effect. For example, the typical respondent was willing to sacrifice environmental certification for the sake of a lower price. Should such a finding hold true for the general population, it would suggest that CFPs cannot command premium prices when marketed in mainstream distribution channels, e.g., big-box retail stores, because for the typical respondent the utility of a lower price outweighs the value of environmental certification. However, it may be possible to market CFPs in a separate distribution channel. The target market for this channel would be the respondents who placed the highest value on the attribute environmental certification. The recent emergence of green building supply retailers are an example of an alternate distribution channel. In the Pacific Northwest, two such stores exist: Environmental Building Supply (EBS) in Portland, Oregon, and The Environmental Home Center in Seattle, Washington.
This study identified several characteristics associated with respondents who value environmental certification more than the other attributes. Again, should these findings hold true for the general population, knowledge of those characteristics could be used by marketers to tailor CFP promotional efforts to maximize their impact on the target market. This includes explicitly stating the benefits that buying CFPs will have on the state of the natural environment. Such a message should resonate strongly with this group, given their higher levels of environmental concern. Second, promotional communications should be targeted to reach audiences that are more likely to be female. Third, not only is environmental certification very important to this group, they reported that they were more likely to be willing to pay at least a 5 percent premium for CFPs. This finding suggests that a premium for certified products may be obtainable in this distribution channel.
Straughan and Roberts (1999) found evidence that altruistic people, i.e., those concerned for the welfare of others, also tend to exhibit ECPB. Presumably this is because the welfare of others is linked to the state of the natural environment. We have no explanation for the results of this study, which found that those who reported themselves as less altruistic were also more likely to rate environmental certification as the most important product attribute.
Our conjoint analysis may be limited by the fact that there is an effect for the number of levels associated with an attribute. An attribute with significantly more levels than the other attributes in a model will be more important than the other attributes. In this study, price was the only attribute with three levels while the others all had two levels. Thus, it is possible that some of the importance in price can be ascribed to the fact that it had one more level than all other attributes.
Perhaps the biggest limitation of this study is that since the population sampled was undergraduate students, definitive inferences from the results cannot be drawn to broader populations of consumers.
Table 1. -- Demographics of sample respondents (% respondents). Gender (n = 263, with 2 missing) Male 45.3 Female 54.7 Total 100.0 Age (n = 262, with 3 missing) 19 1.9 20 19.8 21 32.4 22 21.8 23 11.8 24 2.3 25 2.7 >25 7.3 Total 100.0 Personal income range (n = 258 with 7 missing) Less than $20,000 86.4 $20,000 to $40,000 8.4 $40,001 to $60,000 1.6 $60,001 to $80,000 1.6 $80,001 to $100,000 .8 More than $100,000 1.2 Total 100.0 Parental income range (n = 241 with 24 missing) Less than $20,000 2.5 $20,000 to $40,000 10.0 $40,001 to $60,000 19.9 $60,001 to $80,000 21.2 $80,001 to $100,000 16.6 More than $100,000 29.9 Total 100.0 Table 2. -- Part worth and relative importance estimates from overall group, wood origin group, wood type group, price group, adjustable group, and storage group. All respondents Wood origin subgroup n = 265 n = 55 Part Part worth Importance worth Importance Price 29.5% 18.9% $75 0.874 0.521 $100 0.021 0.036 $125 -0.895 -0.531 Wood type 21.4% 13.3% Solid wood 0.705 0.236 Composite -0.705 -0.236 Wood origin 18.3% 46.3% Certified 0.635 1.745 Non-certified -0.635 -1.745 Shelves 18.9% 12.3% Adjustable 0.723 0.477 Non-adjustable -0.723 -0.477 Storage 11.9% 9.2% 100 CDs -0.399 -0.323 200 CDs 0.399 0.323 100.0% 100.0% Wood type subgroup Price subgroup n = 68 n = 70 Part Part worth Importance worth Importance Price 18.3% 55.1% $75 0.455 2.224 $100 0.039 0.069 $125 -0.494 -2.314 Wood type 49.2% 10.7% Solid wood 1.84 0.252 Composite -1.84 -0.252 Wood origin 10.8% 12.4% Certified 0.283 0.417 Non-certified -0.283 -0.417 Shelves 12.1% 13.4% Adjustable 0.415 0.555 Non-adjustable -0.415 -0.555 Storage 9.6% 8.4% 100 CDs -0.301 -0.271 200 CDs 0.031 0.271 100.0% 100.0% Adjustable subgroup Storage subgroup n = 43 n = 20 Part Part worth Importance worth Importance Price 18.3% 17.4% $75 0.492 0.304 $100 0.074 0.264 $125 -0.480 -0.561 Wood type 11.4% 11.9% Solid wood 0.265 0.408 Composite -0.265 -0.408 Wood origin 10.9% 10.1% Certified 0.299 0.263 Non-certified -0.299 -0.263 Shelves 49.8% 11.6% Adjustable 1.900 0.197 Non-adjustable -1.900 -0.197 Storage 9.6% 49.0% 100 CDs -0.288 -1.895 200 CDs 0.288 1.895 100.0% 100.0% Table 3. -- Parameter estimates for the binary logistic regression model. (a) Variable B S.E. Wald Sig Exp(B) Gender -0.514 0.363 2.000 0.157 0.598 Willingness to pay 0.447 0.140 10.159 0.001 1.563 Environmental concern 0.150 0.050 8.877 0.003 1.162 Altruism -0.073 0.032 5.156 0.023 0.930 Constant -2.690 1.428 3.551 0.060 0.068 (a) Nagelkerke [r.sup.2] = .216 Table 4. -- Classification table for predicting group member. (a) 0 1 Percent correct (%) 0 169 5 97.1 1 34 12 26.1 Overall percentage correct 82.3 (a) 0 = Environmental certification not most important; 1 = environmental certification most important; total cases = 220.
[c]Forest Products Society 2004.
Forest Prod. J. 54(3):42-50.
Anderson, R.C., E. Hansen, and E. Wagner, 2002. Thoughts regarding the forest leadership forum (and forest certification). Unpublished summary of notes taken while in attendance at the Forest Leadership Forum: Collaborative Pathways to Responsible Trade, Atlanta, GA.
Bhat, V.N. 1996. The Green Corporation: The Next Competitive Advantage. Quorum, Westport, CT.
Bigsby, H. and L.K. Ozanne. 2002. The purchase decision: Consumers and environmentally certified wood products. Forest Prod. J. 52(7/8):100-105.
__________, __________, and R.P. Vlosky. 1997. New Zealand consumers and the environmental certification of forest products. In: Proc. Preparing for the 21st century: Value added marketing for value added wood products. Proc. 7279. Forest Prod. Soc., Madison, WI.
Cooper, R.J., S.P. Kalafatis, and M.M.H. Tsogas. 1996. The impact of green issues on the development of timber based furniture--A comparison of two EU countries. In: Proc. Environmental Issues and Market Orientation, Current Topics in Forest Products Marketing. Forest Products Marketing Group, IUFRO XX World Congress. H. Juslin and M. Pesonen, eds. IUFRO, Vienna, Austria.
Dunlap, R.E. 1989. Public opinion and environmental policy. In: Environmental Politics and Policy. J.P. Lester, ed. Duke University Press, Durham, NC. pp. 87-124.
__________ and K.D. Van Liere. 1978. The new environmental paradigm. J. of Environmental Education 9(4):10-19.
Forsyth, K., D. Haley, and R. Kozak. 1999. Will consumers pay more for certified wood products? J. of Forestry 9(2):18-22.
Fowler, G.A. 2002. 'Green' sales pitch isn't moving many products. Wall Street J. March 6.
Green, P.E. and V. Srinivasan. 1990. Conjoint analysis in marketing: New developments with implications for research and practice. J. of Marketing 54(4):3-19.
Groonros, J. and J. Bowyer. 1999. Assessment of the market potential for environmentally certified lumber and wood products among buyers of new homes in Minneapolis/St. Paul and Chicago. Forest Prod. J. 49(6):28-34.
Hair, J., R. Anderson, R. Tatham, and W. Black. 1995. Multivariate Data Analysis with Readings. 4th ed. Prentice-Hall, Upper Saddle River, NJ.
Hartman Group. 1996. Food and the environment: A consumer's perspective--Phase one. The Hartman Group, Inc. Bellevue, WA.
__________. 1997. Food and the environment: A consumer's perspective--Phase two. The Hartman Group, Inc. Bellevue, WA.
__________. 1999. Food and the environment: A consumer's perspective--Phase three. The Hartman Group, Inc. Bellevue, WA.
Hawkins, D.J., R.J. Best, and K.A. Coney. 2001. Consumer Behavior--Building Marketing Strategy, 8th ed. McGraw Hill Co., New York.
Holt, E.A. and R.H. Wiser. 1999. Understanding consumer demand for green power. National Wind Coordinating Committee. http://www.nationalwind.org/
Kahle, L. and P. Kennedy. 1989. Using the list of values (LOV) to understand consumers. J. of Consumer Marketing 6(3):5-12.
Louviere, J.J. 1988. Analyzing decision making: Metric conjoint analysis. Sage Univ. Paper Series on Quantitative Applications in the Social Sciences, Series 07-067. Sage Pub., Newbury Park, CA.
McCullough, D. 2002. A user's guide to conjoint analysis: Before starting out you need to know where the land mines are. Marketing Research 14(2):18-23.
Nimon, W. and J. Beghin. 1999. Are eco-labels valuable? Evidence from the apparel industry. Am. J. of Agricultural Economics 81(4):801.
Oskamp, S., M.J. Harrington, T.C. Edwards, D.L. Sherwood, S.M. Okuda, and D.C. Swanson. 1991. Factors influencing household recycling behavior. Environment and Behavior (23):494-519.
Ottman, J.A. 1993. Green Marketing. 2nd ed. NTC Business Books, Lincolnwood, IL.
Ozanne, L.K. and P.M. Smith. 1996. Wooden household furniture: Does the environment matter to consumers? In: Proc. Environmental Issues and Market Orientation, Current Topics in Forest Products Marketing. Forest Prod. Marketing Group, IUFRO XX World Congress. IUFRO, Vienna, Austria.
__________ and __________. 1998. Segmenting the market for environmentally certified wood products. Forest Sci. 44(3):379-389.
__________ and R.P. Vlosky. 1997. Willingness to pay for environmentally certified wood products: A consumer perspective. Forest Prod. J. 47(6):39-48.
Rametsteiner, E. 1999. Results of the consumer survey. In: Proc. Potential Markets for Certified Forest Products in Europe. B. Pajari, T. Peck, and E. Rametsteiner, eds. European Forest Institute, Brussels, Belgium.
__________. 2002. UNECE-The Forest Products Annual Market Review, 2000-2001. Timber Bull., Chapter 11, Vol. LIV, ECE/TIM/BULL/55/3. www.unece.org/trade/timber/docs/rev-01/chap-10.pdf.
Ramsey, F.L. and D.W. Schafer. 1997. The Statistical Sleuth: A Course in Methods of Data Analysis. 1st ed. Duxbury Press, Belmont, CA.
Reddy, V.J. and R.J. Bush. 1998. Measuring softwood lumber value: A conjoint analysis approach. Forest Sci. 44(1):145-157.
__________, __________, and R. Roudik. 1996. A market-oriented approach to maximizing product benefits: Cases in U.S. forest products industries. In: Proc. Environmental Issues and Market Orientation. Current Topics in Forest Products Marketing. Forest Prod. Marketing Group, IUFRO XX World Congress. H. Juslin and M. Pesonen, eds. IUFRO, Vienna, Austria.
Roper ASW. 2003. Roper Green Gauge report-Summary brochure. Accessed on 3/2003. http://www.roperasw.com/products/consumers/gg.pdf
Rushton, J.P., R.D. Chrisjohn, and G.C. Fekken. 1981. The altruistic personality and the self-report altruism scale. Personality Individual Differences 2(4):293-302.
Schwepker, Jr., C.H., and T.B. Cornwell. 1991. An examination of ecologically concerned consumers and their intention to purchase ecologically packaged goods. J. of Public Policy and Marketing 10(2):77-101.
Smith, P.M. 1999. A Review of studies on consumer attitudes towards forest products marketed with environmental, social and/or sustainability attributes. Food and Agriculture Organization of the United Nations (FAO) Rome, Italy. http://www.fao.org/forestry/fop/foph/marketing/doc/x2421e/x24211e00.htm#/TopOfPage
Straughan, R.D. and J.A. Roberts. 1999. Environmental segmentation alternatives: A look at green consumer behavior in the new millennium. J. of Consumer Marketing 16(6):558-575.
Teisl, M.E., S. Peavey, F. Newmann, J.A. Buono, and M. Herrmann. 2002. Consumer reactions to environmental labels for forest products: A preliminary look. Forest Prod. J. 52(1):44-50.
Wicker, A.W. 1969. Attitudes versus actions: The relationship of verbal and overt behavioral responses to attitude objects. J. of Social Issues 25(4):41-78.
Winterhalter, D. and D. Cassens. 1994. Is the marketplace ready for a certified sustainable forest product? In: Proc. 1993 Society of American Foresters National Convention. SAF, Bethesda, MD.
Roy C. Anderson*
Eric N. Hansen*
The authors are, respectively, PhD Candidate and Associate Professor/Extension Specialist, Dept. of Wood Science and Engineering, College of Forestry, Oregon State Univ., Corvallis, OR 97331-5751. The authors would like to thank Steve Lawton, Associate Professor, OSU, for allowing us to conduct our survey in his classes; Dr. Ernesto Wagner, Faculty Research Assistant, OSU, for his helpful suggestions during the preparation of this manuscript; and the FPJ reviewers for their constructive suggestions and comments. This paper was received for publication in September 2002. Article No. 9550.
*Forest Products Society Member.
|Printer friendly Cite/link Email Feedback|
|Author:||Anderson, Roy C.; Hansen, Eric N.|
|Publication:||Forest Products Journal|
|Date:||Mar 1, 2004|
|Previous Article:||Leaching of flakeboard produced from recycled CCA-treated wood into deionized water.|
|Next Article:||Influence of resin type and flake thickness on properties of OSB.|