The Opportunity Cost of Households´
Recycling Efforts:
The Role of Norms and Warm-Glow Motives
Sara Andersson
Business and Economics, master's level 2020
Luleå University of Technology
ABSTRACT
The overall aim of this thesis is to examine households´ preferences towards relaxing the time demanding aspects of transporting sorted packaging waste. According to welfare economics, households should be willing to pay an amount equal to their reservation wage to benefit from more leisure time. However, specific types of norms, e.g., personal, self-internalized norms, and warm-glow motives may affect this decision. A theoretical model building on the notion that household members prefer to maintain a positive self-image as a responsible person is therefore used. In total 500 surveys were sent out to villas in the municipality of Skellefteå in order to gather information about recycling motives, socio-economic status and time spent on dropping-off packaging waste. The survey also presented a scenario similar to existing curbside recycling schemes in order to examine households’ willingness-to-pay (WTP) to relax the burden of transporting already sorted packaging waste. The empirical results show how both personal norms (giving rise to a feeling of guilty conscious if not recycling) and the warm-glow of giving affect the WTP for responsibility relief. Stronger personal norms implies a higher likelihood that persons have a positive willingness-to-pay for curbside recycling schemes, while strong warm-glow motives instead lead to a lower likelihood. Overall, the average WTP (per hour saved) tends to be lower than the opportunity cost of time measured through the average wage rate after tax.
SAMMANFATTNING
Denna rapport behandlar hushålls uppfattning gällande att minska tidskrävande transport av redan källsorterade förpackningar i hemmet. Individen borde enligt ekonomisk teori vara villig att betala motsvarande hans/hennes reservationslön för att frigöra ytterligare fritid. Däremot kan specifika normer, såsom personliga (självpålagda) normer, och så kallade ”warm-glow” motiv påverka beslutet. En teoretisk modell som bygger på idén att individer föredrar att upprätthålla en positiv självbild som en ansvarsfull person utnyttjas. 500 enkäter skickades ut till villor inom Skellefteå kommun med uppgift att samla in data gällande motiv till källsortering, socioekonomisk status och tid nedlagd på att transportera källsorterade förpackningar. Ett scenario likt nuvarande implementering av fastighetsnära insamling lades fram för att undersöka hushållens betalningsvilja (WTP) för inte behöva transportera källsorterade förpackningar till en återvinningsstation. Resultaten visar hur både personlig norm (som bland annat ger individen dåligt samvete av att inte källsortera) och motivet ”warm-glow” påverkar WTP för minskat ansvar. En starkare personlig norm medför en högre sannolikhet för en positiv betalningsvilja, medan ett starkt ”warm-glow” motiv i stället leder till en motsvarande lägre sannolikhet. Slutligen, den genomsnittliga betalningsviljan (per sparad timme) tenderar att vara avsevärt lägre än den genomsnittliga lönenivån (efter skatt).
ACKNOWLEDGMENTS
I would like to express my sincere gratitude to my supervisor Patrik Söderholm, for all the time he spent reading my drafts, opening survey envelopes, giving valuable advice and sharing his knowledge. It has been very much appreciated, and without his guidance my thesis would not have reached the same level. I would also like to thank my classmates whom commented on my work with open minds. My grateful thanks are also extended to my family for helping me distribute all the surveys. Finally, I wish to thank all survey respondents, who shared their valuable time and contributed to a response rate beyond my expectations.
TABLE OF CONTENTS
CHAPTER 1 INTRODUCTION...1
1.1 Background ... 1
1.2 The Objective of the Study ... 2
1.3 Methods ... 3
1.4 Scope of the Study ... 3
1.5 Outline ... 4
CHAPTER 2 THEORETICAL FRAMEWORK ...5
2.1 Recycling Efforts and the Opportunity Cost of Time... 5
2.2 Norm-Motivated Recycling Behavior ... 9
2.3 A Model of Utility Influenced by Social and Moral Norms ... 10
CHAPTER 3 LITERATURE REVIEW ...15
3.1 Search Strategy ... 15
3.2 The Valuation of Households´ Time in Cost-Benefit Analyses of Recycling . 15 3.3 Households´ Recycling Efforts and the Role of Norms ... 17
3.4 Conclusions Based on the Literature Review ... 22
CHAPTER 4 METHODOLOGY ...26
4.1 Data Collection ... 26
4.1.1 Survey design ... 26
4.1.2 The Contingent Valuation Method ... 27
4.1.3 Survey outline ... 29
4.2 Econometric Specification ... 34
CHAPTER 5 EMPIRICAL DATA AND ESTIMATION RESULTS ...37
5.1 Survey Responses ... 37
5.2 The Determinants of WTP ... 40
CHAPTER 6 DISCUSSION ...43
CHAPTER 7 CONCLUDING REMARKS AND AVENUES FOR FUTURE RESEARCH ... 48
REFERENCES ...51
APPENDIX 1 The Survey, Translated Version ...58
APPENDIX 2 The Survey, Original Version ...62
LIST OF TABLES AND FIGURES
Figure 1: The choice between (paid) work and leisure ... 6
Figure 2: The reservation wage ... 7
Figure 3: Maximum WTP for not having to transport sorted packaging waste ... 13
Table 1: Summary of studies estimating the cost of households’ recycling efforts ... 23
Table 2: Key statistics for Skellefteå municipality ... 27
Table 3: Descriptive statistics for the variables explaining WTP ... 33
Table 4: Key statistics for respondents ... 37
Table 5: Attitudes towards recycling ... 38
Table 6: WTP and the opportunity cost of time ... 39
Table 7: The different stated reasons for non-participation ... 40
Table 8: Heckit estimation results: excluding norms or warm-glow motives ... 40
Table 9: Heckit estimation results: including norms and warm-glow motives ... 42
1
CHAPTER 1 INTRODUCTION
1.1 Background
In 2018, 1 340 300 tons of packaging waste – plastic, paper, glass, metal and newsprint – were generated in Sweden (SCB, 2019-09-17). If all packaging waste would be recycled instead of using new raw material, 37 to 96 percent less carbon dioxide (CO2) would be released into the atmosphere (Hillman et al., 2015).1 Thus, increased household recycling efforts could have significant positive effects on CO2 reduction.
In Sweden, the producers have a liability to manage their packaging waste, and provide recycling alternatives to the consumers (SFS 2018:1462). Despite the producers’ liability, existing Swedish packaging waste recycling scheme are based on several time demanding activities for the individual households. The activities include cleaning the waste, sorting different materials, storage at home, and, for many households, transporting the waste to a recycling station. Regardless of the burdens of recycling, more than 70 percent of the packaging waste is now recycled in Sweden (SCB, 2019-09-17). Comparing household recycling in the European Union, households in Sweden have a high participation rate (EEA, 2019-11-22). Still, there is very little monitoring and control from the authorities if consumers comply with their recycling responsibilities or not.
From a pure private economic perspective, it may not pay-off for households to transport their packaging waste; it takes time that could instead have been spent on other activities, and the environmental benefits accruing to the single household will be very limited. Research states that the perceived utility for the individual recycler does not mainly result in an experienced better climate, but from feeling like a better person and improving their self-image when taking personal responsibility for the climate and the environment by
1 The reduced 𝐶𝑂
2 emissions by material: Glass: 41%; Aluminium: 96%; Plastic: 37%; paper and carbon:
2
adopting a pro-environmental behavior (Nyborg et al., 2006; Bruvoll, 1998; Radetzki, 1999). There are numerous of studies confirming that individuals benefit from recycling, partly because it makes them feel good and improves their self-image and/or the feeling of warm-glow (Berglund and Söderholm, 2009; Andreoni, 1990; Bruvoll and Nyborg, 2004).
In 2015, the European Commission adopted a “circular economy” plan (COM (2015) 614) and a clear development towards easing the households´ recycling burden can now be seen in European municipals (European Environment Agency [EEA], 2019-11-22). Nonetheless, Sweden fails reaching its recycling targets and with the objective to increase household recycling, the Swedish government has introduced new guidelines for curbside recycling as a way of easing households´ recycle burden by taking over the responsibility of transporting sorted packaging waste. By 2021, 60% of all of Sweden’s households need to have access to recycling bins connecting to their dwellings, and by the year 2025, everyone should have access to curbside recycling (GOS, 2018).
In conclusion, to bring sorted packaging waste can be perceived as a true burden with no gain for the households, but still households tend to be willing to commit. It is therefore interesting to investigate how households perceive the burdens of transporting sorted waste, and try to understand how different factors can influence the appearance of these burdens. This thesis will focus on the burdens of transporting the sorted waste regarding the new guidelines from the Swedish government concerning curbside recycling.
1.2 The Objective of the Study
The overall aim of this study is to examine households´ preferences towards relaxing the time-demanding aspects of transporting their already sorted packaging waste to drop-off stations. Specifically, the thesis elicits households´ willingness-to-pay for relief from the responsibility of transporting packaging waste, and examines why household members differ in their preferences towards responsibility relief.
3 1.3 Methods
The theoretical outset is Bruvoll and Nyborg´s (2004) model on how individuals prefer to maintain a positive self-image as a responsible person. Transportation of packaging waste should be perceived as a cost for the households and therefore an opportunity cost of time occurs when transporting packaging waste. Consistent with previous research, the presence of personal norms and the so-called “warm-glow motive” are believed to affect the preferences towards a responsibility relief of transporting packaging waste.
In order to measure the preferences, a handout survey with an open-ended willingness-to-pay question for easing the transport burden was constructed. A hypothetical scenario proposed an extended household service similar to already existing curbside recycling used in many Swedish municipalities (FTI, 2020). The survey was also designed to reveal underlying aspects in attitudes and norms towards the household burden of managing packaging waste in line with previous research on the subject. The three-page long survey was distributed in the Swedish municipality of Skellefteå with a pre-paid return envelope. Hoping the respondents would perceive the survey as lightsome, the survey cut down of questions compared to previous influential surveys which focused on the sorting of waste or different materials making it time demanding for the respondents. A total of 500 surveys was given to households´ mailbox with 285 surveys returned, i.e., a response rate of 57 percent.
The Heckman selection (Heckit) model was used in order to test the hypothesis regarding the influential effect norm-based and warm-glow motives could have on the stated WTP. The stated WTP was expected to differ from the traditional opportunity cost of leisure time (based on the after-tax wage rate).
1.4 Scope of the Study
This report will examine households in Skellefteå municipality’s WTP to ease the recycle burden of transporting their sorted packaging waste and attitudes affecting this. For this reason, the focus is only on the packaging waste already collected on existing recycling stations, in other terms paper, plastic, glass, metal and newsprint. This implies that households’ efforts in terms of sorting and cleaning the packaging waste, compost and hazardous recycling material, and construction materials, will be excluded.
4
The focus on transportation of packaging waste is partly due to the right to have curbside recycling to be imposed by the Swedish government (GOS, 2018). The selected case municipality is therefore entitled to start offering curbside recycling to all households due to the government´s decision. Previous studies have mainly focused on the sorting of the waste and the recycling attitudes amongst different materials (see also Chapter 3). When the transportation and drop-off at recycling stations are included in the average sorting time, the opportunity to announce the gained time of not having to transport the sorted waste becomes obstructed. In order to gain a deeper knowledge, each fraction may benefit being analyzed alone to see if any particular part of handling is more burdensome. This study will contribute with a survey excluding the sorting of packaging waste in order to gain a deeper knowledge of the influencing determinants of transporting already sorted waste. Earlier studies have found the presence of norms to affect household recycling, but they have not tested several specific norms to the same extent as is done in this paper (see further Chapter 3).
1.5 Outline
The thesis is divided into seven chapters. Chapter two will lead to a deeper knowledge in this subject defining the theoretical framework regarding household utility influenced by norms and the burdens of recycling. The third chapter will summarize previous literature concluding with a discussion, highlighting differences and similarities of earlier studies. In chapter four, we introduce the methodology of data collection and the construction of the survey. The fifth chapter summarizes the empirical results from the survey, which are discussed in the sixth chapter. The seventh, final, chapter provides concluding remarks, and outlines some avenues for future research.
5 CHAPTER 2
THEORETICAL FRAMEWORK
This chapter addresses the theoretical framework concerning utility maximization when deciding between alternatives in the context of household recycling efforts, i.e., the time spent on packaging waste recycling (with emphasis on the transportation of waste).
Homo economicus, the rational economic man or woman, always acts based on perfect knowledge, and makes optimal choices in order to maximize his/her utility. The first section will explain the theoretical view of the rational person, and how he/she decides to allocate the available time with an ambition of utility maximization; a simple model of household behavior is used to identify households’ opportunity cost of leisure time. The second section departs from the same basic model, but now adding the complexity of various types of norms effecting the household recycling decision. A specific model of a norm-motivated recycler is specified in the third section; here it is assumed that the individual has a preference for maintaining a positive self-image as a responsible, i.e., norm-compliant, person. This model can be employed to derive households’ willingness-to-pay for being freed from the responsibility to transport waste.
2.1 Recycling Efforts and the Opportunity Cost of Time
Time can be perceived as a scarce resource (Buchanan, 2008), with a limitation of 24 hours per day. Every day individuals make choices what to do with their time, and the economic value of the best alternative not chosen is called opportunity cost. In theory, an individual chooses between leisure time and work, where work occurs either in the labor market or in the form of unpaid household labor, e.g., doing household chores (Ibid). When an individual chooses between two different alternatives, doing a chore such as recycling or enjoy leisure time, opportunity costs will arise, as high as valued leisure time. Accordingly, leisure time is awakened time not spent on work, giving the individual a free choice how to utilize disposable time.
6
Leisure time and work can be perceived as goods with a possibility to substitute between the two alternatives at a diminishing rate (Baye and Prince, 2017). Thus, the alternatives of leisure and work giving utility can be seen as an indifference curve similar to I in Figure 1 and, as mentioned earlier, time is a scarce resource with a limitation of 24 hours per day, thus generating the time constraint T in Figure 1. An individual has the opportunity to spend as much time on work and leisure along line T, an individual who choose not to work would have 24 hours of leisure time, the horizontal intercept in Figure 1. Leisure time may seem more rewarding than work, but working generates income, which also is beneficial for the individual for having a convenient life (ibid). Therefore, the individual attempts to achieve the highest possible utility given the time constraint by balancing time spent on leisure and work. The indifference curve tangents the time limitation in point A, giving the optimal combination of income and leisure possible.
Figure 1: The choice between (paid) work and leisure
Johansson (1993) turns to the labor market in order to value the opportunity cost of time as the worker´s reservation wage. For an unemployed worker, the reservation wage is the wage needed for him/her to accept a job offer; since the worker was unemployed earlier, the employment does not give a negative impact in other sectors as a loss of production, but as a loss of valued leisure time for the individual. Figure 2 illustrates the social cost, SC, of hiring an unemployed individual where the worker´s reservation wage is w; note that the worker´s reservation wage may not be consistent with the paid salary.
7
The social cost occurs due to the lost leisure time the individual can enjoy, making it an opportunity cost to accept the job offer. According to Johansson (1993), the reservation wage for an unemployed individual possibly overstates the true social cost of lost leisure time, due to factors such as the length of unemployment and the presence of wage rates different than the equilibrium wage rate. In addition, employed workers have a higher reservation wage, because of their existing salary. In Figure 2, the reservation wage for an employed worker is 𝑤1 while the unemployed worker has the reservation wage w for giving up leisure time and choose to work.
Figure 2: The reservation wage
The reservation wage is used as an estimate of social cost when performing cost benefit analysis (CBA), and two widely used valuation methods for leisure time are the hourly market wage for purchasing the service to your household and the hourly after-tax wage (e.g., Murphy, 1978).
This estimate is the marginal price of leisure that provides the understanding when an individual would choose leisure before transporting sorted waste, but average time cost is in general not equal to marginal time cost (Becker, 1965). However, the average cost of time is still useful in order to be able to study household behavior, and it gives an indication for how households make decisions and the perceived cost (ibid). When time becomes more expensive, it induces a decline in time-intensive chores and commodities. Time-saving solutions will then be substituted for time-intense chores or commodities in
8
the households´ (ibid). Recycling is perceived to be a time-intensive chore, demanding of household time with little productivity gained for time effort put in. A rise in opportunity cost of time due to a higher salary or less leisure time would decrease time spent on recycling, thus inducing individuals to be less willing to forsake leisure time (Radetzki, 1999). Consequently, reallocation of transportation time in order to gain leisure will only happen if the individual perceive the burden of transporting sorted waste higher than the opportunity cost of time.
Previous recycling studies (reviewed in Chapter 3) have all used the hourly after-tax wage as a proxy for lost leisure time; this approach assumes individuals to be rational and on the marginal yield feel indifferent between spending the last hour on leisure or waste recycling. This implies how the same theoretical approach as the choice between leisure and work will now be a choice between leisure or waste recycling. This is because of the assumption that households only recycle their private packaging waste during leisure time. The opportunity cost of time according to this method is 100 SEK (USD 10) per hour
in Sweden(SCB, 2019-02-21).2
Previous research has however indicated an overestimation when using the after-tax wage compared to the estimates based on the so-called Contingent Valuation Method (CVM) (e.g., Berglund and Söderholm, 2009; Nainggolian et al., 2019), indicating that the opportunity cost of time tends to be lower than the theoretically derived reservation wage.
The upcoming sections in this chapter will provide and explain economic models useful for recycling behavior, not least describing potential reasons why the reservation wage can be higher than the perceived opportunity cost of time. This is done by incorporating the presence of norms to the theory of the rational economic man, influencing the decision of how to allocate given leisure time. Due to the assumption that household recycling only occurs during awaken leisure time, labor supply is assumed to be exogenously given from now on.
2 Based on the average salary in Sweden in the year 2018, divided by the average working hours per year and the average tax-rate with numbers from SCB (2019-02-21b).
9 2.2 Norm-Motivated Recycling Behavior
Andreoni (1990) developed an impure altruism model stating that individuals seldom act solely altruistic. Selfless actions gives utility in form of satisfaction from the pleasure of doing good and contributing to a good cause. This is called the warm-glow of giving and contradicts the neutrality hypothesis stating that rational individuals are indifferent to how a cause is funded. In the context of utility-maximizing recycling behavior, warm-glow implies time spent on recycling also is a gain for the individual. Policy makers use the theory as a commonplace course of action when hoping to design efficient environmental policies (D´Amato et al., 2014).
With the intention of understanding market behavior, the adherence to norms has also been explored by economists (e.g., Akerlof, 1980; Halvorsen, 2008). Social norms are the informal rules in groups and cultures where a person considers what others expect, and desires to gain social acceptance. The fairness norm is also within the spectrum of social expectations, believing everyone should do what is fair to others and creating adverse behavior if something is not perceived as fair. Another kind of norm is the moral reasoning deciding what obligations having to live up to, which can give a bad taste if the individual does not achieve these obligations. Halvorsen (2008) defines this norm as what individuals consider is a moral obligation. The difference compared to the social norm is that now the individual sanctions himself/herself. The moral reasoning, so-called personal norm, is the focus also in the theoretical model of Brekke et al. (2003).
Brekke et al. (2003) argue that the warm-glow theory excludes the complexity of moral reasoning considering individual obligation and social values; individuals do not only gain utility by giving, they must believe giving is beneficial. These authors design a model with the purpose to explain how moral motivation, economic incentives and public policy relate to actual consumer choices. Thus, according to this research, the ideal recycling effort is endogenous, and solely moral norms will influence the individuals’ recycling behavior. Individuals ask themselves what the personal ideal behavior in every situation is, and they then make tradeoffs diverging from this socially responsible choice in order to, for instance, gain leisure. The opportunity costs are measured in consumption units where the individual can buy units of the public good or leisure good. New political guidelines can therefore change actual consumer choices via internal moral deliberation, and explains how this increase required recycling effort.
10
Bruvoll and Nyborg (2004) also reflect how Andreoni´s (1990) model does not explain the mechanisms behind the feeling of warm-glow, and the reason why an advertising campaign or a new guideline would increase recycling participation and/or experienced utility. The authors follow the essence of the Brekke et al. (2003) model with the main difference of the required recycling effort by the norm is given exogenously. They employ a norm requirement, meaning an unchanged norm requirement would yield the same conclusion as Andreoni´s (1990) model whilst a changed norm requirement imposed by a policy change would give other conclusions and possibly increase the perceived burden. The benefits of household recycling are, according to the authors, divided into two parts, the environmental benefits and the benefit of an enhanced so-called self-image. People are expected to have preferences for maintaining an image as responsible, i.e., norm-compliant, persons.
With the aim to examine individuals’ preferences and the norms underlying these, the choice of model is Bruvoll and Nyborg´s (2004). This model includes moral norms as in Brekke et al. (2003), but with the incorporation of a required amount of recycling given exogenously; it explains how a changed policy may alter the self-image directly, making it appropriate for this thesis’s objective. Next section will explain the model of individuals self-image. Then some modifications of the model are made to suit our empirical survey, in part influenced by Halvorsen (2008) and Berglund and Söderholm (2009).
2.3 A Model of Utility Influenced by Social and Moral Norms
The utility for an individual is assumed to be affected by the person’s self-image, S; individuals desire to be a responsible person and will therefore adjust to certain exogenous norms of responsible behavior. Other utility-increasing elements are the consumption of private goods, c, environmental quality, a public good, G, and leisure time, l. Based on these four variables, we can formulate the following general utility function:
U = u(c, l, G, S) (1)
Emphasizing on the time spent on leisure time and recycling activities, income and labor supply are exogenous in the model. The constraint in terms of awaken time is therefore:
11
T is the amount of time left for leisure, l, and recycling activities, 𝑒 = (𝑒𝑇𝑅+ 𝑒𝑆). The time spent on recycling, e, is here separated into two different categories, the sorting of packaging waste (i.e. time spent at home cleaning and handling the waste), 𝑒𝑆, and time needed for transport and dropping-off waste, 𝑒𝑇𝑅. Since focus is on the latter in this thesis, the sorting of waste in the household will be exogenous in the model, thus meaning that a change in 𝑒𝑇𝑅 does not affect the efforts spent on sorting the waste.
Environmental quality is assumed to increase through the provision of recycled packaging waste, either supplied by others exogenously, 𝐺−1, or through recycling efforts made by the individual, g:
𝐺 = 𝐺−1+ 𝑔 (3)
Households in Sweden are required by law to sort and recycle bought packages (SFS 2011:927), but insufficient control of disposable waste and the lack of legal enforcement give the law an insignificant tenor. Consequently, households’ contribution is assumed to be based on voluntary grounds with only the time effort determining the household´s contribution to recycling. Contribution to environmental quality will be zero if the household spend no time on recycling, and the time spent on recycling will benefit the environment at a diminishing rate:
𝑔 = g(e) (4)
Existing policy is assumed to provide the specific norm of recycling effort, g*, to the individual exogenously and the individual strive to achieve the norm in order to gain a positive self-image, S. The individual strives to achieve the expected effort, g*, and if insufficient effort is made, the self-image wanes. The norm is adopted to the extent that the individual feels an obligation to achieve the norm; the lack of legal enforcement does not matter in order to follow the norm. Bruvoll and Nyborg (2004) suggest that the self-image is a function of the difference between the actual contribution, g, and the norm requirement, g*, implying a personal sanction if the norm is not achieved:
12
Bruvoll and Nyborg also assume that the individual’s self-image cannot increase after reaching the norm requirement, g* (i.e. S´ > 0 if g < g* and S´=0 if g ≥ g*). The self-image then reaches a maximum when g=g*; this is essentially how Andreoni´s (1990) theory differentiates from Bruvoll and Nyborg´s. The self-image function (5) unite with Andreoni´s argumentation how individuals contribute to obtain a “warm glow of giving” depending on the level of contribution (Bruvoll and Nyborg, 2004). Nonetheless, in Andreoni´s model self-image never reaches a maximum (always S´ > 0 if g > 0) Hence, if the norm requirement, g*, is unchanged Bruvoll and Nyborgs model would yield the same conclusion as Andreoni´s (ibid).
The fundamentals of the Bruvoll and Nyborg model make it possible to discuss the case of curbside recycling, and the possible welfare effects of shifting one part of the recycling responsibility from the household to a third party (e.g. central authority or an entrepreneur – in Sweden´s case, Packaging´s and Newspaper´s Collector [FTI]). The norm requirement will then decrease as the transport of waste does not avail the households; when 𝑒𝑇𝑅 = 0 the norm, g*, will only contain the sorting of packaging waste, 𝑔∗ = 𝑔(𝑒𝑆∗). When the time used for transportation of waste no longer effects the norm requirement and sorting of waste is expected to stay the same, self-image increase as −𝑑𝑔 ≤ −𝑑𝑔∗.
The households stated compensating variation (CV) for letting others transport packaging waste instead of themselves should equal the reduction in expenditures needed to sustain the same utility when no transport of packaging waste is made by the household. In other words, the stated CV is the amount of money where the household is indifferent between transporting packaging waste, forsaking leisure time, or leaving these chores to others. This means utility effects from recycling efforts can be measured in monetary terms by the compensating surplus. The WTP measured in the survey can be defined indirectly by the equality:
𝑢(𝑐, 𝑇 − 𝑒0, 𝐺, 𝑆0) = 𝑢(𝑐 − 𝑊𝑇𝑃, 𝑇 − 𝑒1, 𝐺, 𝑆1) (6)
the superscripts of e and S denote the initial situation, 0, and the changed situation when the responsibility is shifted from the individual, l. The equality in equation (6) assumes the individual perceives environmental quality to be unchanged despite the shift in
13
responsibility of transporting the packaging waste. Changing to curbside recycling allows the individual to spend more time for leisure when less time is needed for recycling effort (𝑒0 ≥ 𝑒1), and the changed norm is most likely to give a small increase of self-image. People should then, in theory, be willing to pay for more leisure time in order to gain a responsibility relief of having to transport packaging waste. This can be compared to the case where voluntary recycling behavior is based on warm-glow motives; in such a case, people would not be willing to pay anything to gain extra leisure time.
As seen in Figure 3, with leisure time, l, on the vertical axis and recycling effort, e, on the horizontal axis. Starting in point A on the indifference curve I, where the individual has chosen the combination 𝑙0 and 𝑒0. If l would increase from 𝑙0 𝑡𝑜 𝑙1, the individual will obtain more utility, equivalent to point B. However, in order to increase l and maintain the same level of utility the individual have to decrease recycling effort, (𝑒0 𝑡𝑜 𝑒1) and the amount paid is the opportunity cost of time to keep the utility level constant and at the same time enjoy more l in point C.
Figure 3: Maximum WTP for not having to transport sorted packaging waste
Bruvoll and Nyborg´s model gives an explanation how municipalities with the help of advertising can affect the norm of expected recycling efforts and therefore the recycling outcome, due to individuals strive of feeling responsible and maintaining a good self-image. Still, it does not explain why the recycling effort differs across households. Previous research shows how norms could affect the contribution to the public good, and
14
how responsibility relief might decrease individual utility (Berglund and Söderholm, 2009; Halvorsen, 2008). In this empirical investigation, the focus is to find out how different norms can affect the preferences towards curbside recycling. For that reason, the focus is not changing the self-image function (5) in order to explain the perceived recycle burden, but instead investigate if specific norm-motives influence the perceived burden.
If an individual has strong norms and a habit of recycling, making it a perceived low-cost activity, the probability of declining curbside recycling increases (Halvorsen, 2008). The discussion above also opens the possibility to a loss in self-image when receiving responsibility relief, and therefore a low stated WTP. An individual would lose utility if the feeling of warm-glow when recycling is reduced, i.e., the expected outcome is that individuals with warm-glow motives are not willing to pay for responsibility relief. The presence of moral (self-sanctioned) norms would, in contrary, increase the willingness-to-pay for a responsibility relief (Bruvoll and Nyborg, 2004).
15 CHAPTER 3 LITERATURE REVIEW
This chapter will mainly consist of a review of earlier literature with a focus to evaluate the opportunity cost of time and attitudes influencing the perceived cost of household recycling. The emphasis is on the attitudes towards changing to curbside recycling and/or any corresponding alternatives easing the household recycling burden.
3.1 Search Strategy
To collect as many relevant research articles as possible, Ebscohost, Google Scholar and Web of Science were used for the literature study. The main keywords used were curbside recycling, norms, opportunity cost of time, and household burden. Many relevant articles were found with useful incorporated lists of stated preferences regarding recycling behavior and other surveys, which gave numerous other articles in this subject.
3.2 The Valuation of Households´ Time in Cost-Benefit Analyses of Recycling Welfare economics is based on microeconomic theory in order to evaluate well-being in the economy. The principles of how the “rational economic man” acts are used when economic policy is evaluated using cost-benefit analysis (CBA) in order to examine real world trade offs. Many environmental policies are designed and evaluated using this theoretical framework of welfare economics and in the case of CBA of various recycling schemes, the opportunity cost of time is an important post that should be included in CBA using empirical evidence (Gowdy, 2007).
Radetzki (1999) mentions the importance of including the external costs pending from the sorting, cleaning and transporting packaging waste. No one questions whether these costs should be included in CBAs when companies perform these tasks, but still many studies disregard the external costs of recycling if households employ them. Arguments that households voluntarily sort their own waste, and the opportunity cost is equal to zero for inter-household recycling are according to Radetzki exorbitant. The households´ time for sorting waste must also be assigned a positive value. Some households may feel a
16
responsibility to contribute towards the “sustainable society” and would probably recycle without receiving any compensation. However, voluntary contributions should be seen as a scarce resource because with good reason, the willingness to voluntary contribute to other needs will probably decline in relation to the effort employed on recycling. In addition, waste generators are liable to employ recycling in Sweden (SFS 2011:927), therefore it cannot be entirely seen as a voluntary task. An economic valuation of the time spent by households should therefore be included in CBAs of various recycling schemes (Radetzki, 1999).
Bruvoll (1998) compares the costs of four alternative policies for paper and plastic waste, and found the overall positive view on household recycling to be misleading when both environmental and economic costs for recycling exceed other alternatives. The view of recycling representing a way for individuals to be a part of the solution and solving the environmental problems may be one major reason for the popularity of recycling. Still, the private time use is often not assigned an alternative value, and Bruvoll (1998) includes an opportunity cost of time lower than average net wage in Norway. High conventional collection costs and air emissions costs from transporting plastic makes recycling the most expensive alternative for plastic. High environmental and economic transportation costs are similarly the reason why recycling is not the preferable alternative for household paper waste.
The Swedish Environmental Research Institute [IVL] compared two recycling policies using CBA where the evaluation of household time had a significant effect on the preferred choice (IVL, 2010). One of the systems is more dependent on the transportation of waste, making it an expensive system for the households. The cost assessment for household time decides which system is more economically beneficial, making it an important cost item in order to make a proper judgement. Sahlin et al. (2007) developed an index based on the average time spent on recycling from previous studies, and this has been used by IVL. This time is called a “best estimate” of the extra time spent on chores due to sorting and transporting packaging waste. The time spent on transporting waste is computed to be between 33 to 67 hours per ton of packaging waste depending on material according to Sahlin et al. (2007). Whereas the time is valued to be between 6 SEK per hour up to 66 SEK per hour in IVL´s report (2010). The uncertain valuation of time gives a big spectrum determining if a system is preferable or not. IVL stresses the importance
17
to view CBA results with precautions but estimates the results to be the best assumption given.
Consequently, engaging in recycling may be more or less seen as a voluntary chore in Sweden, but it is nonetheless a cost that should be included in CBAs. Households´ time is a scarce resource, and the current recycling scheme is indeed time demanding for the households (Hage et al., 2009; Radetzki, 1999). It is therefore of great importance to have an accurate estimate of the total household cost when comparing recycling schemes.
3.3 Households´ Recycling Efforts and the Role of Norms
When governments try to make individuals contribute to the provision of public goods, e.g., a cleaner environment, the appeal of consumer responsibility is typically a more common approach instead of using economic incentives. Bruvoll and Nyborg (2004) present a theoretical framework introducing norms as an explanation to why campaigns positioning a perceived responsibility on consumers may work in order to increase recycling participation, even though the standard consumer theory does not explain this outcome. Building a theoretical framework explaining an individual’s utility function is the main purpose of the study. However, the findings are illustrated with the help of empirical data gathered from a Norwegian survey. The results indicate a social cost when increasing perceived consumer responsibility, yet, the warm-glow benefit obtained by contributing may outweigh the cost.
Nyborg et al. (2006) attempt to explain the green consumer phenomenon using moral motivation, and its potential influences using previous literature made in the field of economics and social sanctions, such as Akerlof (1980) and Lai et al. (2003). However, Nyborg et al. (2006) stress the gap social sanctions theory has, because it does not explain why a person would engage in a pro-environmental behavior in cases where individual actions are unobservable, such as recycling. Earlier studies disregard aspects of social approval, which typically gives one utility maximizing equilibrium. Whereas Nyborg et al. (2006) argue for a multiple equilibrium in a so-called Nash coordination game, where two different outcomes depend on the participation rate. One of these outcomes is a no-participation equilibrium and the other a high-no-participation equilibrium.
18
A special caveat regarding the use of policies when trying to develop a pro-environmental norm is when no particular institution or individual is formally given the responsibility for the environmental problems, according to Brekke et al. (2003). For example, a policy recommendation such as a tax may give a consumer two different interpretations: (a) a symbolic device reminding the consumers responsibility; or (b) that the government has taken over the responsibility. The policy recommendations given is to make tax rates or subsidy rates high enough to push the behavior to the green choice, at levels that are strictly lower than the standard use of a so-called Pigouvian tax3 (Nyborg et al., 2006). Also, when the adaptation behavior reaches a certain level, such a tax can be removed without any changes in behavior.
Hence, Nyborg et al. (2006) stated how moral motivation could stimulate the demand of consumer goods that are believed to give improved environmental quality as an external benefit. This implies that environmental policies can, and should be, adjusted to only change beliefs about participation rates instead of having to make people purchase the green good (i.e., recycle, buy green products, use electric cars). The authors comment on the possibility of curbside recycling as a way of making recycling behavior visible and gain social status in the community.
Berglund (2006) analyzes households’ perception of recycling activities in order to see if moral motives matter when valuing waste sorting cost. From an economic point of view, recycling should be seen as any other activity, therefore, a utility maximizing person has an opportunity cost of leisure when recycling. Nevertheless, people may also derive benefits when taking responsibility for recycling, i.e., the feeling of joy and satisfaction referred to as “warm-glow” (Andreoni, 1990). Therefore, the time used for household recycling cannot entirely be seen as a cost (see also Sterner and Bartelings, 1999).
In order to estimate household costs, a mail-out survey was sent out to 800 individuals in the city of Piteå, Sweden, with a response rate of 70 percent. Questions about respondents’ attitude towards recycling waste, statements about recycling, the average time spent on recycling and an open-ended question for WTP for someone else to sort their waste was included. The result in Berglund (2006) showed how moral norms may have a significant
3 The Pigouvian tax is where, in the presence of negative external effects, the tax equals the value of the marginal external damages.
19
impact on the costs associated with recycling efforts; the respondents with high moral perceived their cost to be lower.
Berglund and Söderholm (2009) is another Swedish example of integrating norm-motivated recycling behavior into economic analysis. The authors build on the model of norm-motivated household recycling developed by Bruvoll and Nyborg (2004). The objective was to study Swedish households’ preferences towards less responsibility not having to transport the waste to drop-off stations and their WTP. A postal survey to Swedish households was sent out with the aim to measure household benefits and cost in terms of recycling and WTP for relieving the transport burden. The survey used an open-ended contingent valuation (OE-CVM) approach where respondents valued their WTP for a hypothetical change in the transportation burden via an open-ended bid. The utility given the recycler was the benefit of a better environment and better self-image minus the cost, measured in alternative cost of time. The results from the survey of 1115 participants indicate a lower WTP than welfare economics would presume, and if the individual has strong moral norms, the WTP can be negative. The negative WTP can be put on par with McDonald and Oates (2003) results for some non-participants who chose to recycle without a curbside scheme. Motives such as “warm-glow” appear to be an important determinant, aligned with Halvorsen´s (2008) findings. This indicates a mixed blessing when households may lose utility when the government introduces recycling schemes with less responsibility for the households (Berglund and Söderholm, 2009; Nyborg et al., 2006).
Many respondents were satisfied with the existing system in Berglund and Söderholm´s (2009) study, and the majority did not want anyone else to do something that they felt pleased to pursue on their own. Czajkowski et al. (2014) also report respondents wanting to sort their own waste even if costs (i.e. effort and time cost) are higher than if paying someone to do it. In economic terms, despite it being a time-consuming habit, Berglund and Söderholm (2009) conclude the current recycling system seems to be a cost-effective way of contributing to environmental private goods from the household’s perspective.
Almost forty percent agreed with recycling being a pleasant activity in Bruvoll et al. (2002), a Norwegian study which included 1162 interviews which were held year 1999 regarding the extent of, and attitudes towards, their own waste sorting activities. Using
20
earlier studies where the focus mainly was to find a clue on how to increase voluntary recycling, they wanted to see if recycling should be considered as a cost by households (even if it is a voluntary action). In the survey, the respondents were asked to state their WTP for others to handle the sorting of packaging waste using an open-ended CVM approach. According to their study, recycling should not be perceived a voluntary action, especially when the society has policies inferring individual responsibility. This is perceived as a burden for individuals who care about their self-image being morally responsible individuals. Each individual has different costs for the recycle burden depending on moral motivation, which in turn is divided in pleasure of giving and moral responsibility.
Hage et al. (2009) analyze recycling efforts focusing on the motivation among Swedish households using a postal survey. Many environmental policies require the involvement of citizens behavior. That is why the paper aims to find explanations to differences in household participation, and how policies can stimulate households to recycle more. The authors find a significant correlation between motives, such as economic and moral ones, and household recycling rates. Age is the only socioeconomic variable explaining waste recycling behavior. Another survey in the municipality of Kiruna, Sweden, got similar results (Vencatasawmy et al., 2000). Many international studies highlight education and income having an important role (e.g., Kipperberg and Larson, 2012; Huhtala, 2010; Sidique et al., 2010). Self-enforced moral norms play a key role in the understanding of recycling outcomes. Nevertheless, if the recycling scheme involves an easier collection of packaging waste, the importance of norms may diminish (Hage et al., 2009).
Halvorsen (2008) developed a model explaining how social and moral norms may affect a household´s recycling decision through feelings of self-respect, conscience, “warm-glow” and the importance of community. Also. opportunity cost of time affects recycling activities, even though the main motivation for recycling efforts were perceived as non-economic. Data are used from a survey asking about recycling efforts, motivation and WTP for allowing others to sort their waste. The survey has 1162 Norwegian respondents and was handed out in 1999. The results find opportunity cost of time has a negative effect on household recycling. Furthermore, recycling motivation appears to have a large impact on recycling efforts. If recycling is perceived a mandatory task, it does not necessarily decrease recycling motivation.
21
Czajkowski et al. (2017) investigate how selfish, social and moral incentives or pressures reflect individuals’ pro-environmental behavior. The empirical investigation checks Polish choice of waste contracts via a choice experiment where three types of factors are included and compared in the model and influence the recycling decision. These are: economic factors; social pressures; and personal moral sentiments. Respondents´ attitudes were then structured in three main latent variables where two groups capture strong norm-based motivation with a positive view of recycling, and the third having weak norms towards recycling. A substantial share of the respondents prefer to sort on their own rather than handing over the burden to a third party, in this case a central sorting facility. The authors also find a link between WTP and moral or intrinsic motivation, believing sorting at home is more thorough than a central facility would sort.
Aadland and Caplan (2003) investigate different hypothetical biases in order to estimate WTP for curbside recycling. Using CVM, the primary aim is to estimate an accurate relationship between changes in fees and participation rates. The results find several specific characteristics related to WTP for curbside recycling such as gender, age, education and level of ethical responsibility. Young, well-educated women evolved in an environmental organization have the highest WTP. Larger households stated a lower WTP, these households may experience a higher cost implementing recycling behavior among all members in the household. There are findings of statistically significant bias, indicating stated WTP is higher compared to revealed WTP.
Karousakis and Birol (2008) used a choice experiment with the purpose of measuring household preferences for employing curbside recycling in London. Where the purpose of the study was to find the determinants for household recycling behavior and the most valued recycling service attributes. The individual’s utility from recycling depends on the attributes of the service, recycling attitude and various social economic variables of the respondent. Three indices were created, measuring thedeterminants of WTP; recycling constraint, environmental concern and recycling motivation. When evaluating the results, the respondent’s characteristics found to have a great impact on WTP and are statistically significant, in line with welfare economics (Brekke et al., 2010; Heller and Vatn, 2017).
Nainggolan et al. (2019) performed a choice experiment postal survey in Denmark in order to estimate average WTP for not having to sort waste. Reading previous economic
22
behavioral literature, the authors identified five factors important for recycling behavior. The factors are: 1) believing the scheme reduces negative environmental impact; 2) if the recycled material will be taken care of and re-enter as a new product; 3) effort sorting the waste; 4) hassle of using the various schemes; and 5) importance of factors such as self-image, social influence and existing facilities in order to have the motivation. Based on the responses, the respondents were divided into four different segments based on home waste sorting preferences: the frequent waste collection resistors, the sceptics, the sorting enthusiastic and the bio-waste sorting resistors. The findings show households are most likely to lower recycling participation if the waste handling scheme is expensive and with high time burden. Individuals are also more likely to participate if they believe the sorted waste will be recycled. The average WTP was only a fraction of the minimum wage in Denmark, indicating home sorting implies an opportunity cost of time, but cannot be based on the market wage.
3.4 Conclusions Based on the Literature Review
The literature summarized in Table 1 all examine preferences for easing the recycling burden and attitudes towards recycling, mainly by using the CVM approach (Aadland and Caplan, 2003; Bruvoll et al., 2002; Berglund 2006; Berglund and Söderholm, 2009; Czajkowski et al., 2017; Halvorsen, 2008; Bruvoll and Nyborg 2004; Bruvoll et al., 2002) or so-called choice experiments (CEs) (Nainggolian et al., 2019; Karousakis and Birol, 2008). Both methods use a direct valuation approach with an aim to capture peoples valuation in monetary terms for environmental good (Perman et al., 2011). The surveys have been developed using theories from both welfare economics and psychological theory (e.g., the theory of planned behavior), in order to link the factors underlying recycling decisions. This implies different types of econometric models and the determinants for WTP cannot be translated between the different surveys, but can be seen as indications when comparing determinants.
23
Table 1: Summary of studies estimating the cost of households’ recycling efforts
Author(s) and country
Purpose Method Respondents Average WTP Main Conclusions Aadland and Caplan (2003) US Estimate WTP for curbside recycling DC-CVM Approx. 1000 USD 7/ month (curbside)
Cheap talk may reduce hypothetical bias Berglund (2006) Sweden Analyze whether moral motives matter for waste sorting costs OE-CVM 603 USD 27.37/ year Moral motives significantly lower perceived household recycling costs. Berglund and Söderholm (2009) Sweden Household preferences for curbside recycling OE-CVM 1280 USD 19/ year (curbside) Warm-glow is one of the most important determinants of WTP Bruvoll and Nyborg (2004) Norway Social Cost of Recycling Campaigns OE- and DC- CVM 1162 USD 30/ Year For most consumers, recycling effort is considered as a burden Bruvoll et al. (2002) Norway Motives for sorting waste OE- and DC- CVM 1162 USD 20/ year Sorting activities depends on several motives and the desire of a better environment Czajkowski et al. (2017) Poland How selfish moral and social incentives explain behavior DC-CVM 418 USD 13 Recycling participation is linked to moral or intrinsic motivation Halvorsen (2008) Norway How the opportunity cost of time affects recycling activities OE- and DC- CVM 1162 USD 28.16/ year Opportunity cost of time has a significant negative effect on household recycling. Karousakis and Birol (2008) England Examine the determinants of household recycling behavior CE 188 2.678 £/ month Social, economic and attitudinal characteristics determines WTP Nainggolian et al. (2019) Denmark Factors shaping propensity to engage in household sorting CE 1011 2.8-6.3 EUR/h 4 different types of “waste sorters”
24
Non-Swedish studies find a statistically significant relationship between recycling rates and variables such as income and education (e.g., Kipperberg and Larson, 2012; Huhtala, 2010; Karousakis and Birol, 2008), whereas several Swedish studies do not find similar results (e.g., Vencatasawmy et al., 2000; Hage et al., 2009). The present study will employ a questionnaire in order to capture the opportunity cost of time for transportation of waste to a recycling station, and the factors influencing this valuation. The questionnaire will outline a hypothetical scenario where an entrepreneur takes over the entire transportation responsibility, and use the CVM approach to estimate households’ WTP regarding this service.
Compared to foreign households, Swedes tend to have a lower WTP for easing the recycling burden. The recycling activities and the attitudes towards recycling have been rising in Sweden during the last decade (FTI, 2019-01-21), but to the author´s knowledge recent WTP studies for easing the recycle burden in Sweden are absent. Therefore, this study will contribute with up to date information about preferences and attitudes towards easing the recycling burden in line with current policies for curbside recycling. Previous Swedish studies have based the household´s recycling efforts on Bruvoll and Nyborg (2004) and Brekke et al. (2003) model, with some modifications. This study will also base the theory behind recycling effort using Bruvoll and Nyborg´s (2004) model and integrate additional specific norms. This will deepen the knowledge about which specific norms influence the decision, as previous Swedish studies have not examined.
Berglund and Söderholm (2009) have the same fundamentals as this thesis, where the households are relieved from their responsibility for transporting their sorted waste. This thesis will bring up to date information with a similar scenario and focus more on the role of different types of norms, and how these could affect the preferences towards a responsibility relief. Berglund and Söderholm (2009) also received a fairly low response rate (30 percent), and according to the authors this was due to a demanding survey for the respondents. Therefore, this thesis will create a simple and shorter survey, thus hoping to increase the response rate.
While this thesis focuses on the transportation of packaging waste, the majority of the reviewed studies focus on the recycling activity as a combination of several elements, with transportation included amongst sorting and cleaning. Especially sorting of
25
packaging waste has been researched in various studies and is a common approach. Some studies are from the beginning of the 21st century, and the stated WTP might be assumed to be higher, if not simply due to inflation, but since the surveys differ in many other aspects and the WTP estimates should only be seen with precaution and as a guideline, no attempt to create a price index is made. With that said, the included studies are still assumed to bring valuable knowledge leading to the thesis’s objective, and should be compared with the following results coming from the survey.
26 CHAPTER 4 METHODOLOGY
This chapter consists of the discussion of method selection, and a discussion of how the empirical study was designed.
4.1 Data Collection 4.1.1 Survey design
With an aim to conduct a quantitative study, 500 postal surveys with an attached pre-paid envelope were handed out in different neighborhoods in Skellefteå municipality, Sweden. The areas were chosen to attract responses from different socio-demographic levels of the municipality and distances to recycling stations, both in urban areas and villages. In order to do this, official average wages for postal codes were used as a guideline when deciding the participation areas, and seven postal codes were selected representing areas that all had different income levels. Thereafter, the blocks within the chosen postal codes were randomly selected and the surveys was delivered to the villa’s mailboxes by foot.
The municipality of Skellefteå is situated in the County of Västerbotten in the north east of coastal Sweden. The municipality has a population of just over 72 000 out of which almost 36 000 live in urban areas (SCB, 2019-02-21). An average inhabitant in Skellefteå recycle 70 kilos of waste per year, and the Swedish average is 62 kilos of waste recycled per year (FTI, 2020-03-04). Skellefteå´s waste management system resembles those of other Swedish municipalities; the garbage fees are mandatory and volume-based where the inhabitants have one bin for the combustible waste and have the possibility to take care of the compost on their own or add another bin to their waste scheme. Households are also required to sort out and bring the clean packaging waste (paper, glass, metal, and newsprint) to a recycling (drop-off) station maintained by FTI, according to the consumer responsibility (SFS 2011:927). In Table 2, key statistics for Skellefteå municipality are compared with the corresponding data for Sweden as a whole.
27
Table 2: Key statistics for Skellefteå municipality
Subject Skellefteå
municipality
Sweden
Age 43.5 years 41.3 years
Salary (SEK/Year) 302 600 318 200
Employed (share) 64.9% 68.3%
Women (share) 49% 50%
University graduates (share) 16.7% 21.7%
No. household members 2.8 3
Source: SCB (2019-02-21)
As the preference for the transportation of waste is the main purpose with the survey, no apartments or multiple dwellings with a close, defined recycling area specifically for the residents were included in the sample. This is in order to be certain that the presented WTP scenario represents a clear change in behavior for the respondents (see below).
The time and budget limitations made it difficult to hand out more than 500 surveys, and for the same reasons no reminders were handed out. Nevertheless, the number of surveys returned (57 percent) is assumed to give a hint whether transporting packaging waste is a burden for households and if norms affect the preferences. The limitation of only including houses in the empirical investigation may give a sample not representative for the municipality. However, it is a big segment and they can decide over their recycling schemes, making it more realistic to answer the question if they want to participate in curbside recycling or not. However, a realistic scenario can give sample of protest bids and strategic bids if the respondents believe that the bids can be used as guidance when deciding the fee, see following sections why these biases are assumed to be low.
4.1.2 The Contingent Valuation Method
The survey was based on previous recycling literature and previous applications of stated preferences (SP) regarding the household´s burden of recycling. Contingent valuation method (CVM) and choice experiment (CE) are the two main methods for SP (Perman et al., 2011). Both methods can measure non-use values and potential presence of norms influencing SP. CVM describes the hypothetical good or service and respondents are at that time asked to state willingness-to-pay (WTP) or willingness-to-accept (WTA) for a hypothetical change of the good or service. This is a flexible approach, which produces informative and straightforward answers to analyze individuals’ preferences in monetary
28
terms, with downsides of possible strategic behavior and missing scope-sensitivity. CE instead presents a set of attributes trying to capture the good with different levels, thus building on the assumption that consumers get satisfaction from the attributes, and not the good itself (Lancaster, 1966). This approach is more informative if the right attributes and levels are chosen, but the importance of survey design and chosen attributes limits the possibility to compare the results from different surveys. In this study, a CVM survey will be used for assigning an economic value to the benefit, in terms of time saved, of not having to transport sorted waste since Bruvoll and Nyborg´s model is better applicable for CVM. Four additional reasons for using CVM instead of CE are:
• It is important to make the survey as short as possible in order to increase the participation rate (Japec et al., 1997);
• It lowers the risk of unreliable results due to many attributes or scenarios (Hanley et al., 2001);
• The ability to measure any multi-dimensional aspects of the opportunity cost of is not the key purpose in this thesis; and
• The CVM approach facilitates the comparison with other studies (Hanley et al., 2001).
The use of CVM has been debated for a long time, not least since it may introduce some hypothetical bias when asking the respondents to reveal a maximum willingness-to-pay for a good or service not having to pay for it (Aadland and Caplan, 2003). Deviations from the true WTP can arise from moral conflicts, response bias, protest answers, and/or the cognitive burden relating to a hypothetical scenario (Diamond and Hausman, 1994). A desire to influence policy decisions (strategic behavior) may also be a potential deviation from true WTP, but is not assumed to be a risk in this scenario since time does not represent a public good.
Open-ended CVM (OE-CVM) has been criticized for the possibility of free-riding where respondents understate their true WTP (Perman et al., 2011), and for the difficult task of putting a number on a hypothetical question (Bateman et al., 2002). These disadvantages are, it is argued, avoided if the respondents are given dichotomous choices with different values to choose among, providing a smaller cognitive burden for respondents stating WTP. The use of this approach increased substantially after the so-called NOAA panel
29
recommended it in the mid-1990s (Arrow et al., 1993). However, it can lead to starting point bias (McFadden, 1994), and it does not provide WTP surplus, something which makes it a poor choice for CBA purposes (Bateman and Willis, 1999). Empirical evidence revealed the results from dichotomous choice surveys to be larger than OE-CVM surveys, possibly due to so-called yeah saying (Sund, 2010), and the difficulties in designing the price range given to respondents (Mitchell and Carson, 1989). In this thesis, the OE-CVM is used instead of dichotomous choice for four main reasons:
• The OE-CVM approach measures consumer surplus (Bateman and Willis, 1999); • The results can be used for CBA purposes (ibid);
• It prevents the problem of starting point bias (McFadden, 1994); and
• Free-riding and strategic bias is primarily an issue for the direct valuation of public goods.
Still, CVM results can only be seen as guidance; previous studies have found a difference between stated WTP and the true payment. Several respondents can be uncertain about their stated amount and some may be willing to pay for the service but does not know how much. How the CVM question is formulated can have an impact on the result, the concepts of choosing WTP or WTA can be determined with two questions (Perman et al., 2011):
• Is the outcome of the scenario an improvement or a deterioration from the status-quo?
• Is the main purpose to elicit the welfare measures compensating surplus (CS) or equivalent surplus (ES)?
The reviewed literature in Chapter 3 all used WTP as a measurement of estimating the responsibility relief following a change in recycling the scheme. For the present study, the respondents are asked to state how much they value more leisure time, which is assumed to gain utility. Therefore, WTP will be used as measure of the improvement of welfare following the change in the household recycling scheme (see further below).
4.1.3 Survey outline
During the developing phase, the methods and questions used in Berglund (2006), Bruvoll and Nyborg (2004) and Berglund and Söderholm (2009) strongly influenced the final
30
design of the postal survey. In order to make sure it was easy to understand and minimize misinterpretations, the survey was first tested on a small sample of respondents in line with the recommendations of Bateman et al. (2002). These pre-study efforts resulted in some changed questions in order to make sure the content was not misleading. Most of the changes were semantically oriented. An English and a Swedish version of the full survey is provided in Appendix 1 and 2, respectively.
The survey was divided into three parts as recommended by Mitchell and Carson (1989), where the first section was intended to be used as warm-up attitude questions in order to involve the respondents before presenting the WTP-scenario (Bateman et al., 2002). Specifically, the three following sections:
• Statements concerning recycling motivation and norms;
• Introducing a hypothetical waste management scenario, and asking an OE-CVM question; and
• Socio-economic questions (age, gender, income etc.).
The first part has the motive to capture the respondents’ attitudes towards recycling and the motivational factors behind their decisions. Nine motivational statements regarding recycling behavior in a five-degree scale to which the respondents agreed to these statements. “Disagree” was assigned a one (1), “Partly disagree” two (2), “Uncertain” three (3), “Partly agree” four (4) and “Fully agree” five (5). Some of these statements will be used as indicators in order to estimate the importance of social and moral norms as well as warm-glow motives, in turn influencing the perceived opportunity cost of time for recycling activities. An example of motivational statement is: “I want to consider myself as a responsible person”, which indicates how an internalized norm can affect recycling behavior because of striving towards fairness. The specific norms tested in the survey are introduced below.
All statements can be seen in Appendix 1, where some behavioral statements such as “In my household all packaging and newsprints are recycled” also are included in order to map the respondents´ attitudes and recycling efforts. The statements are followed by a question asking the respondent to give an estimation of the time spent (per week) on transporting and dropping-off their packaging waste.
