Italian Consumers’ Willingness to Pay for Eucalyptus Firewood

: Eucalyptus trees cover about 20 million hectares globally and are used to produce pulp, paper and ﬁrewood for domestic uses. From an environmental perspective, these trees have fewer impacts than other crops. In Italy, plantations of eucalyptus can provide a large amount of biomass to satisfy part of the country’s internal demand. However, eucalyptus cultivation is less proﬁtable than cultivation of traditional crops due to the low market prices of wood. This study aims both to analyze the willingness of a sample of Italian consumers to pay for eucalyptus ﬁrewood and to investigate the main factors that may a ﬀ ect this willingness. Data are collected from a sample of 231 consumers using a web-based survey. The double-bounded dichotomous choice contingent valuation model is then applied. The ﬁndings show that information, the energetic density of ﬁrewood, consumers’ interest in environment issues, and the age of respondents are aspects that are positively associated with respondents’ willingness to pay for eucalyptus ﬁrewood. Conversely, interest in both ﬁrewood species and packaging are factors that reduce consumer willingness to pay for eucalyptus ﬁrewood. Even though these results cannot be generalized to the whole Italian population, the ﬁndings may indicate new opportunities for eucalyptus, while growing demand for eucalyptus could o ﬀ er an interesting opportunity for ﬁrms to enter the sector and develop marketing strategies targeted towards speciﬁc market niches.


Introduction
Globally, eucalyptus is the most commonly used species for fast growing plantations (with a 10-16 year rotation) and have the potential to help meet global demand for wood [1,2]. Eucalyptus trees cover about 20 million hectares in more than 90 countries around the world, particularly in countries such as Brazil, India, and China [3]. Eucalyptus is used to produce pulp and paper, charcoal, sawn timber, wood panels for industries, and also firewood for domestic uses [3].
Moreover, eucalyptus management (e.g., tree density, fertilization, harvesting cycles, etc.) is less intensive than the management of conventional agricultural crops but is more intensive than conventional forestry [1], which means that eucalyptus occupies a niche between highly productive forestry and conventional forestry [4]. Eucalyptus is an efficient biomass producer and can produce more biomass than many other tree species [5]. In addition, from an environmental perspective, eucalyptus is less impactful than other crops [6], contributes to the conservation of biodiversity [5,7], and shows high carbon sequestration potential during its growth [8,9]. In fact, eucalyptus can play an important role in mitigating climate change since it is fast growing and can fix more CO 2 by the process of photosynthesis [5]. Moreover, properly managed eucalyptus plantations can control soil erosion, and the litter which accumulates under most eucalyptus plantations can improve soil fertility [5].
In Italy, there are more than 100,000 ha planted with agro-forest species such as poplar, eucalyptus, and acacia [10], and plantations of eucalyptus [11] can provide a large enough amount of biomass to satisfy about 72% of Italian demand [12,13]. Eucalyptus presents similar characteristics to other common firewood species in Italy, such as beech and oak. While focusing on gross calorific value, Pereira [14] reported a range of 18.8-19.2 MJ kg -1 for various eucalyptus species, while a range of 19.3-19.4 MJ kg -1 was detected for oak species [15] and a value of about 19.5 MJ kg -1 was shown for beech firewood [16]. Also, the ash content of eucalyptus firewood is similar to the content for oak and beech. In particular, Pereira [14] reported an ash content range of 0.10-0.18% for eucalyptus species, while 0.3-0.4% was shown for oak firewood [15] and 1.0-2.0% was shown for beech [16].
However, from an economic perspective, eucalyptus cultivation is less profitable for farmers than traditional crops, due to having higher production costs and lower market prices for its wood [6,17]. In other words, short rotation forestry (SRF) biomass diffusion, as the diffusion of eucalyptus, should be possible only with an increase of its market values or with the adoption of new process innovations to reduce its production costs at the farm level [18,19]. Since the price of wood is the most important factor underpinning the profitability of agro-forest farms [6,20,21], it could be useful to study the price that people are willing to pay for eucalyptus firewood to investigate the main factors that may affect it.
To the best of our knowledge, no other study has been conducted concerning consumers' willingness to pay for eucalyptus firewood in Italy. Thus, the current study aims to fill this gap by analyzing the willingness to pay (WTP) for eucalyptus firewood of a sample of Italian consumers. Moreover, this study investigates the main factors that may affect the WTP to address marketing strategies for eucalyptus firewood. This is an interesting case study, given that the technical characteristics of eucalyptus firewood such as its calorific power value are similar to that of other firewood species such as oak [22]. This paper is structured as follows: Section 2 provides a brief review of the literature on the WTP and contingent valuation approach; Section 3 describes the materials and methods used. The results are presented in Section 4 and are discussed in Section 5. In addition, Section 5 concludes also with some hints.

Willingness to Pay and Contingent Valuation Method: An Overview
The contingent valuation (CV) method is part of a wider family of approaches called stated preference methods; the CV method estimates economic values such as willingness to pay (or to accept) using responses to survey questions [23,24]. The CV is a method in which people are asked to express their preference [25] and respondents are asked the maximum price range they are willing to pay (WTP). The economic theory underlying CV assumes that the accepted price yields the highest utility for respondents [26]. There are two approaches for assessing WTP: revealed preference and stated preference. In the first approach, participants bid real money for real goods, but these market data are hardly available [27]; while in the second approach, WTP is elicited based on a hypothetical situation. However, the latter approach is prone to hypothetical bias [27]. The stated preference-based contingent valuation experiment is currently one of the most important WTP and it involves field experiments and survey data collection to elicit the preferences of participants [27]. Another important method, among stated preference approaches, is discrete choice experiments (DCE), where preferences are elicited from responses to hypothetical alternatives, meaning participants' bids are incentive-compatible [27].
Moreover, among stated preference approaches, there are models with the dichotomous choice: the first is the single-bounded model where an individual is asked if he/she is willing to pay a stated amount for a product and he/she answers "Yes" or "No" to that question. In this approach, the individual provides little information about its WTP, and to have an accurate estimation of WTP, large samples are needed. Moreover, this method could lead to hypothetical bias, which means that individuals tend to overstate the amount he/she would be ideally willing to pay for a product as compared to when he/she would actually pay for it. The second approach, with the dichotomous choice model is the double-bounded model (DBDC-CV) proposed by [26,28,29] to improve the efficiency of the estimation. In the DBDC-CV model, the market simulated by the dichotomous-choice approach is very similar to the consumer decision-making process in the real market [30]. The respondent must choose between "Yes" and "No" answers, and this can effectively avoid bias in the model due to unfamiliarity with goods [31]. Unlike the single-bounded model, the DBDC-CV is advisable with small samples [32] and involves two questions: the first on whether the respondent is willing to pay a stated amount for a product; and the second about its WTP for a higher (and lower) amount of the initial bid. The respondent's WTP lies between the two offered bid prices if either response is positive, between the second bid and the limit of the WTP if both responses are positive, and below the second bid if both responses are negative. According to [33], the double-bounded approach shows an internal inconsistency in the response strategies between the first and second bounds by people; while the use of the bid range statement reduces the perceived difference between the two questions by respondents. Moreover, since an unrealistic bid price range can lead to a bias in the double-bounded dichotomous choice model [27,34] proposed realistic bid range to reduce the bias in the model.
For reasons mentioned above, in this study the DBDC-CV model is applied and realistic bid range prices are assigned to minimize bias.

Data Collection, the Sample, and the Questionnaire
A market survey is a research method used to investigate market development and marketing opportunities [42]. In this study, data are collected from an initial sample of 253 consumers in Italy by using a web-based survey administered during the period October-December 2019. Later, 22 respondents were excluded from the sample because they were not firewood consumers. The final sample is of 231 consumers and is not representative of the Italian population, as happens in many studies about consumer behavior (see e.g., [43,44] for wood sector or [45][46][47] for food sector). The survey is implemented through social media, emails, and word of mouth. The choice to use a web-based survey is due to both its wide use in the general literature about consumer choices (see e.g., [44,48,49]) and its undoubted cost advantage [48]. Before starting the survey, a pilot study with a sample of 60 consumers was carried out in order to validate the questionnaire.
The last two sections ask questions by using a five-point Likert scale (from 1 = totally disagree to 5 = totally agree). It is important to underline that the respondents that did not have any opinion about eucalyptus firewood answered to be indifferent (3 in the Likert scale). For the reliability of the scale, Cronbach Alpha coefficient for each item group was used and it was found that the scale had good levels (from 0.60 to 0.86) of reliability.
In the first section of the questionnaire, socio-demographic aspects such as age, gender, area of residence, and education were collected [50][51][52].
In the second part, instead, we investigated the consumers attitudes towards to firewood species and its use (use), ethical aspects of consumers' firewood choice, its geographic provenience (i.e., if firewood comes from tropical countries or Mediterranean ones), and its origin (i.e., if firewood comes from an agro-forestry plant or natural woodland) [53]. The questionnaire aims also at characterizing consumers in terms of their attitude towards collecting information from some sources (such as friends, internet, TV, expert of the sector-i.e, agronomists, forestries, and sellers) (friend_info, internet_info, tv_info, expert_info, and rivend_info).    The third section of the questionnaire investigates respondents' willingness to consume eucalyptus firewood (will) that is set out as a binary choice (Yes vs. No).
Consumers were also asked to indicate their familiarity (fam) with eucalyptus firewood by answering if they have heard about eucalyptus firewood or not (dummy variable). In addition, it was asked if respondents have consumed eucalyptus firewood in the past (pass) [54].
Three different eucalyptus firewood supply methods (forn) were also proposed: loose firewood, firewood arranged in pallets, and firewood in 10-15 kg bags.
Respondents are firstly allowed to choose their preferred supply method and then to indicate their willingness to pay (WTPi) according to the supply method chosen.
The WTPi questions follow a format to which respondents only states Yes (I agree) or No (I disagree), meaning whether their willingness to pay is greater or lower than the bids (Euros X) they are offered.
It is important to underline that the bids were expressed in realistic range prices [27,33] that come from the informal local market [55] since an unrealistic bid price range could have led to a bias in the double-bounded dichotomous choice model [34].
For example, say respondents say "yes", they are willing to pay 11 €/quintal ≤ WTPi ≤ 15 €/quintal of eucalyptus. At this point all we know they are willing to pay at least 11 €/quintal ≤ WTPi ≤ 15 €/quintal. We have no upper bound estimate of their WTPi. If we then ask the respondents a follow up question, such as are they willing to pay WTPi > 15 €/quintal, and the respondent says "yes", we have gained some information: they are willing to pay at least 15 €/quintal, but if the respondents say "no", we know that their WTPi is 11 €/quintal ≤ WTPi ≤ 15 €/quintal. The situation is similar for initial responses that are "no", where the next question uses a lower bid amount.
Another important aspect considered in the questionnaire on eucalyptus firewood acceptance is consumers' motivation to use it: such aspects are measured by asking a number of questions related to appeal, curiosity, to technical characteristics, as well as to environmental aspects [53].

Econometric Modelling
The willingness to pay for eucalyptus firewood is evaluated using the contingent valuation method (CV). Respondents were asked the maximum range of price they are willing to pay.
To estimate the willingness to pay for eucalyptus firewood and to explore factors influencing willingness to pay, the double-bounded dichotomous choice contingent valuation model (DBDC-CV) [26,28,33,56] was used. In addition, following [27,33] studies the realistic bids ranges, come from the informal local market [55] were used.
Respondents are requested to answer a first question like the following "If eucalyptus for firewood cost 11 €/quintal ≤ WTPi ≤ 15 €/quintal, would you agree or disagree to it?", and a second question (for the respondents who agree to the first question) "If the price is > 15 €/quintal, would you agree or disagree to it?" or "If the price is 6 €/quintal ≤ WTPi < 11 €/quintal, would you agree or disagree to it?" (for the respondents who disagree to the first question).

Descriptive Statistics
The sample ( Table 1) is composed of 231 individuals, of which 143 males, with a mean age of about 43 years (S.D. = 12.21; range 25 to 80 years) and 59% of respondents have a low education level (i.e., primary or secondary school). Moreover, 57% of sample come from little towns.
Our findings show ( Table 2) that about 92% of sample buys firewood for domestic use, 38% consume firewood more than 3 times a week and 13% of respondents consume oak as firewood specie.
Based on the percentage of answers reported per each item, the percentage value for the item groups reveals a high attention to origin of wood (64.50%), followed by firewood species (60.61%), and by provenience (56.28%) of wood. Also, ethical aspects are important for 51.95% of respondents.
In addition, the chance to take information on firewood seems important; in this regard, people seemed more interested to take information from sellers (70.14%) followed by friends (70.13%) and from experts of the sector (59.30%). TV and internet as information sources are not highly perceived; in fact, only 10.83% of sample takes information by internet and only 6.50% of respondents by television.
In addition, 65% of the sample is willing to consume eucalyptus firewood, 55% had not heard about it, and 82% of respondents had never consumed eucalyptus in the past. However, about 49% of sample is willing to consume eucalyptus for curiosity and about 71% is willing to consume it if eucalyptus firewood showed a higher energy density than other firewood species. Moreover, under environmental aspects, 73% of respondents are willing to use it if eucalyptus was less impactful than other firewood species.
Finally, 58% of sample would prefer to buy loose firewood, followed by 24% of people that would prefer firewood in 10-15 kg bags. In addition, 46% of sample is not interested to packaging aspects of firewood.

Results of the Econometric Model
The double-bounded dichotomous choice model is performed on the WTPi for loose firewood, since this was the demand method most frequently chosen by respondents. 47% of the sample is willing to pay 6 €/quintal ≤ WTPi < 11 €/quintal of eucalyptus firewood, followed by 20% of respondents are ready to pay WTPi >15 €/quintal. Table 3 shows the results of the model with the estimated coefficients (β), their standard errors, marginal effects, significance levels, and goodness-of-fit statistics. The goodness of fit as measured by McFadden's pseudo-R 2 is equal to 0.26. The model makes it possible to highlight influential variables on the consumers' willingness to pay for eucalyptus firewood.
Our findings show that aspects such as information from friends and experts of the sector (i.e., agronomists, forestries), energetic density of firewood can shape the probability that respondents would be willing to pay.
Also, respondents' age and consumers' attention towards environmental issues are positively associated with their willingness to pay. This aspect could be an interesting point if you consider that firewood extraction is one of the many causes of deforestation and forest degradation at a world level [60].
Moreover, participants who pay attention to both firewood species and packaging are less willing to pay than other consumers.
Finally, in addition to the variables discussed above, the model also considered other variables as determinants of the willingness to pay, but none of these were found to be significant.

Discussion and Conclusions
This study aims to contribute to the current literature on willingness to pay for eucalyptus firewood by investigating the main factors that might affect the Italian consumers' willingness to pay. Although our sample is not representative of the Italian population, and thus the findings cannot be over-generalized, the findings should be interesting hints to address marketing strategies. Unfortunately, there are not many studies evaluating the willingness to pay for eucalyptus firewood that could help us to evaluate the findings. The sample composed of 62% males, with a mean age of about 43 years, confirming the current literature [61] and 40.69% of respondents have a high education level [61]. Moreover, 55% has not heard about eucalyptus firewood and 82% of respondents have never consumed eucalyptus in the past. In this regard, some authors (e.g., [62]) noted that the lack of familiarity with analyzed goods could cause unreliable responses in CV surveys; however, the question is a controversial issue [63]. In fact, [64] showed, according to standard micro-economic theory, that familiarity with goods is not a precondition for decision-making, but that in existing markets consumers also have to make decisions about new goods for which they do not have previous experience. So, lack of familiarity with analyzed goods does not affect the responses in surveys. Similar results are also reached by [65], who showed no relationship between previous experience of people and their willingness to pay.
Moreover, the initial range of prices offered in the study (11 €/quintal ≤ WTPi ≤ 15 €/quintal) was higher than willingness to pay of our sample (6 €/quintal ≤ WTPi < 11 €/quintal), confirming the current literature [6] where the sale price of eucalyptus was of 9 €/quintal even if according to [5] to have a profitable farm of eucalyptus, the sale price of wood should reach a value of 14.50 €/quintal.
Respondents' age is an important driver for willingness to pay; in fact, according to [54,66], the likelihood a respondent is willing to pay increases with age; while according to other authors [67], respondents' age is not statistically significant on the stated preference of paying.
The high willingness to consume eucalyptus firewood (65% of sample) may indicate that people are becoming more receptive towards eucalyptus as a good firewood alternative.
In addition, 73% of respondents would be willing to use it if eucalyptus was less impactful than other firewood species. Our findings showed consumers' attention towards environmental issues may be important drivers for willingness to pay. In fact, those with a higher environmental concern are willing to pay. This is an important aspect if you think that firewood use contributes to deforestation, particularly of native forests [55]. In fact, among researchers there is an unanimous consensus that the current level of biomass consumption threatens the sustainability of forests in many countries [68,69] and that the growth of firewood markets is correlated with environmental impacts like the degradation of forests and deforestation [55,70]. According to [55], there is a negative relationship between increases in firewood consumption and the sustainability of forest resources. In this framework, it could be useful to recourse to short rotation forestry (SRF) biomass (as eucalyptus cultivation) to avoid environmental impacts like deforestation. Moreover, since eucalyptus plantations in some countries (except USA) are managed under the auspices of sustainable forestry certification program [71], it could be interesting to study consumers' attitudes, preferences, and willingness to pay a premium price for certified eucalyptus. This is an area that will require further investigation.
According to [55], the most valued attributes by people are related directly to environmental issues and to technical aspects (energetic density and humidity). These aspects are linked, as stove exchange programs, focused on improvement the technical aspects as efficiency of firewood combustion, can lead to a decrease of environmental impacts like deforestation [72,73]. Improving production processes may help reduce the overexploitation of forests and increase wood combustion efficiency and improve air quality through establishing and enforcing wood quality standards (such as the humidity level) [74,75]. In our case, environmental and technical (as energetic density) aspects can shape the probability that respondents would be willing to pay.
Moreover, it is interesting to notice that the willingness to pay increases when people do not pay attention to firewood species. In other words, people are not loyal to particular wood species and are interested in the energetic density of wood, and they are willing to pay more. Similar behaviors are observed in other studies (e.g., see [76]) where consumers appreciate the intrinsic attributes of products rather than extrinsic ones. This is an interesting result given the calorific power value of eucalyptus is similar to that of the oak [22], and therefore the possible replaceability of eucalyptus to oak.
Finally, information received can also shape the probability that respondents would be willing to pay. In fact, [43] showed that information on firewood significantly affects consumer choices. According to our results, correct information may be a leverage to increase the willingness to pay for firewood, confirming the current literature [55].
The present study has a main limitation, common to most papers dealing with studies on the consumer: the sample is not representative of the Italian population. However, the use of web-based surveys is well established and broadly accepted in the literature on consumer choice. Moreover, on one hand, the result of the double-bounded choice could be affected by an initial bid during the interaction process; on the other hand, it allowed us to have accurate answers due to closed questions [57]. For these reasons, we are aware that the conclusions of this study cannot be over-generalized; however, we believe that the usefulness of pilot studies carried out on market issues should not be dismissed so easily. In fact, the emerged results could open new spaces for eucalyptus firewood, since the quantity and quality of information received, in particular on the environmental and technical aspects (energetic density), could shape the probability that people would be willing to pay. Even though we cannot over-generalize our results to the whole Italian population, we could conclude that a growing eucalyptus demand would offer an interesting opportunity for firms to enter the sector and develop marketing strategies targeted to specific market niches.
Author Contributions: N.P. contributed to the study design, data collection, data analysis, writing, and revising of the whole manuscript. A.S. contributed to the study design. L.P. contributed to the study supervision, project administration, and funding acquisition. All authors have read and agreed to the published version of the manuscript.