Master of Science in International Business and Management “Sustainability in the Beauty Industry and its Effect on Consumers’ Purchase Intention”

(1)

Master of Science in International Business and Management

Master Thesis

“Sustainability in the Beauty Industry and its Effect on Consumers’

Purchase Intention”

Author: Laura Nieto-Márquez González Student number: 3796116

Email: l.nieto-marquez.gonzalez@student.rug.nl Supervisor: Dr. T. Halaszovich

Co-Assessor: Dr. R.W. de Vries Date of Submission: 15th of January, 2020

Word count: 8,173 (excluding tables, references, and appendix)

Faculty of Economics and Business University of Groningen

Duisenberg Building, Nettelbosje 2, 9747 AE Groningen, The Netherlands P.O. Box 800, 9700 AV Groningen, The Netherlands

(2)

ABSTRACT

The beauty industry is one of the longest-standing and highest-grossing creative industries in the world, and it experienced a great surge of growth in the past decade. Whilst there have been gaps in the sustainability management from brands almost from the beginning of the industry’s history, the latest high increase in competitiveness and volume of products and brands in the business has led to consumers knowing they have more than just one choice – as opposed to the couple of renowned conglomerate brands most customers shopped from years ago.

There has been a shift in the beauty consumer, and they demand transparency and more responsibility from brands. However, there still exists a massive wave of consumerism among individuals – how truly sustainable is this endless consumption of beauty products? Does the end consumer truly care about beauty brands’ social and environmental responsibility?

Building on previous research on sustainability and habitual behavior, an online survey was conducted. A final sample of 131 respondents demonstrated that beauty brands’ sustainable (or unsustainable) behavior does affect and helps predict consumers’ purchase intentions; although this relationship was negatively moderated by consumers’ habits. Further research would be needed to see if other variables could amplify this relationship as well.

(3)

LIST OF FIGURES AND TABLES

P.:

Figure 1: Conceptual model 14

Table 1: Frequency table 17

Table 2: Reliability analysis for habitual behavior variables 19

2.1: Cronbach's Alpha 19

2.2: Cronbach's Alpha if item deleted 20

Table 3: Factor analysis of habitual behavior variables 21

3.1: Eigenvalues 21

3.2: Communalities 22

3.3: Rotated factor matrix 23

Table 4: Reliability analysis for CSR variables (pre-removal of Q21) 24

4.2: Cronbach's Alpha if item deleted 25

Table 5: Factor analysis of CSR variables (pre-removal of Q21) 26

5.1: Eigenvalues (Option A) 26

5.2: Communalities (Option A) 27

5.3: Rotated component matrix (Option A) 28

5.4: Eigenvalues (Option B) 29

5.5: Communalities (Option B) 30

5.6: Rotated factor matrix (Option B) 31

Table 6: Reliability analysis of CSR variables 32

(5)

Table 7: Factor analysis of CSR variables 34

7.1: Eigenvalues 34

7.2: Communalities 35

7.3: Rotated factor matrix 36

Table 8: Correlations 38

8.1: Correlations 38

8.2: (cont.) 38

8.3: New correlations 39

Table 9: Regression analysis 40

9.1: Model summary 40

9.2: ANOVA 40

9.3: Coefficients 41

Table 10: Moderated regression analysis 42

10.1: Model summary 42

10.2: Model 42

10.3: Interactions 42

(6)

1. INTRODUCTION

The word ‘cosmetic’ or ‘cosmetics’ first appeared in the 1600s to describe “the art of beautifying, anointing, or decorating the human body” (Lexicon n.d.). However, even though the word as we know it has only existed for a relatively short period of time, cosmetics have been used to enhance one’s natural beauty from Ancient times, thus making the beauty industry one of the currently longest-standing ones – perhaps making it all the more interesting to analyze.

The beauty industry has always been characterized by cycles, much like the fashion industry – from the passion for pallor of skin that remained until the Dark Ages (c. 5th – 10th century) and quickly turned into an unhealthy obsession about the removal of the tiniest skin imperfection to a craze for red amongst the French and English aristocrats until the French Revolution (1789). By the 19th century, the predecessor of the ‘Gibson Girl’ had appeared in Paris (Tungate 2012: 7 – 104), setting the ideal of beauty until World War I: “an independent young woman [...] [who was able] to take part in and enjoy strenuous physical activities [like] bicycling, playing tennis and golf, horseback riding, swimming, and the like” (Library of Congress 2013).

Beauty was usually only for the wealthiest individuals, but it eventually grew to become more affordable, and in Paris pots of rouge started to range more in price (Vigarello 2004) to allow women to appear healthy and fresh. By the end of the century, even the traditionally expensive and luxurious emulsions and cold creams started having to compete with the beauty products that would make way to what we now consider ‘typical’ cosmetics (Tungate 2012).

Despite the Parisian leadership in cosmetic products and its worldwide popularity as the capital of beauty, the outbreak of World War I (1914) catapulted the United States of America to the podium of the world’s beauty market (Jones 2011). The lack of conflict in the USA had allowed for a market expansion whilst European factories either shut down or switched to the manufacturing of weaponry and other war products. In fact, the American cosmetics and toiletries market had reached US$60 million by 1919, nothing short of incredible when taking into account that most creative and exclusive processes had come from Europe until then (Jones 2011).

(7)

Not long after that, the industry skyrocketed. The cosmetics and toiletries (or personal care products) industry is now a multi-billion industry and one of the largest creative industries (Jones 2011), reaching a decade high in 2018 and continuing to grow all along 2019 (Euromonitor 2019). In fact, the leading beauty manufacturer worldwide in 2018, L’Oréal S.A., made a whopping €26.9 bn in the aforementioned fiscal year (L’Oréal S.A. 2018), followed by Unilever, which made €20.6 bn (Unilever 2018). The global cosmetic products market is expected to reach US $863 bn in 2024, which translates into a growth rate of more than 7% in six years, from 2018 to 2024 (Zion Market Research 2018) – it is one of the very few sectors that remains seemingly unaffected by economic crises (Reuters 2018), perhaps because the desire to look young, clean, and beautiful is and has always been something inherently nurtured in every culture around the world, both in women and more increasingly in the past few years, in men.

Not surprisingly, the monetary compensation that the sector had promised since the end of World War I had brought plenty of questionable practices to the industry. But there was a change in consumers: they no longer remained silent against the numerous news that continued to come out about their favorite cosmetics products – the average cosmetics consumer became concerned about their own safety, and grew increasingly skeptical of the companies that sold their products.

This was the start of the ‘green’ movement for the cosmetics industry, and the growing consumer’s preference for organic and natural ingredients in their products. The genesis of the Body Shop in 1976 that would truly revolutionize the beauty industry; it led the movement of fair trade business practices and cruelty-free products – eventually leading to the banning of animal testing, which was made final in the European Union on finished cosmetic products in September of 2004 and on all ingredients in March of 2009 (European Commission 2017) (Jones 2011). This environmentally-conscious trend in consumer preferences continued – a 1993 study already showed a strong, significant concern in people for major environmental issues, and that the respondents would not only be willing to pay higher taxes to fix these issues, but also to make adjustments and restrictions in their daily lives to help as well (Krause 1993). The so-called ‘green consumers’ first appeared in the 1990s after a New York-based survey found that no fewer than 18.8 million American households (Kumar 2005) were environmentally-interested consumers – thus accounting for almost 20% of the US population back then.

(8)

started introducing ‘green’ product lines that claimed to be natural and cruelty-free (Kumar 2005). However, critics were quick to follow from the beginning, too, and people noticed many of these companies’ claims about their environmental concerns were either embellished or promptly fabricated. Part of the problem stems from the fact that consumers can rarely discern the veracity of these claims, and have thus become rather skeptical and cynical about the advertising industry in general, but more so when ‘green marketing’ is involved in any way. Environmental claims in advertising are often seen skeptically, and consumers view them as misleading – what’s more, environmentally-conscious consumers do not even find ‘green’ ads authentic at all (Matthes and Wonneberger 2014).

This change in attitude on the part of the consumer was also favored by the rise of the Internet and how much easier it is for the average consumer to access information﹣and, inevitably, how easy it is to shop for goods and services online nowadays. Millennials are the first tech-savvy generation, and are thus more sophisticated shoppers than previous generations – they are concerned about social and environmental issues, although they are also self-indulgent and have a high spending power (Eastman, Iyer, & Thomas 2013). Such is their interest in online shopping that only in 2018, the percentage of beauty products sold in online channels accounted for 10% of all sales – representing a growth of more than 20% per year (Euromonitor 2019).

But in an already oversaturated market where brands are fiercely competing against each other, – not only for the best supplier prices, but especially for the consumer’s attention – companies see themselves forced to remain relevant and one-up each other by creating bold, interesting packaging, and constant product launches that most consumers simply cannot keep up with – and neither can the environment (Tournois 2014).

(9)

when some of their favorite brands show less than sustainable behaviors. Do consumers, then, really want brands to be sustainable? Or do they want them to keep coming out with products so that they can satisfy their demand for more?

Is the beauty industry’s consumerism simply a product of its capitalist history, and is the end consumer a mere bystander? At the end of the day, more and more products are constantly being launched by companies; endless influencers are even coming up with their own lines of beauty products – yet consumers eventually shop these products (and massively so), despite the avalanche of critics that revolve these usually controversial launches. Are these critics true, or are consumers being hypocritical? Do consumers only care when it does not come at a cost to them, or do they truly care about sustainability?

This paper thus aims to answer the following research question:

In an era when consumers claim to be more concerned about sustainability and demand more transparency from beauty brands and manufacturers; how truthful is this alleged desire for

(10)

2. LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT

2.1. Brands’ environmentally- and socially-responsible behavior

The main premise in this research paper’s question is the desire from consumers for more transparency from beauty brands and manufacturers, or rather, for ‘responsible’ ones. Since the birth and evident success of ‘green’ beauty and cosmetics brands like the Body Shop in 1976, many beauty conglomerates have tried to take advantage of the benefits that being labeled as ‘sustainable’ can bring to a company. Whilst there have been plenty of honest attempts from beauty brands at bettering the world we live in, environmental claims from firms are hardly ever seen as transparent or honest, and the end consumer has become skeptical with the passage of time (Matthes and Wonneberger 2014); perhaps because we as consumers are usually bombarded with advertisements on our daily lives, or perhaps because many of these aforementioned claims are sometimes phrased in slightly dishonest ways that do not portray the exact truth (Kumar 2005).

Since we are discussing socially- and environmentally-responsible behaviors in the beauty industry and how those affect consumers’ behavior, it seems safe to assume that one of the variables that will directly influence people’s purchasing intention is the corporate social responsibility (or CSR) measures applied by the cosmetics companies that manufacture the products they buy and the behavior attached to them – scandals, lies, intentional misinformation; but also embracing cruelty-free or vegan manufacturing lines, reducing packaging options for the consumer, or helping people by employing locals.

(11)

a corporate environmental policy statement (CEP), but there is little proof that companies actually do implement these plans (Ramus and Montiel 2005).

It has been proven that consumers try to shop for beauty products in a conscious way. In 2012 already, an American survey carried out by the online retailer Vitacost found out that 75% of women preferred to purchase products that came with the ‘cruelty-free’ bunny (Sahota 2013: 4). Beauty companies’ behavior is constantly scrutinized, and not without reason – environmental pollution has been associated with cosmetics due to the amount of chemicals used to manufacture products and also to the damage the finished product can cause to our planet; the ingredients are often sourced unethically and are sometimes unsafe for the consumer, and the packaging is excessive and unsustainable (Sahota 2013).

It was already in the 1990s when several surveys showed consumers’ increasingly heightened concern about environmental issues – a New York survey group estimated in 1990 that about 20% of the US population (18.8 million households) were ‘environmentally-interested’, their main concerns being animal rights, clean air, and waste management (Kumar 2005). In 1993, a study also showed a strong, significant concern in people for major environmental issues, and that the respondents would not only be willing to pay higher taxes to fix these issues, but also to make adjustments and restrictions in their daily lives to help as well (Krause 1993).

Companies’ advertised efforts to become environmentally- or socially-friendly under stakeholder pressures are not always seen as truthful, however, and their outward behavior can oftentimes work against them when seen as money-grubbing and cold instead of as an honest attempt to fix their wrongdoings. Critics have always been quick to point out the gaps in cosmetic manufacturers’ claims even from the late 1990s – for example, many so-called ‘biodegradable’ packages cannot truly and 100% degrade in most landfills (Kumar 2005). Environmental claims in advertising are often seen skeptically, and consumers view them as misleading – what’s more, environmentally-conscious consumers do not even find ‘green’ ads authentic at all (Matthes and Wonneberger 2014).

(12)

individual concerned about social and ecological issues will not buy beauty products from a brand that is not inclusive to all skin colors, or that is not vegan.

2.2. Purchase intention

Purchase intention has widely been used in research as a proxy for purchase behavior, and it can be defined as an individual’s willingness to purchase a certain product or brand – or rather, when a consumer intends to become involved in a purchase transaction (Ling, et al. 2010). Positive altruistic motives from consumers, like improving society or the environment, positively and significantly affect consumers’ attitudes towards firms, and thus their intention to buy from said firms (Wongpitch et al 2016). What consumers perceive to be as good ‘CSR’ from companies can positively influence their purchase intentions through consumer ethics (Lee and Lee 2015).

If consumers have a higher purchase intention towards those brands they perceive as more ethical, then it is only logical that their purchase intention will be higher as well with those brands that carry out more socially- and environmentally-responsible behaviors.

As a result, the first hypothesis to be formulated will be:

Hypothesis 1: the better the brand’s environmentally- and socially-responsible behavior, the stronger the individual’s purchase intention

2.3. Habitual behavior

On the other hand, it has also been shown that consumers will have a higher purchase intention with a more familiar brand, or one with a bigger market share that is better known (Hanzaee and Andervazh 2012).

Habit, or habitual behavior, has been described in the literature as “that which has been done before and is done again; long-established patterns of behavior being repeated more or less automatically” and where “there is no deliberation and no choosing” (Kaas 1982). Habits are “learned sequences of acts that have become automatic responses to specific cues” (Fennis and Stroebe 2016: 257), but behavior can only become habitual if such actions are carried out regularly.

(13)

performance impairment (Marien et al. 2018). This so-called automaticity implies that habits are not always executed consciously – sometimes individuals even make associations between stimuli and responses unawarely, too (Marien et al. 2018), although individuals can still reflect on these patterns and make adjustments if needed according to implicit learning research.

Furthermore, research suggests that habit may be context-dependent, and that context cues such as physical location, or time of the day, can impact an individual’s habitual behavior. In this line, behaviors are more likely to be repeated when contexts are stable to the individual, or aligned with personal goals (Mazar and Wood 2018). However, despite certain processes oftentimes being automatic, the individual can suppress the habitual behavior by exerting self-control (Rebar et al. 2018).

Consumers usually purchase the same brands, the same amounts, and even at the same stores (Wood and Neal 2009), although one must not confuse habit with brand loyalty, which is a form of repeat purchasing behavior (Jacoby and Kyner 1973).

The reason why consumers work with habits is partly because by doing so, alternative responses are reduced (Wood and Neal 2009), which makes shopping easier and faster. Additionally, habits are formed and repeated because they are (or at least once were) rewarding to the individual in some way; or because they represent knowledge that the consumer gathered through experience (Wood and Neal 2009). In the case of the beauty industry, there are no standardized regulations that force manufacturers to advertise the shelf-life of products, but it is widely understood that certain products (especially those applied on and around the eye area, like mascara) have shorter shelf-lives than others due to potential eye infections (FDA 2018). With many products, the purchasing frequency is high, either because the consumer finishes the product or because the product is no longer safe to use. Many consumers will rotate their products and try something new, either from the same brand or from a completely different one, but force of habit is very strong with beauty products. In this line, many consumers shop consistently for the same beauty product because it “works for them”, or because it used to at a point in their lives and they refuse to switch to another product or brand, or because changing their beauty routine may entail some in-depth learning that they might not be willing to perform yet.

Research shows, moreover, that when under time constraints, distractions, or other self-control limitations, consumers will resort to acting by force of habit (Wood and Neal 2009).

(14)

Hypothesis 2: the more habitual the purchase is for the individual, the more it reduces the relationship between the brand’s environmentally- and socially-responsible behavior and the

individual’s purchase intention

However, it would not be right to assume that consumers repeatedly purchase the same product purely out of habit – simple preference, income level, values, or brand loyalty may directly impact the individual’s final purchase decision.

2.4. Income

Research has consistently suggested that consumer behavior and purchasing habits vary depending on individual characteristics (eg., age, gender, income, education). In fact, it all points out to the reality that household income does indeed affect consumers’ purchasing intentions – those individuals with higher levels of income spend more on healthier alternatives for food and take organic produce into consideration when shopping (Zhang et al. 2018).

When shopping for socially- and environmentally-responsible products, income is a relevant variable. A low-income consumer will probably not have the means to refuse to purchase certain brands regardless of the company’s stand on animal testing or veganism; whereas higher-income consumers might have a “pricier works better” mentality that can lead them to clinging to luxury brands that certainly carry several environmental and social issues along with them – perhaps because through this purchase the consumer is also buying status and prestige. It has been shown multiple times in research and theoretical models (Duesenberry 1949; Congleton 1989; Rauscher 1993) that income and status-seeking are linked, and that it is highly socially encouraged, too. In fact, as much as status is something sought by those consumers with higher income, many products – green or no – can be seen as prestigious. Today’s green consumer is a very complicated one, and for them green is a status symbol and a core belief (Sheehan 2013: 289), especially to those in urban areas.

(15)

consistently and positively linked to environmental sensitivity – at higher levels of income, the more individuals can afford the marginal costs associated with the support of environmentally- and socially-responsible causes (Straughan and Roberts 1999).

However, it is important to keep in mind that an individual’s attitude towards a product or brand – which basically describe a person’s estimations and feelings about a product’s quality (Kotler and Armstrong 2018, p. 173) – is what eventually determines whether they are more likely to purchase it or not, so one cannot solely predict an individual’s purchase intention based on their income, but it is undeniable that it plays a big part in what a person can and cannot afford and therefore which products they can access. Can this person put environmental problems at the top of their list when shopping for moisturizers, or are they on a strict budget and must they remain frugal and have price as their top priority?

Therefore, the next hypothesis will be as follows:

Hypothesis 3: the higher the income of the individual, the more it amplifies the relationship between the brand’s environmentally- and socially-responsible behavior and the individual’s

purchase intention FIGURE 1

(16)

3. METHODOLOGY

A survey was conducted and distributed online in order to obtain information about consumers’ habits and their purchase intentions when shopping for beauty products.

The questionnaire consisted of a total of twenty-six multiple choice, Likert scale questions, and it was redacted both in English and in Spanish, two of the most widely spoken languages in the world. Additionally, it is to be noted that Spanish is my mother tongue, and nothing was lost in translation when transposing the questions from one survey to another. This fact was also the reason why this language was selected as a secondary one.

Respondents started by answering a filter, dichotomous question where they were asked whether they used beauty products routinely in their daily lives; the survey stopped immediately after answering all demographic questions for those respondents who answered “no”. It is important to highlight the fact that everyone that took the survey read through a very short notice where it was explicitly explained that ‘beauty products’ do not entail solely makeup and that, as explained in this paper, the category includes multiple types of products that almost everyone uses every day. See Appendix 1 to see the survey questions.

Thereafter followed six very basic demographic questions regarding gender, age, income level, education, employment situation, and marital status. The ‘income’ question (Q4) would later be one of the predictor variables in the conceptual model.

(17)

one that was clearly unethical and unsustainable, but one that, on the other hand, had very affordable and efficient products that the respondent was supposed to have been using for years.

(18)

4. RESULTS

In total, 149 people responded to the survey, of which only 11.4% (17 people) failed to pass the filter question and is thus unusable. When analyzing the data pertaining to these individuals in depth, it was found that, in line with expectations since women tend to use more beauty products (or at least be more open about it), 70.58% of them were male. 58.82% were between 18 and 24 years old, so the majority of these cases were quite young – also in line with expectations, as one could argue that the younger the individual is, the less they could feel like they need beauty products of any sort.

The total number of usable responses were thus 131 (88.6% of the original sample), and the other seventeen cases were thus deleted and will not be appearing in further calculations.

Out of the usable responses, and contrary to expectations, the majority of them (54.2%) were male. The sample (N=131) was fairly young, with most of them being between 18 and 24 years old (32.1%), or even between 25 and 34 years old (29%). The greater percentage of the sample was at a lower range of income level, and most of them (45%) were full-time workers who had at least started a bachelors’ degree (42%).

TABLE 1

Frequency tables of cases

Frequency Percent

Valid (Gender) Male 71 54.2

Female 59 45

Non-Binary 1 0.8

Total 131 100

Valid (Age) Younger than 18 years old 2 1.5

18 - 24 years old 42 32.1

25 - 34 years old 38 29

35 - 44 years old 22 16.8

(19)

55 - 64 years old 12 9.2

Older than 65 years old 1 0.8

Total 131 100

Valid (Income) Less than €20,000 annually 56 56

Between €20,000 - €34,999 annually 41 31.3 Between €35,000 - €49,999 annually 20 15.3 Between €50,000 - €74,999 annually 9 6.9 Between €75,000 - €99,999 annually 3 2.3

More than €100,000 annually 2 1.5

Total 131 100

Valid (Education) Less than a high school degree 4 3.1

High school degree or equivalent 17 13

Technical school 17 13

Bachelor's degree 55 42

Master's degree 30 22.9

PhD 6 4.6

Prefer not to say 2 1.5

Total 131 100

Valid (Employment) Full-time worker 59 45

Part-time worker 12 9.2

Part-time worker and student 11 8.4

Student 34 26

Unemployed and student 1 0.8

Unemployed 7 5.3

Retired 6 4.6

(20)

Total 131 100

For the predictor variable ‘habitual behavior’ there was a total of nine scale variables, all of them measured on a five-point Likert scale that went from “strongly agree” to “strongly disagree”. Two of those variables (Q13 and Q15, which accounted for “attention to other beauty brands when shopping for beauty products” and “regularly changing beauty brands”, respectively) had to be recoded into reversed variables.

A reliability analysis was first performed on the variables to check for internal consistency amongst them. The test showed a Cronbach’s alpha value of .88, which indicates a high value of internal consistency for the scale within the specific sample, as well as an adequate level of inter-item reliability. Additionally, the analysis confirmed that this value would not improve whatsoever if any of the variables were to be removed from the sample.

TABLE 2.1

Reliability analysis for habitual behavior variables: Cronbach’s Alpha Reliability Statistics

Cronbach's Alpha

Cronbach's Alpha Based on

Standardized Items N of Items

(21)

TABLE 2.2

Reliability analysis for habitual behavior variables: Cronbach’s Alpha if Item Deleted Item-Total Statistics Scale Mean if Item Deleted Scale Varianc e if Item Deleted Corrected Item-Total Correlation Squared Multiple Correlation Cronbach's Alpha if Item Deleted

Q9 Reluctance to try new beauty products until finding out efficacy

20.97 57.507 .645 .489 .860

Q11 Avoidance of purchase unless they are on shopping list

20.66 56.194 .682 .534 .856

Q12 Automatically knowing which brands to buy

21.40 63.135 .455 .253 .875

Q13 Attention to other beauty brands when shopping

19.94 60.412 .523 .364 .870

Q14 Rarely buying from brands out of uncertainty of performance

21.23 55.701 .753 .598 .850

Q15 Regularly changing beauty brands 20.37 56.034 .658 .544 .858

Q16 Rarely switching brands if the individual really likes a beauty brand

21.09 60.653 .560 .439 .867

Q17 Usually purchasing only the beauty products strictly needed

20.90 55.675 .661 .649 .858

Q18 Carefully watching expenditure on beauty products

21.04 57.899 .596 .566 .864

(22)

Analysis methods available on SPSS. Immediately at first glance, the analysis found two underlying factors, the remaining seven thus considered ‘scree’.

TABLE 3.1

Factor analysis of habitual behavior variables: Eigenvalues Total Variance Explained

Factor

Initial Eigenvalues

Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumul ative % Total % of Variance 1 4.534 50.374 50.374 4.104 45.604 45.604 2.690 29.889 2 1.117 12.416 62.790 .735 8.166 53.770 2.149 23.881 3 .768 8.533 71.322 4 .722 8.022 79.345 5 .591 6.564 85.908 6 .391 4.349 90.258 7 .337 3.744 94.002 8 .314 3.484 97.486 9 .226 2.514 100.000

Extracting method: Principal Axis Factoring.

(23)

TABLE 3.2

Factor analysis of habitual behavior variables: Communalities Communalities

Initial Extraction

Q9 Reluctance to try new beauty products until finding out efficacy .489 .516

Q11 Avoidance of purchase unless they are on shopping list .534 .547

Q12 Automatically knowing which brands to buy .253 .280

Q13 Attention to other beauty brands when shopping .364 .335

Q14 Rarely buying from brands out of uncertainty of performance .598 .651

Q15 Regularly changing beauty brands .544 .616

Q16 Rarely switching brands if the individual really likes a beauty brand .439 .425

Q17 Usually purchasing only the beauty products strictly needed .649 .827

Q18 Carefully watching expenditure on beauty products .566 .643 Extracting method: Principal Axis Factoring.

(24)

TABLE 3.3

Factor analysis of habitual behavior variables: Rotated factor matrix Rotated Factor Matrixa

Factor

1 2

Q9 Reluctance to try new beauty products until finding out efficacy .651 Q11 Avoidance of purchase unless they are on shopping list .579

Q12 Automatically knowing which brands to buy .506 Q13 Attention to other beauty brands when shopping .530 Q14 Rarely buying from brands out of uncertainty of performance .664

Q15 Regularly changing beauty brands .754

Q16 Rarely switching brands if the individual really likes a beauty brand .620 Q17 Usually purchasing only the beauty products strictly needed .871

Q18 Carefully watching expenditure on beauty products .767

Extraction Method: Principal Axis Factoring.

Rotation Method: Varimax with Kaiser Normalization.a a. Rotation converged in 3 iterations.

When taking a look at the factors, it was obvious that Factor 1 had to do with consumers’ reluctance to purchase a product that they had never tried before, or to switch brands. Factor 1 was therefore named ‘uncertainty avoidance’. Factor 2, on the other hand, accounted for individuals’ spending, and their hesitance to buy something they had not accounted for before. Factor 2 was thus dubbed ‘budgeting’.

The nine different ‘habitual behavior’ dimensions were therefore reduced into two. Factor 1 (uncertainty avoidance): Q9, Q12, Q13, Q14, Q15, Q16

(25)

For the second predictor variable, ‘socially- and environmentally- responsible behavior of brands’, or simply ‘CSR’, the sample also had nine scale variables, measured on a five-point Likert scale as well. As with the first predictor variable, there were two of them that had to be recoded because the values were reversed; these variables were Q21 and Q30 (“advertisements from manufacturers presenting a true picture of products”, and the second fictional scenario, where the company “Beauty by Z” was being unsustainable, respectively).

A reliability analysis was performed on the variables before carrying out a factor analysis to reduce the number of variables in order to test for internal consistency. Cronbach’s alpha value was .827 which, again, proved that there was a high level of internal consistency for the scale within the specific sample. However, and as opposed to the previous predictor variable, this value did improve with the removal of one of the variables; if the variable Q21 (“advertisements from manufacturers presenting a true picture of products”) was removed, Cronbach’s alpha would go from .83 to .85. Granted, that is not too significant of an increase, however the ‘corrected item-total correlation’ value for that specific variable was at a very low .08. Cronbach’s alpha value does not increase with the removal of any other variables.

TABLE 4.1

Reliability analysis of CSR variables: Cronbach’s Alpha (pre-removal of Q21) Reliability Statistics

Cronbach's Alpha

Cronbach's Alpha Based on Standardized

Items N of Items

(26)

TABLE 4.2

Reliability analysis of CSR variables: Cronbach’s Alpha if Item Deleted (pre-removal of Q21) Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Squared Multiple Correlation Cronbach's Alpha if Item Deleted

Q19 Beauty product knowledge 16.36 35.924 .464 .491 .819

Q20 Use of beauty product knowledge to verify advertising claims by manufacturers and brands

16.84 35.213 .573 .570 .804

Q21 Advertisements from manufacturers presenting a true picture of products

17.16 42.920 .082 .143 .854

Q22 Thinking commercial advertising in beauty products should be forced to mention the ecological disadvantages of products

17.61 36.270 .681 .662 .796

Q23 Thinking excess packaging is one source of pollution that could be avoided if cosmetics manufacturers were more environmentally aware

17.98 38.161 .656 .584 .803

Q24 Worrying about problems associated with the lack of sustainability of beauty products

17.21 34.119 .640 .615 .796

Q25 Belief that all consumers should be interested in the environmental consequences of the cosmetic products they purchase

17.73 35.428 .765 .760 .787

Q29 Skincare X (ethical and sustainable but inconvenient and expensive for the consumer)

17.26 34.640 .593 .538 .802

Q30 Beauty by Z (unethical and unsustainable but convenient and affordable for the consumer)

17.09 35.222 .497 .336 .815

Cronbach’s Alpha=.827

(27)

Given the circumstances, a preliminary principal component analysis was performed on SPSS before deleting that variable altogether. The analysis resulted in two possible results: either the nine variables were reduced into three different factors, or into two. If the variables were reduced into three factors, the variable Q21 would only load on one of them, whilst the other two would carry the rest of the variables. On the other hand, if SPSS was forced into factoring the variables into two, then the variable Q21 would not load on either of them; its communality value would additionally sit at a meek .009.

TABLE 5.1

Factor analysis of CSR variables: Eigenvalues, option A (pre-removal of Q21) Total Variance Explained

Initial Eigenvalues Extraction Sums of Squared Loadings

Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance 4.249 47.206 47.206 4.249 47.206 47.206 3.610 40.109 1.213 13.476 60.683 1.213 13.476 60.683 1.828 20.316 1.160 12.886 73.569 1.160 12.886 73.569 1.183 13.143 .683 7.589 81.158 .608 6.758 87.916 .411 4.569 92.485 .274 3.039 95.524 .230 2.561 98.085 .172 1.915 100.000

(28)

TABLE 5.2

Factor analysis of CSR variables: Communalities, option A (pre-removal of Q21) Communalities

Initial Extraction

Q19 Beauty product knowledge 1.000 .871

1.000 .829

Q21 Advertisements from manufacturers presenting a true picture of products 1.000 .802

1.000 .751

1.000 .668

1.000 .709

1.000 .843

1.000 .613

1.000 .534

(29)

TABLE 5.3

Factor analysis of CSR variables: Rotated component matrix, option A (pre-removal of Q21) Rotated Component Matrixa

Component

1 2 3

Q19 Beauty product knowledge .911

.853

Q21 Advertisements from manufacturers presenting a true picture of products .894

.858

.745

.738

.906

.675

.623

Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.a

(30)

OPTION A: Q21’s communality=.8

Number of factors (“components”)=3 Factor 1: Q22, Q23, Q24, Q25, Q29, Q30

Factor 2: Q19, Q20 Factor 3: Q21

TABLE 5.4

Factor analysis of CSR variables: Eigenvalues, option B (pre-removal of Q21) Total Variance Explained

Initial Eigenvalues

Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance 4.249 47.206 47.206 3.882 43.138 43.138 3.082 34.250 1.213 13.476 60.683 .911 10.127 53.265 1.711 19.015 1.160 12.886 73.569 .683 7.589 81.158 .608 6.758 87.916 .411 4.569 92.485 .274 3.039 95.524 .230 2.561 98.085 .172 1.915 100.000

(31)

TABLE 5.5

Factor analysis of CSR variables: Communalities, option B (pre-removal of Q21) Communalities

Initial Extraction

Q19 Beauty product knowledge .491 .488

.570 .906

Q21 Advertisements from manufacturers presenting a true picture of products .143 .009

.662 .685

.584 .531

.615 .535

.760 .900

.538 .435

.336 .305

(32)

TABLE 5.6

Factor analysis of CSR variables: Rotated factor matrix, option B (pre-removal of Q21) Rotated Factor Matrixa

Factor

1 2

.924

Q21 Advertisements from manufacturers presenting a true picture of products

.800

.687

Q24 Worrying about problems associated with the lack of sustainability of beauty products .640 Q25 Belief that all consumers should be interested in the environmental consequences of the

cosmetic products they purchase

.932

Q29 Skincare X (ethical and sustainable but inconvenient and expensive for the consumer) .573 Q30 Beauty by Z (unethical and unsustainable but convenient and affordable for the

consumer)

.529

Rotation Method: Varimax with Kaiser Normalization.a

(33)

OPTION B: Q21’s communality=.009 Number of factors=2 Factor 1: Q22, Q23, Q24, Q25, Q26, Q27, Q29, Q30 Factor 3: Q19, Q20 Variables missing: Q21

The variable Q21 was thus excluded from the sample, seeing as it was not reliable and it did not have any kind of consistency with the rest of the variables.

A new reliability analysis proved that Cronbach’s alpha value had indeed increased to .85; the value would now only seem to increase with the removal of variable Q19 (“beauty product knowledge”), but this increase would be insignificant (from .854 to .856) and the variable’s corrected item-total correlation was still higher than .40 (.45) so the removal of this variable was deemed unnecessary.

TABLE 6.1

Reliability analysis of CSR variables: Cronbach’s Alpha Reliability Statistics

Cronbach's Alpha

Cronbach's Alpha Based on

Standardized Items N of Items

.854 .869 8

(34)

TABLE 6.2

Reliability analysis of CSR variables: Cronbach’s Alpha if Item Deleted Item-Total Statistics Scale Mean if Item Deleted Scale Variance if Item Deleted Corrected Item-Total Correlation Squared Multiple Correlation Cronbach's Alpha if Item Deleted

Q19 Beauty product knowledge 14.11 34.102 .451 .471 .856

14.60 32.935 .597 .563 .836

15.37 34.095 .699 .656 .827

15.73 36.213 .645 .582 .837

14.96 31.791 .671 .614 .827

15.49 33.467 .764 .752 .821

15.02 32.261 .626 .522 .832

(35)

An exploratory factor analysis was then performed on the remaining eight ‘CSR’ variables through the principal axis factoring method. Q19’s communality value sat at an acceptable .52, and the variable loaded certainly heavily onto Factor 2 at a value of .69, so the removal of this variable was ratified as irrelevant.

TABLE 7.1

Factor analysis of CSR variables: Eigenvalues Total Variance Explained

Factor

Initial Eigenvalues

Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings Total % of Variance Cumulative % Total % of Variance Cumulative % Total % of Variance 1 4.240 53.003 53.003 3.873 48.412 48.412 3.174 39.674 2 1.204 15.056 68.059 .900 11.250 59.662 1.599 19.988 3 .825 10.312 78.371 4 .609 7.611 85.982 5 .423 5.287 91.269 6 .284 3.551 94.821 7 .238 2.971 97.792 8 .177 2.208 100.000

(36)

TABLE 7.2

Factor analysis of CSR variables: Communalities Communalities

Initial Extraction

Q19 Beauty product knowledge .471 .516

.563 .868

.656 .691

.582 .523

.614 .542

.752 .894

.522 .440

.322 .299

(37)

TABLE 7.3

Factor analysis of CSR variables: Rotated factor matrix Rotated Factor Matrixa

Factor

1 2

.892

.812

.688

Q24 Worrying about problems associated with the lack of sustainability of beauty products .662 Q25 Belief that all consumers should be interested in the environmental consequences of the

cosmetic products they purchase

.935

Q29 Skincare X (ethical and sustainable but inconvenient and expensive for the consumer) .592 Q30 Beauty by Z (unethical and unsustainable but convenient and affordable for the

consumer)_R

.527

Rotation Method: Varimax with Kaiser Normalization.a a. Rotation converged in 3 iterations.

After removing Q21 from the sample, the variables were easily reduced into two factors. It became apparent that variables Q22, Q23, Q24, Q25, Q29, and Q30 loaded on Factor 1; whilst variables Q19 and Q20 loaded on Factor 2. Again, no cross-loadings were found on the rotated factor matrix. The factor loading threshold remained at .50 since the sample size did not change.

(38)

Factor 1 (Ethics): Q22, Q23, Q24, Q25, Q29, Q30 Factor 2 (Beauty knowledge): Q19, Q20

The last predictor variable was ‘income’, which came directly from one of the very first demographic questions thrown at the respondents. No factor analysis was needed, since this was simply one variable.

Lastly, the outcome variable came from the two variables that measured individuals’ willingness to modify their purchasing habits or behavior -- either by spending more to promote sustainability (Q26), or by no longer supporting certain brands that are not sustainable (Q27). They were computed into one same variable since the scaling system was the same.

(39)

TABLE 8.1

Correlations

Uncertainty

Avoidance Budgeting Ethics

Beauty Knowledge

Uncertainty Avoidance

Pearson Correlation 1 .614** _-.202* _-.227**

Budgeting Pearson Correlation .614** 1 -.192* -.437**

Ethics Pearson Correlation -.202* _-.192* ₁ _.454**

Beauty Knowledge Pearson Correlation -.227** -.437** .454** 1

Income Pearson Correlation -.161 .018 .129 -.055

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

TABLE 8.2

Correlations (Cont.)

Correlations

Income

Uncertainty Avoidance Pearson Correlation -.161

Budgeting Pearson Correlation .018

Ethics Pearson Correlation .129

Beauty Knowledge Pearson Correlation -.055

Income Pearson Correlation 1

(40)

the model, the two factors were from this point onwards jointed into the variable ‘Habitual Behavior’.

TABLE 8.3

New correlations

Correlations

Ethics Beauty Knowledge Income

Habitual Behavior

Ethics Pearson Correlation 1 .454** .129 -.218*

Beauty Knowledge

Pearson Correlation .454** ₁ _-.055 _-.384**

Income Pearson Correlation .129 -.055 1 -.067

Habitual Behavior

Pearson Correlation -.218* _-.384** _-.067 ₁

As can be seen in Table 8.3, there were no further issues with multicollinearity among the predictor variables.

A multiple regression analysis was conducted next in order to check for the efficacy of the model. The analysis consisted of a dependent variable called ‘Purchase Intention’, and five predictor variables, ‘Income’, ‘Uncertainty Avoidance’, ‘Budgeting’, ‘Ethics’, and ‘Beauty Knowledge’. The last four aforementioned variables were the ones derived from previous factor analyses.

(41)

TABLE 9.1 Regression analysis Model Summaryb Model R R Square Adj. R Square Std. Error of the Estimate Change Statistics R Square Change F Change df1 df2 Sig. F Change 1 .928a _.861 _.856 _.414 _.861 _194.906 ₄ ₁₂₆ _.000

a. Predictors: (Constant), Habitual Behavior, Income, Ethics, Beauty Knowledge b. Dependent Variable: Purchase Intention

TABLE 9.2 Regression analysis ANOVAa Model Sum of Squares df Mean Square F Sig. 1 Regression 133.806 4 33.452 194.906 .000b Residual 21.625 126 .172 Total 155.431 130

a. Dependent Variable: Purchase Intention

(42)

TABLE 9.3 Regression analysis Coefficientsa Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) -.253 .192 -1.322 .188 Ethics 1.156 .047 .937 24.730 .000 Income .030 .032 .032 .935 .352 Beauty Knowledge -.024 .037 -.026 -.658 .512 Habitual Behavior .012 .040 .011 .301 .764

a. Dependent Variable: Purchase Intention

The adjusted R2 value indicated that the model could predict up to 85.6% of the variance

in the outcome variable at a very significant level (p<0.1) (Table 9.1).

F(4,126)=194.906, p<01 (Table 9.2) (Ethics) b=1.16, t(126)=24.730, p<0.01 (Income) b=.03, t(126)=.935, p=.35 (Beauty Knowledge) b= -.02, t(126)= -.66, p=.51

(Habitual Behavior) b=.01, t(126)=.3, p=.76

Ethics appeared to be the only variable to be able to predict any variance in purchase intention at a significance level, whilst the others did not manage to do so (Table 9.3).

(43)

TABLE 10.1

Moderated regression analysis: Model Summary

R R2 _MSE _F _df1 _df2 _p

.9303 .8654 .1687 132.8478 6.0000 124.0000 .0000

TABLE 10.2

Moderated regression analysis: Model

Coeff se t p LLCI ULCI

constant .0018 .3104 .0057 .9955 -.6126 .6161 Ethics 1.0241 .1311 7.8112 .0000 .7646 1.2836 Habitual Behavior -.1336 .0842 -1.5863 .1152 -.3002 .0331 Int_1 .0835 .0420 1.9861 .0492 .0003 .1667 Income .0600 .0816 .7358 .4632 -.1014 .2215 Int_2 -.0166 .0352 -.4725 .6374 -.0863 .0530 Beauty Knowledge -.0315 .0371 -.8481 .3980 -.1049 .0420 TABLE 10.3

Moderated regression analysis: Interactions

Product terms key:

Int_1 : Ethics x Habitual Behavior Int_2 : Ethics x Income

Test(s) of highest order unconditional interaction(s):

R2-chng F df1 df2 p

X*W .0043 3.9445 1.0000 124.0000 .0492

X*Z .0002 .2232 1.0000 124.0000 .6374

(44)

TABLE 10.4

Moderated regression analysis: Effects

Focal predict: Ethics (X)

Mod var: Habitual Behavior (W) Mod var: Income (Z)

Conditional effects of the focal predictor at values of the moderator(s):

H.B. Income Effect se t p LLCI ULCI

1.6142 1.0000 1.1423 .0721 15.8385 .0000 .9995 1.2850 1.6142 1.9924 1.1258 .0529 21.2695 .0000 1.0210 1.2305 1.6142 3.1392 1.1067 .0554 19.9920 .0000 .9971 1.2163 2.5999 1.0000 1.2246 .0733 16.7155 .0000 1.0796 1.3696 2.5999 1.9924 1.2081 .0542 22.2734 .0000 1.1007 1.3154 2.5999 3.1392 1.1890 .0564 21.0999 .0000 1.0775 1.3005 3.5856 1.0000 1.3069 .0947 13.8008 .0000 1.1194 1.4943 3.5856 1.9924 1.2904 .0807 15.9846 .0000 1.1306 1.4501 3.5856 3.1392 1.2713 .0820 15.5072 .0000 1.1090 1.4336

In this model, Moderator W is ‘Habitual Behavior’ (which in Table 10.3 has been shortened to ‘H.B.’ for the sake of clarity), whilst Moderator Z is ‘Income’. The two predictors were the variables ‘Ethics’ and ‘Beauty Knowledge’, the latter in the form of a covariate; the outcome variable remains ‘Purchase Intention’.

The model summary (Table 10.1) suggested that all the predictors are useful at predicting the outcome variable. 86.54% of variance in consumers’ purchase intention can be attributed to the predictors in the model.

R2=.86

(45)

Out of all the predictors, it seems that, in line with the previous regression analysis, only ‘Ethics’ predicted overall purchase intention (Table 10.2). Neither ‘Income’, ‘Habitual Behavior’, nor ‘Beauty Knowledge’ predict overall purchase intention.

(Ethics) b=1.02, t(124)=7.81, p<.01 (Habitual Behavior) b= -.13, t(124)= -1.59, p=.11

(Income) b=.06, t(124)=.74, p=.46 (Beauty Knowledge) b= -.03, t(124)= -.85, p=.4

As for the interactions between the predictors and the moderators (Int_1 and Int_2, Table

10.3), the addition of either of them does not result in a great change in R2. However, the second

interaction is significant at two levels (α=.10 and α=.05) whilst the first one is not statistically significant.

(Int_1) F(1,124)=3.94, p=.049; δR2_=.005

(Int_2) F(1,124)=.22, p=.64; δR2=.0002

Table 10.4 showed that at low levels of habitual behavior, increasing income levels of consumers led to ‘CSR’ predicting purchasing intentions of consumers. What can also be observed on this table is that at higher levels of habitual behavior, the effect seemed to increase as well, leading to a stronger predicting behavior from ‘CSR’ on the outcome variable. It should be noted that the effect that habitual behavior has on the model is negative.

4.1. Hypotheses testing

(46)

and consumers’ purchase intentions. The positive, significant (p<.01) effect of ‘Ethics’ on ‘Purchase Intention’ on the model (Tables 9.3 and 10.2) confirms Hypotheses 1a and 1b. However, it should be noted that ‘Beauty Knowledge’ does not seem to adjust this relationship (Table 10.2), nor is its influence significant to the model (table 9.3). Being knowledgeable about product qualities or manufacturers’ claims does not affect consumers’ purchase intention in a statistically significant way (p>.10).

Hypothesis 1 implies that the better the social and environmental behavior of a brand or company, the stronger the individual’s purchase intention will be towards that specific brand. The fact that this hypothesis was confirmed suggests that consumers do indeed care for social and environmental issues and would in fact alter their purchasing habits in order to align with their core beliefs.

Hypothesis 1: the better the brand’s environmentally- and socially-responsible behavior, the stronger the individual’s purchase intention

I additionally hypothesized that when purchases become habitual–or even automatic–for the consumer, these serve as some sort of leverage for those brands that do not behave in such ‘ethical’ or sustainable ways. It is indeed hard to change habits; consumers tend to shop the same amounts, at the same shop (Wood and Neal 2009), and that kind of automaticity saves individuals time when shopping for any type of product. In the case of beauty products, repeated purchases often happen due to fear of other brands being subpar, or simply because a product ‘works for you’. Habitual behavior is deeply ingrained in beauty products.

Hypothesis 2 speculated that the more habitual a purchase was to the individual, the more the relationship between brands’ socially- and environmentally-responsible behavior and consumers’ purchase intention would be reduced.

The moderating effect habitual behavior was expected to have on this relationship was therefore negative, which was proven in Table 10.4. On its own, habitual behavior does not influence purchasing behavior (p>.10) (Table 9.3), but it does interact with ‘Ethics’ significantly (Table 10.3) and it does moderate in a negative and statistically significant way (Table 10.4) the relationship between ‘CSR’ and consumers’ purchase intention.

(47)

Hypothesis 2: the more habitual the purchase is for the individual, the more it reduces the relationship between the brand’s environmentally- and socially-responsible behavior and the

individual’s purchase intention

Last but not least, I argued that the higher the income levels of the individual, the more the relationship between the brands’ socially- and environmentally-responsible behavior and the individual's purchase intention would be amplified; whereas the lower the income, the more this relationship would be reduced. Research has suggested that those with greater purchasing power tend to spend more on healthier and organic alternatives (Zhang et al. 2018), and income and status are highly correlated (Duesenberry 1949). Oftentimes the most sustainable beauty products are more expensive because it takes longer and more effortful processes to manufacture, so those individuals at lower levels of income would not be able to afford those kinds of products.

As happened with the previous hypothesis, ‘Income’ on its own did not serve its purpose to predict any variance in consumers’ purchasing habits (p>.10) (Table 9.3); however, this one did not interact significantly with the predictor, either (p>.10) (Table 10.3). It did seem to have a moderating effect on the variables (Table 10.4), but since its interaction with the predictor was not significant, Hypothesis 3 is rejected.

Contrary to expectations, it seems that income does not influence the relationship between brands’ socially- and environmentally-responsible behaviors and consumers’ purchase intentions. This might be because individuals’ beliefs and morals are put before economic needs, but further research would need to be done in this area to confirm this.

Hypothesis 3: the higher the income of the individual, the more it amplifies the relationship between the brand’s environmentally- and socially-responsible behavior and the individual’s

purchase intention

5. CONCLUSIONS

(48)

In line with the results, it can be said that consumers’ desire for social and environmental responsibility from beauty companies is far from hypocritical; it seems like individuals are striving for more sustainable ways of shopping their favorite products and are actively trying to move towards greener alternatives. Whether this means spending more money on certain products or removing certain companies or brands from their shopping lists, it appears that consumers are in fact more than willing to go the extra mile to be sustainable in their shopping for beauty products.

While consumers’ habits influence this desire, their income level appears to have no moderating effect on this relationship. Perhaps people’s eagerness to become more environmentally-friendly surpasses their budget because individuals’ belief systems and morals tend to be deeply ingrained within their personalities; perhaps the world’s increasingly deteriorating state at an alarming rate is eventually setting into people’s minds in the shape of fear. Or perhaps it is simply that consumers do not mind spending more on beauty products if it means spending it on better and more sustainable alternatives while they save money on other types of products.

In essence, it seems like despite being one of the fastest-moving industries,–and certainly one of the longest-standing, and one of the ones that make the most money–the beauty industry average consumer is finally moving towards a ‘greener’, more sustainable movement; one that could perhaps make up for the excessive consumerism that has reigned in the business during the past couple of decades. It is yet to be seen if this change in consumers’ purchasing intentions will result in a definite change for the better in the bigger companies in the game.

5.1. Managerial implications

(49)

However, it is not a good idea for brands to try and compete against well-established companies that already mean a habitual purchase for consumers with the same type of product. This paper finds habitual purchases negatively moderate the relationship between brands’ socially- and environmentally-responsible behavior and consumers’ purchase intention. Since there is a strong, positive relationship between brands’ ‘CSR’ and individuals’ purchase intention, it would best be suggested to come up with a more sustainable, cleaner, greener alternative. This will have a better effect on consumers than simply competing against their go-to products.

It looks like this shift in consumer demand towards greener and more sustainable alternatives is not going away – the ‘green’ consumer was born in the 1990s (Kumar 2005), after all, and this paper has shown consumers still care, almost more than thirty years later. Beauty companies should adapt to this trend in the industry and respond to customers’ needs to find sustainable beauty products to use, especially in a technological era when most people have easy access to information online from every brand they shop from.

5.2. Limitations and implications for future research

The limitations of this study firstly come from the methodology – the sample is fairly

small (N=131), and while it is diverse in terms of gender, it is also quite a young sample. Most of the respondents were between 18 and 34 years old (61.1%, in total), while the rest were spread out across other age ranges.

Additionally, yet another limitation of this study has to do with respondents’ ability or willingness to read through the whole survey. Proof of this is that there was a disclaimer before the filter question (see Appendix 1) where participants were warned that ‘beauty products’ did not solely mean makeup products and they were given examples of other typical beauty products that

most people use on a daily basis – yet 17 people still answered no, arguably because most people

are in a rush when filling out online surveys. This is, however, an issue with every survey, and not just this study in particular.

(50)

(51)

REFERENCES

Ajzen, I. 1991. The Theory of Planned Behavior. Organizational Behavior and Human Decision

Processes, 50(2): 179–211.

Ajzen, I. 2011. The theory of planned behaviour: Reactions and reflections. Psychology & Health, 26(9): 1113–1127.

Bloemer, J. M.M., & Kasper, H. D. 1995. The complex relationship between consumer satisfaction and brand loyalty. Journal of Economic Psychology, 16(2): 311–329.

Carroll, A. B. 1999. Corporate Social Responsibility. Business & Society, 38(3): 268–295. Congleton, R. D. 1989. Efficient status seeking: Externalities, and the evolution of status games.

Journal of Economic Behavior & Organization, 11(2): 175–190.

Duesenberry, J. S. 1949. Income, Saving, and the Theory of Consumer Behavior. Southern

Economic Journal, 16(4): 485.

Eastman, J., Iyer, R., & Thomas, S. 2013. The Impact of Status Consumption on Shopping Styles: an Exploratory Look at the Millennial Generation. Marketing Management Journal, 23(1): 57– 73.

Elster, J. 1989. Social Norms and Economic Theory. Journal of Economic Perspectives, 3(4): 99– 117.

Master of Science in International Business and Management “Sustainability in the Beauty Industry and its Effect on Consumers’ Purchase Intention”