The Effects of Pro-Social and Pro-Environmental Orientation on Crowdfunding Performance

Reward-based crowdfunding is an alternative type of project financing in which a large and dispersed online crowd contributes relatively small financial amounts in exchange for innovative products or services. The crowd is driven by a broad set of motivations that also comprises sustainability awareness. However, empirical research on crowdfunding projects that feature social or environmental considerations provides inconclusive results. In our study, we enhance the understanding of whether a pro-social and pro-environmental orientation affects the performance of reward-based crowdfunding. We draw on the literature stream of social movements to explain how linguistic framing mobilizes individuals and relate this to how selection is enabled and action is guided in a crowdfunding setting. Based on a sample of 1049 projects from Kickstarter, we employ computer-aided text analysis (CATA) to capture the pro-social and pro-environmental orientation of the project descriptions and transcribed video pitches as linguistic constructs. We found that the level of pro-social or pro-environmental orientation has an inverted U-shaped effect on crowdfunding performance. Moreover, this relationship differs when crowdfunding projects feature a creative product or service idea. Our results suggest that entrepreneurs need to delicately balance a pro-social or pro-environmental orientation and find the “right” level of emphasis to create a competitive advantage.


Introduction
In reward-based crowdfunding as an alternative type of venture financing, project backers are entitled to a non-financial reward in exchange for their contribution. Typically, the reward is the product or service, which the crowdfunding project aims to develop or market through the funding [1]. Thus, project backers intend to consume the products or services from the projects they finance, which is why they are coined as a "different kind of investor" [2]. Since backers behave like typical consumers, they are driven by a broad set of motivations distinct from that of traditional investors [3,4], also in terms of sustainable considerations [5]. However, empirical evidence on sustainability in crowdfunding is limited and inconclusive. Researchers found that crowdfunding projects that promote a sustainable orientation can increase the probability of crowdfunding success [6][7][8][9], while other researchers find no such effect [3] or even a negative relationship [10]. Our study aims to help reconcile the contradictory findings on the relationship between a sustainable orientation and the outcome of crowdfunding projects. In detail, we shed light on how promoting a sustainable orientation affects crowdfunding performance. This helps to better understand the pledging behavior of backers and contributes to the discussion about resource mobilization in reward-based crowdfunding.
We identify a sustainable orientation in a crowdfunding project when it demonstrates the awareness of social and environmental values, such as social and ecological fairness and justice [11]. Our selection of social and environmental categories aligns with prior literature about sustainability and meets the conditions to account for sustainable development [12]. In entrepreneurship research, the goal of sustainable development is to sustain communities (social dimension) as well as life support systems and nature (environmental dimension) [13]. To conceptualize the effects of pro-social and pro-environmental orientation on crowdfunding performance, we employ the theoretical lens of framing. Frames function as cognitive shortcuts that enable individuals to "locate, perceive, identify, and label" occurrences and interpret information in any given situation [14]. More precisely, we draw on the meso level literature stream of social movements to explain how linguistic framing mobilizes individuals. Similar to social movements, the success of crowdfunding campaigns depends on collective action by a dispersed crowd [15]. Just as social movement leaders try to engage a crowd around a cause to mobilize resources, a crowdfunding project creator seeks to find support for his or her crowdfunding campaign [16]. Following the call for research by McKenny et al. [17] to incorporate resource mobilization theory regarding social movements in crowdfunding, we apply linguistic framing as outlined by social movement theorists to examine how the selection is enabled and action is guided in a crowdfunding setting.
In particular, our study offers three contributions. First, we add to literature concerned with factors that determine crowdfunding success and demonstrate how the emphasis of a pro-social or pro-environmental orientation affects the outcome of a crowdfunding campaign. By treating the social and environmental dimensions as psychologically distinct, we contribute to a more nuanced understanding of sustainability in crowdfunding. Second, we improve the methodological approach to measure pro-social or pro-environmental orientation in crowdfunding narratives. We collect both the textual description and transcribed video pitches of crowdfunding campaigns to minimize measurement errors that potentially occur in content analysis. Third, we support the findings of previous studies about the effect of entrepreneurs' language on resource mobilization. Furthermore, we advance the understanding of how the relationship between linguistic framing and the outcome of a crowdfunding project depends upon the product or service characteristics under which the framing is initiated and implemented.
Our study captures the social and environmental orientation as linguistic constructs using a content analysis algorithm that assesses the pro-social or pro-environmental framing of a project based on using cues expressing a pro-social or pro-environmental orientation. We employ the computer-aided text analysis (CATA) tool CAT Scanner [18] for a sample of 1049 reward-based crowdfunding projects from Kickstarter. We find that the level of pro-social or pro-environmental orientation has an inverted U-shaped effect on crowdfunding performance. Thus, framing a crowdfunding project as pro-social or proenvironmental can be beneficial. Yet, entrepreneurs should be aware that overemphasizing a pro-social or pro-environmental orientation may backfire and reduce the probability of crowdfunding performance. This suggests that backers prefer moderate levels of pro-social or pro-environmental orientation in crowdfunding projects. Moreover, we demonstrate that the inverted U-shaped relationship between the level of pro-social orientation and crowdfunding performance differs between projects that do not demonstrate a creative product or service idea and those that do. Our findings help entrepreneurs craft their campaign narratives and create an appealing entrepreneurial narrative for both the textual description and the video pitch. This may contribute to sustainable crowdfunding projects being more successful and thus develop their full potential for society and the environment.
The following section presents an overview of the extant literature on sustainable orientation in crowdfunding and the theoretical background, which serves as a basis to develop our research hypotheses. We describe our dataset and methods in Section 3 and present the results of the empirical analysis in Section 4. In Section 5, we compare our findings to previous research, address our contributions, and shed light on the limitations of this study that offer opportunities for future research.

Crowdfunding and Sustainability
The antecedents of crowdfunding performance have been the subject of entrepreneurship scholars for the last decade [19]. Extant research focused on the multifaceted motivations of backers and the different crowdfunding models to explain how project orientation affects funding success [20]. One research area that is particularly contested is how intrinsic motivations and a pro-social or pro-environmental project orientation influences backers' support. For instance, Allison et al. [6] assess how linguistic cues that frame a venture as either a business opportunity or an opportunity to help others affect crowdfunded microfinancing. They propose that in a microlending context, lenders combine extrinsic factors representing traditional investment motives and intrinsic factors attached to altruistic decisions. By analyzing microloans placed on the crowdfunding platform Kiva.org, they find that lenders prefer ventures highlighting a pro-social orientation over ventures that demonstrate a business opportunity. Building on these findings, Moss et al. [8] explore if crowdfunding lenders prefer ventures that communicate economic and pro-social values in their narratives simultaneously or if they prefer a clear positioning in either social or economic realms. Their results demonstrate that ventures that position themselves through linguistic cues in a specific context are more successful with allocating resources.
However, the results of the studies by Allison et al. [6] and Moss et al. [8] reflect funding motivation based on language cues on the crowdfunding-based microlending platform Kiva. On this platform, lenders are not protected against loan default and do not receive any interest payments. Since loans on Kiva are intended to help the disadvantaged and lenders indirectly donate their interests to cover the platform's overhead, Galak et al. [21] argue that lending motivation on Kiva is inherently pro-social. Thus, these findings are embedded in a unique context that cannot be generalized for other forms of crowdfunding. Consequently, Cholakova and Clarysse [3] find different results using an experimental setting to explore how crowdfunding pledging decisions are influenced by the presence of financial and non-financial motivations in equity-and reward-based crowdfunding. The experiment participants were to decide whether they would pledge (as in reward-based crowdfunding) or invest (as in equity-based crowdfunding) in a fictive crowdfunding campaign. The authors find no empirical evidence that the decision to fund or pledge to ventures is altruistically motivated by the willingness to help others.
Hörisch [10] finds similar results for reward-based crowdfunding campaigns initiated and marketed as environmentally oriented on the platform Indiegogo. The author does not find a positive connection between environmental orientation and crowdfunding success but rather suggests that featuring an environmental orientation can be detrimental. While Hörisch [10] identifies environmental campaigns solely on the basis that they were assigned to the environmental category on Indiegogo, Calic and Mosakowski [7] follow a different approach and examine campaigns on the reward-based platform Kickstarter through a coding scheme to identify a social or environmental orientation. By social or environmental orientation, the authors understand that crowdfunding campaigns indicate primarily social or environmental objectives. Calic and Mosakowski [7] find that a social or environmental orientation positively affects the funding success of crowdfunding campaigns in the technology category. However, the authors divide their sample into campaigns assigned either to social entrepreneurship or to traditional for-profit ventures. Hence, their results show how social entrepreneurship takes a stand in reward-based crowdfunding compared to supposedly for-profit ventures. In this case, "supposedly" refers to what is lost in the approach of Calic and Mosakowski [7] in the continuum between zero (business venture) and one (social enterprise): for-profit crowdfunding projects that incorporate pro-social or pro-environmental elements.
Defazio et al. [9] address this shortcoming by using scores that also allow conclusions drawn on the effect of different levels of a pro-sustainable orientation. For this, the authors measure the frequency of words related to a sustainability orientation in reward-based crowdfunding campaigns and weigh them in relation to the text length. They find that the probability of successful funding increases for small levels and decreases for large levels of pro-sustainable orientation. In summary, the findings of Defazio et al. [9] suggest that the relation between pro-sustainable orientation and crowdfunding success is curvilinear, that is, positive for moderate emphasis and negative for high emphasis on pro-sustainable orientation.
This brief overview of prior research on social and environmental orientation in crowdfunding campaigns reveals two research gaps our study aims to address. First, the studies depicted above focus on either a social (e.g., Cholakova and Clarysse [3]) or an environmental (e.g., Hörisch [10]) dimension or blend the two dimensions (e.g., Allison et al. [6]; Defazio et al. [9]; Moss et al. [8]). Moreover, the studies are set in different crowdfunding forms (e.g., reward-based and lending-based) on different platforms (e.g., Kickstarter and Kiva), which does not allow their findings to be aggregated into one conclusive image for reward-based crowdfunding, as the underlying motivations to participate in the respective forms vary from one another. Although Calic and Mosakowski [7] indicate the presence of a social or an environmental orientation separately, they merely use their respective presences as an indicator to assign these campaigns to the social entrepreneurship category [22] and compare them to their for-profit counterparts. In conclusion, to our knowledge, no study to date separately evaluates the conditions for supporting reward-based crowdfunding projects that feature a pro-social or pro-environmental orientation. We believe that treating the social and environmental dimensions as psychologically distinct contributes to a more nuanced understanding of sustainability in crowdfunding. While Choi and Ng [23] find that prior research "does not offer an examination of the notion that different dimensions of sustainability can exist in the minds of consumers," studying Catlin et al. [24] demonstrates that consumers evaluate the two dimensions differently. While a social dimension is associated with an orientation that is short-termed and local, an environmental orientation is perceived to be long-termed and global. Compared to researchers, who operationalize sustainability as a unidimensional construct, taking into account the multidimensionality of sustainability and measuring their relative importance allows a better understanding of backers' pledging behavior.
Second, the studies of Allison et al. [6], Moss et al. [8], and Defazio et al. [9], which are based on language analysis, all make use of the campaign narratives. In a crowdfunding context, narratives can be referred to as the story told about the entrepreneur and his or her project [25]. Allison et al. [6] and Moss et al. [8] assess crowdfunding projects on the micro-lending platform Kiva, whose homepage structure only allows campaigns to publish text and a picture [26]. Consequently, the authors use the textual campaign descriptions as narratives to apply their respective linguistic cues. Besides a textual description and pictures, the reward-based crowdfunding platform Kickstarter also allows a video pitch to be incorporated into the campaign [27]. Video pitches are used to introduce the entrepreneur behind the campaign, the development and current state of the project, and the features and characteristics of the offered product or service. As such, video pitches are an important means to convey the campaign's story and thereby are part of the campaign's narrative [28]. Defazio et al. [9] state that crowdfunding backers source information about a project both from textual descriptions and video language. Disregarding video language in the linguistic analysis could, therefore, lead to measurement errors, as the data basis would differ from the information basis that a potential project backer considers. We take full advantage of the information available in crowdfunding campaigns and use both textual descriptions and video language as narratives for our entire sample. To our knowledge, this is the first study that evaluates the conditions for supporting reward-based crowdfunding campaigns that feature a pro-social or pro-environmental orientation based on the full narratives.

Framing and Frames in Social Movement Theory
Frames function as cognitive shortcuts that enable individuals to "locate, perceive, identify, and label" occurrences and thus interpret information in any given situation [14]. Framing as a theoretical construct is widely applied in organizational and management theory to explain individual decision-making on a micro level, interactive meaning construction on a meso level, and institutional changes on a macro level [29]. This study draws on the meso level literature stream of social movements in which framing is used to explain how meaning is co-constructed through human interaction [30]. In detail, social movement literature supports our study in two respects: first, it helps to explain how linguistic framing mobilizes individuals. Similar to social movements, the success of crowdfunding campaigns depends on collective action by a dispersed crowd [15]. Just as social movement leaders try to engage a crowd around a cause to mobilize resources, a crowdfunding project creator seeks to find support for his or her crowdfunding campaign [16]. Following the call for research by McKenny et al. [17] to incorporate resource mobilization theory regarding social movements in crowdfunding, and in line with Defazio et al. [9] and Nielsen and Binder [16], we apply linguistic framing as outlined by social movement theorists to examine how the selection is enabled, and action is guided in a crowdfunding setting.

Meaning Construction in Social Movements
The framing perspective in social movement theory follows the principle of interactionism that meanings or labels for objects, experiences, or events surrounding us do not exist a priori but arise from interpretive processes based on interaction [31]. While social movements in the traditional sense were outlined as carriers of ready-made ideas and beliefs, the framing perspective understands movement actors as signifying agents that produce and maintain meanings for adherents, antagonists, and observers [32].
The process of meaning construction through human interaction described above is conceptualized as framing. The signifying work of framing is an active (something is being done) and dynamic (evolves through interaction) process that entails contention, as what is being done and how it is evolving is resulting in interpretative frames that are new or challenge existing ones. The outcome of this framing activity is denoted as a collective action frame [33]. The shared fundamental understanding of framing and frames in social movement theory is based on Goffman's definition of frames as "schemata of interpretation" that assign meaning to social constructions, and framing is the active part of extending frames or creating new ones [14]. By that, frames perform two tasks: first, they demarcate what is relevant and what is not, that is, what is inside the frame and what is out of the frame. Second, frames condense information and label occurrences to help make sense of the situation one is confronted with. This twofold interpretive function to render events is also performed by collective action frames, but in a way that calls for action, e.g., to mobilize support and resources, such as time and money [32].
Collective action frames are continuously created and elaborated in dynamic framing processes. This ongoing process of sense-making is embedded in the overarching concept of discursive fields in which discussions, decisions, and actions take place [34]. Discursive fields evolve when a set of actors with congruent ideologies, that is, with congruent values and beliefs, debate events they consider problematic. The interaction inside discursive fields plays a central role in creating collective action frames and developing identities. In what is called core framing tasks by social movement theorists, like-minded individuals inside the discursive field negotiate a common understanding of problematic situations they try to change. Core framing tasks consist of three components: Diagnostic framing refers to identifying the problem and the attribution of blame. Prognostic framing proposes solutions to the problem at hand and specifies execution strategies. Motivational framing is the mobilizing framing task, which calls for the engagement in collective actions alongside a corresponding "vocabulary of motive." In conclusion, diagnostic and prognostic framing seeks to find agreement, while motivational framing fosters action [32].
By pointing to the problem and its presumed causes, diagnostic framing additionally serves as a marker to construct identity fields and thus assigns roles to the actors relevant to the issue, that is, whether they are subsumed under the field of movement protagonists (problem solvers), antagonists (movement opponents), or audience (neutral in-dividuals) [35]. To attract the audience and garner additional support for the movement on a discursive and direct level, the protagonists try to align the audience's personal identities with the movement's collective identity (e.g., its goals or tactics) through identity talks. On a strategic level, protagonists try to align or link the movement's frames with ideologically related movements and public opinion clusters to mobilize support on a large scale. How strong the mobilizing potency of a frame is among these groups, if it is noticed and "strikes a responsive chord," depends on the frame's salience and resonance [36]. Here, salience refers to how noticeable and accessible a frame's vocabulary is to an audience. An increase in salience enhances the chances that the audience perceives and processes the frame and consequently includes it into their judgment and decision-making process [37,38]. Resonance is a second variable factor that affects the mobilizing potency of frames. A-frame resonates with an audience when it aligns with their values and beliefs, which leads the audience to feel personally connected to the frame and thus form opinions based on the frame and act in its spirit [39].
Framing activities occur in various forms, with linguistic framing being one form to attract considerable attention from social science scholars (e.g., Giorgi and Weber [40]; Loewenstein et al. [41]). Linguistic framing entails using a selective vocabulary set to create specific meanings that influence the audience's understanding and guide their behavior [14,37]. As described above, in forming collective action frames, vocabularies of motive evolve that equip the frame with convincing linguistic rationales for audiences to participate in the movement and act in its spirit [42]. We apply these insights to rewardbased crowdfunding, where entrepreneurs seek to present their product or service through a campaign description and a video pitch using a vocabulary that motivates backers to pledge to their campaign [43].
On reward-based crowdfunding platforms, backers can choose from many pledging options that all compete for attention [44]. To facilitate the backers' investment decision, entrepreneurs present their campaigns via a textual description and usually a video pitch that offers compelling arguments why their offering has a clear added value over other products or services on the crowdfunding platform [16]. Together with personal details about the entrepreneur and the venture, these arguments represent the campaign's message or "entrepreneurial narrative" [25], which allows entrepreneurs to communicate a strategic selling proposition [43]. Backers base their decision on whether to pledge to a campaign or not mainly on the entrepreneurial narrative, with limited options to verify or monitor the given information [45]. Furthermore, entrepreneurs on crowdfunding platforms usually cannot build on the advantages of a brand history that assures the backers of a certain kind of product or service quality, nor can they guarantee their future prospects [46]. This is relevant since strategic linguistic frames help compete for the audience's attention and appreciation, especially in situations of incomplete information and uncertainty [39]. Thus, crowdfunding literature suggests that the way an entrepreneurial narrative is framed affects the audience's interpretation of the information [6] and their decision to support a campaign [43,47]. Building on these findings, Nielsen and Binder [16] employ an experimental setting to show that crowdfunding campaigns using strategic framing outperform those whose message is solely framed descriptively. In summary, we argue that strategic framing directs collective action because it attracts attention to the object being framed among a choice set and guides action when there is limited information.
For example, by framing a product or service in the entrepreneurial narrative as "fair" (e.g., embracing gender equity) or "green" (e.g., using renewable energy sources), entrepreneurs evoke product or service features associated with social consciousness or environmental friendliness. In that way, the product or service is demarcated from other non-socially or non-environmentally oriented offerings on the platform. This demarcation is intended to attract socially or environmentally concerned backers and consequently motivate them to support the campaign [9].

Framing and Sustainable Consumption
Backers in reward-based crowdfunding are coined as a "different kind of investor" because they intend to consume the products or services from the projects they support [2]. Since backers behave like typical consumers, they are driven by a broad set of motivations that is distinct from that of traditional investors [3,4], also in terms of social or environmental considerations [5]. Since the millennium, the effects of pro-social and pro-environmental framing on consumer response have gained traction with marketing and business ethics scholars. They find that customers react positively to such framing efforts, which is beneficial for organizational performance [48][49][50][51]. In particular, since consumers become increasingly involved in sustainable consumption [52], they request socially conscious product options and potentially boycott ("buycott" in the social movement jargon [53]) firms that promote products or services that are "unfair" [54]. As a result, firms that offer socially conscious products can evoke positive consumer responses [55]. Environmentally sustainable products, for their part, can change the brand attitude of customers: brands that introduce "green" products are evaluated more positively by their consumers compared to brands that solely offer conventional product counterparts [50].
The consuming behavior described above intends to change the market or institutional paradigms by incorporating non-financial values, such as fairness, justice, or ecological awareness in purchase decisions and is discussed in research as political consumerism [56], ethical consumption [52], or sustainable consumption [57]. While focusing partly on different underlying motivations, according to Long and Murray [58], these concepts all refer to "consumption practices that consider factors beyond the material use-values of the product" with the intention of changing existing market practices [56]. For sustainable consumption, this means that consumers buy products with the awareness of the consequences and responsibilities that one's buying behavior has for the future of the society and the environment [59], and potentially choose one product over another due to these considerations [52]. Thus, sustainable consumption can be defined as making socially or environmentally conscious purchase decisions to support changes in market practices that are seen as unjust.
Sustainably consuming shares similarities to participating in a social movement. Both involve active engagement to change events that, through the lens of individual and, respectively, collective identity, are seen as problematic. Yet, sustainable consumption is not a social movement by definition since consuming primarily pursues self-interest. Instead, it is a mix of citizen engagement that combines individual and public interests that have been conceptualized by Micheletti [56] as individualized collective action: "a group of individual actors making similar decisions based on a perceived shared ideology, while simultaneously meeting their own personal needs" [58].

Research Hypotheses: Emphasis of Pro-Social and Pro-Environmental Framing on Crowdfunding Performance and the Role of Creativity
A central research subject of framing literature deals with defining an effective communication structure to convey a frame to an audience. Studies that address linguistic framing find that frames are effectively conveyed through cues that refer to central concepts of the frame [40,50]. In particular, cues make a frame salient and thus noticeable and accessible to an audience [37]. Moreover, the quantity of cues determines how much emphasis is put on a frame [60]. The "right" amount of emphasis depends on the situation and context and is difficult to generalize. For example, Entman [37] claims that a single remark in a text may be sufficient to set a frame if the frame resonates with the audience. Similarly, Hertog and McLeod [60] find that one or two cues can effectively convey a frame even in a large amount of text.
Concerning overemphasizing, scholars find that "too many" cues can decrease a frame's effectiveness. From a marketing standpoint, Olsen et al. [50] argue that brands using a greater quantity of cues to highlight the environmental value of a product obstruct the efficacy of their green marketing efforts. As for entrepreneurship research, Parhankangas and Ehrlich [61] present similar results while analyzing investment proposals. The authors find empirical evidence that business angels are less likely to invest in startups that promote themselves as overly innovative and use high levels of positive language.
In reward-based crowdfunding, projects that overuse accountability language disclose too much information and consequently reduce their chances of achieving their funding goal [62]. Finally, Defazio et al. [9] show that overemphasizing a sustainable orientation in crowdfunding projects negatively affects the project's success. These findings are due to two main factors. First, when a frame is overemphasized in a market context, the audience tends to become more skeptical about the offering [63]. Too much emphasis causes the audience to doubt a frame's credibility, which in turn negatively influences the frame's effectiveness [62,64]. Hence, it can backfire if an entrepreneur uses too many cues to set a frame in a reward-based crowdfunding campaign since potential backers may not believe that the claims are true.
Second, frames highlight certain characteristics of a context and, in turn, cover up other ones [65]. As a result, the audience becomes uncertain about an offering when it feels that product or service features could be missing [66], especially when these features are typically present in comparable offerings [67]. From a consumer standpoint, a product is a complex bundle of attributes and benefits. While core benefits stem from the product's value-giving attributes that ensure its basic functionality, added benefits are based on sustainable considerations, for example, [68]. The basic functionalities are the driving force for purchase decisions, and even highly sustainability-conscious consumers are unwilling to trade off social or environmental attributes at the expense of product or service functionality [69][70][71]. In this vein, crowdfunding scholars provide evidence that backers in reward-based crowdfunding are mainly driven by self-interest since the anticipation of a reward in exchange for their contribution appears as a strong motivator when funding campaigns [72,73]. Hence, when an entrepreneur in a crowdfunding project strongly emphasizes a pro-social or pro-environmental orientation in the entrepreneurial narrative, potential project backers may be deflected from basic product or service functionalities they expect to be present.
We conceptualize that a pro-social or pro-environmental framing affects the outcome of a crowdfunding project in two different ways via countervailing latent effects. The first latent effect is positive and reflects that a pro-social or pro-environmental orientation corresponds to backers' requests for socially conscious or environmentally sustainable products or services. As we discussed above, firms that introduce such offerings are evaluated more positively by consumers and outperform competitors who offer conventional products or services. In contrast, the second latent effect harms the outcome of a crowdfunding project. As laid out above, a strong emphasis on social or environmental framing increases the chance that backers doubt the frame's credibility and that it hides other product or service features that backers deem important.
Both the positive and the negative latent effects grow with the level of social or environmental framing. However, the two latent effects grow at different rates. The positive latent effect takes on a concave shape, and thus its slope decreases with a growing level of social or environmental framing. This is because an upper limit exists on how a crowdfunding campaign can promote a social or environmental orientation. In our study, we derive scores by dividing the amount of social or environmental rhetoric by the word length of the narratives. Accordingly, this upper limit is the point (namely a score of 100) where the project would only consist of social or environmental rhetoric, leaving no room to describe other details of the project, e.g., technical specifications. We thus conceptualize that the marginal positive latent effect will get smaller as the level of social or environmental framing increases. For the negative latent effect, we propose that it takes on a convex shape and that its slope increases with social or environmental framing.
In conclusion, one or two references or cues may be enough to successfully convey a frame to an audience. Hence, when the level of social or environmental framing is relatively low, this may not hide other important information and may not jeopardize the frame's credibility. However, as the level of social or environmental framing grows, escalating negative effects come into play, such as skepticism and the covering of relevant product or service information. Since the marginal negative latent effect will become stronger as the level of social or environmental framing increases, we conceptualize that after a certain point, skepticism and the concealment of other information start to dominate the concavely shaped benefits of meeting backers' social or environmental product requests (see Figure 1). that the marginal positive latent effect will get smaller as the level of social or environ mental framing increases. For the negative latent effect, we propose that it takes on a co vex shape and that its slope increases with social or environmental framing.
In conclusion, one or two references or cues may be enough to successfully convey frame to an audience. Hence, when the level of social or environmental framing is rel tively low, this may not hide other important information and may not jeopardize th frame's credibility. However, as the level of social or environmental framing grows, esc lating negative effects come into play, such as skepticism and the covering of releva product or service information. Since the marginal negative latent effect will becom stronger as the level of social or environmental framing increases, we conceptualize th after a certain point, skepticism and the concealment of other information start to dom nate the concavely shaped benefits of meeting backers' social or environmental produ requests (see Figure 1). Thus, we predict an inverted U-shaped effect of the level of pro-social or pro-env ronmental orientation on crowdfunding performance. Therefore, we hypothesize the fo lowing.
Hypothesis 1. Pro-social orientation will have a curvilinear, inverted U-shaped relatio ship with crowdfunding performance. Hypothesis 2. Pro-environmental orientation will have a curvilinear, inverted U-shape relationship with crowdfunding performance. Thus, we predict an inverted U-shaped effect of the level of pro-social or pro-environmental orientation on crowdfunding performance. Therefore, we hypothesize the following. Hypothesis 1. Pro-social orientation will have a curvilinear, inverted U-shaped relationship with crowdfunding performance.

Hypothesis 2.
Pro-environmental orientation will have a curvilinear, inverted U-shaped relationship with crowdfunding performance.
Prior studies suggest that the relationship between linguistic framing and the outcome of a crowdfunding project depends upon the product or service characteristics under which the framing is initiated and implemented [74]. An important product or service characteristic concerning the outcome of crowdfunding projects is creativity [7,74,75]. Accordingly, we expect that a creative product or service idea will likely affect the relationship between the framing of an entrepreneurial narrative and the project outcome.
Kickstarter's mission is "to help bring creative projects to life" [76]. In a market context, a product or service is considered creative when it is both novel and useful [77]. Putting a creative product or service into practice requires an extensive set of procedural, technical, and intellectual knowledge [78,79]. This specialized knowledge combined with lateral thinking enables entrepreneurs to combine information from their prior experiences and their present environment into creative product or service solutions [80,81]. Thus, entrepreneurial creativity is being able to "rapidly recognize the association between problems and their purported solutions by identification of non-obvious associations and/or by reshaping or reforming available resources in a non-obvious way" [82]. We expect that this ability is beneficial when featuring a sustainable orientation in a market context and facing its accompanying challenges. For instance, entrepreneurs need to address environmental, social, and economic outcomes simultaneously, which involves improving "the general welfare of society" [83] and pursuing financial performance goals simultaneously [84]. This implies taking a wide range of different stakeholders and their often conflicting demands into account [85,86].
Moreover, a pro-social or pro-environmental orientation implies a long-term orientation of the business [87]. Thus, compared to financially driven short-term decision-making of traditional businesses [88,89], entrepreneurs include the needs of future generations in their decision-making and typically feature a more global perspective [83].
In conclusion, when featuring a pro-social or pro-environmental orientation, entrepreneurs are confronted with situations in which they need to address multiple conflicting outcomes simultaneously and satisfy demands at the firm and societal level with different time horizons and different agendas [84]. We argue that creativity helps in handling these tasks as it drives entrepreneurs to, e.g., develop or identify new raw materials, figure out new ways of production, or find new business models to fulfill the expectations above. As a basis for our analysis, we suggest that entrepreneurs are more likely to fulfill the claims about a social or environmental orientation when the project or service idea demonstrates creativity. Accordingly, we speculate that backers assume that the pro-social or pro-environmental claims in a crowdfunding project are more likely to be realized when the project or service idea is creative. This results in less skepticism about the feasibility of the social or environmental claims.
As discussed above, the overall impact of the level of social or environmental framing on project outcome is the net result of the positive and negative latent effects of the framing. How creativity moderates the overall impact of social or environmental framing on project outcome will consequently be determined by how creativity affects the negative effect of social or environmental framing. We argue that creativity will strengthen or weaken the negative effect of social or environmental framing for the following reasons. If the project does not demonstrate a creative product or service idea, backers might doubt that the entrepreneur can manage the complex task of addressing the multiple conflicting outcomes that emerge with a social or environmental orientation. This might intensify skepticism about the social or environmental claims of the entrepreneur, which will, in turn, strengthen the negative latent effect that backers might doubt the frame's credibility. In contrast, demonstrating a creative product or service idea might lessen backers' skepticism about the social or environmental claims. Since crowdfunding projects are built on projections of a future that only exists if the campaign generates enough funding [62], claims about the project need to be plausible [90]. Drawing from the literature above, this may reduce skepticism about the social or environmental claims, which will, in turn, weaken the negative latent effect that backers may doubt the frame's credibility (see Figure 2).  Consequently, we argue that the inverted U-shaped relationship between the level of social or environmental orientation and being successfully funded, receiving higher total funding amounts, and attracting a higher number of project backers may differ when projects feature a creative product or service idea. We, therefore, hypothesize the following.

Hypothesis 3.
The inverted U-shaped relationship between a social orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative.

Hypothesis 4.
The inverted U-shaped relationship between an environmental orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative.

Data
We used data from the Kickstarter website that was collected by a web-crawling algorithm. The algorithm retroactively collected all successful and failed projects that went live on Kickstarter in 2018. To answer our research question, we focused on technology projects. Since Kickstarter makes special demands for projects posted in the technology category, such as a manufacturing plan and a delivery date for rewards, technology projects are more likely to become a long-lasting enterprise [91]. The focus on a single category also improves project comparability. Furthermore, we adhere to limiting the population of projects to those whose funding goal is at least 5000 USD. This funding level threshold addresses the problem that projects on Kickstarter have a wide range of funding goals and that the underlying motivation for a 100 USD project might differ from that of a project whose funding goal is 10,000 USD. We try to eliminate these discrepancies in project concepts by only including relatively large projects into the sample. In addition, focusing on relatively large projects allows us to better compare our findings to ventures funded through traditional financing options [27]. Consequently, we argue that the inverted U-shaped relationship between the level of social or environmental orientation and being successfully funded, receiving higher total funding amounts, and attracting a higher number of project backers may differ when projects feature a creative product or service idea. We, therefore, hypothesize the following.

Hypothesis 3.
The inverted U-shaped relationship between a social orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative.

Hypothesis 4.
The inverted U-shaped relationship between an environmental orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative.

Data
We used data from the Kickstarter website that was collected by a web-crawling algorithm. The algorithm retroactively collected all successful and failed projects that went live on Kickstarter in 2018. To answer our research question, we focused on technology projects. Since Kickstarter makes special demands for projects posted in the technology category, such as a manufacturing plan and a delivery date for rewards, technology projects are more likely to become a long-lasting enterprise [91]. The focus on a single category also improves project comparability. Furthermore, we adhere to limiting the population of projects to those whose funding goal is at least 5000 USD. This funding level threshold addresses the problem that projects on Kickstarter have a wide range of funding goals and that the underlying motivation for a 100 USD project might differ from that of a project whose funding goal is 10,000 USD. We try to eliminate these discrepancies in project concepts by only including relatively large projects into the sample. In addition, focusing on relatively large projects allows us to better compare our findings to ventures funded through traditional financing options [27].
We first drew a random sample of 1099 US-based technology projects from the Kickstarter website. We adjusted our sample for 44 projects, which falsely indicated a US location, but were actually based in China. Furthermore, we removed six projects for which the campaign data were incomplete, e.g., when the Kickstarter account was deactivated so that the gender of the entrepreneur could not be collected. This resulted in a final dataset of 1049 US-based technology projects.
In addition to the project descriptions, we collected the project video pitches to ensure that we capture the complete entrepreneurial narrative of the projects. Since video pitches are an important source of project information for backers, we extracted all linguistic information from the videos using otter.ai's automated voice recognition algorithm. To ensure the accuracy of the automated transcription, 60 video pitches were transcribed manually and compared to their AI-generated counterparts, resulting in a matching rate of over 90%. In other words, more than 90% of linguistic information was captured correctly by otter.ai. Afterward, all automated transcriptions were manually checked for inconsistencies, such as double entries or the incorrect capturing of special characters or proper names. We then merged the transcribed pitches with the projects' textual descriptions and general characteristics, such as project duration, total amount of capital raised, and number of project backers.
Collecting both the textual description and transcribed pitches addresses common methodological issues related to content analysis. Typically, crowdfunding studies that employ content analysis only use the textual description to measure their linguistic constructs, irrespective of the existence of a project video pitch. This approach could potentially entail measurement errors because the data collected are only a fraction of the information available to backers. In our sample, entrepreneurs used 388 words in their video pitches on average. Compared, the average textual description consists of 842 words. Thus, a large share of linguistic information goes unnoticed when disregarding video pitches. In total, we analyzed approximately 1.14 million words.

Dependent Variables
Our dependent variable is crowdfunding performance, which comprises different aspects of a successful crowdfunding project. Kickstarter works on an all-or-nothing basis: If a project does not reach its goal by the deadline, no funds will be collected. However, if a project matches or exceeds its goal, it receives all of the money pledged [92]. This results in different approaches to operationalizing crowdfunding success using an indicator if a project met its goal [93], complemented by the total amount of capital raised [7] and the number of individual investors [94]. We follow the approach of Anglin et al. [94] and operationalize the success of a crowdfunding project as crowdfunding performance and include the three aforementioned success aspects in our study. For this purpose, our dichotomous variable successfully funded indicates if a campaign at least met its preset goal. Successfully funded was coded as "1" for campaigns that met their goal during the duration of the campaign and as "0" otherwise. Second, the amount pledged accounts for the total capital raised during the campaign, irrespective of whether or not the preset funding goal was met. This continuous measure is important for our study because it differentiates between campaigns that did not generate any funding and campaigns that only just missed their preset goal by a narrow margin. At the other end of the spectrum, this variable considers if a campaign barely met the funding goal or "over performed" and raised far more than the funding goal. Since the amount pledged has a high degree of skewness, we used the natural log of the variable to correct for the influence of extreme outliers [95]. Lastly, the number of backers is a continuous variable for the number of individual investors supporting a project. Studies show that as opposed to a few large contributions, a higher number of backers with smaller contributions is crucial for crowdfunding success [96,97]. Since this is applicable across different crowdfunding platforms, the variable number of backers helps to generalize our findings for reward-based crowdfunding in general. Again, we transformed the variable using the natural log to correct for extreme outliers.
To estimate our models for the dichotomous variable successfully funded, we use logistic regression and report the coefficients as odds. For log of amount pledged and log of number of backers, we employ linear regression.

Independent Variables and Model Estimation Procedures
In this study, the independent variables pro-social orientation and pro-environmental orientation represent constructs that demonstrate the awareness of social and environmental values, such as social and ecological fairness and justice [11]. Our selection of the social and environmental categories aligns with prior literature about sustainability and meets the conditions to account for sustainable development [12]. In entrepreneurship research, according to Shepherd and Patzelt [13], the goal of sustainable development is to sustain communities (social dimension) as well as life support systems and nature (environmental dimension). Our study captures the social and environmental orientation as linguistic constructs using a content analysis algorithm that assesses the pro-social or pro-environmental framing of a project based on using cues expressing a pro-social or pro-environmental orientation. We employed the computer-aided text analysis (CATA) tool CAT Scanner [18]. CAT Scanner measures the salience of a construct based on the frequency of words, word stems, and phrases [98].
Using CATA has methodological benefits when analyzing large amounts of text because compared to manual coding, it minimizes the risk of coder fatigue and disagreement between coders [98]. To measure our independent variables, we used content analytic dictionaries developed by Pencle and Mălăescu [99]. In their study, the authors focus on developing dictionaries that capture the multidimensional construct of corporate social responsibility and validate these dictionaries using sustainability reports from initial public offering prospectuses. The dictionaries employee, human rights, social and community, and environment were generated deductively by synthesizing the frameworks of the major corporate social responsibility guidelines and inductively by assessing a subsample of sustainability reports. In detail, the dictionaries adopt the following definitions: Employee relates to the awareness and care of the organizations' internal stakeholders. Human rights consider the individual and collective rights of all stakeholders (e.g., minorities) and promote inclusiveness. Social and community focus on the needs of the local community, native communities, and developing society. Finally, environment relates to protecting natural resources and the responsible use of materials. We consider the dictionaries, which were initially designed for corporate social responsibility measures, to be suitable for our study. Pencle and Mălăescu [99] state that corporate social responsibility and sustainability are two deeply interwoven constructs, and they use the two terms interchangeably when referring to their sample. Furthermore, the authors made the dictionaries freely available and strongly encouraged researchers to refine or consolidate the dictionaries according to the needs of the study. We follow this recommendation and subsume employee, human rights, and social and community under the dimension social (the other being environment), and we detected and deleted all double entries among the dictionaries to prevent distortion from overlap. This refinement is to ensure the reliability of our measure in the new context [100]. As stated above, selecting these two categories is informed by the prevalent understanding that sustainability consists of a social and environmental dimension. However, to ensure the fit of the intended measures by our independent variables with the consolidated dictionaries, we utilized the detailed definitions of social and environmental sustainability from the 'Wellbeing of Nations: A Country-by-Country Index of Quality of Life and the Environment'. Here, Prescott-Allen [101] further specifies sustainable development beyond the term social and environmental and provides specific elements that, respectively, constitute the two dimensions. Table 1 lists the elements of the social dimension in detail and gives examples of words or phrases from the dictionaries that correspond to these elements. The last column gives the content analytic dictionary the examples originate. Following this approach, Table 2 features the elements that aggregate to the environmental dimension, again with examples stemming from the dictionaries.

Social Elements Dictionary Examples CATA Dictionary
Health and population "health benefits" "employee wellbeing" "benefit the masses." Employee Employee Social and community Wealth (e.g., household wealth) "employee welfare" "affordable housing" Employee Social and community Knowledge and culture "human development" "cultural preservation" "educational programs" Human rights Social and community Employee Community (e.g., freedom) "community projects" "civic engagement" "inclusiveness" Social and community Social and community Human rights Equity (e.g., gender equity) "gender diversity" "employee equity" Human rights Employee Table 2. Elements related to the environmental dimension and dictionary examples.

Environment Elements Dictionary Examples CATA Dictionary
Land "land conservation" "rainforest"

Environment Environment
Resource use "renewable energies" "energy efficiency"

Environment Environment
We consider that longer campaign narratives potentially increase the probability that social or environmental rhetoric is captured, which is why our independent variable may be endogenous. We follow prior content analysis studies and remove the influence of narrative length by dividing the amount of social or environmental rhetoric by the word length of the narratives [94]. This score is multiplied by 100, so a score of, e.g., 10 reflects that 10% of the language used relates to the social or environmental dimension.

Moderating Variable
Creativity accounts for developing products or services that are useful and novel [77]. However, understanding what is novel and useful varies significantly within and across cultures, leading to construct validity problems [102]. We aim to create an indication that takes different cultures and underlying understandings of creativity into account and, consequently, matches the global nature of the crowd that usually is geographically and socially distant from the project creators [7,103]. The project's creativity was assessed by two coders that attended the university's master's program and participated in entrepreneurship courses. We followed the approach of Loewenstein and Mueller [102] and used laypeople, since investment-wise reward-based crowdfunding backers are mainly lay people [104], but instructed the coders on the same implicit theory of creativity. We provided cues for a novel (e.g., paradigm shift, breakthrough, surprise) and useful (e.g., functional, intuitive) that are tried and tested to hold true for different cultures and understandings of creativity [102]. The coders assessed the project descriptions and visuals, such as videos and pictures, unaware if the campaign was successful or how much funding it received. We calculated Cronbach's alpha [105] to measure interrater reliability, which at 0.84 is above the commonly stated critical threshold of 0.7 and leads us to assume that the coders had a common understanding of creativity [106]. The campaigns the coders disagreed on were put to discussion and re-evaluated until they agreed on a final judgment.

Control Variables
With a longer duration decreasing the chances of success, we control the number of days a project is live for funding on the Kickstarter website [27]. Furthermore, we included a dichotomous variable whether the project featured a prototype gallery or not. A prototype gallery is a series of photos that technology projects can integrate above their project description. This allows the current state of a project to be presented and signals transparency [7]. We control for differences in project goals since prior empirical work has shown the influence of funding goals on crowdfunding success [107]. Cordova et al. [108] find that a higher funding goal level of a given project decreases its probability of being funded successfully. Moreover, backers draw important information from funding goal levels and use them as a decision-making tool [107].
As backer behavior is affected by the gender of the project initiator [109], we indicate if the project initiator is male or female. For instance, Greenberg and Mollick [110] examined if gender-based differences in financing are present, as women are disadvantaged when trying to access external funding sources. We account for the role of external social capital in crowdfunding and include whether the Kickstarter account of the project initiator is connected to Facebook [27]. Being connected to social networks has been shown to be an effective proxy for social capital in the form of third-party ties in online communities. Extant studies find evidence that linking the campaign to social networks increases funding success probability [111,112].

Empirical Results
We tested Hypotheses 1-4 by constructing a series of logistic and linear regression models with crowdfunding success as a dependent variable. To account for multicollinearity, we assessed the values of our correlation coefficients and the variance inflation factors [113]. Table 3 provides the descriptive statistics for our sample, and Table 4 presents the variance inflation factors (VIF) and Pearson's correlation coefficients for the independent variables. With all pairwise correlation coefficients below 0.2 and the variance inflation factors ranging from 1.01 to 1.04, we conclude that multicollinearity does not affect our regression models [114].  This table presents the variance inflation factors (VIF) and the pairwise Pearson's correlation coefficients of the independent variables used in the regression models. Note: SO = sustainability orientation.
Hypothesis 1 proposed that a social orientation will exhibit an inverted U-shaped relationship with crowdfunding performance. To test hypothesis 1, we entered a linear term and a quadratic term of the independent variable social to the logistic and linear regression equations. As we proposed in our hypothesis, we could observe positive coefficients for the linear terms and negative coefficients for the squared terms, all statistically significant at the 1% level. This suggests that a pro-social orientation in crowdfunding projects increases the probability of being funded successfully, receiving more pledges, and mobilizing a higher number of backers up to a certain point. After this optimal point, any further increase in pro-social orientation diminishes entrepreneurs' likelihood of crowdfunding performance. We expected this result since backers respond positively to social-oriented products or services and proactively support projects that feature a pro-social orientation. However, promoting very high levels of social orientation creates skepticism among backers, who start to doubt the credibility of the social claims and feel that important product or service features are missing. Figure 3a-c provides plots of the relationships for successful funding, amount pledged, and number of backers and are consistent with the proposed inverted U-shaped relationship. Hence, hypothesis 1 is supported. To interpret our results, we used the marginal effects to calculate the turning points. This allows us to define the level of social orientation that is most beneficial and also aims to provide a threshold beyond which pro-social orientation reduces the respective dependent variable. The turning point for the dependent variable successful funding is at a score of 1.5, for the amount pledged at 1.1, and for the number of backers at 1.4. For a crowdfunding campaign, this means that promoting a pro-social orientation is only advantageous to an optimal score of 1.5, that is, 1.5 words associated with promoting a pro-social orientation per 100 words. From this point, promoting a pro-social orientation anyfurther diminishes the probability of successful funding. Respectively, this applies to receiving pledges at a score of 1.1 and mobilizing backers at a score of 1.4.
Analog to the social dimension, Hypothesis 2 proposes that a pro-environmental orientation will exhibit an inverted U-shaped relation with crowdfunding performance. Our results show positive coefficients for the linear term and negative coefficients for the squared terms, all significant at the 1% level. The plots, as illustrated in Figure 3d-f, indicates that the results align with the proposed U-shaped relation. Thus, Hypothesis 2 is supported. The turning points for the dependent variables are at the following scores: successful funding: 1.74; amount pledged: 3.1; the number of backers: 3.0. Concerning the dependent variable of successful funding, the results indicate that promoting an environmental orientation evokes a positive relation up to a score of 1.74. In line with this argument, the optimal point for receiving pledges is at a score of 3.1 and for mobilizing backers at a score of 3.0. We expected this result for environmentally sustainable products or services, which backers view as more desirable compared to "non-green" options. Yet, overemphasizing a pro-environmental orientation is problematic because backers will not believe the claims and fear that the product or service does not fulfill standard functionalities.
Compared to the social orientation, where the optimal emphasis ranges between scores of 1.1 and 1.4, backers attach more relevance to the emphasis of an environmental orientation, which is reflected in scores ranging between 1.74 and 3.1. However, our sample shows that entrepreneurs put considerably less emphasis on environmental orientation, on average approximately 0.8 words per 100 words. In contrast, entrepreneurs use on average 1.48 words associated with social orientation per 100 words, which is roughly what we identified as the optimal amount of emphasis.
Tables 5-7 provide the results of our hypothesis tests for our dependent variables successful funding, log of amount pledged, and log of number of backers, respectively.  This table presents the results of the logit regression models to examine the effects of pro-social and pro-environmental orientation used in crowdfunding narratives on being successfully funded. Standard errors are in parentheses. The symbols *, **, and *** denote statistical significance at the 10%, 5%, and 1% levels, respectively. Note: SO = sustainability orientation.  This table presents the results of the linear regression models to examine the effects of pro-social and pro-environmental orientation used in crowdfunding narratives on the amount pledged. Standard errors are in parentheses. The symbols **, and *** denote statistical significance at the 5%, and 1% levels, respectively. Note: SO = sustainability orientation.   Hypothesis 3 proposed that the inverted U-shaped relationship between a pro-social orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative. We can support Hypothesis 3 as the interaction of social orientation and creativity has positive coefficients for the linear term and negative coefficients for the squared term, all significant at the 5% level. Moreover, Figure 3g-i is consistent with the idea that the inverted U-shaped relationship between the level of pro-social orientation and crowdfunding performance differs for projects demonstrating a creative product or service idea. Our results indicate that skepticism about the credibility of the social claims is indeed intensified if the project does not demonstrate a creative product or service idea since backers doubt that the entrepreneur can manage the complex tasks accompanied by a pro-social orientation. This results in a weakening of the inverted U-shaped relationship, as shown in Figure 3g-i. However, if the project or service idea is creative, backers' skepticism about the social claims is reduced as they believe that the entrepreneur is capable of simultaneously addressing multiple conflicting outcomes that emerge with a pro-social orientation. The inverted U-shaped relationship is thus steepened, which is illustrated in Figure 3g-i.
We find no statistical significance to support Hypothesis 4, which proposed that the inverted U-shaped relationship between an environmental orientation and crowdfunding performance will be flattened when the project or service idea is not creative and steepened when the project or service idea is creative. that the entrepreneur can manage the complex tasks accompanied by a pro-social orientation. This results in a weakening of the inverted U-shaped relationship, as shown in Figure 3g-i. However, if the project or service idea is creative, backers' skepticism about the social claims is reduced as they believe that the entrepreneur is capable of simultaneously addressing multiple conflicting outcomes that emerge with a pro-social orientation. The inverted U-shaped relationship is thus steepened, which is illustrated in Figure 3g  (e) (f)

Discussion
In recent years, online-based crowdfunding has emerged as a valuable alternative funding source for projects and new ventures [115]. Entrepreneurs present projects on crowdfunding platforms where potential project backers can provide funding. With our study, we enhance the understanding of whether a pro-social or pro-environmental orientation affects success in the context of reward-based crowdfunding. We hypothesized and empirically found that the level of pro-social or pro-environmental orientation has an inverted U-shaped effect on crowdfunding performance. This finding aligns with our theoretical conceptualization that the overall impact of a pro-social or pro-environmental orientation is the net outcome of the respective latent positive and latent negative effects of these orientations. The results suggest that entrepreneurs need to delicately balance a pro-social or pro-environmental orientation and find the "right" level of emphasis to create a competitive advantage over comparable crowdfunding projects that feature conventional products or services. Moreover, we demonstrate that the inverted U-shaped relationship between the level of a pro-social orientation and crowdfunding performance differs between projects that do not demonstrate a creative product or service idea and those that do. As we conceptualized, backers seem to feel less skeptical of pro-social claims if the product or service idea is creative. For the same level of pro-social orientation, this means a higher funding probability, more pledges and more backers compared to "non-creative" projects.
Our results challenge prior observations by Cholakova and Clarysse [3], who surveyed participants in an experimental setting on their pledging and funding motivation and captured their altruistic behavior. The authors find no evidence that the decision of backers to pledge is pro-socially motivated. However, measuring different levels of pro-social orientation unravels that promoting such an orientation can indeed be beneficial up to a certain point. Furthermore, Hörisch [10] suggests that indicating a proenvironmental orientation can be detrimental to the success of reward-based crowdfunding projects. Since we provide a more fine-grained view by capturing different levels of pro-environmental orientation, our results indicate that a moderate amount increases the probability of crowdfunding performance.
We must acknowledge the limitations of our methodological approach. We capture the social and environmental orientation as linguistic constructs using content analytic dictionaries designed to measure the multidimensional construct of corporate social responsibility. Sustainability and corporate social responsibility are two deeply interwoven constructs, and we refined and consolidated the dictionaries according to Pencle and Mălăescu [99] to ensure that they fit our study. Yet, the dictionaries lack external validation, which is why we cannot guarantee the reliability of our measure in the new context [100].
Another limitation concerns the indication of creativity. We used laypeople as coders that were instructed on the same implicit theory of creativity and decided whether a project is creative or not. Although we believe that our approach matches the multifaceted and international nature of participants in crowdfunding, we are aware of the fact that creativity comes in more forms than zero and one, which we did not account for in our study.
Lastly, we aimed to use the textual information available and analyzed the crowdfunding campaign descriptions and the transcribed video pitches. However, there is text that we did not capture with the web-crawling algorithm, especially text that is embedded in graphics and text from the comment and update section of Kickstarter. Thus, our findings are limited by potential measurement errors because we did not capture the entire information basis.

Conclusions
We employ linguistic framing as outlined by social movement theorists to examine how the selection is enabled and action is guided in reward-based crowdfunding. In detail, we shed light on how promoting a sustainability orientation affects crowdfunding performance. These results suggest that the level of pro-social or pro-environmental orientation has an inverted U-shaped effect on crowdfunding performance.
In this context, we contribute to the existing literature on sustainability in rewardbased crowdfunding in three main ways. First, we evaluate the effect of pro-social and pro-environmental orientation on crowdfunding performance separately. While previous researchers blend the two dimensions to account for sustainability [6,8,9], our results demonstrate that backers differentiate between a pro-social and a pro-environmental orientation. These findings align with Catlin et al. [24], who show that consumers perceive the social and environmental dimensions of sustainability as distinct. Thus, our study helps to better understand the pledging behavior of backers and contributes to the literature concerned with factors that determine crowdfunding success.
Second, we improve the methodological approach to measure pro-social or proenvironmental orientation in crowdfunding narratives. For instance, Calic and Mosakowski [7] indicate a social or an environmental orientation through linguistic cues. Yet, they only use these cues as an indicator to label campaigns to the social entrepreneurship category and consequently measure a pro-social or pro-environmental orientation through a binary variable. Defazio et al. [9] address this shortcoming and measure the frequency of the cues to observe the level of a pro-sustainability orientation. However, they do not integrate video language from the entrepreneur's pitch in the linguistic analysis for their entire sample. Accordingly, the data basis differs from the information basis a potential project backer considers, leading to measurement errors as the actual amount of emphasis is not captured. Thus, we collect both the textual description and transcribed pitches to address common methodological issues related to content analysis following the advice of McKenny et al. [100].
Third, we support the findings of previous studies about the effects of entrepreneurs' language on resource mobilization. In conformity with Parhankangas and Ehrlich [61] and Defazio et al. [9], our results confirm that the emphasis on language related to sustainability must be delicately balanced when entrepreneurs try to secure funding. In this regard, we add to the discussion that backers approve more emphasis on the pro-environmental orientation than on the pro-social orientation and that the appreciation of a pro-social orientation depends upon the creativity of the product or service idea. Accordingly, we advance the understanding of how the relationship between linguistic framing and the outcome of a crowdfunding project depends upon the product or service characteristics under which the framing is initiated and implemented.
Our study opens up opportunities for future research. We evaluated the influence of a pro-social or pro-environmental orientation on crowdfunding performance for crowdfunding projects that went live on Kickstarter in 2018, which is a relatively short period. We would be interested in longitudinal studies that demonstrate how the perception of sustainability changes over time. This would provide insights into how, e.g., a rising awareness of social problems in the general public is reflected in the level of pro-social emphasis in crowdfunding projects and if backers' response changes according to that. Longitudinal studies would thus broaden our understanding of consumers as a "different kind of investor" [2]. In this vein, an interesting avenue for future research would be comparing professional early-stage investors, such as business angels, with crowdfunding backers regarding their sensitivity towards sustainability. Future research is also challenged with replicating our findings in equity-based crowdfunding. We would like to see how a pro-social or pro-environmental orientation affects performance in a setting where backers are more driven by financial motivations and invest money for the long term [116]. In this context, it would be interesting to highlight how creativity affects this relationship. Lastly, our transcription approach "boils down" the video pitches to just language. Since the pitch setting typically appears to be the entrepreneur presenting the product or service in front of a camera, valuable body language and paralinguistic cues, such as mimics, tones, or gestures, are lost. We consider this information valuable as it could either accentuate or trivialize a pro-social or pro-environmental orientation.
Future studies could include body language and paralinguistic cues into their sample and use AI-generated coding to empirically assess their influence on crowdfunding performance. Our study also has practical implications. Entrepreneurs may use our findings to better craft their campaign narratives. Framing a crowdfunding project as pro-social or pro-environmental can be beneficial, especially for crowdfunding projects that promote a pro-social orientation and feature a creative product or service idea. Yet, entrepreneurs should be aware that overemphasizing a pro-social or pro-environmental orientation may backfire and reduce the probability of crowdfunding performance. In this vein, crowdfunding platforms could use the results of this study to guide entrepreneurs when advising them on how to create an appealing entrepreneurial narrative for both the textual description and the video pitch. This may contribute to sustainable crowdfunding projects being more successful and thus develop their full potential for society and the environment. Data Availability Statement: The data in this study is publicly accessible on Kickstarter and was retrieved in accordance with Kickstarter's terms of use.

Conflicts of Interest:
The authors declare no conflict of interest.