Consumer behavior in the sustainable clothing industry PDF Free Download
1 / 57/57
100%
RADBOUD UNIVERSITY
Nijmegen School of Management
Master Thesis
Consumer behavior in the sustainable clothing industry
What factors are influencing to what extent Dutch consumers are willing to pay more for
sustainable clothes?
Several problems do exist in the clothing industry. Problems with the environment, for people who
work in the clothing industry, and people who live around clothing factories. Consumers are mainly
responsible for the demand side of this problem. More research is needed on consumer behavior in the
sustainable clothing industry to help solve this problem. This research contributes to this, more
specifically to the willingness to pay by consumers for sustainable clothes. A questionnaire is used to
examine this willingness to pay and which factors are influencing this willingness to pay to what
extent. The five independent variables which are analyzed using a regression are: gender, generation,
education level, information level, and knowing people who are concerned with sustainability. It is
found that older generations, women and people who know more people who are concerned with
sustainability have a higher willingness to pay. The other independent variables showed no significant
effect on willingness to pay. Because the willingness to pay by consumers knowing more people who
are concerned with sustainability is higher, clothing brands could use influencers to sell their
sustainable clothes.
Supervisor: Anita Kopányi-Peuker
Maureen Haverkamp (s4630106)
Master: Financial Economics
2
Table of contents
1. Introduction ..................................................................................................................................... 3
2. Literature overview ......................................................................................................................... 5
2.1 Sustainable investments ................................................................................................................ 6
2.2 Willingness to pay for sustainable clothes .................................................................................... 7
2.2.1 Age ......................................................................................................................................... 8
2.2.2 Gender .................................................................................................................................... 9
2.2.3 Education ................................................................................................................................ 9
2.2.4 Knowledge sustainability ..................................................................................................... 10
2.3 Research question ........................................................................................................................ 11
3. Data and methodology ................................................................................................................... 11
3.1 Formulating hypotheses .............................................................................................................. 11
3.2 Survey design .............................................................................................................................. 13
3.3 Population and sample ................................................................................................................. 16
3.4 Analyzing data ............................................................................................................................. 17
4. Results ........................................................................................................................................... 18
4.1 Introduction data .......................................................................................................................... 18
4.2 Summarizing data on WTP ......................................................................................................... 19
4.3 Correlation analysis ..................................................................................................................... 22
4.3 Regression Analysis .................................................................................................................... 24
4.3.1 Hypothesis testing on generation .......................................................................................... 25
4.3.2 Hypothesis testing on gender ................................................................................................ 26
4.3.3 Hypothesis testing on education ........................................................................................... 27
4.3.4 Hypothesis testing on information ........................................................................................ 27
4.3.5 Hypothesis testing on knowing people ................................................................................. 28
4.3.6 Other results ......................................................................................................................... 28
4.3.7 Conclusion regression analysis ............................................................................................. 31
5. Discussion ..................................................................................................................................... 32
5.1 Interpretations .............................................................................................................................. 32
5.2 Limitations................................................................................................................................... 35
5.3 Further research ........................................................................................................................... 36
5.4 Recommendations ....................................................................................................................... 38
6. Conclusion ..................................................................................................................................... 39
7. References ..................................................................................................................................... 41
Appendix 1 - Survey .............................................................................................................................. 47
Appendix 2 - Assumption tests ............................................................................................................. 52
Appendix 3 - Linear regressions............................................................................................................ 53
Appendix 4 - Regression on statement willingness to pay .................................................................... 57
3
1. Introduction
The clothing industry is one of the most polluting industries of the world. Many
environmental issues are involved in the clothing industry. Issues which have a big impact are
that the textile industry uses large quantities of water, energy and chemicals, as well as
generating waste, pollution, and effluents. Besides the negative effects on the environment,
the clothing industry also has a negative impact on our society. The reason for this is that it
affects the people who work in this industry, since most of our clothes are produced in
countries where working conditions are poor (Veldhoven, 2020). Several examples of poor
working conditions are now given. China accounts for 30% of the world’s apparel export,
while here the working conditions often are poor. Some Chinese workers even make just 12-
18 cents per hour (Claudio, 2007). Possibly the worst example is that a large part of the 170
million children worldwide engaged in child labor are found in the clothing industry
(Grondelle, 2018). Mainly due to bad regulation in the clothing industry, the Rana Plaza
building in Bangladesh collapsed in April 2013, killing 1129 people working in this clothing
factory (Huq et al, 2016). Several chemicals are used in producing clothes. This also causes
health issues for the workers in the clothing industry. Many chemicals that are used in textile
production are found to cause many diseases like cancer, kidney, lung and liver issues, and
neurotoxicity in some people (Wittcoff et al., 2012). Other examples are that workers often
have 80 hours workweeks, while many of them do not get paid for doing overtime, workers
suffering from physical abuse, especially women, and salaries often half of the minimum
wage, if even some sort of minimum wage exists in that country (Grondelle, 2018). The
industry does not only affect the people who work in this industry, but also people in poor
countries who live around factories of this industry. For example the 5 million people who
live around the Citarum River in Indonesia and rely on it for drinking and bathing. Along the
banks of this river, 200 textile factories are established. These factories release dyes and other
chemicals into the water of the river, which causes the river to change color and devastating
the local ecosystem (MAKE.GOOD, 2020).
It is clear that the clothing industry is a very polluting industry, therefore the industry has to
change. A positive change would be for the industry to become more circular. This is a radical
rethink of the relationships between markets, customers, and natural sources (Lacy, 2016).
For a circular clothing industry, rigorous changes need to be made. Non-renewable resources
must be replaced by renewable, regenerative inputs. The main principles of a circular clothing
4
industry are based on a circular economy and sustainable development. Circular clothes
essentially have four typical features: Reuse textiles, repair or upgrade textiles, collect and
remanufacture textiles, and all textiles are recyclable (Veldhoven, 2020). This does not only
need to happen to achieve a less polluting industry. It also represents a huge opportunity for
companies to create a competitive advantage, by producing through digital technologies,
innovative business models and engineering (Lacy, 2016). After China, Europe is the biggest
exporter of textiles. Therefore the European Union considers the clothing industry as one of
the seven priority sectors for which its waste has to be reduced by 2030 with 50%. In the
Netherlands the goal is to have a circular economy by 2050. “The government has launched a
policy programme which aims to use at least 30% recycled material in new clothing by 2030
and have halved the environmental footprint of the textile sector by 2035” (Veldhoven, 2020,
p. 20). This means that besides the opportunity for clothing companies to create a competitive
advantage, another reason for them to become (partly) circular is that more laws will exist in
the future.
Consumers are mainly responsible for the demand side of sustainability in the clothing
industry. Consumers have become increasingly detached from textile and clothing production
contexts, therefore a greater effort is required to inform them about the sustainable impact of
their consumption practices (Boström, 2016). However, sustainability has recently become an
important new driver in consumers’ purchasing decisions (Gazzola, 2020). This means a
tension exists between consumers being detached from textile and clothing and stating they
started to care about sustainability. It might not be clear due to this what their actions are in
the (sustainable) clothing industry. Since consumers are mainly responsible for the demand
side of sustainability in the clothing industry, it is important to have more information about
consumers’ behavior in the sustainable clothing industry. It is critical that consumers have the
intention to pay more for sustainable clothes, since researchers found that sustainable products
are usually priced higher, which can be a barrier to consumption (Brucculieri, 2018).
This study aims to make important contributions to the literature of consumer behavior in the
(sustainable) clothing industry. The main focus of consumer behavior is the willingness to pay
for sustainable clothes. Several studies showed that consumers are willing to pay more for
sustainable clothes (e.g. Shahbandeh, 2018; Ciasullo, 2017; Nosto, 2019). This will further be
discussed in the literature overview. From the literature overview a research question has been
formulated: “What factors are influencing to what extent Dutch consumers are willing to pay
5
more for sustainable clothes?” Some hypotheses have been formulated to help answer this
research question. The hypotheses have been investigated with the help of a survey, where the
respondents are Dutch consumers. In 2011 the CBS already showed that Dutch consumers are
buying more sustainable products. Mainly clothes, electronics and cars (CBS, 2011). However
not much data exists on how much Dutch consumers are willing to pay for sustainable
clothes. Therefore it is important to do research on the consumer behavior of Dutch
consumers. The Paris Agreement is the first universal legal global climate change agreement,
adopted in December 2015. The Dutch government also agreed on this Paris agreement,
which means that they have to support climate action to reduce emissions (European
Commission, 2021). Sustainable clothes will reduce emissions and therefore is important for
the Netherlands. As discussed, the Dutch government has launched a policy program for the
clothing industry to become more sustainable. This is another reason that it is important to
know more about the willingness to pay for sustainable clothes by Dutch consumers.
The results of this research are that three examined independent variables are influencing the
willingness to pay for sustainable clothes. These variables are gender, generation and
knowing people who are concerned with sustainability. On average a Dutch consumer is
willing to pay 43% more for sustainable clothes than for non-sustainable clothes. More
research is needed to find more variables influencing the willingness to pay.
The structure of this research is as follows: first a literature overview is done to have more
knowledge about literature on willingness to pay for sustainable clothes. Second, the
methodology of this research is discussed, in which also the hypotheses are formulated. Third,
the results of the collected data are discussed. Fourth, a discussion about the research is done,
in which also the interpretations of the results are made. Last, a conclusion is made on the
research.
2. Literature overview
This chapter aims to connect different strands of literature, about the sustainable clothing
industry, and also to contribute to the behavior finance literature regarding the sustainable
clothing industry.
6
2.1 Sustainable investments
In Europe, sustainable investments taking into account social, ecological, and/or ethical
concerns play an increasing role in financial markets. There has been invested approximately
11 trillion euro in sustainable investments in Europe in 2016 (Blankenberg, 2018). Research
carried out by Gutsche and Ziegler (2019) indicated that a traditional finance theory states
that sustainable investments would only be made if they are in terms of risk and return at least
as attractive as other investments. However, Gutsche and Ziegler found out that in fact some
other studies show that other non-financial motives are relevant for sustainable investments.
In their own research Gutsche and Ziegler found some non-financial motives why investors
have a higher mean willingness to sacrifice returns for sustainable investments. These motives
are: affinity to left-wing parties, high feeling of warm glow from sustainable investments, and
a strong environmental awareness (Gutsche, 2019). Their results are based on a survey among
German private financial decision makers. Buying sustainable clothes is also a kind of
investment. They most often have a longer lifetime which results in buying less clothes. This
is the case with sustainable clothes, because extending clothing life has been found to be the
single largest opportunity to reduce water, carbon, and waste footprints (Cooper, 2013).
Therefore it is important to know whether the traditional finance theory of only making
sustainable investments if they are financially as attractive as other investments holds for
buying sustainable clothes, or that in line with sustainable investments other non-financial
motives are found.
Next to these non-financial motives in investment decisions, in behavioral finance, multiple
biases which also affect people to make investment decisions do exist. One of these biases is
the herd mentality bias. According to this bias, investors tend to copy what other investors are
doing (CFI, 2021). Research carried out by Chang and Lin (2015) indicated that investors
who tend to herd more are investors who are excessively optimistic and overconfident, live in
countries with high masculinity indexes, and show low uncertainty avoidance. Research
carried out by Park and Sabourian (2009) showed that herding only occurs when the number
of sales during an episode of buy herding is not too large, and when the information of
someone is sufficiently dispersed in a way that they consider extreme outcomes more likely
than moderate outcomes. They also found that this herding behavior in a standard sequential
security trading model does lead to more volatile prices, and they lower liquidity (Park et al,
2009). Another research by Filiz et al (2019) found that the mood of investors influences the
7
occurrence of herding behavior, where the tendency to show herding behavior is stronger in a
neutral mood than in a positive or negative mood. Multiple researchers did demonstrate that
the herding bias does occur in financial markets, as discussed even some factors are found
that affect investors to show herding behavior. Nowadays, investment platforms even make it
possible to copy what other investors are doing exactly. This copy trading allows investors to
receive the information of successes of other investors/agents in financial markets and even to
directly copy their portfolios (Apestequia et al, 2020). It is important to know whether this
bias also holds for buying sustainable clothes. It has been found that millennials prefer
receiving advice through social media influencers than from established institutions, therefore
often new products are introduced by influencers (Schoenmueller, 2021). This is another sign
that people (investors) tend to copy other people (investors). If this bias does hold for buying
sustainable clothes, then brands selling sustainable clothes could look more into introducing
their sustainable clothes through influencers.
2.2 Willingness to pay for sustainable clothes
Clothing brands are under pressure to address the growing demand for the industry to face the
sustainability agenda head-on. The clothing brands that will be successful will be the ones that
make adaptations early (McKinsey, 2020). For the clothing brands who want to adapt, it
would be good to know more about the willingness to pay for sustainable clothes by
consumers, such that they can make financial indications. However, with the political
instability complicating the perspective for the global fashion industry in 2020, the global
economy is under pressure (McKinsey, 2020).
Multiple researchers found that a relatively large part of the consumers are willing to pay
more for sustainable clothes. Research carried out by Shahbandeh indicated how much more
consumers living in Hong Kong, Shanghai, London, New York, and Tokyo were willing to
pay in 2018 for sustainable clothes than for normal clothes. The survey presented that all
respondents wanted to pay more for sustainable clothes (Shahbandeh, 2020). Most consumers
are willing to pay more, however how much they relatively want to pay differs for the
consumers. Research carried out by Nosto in 2019 indicated that although most consumers
prefer sustainable clothes over non-sustainable clothes, only 29 per cent of consumers are
willing to pay more for a sustainable-made version of the same item. Their respondents were
8
50% from the United States and 50% from the United Kingdom (Nosto, 2019). This outcome
is very different from the outcome of Shahbandeh, where 100% of the respondents wanted to
pay more for sustainable clothes. Research carried out by Ciasullo et al (2017) also indicated
the willingness to pay a premium price in 2016 for a sustainable fashion product. Their
respondents were a sample of students at the University of Salerno. In their survey 4% of the
respondents wanted to pay 0% premium for a sustainable fashion product, and the other
respondents wanted to pay a certain premium above zero for a sustainable fashion product
(Ciasullo et al, 2017). Research carried out by Aryal et al (2009) indicated that all their
respondents were willing to pay a price premium for organic products. Their respondents
were consumers buying organic products in Kathmandu Valley (Nepal) (Aryal et al., 2009).
Research carried out by Lin indicated that more than half of the respondents want to pay more
for sustainable clothes based on health and environmental issues. The respondents of this
research were Hawaii consumers (Lin, 2009). A research on Dutch consumers indicated that
49% of younger people are willing to pay more for sustainable produced clothes (Kamphuis et
al, 2020).
The findings of the five researches are different, however a simple explanation for this might
be that their respondents also differ. What the studies have in common is that most consumers
regardless of the place they live are willing to pay more for sustainable clothes. Besides this,
some factors are influencing the willingness to pay for sustainable clothes by consumers. The
aim of this research is to examine what factors influence to what extent Dutch consumers are
willing to pay for sustainable clothes. This is done to achieve more insight in Dutch
consumers’ behavior regarding sustainable clothes. In the next sections, literature on factors
influencing willingness to pay are discussed. The factors which are discussed are age, gender,
education level and knowledge about sustainability.
2.2.1 Age
Research has been done on the influence of age on consumers’ willingness to pay for
sustainable clothes. The younger generations are paying growing attention to sustainability
(Gazzola, 2020). Young consumers perceive sustainability as important in the clothing
industry, they recognize their own ability to make a difference with their consumption choices
(Han, 2016). The survey of McKinsey (2020) also showed that younger generations are
9
willing to pay more for sustainable clothes. It is a survey in the United States conducted from
consumers in 2019. Their results show that 12% of the Boomer generation (1646-1964) would
pay more for sustainable clothes, 17% of Generation X (1965-1981), 26% of the Millennial
generation (1982-1995), and 31% of Generation Z (1996-2019). This means that younger
generations are willing to pay more for sustainable clothes. However, contradictory research
does exist. Notaro and Paletto (2021) found that young people are more concerned about the
environmental impact and price when making a purchase decision, whereas older people are
mostly concerned about human health and not so much about price. Older people are more
willing to buy eco-friendly products, especially clothes and furniture, and young people place
greater trust in environmental brands (Notaro and Paletto, 2021). These results would mean
that older people are willing to pay more, however more evidence exists for the direction of
younger people willing to pay more. Namely, Gazzola (2020) and Han (2016) found that
younger generations are paying growing attention to sustainability and recognize their ability
to make a difference, and McKinsey (2020) found that younger generations are willing to pay
more for sustainable clothes.
2.2.2 Gender
Another factor for which research has been done is the influence of gender. Multiple studies
have shown that women’s strong sense and high sensitivity of female altruism can influence
the gap of sustainable choices between men and women (Gazzola, 2020). Another study has
shown that men tend to avoid sustainable choices, because buying sustainable products is
judged as feminine (Brough, 2016). However, it has been found that women make sustainable
choices with buying clothes due to the negative impact on health and animals, while men
make sustainable choices due to the negative impact on water quality, forests, and air quality
(Notaro and Paletto, 2021). If the respondents of the survey, so Dutch consumers, see the
negative impact on water quality, forests, and air quality as most important, it might be that
men are willing to pay more.
2.2.3 Education
Not much research about the influence of education on the willingness to pay by consumers
exists yet. It currently has not been proven by researchers that people who are higher educated
10
are willing to pay more. It is even found that education is positively correlated with
unsustainable living (United Nations, 2012). Besides this, in a similar study about
consumers’ behavior, regarding sustainable beer instead of sustainable clothes, it has been
found that younger people and people with lower levels of education are willing to pay more
for sustainable beer (Carley, 2018). This finding indicates that it might be the same with
sustainable clothes. Education is highly correlated with income (Tolley, 1971), therefore it is
also important to look at the findings on income. Some previous researchers found that richer
people spend more on sustainable goods (e.g. Taylor, 2014; Pedrini, 2014). However, not all
studies found the same results. The study of Kendall Cox Park did not find enough evidence
for their hypothesis that wealthier people are willing to pay more for ethical goods (Kendall
Cox Park, 2018). So this means that with research on education and income, conflicting
outcomes do exist.
2.2.4 Knowledge sustainability
As found in numerous researches, environmental knowledge is an important factor in
predicting environmentally sustainable behavior (Connell, 2014, p. 45). This indicates that it
is an important factor for this research to look at, because it does affect sustainable behavior.
Education for sustainable development does help to motivate people to take actions that
promote sustainable development. This is the case because it causes more awareness about the
complex and dynamic issues and helps people understand sustainable development better
(United Nations, 2012). This indicates that if consumers are better educated about the new
sustainable developments in the clothing industry, and are more aware about the complex and
dynamic issues in the clothing industry, they will take action to promote the sustainable
developments. Which most likely will also lead them to be willing to pay more for sustainable
clothes. Environmental knowledge is an important driver to engage consumers to consume
clothes in an environmentally sustainable way (Connell, 2014). However, besides the findings
that environmental knowledge and sustainable consumption are positively correlated,
literature observing problems with this also exist. Some research, like Jacobs et al (2018) and
Wiederhold and Martinez (2018), showed that a gap between attitude and behavior has been
observed. This means that environmental concerns do not always lead to environmental
behavior (Dhir, 2021).
11
2.3 Research question
It has been proven by some researchers that willingness to pay for sustainable clothes is
higher than the willingness to pay for non-sustainable clothes. Some literature has been
discussed on factors influencing the willingness to pay positively or negatively. Not much
research has been done on this already and the outcomes of the researches are not always in
line with each other. Therefore this literature overview led to the formulation of the following
research question;
“RQ: What factors are influencing to what extent Dutch consumers are willing
to pay more for sustainable clothes?”
3. Data and methodology
The literature overview gave more insight into the factors influencing the willingness to pay
for sustainable clothes by consumers. In this chapter, first the hypotheses for this research are
formulated. Second, how the data for this research is collected is discussed. Third, the
population and sample of this research is discussed. Last, it is explained how this collected
data will be analyzed and how the hypotheses will be tested.
3.1 Formulating hypotheses
In the literature review, some factors which might affect the willingness to pay for sustainable
clothes by Dutch consumers have been discussed. From this literature the next hypotheses are
formulated:
Hypothesis 1: The younger generations are willing to pay more for sustainable clothes
This hypothesis for age is formulated based on the papers from Gazzola (2020), Han (2016),
McKinsey (2020), and Notaro and Paletto (2021). These papers mainly show evidence for
younger generations to be willing to pay more. Even though there are mixed results, the
results of the younger generations willing to pay more are more common than the other
directions,
12
Hypothesis 2: Women are willing to pay more for sustainable clothes
All papers on gender show results in this direction (Gazzola, 2020; Brough, 2016; and Notaro
and Paletto, 2021). Therefore this hypothesis is expected to be correct.
Hypothesis 3: Higher educated people are not willing to pay more for sustainable
clothes
Currently it has not been proven by researchers that higher educated people are willing to pay
more for sustainable clothes. Some studies exist which show it is most likely that the lower
educated people are willing to pay more (United Nations, 2021; Carley, 2018). Income and
education are highly correlated with each other (Tolley, 1971). Most higher educated people
have a higher income and some previous researchers found that richer people spend more on
ethical goods (e.g. Taylor, 2014; Pedrini, 2014). This could mean that higher educated people
are willing to pay more for sustainable clothes. However the study by Kendall Cox Park did
not find enough evidence for their hypothesis that wealthier people are willing to pay more for
ethical goods (Kendall Cox Park, 2018). Therefore this hypothesis is formulated. Since the
two variables income and education are interdependent, a separate hypothesis for income has
not been formulated.
Hypothesis 4: People who are better informed about the problems in the clothing
industry are willing to pay more for sustainable clothes
This hypothesis on better being informed about the problems in the clothing industry is based
on studies from Connell (2014) and United Nations (2012). The researches both showed that
education for sustainable development, and thus environmental knowledge, causes people to
be motivated to take actions that promote sustainable development. In this research this would
mean that with more knowledge about the environment in the clothing industry, people would
be motivated to buy sustainable clothes to stimulate sustainable development in the clothing
industry.
Hypothesis 5: People who know more people who are concerned with sustainability
are willing to pay more for sustainable clothes
As discussed in the literature overview, it is important to know whether the herd mentality
bias holds with buying sustainable clothes. According to the herd mentality bias, investors
tend to copy other investors (CFI, 2021). Therefore it is expected that people who know a lot
13
of people who are concerned with sustainability, are also more concerned with sustainability
and therefore are willing to pay more for sustainable clothes. They copy the behavior of other
people by taking actions (buying sustainable clothes) that indicate being concerned with
sustainability. Therefore this hypothesis is formulated and might be a proxy for the herding
bias. Besides this, this hypothesis could capture other motives as well. For example that
people tend to have friends with the same interests or that it might be affected by a certain
upbringing of people. The motive of the herding bias is most important for this research.
3.2 Survey design
The dataset of this research is gathered through a survey. The type of survey used is a
questionnaire, where a list of questions is distributed online, using voluntary response
sampling. The language of the survey is Dutch, since the research is about Dutch consumers.
The survey is designed with the use of Qualtrics. This is an online research and data collection
tool for which Radboud University has a subscription. The survey of this research can be
found in Appendix 1. This section gives insight into how each variable of the hypotheses is
measured.
The dependent variable of this research is willingness to pay. In the survey the respondents
were asked to keep a specific type of unsustainable clothing piece of their own closet in their
mind. A given price for this piece of clothes was notated and the respondents had to say for
themselves how much they would pay for this specific clothing piece when it would be
produced sustainable. The relative difference in price between the non-sustainable and
sustainable clothing piece is used as the willingness to pay for this research. Different clothing
pieces (jeans, T-shirt, socks) are used to collect more willingness to pay information per person.
However, an anchoring effect could be a problem with this method. This means that the specific
price notated might affect the willingness to pay by the consumers. Simonson and Drolet (2003)
gave an example of this problem: if a consumer is asked to consider a price of $20 for a food
processor, which is a very low price for a food processor, then the consumer will be likely to
be willing to pay that price. If this consumer is asked how much his or her maximum willingness
to pay is for that food processor, then the willingness to pay is expected to be influenced by the
provided anchor price of $20 and thus will be low. If this anchor price would have been $80, it
14
is expected that the willingness to pay is higher for the same consumer (Simonson, 2003).
Therefore the three different clothing pieces have two different specific notated prices. The
three different clothing pieces have a low anchoring price and a high anchoring price. The prices
of the shirts are €10 and €30. The prices of the jeans are €30 and €75. The prices of the socks
are €7,50 and €15. The relative difference is prices per clothing piece varies from 2-3. The six
items with the different prices are randomized in the survey. Respondents saw all six items and
not only the high or low anchoring price items. This is done because when you make two
samples, one for the low anchoring prices and one for the high anchoring prices, most likely
more respondents are needed. Besides this, it is also interesting for this research to know the
difference between willingness to pay with low and high anchoring prices per person.
Another problem that could occur with examining willingness to pay is the experimenter
demand effect. This effect occurs when respondents change their behavior due to what they
expect to be appropriate behavior (Zizzo, 2010). The respondents might say they pay more for
the sustainable clothes, because they think this is more appropriate behavior. To partly solve
this problem, a method has been used which is often used by research on willingness to pay
with sustainable products, namely the dissonance minimizing referendum
1
. This method uses
an extra question, which causes the respondents to be able to express their support for a specific
program without having to state that they are willing to pay more for this program (Uehleke,
2016). In this research the respondents were asked to which statement they agree more: I
support sustainable clothing and am willing to pay more for it; I support sustainable clothing,
but I am not willing to pay more for it; and I do not support sustainable clothing regardless of
the price. By offering the respondents to agree with the second statement, they can choose to
support sustainable clothing and thus show appropriate behavior, but still not pay more.
Besides this independent variable, also independent variables are measured. How they are
measured is now discussed.
1
This problem is only solved partly, therefore it is a limitation of this research. This is further discussed in 5.2
Limitations.
15
- Younger generations:
In the survey it was asked in which generation the respondents belong. Four generations are
used: Boomer (1946-1964), Generation X (1965-1981), Millennial (1982-1995), and
Generation Z (1996-2019).
- Women:
Respondents were asked to state their gender from the following options: man, woman, and
other.
- Higher educated:
Respondents had to state their highest education level they have completed from the following
education levels: only elementary school, only secondary school, Vocational education and
training, Bachelor’s degree, and Master’s degree or higher.
- Better informed:
Respondents had to indicate for themselves how well they are informed about the problems in
the clothing industry. They could choose between 1-5 for their level of information, where 5 is
highly informed and 1 is not informed. There has been chosen for 1-5 because then the
respondents could choose to have average information when they choose 3.
- Knowing more people who are concerned with sustainability:
Respondents had to indicate for themselves how many people they approximately know who
are concerned with sustainability. They could choose in a range of 1-5. Where 1 means that
they do not know people who are concerned with sustainability, with 3 they know
approximately as many people who are concerned with sustainability as people who are not
concerned with sustainability, and 5 means that all people they know are concerned with
sustainability. Being concerned with sustainability is very broad and therefore examples of
being concerned with sustainability were given with this question in the survey. Some examples
are: buying sustainable clothes, eating less meat, voting for a political party because they care
about sustainability, flying less, trying to use less energy, and prefer cycling over driving with
a car.
- Annually bought clothes
Respondents were asked to indicate how many clothes they buy annually. This is asked with an
open question.
16
Besides these variables, a question has been used in the questionnaire to make clear what
sustainable clothes means in this research:
- “During this survey, a number of questions involve the topic of sustainably produced
clothing. It is therefore important to know what is meant by this in this survey.
Aspects that fall within sustainably produced clothing are: fair wages, high quality,
social responsibility, less harmful for the environment and sustainably selected raw
materials such as organic cotton (Pookulangara and Shephard, 2013). When talking
about sustainably produced clothing, these aspects are met. Is this clear to you?”
Respondents were asked to answer this question with yes or no.
3.3 Population and sample
The population of the Netherlands born since 1946 was selected as the survey population.
There has not been filtered on that, since the survey was spread online. Only one respondent
stated to be born 2 years earlier. This respondent stated to be born between 1946-1964,
nothing has done with this, since it most likely does not affect the outcome. To stimulate
response, the survey was made as short as possible, without missing important information.
The average duration of filling in this survey was 3.5 minutes. Completed questionnaires were
received from 414 respondents. Some of these 414 responses were dropped
2
. Finally, 404
valid questionnaires were available for the data analysis. Table 1 summarizes the
demographic characteristics of the respondents.
Characteristic
Respondents
Percentage
Gender
Male
177
43.8%
Female
226
55.9%
Other
1
0%
Generation
1946-1964 (Boomer)
123
30.4%
1965-1981 (Generation X)
148
36.6%
1982-1995 (Millennial)
51
12.6%
1996-2019 (Generation Z)
82
20.3%
Education
Only elementary school
1
0%
2
Why they were dropped is explained in detail in 4.1 Introduction data.
17
Only secondary school
51
12.6%
Vocational education and training
123
30.4%
Bachelor’s degree
143
35.4%
Master’s degree or higher
86
21.3%
Table 1: Demographic characteristics of the respondents.
3.4 Analyzing data
The data achieved from the survey is analyzed with Stata. A multiple linear regression is used
to estimate the relation between the dependent variable (willingness to pay for sustainable
clothes) and the independent variables (age, gender, education level, information level of
problems in clothing industry, and knowing people who are concerned with sustainability).
The multiple linear regression formula in this case will be:
• = willingness to pay
• = age
• = gender
• = education level
• = information level of problems in clothing industry
• = knowing people who are concerned with sustainability
• = model error
Stata is used to do a regression analysis to calculate the effect that the independent variables
have on the dependent variable. The variables are tested on multicollinearity, correlations, and
heteroskedasticity. After these controls, the hypotheses are tested using the regression
analysis.
From the dependent variable willingness to pay, a sample split has been made. With doing so,
a broader analysis of the collected data can be made. The variable has been split in the
following samples:
- Willingness to pay
18
- Willingness to pay with a low anchoring price
- Willingness to pay with a high anchoring price
- Willingness to pay for shirts
- Willingness to pay for socks
- Willingness to pay for jeans
These samples all have been used for models. These models are also used to test the
hypotheses. Therefore it is also examined what kind of effect the independent variables have
on the other samples of willingness to pay. In the next chapter the results of this research are
discussed.
4. Results
In this chapter, the results of the data analyses are discussed. First, the data is introduced.
Second, the data on willingness to pay is summarized. Third, the data correlation analysis is
discussed. Fourth, the regression analysis is discussed, in which the hypotheses of the
research are tested. Last, the results are summarized.
4.1 Introduction data
As discussed in chapter 3, 414 respondents completed the questionnaires. A response was
compulsory for all the questions, which resulted in no missing values in the completed
questionnaires. However, some of the 414 respondents were dropped, because they were
outliers. An observation is considered an outlier when it is inconsistent with the majority of
the data. An outlier can dramatically change the values of both mean and variance of a
distribution (Salkind, 2010). For the dropped observation, this was also a problem
3
. When the
data analyses started, the sample size of this research was N=404. For the category gender,
only one respondent identified itself as different. Therefore this one was left out and the final
sample size of this research changed to N=403. The collected data has been analyzed using
Stata.
3
Two reasons did exist to drop some observations. Most of the observations were dropped, because the
questions on willingness to pay were open questions. Respondents did not state an amount they are willing to
pay, but answers like “I do not know or I do not have a jeans of €20.” Another respondent was dropped because
the respondent stated to buy 5000 clothes per year, which influences the mean too much and the respondent
might not have taken the questionnaire seriously.
19
4.2 Summarizing data on WTP
The dependent variable of this research, willingness to pay, is most important. Therefore this
summary of the data starts with a summary of the most important data on willingness to pay.
See table 2 below for this summary.
Willingness to pay
Mean
43%
Minimum
-22.%
Maximum
232%
Mean shirt €30
29%
Mean shirt €10
68%
Mean socks €15
30%
Mean socks €7.50
55%
Mean jeans €75
20%
Mean jeans €30
53%
Willingness to pay
Amount of respondents
<0
3 out of 403
=0
24 out of 403
>0
376 out of 403
Table 2: Summary willingness to pay
As explained in chapter 3, an anchoring effect could occur. Therefore for each clothing item
two different prices were chosen, a high and a low price. From this table it can be concluded
that it is good that this has been done. The difference between the willingness to pay with low
and high prices is high. The willingness to pay is much higher for the lower prices. Whether a
significant difference between willingness to pay for low and high anchoring prices exists has
been tested with the Wilcoxon signed-rank test. The results of this test indicate that we can
reject the null hypothesis at a significance level of 0.01, where the null hypothesis is that both
distributions are the same. This means a significant difference does exist. The mean
willingness to pay is 43%, most people are willing to pay more for sustainable clothes. For a
shirt of €10, the mean willingness to pay is the highest and for jeans of €75, the mean
willingness to pay is the lowest. In this research, 93% of the respondents stated to be willing
20
to pay more for sustainable clothes. This is much higher than the 49% found by Kamphuis et
al (2020).
Besides the information in table 2, it is also important to look at the histogram of willingness
to pay, to have a first look on whether it has a normal distribution. With a sample split of
willingness to pay in: willingness to pay with low anchoring price, willingness to pay with
high anchoring price, willingness to pay for shirts, willingness to pay for socks, and
willingness to pay for jeans a broader analysis of the data can be done. Therefore it is also
important to look at the histograms of these samples of willingness to pay. See figure 1 below
for the histograms on the samples of willingness to pay.
Figure 1: Histograms of willingness to pay from left above to right under: willingness to pay, willingness to pay
with low anchoring price, willingness to pay with high anchoring price, willingness to pay for shirts, willingness
to pay for socks, and willingness to pay for jeans.
21
These histograms indicate that the variables are not normally distributed. This is tested for all
the variables with the Shapiro Wilk Test. See the outcome of this test in table 3 below.
Variable
Prob>z
Willingness to pay
0.00
Willingness to pay with low anchoring price
0.00
Willingness to pay with high anchoring price
0.00
Willingness to pay for shirts
0.00
Willingness to pay for socks
0.00
Willingness to pay for jeans
0.00
Table 3: Shapiro Wilk Test to test normal distribution willingness to pay samples.
From this table it can be concluded that all the willingness to pay variables are not distributed
normally. However, the assumption of normality is sufficient, but not necessary for the
validity of many hypotheses testing (Salkind, 2010). Therefore the regression analysis can
still be done with the variables.
This research aims to examine the influence of some variables on willingness to pay. It is
important to know how these variables influence willingness to pay. The variables which are
examined are gender, generation, education, information level, and knowing people who are
concerned with sustainability. See table 4 below for the influence of these variables on
willingness to pay for sustainable clothes.
Characteristic
Respondents
Mean WTP
Gender
Male
177
38.9%
Female
226
45.5%
Generation
1946-1964 (Boomer)
123
49.2%
1965-1981 (Generation X)
148
42.3%
1982-1995 (Millennial)
51
31.6%
1996-2019 (Generation Z)
82
40.2%
Education
Only elementary school
1
28.9%
Only secondary school
51
41.3%
22
Vocational education and
training
123
42.0%
Bachelor’s degree
143
39.3%
Master’s degree or higher
86
50.0%
Information
No information at all
7
25.1%
Little information
110
39.5%
Average information
205
40.3%
A lot of information
73
53.0%
All possible information
9
63.5%
Knowing
No people at all
12
21.6%
Few people
181
37.5%
Just as many people do or
do not
121
47.8%
Many people
87
48.6%
All the people
3
56.7%
Table 4: Mean willingness to pay for different variables.
This table gives a clear overview of the collected data of this research. It shows that females,
Boomers, people with Master’s degree or higher, people with all possible information, and
people who only know people who are concerned with sustainability have the highest average
willingness to pay. In the next sections it will be tested whether these observations are
significant.
4.3 Correlation analysis
In order to analyze the collected data, it should be tested whether the variables are correlated.
For this research the Spearman rank correlation test has been used. See table 5 for the
correlation matrix of the variables.
Variable
WTP
Gender
Generation
Education
Information
Knowing
people
WTP
1
Gender
0.1303*
1
Generation
-0.1016*
0.0287
1
23
Education
-0.0052
-0.0894
0.0929
1
Information
0.1254*
-0.0363
-0.0871
0.1049*
1
Knowing people
0.2133*
0.0689
0.0104
0.2510*
0.3517*
1
*Correlation is significant at the 0.05 level
Table 5: Correlation matrix.
Table 5 shows some significant correlations, however they are not extremely high. Significant
positive correlations were found between the dependent variable, willingness to pay, and three
independent variables. These three variables are; gender, information and knowing people.
This means in this research that women, people who have more information about the
problems in the clothing industry, and people who know more people who are concerned with
sustainability, have a higher willingness to pay. Also one significant negative correlation was
found, between generation and willingness to pay. In this research this means that the younger
the generation, the lower the willingness to pay. The implications of these correlations were
tested through regression analysis and are discussed in the next section.
Besides correlations between independent and dependent variables, also some independent
variables seem to be correlated. Education is positively correlated with both the level of
information someone has about the problems in the clothing industry and how many people
someone knows who are concerned with sustainability. The variable information level about
the problems in the clothing industry and knowing people who are concerned with
sustainability are also positively correlated with each other. These correlations are not
surprising. First, the correlation between knowing people and level of information is not
surprising, namely if someone has a lot of information about the problems in the clothing
industry and shares this with people around them, the chance of them being more concerned
with sustainability is higher. It also works the other way around, if you have more people
around you who are concerned with sustainability, they might share their information why
they are concerned with sustainability with you, which might also include their information
on the problems in the clothing industry. Second, the correlation between education and
knowing people and information level is not surprising, because as seen, the lower the
education level, the lower the willingness to pay. People who for example have the education
level ‘Vocational education and training’ most likely know a lot of people from this education
level, since they were in class together. Which means they know a lot of people who have a
lower willingness to pay than average. This lower willingness to pay may be caused by little
24
information about the problems in the clothing industry as well as knowing few people who
are concerned with sustainability.
4.3 Regression Analysis
To test the five hypotheses of this research, regression analyses were done. With this
regression analyses, the linear relation between the dependent variable and independent
variables are examined. In this section the tested hypotheses are discussed.
Before testing the five hypotheses, the assumptions of multicollinearity and heteroskedasticity
are checked. They are presented in Appendix 2. These tests showed that only one assumption
appeared to be a problem, which is heteroskedasticity. This problem can be solved by adding
the robust standard errors to the model (Astivia et al, 2019). Therefore this has been done for
this research.
A multiple linear regression was calculated to predict the effect of gender, generation,
education, information and knowing people on willingness to pay. Results of the linear
regression are represented in table 6, where one observation is the average willingness to pay
for one person. This model is statistically significant F (7, 395) = 5.32; p < .05 (p = .00). This
linear regression is also calculated for willingness to pay with a low anchoring price, with a
high anchoring price, for shirts, for socks, and for jeans. These models are also statistically
significant, see F-test in table 6. The results of these regressions are also represented in table
6. All linear regressions can be found in Appendix 3. Most independent variables are treated
as continuous, except for generation. As a control, the regression was done with all the
independent variables as factors. The outcome was that only the variable generation had to be
treated as a factor variable, since this affects the conclusions. Besides this, for education two
groups have been pooled together. The education levels ‘only elementary school’, ‘only
secondary school’, and ‘Vocational education and training’ are treated as low education
levels. The education levels ‘Bachelor’s degree’ and ‘Master’s degree or higher’ are treated as
high education levels. Since these levels are pooled together as two different groups, this
variable could also be treated as continuous. In the next sections the regressions are used to
test the hypotheses.
25
Variables
WTP
WTPlow
WTPhigh
Coef.
Stand.
Error
Coef.
Stand.
Error
Coef.
Stand.
Error
Female
.061*
.035
.124**
.051
-.003
.022
Generation X
-.061
.047
-.097
.070
-.025
.028
Millennials
-.150***
.052
-.201***
.074
-.098***
.034
Generation Z
-.091*
.048
-.086
0.73
-.096***
.029
Education
high
.003
.035
-.004
.053
.011
.022
Information
.044
.029
.071
.046
.018
.015
Knowing
.050**
.020
.067**
.030
.033***
.012
Num. of obs.
403
403
403
F (7, 395)
5.41
5.19
6.01
WTPshirt
WTPsocks
WTPjeans
Coef.
Stand.
Error
Coef.
Stand.
Error
Coef.
Stand.
Error
Female
.115**
.045
.030
.040
.038
.035
Generation X
-.029
.061
-.090
.056
-.064
.048
Millennials
-.095
.060
-.227***
.059
-.127**
.055
Generation Z
-.005
.059
-.182***
.055
-.087*
0.50
Education
high
.021
.045
-.020
.022
.009
.035
Information
.075
.050
.014
.028
.043
.027
Knowing
.051*
.028
.053**
.022
.046**
.018
Num. of obs.
403
403
403
F (7, 395)
3.54
5.86
4.14
***p<0.01, **p<0.05, *p<0.1
Table 6: Regressions on willingness to pay overall, with low anchoring price, with high anchoring price, on
shirts, on socks, and on jeans.
4.3.1 Hypothesis testing on generation
The hypothesis on generation is: The younger generations are willing to pay more for
sustainable clothes. The regression on willingness to pay showed a negative relation between
26
generation and willingness to pay. Where the generations are coded as follows: 1946-1964
(1)/ 1965-1981 (2)/ 1982-1995 (3)/ 1996-2019 (4). The first generation are the Boomers, the
second are Generation X, the third are Millennials, and the last one are Generation Z. The
results show that Generation X on average is willing to pay 6.1%-point less than Boomers,
however this is not significant (p=0.196). Millennials are willing to pay 15.0%-point less than
Boomers (p=0.004). Last, Generation Z is willing to pay 9.1%-point less than Boomers
(p=0.055). This is not in line with the hypothesis. Therefore this hypothesis should be
rejected.
The hypothesis is also tested with the help of the regressions of the other samples of
willingness to pay. With all the regressions this hypothesis should be rejected. All regressions
show a negative correlation, which is not in line with the hypothesis. Also some variables
show to be insignificant. See table 6 for the coefficients and their significance.
4.3.2 Hypothesis testing on gender
The hypothesis on gender is: Women are willing to pay more for sustainable clothes. From
the regression on willingness to pay it has been found that women on average are willing to
pay more for sustainable clothes. On average, they are willing to pay 6.1%-point more than
men. However, the variable is weakly significant (p=0.080). So there is some evidence that
women are willing to pay more than men, but it is not significant at the 95% confidence level.
Therefore this hypothesis can only be accepted at 90% confidence level.
The respondents of this research could choose between three answers when being asked for
their gender: man, woman and other. Only one of the respondents chose the answer ‘other’.
This observation was deleted, because you cannot take conclusions from one observation.
This observation does thus not affect the outcome of this hypothesis. Since only two groups
exist, men and women, with a Mann Whitney test, it can be tested whether there is a gender
difference in this sample without controlling for other characteristics. The results of this test
show that men and women are statistically different at a 0.10 significance level.
Testing the hypothesis with the other regressions, it can be accepted at 95% confidence level
for two samples of willingness to pay. Namely for willingness to pay with a low anchoring
27
price and for shirts. Where respectively women are willing to pay 12.4%-point and 11.5%-
point more than men. For the other three samples, willingness to pay with a high anchoring
price, for socks, and for jeans, the variable gender is not significant.
4.3.3 Hypothesis testing on education
The hypothesis on education is: Higher educated people are not willing to pay more for
sustainable clothes. The five different education levels are pooled into two different groups.
Where the education levels ‘only elementary school’, ‘only secondary school’, and
‘Vocational education and training’ are treated as low education levels and the education
levels ‘Bachelor’s degree’ and ‘Master’s degree or higher’ are treated as high education
levels. The regression on willingness to pay showed that education does not have an effect on
willingness to pay. The variable is by far not significant (p=0.925). Therefore this hypothesis
should be rejected.
Testing the hypothesis with the regressions of the other samples of willingness to pay, also
show that the hypothesis should be rejected. The variable is not significant in one of the
regressions.
4.3.4 Hypothesis testing on information
The hypothesis on information is: People who are better informed about the problems in the
clothing industry are willing to pay more for sustainable clothes. The regression on
willingness to pay showed some evidence that the level of information is positively correlated
with the willingness to pay. Where the level of information is coded as follows: no
information at all (1)/ little information (2)/ average information (3)/ a lot of information (4)/
all possible information (5). When the level of information increases with one, the willingness
to pay on average increases with 4.4%-point. So this means that people who are better
informed about the problems in the clothing industry are willing to pay more for sustainable
clothes. However, the variable is not significant (p=0.133). This means that although the
positive correlation is in line with what the hypothesis suggests, the hypothesis should be
rejected. If the amount of respondents would have been higher, then it could be that the
28
variable would be significant, since it is close to being significant with the amount of
respondents of this research.
With the other regressions, the variable information level is also insignificant. Therefore this
hypothesis should also be rejected using the regressions of the other samples of willingness to
pay.
4.3.5 Hypothesis testing on knowing people
The hypothesis on knowing people is: People who know more people who are concerned with
sustainability are willing to pay more for sustainable clothes. From the regression, it has been
found that there exists a positive relation between how many people someone knows who are
concerned with sustainability and the willingness to pay. How many people someone knows
is coded as follows: No people at all (1)/ few people (2)/ just as many people do or do not (3)/
many people (4)/ all the people (5). When the level of knowing people increases with one, on
average the willingness to pay increases with 5.0%-point. This effect is significant (p=0.012) .
Therefore this hypothesis can be accepted.
Using the other regressions, the hypothesis should also be accepted. These regressions are
willingness to pay with a low anchoring price, with a high anchoring price, for shirts, for
socks, and for jeans. Respectively when the level of knowing people increases with one, on
average the willingness to pay increases with 6.7%-point, 3.3%-point, 5.1%-point, 5.3%-
point, and 4.6%-point. The variable is significant at 99% confidence level for the model using
willingness to pay with a high anchoring price. For the models with willingness to pay with a
low anchoring price and for socks and jeans the variable is significant at 95% confidence
level. Last, with the model using willingness to pay for a shirt, the variable is significant at
90% confidence level.
4.3.6 Other results
Besides the results of the hypotheses, some other important results were found with the data
analyses. The respondents of the questionnaire were asked how many clothes they think they
buy for themselves per year. This was asked because overall consumption, so also
consumption of clothes, is the root cause of the current environmental crisis (Schanes et al,
2016). Research carried out by Maldini et al (2017) indicated with the use of a survey that
29
Dutch consumers on average buy 46 pieces of clothes per year. There is a remarkable
difference between this average and the average of the respondents of this research. The
average of this research is 20. See figure 2 below for the results of this research.
Figure 2: Number of purchased clothes per consumer per year.
Multiple possible explanations exist for the difference in the average of Dutch consumers and
the average of the respondents of this research. One of the possible explanations is that a
larger proportion of the respondents belong to the older generations. The average amount of
clothes purchased by the Boomers (1946-1964) is 15 clothes, by people of Generation X
(1965-1981) 21 clothes, by Millennials (1982-1995) 20.5 clothes, and by people of
Generation Z (1996-2019) 24.5 clothes. However, even the average of the youngest
generation is much lower than the average of the Dutch consumer. Another possible
explanation is that the respondents based their answer on their buying behavior during
COVID-19. This is different from before COVID-19, because consumers intend to shift their
spending to essentials, like grocery, and cut back on products like clothes (Arora et al, 2020).
Another explanation might be that consumers underestimate the amount of clothes they buy.
Besides these explanations where the consumers show their true behavior, the outcome also
might be affected by the experimenter demand effect. This means that the respondents
changed in behavior, because they thought buying less clothes is more appropriate behavior
(Zizzo, 2010). However, more research is needed on the average clothing items bought by
Number of purchased clothes per year
0-9 10-19 20-29 30-39 40+
30
Dutch consumers. The research of Maldini et al (2017) only included 50 respondents, which
is not many and therefore conclusions on their findings should be treated carefully.
Besides looking at this average amount of clothes bought annually, it is also important to
make a regression on willingness to pay including the amount of clothes bought annually as
an independent variable to see whether it affects willingness to pay. See table 7 for the
outcome of this regression.
Variables
WTP
WTP
Coef.
Stand. Error
Coef.
Stand. Error
Female
.061*
.035
.061*
.035
Generation X
-.061
.047
-.060
.048
Millennials
-.150***
.052
-.152***
.052
Generation Z
-.091*
.048
-.087*
.049
Education high
.003
.035
.005
.036
Information
.044
.029
.043
.029
Knowing
.050**
.020
.049**
.020
Amount of clothes
-.001
.001
Num. of obs.
403
397
F (7, 395)/F (8, 388)
5.41
4.67
***p<0.01, **p<0.05, *p<0.1
Table 7: Regressions on willingness to pay with and without an independent variable amount of clothes bought
annually.
It can be concluded that the amount of clothes bought annually by Dutch consumers does not
significantly (p=0.454) influence the willingness to pay. If the variable would have been
significant, then the correlation between willingness to pay and buying clothes annually
would be negative.
Besides the low average amount of purchased clothes per year in this research and the
regression including the amount of clothes bought annually, another remarkable result on
willingness to pay was found after doing the data analysis. To deal with the experimenter
demand effect in this research the respondents were asked to state to which statement they
agree more: I support sustainable clothing and am willing to pay more for it; I support
sustainable clothing, but I am not willing to pay more for it; and I do not support sustainable
31
clothing regardless of the price. A remarkable outcome was that most people who say to agree
most with the statement: I support sustainable clothing, but I am not willing to pay more for it,
are still willing to pay more for sustainable clothes. However, the willingness to pay is the
highest for people who agree most with the first statement. The willingness to pay for the
second statement is 34.2%-point lower than for the first statement (p=0.000) and the
willingness to pay for the third statement is 52.1%-point lower than for the first statement
(p=0.019). See Appendix 4 for the regression. Therefore the fact that some people agree most
with the statement including not to pay more, but still are willing to pay more, might be
explained by them willing to pay more, but not much more. They feel more in line with this
statement and not with the statement of willing to pay more. They most likely associate
paying more with a higher percentage than they would like to pay more. For a clear overview
of the observations and average willingness to pay per statement, see table 8.
Statement
Observations
Average WTP
1
292
52.1%
2
109
17.9%
3
2
0%
Table 8: Number of observations and average willingness to pay per statement.
4.3.7 Conclusion regression analysis
In this section, the five hypotheses of this research have been tested with a linear regression
analysis. Only one of the five hypotheses could be accepted. The other four hypotheses are
rejected for different reasons. See table 9 for an overview of all researched hypotheses and
their findings.
Hyp.
Variable (p-value)
Expected influence on WTP
Actual influence on WTP
1
Generation
(p=0.168/0.003/0.053)
Younger generations higher WTP
Older generations higher WTP
2
Gender (p=0.073)
Woman higher WTP
Woman higher WTP
3
Education (0.440)
Higher educated people higher
WTP
No evidence due to
insignificance
4
Information (0.135)
More informed higher WTP
No evidence due to
insignificance
32
5
Knowing people
(0.019)
Knowing more people higher
WTP
Knowing more people higher
WTP
Table 9: Overview of the researched hypotheses.
Hypotheses 3, and 4 are rejected because these independent variables are not significant. The
hypothesis on generation is rejected because the influence on willingness to pay is the
opposite of expected. The variables knowing people and gender are significant, from the
variable generation only generation 2 (Generation X) is not significant
4
. In the next chapter it
is discussed what these results mean for the conclusion of this research.
5. Discussion
In this chapter first the interpretations of the results are discussed, so what the results of this
research mean. Second, the limitations of this research are discussed. Last, it is discussed
what further research could be done after this research.
5.1 Interpretations
The results of the data analysis are discussed in the previous chapter. In this section it is
discussed what these results mean.
Only one of the five hypotheses could be accepted at the 95% confidence level after the data
analysis. The hypothesis that has been accepted is: People who know more people who are
concerned with sustainability are willing to pay more for sustainable clothes. This hypothesis
was formulated to find out whether the bias herding behavior also occurs by the willingness to
pay for sustainable clothes. When you know more people who are concerned with
sustainability, you are more likely to act in line with them. This is the same as with
investments, if a lot of people invest in a certain asset, other people will follow. This
phenomenon is known as herding behavior and the acceptance of this hypothesis indicates
that this phenomenon also occurs by consumers buying sustainable clothes. With the other
4
Multiple hypotheses are tested with the regression. For the model using the dependent variable willingness to
pay a correction has been made for this. The Holm-Bonferroni test is used to do this. This method is more
powerful than the Bonferroni test (Glen, 2021). The formula of the Holm-Bonferroni test is:
(Salkind, 2010). The correction showed that the variables are not significant at 95% confidence level. This
correction suggest that there might be a power issue with the results, thus the results should be interpreted
cautious.
33
five models using the other samples of willingness to pay, the hypothesis could also be
accepted. As discussed, this hypothesis could capture other motives as well, like having
friends with the same interests and being affected by your upbringing.
Another hypothesis could be accepted at 90% confidence level. This hypothesis is: Women
are willing to pay more for sustainable clothes. Using the models including willingness to pay
with a low anchoring price and for shirts, the hypothesis could be accepted at 95% confidence
level. For the other three models, the variable is not significant. This outcome is in line with
the outcomes of the discussed researches of Gazzola (2020) and Brough (2016).
Two of the other five are rejected due to insignificancy. These two hypotheses are: Higher
educated people are not willing to pay more for sustainable clothes, and people who are
better informed about the problems in the clothing industry are willing to pay more for
sustainable clothes. The two independent variables: education and level of information are not
significant. However, the level of information is close to being significant (p=0.135). This
means that it is important to keep in mind that some evidence does exist that the hypothesis of
this variable could have been accepted if more respondents were achieved for the
questionnaire, because the coefficient is positive. In this case this positive coefficient means
that some evidence exists that people who are better informed about the problems in the
clothing industry are willing to pay more for sustainable clothes. This is in line with the
studies of Connell (2014) and United Nations (2012), because they found that more
knowledge about environmental problems causes people to promote sustainable development.
With buying and willing to pay more for sustainable clothes, the consumers promote and help
the clothing industry to become more sustainable. For the other models, the two independent
variables education and information level are also insignificant.
Only one hypothesis has been rejected while being significant. This hypothesis is: The
younger generations are willing to pay more for sustainable clothes. The regression showed a
negative coefficient. In this case this implies that the younger generations are willing to pay
less than the older generations. This is the opposite of what the hypothesis suggests, therefore
this hypothesis has been rejected. This outcome might be caused by the fact that older
generations earn more money and thus are able to be willing to pay more for sustainable
clothes (CBS, 2019). Consumers with a very low income are also not the target population of
this research, since they cannot afford to buy many clothes. It might be that some of these
34
people are still included in mainly the youngest two generations, which might also affect the
outcome of this hypothesis. The outcome that older generations are willing to pay more is not
in line with most literature discussed in chapter 2 (e.g. Gazzola, 2020/Han, 2016/McKinsey,
2020). It is in line with the research of Notaro and Paletto (2021) that also has been discussed
in chapter 2. They found that older people are more willing to buy eco-friendly products and
especially clothes and furniture, they care less about price than younger people (Notaro and
Paletto, 2021). With the other models using the different samples of willingness to pay, the
hypothesis is also rejected, because they all show a negative correlation.
When comparing the three significant variables to which extent they influence the willingness
to pay for sustainable clothes by Dutch consumers, then generation has the highest influence,
namely for Millennials -15.0%-point and for Generation Z -9.1%-point. Knowing people who
are concerned with sustainability does influence willingness to pay positively with 5.0%-point
by every step higher in knowing people and gender does influence it more with 6.2%-point.
Information level is close to being significant and thus important to also take into account.
The influence is 4.4%-point per step of information level, which is the lowest influence of
these four variables. The coefficients are not tested for difference, so the conclusion is only
observational.
Another remarkable outcome of the data analysis is that people estimate their number of
purchased clothes per year much lower than the actual average of clothes Dutch consumers
buy. The average of this research is 20, while the average of Dutch consumers according to
other research is 46.
Concluding this: in this research five hypotheses have been tested. Two of these hypotheses
have been rejected, because the independent variables are not significant. The independent
variables of the other three hypotheses are significant. The hypothesis on younger generations
willing to pay more is rejected at 95% confidence level, the hypothesis on people who know
more people who are concerned with sustainability are willing to pay more is accepted at 95%
confidence level and the hypothesis on women willing to pay more is accepted at 90%
confidence level.
35
5.2 Limitations
Several previous studies on WTP are based on real investment, and thus based on revealed
preferences data. This study is not based on revealed preferences data. An advantage of
revealed preferences data is that it is based on actual decisions and choices in real-world
situations. However, stated preferences data, which is used in this study, are generally
collected in studies with experiments or surveys. This data is based on situations where a
decision or choice is made by considering hypothetical scenarios. A general advantage of this
data compared to revealed preferences data is that they are not limited to products, situations,
and attributes of alternatives that do exist or have existed, but can also refer to new products
or investments, products with low market penetration, or new attributes of products or
investments (Gutsche and Ziegler, 2019). Sustainable clothes are also relatively new products,
therefore stated preferences data is used in this study. Another reason for choosing stated
preferences data is that the questions in the survey regarding this method are easier to
understand for the respondents, which makes it also easier to get enough respondents.
However, one of the limitations of this study is thus that the data is not based on actual
decisions and choices in real-world situations.
As discussed in 3.2 another problem that could occur when doing research to willingness to
pay is the experimenter demand effect. This problem occurs when respondents change their
behavior, because they think this is appropriate behavior. In this research, this problem has
partly been solved with the dissonance minimizing referendum method. However, it still is a
limitation of this research that you do not know whether the respondents show their actual
behavior or the behavior they think is appropriate. In this situation the use of revealed
preferences data might also solve the problem, since this is based on actual decisions and
choices in real-world situations. Still, when respondents know they are participating in an
experiment, they could still show behavior that they think is appropriate behavior. Therefore a
situation needs to be created in which people do not know they are participating in an
experiment.
For this research it has been chosen to make use of the method of voluntary response
sampling and online data collection. Unfortunately this method is always at least somewhat
biased, since some people are more likely to volunteer for filling in a questionnaire than
others. This is then also a limitation of this research, but this would also have been the case
with lab experiments. With voluntary response sampling it is also difficult to get equal
36
amounts of respondents in the subgroups. For example in this research more respondents from
older generations do exist. See table 10 on the next page for the distribution of this group.
Generation
1946-1964 (Boomer)
123
30.4%
1965-1981 (Generation X)
148
36.6%
1982-1995 (Millennial)
51
12.6%
1996-2019 (Generation Z)
82
20.3%
Table 10: Example not equal number of respondents in subgroups.
This limitation implies that the conclusions made in this research about the population are
weaker than with probability samples. Therefore careful considerations should be made to
factors such as respondent gender, age, interest, etc. (Lefever et al., 2007). These factors are
the independent factors of this research, thus it should be mentioned that the conclusions of
this research should be treated carefully. Despite the limitations of this method, it provides
researchers with an unique opportunity for collecting data, because they can potentially have
access to a large and geographically distributed population, while being time and cost
efficient (Lefever et al., 2007).
The last limitation of this research is the sample size. Although the sample size is not small, if
the sample size would have been larger, then most likely more independent variables would
have been significant. More time and if possible money is needed to test the hypotheses with a
larger sample size. It would give the research more power.
5.3 Further research
While doing the regression analysis, it was found that there is a very low R-squared. The
regression can be found in Appendix 3. The R-squared of this model is 0.0660, which implies
that 6.6 percent of the total variance in willingness to pay for sustainable clothes is explained
by the independent variables of this model. The independent variables of this model are
gender, generation, education, information level and knowing people who are concerned with
sustainability. So the percentage is relatively low, which means that this model has a
relatively low explanatory power. However, this is not really a problem, because the model is
not intended to predict the willingness to pay. What is a problem is that this research has been
done to find out more about which factors are influencing the willingness to pay for
37
sustainable clothes. The low R-squared indicates that a lot of other factors do influence this
willingness to pay. This means that more research needs to be done to find more factors
influencing the willingness to pay for sustainable clothes. The influence of income is not
examined in this research, because of the correlation with education, however this should still
be examined. Something else that could influence the willingness to pay is the kind of
problem in the clothing industry the consumers care about most. For example: do they care
more about the environment or do they care more about the people working in the clothing
industry? Notaro and Paletto (2021) also did research on this for sustainable choices, they
found that men make sustainable choices based on the negative impact on water quality,
forests, and air quality and women based on the negative impact on health and animals. This
might also influence the willingness to pay for sustainable clothes. This could be tested by
asking a question about what negative impact of producing non-sustainable clothes motivates
someone most to buy sustainable clothes. When knowing more about this, it can be taken into
account with producing and selling sustainable clothes in a way that more people are willing
to buy sustainable clothes. For example if the outcome of a research on this would show the
same result as Notaro and Paletto (2021) about women caring a lot about animals, then this
should be taken into account when producing and selling clothes. Animals should not be
affected when producing the clothes and when selling these clothes this should be mentioned
to give women an extra incentive to buy these clothes.
Two of the independent variables, gender and level of information, are very close to being
significant. This problem could be solved with more respondents for the questionnaire.
Therefore future research could be done on this topic with more respondents. To see whether
the variables are significant then and maybe the two hypotheses on these variables could then
be accepted. They both show a positive correlation with willingness to pay, this is in line with
the expectations. So if the variables are significant, the hypothesis could be accepted.
The hypothesis regarding younger generations willing to pay more for sustainable clothes is
rejected, because the regression showed that older generations have a higher willingness to
pay. As discussed, this may be caused by the fact that older generations have a higher income.
Further research is needed to see whether this is correct or whether other reasons do exist for
older generations to pay more.
38
5.4 Recommendations
As discussed in chapter 4, two hypotheses could be accepted after the data analysis. One of
the hypotheses that has been accepted is: People who know more people who are concerned
with sustainability are willing to pay more for sustainable clothes. This indicates a kind of
herding behavior in sustainable consumer behavior. As discussed in 2.1, it has been found that
millennials prefer receiving advice through social media influencers. Millennials have the
lowest significant willingness to pay found compared to Boomers. So, on one hand there
should not be focused on them, because their willingness to pay is low, on the other hand
there should be focused on them, because their willingness to pay has a lot of potential to
grow. Another reason to still focus on Millennials is that only two of the Millennials of this
research did not want to pay more for sustainable clothes and only one of them wanted to pay
less for sustainable clothes compared to non-sustainable clothes, while the other 48
Millennials did want to pay more for sustainable clothes. This means most Millennials are
willing to pay more for sustainable clothes. Therefore the recommendation for clothing brands
is that they should introduce their sustainable clothes through influencers. Herding behavior
means that people act like other people, if consumers act like the influencers, the clothing
brands can sell more sustainable clothes. This would help partly solve all the problems in the
current clothing industry. The other hypothesis that could be accepted is: Women are willing
to pay more for sustainable clothes. Therefore another recommendation is that clothing
brands should focus most on women with producing and selling sustainable clothes. This
could also be done with the help of social media influencers. Then there should be focused on
social media influencers with relatively more female followers.
The recommendation for the government regarding the findings of this research is to make
sustainable clothes more common. The Dutch government did already do research to examine
how to best do this. They are willing to do this by changing the mindset of consumers and
changing behavior of consumers. With changing mindset they want to let consumers influence
each other with their behavior. With their behavior change model they want to systematically
and efficiently change behavior effectively (Rijksoverheid, 2020). The findings of this
research are reason to recommend to indeed do this.
39
6. Conclusion
Several problems do exist in the clothing industry. Problems for the environment, for people
who work in the clothing industry, and even for people who live around clothing factories.
Consumers are mainly responsible for the demand side of this problem. It is important to
know more about the consumer behavior in the clothing industry. Sustainable clothes are
more expensive to produce. Therefore it is important to know the willingness to pay of
consumers for sustainable clothes. Besides the importance of this willingness to pay, it is also
important to know which factors are influencing the willingness to pay by consumers for
sustainable clothes. If the factors influencing the willingness to pay are known, then clothing
brands can use this to sell sustainable clothes. In this research a questionnaire has been used to
do further research on this.
Factors influencing the willingness to pay which are examined with the help of this
questionnaire are generation, gender, education level, level of information about the problems
in the clothing industry, how many people someone knows who are concerned with
sustainability, and annually bought clothes. The data collected with the questionnaire has been
analyzed with the help of Stata. With the help of a regression analysis the five hypotheses of
this research have been tested. Only two of the five hypotheses could be accepted, which are:
People who know more people who are concerned with sustainability are willing to pay more
for sustainable clothes and women are willing to pay more for sustainable clothes. The
finding of knowing people indicates that some kind of herding bias does occur in the clothing
industry, just like in the financial market. Clothing brands could use these findings about
these factors influencing willingness to pay by using influencers introducing sustainable
clothes and probably focusing more on women first.
Two of the hypotheses have been rejected because the influence of the independent variables
were not significant. These two hypotheses are: Higher educated people are not willing to pay
more for sustainable clothes and people who are better informed about the problems in the
clothing industry are willing to pay more for sustainable clothes. The variable information is
close to being significant, and showed a positive correlation, as expected. If this research had
more respondents, it might be the case that the variable would have been significant.
Therefore it is important to mention that some evidence does exist that the hypothesis is
correct. Further research is needed to find whether this is correct.
40
Only one of the hypotheses has been rejected while being significant, which is: The younger
generations are willing to pay more for sustainable clothes. Boomers are willing to pay most,
then Generation Z and last Millennials. About Generation X no conclusions can be made due
to insignificancy. A possible explanation found for the Boomers willing to pay most is that
older generations earn more, and therefore are able to pay more for sustainable clothes.
Further research is needed to find out whether this is correct.
The regression analysis on willingness to pay with the five independent variables showed a
very low R-squared, namely 0.0660. This implies that 6.6% of the total variance in
willingness to pay for sustainable clothes is explained by the independent variables of this
model. This is lower than expected and further research is needed to find other factors
influencing the willingness to pay.
The research question of this research is: What factors are influencing to what extent Dutch
consumers are willing to pay more for sustainable clothes? With the help of the hypotheses it
has been found that gender, generation, and knowing people who are concerned with
sustainability do influence the willingness to pay for sustainable clothes by Dutch consumers.
Gender does influence willingness to pay positively with 6.2%-point, which means that
women on average are willing to pay 6.2%-point more than men for sustainable clothes.
Generations do influence willingness to pay negatively, which means that Generation X on
average is willing to pay 6,1%-point less than Boomers, Millennials 15.0%-point less than
Boomers, and Generation Z 9.1%-point less than Boomers. However the influence of
Generation X is not significant. Knowing people who are concerned with sustainability does
influence willingness to pay positively with 5.0%-point. which means that by every increase
of the five steps of how many people someone knows, on average the willingness to pay
increases with 5.0%-point. The mean willingness to pay by Dutch consumers is 43%, and
93% of the respondents were willing to pay more for the sustainable clothes. Some evidence
does exist that level of information about the problems in the clothing industry does also
influence to what extent Dutch consumers are willing to pay more. So some factors have been
found, however further research is needed to find more factors influencing the willingness to
pay for sustainable clothes by Dutch consumers. This might help get the clothing industry to
become more sustainable.
41
7. References
Apesteguia, J., Oechssler, J. and Wiedenholzer, S. 2020. “Copy trading.” Management
Science.
Arora, N., Charm, T., Grimmelt, A., Ortega, M., Robinson, K., Sexauer, C., Staack, Y.,
Whitehead, S. and Yamakawa, N. 2020. “A global view of how consumer behavior is
changing amid COVID-19.” Retrieved from: covid-19-global-consumer-sentiment-20200707.pdf
(mckinsey.de)
Aryal, K., Chaudhary, P., Pandit, S. and Sharma, G. “Consumers’ Willingness to Pay for
Organic Products: A Case From Kathmandu Valley.” Journal of Food and Agriculture and
Environment, vol. 10, p. 12-21.
Astivia, O. and Zumbo, B. 2019. “Heteroskedasticity in Multiple Regression Analysis: What
it is, How to Detect it and How to Solve it with Applications in R and SPSS.” Practical
Assessment, Research, and Evaluation, vol. 24.
Blankenberg, A. and Gottschalk, J. 2018. “Is socially responsible investing (SRI) in stocks a
competitive capital investment? A comparative analysis based on the performance of
sustainable stocks.” Center for European, Governance and Economic Development Research
(cege).
Boström, M. and Micheletti, M. 2016. “Introducing the Sustainability Challenge of Textiles
and Clothing.” Springer, Journal of Consumer Policy, p. 367-375.
Brough, A., Wilkie, J., Ma, J., Isaac, M. and Gal, D. 2016. “Is eco-friendly unmanly? The
green-feminine stereotype and its effect on sustainable consumption.” Journal of Consumer
Research, p. 567–582.
Brucculieri, J. 2018. “Sustainable fashion brands explain that yes, they can be profitable.”
Huffington Post.
42
Carley, S. and Yahng, L. 2018. “Willingness-to-pay for sustainable beer.” PLoS ONE,
journal.pone.0204917.
CBS. 2011. “Consument koopt meer duurzame goederen.” Retrieved from: Consument koopt
meer duurzame goederen (cbs.nl)
CBS. 2019. “Welvaart in Nederland, 2019.” Retrieved from: WelvaartinNederland2019_web.pdf
CFI. 2021. “What is Herd Mentality Bias?” Retrieved from: Herd Mentality - Overview,
Examples, Impact of Social Bias (corporatefinanceinstitute.com)
Chang, C. and Lin, S. “The effects of national culture and behavior pitfalls on investors’
decision-making: Herding behavior in international stock markets.” International Review of
Economics & Finance, vol. 37, p. 380-392.
Ciasullo, M., Troisi, O. and Torre, C. 2017. “What about Sustainability? An Empirical
Analysis of Consumers’ Purchase Behavior in Fashion Context.” Sustainability, MDPI.
Claudio, L. 2007. “Waste Couture: Environmental impact of the Clothing Industry.”
Environmental Health Perspectives, vol. 115.
Connell, K. and Kozar, J. 2014. “Environmentally Sustainable Clothing Consumption:
Knowledge, Attitudes, and Behavior.” Roadmap to Sustainable Textiles and Clothing, p. 41-
61.
Cooper, T., Hill, H., Kininmonth, J., Townsend, K. and Hughes, M. 2013. “Design for
Longevity. Guidance on increasing the active life of clothing.” Working together for a world
without waste.
Czibor, E., Jimenez-Gomez, D. and List, J. 2019. “The Dozen Things Experimental
Economists Should Do (More of).” Southern Economic Journal, vol. 86, p. 371-432.
43
Dhir, A., Sadiq, M., Talwar, S., Sakashita, M. and Kaur, P. 2021. “Why do retail consumers
buy green apparel? A knowledge-attitude-behaviour-context perspective.” Journal of
Retailing and Consumer Services, vol. 59.
European Commission. 2021. “Paris Agreement.” Retrieved from: Paris Agreement | Climate
Action (europa.eu)
Filiz, I., Nahmer, T. and Spiwoks, M. 2019, “Herd behavior and mood: An experimental
study on the forecasting of share prices.” Journal of Behavioral and Experimental Finance,
vol. 24.
Gazzola, P., Pavione, E., Pezzetti, R. and Grechi, D. 2020. “Trends in the Fashion Industry.
The Perception of Sustainability and Circular Economy: A Gender/Generation Quantitative
Approach.” Sustainability, MPDI, vol. 12.
Glen, S. 2021. “Holm-Bonferroni Method: Step by Step.” StatisticsHowTo.com: Elementary
Statistics for the rest of us! Retrieved from: Holm-Bonferroni Method: Step by Step - Statistics
How To
Grondelle, V. 2018. “A Negative Duty to Not Buy Unfairly Produced Clothes.” University
Pompeu Fabra. Retrieved from: A-Negative-Duty-To-Not-Buy-Unfairly-Produced-Clothes.pdf
(researchgate.net)
Gutsche, G. and Ziegler, A. 2019. “Which private investors are willing to pay for sustainable
investments? Empirical evidence from stated choice experiments.” Journal of Banking &
Finance, vol. 102, p. 193-214.
Han, J., Seo. Y. and Ko, E. 2016. “Staging luxury experiences for understanding sustainable
fashion consumption: A balance theory application.” Journal of Business Research, vol. 74,
p. 162-167.
Huq, F., Chowdhury, L. and Klassen, R. 2016. “Social management capabilities of
multinational buying firms and their emerging market suppliers: An exploratory study of the
clothing industry.” Journal of Operations Management, vol. 46, p. 19-37.
44
Jacobs, K., Petersen, L., Hörisch J. and Battenfeld, D. 2018. “Green thinking but thoughtless
buying? An empirical extension of the value-attitude-behaviour hierarchy in sustainable
clothing.” Journal of Cleaner Production, vol. 203, p. 1155-1169.
Kamphuis, A., Baart, L. and Roelofs, S. 2020. “Jongeren & duurzame kleding.” CNV
Internationaal.
Kendall Cox Park. 2018. “Understanding ethical consumers: willingness-to-pay by moral
cause”. Journal of Consumer Marketing, vol. 35, p. 157-168.
Lacy, P. and Rutqvist, J. 2016. “Waste to Wealth: The Circular Economy Advantage.”
Accenture strategy, Palgrave Macmillan.
Lefever, S., Dal, M. and Matthiasdottir, A. 2007. “Online data collection in academic
research: advantages and limitations." British Journal of Educational Technology, vol. 38, p.
574-582.
Lin, S. 2009. “Exploratory evaluation of potential and current consumers of organic cotton in
Hawaii.” Asia Pacific Journal of Marketing and Logistics, vol. 21, p. 489-506.
MAKE.GOOD. 2020. “How fashion is polluting global waterways.” Retrieved from: River
Blue The Movie | Knowledge Hub | MAKE GOOD
Maldini, I., Duncker, L., Bregman, L., Piltz, G., Duscha, L., Cunningham, G., Vooges, M.,
Grevinga, T., Tap, R. and Balgooi, F. 2017. “Measuring the Dutch clothing mountain: Data
for sustainability oriented studies and actions in the apparel sector.” Saxion, Academie ACT.
McKinsey & Company. 2020 “The State of Fashion 2020.” The Business of Fashion.
Retrieved from: The State of Fashion 2020 (mckinsey.com)
Nosto. 2019. “Consumer survey: Sustainability in fashion retail.” Nosto. Retrieved from:
EN-Sustainability-Summary-2019 (1).pdf (nosto.com)
45
Notaro, S. and Paletto, A. 2021. “Consumers’ preferences, attitudes and willingness to pay
for bio-textile in wood fibers.” Journal of Retailing and Consumer Services, vol. 58,
Park, A. and Sabourian, H. 2009. “Herding and Contrarian Behavior in Financial Markets.”
Econometrica, vol. 79, p. 973-1026.
Pedrini, M. and Ferri, L. 2014. “Socio-demographical antecedents of responsible
consumerism propensity.” International Journal of Consumer Studies, vol. 38, p. 127-138.
Pookulangara, S. and Shephard, A. 2013. “Slow fashion movement: Understanding consumer
perceptions – An exploratory study.” Journal of Retailing and Consumer Services, vol. 20, p.
200-206.
Rijksoverheid. 2020. “Gedragsonderzoek kleding.” Ministerie van Infrastructuur en
Waterstaat. Retrieved from: Onderzoek+naar+kleding+en+gedrag+-
+uitgevoerd+door+D&B+in+opdracht+van+MinIenW.pdf
Salkind, N. 2010. “Encyclopedia of research design- chapters outlier, normal distribution
and Holm-Bonferroni.” Vol. 1. SAGE Publications, retrieved from: SAGE Reference -
Encyclopedia of Research Design (oclc.org)
Schanes, K., Gilium, S. and Hertwich, E. 2016. “Low carbon lifestyles: A framework to
structure consumption strategies and options to reduce carbon footprints.” Journal of Cleaner
Production, vol. 139, p. 1033-1043.
Schoenmueller, V., Libai, B. and Kogan, S. 2021. “The Rise and Fall of Influencers:
Evidence from Social Trading.” SSRN.
Shahbandeh, M. 2020. “Extra amount global consumers were willing to pay for sustainable
clothing 2018.” Statista. Retrieved from: • Premium consumers willing to pay for sustainable
fashion products worldwide 2018 | Statista
Simonson, I. and Drolet, A. 2003. “Anchoring Effects on Consumers’ Willingness-top-Pay
and Willingness-to-Accept.” Research Paper Series, Stanford graduate school of business.
46
Taylor, J. and Boasson, V. 2014. “Who buys fair trade and why (or why not?) A random
survey of households.” Journal of Consumer Affairs, vol. 48, p. 418-430.
Tolley, G. and Olson, E. 1971. “The interdependence between income and education.”
Journal of Political Economy, vol. 79.
Uehlike, E. 2016. “Convergent validity in contingent valuation: an application to the
willingness to pay for national climate change mitigation targets in Germany.” Journal of
Agricultural and Resource Economics, vol. 61, p. 76-94.
United Nations Economic Commission for Europe. 2012. “Environment for Europe, and
Sustainable Development Section.” UNECE.
Veldhoven, S. 2020. “From Linear to Circular in the Textile and Apparel Industries.”
Netherlands Enterprise Agency and Holland Circular Hotspot.
Wiederhold, M and Martinez, L. 2018. “Ethical consumer behaviour in Germany: the
attitude-behaviour gap in the green apparel industry.” International Journal of Consumer
Studies, vol. 42, p. 419-429.
Wittcoff, H., Reuben, B. and Plotkin, J. “Industrial organic chemicals.” Second Edition.
Retrieved from: Industrial_Organic_Chemistry.pdf (d1wqtxts1xzle7.cloudfront.net)
Zizzo, D. 2010. “Experimenter demand effects in economic experiments.” Experimental
Economics, vol. 13, p. 75-98).
47
Appendix 1 - Survey
Survey sustainable clothes
Start of Block: Default Question Block
You have been invited to participate in this study on consumer behavior within the sustainable
clothing industry. I am doing this research for my master's degree in financial economics at Radboud
University. The questionnaire will take approximately 3-5 minutes. Don't think too long about the
answer to the questions, there are no right or wrong answers. All information obtained during the
questionnaire will be treated in strict confidence. You remain anonymous and the data cannot be
traced back to individual participants.
Participation in the questionnaire is completely voluntary, you can stop at any time during the
questionnaire. If you don't stop, I kindly ask you to answer all the questions. The information obtained
from this questionnaire will be of great help to me for my research. Not much information is yet
available about consumer behavior within the sustainable clothing industry. I would therefore like to
thank you for taking the time to complete my questionnaire.
If you have any questions after completing the questionnaire, you may send them to
m.haverkamp@student.ru.nl.
Page Break
48
During this survey, a number of questions involve the topic of sustainably produced clothing. It is
therefore important to know what is meant by this in this survey. Aspects that fall within sustainably
produced clothing are: fair wages, high quality, social responsibility, less harmful for the environment,
and sustainably selected raw materials such as organic cotton. When talking about sustainably
produced clothing, these aspects are met. Is this clear to you?
o Yes (1)
o No (2)
Which of the statements below best describes your opinion?
o I support sustainable clothing and I am willing to pay more for clothing that is produced
sustainably. (1)
o I support sustainable clothing, but I am not prepared to pay more for clothing that is produced
sustainably. (2)
o I do not support sustainable clothing, regardless of the price. (3)
Page Break
49
The questions on this page are randomized.
Consider a non-sustainable shirt that you have in your closet for around €30. Now suppose that this
shirt cost exactly €30 for you. How much would you be willing to pay for it if it was produced
sustainably?
________________________________________________________________
Consider a non-sustainable shirt that you have in your closet for around €10. Now suppose this shirt
cost you exactly €10. How much would you be willing to pay for it if it was produced sustainably?
________________________________________________________________
Consider non-sustainable socks (6 pairs) that you have in your closet for around €15. Now suppose
these socks cost you exactly €15. How much would you be willing to pay for them if they were
produced sustainably?
________________________________________________________________
Consider non-sustainable socks (6 pairs) that you have in your closet for around €7.50. Now suppose
that these socks cost you exactly €7.50. How much would you be willing to pay for them if they were
produced sustainably?
________________________________________________________________
Consider a pair of non-sustainable jeans that you have in your closet for around €75. Now suppose that
these jeans cost you exactly €75. How much would you be willing to pay for it if it was produced
sustainably?
________________________________________________________________
Consider a non-sustainable pair of jeans that you have in your closet for around €30. Now suppose that
these jeans cost you exactly €30. How much would you be willing to pay for it if it was produced
sustainably?
________________________________________________________________
50
Page Break
In what period were you born?
o 1946-1964 (1)
o 1965-1981 (2)
o 1982-1995 (3)
o 1996-2019 (4)
What is your gender?
o Man (1)
o Woman (2)
o Different (3)
What is your highest level of education?
o Elementary School (1)
o Secondary school (2)
o Vocational education and training (3)
o Bachelor (4)
o Master or higher (5)
51
How well informed do you think you are about the problems in the clothing industry? This includes:
poor working conditions, loss of biodiversity, depletion of raw materials, high CO2 emissions and an
enormous amount of waste.
No information
at all (1)
Little
information (2)
Average
information (3)
A lot of
information (4)
All possible
information (5)
Clothing
industry
information: (1)
o
o
o
o
o
How many people do you know who are involved in sustainability? Think of: people who buy
sustainable clothing, consciously eat less meat, vote for a political party that is committed to
sustainability, consciously fly less, try to use less energy and consciously prefer to use the bicycle
rather than the car.
No people at
all (1)
Few people (2)
Just as many
people do or
don't (3)
Many people
(4)
All the people
(5)
People
concerned with
sustainability:
(1)
o
o
o
o
o
How many items of clothing do you estimate to buy annually?
________________________________________________________________
End of Block: Default Question Block
52
Appendix 2 - Assumption tests
- VIF test for Multicollinearity
All the VIF values are below 10. This indicates that none of the five independent variables shows a
sign of multicollinearity, and no correction is needed for multicollinearity.
- Cook-Weisberg test for Heteroskedasticity
The null hypothesis is rejected in this case, because the P value is smaller than 5% significant level.
This means that there has been a problem with heteroskedasticity. Therefore robust standard errors
have been used in the regression.
53
Appendix 3 - Linear regressions
- Regression on willingness to pay
- Regression on willingness to pay with low anchor price
.
54
- Regression on willingness to pay with high anchor price
- Regression on willingness to pay for shirts
55
- Regression on willingness to pay for socks
- Regression on willingness to pay for jeans
56
- Regression on willingness to pay including variable amount of clothes
57
Appendix 4 - Regression on statement willingness to pay
