Good Greed
Incentivising Social Support via Everyday Consumption
Name: Jessica Exton Student Number: 11373806
Course: Masters of Economics, Behavioural Economics and Game Theory Academic Year: 2016/2017
Introduction ... 7
Literature Review ... 11
Altruism ... 11
Warm Glow ... 13
Impure Altruism ... 15
The Context of Social Business ... 20
Study One... 28
Methodology ... 28
Experimental Design Procedures ... 29
Hypotheses Formulation ... 32
Results ... 33
Study Two ... 38
Methodology ... 38
Experimental Design Procedures ... 39
Hypothesis Formulation ... 43
Results ... 44
Discussion ... 59
Recommendations for Future Research ... 62
Conclusion ... 67
References ... 68
Appendix ... 77
List of Figures Figure 1 Depiction of a Pro-Social Consumption Decision ... 18
Figure 5 Study One - Overview of Data ... 35
Figure 6 Study One - Daily Store Views for Social and For-Profit Stores ... 36
Figure 7 Study One – Test of the Significance of the Difference between the Social
and For-Profit Stores, Regression Analysis ... 37
Figure 8 Study Two – Summary Statistics ... 45
Figure 9 Study Two – Overview of Data ... 47
Figure 10 Study Two – Average Change in Willingness to Pay Given Introduction of
Social Impact ... 48
Figure 11 Study Two - Willingness to Pay Analysis between Questionnaire A and
Questionnaire B ... 49
Figure 12 Study Two – Subject Willingness to Pay Increase, Bar Graph,
Questionnaire A ... 51
Figure 13 Study Two – Equality Test of Value Before and After the Introduction of
Social Impact, Signrank Test, Questionnaire A ... 52
Figure 14 Study Two – Significance Test of Increase in Willingness to Pay given
Introduction of Social Impact, Welch Unequal Variances T-Test, Questionnaire A ... 53
Figure 15 Study Two – Subject Willingness to Pay Increase, Bar Graph,
Questionnaire B ... 54
Figure 16 Study Two – Equality Test of Value Before and After the Introduction of
Social Impact, Signrank Test, Questionnaire B ... 55
Figure 17 Study Two – Significance Test of Increase in Willingness to Pay given
Introduction of Social Impact, Welch Unequal Variances T-Test, Questionnaire B ... 56
Figure 18 Study Two, Test of Significant Difference given the Presence of Altruism,
Questionnaire A and Questionnaire B ... 57
Figure 19 Study Two – Aggregate Change in Value Attribution to Specific Altruistic
and Warm Glow Motivation ... 58
Figure 21 Study Two - Images of the Online Store Headers ... 77
Table 4 Study Two - Questionnaire Content, Data Collected ... 87
Table 5 Study Two - Questionnaire Content, Data Variables ... 88
Table 5 Study Two - Questionnaire Content, Data Categorisation ... 89
Statement of originality
This document is written by Jessica Exton (UvA student) who declares to take full responsibility for the contents. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
6
Abstract
The sum of reported contributions to charities and social cause is large enough to significantly address costly public good challenges such as international development or environmental sustainability. Innovative methods for encouraging individual giving are therefore continually being developed. One such innovation is the growing market for social goods; products which combine the private benefit of physical consumption with contributions to social outcomes.
This paper contributes to our understanding of consumer motivations within the context of social good consumption by investigating how consumers respond to the idea of social impact within the purchase environment. Given the consumer is an impure altruist and motivated by a combination of altruistic concern for others and an internal warm glow benefit which stems from completing an act considered to be moral, the relative influence of altruism and warm glow on consumption decision-making is also analysed. Findings indicate that social impact alone does not change levels of consumer interest and that internal warm glow benefits are more influential as a motivation for acting pro-socially, compared to an altruistic concern for helping others.
7
Introduction
This paper is broken into distinct sections. Initially I will provide an introduction by relaying the motivation of this research and intended contributions via the specified research questions. I will then detail a review of the literature on altruism, warm glow and impure altruism, before applying these insights within the context of social business. Next, I will provide the methodology, research design, hypothesis and results for two separate studies concerned with consumer motivation in the social business environment. Finally I will discuss the applicability of results and possible study limitations, before the key conclusions. The introduction is provided below.
Motivation for Research
It has been reported that achieving global access to clean water will cost $28.4 billion, meeting the recently signed sustainable development goals will require $1.4 trillion and eradicating world hunger by 2030 will take $160 per year for each of the 1.3 billion living in extreme poverty (D’Urso 2015, Hutton & Varughese 2016, Anderson 2015). Due to the substantial costs associated with social challenges such as these, finding efficient sources of funding is a topical challenge.
In a globally connected world, individuals contribute to socially beneficial outcomes both locally and internationally, with access to and understanding of social challenges that were previously inaccessible (Jackson 2014). In 2015-16 it was estimated that 80.8% of the Australian adult population gave $12.5 billion to charities and not-for-profit organisations, 2.1% of the US $17.95 trillion total GDP was made up of charitable giving, and within the European Union, individual giving totalled almost €25 billion (Philanthropy Australia 2016, Giving USA 2016, Fondation de France 2015). With individual contributions to social causes so significant at the aggregate level, innovative approaches to sourcing individual donations are increasingly being developed and tested within the market. Social business models, such as the allocation of a percentage of profits to support social cause or ensuring that each sale results in the donation of a needed item to a given cause, are such innovations.
8 Social business
Social businesses typically apply a standard operating model within the retail market but are led by a mission consistent with public benefit, and trade to fulfil this mission by reinvesting their profits in pro-social outcomes (Social Traders 2016). Within the context of social business, the term pro-social can be used with reference to any positive outcome that influences groups external to the consumer. To name only a few examples, pro-social outcomes might include environmental sustainability, access to education for girls in India, or raising funds for medical research.
Many businesses in a range of industries are incorporating efforts to support social cause by adopting socially conscious business practice linked to consumer purchase (Strahilevitz 1999). Within America, the Gap, Motorola, Apple, Hallmark and a number of other influential firms participate in a campaign which sells merchandise to support relief from disease in Africa, American Express links its credit card offers to social outcomes and Ben and Jerry’s Ice Cream supports a variety of environmental and social causes (Elfenbein & McManus 2010). In Australia, “TOM’s” has provided 35 million pairs of shoes to children in need over the past 10 years through its one-for-one social operating model (Naeini et al. 2015) and another such social enterprise called “Thankyou.” uses 100 percent of its profits from the sale of bottled water and select supermarket products to support social outcomes (Thankyou. 2017). While the prevalence of social businesses practice continues to grow, the market’s ability to conduct effective fundraising increases.
Regardless of sector or specialisation, the common feature of social businesses is the delivery of social goods; products which combine the private benefit of physical consumption with social outcomes which help others, linking everyday purchase decisions to social support (Elfenbein & McManus 2010). When selecting a social good, consumers benefit not only from their purchase, but also from the social business’ mission and the knowledge that they contributed to that social intent. Individuals report greater happiness when spending on others rather than themselves and also indicate they favour retailers or brands associated with a social outcome (Dunn, Aknin & Norton 2008, Meyer 1999). This channel therefore has the potential to unlock efficient funding sources for social benefit.
9 The global success and growth of the social business model appears to stem from both the improved social outcomes which motivate altruistic behaviour and from an individual preference for contributing to pro-social activity. Previous research into the motivation behind contributing to social cause identifies a distinct difference between the external motivation to achieve a positive social outcome which benefits others, known as “altruism”, and the internal incentive of feeling good about having completed an action considered to be moral, termed “warm glow” (Andreoni 1989). The majority of individuals are known as “impure altruists” and are motivated by a combination of both altruism and warm glow (Andreoni, Harbaugh & Vesterlund 2007). For example, in purchasing a consumer item such as environmentally friendly washing powder, customers demonstrate that they care both for how they feel about having contributed to the cause, and for the social benefits of environmental preservation. An understanding of which motivator is more important to consumers is essential to the success of social business. Given that altruistic and warm glow motivations are observed within the social business environment, these influence not only how individuals value social outcomes but also how much they are willing to pay for a consumption item in order to indirectly support positive social impact.
Contributions from this Paper
Despite extensive research into impure altruism, the understanding of how it applies to social business remains incomplete. Without a comprehensive understanding of the ratio of intrinsic warm glow to external altruistic motivation (Lilley & Slonim 2014), efficient use of business as a channel for social support is difficult. An improved understanding of the relative importance of altruism and warm glow within the social business environment may enable improved operational design and a more effective way to combine social and personal benefits in a positive customer experience.
This paper will investigate two distinct ideas within the context of social business. Firstly, it will compare an online social business with a for-profit replicate, in order to ascertain whether the presence of social impact causes a change in consumer interest. This will be conducted within an online sales platform with experimental control over the design, delivery
10 and messaging relating to social impact. Secondly, this paper will run a between-participants experiment to test the respective influences of warm glow and altruistic motivations on a consumer’s willingness to pay for an item. Findings will demonstrate whether the introduction of social impact causes a significant change in the level of consumer interest in a given product range, and will also provide an indication of whether altruism or warm glow dominates the consumer decision-making process. As the social business market continues to grow, this research will enhance our understanding of consumer motivations, allowing business to model their social offering in a way that emphasises the outcomes that the consumer considers most important.
Research questions
To gain a deeper understanding of the motivations behind social good consumption, this paper addresses two research questions.
Research question one: Does the potential for social impact influence levels of market interest in a given range of consumer products?
Research question two: How does willingness to pay for a non-essential consumer item with social impact differ between consumers driven by impure altruistic motivation and those driven by pure warm glow?
Set up structure of thesis
In the following section I will present an extensive account of past research on altruistic and warm glow motivation, before using a simple model of impure altruism to represent the impact of pro-social behaviour within the social good consumption setting. Subsequently, I will provide the methodology, research design, hypothesis and results for two separate studies which will address the two research questions, by investigating altruistic and warm glow motivations within the context of a single social business. Finally I will provide a discussion about the applicability of results and the key conclusions.
11
Literature Review
In the following section, I will explore the existing literature concerning altruism, warm glow and impure altruism, in order to provide a background understanding of what motivates pro-social activity. I will then utilise a simple model of impure altruism to explore the current literature relating to consumer decision-making within the context of social business, before addressing the intended contributions of this paper.
Altruism
Altruism can be defined as the motivation to exert personal energy for the benefit of others and is seen across all human cultures, locations and races (Batson 2010). Pure altruists are solely motivated to act in the interests of others while impure altruists are incentivised by both the external benefits experienced by those other than themselves as well as by an internal or personal outcome (Andreoni 1989). While some arguments suggest that altruism may be cultural or acquired during psychological development and socialisation as a child (Roth et al. 1991, Henrich et al. 2001), the frequency and consistency with which altruistic behaviour is demonstrated, suggests that this is not simply a reflection of learning, wealth or capacity, but in fact that people have a taste for it (Strahilevitz 1999, Bernheim 1986). Socially beneficial behaviour includes not only the act of sharing one’s wealth of money or time with a group or individual but also taking actions or making consumption choices that deliver secondary benefits to others. Social benefit can describe any positive outcome that influences a small or large group of individuals and within the context of business, might include improved environment conditions resulting from sustainable production practices, enriched livelihoods due to fair trade production or reduced inequality via the delivery of a percentage of each sale value to support a specified cause.
To develop an understanding of why people practice socially beneficial behaviour, researchers have studied this activity across multiple environments and subjects. To name a few, studies have been conducted on the choice of efficient energy consumption (Menges, Schroeder & Traub 2005), recycling practices (Kinnaman 2006), selecting between donations of time or money (Lilley & Slonim 2014, Andreoni, Gale & Scholtz 1996), socially
12 beneficial consumption under social and private environments (Ariely, Bracha & Meer 2007), investment with varying impact scope (Hsee & Rottenstreich 2004) and how the motivations behind giving blood change with frequency (Ferguson et al. 2012). Research into altruism stems from the desire to understand what motivates people to disregard classic utility maximisation theory in order to exert their own energies for the gain of others. Arguably the simplest explanation for why this occurs is that individuals act as pure altruists and gain a benefit from assisting others via another’s improved wellbeing (Strahilevitz 1999). This approach is consistent with the notion that altruism involves practicing pro-social activities solely because one values the beneficial outcome experienced by someone else.
One approach to understanding giving by pure altruists is to see giving as equivalent to the purchase of a unit of social benefit, similar to any other commodity (Vesterlund 2006). The value of this unit of social benefit differs between individuals, who continue to consume until they receive no additional gain from an extra unit obtained (Vesterlund 2006). When all individuals purchase at a level where they receive maximum satisfaction from their consumption and they have no desire to purchase any more or less, there is an efficient market distribution of the social benefit commodity.
Bergstrom et al (1986), Cornes & Sandler (1984), and Warr (1982) were some of the first to use this approach to study and model pure altruism in an attempt to explain what motivates individuals to act pro-socially and how this impacts the environment. A purely altruistic individual will never give to a social cause if they are unable to confirm that their personal contribution will directly impact the total contribution, as in such a case, they would receive no benefit (Vesterlund 2006). Altruistic individuals therefore demand to understand the return on their social investment, highlighting that the assured satisfaction of the donor, as a result of social impact, is fundamental to their involvement.
For pure altruists, government spending or any additional contribution to the total pool of social benefit has a crowding out effect, meaning that in response to an increase in the total contribution to social benefit, pure altruists will reduce their personal donations to restore the previous aggregate level (Bergstrom et al 1986). For example, given increased government
13 funding for the construction of buildings to house the homeless, an individual who is purely motivated by altruism would stop or reduce their contributions to local services who deliver temporary housing, given that this service can now be fulfilled by governmental efforts. When individuals only care about the overall level of social impact as in this example, if the total donation amount increases above the efficient and optimum amount, they will reduce their individual contributions to achieve market efficiency (Abrams & Schitz 1978). The pure altruism model predicts an efficient allocation of social benefit from individuals who value it most (Lilley & Slonim 2014). The more agents are motivated by altruism the more efficiently they give and the more likely they will reduce their individual donations if the total pool increases.
While altruism is proven to motivate socially beneficial activity, in reality giving is more complex than the purely altruistic model because people are not solely motivated by the social outcome of their actions (Crumpler & Grossman 2008). Research shows that while altruism exists and plays a significant role in decision-making for many people, motivations for giving may actually be partially ‘strategic or selfish’ (Becker 1976). While individuals are motivated by a personal as well as altruistic benefit, alternative theories, such as that of warm glow, more accurately represent why people act pro-socially.
Warm Glow
Individuals consistently reveal that they are internally motivated to act in the interests of others, demonstrating that they receive a personal benefit from acting pro-socially and therefore cannot be purely motivated by altruistic outcomes (Andreoni 1989). This personal benefit is termed warm glow and can be defined as a positive internal feeling experienced by the giver which is associated with feeling good about having completed an action considered to be moral (Fehr & Schmidt 2003, Vesterlund 2006, Meier 2007). Warm glow motivations provide the giver with benefits beyond the altruistic concern for aggregate levels of social welfare and therefore influence the individual’s willingness to act in the interests of others. Warm glow motivations can lead to a lack of concern for the realisation of social benefit, as demonstrated by individuals who contribute to social causes despite knowing that their
14 donations have limited or no impact (Abbott, Nandeibam & O’Shea 2013). In this case, it is the act of giving itself that provides benefit to the donor and not the outcome of any donation (Crumpler & Grossman 2008). Experimental studies show that people willingly give to others despite no social result of their action and when they do so, they benefit only from internal warm glow.
The presence of warm glow benefit also causes individuals to value their own social contribution more highly than others’ and therefore to contribute with reduced concern for the total donation amount (Null 2011). Donors who demonstrate this trait make pro-social consumption choices despite being unable to directly monitor what happens to their money. For example, they may purchase a relatively expensive brand of t-shirt because it is made using fair trade practice, without qualifying the truth to this claim or understanding what social improvement fair-trade delivers. In this situation the additional value that excuses a higher price is the fair-trade production, however this benefit is virtually intangible because the donor may never experience what social benefit fair-trade delivers (Arnett, German & Hunt 2003). The fact that this exchange is purely reliant on trust, yet people continue to pay increased prices for fair-trade items, suggests that this exchange delivers a benefit to the consumer that is greater than the intangible and unreported social impact (Arnett, German & Hunt 2003). Because individuals are motivated by internal warm glow and therefore do not value other’s donations equal to their own, they appear willing to contribute to social cause without express concern for the realisation of social impact.
Warm glow is also evident in a demonstrated lack of sensitivity to the scope of social impact (Linardi & McConnell 2011, Imas 2013). For example, individuals demonstrate a willingness to donate the same value of financial support to rescue one animal as to rescue four (Hsee & Rottenstreich 2004) and others will spend the same amount to benefit one person as to benefit ten (Small, Loewenstein & Slovic 2007). The internal benefits that contribute to warm glow include signalling one’s wealth to others via the contribution of large donations (Becker 1976, Glazer & Konrad 1996), a satisfaction of addressing feelings of moral obligation or one’s personal standards and expectations (Kahneman & Knetsch 1992, Schwartz 1977, Dawes & Thaler 1998, Abbott, Nandeibam and O’Shea 2013), enjoying increased social
15 approval (Hollander 1990), feelings of being responsible for positive outcomes experienced by others (Andreoni 1989 & Dunn, Aknin & Norton 2008), viewing oneself as good or kind (Walster, Berscheid & Walster 1973) and avoiding negative reputational consequences (Conrad 2005). Individuals continually demonstrate that the value of social outcome is relatively unimportant compared with the value they donate, hence internal motivation is evident.
Because there are numerous types of warm glow benefits, it is possible that they may be a result of the situational elements of a pro-social setting (Shang, Reed & Croson 2008). These elements include the social norms and rules of a group or community, the contribution behaviours of others and the characteristics of a social cause such as any uncertainty regarding channelling support or whether the social cause is evaluated by itself or within the context of external sectors or social groups (Conrad 2005, Fehr & Fischbacher 2003, Kahneman & Knetsch 1992). While these elements influence individual decisions, when an individual is making a decision with the potential to positively impact others, they will most commonly be influenced by a combination of both the internal warm glow linked to their decision and altruistic ties to the social outcome they will contribute to (Menges, Schroeder & Traub 2005). This motivation is termed impure altruism as it combines both a concern for others and for oneself (Andreoni 1989).
Impure Altruism
Impure altruists are influenced by a combination of both internal warm glow and external altruistic motivations for acting in the interests of others and therefore, in contrast to a pure altruist, are motivated by both factors that increase the efficient allocation of social benefit and those that make them feel good about completing what they believe to be a moral action (Andreoni & Rao 2010). This suggests that an action that benefits others also benefits the giver, aligning more closely with traditional individual utility maximisation theories of economics (Menges, Schroeder & Traub 2005). Testing impure altruism involves analysing the decisions of subjects who are in a position that allows them to influence or determine both the level of benefit experienced by others and by themselves. This differs to testing purely warm glow, which investigates decision making by subjects who are unable to change the
16 level of external social benefit, but by increasing their levels of giving, are able to increase their personal contribution and hence influence their internal warm glow benefits.
Impure altruism has been tested under many conditions including liquidity constraints (Hubbard & Judd 1986, Altig & Davis 1989), reciprocity (Sugden 1984), lifetime uncertainty (Blanchard 1985), strategic giving (Bernheim, Shleifer & Summers 1985, Cox 1987), large economies (Andreoni 1989), varying the relative benefits of private and public goods (Palfrey and Prisbrey 1997, Steinberg 1987) and myopia (Thaler & Shefrin 1981). Impure altruism theory predicts that the complete crowding out of individual donations resulting in optimum levels of social support, such is seen with pure altruists, will not occur because of the presence of warm glow (Tonin & Vlassopoulos 2010, Steinberg 1991). This is an essential component of understanding what motivates individuals to act in the interests of others (Imas 2013) and is therefore an important addition to the altruistic model.
Businesses, Governments and Charitable Organisations
With consumers and donors exhibiting impure altruistic motives, businesses, governments and charitable bodies who rely on social support, continue to apply creative approaches to eliciting donations from individuals (Ariely, Bracha & Meier 2007). For example, charities who publish their donor’s names in brackets based on contribution amount, increase the value of warm glow by introducing reputational benefits, encouraging individuals to increase their support and social standing by donating just enough to move up a donation bracket (Harbaugh 1997). Similarly, subsiding fuel-efficient cars can influence consumer activity by reducing the reputational and warm glow benefits of purchasing such a vehicle and thereby impacting demand (Ariely, Bracha & Meier 2007). The most important factor that convinces individuals to support an organisation with social impact is the balance between how much their contributions will assist others and how much their action will benefit themselves (Brunel & Nelson 2000).
Modelling Giving in the Context of Warm Glow
Understanding that individuals are impure altruists and not indifferent to the social support that comes from themselves or from others, allows for a more comprehensive model of
17 giving behaviour, compared to the purely altruistic approach (Bernheim 1986). Modelling social behaviour therefore requires incorporating internal warm glow benefits into standard economic models where agents have and display preferences over outcomes. It also requires the understanding that agents have limited tolerance for taking actions that are suboptimal in terms of their preference but that when agents do make suboptimal choices in terms of outcome, they do so intentionally and in pursuit of a known aspiration, namely the warm glow benefit (Feddersen & Sandroni 2009). For example, when individuals purchase a car that is relatively expensive due to improved fuel efficiency and reduced environmental impact, they do so because they value the reputational warm glow associated with being a responsible consumer as equal to the additional costs of this purchase.
A decision-making model with warm glow should accommodate these assumptions and have three elements (Feddersen & Sandroni 2009);
(1) A reference over a finite set of alternatives such as the choice of a range of environmentally friendly and environmentally destructive cleaning products;
(2) A tolerance that identifies a least preferred alternative, for example knowing which cleaning product provides the worst cleaning finish making it least preferred, and
(3) Aspirations that identify the expensive environmentally friendly and therefore socially beneficial option that the consumer hopes to purchase but also the less expensive but environmentally destructive alternative that will be selected if the aspirational socially beneficial product is unaffordable.
Under this set up, individuals will make a decision between the given cleaning products based on their preferences and aspirations (Feddersen & Sandroni 2009). If the internal warm glow benefits combined with the external environmental benefits of the aspirational purchase outweigh the additional costs of the environmentally friendly product, individuals will purchase the more expensive option to maximize their aspirational outcomes. But if this cost outweighs the sum of the internal and external benefits, then the more affordable but environmentally destructive option will be selected.
18 This idea is depicted in Figure 1 below, which demonstrates the choice between three cleaning products. When the consumer makes an aspirational purchase and selects the green, most preferred option, they demonstrate that they value the increased costs associated with a more expensive product less than the value combination of the external altruistic environmental preservation and internal warm glow benefits. When they purchase the second preferred item they select the item that is cheaper in price and provides a moderate cleaning finish because the combined warm glow and altruistic benefits are valued at less than the price difference of $3 ($10 - $7).
Figure 1 Depiction of a Pro-Social Consumption Decision
A representation of this decision can be modelled in the below equation (Figure 2) which relates consumer utility (Ui) to a function of the consumption of the purchased product (Xi), contribution to the external social outcome (Gi) and the internal personal satisfaction of doing so (gi). In this example, the internal satisfaction (gi) might be caused by a range of outcomes such as;
• Friends and family recognising and praising the fact that the person has selected a product
that promotes environmental sustainability resulting in increased social approval;
• Avoidance of any negative social connotations linked to the environmental damaging
product; Pro-Social Purchase Conditions Product A Best cleaning finish $10 No environmental Impact Product B Moderate cleaning finish $7 Moderate environmental impact Product C Worst cleaning finish $5 Harsh environmental impact Most
19
• Internal belief that one has made a decision that has directly helped others and as a result
viewing oneself as good or kind; or
• Personal satisfaction resulting from the belief that the best social choice was made.
This model highlights that for impure altruists, it is a combination of internal warm glow, external altruistic and product characteristics that spurs the purchase decision.
Ui = Ui(Xi, Gi, gi)
Figure 2 Benefits Model - Consumption of Socially Beneficial Goods
A pure altruist experiences no internal warm glow satisfaction so this variable (gi) would be removed from the model (Ui = Ui(Xi, Gi)), while a consumer who is purely motivated to advantage themselves, experiences no benefit from socially minded outcomes so (Gi) would be removed (Ui = Ui(Xi, gi)). The relationship between altruism and warm glow benefits, as well as the value of the consumer product, determine the total utility gained from consumption. It is this utility that motivates consumption decisions made by the individual within the social business context.
Additional Approaches to Modelling Warm Glow
A number of additional models are also interesting to note when incorporating warm glow within the decision-making environment. While a variety of warm glow benefits have been noted, Harbaugh (1997) suggests it’s particularly useful to divide these individual motivations into two broad components; warm glow and prestige. In his model, prestige is specifically defined as the reputational effect of pro-social actions and in the model above, is incorporated within the internal warm glow benefit (gi). DellaVigna, List and Malmendier (2012) also categorise the benefits of social actions into altruism, warm glow and reputational outcomes, demonstrating that social pressure is a significant determinant in whether individuals contribute to door-to-door giving campaigns. Consistent with this categorisation, Ariely, Bracha & Meier (2007) investigate warm glow, altruism and reputation within the market for fuel-efficient cars and Menges, Schroeder and Traub (2005) apply this same approach to green energy consumption and obtain similar findings.
20 Just as defining a type or category of warm glow motivation provides insights into why individuals act in the interests of others, the differentiation of specific altruistic motivations is also useful. Ferguson et al. 2012 separate reluctant altruism from pure and impure altruism, introducing the new term to separately account for the benefits received from social actions that are motivated by a lack of trust in others. Within the setting of a voluntary blood bank, first-time donors reveal that they are motivated by impure and reluctant altruism, while repeat donors experience warm glow and pure altruism. While there are multiple versions of the warm glow model, each is consistent in separating the internal warm glow and external altruistic outcomes of acting in the interests of others, as can be applied within the context of consumer decision-making and social business.
The Context of Social Business Social Business
While to some extent the concept of a social business is still ‘under construction’, Yunus, Moingeon & Lehmann-Ortega (2010) provide the useful definition of a ‘self-sustaining company that sells goods or services and relays it’s owners investments, but whose primary purpose is to serve society and improve the lot of the poor’ (P. 309). Under this definition, a social business is a business in every sense, despite it often being challenging to merge for-profit activity and social objectives (Yunus, Moingeon & Lehmann-Ortega 2010). Social businesses are unique due to their popular association with pro-social outcomes but while this concept has obvious appeal, the market impact of this growing sector is yet to be fully measured (Webb & Mohr 1998, Andrews et al. 2014, Massetti 2011). Figure 3 below shows a visual representation of the motivations behind a social business compared with not-for-profit organisations and for-not-for-profit businesses, as provided by Yunus, Moingeon & Lehmann-Ortega (2010).
21 Social Goods
Within the context of social business there has been some research into what motivates and influences purchase decisions associated with raising funds for cause. The growing prevalence of both social businesses and social goods is evidence of the 80% of consumers who state that they would switch to a brand that supported a social cause, if quality and price were maintained (DaSilva & Kerkian 2010). Bundled private and public goods, those that have an internal consumption benefit as well as a positive social impact, receive higher willingness to pay values from experimental participants (Frackenpohl & Poenitzsch 2013) and two out of three of customers reported they would favour retailers or brands associated with a good cause (Meyer 1999). Additional research has found that socially responsible consumption behaviour is commonly maintained against varying market characteristics such as limited consumer information (Bartling, Weber & Yao 2014) and that social impact increases the benefit of purchase for the consumer.
22 The impact of altruistic and warm glow motivations has been demonstrated by consumers who donate greater amounts when purchasing social goods compared with offering direct donations and who select social goods over for-profit alternatives (Cornes & Sandler 1984, Kotcher 2006, Casadesus-Masanel et al. 2009). This activity is consistent across online auctions where individuals bid earlier and higher valuations for social products than for others, and within lottery environments which generate greater donation amounts, compared with requests for direct donations (Elfenbein & McManus 2010, Landry & Price 2006). Evidence suggests that the introduction of social impact directly linked to purchase influences consumer decision-making across a range of purchase environments.
When business is used as a channel for pro-social behaviour, marketing and the communication of product value impacts the tendency of customers to purchase (Sargeant, Hudson & West 2008). Advertising social outcomes to consumers can affect individual decision-making because this impacts the reputation of the organisation and the customer’s understanding of consumption benefits (Ross, Patterson & Stutts 1992). For example, marketing has the potential to highlight warm glow benefits over external altruistic outcomes and enhance how the customer personally feels about having purchased a social product. Advertising only the internal benefit of purchase has been found to have a positive influence on donation values, as was shown in a popular Red Cross campaign which claimed ‘Feel Good, Give Blood’ in an attempt to encourage increased blood donation by focusing solely on the personal benefit to the donor (Anik et al. 2009). This marketing approach causes warm glow in potential customers (Sweeny 2007) and highlights specific benefits to the consumer. Social organisations therefore invest significant portions of their budget in branding and communications, based on the strong evidence that the way customers view social businesses is directly linked to income generation from donors (Dixon 1997, Kennedy 1998, Grounds & Harkness 1998). Because of this, not-for-profit brands are becoming increasingly valuable tools for eliciting social support.
While marketing that concentrates on the social cause of a business can benefit the organisation, effective messaging requires accounting for the consistency of community support provided by the organisation, the quantity of social support relative to consumer
23 contributions, the fairness of the social efforts and any perceived trade-off between product quality and social impact (Barone, Miyazaki & Taylor 2000, Webb & Mohr 1998, Varadarajan & Menon 1988). Because factors such as these are pivotal to consumer decision-making, marketing efforts have the potential to both positively and negatively impact consumer contributions.
Campaigns such as the delivery of donation subsidies or the bundling of donations with low-value tokens, while intended to stimulate increased short-term support, can reduce the low-value of warm glow benefits in the long-term. Social organisations flirt with potential risk when employing campaigns that involve the bundling of donations with worthless items, as while initially this might stimulate support, the value of longer-term warm glow benefits may be affected (Imas 2013). In the case where a consumer highly values the internal benefits they receive from helping others, when an organisation associates pro-social activity with a low value reward, the value of warm glow is then limited to this imposed reward amount and the individual’s benefit consequently reduced (Gneezy & Rustichini 2000, Heyman & Ariely 2004). Because consumers also experience diminishing marginal utility from warm glow (Lilley & Slonim 2014), the long-term tendency to act pro-socially is possibly limited by campaigns such as these.
Another possible limitation of social business is consumer scepticism regarding the motivations of business activity. While customers are generally positive towards linking social impact to their consumption activity, they aren’t always convinced by the validity of social marketing, particularly when it is used with the intent of increased sales (Webb & Mohr 1998). While negativity can stem from scepticism such as this (Sweeny 2007), the majority of customers do however admit that cause-related marketing has an overall positive impact because they appreciate that an organization is incorporating a worthy cause into their mission and activity.
In addition to the marketing and communication of social business, the characteristics of the social good available for purchase are an additional important influence on consumer decision-making. Research suggests that the type of product being sold to generate social
24 benefit affects consumer preferences when purchases are relatively large, but that the presence of social impact generally does not affect what the consumer decides to purchase when the size of the donation or the cost of the item is relatively small (Strahilevitz & Myers 1998). Social impact incentives are found to be relatively more successful when applied to frivolous products and when the product is non-essential or a once-off purchase, compared with a staple and repeated purchase (Strahilevitz & Myers 1998). When the purchase criteria of the product plays a significant role in consumption decisions , such as when the consumer only desires a blue cup with a white handle and no other colour will suffice, social impact becomes less of a concern , however when the consumer has some flexibility over what and how they purchase, for example if they simply require a cup of any colour, there is more consideration for external impacts (Webb & Mohr 1998) and opportunities for the consumer to contribute to social cause.
In addition to the characteristics of the product, the reputation or self-image of an individual within the social purchasing environment plays a role in donation decisions (Winterich & Barone 2011) and the presence of a social norm effect is observed to increase socially beneficial behaviour (Abbot, Nandeibam and O’Shea 2013). Social environments do not always have a positive influence on the level of social activity however, as the demand for social goods differs under private and public decision-making (Conrad 2005). When a purely altruistic act of giving is ‘corrupted’ by advertising the benefits received by the individual giver within a social setting, a reduction of the pro-social behaviour can occur (Ariely, Biacha & Meer 2007). For example, if consumers assume that purchasing a social good requires them to compromise between quality and social impact, they demonstrate the requirement for additional warm glow benefits such as social approval.
Another distinct and important feature of the purchase decision is the relationship between donation amount and a consumer’s willingness to pay for the product (Koschate-Fischer, Stefan & Hoyer 2012). This relationship is influenced by the fit of the social initiative with the organisation’s mission and with the type of product for sale, as well as the timing of support delivery (Becker-Olsen & Hill 2005). Proactive, high-fit initiatives can lead to improved consumer attitudes and intentions, while reactive, low-fit initiatives can have a
25 negative effect on social contributions. While discounting items also has the potential to reduce value perception, price sensitivity and brand equity, inclusion of a donation to charity as a promotional tool does not necessarily cause this negative outcome (Arora & Henderson 2007). For example, if a company is discontinuing the production of a specific pair of shoes, rather than discounting the remaining stock which can signify a reduction in value, maintaining the current price and linking the product to a social outcome can introduce additional value for the consumer, stimulating purchase without negatively impacting brand value or reputation.
While the effective design and marketing of social business is arguably complex, when businesses sell goods bundled with a promised social benefit, sale success is largely a reflection of how ‘good’ this makes the customer feel (Strahilevitz 1999). Evidence suggests that the willingness to pay for a charity-linked product is dependent on the interaction between the size of the donation and the value of the product, as well as the warm glow that the consumer receives in addition to the benefit of owning the new product. Overall, bundling public and private goods intensifies consumer demand through the introduction of benefits for the consumer and the wider social group.
What this will contribute to the literature
Within the literature we see that the majority of individuals act as impure altruists, motivated by both the altruistic desire to assist others and the warm glow outcome of being a provider. Within the context of social business, individuals make purchase decisions linked to socially beneficial outcomes, the presence of which influences the decision-making process. Current findings suggest that potential consumers perceive social businesses positively, are willing to pay more for social goods and will select the socially beneficial option when provided with a choice. While consumers appear to value social outcomes and are influenced by a product’s characteristics and the purchase environment, the influence of social impact on consumer interest outside of the experimental setting and whether altruism or warm glow is most influential within the consumer decision-making process, are two areas that are yet to be completely understood. Within the context of individual giving via social business channels, this paper will therefore investigate these two areas.
26 Firstly the paper will compare an online social business, which uses 100% of its profits to deliver educational resources to girls in India, with an identical replicate that differs only with regard to the mention of social impact. This will be done in order to understand whether the presence of social impact changes the level of consumer interest in a given product range within the online market setting and will contribute to the literature by testing the influence of the presence of social impact in the market.
Understanding whether the introduction of social impact influences consumer interest levels will provide insight into whether consumers are searching for the presence of social impact when shopping online. If the two stores receive similar levels of interest however, one may conclude that it is the product rather than social impact that causes consumer interest and that social impact might be a secondary concern within the consumer decision-making process. This would further enhance our understanding of consumer decision-making by indicating whether consumers make the choice to search for and view social goods when simply shopping for a desired item and not when they are given options from which to select from. Secondly this paper will run a between participants experiment to test the presence of warm glow and altruistic motivations within a consumer’s willingness to pay for a non-essential consumer item with social benefits. This will be completed within two separate settings; (1) impure altruism will be tested by analysing the decisions of subjects who can influence or determine both the level of benefit experienced by others and by themselves and (2) pure warm glow will be tested by investigating decision making by subjects who are unable to change the level of external social benefit, but by increasing their levels of giving are able to increase their personal contribution and hence influence their internal warm glow benefits. At an aggregate level these results will be compared to understand whether the presence of altruism significantly changes the willingness to pay for social impact.
With increased knowledge of consumer motivation, the communication of social business can be tailored to highlight those benefits that are most important to the customer, thereby contributing to improving the appeal and performance of the social business model. For
27 example, given consumers are motivated to a larger degree by warm glow rather than altruism, advertising the internal benefits of contributing to a social cause may improve performance. Being able to distinguish between the altruistic and warm glow motivations, in addition to the appeal of the given consumption item, may enable social businesses to better understand their customers and to improve the design of their operating model.
With regards to contributions to the literature on impure altruism, the paper seeks to provide real-word application of findings. The results of these two studies are important not only for social businesses but all other organisations designed to generate some form of pro-social outcome, because they may provide effective and efficient insights into improving customer appeal, without the necessity for increased costs or effort. Purely altering the approach to engaging and communicating with consumers might be enough to appeal to the motivations of individuals to make pro-social consumption choices.
The following section of the paper is broken into two main areas of study; ‘study one’ covers the methodology, research design and results for the investigation into the influence of social impact on online consumer interest levels while ‘study two’ provides the methodology, research design and results for the test that evaluates the relative influence of altruism and warm glow within the consumer decision-making process. Following these sections I will provide a discussion of the results.
28
Study One
In this section I will provide the methodology, research design and hypothesis relevant to study one, before also providing a brief discussion of the results.
Methodology
Study one involved an online field study and investigated whether the presence of social impact directly linked to the sale of non-essential consumer items caused increased interest from online customers. Study one addressed research question one which asked; does the potential for social impact influence levels of market interest in a given range of consumer products? Findings from this study indicated whether the presence of social impact significantly increased online consumer interest in a given product range.
A registered Australian charity called Ruby Silver, which functions as a social business, was engaged as the research platform. Ruby Silver retails and wholesales sterling silver jewellery which is considered to be a non-essential consumption item as regular purchase is not vital to the physical or emotional wellbeing of the consumer.
Ruby Silver identifies as a social business as it uses 100 per cent of its profits to purchase and deliver educational resources to girls in India. The girls who receive support from Ruby Silver live in either New Delhi or Rishikesh, and rely on Non-Government Organisations (NGOs) in their community to access education. These NGOs work closely with local families to communicate the importance of obtaining an education and to gain support for girls to attend school. They also cover the costs of school and educational supplies, and provide a place for the girls to do their homework or play together in the afternoons.
As per the organisation’s description and business objectives, Ruby Silver profits are used to purchase school supplies such as uniforms, blankets and schoolbooks. Ruby Silver runs an effective model with aspirations consistent to that of other small social businesses, to maximise its profits and to channel these funds to their social objective. Access to Ruby Silver’s operational practices has enabled the completion of this study.
29 Experimental Design Procedures
Online Store Comparison
A field study was conducted to ascertain whether the advertisement of social impact directly linked to the consumption of non-essential consumer items, resulted in increased customer interest. To do this, two new Ruby Silver online stores were created and placed on a global peer-to-peer e-commerce selling platform. The stores were created identically except for the presence of cause-related messaging; the social store advertised that it delivered social benefit as a direct result of each customer purchase, using 100% of the profits to deliver school resources to girls, while there was no mention of social impact within the for-profit replicate. The performance of these two stores over a 118 day period was recorded and compared to ascertain whether there was any significant difference in levels of customer interest.
Platform Selection
The platform that was selected to host the two Ruby Silver stores was Etsy, a peer-to-peer e-commerce website designed as an online marketplace where ‘millions of people around the world connect… to make, sell and buy unique goods’ (Etsy 2017). Etsy was selected because it has significant customer traffic reporting 28.6M active buyers, the sales environment complemented the nature of the social business’ hand-crafted products, and the platform was able to host two almost identical stores simultaneously. Etsy typically attracts a wide range of different customer demographics given it is a world-wide marketplace selling millions of different products.
Store development
The two Ruby Silver online stores were created and went live one day apart, the social store on 8PP
th
PP March 2017 and the for-profit store on 9PP
th
PP March 2017. Information about the social impact associated with purchase was included at a number of specific locations within the social store, while this was specifically removed in the for-profit setting. Table 1 below lists the store features that were used in the Ruby Silver stores and indicates any differences
30 between the social and for-profit environment. For visual images of the stores headers please see Appendix 1.
Store feature Identical? Explanation of difference (if relevant)
Product images Yes Store owner description Yes
Prices Yes
Branding Yes
Store logo Yes
Product availability Yes Shipping terms Yes Featured stock Yes Count of stock Yes Links to social media & website
Yes
Payment options Yes Terms & Conditions of sale
Yes
Product descriptions No Social store included mention of social impact Product names No Social store included mention of social impact
Store name No One store was called Ruby Silver and the other Silver Ruby Store description No Social store included mention of social impact and images of
social impact
Store backdrop image No Image the same except social store included mention of social impact
Store slogan No Social store included mention of social impact
Table 1 Study One - Online Store Features
While only one store communicated social impact to those viewing the platform, in reality both stores were run by the social business and therefore 100 per cent of the profits from both stores were used to support the education of girls via the delivery of educational resources. No changes were made to the stores for the duration of the period, once the stores were both live.
31 Data Collection
The Etsy platform captures data on a large number of store performance variables. Data from the two Ruby Silver stores was collected from the day that both stores were live, 9PP
th
PP March 2017. Data collection occurred at only two occasions throughout the 118 day timeframe so as not to influence the store viewing statistics, once when the stores had been live for 3 weeks, to initially view the type of data available, and again at the end of the collection period. Data collection involved viewing the store performance metrics on the Etsy platform and manually transferring these variables into Excel format.
Data Description
Table 2 below includes the data variables available from the Etsy platform an indication of
variable type and description.
Data Available Variable type Description
Shop views Explanatory variable
The number of separate times that any webpage associated with the store is viewed online. This is a numerical value.
Listing views Control variable The number of separate times that the webpage specific to each product listing within the store is viewed online. This is a numerical value.
Shop favourites Control variable On the Etsy platform customers are able to ‘favourite’ a store by selecting the favourite button on the store homepage. This saves the store in their ‘favourite’s folder’. This is a numerical value listed against any store that has been favourite’ed.
Listing favourites Control variable On the Etsy platform customers are able to ‘favourite’ an item by selecting the favourite button on each of the item pages. This saves the product in their ‘favourite’s folder’. This is a numerical value listed against any product that has been favourite’ed.
Traffic sources – shop view
Control variable The website that customers were on before selecting to move to the store page. E.g. if the entry reads ‘etsy.com’ the customer found the store from a page within the Etsy platform. This is a list of pages internal and external to the Etsy platform.
32 customer
Traffic sources – listing view
Control variable The website that customers were on before selecting to move to a listing page. E.g. if the entry reads ‘etsy.com’ the customer found the listing from a page within the Etsy platform. This is a list of pages internal and external to the Etsy platform.
Most active Supplementary informational source
A list of the most viewed products within the store including the number of views per product listing. This is a numerical value.
Customer conversations
Supplementary informational source
The number of customer enquiries received. This is a numerical value.
Table 2 Study One - Data Descriptions and Variable Type
The number of shop views per day was selected as the control variable as a direct measure of store popularity. Additional control variables were selected based on the frequency they were recorded by Etsy and the ability to link values to individual consumer actions. Data was recorded on a daily basis, per individual action.
Data Management
Data that had been transferred from Etsy to Excel was initially reviewed for any outliers before the ‘Location of Customer’, ‘Search Source – Shop View’, and ‘Search Source – Listing View’ variables were allocated dummy variables of either (0) or (1) to indicate the relevant control variables per day. An additional dummy variable was added to indicate whether the data collected had come from the social store (1) or the for-profit environment (0). Data was then transferred into STATA for analysis.
Hypotheses Formulation
Research question one asked does the potential for social impact influence levels of market interest in a given range of consumer products? and the design of study one was such that it would provide indication of whether the presence of social impact increased, decreased, or did not change the level of consumer views, compared with the for-profit alternative.
33 Understanding whether consumers were searching for socially minded products and therefore viewing them more often within the online platform would indicate whether the presence of social impact was important to consumers as a condition of purchase. Specifically avoiding asking consumers to select their preference between social and for-profit items, this approach looked purely at whether the social trait was searched for by potential online consumers and caused increased consumer traffic. Given that consumers were required to identify the stores themselves within the global Etsy platform, this would provide enhanced understanding of consumer behaviour and grow the literature in this field.
Based on this approach, hypothesis one is as follows:
• Hypothesis one: The presence of social impact significantly increases market interest in a given range of non-essential consumer goods.
Results
In this section I will provide the relevant summary statistics before detailing the descriptive results.
Summary Statistics
A high-level overview of the data values obtained in study one is provided in Figure 4 below. The social store had an average shop view per day of 1.02 while the for-profit store had an average of 0.97. The standard deviation of shop views across the social and for-profit stores was similar at 1.42 and 1.33 respectively. The minimum and maximum values per day were identical between the two stores, at 0 and 7 respectively.
The social store received a slightly higher average number of listings selected as a customer favourite per day but a lower number of average times that a customer selected the shop as their favourite, compared with the for-profit store. These values are relatively small given the low number of favourite activities per customer. Overall there were no obvious differences between the social and for-profit stores.
34
Figure 4 Study One Summary Statistics
Descriptive Statistics
A descriptive summary of the data from study one is provided below in Figure 5 below. With data from both the social and for-profit stores combined, the number of items viewed on any day was between 0 and 15. Customers selected the stores as a favourite 5 times and individual product listings were selected as customer favourites 34 times.
The most common customer location prior to viewing the store was the internal Etsy platform with 46 (40%) occurrences, followed by traffic direct to the site with 30 (26%) occurrences and 23 (20%) from external search engines. Etsy also provided the main source of customers who viewed individual product listings with 224 (75%) customers coming from this online location. 72 customers (24%) came from the Etsy App and just 2 (1%) from direct traffic to the store platform. The majority of customers who viewed the stores were located in the United States or Canada with 171 views (41%) from this area. Australia also had a relatively high number of views with 142 (34%), followed by Europe with 71 views (17%). 4% of the sample was made up of unknown locations.
Shop Views Shop Favourite Listing Views Listing Favourite
Mean 1.02 0.02 2.25 0.46 Standard Deviation 1.42 0.13 2.05 0.93 Minimum 0 0 0 0 Maximum 7 1 8 4 Mean 0.97 0.07 2.41 0.12 Standard Deviation 1.33 0.25 2.69 0.38 Minimum 0 0 0 0 Maximum 7 1 15 2 Mean 0.99 0.04 2.33 0.29 Standard Deviation 1.37 0.2 2.38 0.73 Minimum 0 0 0 0 Maximum 7 1 15 4 Social For-profit Total
35
Figure 5 Study One - Overview of Data
Figure 6 below shows the distribution of the number of customer views per day within the
social store (blue) and for-profit store (grey). On 27 of the days, making up 46% of recorded time, the social store received 0 views. On 20 of the days (34%) the social store received one view and on the remaining 20% of time the store received 2 or more views per day. As occurred within the social store, on 27 of the days, making up 46% of recorded time, the for-profit store received 0 views. On 18 of the days (31%) the for-for-profit store received one view and on the remaining 24% of time the store received 2 or more views per day.
Total observation days: 118 frequency in % Min Max
Store views (average per day) 1 0 7 Listing views (average per day) 2 0 15 Store favourites (sum) 5 0 1 Listing favourites (sum) 34 0 4 Location of customer US/Canada (sum) 171 41% Australia (sum) 142 34% Europe (sum) 71 17% Asia (sum) 6 1% S_America (sum) 10 2% Unknown (sum) 15 4% Search Source - Shop view Etsy (sum) 46 40% EtsyApp (sum) 9 8% External search engine (sum) 23 20% DirectTraffic (sum) 30 26% Rubysilver.com (sum) 6 5% Google (sum) 2 2% Search Source - Listing view Etsy (sum) 224 75% EtsyApp (sum) 72 24% DirectTraffic (sum) 2 1%
36
Figure 6 Study One - Daily Store Views for Social and For-Profit Stores
Note: The left vertical axis measures the total number of days within the 118 days analysis period on which the given store received a specified number of customer views and the right vertical axis measures the cumulative percentage of this total number of days. The horizontal axis measures the total number of customer visits to the store per day.
Econometric Analysis
A regression analysis was completed to ascertain whether there was a significant difference between the number of shop views in the social store and for-profit store. This was required to understand whether any difference seen in the data was statistically significant and therefore relevant for analysis. Without this approach, any numerical difference could not be stated to be relevant.
The regression analysis in Figure 7 examined whether the dependant variable (‘shopviews’), which measured the number of shop views per day, was significantly related to the dummy control variable (‘social’), which indicated whether the views occurred within the social or for-profit store. The null hypothesis stated that there was no significant relationship between the store type and the number of views per day. Given the regression p-value of 0.8401 the null hypothesis could not be rejected at any level of significance. We therefore concluded that
0% 20% 40% 60% 80% 100% 120% 0 5 10 15 20 25 30 0 1 2 3 4 5 6 Cu m ul a0 ve p er ce ntag e of n um be r o f d ay s N um be r o f d ay s Number of store views per day
Social & For-Profit Histogram of Store Views
37 there was no significant relationship between the number of shop views per day and whether the store was for-profit or social. This finding lead to the conclusion that in this case, the presence of social impact within the online stores had not stimulated greater customer interest.
Figure 7 Study One – Test of the Significance of the Difference between the Social and For-Profit Stores, Regression Analysis
Note: The table presents coefficients from regression analysis. The dependent variable is the number of shop views per day and the independent variable is the dummy variable that indicates whether the store was social or for-profit.
To further test this model, control variables were added to ascertain whether these influenced the findings. When the continuous control variables indicating the number of listing views and the number of listing favourites, as well as the dummy variable indicating the location of the viewer were included within the model, the ‘social’ variable indicating whether the views occurred within the social or for-profit store remained insignificant. This insignificant finding was also seen when using listing views as the dependant variable and when the location dummy variable was replaced with the dummy variable indicating the online source of the product viewer. For a data output relating to the first of these further tests see Appendix 2. Given these additional insignificant findings we concluded that the presence of social impact within the online sales environment did not significantly influence the level of consumer interest. Number of Obs 118 F (1,116) 0.04 Prob > F 0.8401 R-squared 0.0003 Root MSE 1.3738
Shopviews Coef. Robust
Standard Error t P > |t| Social -0.51 -0.25 -0.2 0.84 -0.55 0.45 Const. 1.02 0.18 5.5 0 0.65 1.38 (95% conf. interval) Regress shopviews social, robust Linear Regression
