Shark bait; do investors on television’s Shark Tank discriminate
against African-Americans?
Master’s Thesis Ignacio Giménez
S3444147
MSc BA Small Business and Entrepreneurship June 25, 2018
Supervisor: Dr. Florian Noseleit Co-Assessor: Dr. Aard Groen
Abstract
Racial inequality in the United States continues to hamper the advancement of African-Americans. One field where the gap is notable is entrepreneurship, with the African-American self-employment rate half that of whites (Gold, 2016). This is a concern because entrepreneurship is an effective way for
communities to improve their economic lot (Fairlie & Robb, 2008). A lack of access to startup capital is a major contributor to the underrepresentation of African-Americans as entrepreneurs, and discrimination against African-Americans has been shown in credit markets and on crowdfunding platforms (Bewaji et al., 2015; Younkin & Kuppuswamy, 2017). Whether African-Americans face the same discrimination from private investors, however, is an open question. Using six seasons of the popular American television program Shark Tank, the investment decisions of the Sharks, a panel of investors who invest their own money in pitch teams who present on the show, are analyzed for evidence of racial
discrimination. The show provides a novel context for searching for signs of discrimination, as the process used on the show closely mirrors investing by private investors in real life, albeit that on Shark Tank, the decisions made by the investors are public. The analysis of the data shows no evidence of discrimination. Support is provided for the theories that this is due to financial incentives, a robust screening process of the pitch teams, and the public nature of the investment decisions made on the show.
1. Introduction
Since the Civil Rights movement of the 1960s, significant steps have been made in the United States towards achieving equality for African-Americans, but as studies in multiple disciplines show, there is still work to be done (Gold, 2016; Pew Research Center, 2016). African-Americans continue to face inequality in a wide range of fields, such as employment, education and housing (Pew Research Center, 2016). In the business world, the situation is the same. Self-published data from Google, Yahoo, Facebook and Apple show that African-Americans are underrepresented at those leading companies, making up just 2 percent of the workforce (Kang, 2015). African-Americans are not only
underrepresented at top companies, however. In the field of entrepreneurship, a similar picture emerges; the rate of self-employment among African-Americans is less than half of that of whites (Gold, 2016). This is a concern because entrepreneurship has been used as an effective vehicle for economic
improvement for communities and ethnic groups (Gilbert et al., 2004; Fairlie & Robb, 2008). One driver for the gap in entrepreneurship between blacks and whites in the United States may be lack of access to startup capital for new ventures. Entrepreneurship literature has shown that access to financial capital in a firm’s early stages is crucial for firm success (Bewaji et al., 2015). For both angel investors (AI) and venture capital funds (VC), the personal characteristics of the entrepreneur have been found to be
influential in the funding decision, and race, a visible personal characteristic, may play a role (MacMillan, 1985; Hsu, 2007; Greene, et al.; 2001, Rajan, 2012; Paul et al., 2007; Feeney et al., 1999). Research shows that blacks are disadvantaged in gaining access to startup capital, and thus their ability to successfully launch and grow new ventures is similarly curtailed (Bewaji et al. 2015; Gold, 2016; Younkin and Kuppuswamy, 2017).
discrimination is determining where discrimination is present. Then, the specific context can be analyzed and suggestions for improvements can be made.
The U.S. television show Shark Tank provides an opportunity to search for evidence of
discrimination against African-Americans by investors making funding decisions. On the show, business owners ask for an investment from a panel of “Sharks”, investors who choose whether to invest their own money in the proposed venture (About Shark Tank). This context is relevant for studying investor
behavior because it has both important similarities but also one major difference with the process by which investors make investment decisions in real life. Researchers who have analyzed data from the show point out that the process through which pitch teams seek investments on Shark Tank is broadly similar to how investments decisions are made in real life (Poczter & Shapsis, 2018). However, VCs and AIs in most cases are not compelled to make their decisions public, and this may affect the decisions that they make. For that reason, other researchers have indicated that Shark Tank is most similar to other public pitch competitions (Smith & Viceisza, 2018). The Sharks may show evidence of discrimination in their investment decisions if they are influenced by taste-based, statistical or implicit bias, as many investors are (Younkin & Kuppuswamy, 2017). On the other hand, there are some reasons to think the Sharks will not discriminate in their funding decisions. First, the desire to enjoy high returns on investments might motivate the Sharks to look past any conscious or unconscious bias; research into discrimination by investors against female entrepreneurs suggests that the financial incentives can be sufficient to discourage discrimination (Buttner & Rosen, 1989). This may be the case for African-Americans as well. Second, under the microscope of television audiences in the United States, a racially-conscious society, they may racially-consciously strive to eliminate any traces of discriminatory preferences from their investing decisions. The aim of this thesis, then, will be to determine whether the investment decisions of the Sharks are discriminatory against African-Americans. If so, the literature on anti-discrimination measures will be used to suggest corrective mechanisms. If not, the structure and context of the show will be analyzed to determine which elements may help in eliminating discrimination, and how these can be applied in other contexts to reduce the deleterious effects of discrimination in new venture funding decisions on African-Americans.
Therefore, this thesis will seek to address the research question of whether angel investors viewing in-person pitches discriminate against African-American entrepreneurs. Additionally, if evidence of
discrimination is found, the thesis will seek to address what measures can be taken to reduce it. If evidence of discrimination is not found, the thesis will seek to address what anti-discrimination
open question whether racial inequalities are simply the vestiges of the discrimination of years past, or whether discrimination is ongoing (Levitt, 2003). Therefore, determining if and where discrimination is perpetuated is important. The question remains open in part because measuring discrimination is hard. There is a stigma attached to racist attitudes, and because of this, research subjects are unlikely to report them (Levitt, 2003). How investors make decisions and the influence of the personal characteristics of the entrepreneurs that they invest in has also been a subject of considerable study. This is complicated by the fact that data on investment decisions by AIs and VCs is hard to come by, due mostly to the private nature of these investors and their aversion to publicity (Poczter & Shapsis, 2018). Shark Tank provides a rare opportunity to examine both issues. Until now, the question of whether AIs discriminate based on race remains unanswered, and this study aims to contribute to that knowledge.
2. Literature Review
This section provides a review of the literature surrounding racial inequality in the United States, black underrepresentation in entrepreneurship and the potential drivers of that underrepresentation.
Racial Inequality in the United States and in Entrepreneurship
Much research has been focused on the gaps between whites and African-Americans in the United States. The disparities are present across fields that touch virtually every aspect of life. A 2016 Pew Research Center report provides some perspective on the problem. In terms of income, the gap is significant; in 2014, the median adjusted income for African-American households was $43,300, compared with $71,300 for white households (Pew Research Center, 2016). African-Americans are also more than twice as likely to live in poverty (Pew Research Center, 2016). These gaps translate to wealth generation; African-Americans in the United States have levels of wealth one-eleventh that of whites (Robb & Fairlie, 2007).There is also a gap in entrepreneurship and self-employment; 11% of whites in the United States are self-employed, while only 5.1% of African-Americans are (Gold, 2016). Fairlie and Robb (2008) find that African-Americans are substantially less likely than whites and Asian-Americans to own a business. Unfortunately, there is also a performance gap between African-American owned firms and those owned by other ethnic groups. African-American owned businesses have lower sales and profits, employ fewer people and have higher closure rates than white-owned businesses (Fairlie & Robb, 2008).
Praag & Versloot et al., 2007). Entrepreneurship can positively impact communities through direct and indirect effects. High-quality new businesses positively impact employment growth directly, through the new jobs they create, and well as indirectly, by increasing competition and productivity (Fritsch &
Noseleit, 2012). The positive effect of entrepreneurship has not gone unnoticed in the United States. Since the 1980s, local, state and the federal government have shifted their policy focus from regulating big businesses to promoting entrepreneurship. Recognizing the ability to create industry clusters and reap the benefits of spillover, these governments looked to create policies that would foment the founding and development of new businesses and benefit their constituent economies (Gilbert et al., 2004). Beyond government initiatives, communities have also rallied around themselves to lift themselves up through entrepreneurship. The success of the Asian-American community in the United States if an oft-cited example. The Asian-American business community is widely considered the most successful example of minority entrepreneurship in the United States (Bates & Dunham, 1993). Despite often being newer arrivals to the country, Asian-American owned businesses have outperformed those owned by other minority groups (Fairlie & Robb, 2008). Some studies have shown that businesses owned by Asian-Americans even outperform white-owned businesses (Bates & Dunham, 1993).
Why then, have some regions and communities been able to reinvent themselves through
entrepreneurship, while others have not? Sociologists have focused on ethnic ties, cultural norms and community resources to explain why Asian-Americans, have outperformed other minority groups in the field of entrepreneurship (Bates, 2011; Bates & Dunham, 1993). Economists, however, have focused on resource utilization to explain the gap (Bates, 2011). Several studies have found that the gap in
entrepreneurial achievement between African-Americans and other groups is driven by a lack of access to capital (Fairlie & Robb, 2008; Robb & Fairlie, 2007; Bates, 2011; Blanchflower et al., 1998). In a 2015 study on funding accessibility, Bewaji et al. (2015) analyzed the difficulties faced by minorities in the United States in receiving new-venture funding. They point out that access to financial resources is crucial for entrepreneurs and is a predictor of new-venture performance, and that minorities are constrained in accessing startup capital. Gold (2016) also supports the view that access to financial resources is necessary for entrepreneurs to succeed, that African-Americans are disadvantaged in terms of access to these resources, and that with greater access to financial resources African-American
businesses. They found that Black-owned small businesses are three times more likely to have a loan request denied and that Black-owned firms pay higher interest rates. Robb and Fairlie (2007) also found that African-Americans have constrained access to capital and face discrimination in lending. This, in turn, leads to lower rates of African-American business ownership, higher failure rates and lower
employment (Robb & Fairlie, 2007). While cultural difference may play a role, the evidence shows that a main driver of the gap in entrepreneurship between African and Asian-Americans is a lack of access to startup capital (Fairlie & Robb, 2008). While discrimination in credit markets has been shown in multiple studies, as seen above, whether AIs and VCs are subject to the same prejudices is an open question.
The Role of Personal Characteristics in Investor Decision-Making
emphasis on entrepreneurs’ personal characteristics, and use those characteristics as indicators of the quality of the investment. As such, the possession or lack of key characteristics can make or break entrepreneurs’ chances of receiving investments. For African-Americans, their race is a visible personal characteristic that some investors may use to gauge the quality of their companies. Younkin and
Kuppuswamy (2017) find that African-Americans are less likely to receive funding for new ventures, and explain the mechanisms that may lead to their projects being perceived as lower quality. Investors may be motivated by statistical discrimination; if they believe African-American-led companies are less likely to grow quickly or be acquired at a high valuation, they may avoid investing in them to maximize their returns. Alternately, they may be motivated by taste-based discrimination; since most investors are white men, they may prefer investing in founders of their own race, even if this results in economic
inefficiencies. Finally, the discrimination may be a result of implicit bias, which effects decision-making at an unconscious level and may occur without the investors’ intention. The potential sources of
discrimination are elaborated upon in the following section.
Models of Discriminatory Behavior
stereotypes are deeply embedded in American culture, they can be used to justify discriminatory hiring decisions. Some studies have attempted to determine whether taste-based or statistical discrimination is the driver of discrimination in markets. Carlsson and Rooth (2012) examined hiring decisions in Sweden and determined that discrimination against minorities was due to taste-based, and not statistical
discrimination. Magrom et al. (2016) found evidence of discrimination on the crowdfunding site
Kickstarter and used a survey to determine that this was due to taste-based, not statistical discrimination. Determining the source of discrimination, however, is outside the scope of this study, which aims only to determine whether discrimination exists. In all cases of taste-based or statistical discrimination one assumption is constant; the decision-maker makes a conscious decision to discriminate.
In contrast, implicit bias affects our actions on an unconscious level (Kirwan Institute for the Study of Race and Ethnicity, 2017). Because of this, implicit bias can lead us to act in ways which directly
contradict our beliefs (Greenwald & Krieger, 2006). Researchers interested in the topic have uncovered the discriminatory effects of implicit bias in a multitude of real-world situations, including criminal justice, education, healthcare and housing (Kirwan Institute for the Study of Race and Ethnicity, 2017). Research into implicit bias has been aided by the development of the Implicit Association Test (IAT). The IAT measures implicit bias in individuals through a computer-based test and has been shown to
accurately measure implicit bias (Bertrand et al., 2005). Findings from the use of the test show that most Americans display a pro-white, anti-African-American bias (Kirwan Institute for the Study of Race and Ethnicity, 2014). As Greenwald and Krieger (2006) state, the evidence shows that implicit bias is widespread in American society, and is associated with discrimination against African-Americans. Increased implicit bias in physicians was found to be associated with a higher likelihood of providing cardiovascular treatment to whites, and lower likelihoods of providing the same treatment to blacks (Green et al., 2007). Implicit bias was also found to affect hiring decisions, with implicit bias being associated with less favorable judgements of minority job applicants (Perrewe et al., 2006). Evidence of the effect of implicit bias has also been shown in day-to-day interactions; individuals with more implicit bias were found to speak longer, smile more and have more social interactions with whites as compared to African-Americans (Jolls and Sunstein, 2006). Studies have shown that even African-Americans display pro-white, anti-African-American implicit bias. Despite this, they display the bias at much lower rates (34%) than whites (71.5%) (Greenwald & Krieger, 2006).
The literature shows that discrimination against African-Americans is pervasive in many fields, including the funding of new ventures. The problem is deeply entrenched, due in part to the varied
crowdfunding campaigns (Bewaji et al., 2015; Younkin & Kuppuswamy, 2017). Because racial inequality is still prevalent in the United States, including in new-venture funding, it is expected that
African-Americans seeking funding on Shark Tank will also be faced with discrimination. This leads the following hypotheses:
H1a: Shark Tank pitch teams with African-American members are less likely to receive an offer for funding from the Sharks than teams with no African-American members
H1b: Shark Tank pitch teams with African-American members receive fewer offers for funding from the Sharks than teams with no African-American members
H1c: Shark Tank pitch teams with African-American members are less likely to receive an investment (i.e. accept an offer) for funding from the Sharks than teams with no African-American members
Racial and Social Issues in the Spotlight
Shark Tank was first aired in 2009 and has run for 9 seasons, the most recent concluding in February 2018. During that time, America experienced the presidency of its first black President, Barak Obama, the emergence of the Black Lives Matter (BLM) movement and an increased focus on racial equality. In his 2016 New York Times article, “The State of Race in America”, Charles M. Blow writes that in the United States, the issue of racial inequality is as pressing as it has ever been. The rise of social media has meant that activists are able to rally around and draw attention to issues, whereas in the past they would not have had access to the public sphere (Carney, 2016). This can be illustrated by the rise of the BLM movement. BLM was born in 2012, after the shooting death of teenager Trayvon Martin by George Zimmerman, a member of the neighborhood watch in his Sanford, Florida community. The death of Trayvon Martin became a rallying point for activists on social media, and using the Twitter hashtag #BlackLivesMatter, activists initiated the BLM movement, seeking to draw attention to a perceived tolerance in American society for the killing of African-Americans (Carney, 2016). The movement picked up steam in 2014, after the shooting death of Michael Brown and subsequent protests in Ferguson, Missouri (Carney, 2016). Again, Twitter served as the meeting point for like-minded activists to reignite interest in the movement (Carney, 2016). The conversation surrounding race reached a crescendo, according to Blow, in 2015, when literature, film, television, and newspapers were dominated by the discussion of continued racial inequality in the country and how to fix it (Blow, 2016).
Valley (Kang, 2015). Beyond the need for diversity within companies, Smith also discussed inequality in entrepreneurial funding; “three percent of venture funding is going to women and less than one percent to people of color…we need to support VCs to overcome their biases” (Kang, 2015). As shown by Smith’s comments, race in America continues to be a front-page issue, and ameliorating racial inequities is a priority for many in the country, including in the business community. Furthermore, the need for business owners and investors to consider the societal impact of their actions has recently received increased focus in the investing world. In 2018 Laurence D. Fink, founder and chief executive of BlackRock, one of the world’s foremost investment firms, penned a letter on the subject. Fink indicated that going forward, BlackRock would look to invest only in firms that were making positive social contributions, and that society is looking to the business world to contribute to solving social problems (Sorkin, 2018). This is demonstrative of a shift in focus in the United States, from investing purely for financial reasons to considering social goals alongside financial ones when making investment decisions. Another example is of the prominent investor Melinda Gates, wife of billionaire Bill Gates. Gates has put her money where her mouth is, with a strategy of supporting VC funds that invest in women and minority companies, as well as funds that are run by women and minorities (Bort, 2018). She considers that directing capital towards these funds is not only beneficial socially, but can lead to profitable outcomes for her as an investor, as the businesses they support may be missed by traditional funds. Taking a shot at the entrenched investing class in the United States, she says that they are overly enamored with businesses run by a “white guy in a hoodie” (Bort, 2018). While it remains to be seen whether Gates’ and Fink’s strategies will pay off, one thing is clear; investors are increasingly willing to publicize their dedication to social issues, including ameliorating gender and racial equality.
The focus on social issues, particularly equality-related issues, is not restricted to activists and investors, however. Increasingly, public figures in the United States are incentivized not to create racially unequal outcomes. This is the case especially in the entertainment industry, where the public nature of the business allows for quick recognition of any racial disparities and where pressure to correct these
expected that any discriminatory effect found in the Sharks’ decisions will be reduced over time. This leads to the following hypotheses:
H2a: Over time, the effect of discrimination on the likelihood of pitch teams with African-American members receiving an offer from Shark Tank investors will be reduced
H2b: Over time, the effect of discrimination on the number of offers received by pitch teams with African-American members from Shark Tank investors will be reduced
H2c: Over time, the effect of discrimination on the likelihood of pitch teams with African-American members receiving an investment (i.e. accepting an offer) from Shark Tank investors will be reduced
Conceptual Model
The discussion above leads to the following conceptual model:
3. Methods
This section describes the methods used to collect data, the independent, dependent and control variables used, and the statistical methods employed to test the hypotheses. The reliability and validity of the study are also discussed.
Data Collection
Data for the analysis of the hypotheses was collected from the television show Shark Tank. Shark Tank is produced by ABC in the United States, and is the American version of a popular reality show format first created in Japan. The format was later popularized throughout the world with versions in several different countries, such as the BBC’s Dragon’s Den in the UK. On the show, entrepreneurs pitch
Presence of an African- American on the pitch team
Likelihood of obtaining funding
Time -
their businesses to the Sharks, and ask for an investment in the form of equity in exchange for cash. The entrepreneurs can pitch existing businesses or business ideas, and may pitch in teams or individually. The Sharks are a panel of 5 investors who can invest their own money in the ventures. The pitches typically last about an hour, but are edited down to shorter segments for television (Feloni, 2016). During the pitches, the Sharks can ask questions, make offers or decline to invest in the business. The Sharks may renegotiate the terms of the investment, but must make an investment at minimum equal to the cash amount requested by the entrepreneur. The Sharks may take equity positions, make loans, or request royalties in exchange for the cash investments they provide. The pitch is over once the business-owner accepts an offer, or all the Sharks decline to invest. A typical episode of the program is 42 minutes long and contains 4 pitches. As of the 2017-2018 season, the Sharks have invested over $100 million
combined through the program (About Shark Tank). This means that while the context for this study is a reality television program, the data supplied in this context can be analyzed to answer questions about investor behavior, because the investment decisions made on the show are real. The Sharks are often quick to point out that they are also valuable coaches and mentors for the businesses they invest in. Barbara Corcoran, for example, hosts periodic meetings with the entrepreneurs she has invested in through the show, creating a network for them to help each other and get advice from her. Similarly, Lori Greiner looks to have the products she invests in featured on the home shopping network QVC and arranged for them to be displayed together in the major American retailer Bed Bath & Beyond. This combination of funding and coaching is the same method used by VCs, who maximize their investments through a combination of scouting, i.e. only investing in companies with good chances for success, and coaching, i.e. providing additional management skill to increase company performance (Baum & Silverman, 2004). Since the Sharks invest their own money and behave similarly to investors outside the program, their decisions on the program can be analyzed to answer questions about investor behavior.
Dependent Variables
they do accept the offer, however, an agreement is made for the investment to be completed, subject to further negotiation (Feloni, 2016). For the purposes of this study, the results of the later negotiations are not available, so if a pitch team accepts an offer from a Shark, it is considered that an investment has been made and the DummyInvestmentMade variable is coded accordingly. The variable is coded as 1 if the business-owner agrees to receive funding from a Shark, and coded as 0 if the business-owner does not receive any offers or receives one or more offers but does not accept any.
Independent Variable
The independent variable concerns whether the pitch team contains an African-American member. A binary variable, DummyBlack, indicates whether the pitch team contains an African-American. The variable is coded 1 if the pitch team contained an African-American and 0 if not. Since pitches can be made by individuals or teams, another variable, CountBlack, indicates how many African-Americans are present on the team. This is a discrete, numerical variable ranging from 0 to an upper limit specified by the number of team members of any race that are on the team in question. Since the show takes place in the United States, with American business-owners and Sharks, the definition of African-American used is the common African-American definition; any person with any African-African-American ancestry (Davis, 1991). It is important to note the definition used, because as the sociologist F. James Davis writes, the category of African-American in the United States is socially constructed. The United States has traditionally used the “one-drop rule”, meaning that any person with any known African-American ancestry is considered African-American. This means that individuals with different backgrounds and appearances are all considered African-American (Davis, 1991). This categorization system is unique to the United States and contrasts with racial categories used in other countries, and is thus important to note.
Control Variables
as a control variable because investors may prefer pitches made by men over those made by women. For example, when analyzing the results of pitch competitions, Brooks et al. (2014) found that investors preferred pitches made by men over those made by women, even when the content of the pitch was the same. The size of the pitch team was also recorded using the variable CountTeamMembers. The number of team members was recorded as a discrete, numerical variable. The size of the team is of interest because it may affect the Sharks’ perception of the pitch. Chowdhury (2005) writes that the demands of running a new venture are better managed by an entrepreneurial team instead of a lone entrepreneur. The details of the business-owners request to the Sharks were also recorded and subsequently their impact on the dependent variable was tested through regressions. These are the equity stake offered, the amount of money requested in exchange for the equity stake and the valuation of the company, which is derived by dividing the amount of money requested by the decimal conversion of the equity stake offered. The variables are named EquityOffered, CashAmountRequested and Valuation, respectively. The use of these control variables follows Wanner’s (2018) research using Dragon’s Den, the UK version of the show. Finally, each pitch is classified as offering an investment in a company that commercializes a good or a service. Research has indicated that consumers perceive goods and services differently, and they may thus require different marketing strategies, with services potentially subject to longer consumer adoption processes (Murray & Schlacter, 1990). Additionally, the Sharks look to invest in businesses in which they can leverage their previous business experiences. Some, like Daymond John, founder of the clothing line FUBU, are more experienced in goods. Others, like Robert Herjavic, whose expertise is in internet services, may feel more comfortable investing in companies that provide services. While, as Murray and Schlacter (1990) indicate, products exist on a goods/service continuum, the variable is binary due to the complexity and ambiguity of placing products on the goods/services continuum. The definition for a good used is “a physical entity composed of tangible attributes” (Murray & Schlacter (1990), pg. 53). Products fitting this definition are receive the value 1 for the variable DummyGoodService, all products not fitting this definition receive the value 0. Research shows that other personal characteristics influence investment decisions. These include management experience, leadership experience and education (Monika, 2015; Fried & Hisrich, 1994; MacMillan, 1985; Hsu, 2007). Unfortunately, however, these characteristics were not observable on the show, and thus these variables are not included in the model.
Example Pitch and Data Collection
1. Cameron Sheldrake enters the Shark Tank, introduces himself and indicates that he is seeking $100,000 for 15% of his company. The size of the pitch team and the race and gender of the participant are recorded.
2. Cameron proceeds to give his pitch. He explains the difference between sweet corn and grain corn and explains that his company, Off the Cob, makes the only commercially available tortilla chips made out of sweet, rather than grain corn. The company is recorded in the data set as producing a good.
3. Cameron hands out samples of the chips and answers questions from Sharks regarding his product, sales and distribution network. The Sharks ask questions about the cost of production and wholesale and retail prices of the chips.
4. The Sharks then comment on their interest in investing in the business. Kevin O’Leary is the first to indicate that he is not interested. Mark Cuban asks some additional questions but then also indicates that he is not interested in investing. Daymond John then steps in and expresses
concerns about the industry and indicates that he too is “out”, or not interested in investing. Next, Lori Grenier indicates that she is also not interested. Finally, guest host Nick Woodman expresses that he enjoys the product and thinks the marketing is well done, but is not interested in investing as he cannot see a path to growing the business.
Figure 1
Cameron Sheldrake of Off the Cob delivers his pitch in the Shark Tank, hoping that a Shark will invest $100,000 for 15% of his company.
Statistical Analysis
The first step of the statistical analysis involves reporting descriptive statistics. The descriptive statistics are analyzed for the entire data set, as well as split into one data set consisting of all teams with at least one African-American team member (DummyBlack = 1), and all other teams that have no African-American team member (DummyBlack = 0). A difference in means test is then applied to the resulting descriptive statistics to analyze whether there are differences between the two groups that are noteworthy given the context of the study. Then, the variables are checked for multicollinearity. If there are no issues of multicollinearity, the hypotheses can be tested using regression analysis.
Two regression models are employed to test the effect of the independent on the dependent variables. The model employed depends on the type of dependent variable. For regressions that model a binary outcome, a Probit model is appropriate (Hoetker, 2007). Hypotheses 1a, 1c, 2a and 2c all employ a binary dependent variable, and therefore a Probit model is used to test these hypotheses.
binomial or Poisson regression are appropriate to test them, depending on the presence of overdispersion. Checking for overdispersion will be done by running negative binomial and Poisson regressions with the same dependent and independent variables and comparing the results. If the results are similar, there are no issues of overdispersion and a Poisson regression is used. If the results differ, there are issues of overdispersion and the negative binomial regression is appropriate.
Hypotheses 2a, 2b and 2c propose that over time, the effects of discrimination will be reduced. In this case, the “effect of discrimination” is measured by the same dependent variables used in testing Hypotheses 1a, 1b and 1c. To test whether these effects are reduced over time, the season in which the pitch aired is used as a proxy for time. An interaction effect between the season and DummyBlack
categorical variables is created which allows for the effects of the presence of an African-American on the pitch team to be analyzed on a season-by-season basis.
Reliability and Validity
The research design of this study does not pose issues of reliability. The findings of the study are easily reproducible by any researcher with access to the first six seasons of the show, and knowledge of the data collection method used. The dependent, independent and control variables are all unambiguous and recorded directly from the show, meaning that any person with knowledge of the variables and statistical models employed will be able to reproduce the results of this study. The only threat to the repeatability of the results, then, would be errors made in the data collection process. To ensure that errors did not pose a threat to the reliability of the study, an audit was made of the 495 pitches viewed; 10% were re-watched and the data was checked. During the audit, no issues with the integrity of the data were found.
Dragon’s Den have repeatedly been used as sources of data to analyze investor decision making (Maxwell et al., 2011; Poczter & Shapsis, 2018; Smith & Viceisza, 2018; Wanner, 2018). Levitt (2003), however, looked for evidence of racial discrimination on the television show Weakest Link and acknowledged that those featured on the show may be hesitant to display discriminatory behavior, knowing that their decisions would be televised. Similarly, Poczter and Shapsis (2018) analyzed decision-making on Shark Tank and indicated that the Sharks might change their normal behavior to avoid being perceived as prejudiced against women. It may be that the Sharks behavior is affected by the public nature of their decisions. If this is the case , the generalizability of the results of may be restricted to other scenarios where investors’ decisions are similarly public. The structure of the show is similar to how investors make investments in real life, and particularly similar to other public pitch competitions (Poczter & Shapsis, 2018; Smith & Viceisza, 2018). This supports the generalizability of the results. Additionally, entrepreneurs are frequently invited to pitch to potential investors in public situations, for example at university entrepreneurship contests. In these situations, the investors’ decisions are public as well, albeit without the same level of viewership as those made on Shark Tank. Therefore, it is reasonable to believe the results of the study can be generalized to other investment contexts where investors are pitched investments in person and are required to make public, time-constrained decisions on whether to invest.
4. Analysis
This section presents the descriptive statistics and the results of the tests of hypotheses described above.
Descriptive Statistics
495 pitches were analyzed for this study. These were all the pitches that were aired as part of the series from seasons 1 to 6. These seasons of the show were aired between 2009 and 2016. Of the 495 pitches, 62% received an offer from the Sharks, and 52% accepted an offer. On average, the contestants offered about 17.5% equity stakes in their companies, and requested just under $260,000 in return. This means that the mean valuation of Shark Tank businesses is $2,165,995. Most of the pitch teams pitched businesses involved in the commercialization of goods; almost 80% of the pitch teams pitched businesses based on selling goods. This contrasts with the general trend in the U.S. economy, which has been steadily moving away from a focus on the production of goods to the provision of services since the middle of the 20th century (Bell, 1976). The reason for the discrepancy may have to do with television
production value. While Shark Tank provides a venue for entrepreneurs to get funding for their
the show’s viewership. It may be that interesting and novel goods grab and hold viewers’ attention more than useful services, and thus the shows’ producers pre-selected goods-based businesses more frequently than services-based businesses. The pitches were delivered by pitch teams with a total of 748 team members. Females are underrepresented on the show, with 40% of pitch teams including a female
member. Of the 495 pitch teams, 8% contained an African-American. Overall, 9% of the total participants in the program were African-American. The descriptive statistics are presented in Table 1.
Table 1: Descriptive Statistics
Variable Mean Std. Dev. Min Max
CountTeamMembers 1.511 0.645 1 6 DummyBlack 0.081 0.273 0 1 CountBlack 0.093 0.336 0 3 DummyOfferReceived 0.622 0.485 0 1 CountOfferReceived 1.097 1.105 0 5 DummyInvestmentMade 0.521 0.524 0 4 CountMale 0.984 0.755 0 4 CountFemale 0.517 0.722 0 5 DummyFemale 0.404 0.491 0 1 ShareFemale 0.330 0.429 0 1 EquityOffered 17.538 10.015 3 100 CashAmountRequested 259854.500 461708.400 10000 5000000 DummyGoodService 0.794 0.405 0 1 Valuation 2165996.000 3756644.000 40000 3.00E+07
Means T-tests were conducted for to determine whether these differences are statistically significant. The Difference in Means T-Test tests the null hypotheses Ho for each variable in question:
Ho: (mean(DummyBlack=0) - mean(DummyBlack=1)) = 0
That is, the null hypothesis is that there is no statistically significant difference between the means for the pitch teams with African-American members and the pitch teams without African-American members for each of the relevant parameters, Valuation, EquityOffered, and CashAmountRequested. The results of the Difference in Means T-Tests are presented in Table 2. They indicate that the difference in mean valuation between pitch teams with African-American members and the pitch teams without African-American members is statistically significant at the 5% significant level.
Table 2: Results of Difference in Mean T-Test
Value DummyBlack Observations Valuation EquityOffered CashAmountRequested
0 455 2264063 17.38 269094.50
1 40 1050486 19.38 154750.00
Pr(T< t)
Ha = (mean(0) - mean(1)) ≠ 0 0.05** 0.227 0.133
Ho: diff = 0. diff = mean(0) - mean(1). ***Statistically significant at the 1% level ** Statistically significant at the 5% level * Statistically
significant at the 10% level
Variable Correlations
number of team members is the sum of male and female team members, and not a cause for concern. DummyInvestmentMade is highly correlated with DummyOfferReceived. This is to be expected and not a cause for concern, since an offer from the Sharks if required for an investment to be made. The
independent variables CashAmountRequested and Valuation have a high correlation. This is to be
expected, since the amount of the request is used to calculate the company's valuation, and not a cause for concern.
Regression Analysis
Probit and Poisson regression were run to test the hypotheses. The results of the tests of Hypotheses 1a, 1b and 1c are presented in Table 3. A Probit regression with dependent variable DummyOfferReceived was used to test Hypotheses 1a. Hypothesis 1a proposes that Shark Tank pitch teams with African-American members are less likely to receive an offer for funding from the Sharks than teams with no African-American members. Based on the results of the Probit regression analysis,
Hypothesis 1a is not supported. The independent variable DummyBlack, which takes the value of 1 if there is an African-American present on the pitch team and 0 if not, has a negative but not significant relationship with the dependent variable. This result indicates that the effect of the independent variable on the dependent variable is not statistically significantly different from 0. The control variable
EquityOffered. Hypothesis 1c proposed that Shark Tank pitch teams with African-American members are less likely to receive an investment (i.e. accept an offer) for funding from the Sharks than teams with no African-American members. A Probit model was used to test this hypothesis, as the dependent variable, DummyInvestmentMade, is binary. Again, the results of the regression analysis indicate that the hypothesis is not supported. The same control variables are employed in this regression and the results show similarities to the results of the test of Hypothesis 1a. Again, DummyBlack has a negative but not significant relationship with the dependent variable. Again, CountTeamMembers, DummyGoodService and EquityOffered have significant positive, positive and negative relationships with the dependent variable, respectively. The results of the tests of Hypotheses 1a, 1b, and 1c are clear. Throughout the first 6 seasons of the program, pitch teams with African-American members are at no disadvantage; the Sharks are equal opportunity investors. Some clear preferences among the Sharks, however, do emerge from the data. The Sharks prefer to invest in teams with more members, goods instead of services, and are more likely to make offers and investments when the equity stake offered is lower.
The question addressed in Hypotheses 2a, 2b, and 2c is whether any discriminatory effect might be mediated by time. That is, because the issue of discrimination against African-Americans in the United States has increasingly become a hot button issue in the years between 2009 and 2015; it may be that the Sharks started out their investing subject to the same forces that have caused widespread inequality, but as American society came to increasingly focus on racial equality in the subsequent years, they changed course. The tests of Hypotheses 2a, 2b, and 2c use the same models as Hypotheses 1a, 1b and 1c, respectively, but include an interaction effect between DummyBlack and Season, which indicates which season the pitch took place in. The results of the tests of Hypotheses 2a, 2b and 2c, with interaction effects, are presented in Table 4. The results indicate that Hypotheses 2a, 2b and 2c are not supported. For the hypotheses to be supported, the regression results should show a pattern of negative and significant relationships in early seasons and either non-significant or positive and significant relationships in later seasons. None of these patterns are manifested in the results. This lends further support to the idea that the Sharks are non-discriminatory. They have been consistently non-discriminatory throughout the life of the program, indicating that the non-discriminatory nature of the investments made on the show are
Table 3: Regression results of the tests of Hypotheses H1a, H1b and H1c
Dependent Variable DummyOfferReceived CountOfferReceived DummyInvestmentMade
Independent Variables CountTeamMembers 0.238** (0.101) 0.097 (0.063) 0.234** (0.094) DummyBlack -0.015 (0.214) -0.124 (0.179) -0.030 (0.210) ShareFemale -0.0009 (0.138) 0.010 (0.104) 0.120 (0.134) DummyGoodService 0.432*** (0.145) 0.383*** (0.118) 0.402*** (0.144) EquityOffered -0.031*** (0.007) -0.035*** (0.006) -0.015** (0.006) Model Pseudo R2 0.061 0.044 0.033
Table 4: Regression results of the tests of Hypotheses H2a, H2b and H2c
Dependent Variable DummyOfferReceived CountOfferReceived DummyInvestmentMade
Independent Variables CountTeamMembers 0.240** (0.103) 0.099 (0.063) 0.226** (0.095) ShareFemale -0.046 (0.141) -0.025 (0.104) 0.084 (0.136) DummyGoodService 0.422*** (0.149) 0.357*** (0.119) 0.396*** (0.146) EquityOffered -0.031*** (0.007) -0.034*** (0.006) -0.013 (0.006)** Dummy Black x Season 0 2 0.405 (0.287) 0.311 (0.206) 0.192 (0.277) 0 3 0.019 (0.243) 0.022 (0.194) -0.035 (0.241) 0 4 0.129 (0.214) -0.034 (0.172) 0.173 (0.211) 0 5 0.063 (0.212) 0.008 (0.168) 0.149 (0.209) 0 6 0.261 (0.217) 0.190 (0.164) 0.393 (0.213) 1 1 -0.066 (0.623) -0.203 (0.595) 0.134 (0.608) 1 2 -0.476 (0.761) -0.969 (1.011) -0.182 (0.766) 1 3 1.124 (0.660) 0.425 (0.404) 0.736 (0.573) 1 4 -0.324 (0.616) -0.813 (0.721) -0.054 (0.611) 1 5 0.770 (0.477) 0.460 (0.279) 0.467 (0.416) 1 6 -0.516 (0.445) -0.761 (0.468) -0.394 (0.451) Model Pseudo R2 0.080 0.057 0.046
5. Discussion
This thesis aimed to investigate whether AIs viewing in-person pitches discriminate against African-American entrepreneurs. This is an important question to address because African-Americans remain economically disadvantaged in relation to other ethnic groups in the United States (Pew Research Center, 2016). Entrepreneurship is an effective way to drive economic and employment growth, but African-Americans are also underrepresented as entrepreneurs, and are less successful when they do start their own new ventures (Robb & Fairlie, 2007). One of the drivers of this gap is a lack of access to startup capital, and it has been shown that African-Americans are discriminated against when they seek to access startup capital, including in credit markets and on crowdfunding platforms (Fairlie & Robb, 2008; Younkin & Kuppuswamy, 2017). Investors have been shown to weigh founders’ personal characteristics more heavily than purely business-related parameters like sales and profitability when making
investments (Rajan, 2012). If race, an inescapable, visible personal characteristic, makes investors less likely to invest in African-American businesses, there would be a compelling reason to try to correct this discrimination. On the contrary, it has been suggested that economic concerns; the desire to make an attractive return on an investment, are enough to compel investors to overcome any potential biases against women, and this study provides some evidence that this may be the case for African-Americans as well (Butter & Rosen, 1989). After collecting and analyzing data from six seasons of Shark Tank, the evidence shows that the Sharks do not discriminate against African-American entrepreneurs. Therefore, it is appropriate to turn to the sub question, what mechanisms specific to Shark Tank help contribute to this desirable outcome? To that end, three possible anti-discrimination mechanisms are proposed and
elaborated on in this section; that the desire to make good returns on investment disincentivize
discrimination, that the producers’ screening role helps eliminate discrimination, or that the public nature of the decisions made on the show disincentivize discrimination.
investment on the show. This may be sufficient motivation for them to overcome implicit as well as taste-based bias. In the taste-taste-based discrimination model, actors are driven to discriminate by personal animus towards an out-group. Taste-based discrimination is economically inefficient, the discriminating actor is willing to pay a price to avoid interacting with the out-group; call it the “racism premium”. In this case, since there is a stringent screening process that ensures that the presenters are qualified before they gain access to the Sharks, it may be that the racism premium is too high. A Shark who wishes to avoid investing in African-American businesses for taste-based reasons would have to forgo potentially lucrative rents, and may not be willing to do so. Again, however, it is possible that there are no outright racist Sharks, and this seems likelier than the proposition that none of the Sharks bear any implicit bias. Finally, it is important to note another point in support of the theory that financial motivators are behind the lack of discrimination on the show. As discussed in the literature review, extensive evidence of discrimination against African-Americans has been found in lending (Blanchflower et al., 1998; Robb & Fairlie, 2007; Fairlie & Robb, 2008; Bewaji et al., 2015). Lenders, however, have less personal financial disincentive to discriminate than do AIs, who invest their own money. The bank employees who process and approve loans, after all are usually not major shareholders in the banks that they work for.
Additionally, even if they were exceptionally loyal employees who wanted to make the already
astronomically wealthy anonymous beneficiaries of their hard work even richer, loans can only be paid back on time and with the agreed upon interest. That is, there is less of an upside than that of AIs, who can hope to make many times their investment if the company is successful. Therefore, the incentives to overcome implicit and taste-based discrimination that exist for AIs simply do not for lenders.
they invest 20 times more often than the baseline rate indicates that they trust that they are being brought quality businesses. This may help the Sharks overcome statistical discrimination. Statistical
discrimination models, in contrast with implicit bias and taste-based discrimination models, argue that discrimination is the result of rational decision-making; if investors think that African-American
businesses are more likely to fail, they may avoid investing in them (Guryan & Charles, 2013). Of course, as we have seen, African-American businesses in the US are more likely to fail, and they also have lower sales and profits (Robb & Fairlie, 2007). The Sharks may know these figures and be tempted not to invest in African-American businesses, but being cognizant of the stringent vetting process that presenters on the show must go through, are assured that there is no need for statistical discrimination. As Paul and Whittam (2009) note, besides screening, gatekeepers also make networking a key function of their job, so that they are able to get the angels they represent access to entrepreneurs that they otherwise might miss out on. The producers of Shark Tank act in a similar role, and it appears that they have prioritized finding qualified African American entrepreneurs. Overall, 9% of the total participants in the program were African-American. African-Americans, then, are underrepresented on the show as compared with their share of the U.S. population, which was estimated by the U.S. Census Bureau in 2017 to be 13.3% (United States Census Bureau, 2017). However, as compared with the overall rate of African-American self-employment in the United States, 5.1% African-Americans are over represented on the show (Gold, 2016). This may be because the producers of the show actively seek out African-Americans to participate. Episode 11 of Season 5 of the show shows Sharks Mark Cuban and Daymond John attending the
Kingonomics Conference for minority entrepreneurship and speaking on a panel dedicated to encouraging minority entrepreneurship. Additionally, the episode shows that the Sharks were accompanied by casting staff in an effort to increase African-American participation on the show. The combination of the
producers seeking out qualified African-American entrepreneurs and the Sharks trust in the stringent vetting process may result in those African-American entrepreneurs who do appear on the show being at no disadvantage compared to other ethnic groups.
The third and final reason that the Sharks do not discriminate may be due to the public nature of their decisions. Shark Tank is a nationally-broadcast show, with viewers throughout the country and around the world. Additionally, the Sharks are public figures with widely-known ventures, like Mark Cuban’s NBA franchise the Dallas Mavericks. As Levitt (2003) noted, participants on televised
competitions are disincentivized to show any signs of racism by the knowledge that their decisions will be made public. Especially in today’s social media age in the United States, where the perception of
noted by giants in the investing world like Melinda Gates and Laurence D. Fink. The Hawthorne Effect impacts studies where participants know that they are being observed, and therefore alter their behavior (Poczter & Shapsis, 2018). Could it be that a similar effect leads the Sharks to consciously strive to eliminate discrimination from their investing decisions? Poczter and Shapsis note that while the Sharks may alter their decision-making to avoid being seen as discriminatory, the scope of this effect is likely limited by the fact that they are investing their own money. So, while they may strive to look fair in front of television audiences and invest in more African-American businesses, they are unlikely to invest in a company that would have no chance of receiving an investment outside of the Shark Tank. Further research is needed to determine whether the non-discrimination effect is present in other investing scenarios where investors’ decisions are not made public. The public nature of the decisions made on Shark Tank also contrasts with the scenarios in which discrimination against African-Americans in new venture funding has been found; in seeking loans from banks and on crowdfunding platforms. In both cases, the decision-makers can be relatively sure that their decisions will not be made public, and they will not be associated with them.
Finally, some other notable points from the results study are presented. It was found that teams with American members assign lower valuations to their businesses than teams without African-American members. This may be because African-African-American-owned businesses tend to have lower sales and profits than white-owned businesses, leading to lower valuations (Fairlie & Robb, 2008). Alternately, it may be that African-Americans undervalue their businesses for other reasons, and further research might elucidate why. The factors which were found by this study to positively influence the likelihood of receiving an investment were the size of the pitch team, commercialization of goods instead of services, and offering lower equity stakes. That the Sharks preferred larger teams is not surprising, as it has become widely recognized that well-built teams are most likely to be able to meet the demands of starting and scaling new ventures (Chowdhury, 2005). The preference for goods over services might be explained by the expertise of the Sharks on the show. The Sharks often comment that they prefer to invest in businesses that are similar in nature to those in which they made their fortunes, and most of the Sharks have
backgrounds in goods-related fields, educational software products, real estate, direct-to-consumer television marketing, and clothing (About Shark Tank). Investors tend to stick to what they know,
equity, have less equity available to offer. Both types of companies would make for attractive
investments, and this may why the Sharks prefer to invest in pitch teams who offer lower equity stakes.
6. Conclusion
The results of the study present both managerial and theoretical implications. For African-American managers, the encouraging news is that Shark Tank is non-discriminatory. This may also mean that investment scenarios with similar parameters are also non-discriminatory. Since African-Americans face discrimination in lending, and lack of access to capital is an impediment for their businesses, this could present a useful avenue for seeking funding. They study provides support for the idea that African-Americans should seek out investment scenarios with private investors where the businesses are pre-screened, and investment decisions are public, like public pitch competitions. On the other side, there are some implications for investors. As we have seen, there is an increased focus in the United States on how investors can help solve societal issues. Investors like Melinda Gates who are interested in promoting minority owned businesses can promote measures that are used on the show, like the screening process, active search for African-American entrepreneurs and publication of investments to complement their efforts. Alternately, if we believe that the motivation of potential profits is enough to eliminate discrimination for private investors, there may not be a need to steer additional resources towards correcting a perceived imbalance. If this is the case, efforts to correct the underrepresentation of African-Americans in entrepreneurship would be better used elsewhere, for example in promoting skills education or helping lenders to overcome their documented biases. The study also contributes to the theoretical knowledge about discrimination in entrepreneurship. Much has been written about the gap between African-Americans and other ethnic groups in entrepreneurship. Sociologists and economists have proposed their own theories for this gap, and much of the economics theory surrounding it has focused on resource constraints, and discrimination in lending has been shown by multiple researchers (Blanchflower et al., 1998; Robb & Fairlie, 2007; Fairlie & Robb, 2008; Bewaji et al., 2015). This study, however, failed to find evidence for discrimination by angel investors. Further research is needed to determine why that is.
differential between the initial ask from the pitch team and the final offer from the Shark, in the event of an offer. Second, if the non-discrimination is due in part to the screening mechanism by the producers, the results of this study need not apply to investing scenarios where there is no gatekeeper providing access to the investors. Structured angel groups, groups of angels who work together to invest in new companies, are increasing in prominence and beginning to account for a larger proportion of angel investors, and these groups may employ gatekeepers (Paul & Whittam, 2009). However, many angels are still
independent, and seek out and vet potential investments on their own. In this case, the theorized effect of the gatekeepers to reduce discrimination would not exist. Further research should look into whether discrimination exists in investment scenarios where there is no screening mechanism between the pitch team and the investor. Third and last, the results of this study are limited in generalizability to investing situations that are broadly similar to Shark Tank. The key features of the show; a pitch to a panel of investors, a time limit in which the investors need to decide to invest, the public nature of the decision, etc., are not present in all investing scenarios. To the extent that these features of the show impact the decision-making of the Sharks, they limit the generalizability of the study. Data on the investment decisions of angels, however, is hard to come by (Paul & Whittam, 2009). This is due to the private nature of the investments, in which the decision-making process is rarely publicized. More and novel ways of collecting data on private investors should be sought out to better understand the decision-making process.
References
About Shark Tank. (n.d.). Retrieved from www.abc.go.com.
Bates, T. (2011). Minority entrepreneurship. Foundations and Trends in Entrepreneurship, 7(3-4), 151-311. doi:10.1561/0300000036
Bates, T., & Dunham, C. (1993). Asian-American success in self-Employment. Economic Development Quarterly, 7(2), 199-214. doi:10.1177/089124249300700206
Baum J., & Silverman, B. (2006). Picking winners or building them? alliance, intellectual, and human capital as selection criteria in venture financing and performance of biotechnology startups. Economics of Biotechnology, 2
Becker-Blease, J., & Sohl, J. (2007). Do women-owned businesses have equal access to angel capital? Journal of Business Venturing, 22(4), 503-521. doi:10.1016/j.jbusvent.2006.06.003
Bell, D. (1976). The coming of the post-Industrial society. The Educational Forum, 40(4), 574-579. doi:10.1080/00131727609336501
Bertrand, M., Chugh, D., & Mullainathan, S. (2005). New approaches to discrimination - implicit discrimination. The American Economic Review, 95(2), 94.
Bewaji, Yang, & Han (2015). Funding accessibility for minority entrepreneurs: An empirical analysis. Journal of Small Business and Enterprise Development, 22(4), 716-733. doi:10.1108/JSBED-08-2012-0099
Blanchflower, D., Levine, P., Zimmerman, D., & National Bureau of Economic Research. (1998). Discrimination in the small business credit market (NBER working paper series, working paper 6840). Cambridge, MA: National Bureau of Economic Research.
Blow, C.M. (2016, June 30). The State of Race in America. The New York Times. Retrieved from
www.nytimes.com.
Bort, Julie. (2018, May 31) Melinda Gates has sharp words for the VC industry: Enough with your love for 'the white guy in a hoodie'. Business Insider. Retrieved from https://amp.businessinsider.com. Brooks, A., Huang, L., Kearney, S., & Murray, F. (2014). Investors prefer entrepreneurial ventures
pitched by attractive men. Proceedings of the National Academy of Sciences of the United States of America, 111(12), 4427-31. doi:10.1073/pnas.1321202111
Buttner, E., & Rosen, B. (1989). Funding new business ventures: Are decision makers biased against women entrepreneurs? Journal of Business Venturing, 4(4), 249-261. doi:10.1016/0883-9026(89)90015-3
Carlsson, M., & Rooth, D. (2012). Revealing taste-based discrimination in hiring: A correspondence testing experiment with geographic variation. Applied Economics Letters, 19(18), 1861-1864. doi:10.1080/13504851.2012.667537
Chowdhury, S. (2005). Demographic diversity for building an effective entrepreneurial team: Is it important? Journal of Business Venturing, 20(6), 727-746. doi:10.1016/j.jbusvent.2004.07.001 Clark, C. (2008). The impact of entrepreneurs' oral ‘pitch’ presentation skills on business angels' initial
screening investment decisions. Venture Capital, 10(3), 257-279.
Dargis, M., Morris, W., Scott, A.O. (2016, January 15). Oscars So White? Or Oscars So Dumb? Discuss. The New York Times. Retrieved from www.nytimes.com.
Davis, F. (1991). Who is black? : One nation's definition. University Park, Pa.: Pennsylvania State University Press.
Fairlie, R., & Robb, A. (2008). Race and entrepreneurial success : Black-, asian-, and white-owned businesses in the united states. Cambridge, Mass.: MIT Press. doi:10.1080/13691060802151945 Feeney, L., Haines, G. H., & Riding, A. L. (1999). Private investors’ investment criteria: insights from
qualitative data. Venture Capital, 1, 121 – 145.
Feloni, R. (2016, September 22) 15 Behind-the-Scenes Secrets You Didn't Know About 'Shark Tank'. Business Insider. Retrieved from www.businessinsider.com
Fried, V., & Hisrich, R. (1994). Toward a model of venture capital investment decision making. Financial Management, 23 (3), 28-28. doi:10.2307/3665619
Fritsch, M., & Noseleit, F. (2013). Start-ups, long- and short-term survivors, and their contribution to employment growth. Journal of Evolutionary Economics, 23(4), 719-733. doi:10.1007/s00191-012-0301-5
Gilbert, B., Audretsch, D., & McDougall, P. (2004). The emergence of entrepreneurship policy. Small Business Economics : An International Journal, 22(3-4), 313-323.
doi:10.1023/B:SBEJ.0000022235.10739.a8
Gold, S. (2016). A critical race theory approach to black american entrepreneurship. Ethnic and Racial Studies, 39(9), 1697-1718. doi:10.1080/01419870.2016.1159708
Green, A., Carney, D., Pallin, D., Ngo, L., Raymond, K., Iezzoni, L., & Banaji, M. (2007). Implicit bias among physicians and its prediction of thrombolysis decisions for black and white patients. Journal of General Internal Medicine, 22(9), 1231-1238. doi:10.1007/s11606-007-0258-5 Greene, P., Brush, C., Hart, M., & Saparito, P. (2001). Patterns of venture capital funding: Is gender a
Greenwald, A., & Krieger, L. (2006). Symposium on Behavioral Realism - Implicit Bias: Scientific Foundations. California Law Review, 94(4), 945.
Guryan, J., & Charles, K. (2013). Taste-based or statistical discrimination: The economics of
discrimination returns to its roots. Economic Journal, 123(572), 432. doi:10.1111/ecoj.12080 Haagsma, R. (1993). Is statistical discrimination socially efficient? Information Economics and Policy,
5(1), 31-31.
Herring, C. (2009). Does diversity pay? Race, gender, and the business case for diversity. American Sociological Review (print), 2009(74):2, 208-224
Hoetker, G. (2007). The Use of Logit and Probit Models in Strategic Management Research: Critical Issues. Strategic Management Journal, 28(4), 331-343.
Hsu, D. (2007). Experienced entrepreneurial founders, organizational capital, and venture capital funding. Research Policy, 36 (5), 722-741. doi:10.1016/j.respol.2007.02.022.
Jolls, C., Sunstein, C., & John M. Olin Center for Law, Economics, and Business. (2006). The law of implicit bias (Discussion paper, no. 552). Cambridge, MA: Harvard Law School, John M. Olin Center for Law, Economics, and Business.
Kang, C. (2015, July 9). Obama’s top tech adviser takes fight for Silicon Valley diversity to Washington. The Washington Post. Retrieved from www.washingtonpost.com.
Kirwan Institute for the Study of Race and Ethnicity. (2014). State of the science: Implicit bias review 2014. Columbus, Ohio: Kirwan Institute for the Study of Race and Ethnicity, The Ohio State University.
Kirwan Institute for the Study of Race and Ethnicity. (2017). State of the science: Implicit bias review 2017 Edition. Columbus, Ohio: Kirwan Institute for the Study of Race and Ethnicity, The Ohio State University.
Lindén, A., & Mäntyniemi, S. (2011). Using the negative binomial distribution to model overdispersion in ecological count data. Ecology, 92(7), 1414-1421. doi:10.1890/10-1831.1
Levitt, S. (2003). Testing theories of discrimination : Evidence from weakest link (NBER working paper series, no. 9449). Cambridge, Mass.: National Bureau of Economic Research.
Macmillan, I., Siegel, R., & Narasimha, P. (1985). Criteria used by venture capitalists to evaluate new venture proposals. Journal of Business Venturing, 1 (1), 119-128.
doi:10.1016/0883-9026(85)90011-4.
Marom, D., Robb, A., & Sade, O. (2014). Gender dynamics in crowdfunding (Kickstarter): Evidence on entrepreneurs, investors, deals and taste based discrimination. SSRN Electronic Journal, (2014). doi:10.2139/ssrn.2442954
capital. 7, 153–172.
Maxwell, A. L., Jeffrey, S. A., & Lévesque, M. (2011). Business angel early stage decision making. Journal of Business Venturing, 26, 212–225.
Monika, & Sharma, A. (2015). Venture capitalists’ investment decision criteria for new ventures: A review. Procedia - Social and Behavioral Sciences, 189 , 465-470.
doi:10.1016/j.sbspro.2015.03.195.
Murray, K., & Schlacter, J. (1990). The impact of services versus goods on consumers’ assessment of perceived risk and variability. Journal of the Academy of Marketing Science : Official Publication of the Academy of Marketing Science, 18(1), 51-65. doi:10.1007/BF02729762 Paul, S., Whittam, G., & Wyper, J. (2007). Towards a Model of the Business Angel Investment Process.
Venture Capital, 9(2), 107 – 125.
Perrewe, P., & Ferris, G. (2006). Implicit sources of bias in employment interview judgments and decisions. Organizational Behavior and Human Decision Processes, Vol. 101 No. 2 (nov. 2006), P152-167
Pew Research Center, June 27, 2016. “On Views of Race and Inequality, Blacks and Whites Are Worlds Apart.”
Poczter, S., & Shapsis, M. (2018). Gender disparity in angel financing. Small Business Economics : An Entrepreneurship Journal, 51(1), 31-55. doi:10.1007/s11187-017-9922-2
Rajan, R. G. (2012). Presidential Address: The Corporation in Finance. Journal of Finance, 67(4), 1173– 1217.
Robb, A., & Fairlie, R. (2007). Access to financial capital among U.S. businesses: The case of african american firms. The Annals of the American Academy of Political and Social Science, 613(1), 47-72.
Robinson, J., Blockson, L., & Robinson, S. (2007). Exploring stratification and entrepreneurship: African American women entrepreneurs redefine success in growth ventures. The Annals of the American Academy of Political and Social Science, 613(1), 131-154.
Smith, B., & Viceisza, A. (2018). Bite me! ABC’s Shark Tank as a Path to Entrepreneurship. Small Business Economics : An Entrepreneurship Journal, 50(3), 463-479. doi:10.1007/s11187-017-9880-8
Sorkin, A.R. (2018, January 15). BlackRock’s Message: Contribute to Society, or Risk Losing Our Support. The New York Times. Retrieved from www.nytimes.com.
United States Census Bureau (2017, July 1). QuickFacts UNITED STATES. United States Census Bureau. Retrieved from www.census.gov.
van Aken, J., Berends, H., & Van der Bij, H. (2012). Problem solving in organizations: A methodological handbook for business and management students. Cambridge University Press.
van Praag, M., & Versloot, P. (2007). What is the value of entrepreneurship? : A review of recent research (Discussion paper / tinbergen institute, tI 2007-066/3). Amsterdam etc.: Tinbergen Institute.
Wanner, T. (2018). Do Angel Investors Prefer Dominant Entrepreneurs? (Unpublished master's thesis). University of Groningen, Groningen, Netherlands
Yang, Y., Narayanan, V., & Zahra, S. (2009). Developing the selection and valuation capabilities through learning: The case of corporate venture capital. Journal of Business Venturing, 24(3), 261-273. doi:10.1016/j.jbusvent.2008.05.001
Younkin, P., & Kuppuswamy, V. (2017). The colorblind crowd? founder race and performance in crowdfunding. Management Science, (20170531). doi:10.1287/mnsc.2017.2774
