The Bullshit Economy

Estimating the potential ad revenue for Pseudoscience Youtube Channels.

Benjamin Maclaren | Center for the Public Understanding of Science | The Australian National University, Australia | 2019

This was my final research project part of my undergraduate degree at the Australian National University.

Abstract:

The spread of pseudoscience has become a major concern for the public. As measles outbreaks escalate in the US, the World Health Organisation has labelled vaccine hesitancy as one of the ten threats to global health. Social media plays a critical role in the presence of misinformation, with the majority of the global populace use social media as their main sources of news and information. Pseudoscience communities have been using social media to form isolated communities that reinforce beliefs, groom potential advocates and otherwise selectively spread misinformation throughout the public. Youtube has played its part in the spread of pseudoscience. Providing a platform for content creation the Youtube recommendation system has come under fire for promoting conspiracy videos. Whilst their ad revenue system has enabled pseudoscience channels to earn revenue for spreading misinformation. To help us understand how these communities are sustaining themselves, we analyse channel and video statistics for 106 youtube channels and calculated estimated yearly ad revenue. It was found that there was a large potential for pseudoscience channels to earn significant revenue from spreading pseudoscientific misinformation resulting in businesses unknowingly supporting the spread of harmful information.

1. Introduction

To visualise the staggering numbers of children that died from measles every year in the 1990s, A single Boeing 747 plane can fit 524 people on board. Imagine every day, three Boeing 747 planes full of children, crashing, killing 100% of the children on board — that is the reality of measles. From January to April to December, on Christmas day, Easter and birthdays, year after year, leading to a total of 550,000 deaths every year globally. (Duprex, 2019).

Yet just 19 years ago, Measles was declared eliminated by the Centre for Disease Control thanks to a long campaign in vaccination efforts (“Measles CDC,” 2019). The adoption of the measles vaccine resulted in an 80% drop in measles-related deaths from 2000–2017 worldwide. This saved an estimated 21.1 million lives, making the measles vaccine one of the best ever buys in public health (“Measles,” 2019). Yet over the past 5 years, measles outbreaks recommenced and spread across the U.S, with a 30% rise in measles cases globally (“Ten health issues WHO will tackle this year,” 2019 ).

After a long and arduous campaign, with millions of lives saved, measles has returned. Many of these outbreaks appear to have been the result of parents refusing to vaccinate their children, often believing that vaccines cause autism, a belief that stemmed from a long-debunked, discredited and withdrawn paper in the early 90s (Mason, 2019). These beliefs on the danger of vaccines persevere in the face of overwhelming scientific consensus, causing harm to vulnerable communities. It appears the eradication of measles gave rise to an ignorance of the reality of measles. This ignorance resulted in some people misguidedly seeing vaccines as unnecessary in this modern age (Duprex, 2019). These pseudoscience communities still continue to produce and spread scientific misinformation under the guise of truth.

“The Earth is a flat disk centred at the north pole and surrounded by a wall of ice on all sides that hold the oceans back” — or so the English writer Samuel Rowbotham of the book “Earth is Not a Globe “would have people believe. After the writer’s death in 1884 the “Universal Zetetic Society” was established, eventually, it was succeeded in 1956 by the “International Flat Earth Research Society” which subsequently experienced declined in popularity. In 2004 it was reborn as an internet forum based organisation The Flat Earth Society (“Modern flat Earth societies,” 2019).

One of the current leading flat earth theories believed today proposes that NASA employees guard the great wall of ice that surrounds the disk earth. Much like watchmen of the night, that photos of the earth as a globe are photoshopped and that GPS devices are rigged to aeroplanes to fake flying in straight lines (Wolchover et al., 2017).

Many of these Flat Earth theories run in parallel to conspiracy theories. At the Flat Earth Conference in 2017 many of the speakers actively promoted other conspiracies with talks titled “NASA and Other Space Lies”, “Waking Up to Mainstream Science Lies” and “Flat Earth & The Bible”. But whilst flat earth pseudoscience experienced a rise, It was not uncommon for speakers at the conference to talk about the hardship they experienced. Members have been ostracized by their churches or have lost their jobs as a result of their beliefs in a flat earth (Wolchover et al., 2017).

Historically, advocating for pseudoscience, is thankful, an uphill battle. Sustaining a pseudoscience campaign is a great monetary, time-intensive and often criticized burden, requiring the purposeful investment of time and money from willing people. One of the main catalysts for the growth of flat earth pseudoscience was Youtube. In 2014 videos had begun to spread and communities forming resulting in a larger audience and wider proliferation of disinformation than what traditional media facilitated. Youtubes freedom of content creation and it’s next video recommendation system, opened the doors to a wider pool of audiences being exposed to pseudoscientific content. This coupled with YouTube’s ad revenue system incentivised and supported content creators regardless of whether their content was scientifically accurate or not (Mitchell, 2017).

While there exists a number of studies on science on the Youtube platform (Shapiro and Park, 2015, Welbourne and Grant, 2016, Amarasekara and Grant, 2019). There is a lack of research specifically, on the financial support of Youtube’s pseudoscience content. Ad revenue remains a key feature of the Youtube platform, incentivising channels to continue developing attractive consumer content that doesn’t necessarily have to have any relationship with the truth or scientific fact. This presents an important question: Is it possible that pseudoscientific Youtube channels are earning enough income from ad revenue to support the spread of pseudoscience and misinformation?

This study aims to investigate the potential revenue that pseudoscientific YouTube channels may be earning, and to determine if Youtube’s revenue-sharing model providing a supportive pathway for dangerous pseudoscience. To address this question, we identified a range of pseudoscientific channels, and conducted an analysis of channel and video content statistics, to estimate the potential revenue these channels could be earning.

This investigation represents a step in discovering how businesses and platforms may be unknowingly supporting, promoting and otherwise aiding in the spread of harmful misinformation and pseudoscience. It should be acknowledged that there are many factors that go into a Youtube channels revenue — this current paper is intended to give insight into the possibilities and should be built on in future work. We will start with discussing how to separate science from pseudoscience, the presence of scientific misinformation and its modern use, followed by the rise of the internet and social media then narrowing down into youtube, youtube misinformation, earning money on youtube and finishing with our data method, results and conclusion.

Separating Science and Pseudoscience

What exactly is pseudoscience? The issue of distinguishing science from pseudoscience (known in philosophy as the demarcation problem), has a significant influence on many areas in society from education, population health, food and drug safety to expert testimony in court, medical treatments and research funding (Resnik, 2000).

Pseudoscience is a complex problem, we do not call Isaac Newton a pseudoscientist, even though he believed light to be composed of the corpuscle. Before Einstein, scientists had thought that Newton had deciphered the very foundational laws of the universe, only to then to have many of his theories replaced with more accurate ones. Scientific theories can be proven wrong yet those who once believed in them are not always considered pseudoscientists (Pigliucci and Boudry, 2013).

There have been many attempts over the past centuries to come up with valid criteria to separate science from pseudoscience. To find a complete criterion that separates pseudoscience from science is beyond the scope of this paper. However, a possible description is explored by Lakatos, 1978, claiming that scientific achievements are not individual hypothesis’ but a research plan and that science is not simply trial and error or an argument of logic against logic.

“Newton’s theory of gravitation, Einstein’s relativity theory, quantum mechanics, Marxism, Freudianism, are all research programmes, each with a characteristic hardcore stubbornly defended, each with its more flexible protective belt and each with its elaborate problem-solving machinery. Each of them, at any stage of its development, has unsolved problems and undigested anomalies. All theories, in this sense, are born refuted and die refuted.” (Lakatos, 1978)

So then how do we distinguish science from pseudoscience?. Lakatos, 1978 and W. Lack and MA, 2016, suggest two simple possible criteria to the issue of demarcation: Prediction — A theories ability to make stunning and importantly unexpected predictions, and openness to correction — A plans willingness to change and adapt based on results or be discarded if proven wrong. (Lakatos, 1978) This is quite apparent in the flat earth movement, where despite experiment results demonstrating a spherical earth and predictions of a flat earth failing, rather than changing the theory to fit the results, the results were instead ignored and experiments that confirmed the theory were sought after instead (Kottke, 2019).

There is no absolutely clear line that separates legitimate science and pseudoscience, the difference stems not from the topic but from the process (Novela, 2013). Blood-letting or (phlebotomy)is an effective treatment for polycythemia — too many red blood cells — , however, blood-letting, as practised in Galenic medicine — most sicknesses was thought to occur from too much blood — is pseudoscience (Novela, 2013). One process works, whilst the other doesn’t even if the topic is the same. Many ideas of science today began as a fringe science and many processes in pseudoscience like mindfulness have been adopted and supported by science (Powell, 2018).

The anti-vaccination movement isn’t the first case of pseudoscience in the world nor the first case of harm being caused by misinformation. People have been believing weird things for centuries. Religion, conspiracies, superstitions and some traditions are all considered to be an avenue of pseudoscience to some degree. Misinformation likewise has been around for decades from the earliest snake oil miracle cure to the spam and scam emails that fill our email junk filters (Shermer, 1997).

People are faced with real questions every day that require scientifically accurate information, should creation science be taught in public schools? should an insurance company pay for visits to a chiropractor? can we trust that a building’s design won’t collapse in an earthquake or should medical research into acupuncture be funded? (Resnik, 2000). As demonstrated in the case of the anti-vaccination movement, when we can’t distinguish between pseudoscience and science, decisions are made that can have very real and sometimes deadly consequences (Hansson, 2017). We must be able to distinguish between pseudoscience and science to make practical decisions.

“The demarcation between science and pseudoscience is not merely a problem of armchair philosophy: it is of vital social and political relevance.”(Lakatos, 1978)

Whilst the boundary between pseudoscience and science is fuzzy and difficult to assess, there are a number of defined categories and labels that are used by scientists and in public discourse to lump topics into the pseudoscience category. Often when talked about in the public eye, we often refer to specific topics of pseudoscience with topic derived labels to a pseudoscience believer like anti-vaccination and (Anti-Vaxxers), Flat Earther and (Flat Earthers) or Creation Science and “Creationists” (“Advocates,” 2019; Wolchover et al., 2017; ). To make matters simple, this paper will only focus on categories and labels that have been communally documented on the Wikipedia pseudoscience categories and pseudoscience advocates page with an emphasis on socially relevant categories and excluding pseudoscience that lacks coherent communities.

Misinformation, and the Modern Day

Over the past few years, there has been a growing concern about the effect of misinformation on society, both in science and politics (Scheufele and Krause, 2019). In the current media climate, the roles of creating and curating media and the responsibilities to fact check information has moved from professional news journalists, trained experts and news-specific industries over to a broader media industry and the public. Single authoritative organisations are no longer the main trusted source for information (Korsunska, 2019).

The spread of misinformation isn’t harmless. It can have serious and disastrous impacts on society. In 2013 an erroneous tweet from the associated press claiming that the then U.S. President Obama was injured in an explosion resulted in $130 Billion dollars in AP stock value dropping in minutes. Whilst the stock prices recovered shortly after — this demonstrated how news on social media can be used to manipulate systems (Rapoza, 2019). In a similar vein, a reporter was arrested in 2013 for being bribed to fabricate news stories throughout 2012–2013. This led to widespread criticism of the construction company Zoomlion, resulting in a 26.9% decrease in the companies stock price(Rapoza, 2019).

“If you can lower the price of a stock by one percent by purposefully manipulating the news flow by producing content and if you have the right trading mechanism in place, you can capitalize on that,” (Rapoza, 2019).

One of the major instigators of the concern of misinformation was the 2016 United States presidential election. A few months prior to the election, many U.S. citizens were exposed to false stories with dubious for-profit articles being shared by millions of people. Post-election results showed that many people who read those stories believed them to be true resulting in misinformed voting. (Guess and Nyhan, 2018).

The use of misinformation to discredit an opponent or sway public opinion isn’t new. It extends all the way back to Roman times when Antony met Cleopatra. Octavian launched a propaganda war against his political opponent Antony, in the form of short sharp slogans written on coins now akin to modern day tweets. They denounced Antony as a puppet of Cleopatra’s, a womaniser and a drunk. This attack was very effective at discrediting Antony and resulted in Octavian becoming the first emperor of Rome. “Fake news had allowed Octavian to hack the republican system once and for all.” (Kaminska, 2017).

This wouldn’t be the first time a political member would use misinformation to discredit a rival and win the proverbial throne. In a similar fashion to Octavian, participants of the anti-vaccination movement cherry pick scientific information from open-access journals and used them to increase vaccine uncertainty in the general populace (Schalkwyk and Joubert, 2019).

Information Economy

In human history, there have been three main periods of communications that facilitated the effectiveness of the spread of misinformation and information. These periods represent major changes to the economy of misinformation: the first was our most basic form of communication oral traditions where information was communicated through sounds, lights and basic symbols. The second was the era of literacy ranging from first pictograms up to the invention of the printing press in 1493, which allowed misinformation and disinformation to spread with ease through mass production print (“Printing Press, Digital Age, and Social Movements,” 2017).

The third and current phase is the electronic communications era, born from the invention of the telegraph in 1844. Due to the nature of the more traditional printing press form of media from the second era, the communication power was held by governments and large organisations and was subject to filtering, censorship and reviews before being released to the public. By the dusk of the industrial revolution more modern journalism had begun to take root (“Printing Press, Digital Age, and Social Movements,” 2017).

The information age, also known as the digital age or the computer age, is often credited to the inventors of the transistor that revolutionised electronics. It is the time in the 21st century where the world transitioned from the traditional industry that was the result of the industrial revolution, into an economy based upon information technology. whilst the computer is a big part of the information age, globalisation — the integration of markets, nation-states and technologies to a degree never seen previously (Friedman, 1999), and informationization — the transformation of economic and social relations to such an extent that cultural and economic barriers are minimized (Kluver, 2000). Both form the foundation of changing information economy (“Information Age,” 2019).

The introduction of the internet subverted past media platforms and was established as the dominant information platform of the world that integrated a variety of media. Now anyone with a computer can distribute and provide information on the internet with ease, regardless of class, education, expertise, location or finances (“Printing Press, Digital Age, and Social Movements,” 2017).

Online Media and Social Networks.

As of 2019, 54% of global news is consumed through online sources. The vast majority of the public depend on a mix of media formats, with 60% of Americans relying on the internet for their primary source of science information (Su et al., 2015). Media sources have changed in comparison to before the internet, On the heels of the changing media landscape, 17% of consumers use social media as their main news source with 45% of global consumers being concerned about what is real or fake on the internet ( Glen Fuller, Caroline Fisher, 2019).

However, despite this high level of concern, only 41% of global consumers are likely to check the truthfulness of a story by checking alternative sources. These changes in attitudes reflect the changing information economy. The move to online news has weakened the relationship between readers and publishers, resulting in the modern news reader sourcing information from many news sources (Glen Fuller, Caroline Fisher, 2019).

Most online media isn’t under the scrutiny of quality control or have the same editorial gatekeepers that traditional mass media had. Content is generated by both experts and non-experts, blurring the lines between opinions and facts. One of the things that makes social media unique is that it enables people to not only connect with people across the world, culture and class but that it enables the visibility and sharing of people’s social networks all determined and designed by private companies (Boyd and Ellison, 2007).

Complicating this further, people are able to create virtual communities around like-minded people who seek the same sources of information to confirm held beliefs and opinions. This results in an echo chamber effect — where people receive more information on what they agree with and are exposed to fewer alternative or opposing views and communities (Korsunska, 2019).

Social Network sites or as most commonly known in public discourse, as Social media, started to emerge around 15–20 years ago with the maturity of the internet and the interconnectivity from globalization and informatization. Social Media is roughly defined as “A group of Internet-based applications that build on the ideological and technological foundations of the Web 2.0 that allow the creation and exchange of user-generated content” (Kaplan and Haenlein, 2010). Before the platforms of Instagram, Facebook and Youtube, Social media first began with the early communities that developed using Bulletin Board Systems. BBS systems over a span of 20 years would eventually evolve into chat-rooms and forums then into the social platforms we know today (Brown, 2012).

Social media platforms have been struggling with the growth of misinformation communities (Wong, 2019). This has lead technology companies to attempt to limit the spread of misinformation on their platforms (Roose, 2019) In March this year, Facebook banned anti-vaccination ads and removed anti-vax groups from their recommendation system (Wong, 2019). Additionally, in June 2019 Facebook removed the conspiracy facebook page “Natural News” a page that had garnered over 3 million followers and advocated disinformation about vaccines, and promoted natural remedies and conspiracy theories (Weill, 2019). Pinterest — a visual social network, explicitly banned the “promotion of false cures for terminal or chronic illnesses and anti-vaccination advice.” (“Health misinformation,” 2019) Youtube in January, announced they would be reworking their recommendation systems to attempt to reduce the number of videos that could misinform viewers. This was followed by a number of prominent anti-vaccination Youtube channels being demonetized in an attempt to remove financial incentives (Wong, 2019).

Youtube, Controversy and Fake news on Youtube

The consumption of online videos has grown the past decade, YouTube is the biggest online video platform hosting user-generated content — totalling over 1.9 billion regular monthly users across 91 countries, available in 80 different languages (“Press — YouTube,” 2019). First founded in 200,5 since then YouTube was acquired by Google in 2006. Over a period of 5 years, YouTube rose from their first published video to more than 400 Hours of videos being uploaded every minute with over 2 billion views every day, they are the second largest search engine and the third most visited site (Smith, 2019).

Youtube has been the focus of a number of misinformation issues over several years. On its launch, the platform was fairly minimalist, people could publish videos of any kind (with the exception of sexually explicit or excessively violent material) with almost no regulation, curation or control (Burgess, 2012).

Over time, YouTube developed a system colloquially known as “The Algorithm” or “The YouTube Algorithm” (YA), this most commonly refers to YouTube’s search, discovery and recommendation system. The YA’s is designed to get people to watch more videos they always enjoy so they come back to YouTube regularly, it is designed to customise the search results and video recommendations to individual users likes, dislikes and habits, The algorithm follows the audience, it tracks and predicts a variety of audience statistics including what users watch and don’t watch, how much time is spent watching a video and likes and dislikes (Green, 2015; “YouTube Search and Discovery,” 2019).

In 2012 Youtube made a number of changes to the recommendations algorithm, previously the algorithm had been designed to maximize views, recommending videos that the system thinks a user would most likely click on. The original system however, led to “click-baiting”- videos with exaggerated titles and thumbnails designed to fool and entice you into clicking them, despite whether they accurately represented the video’s content or not. (“YouTube Search and Discovery,” 2019)

Instead, the new updated system emphasised watch time, encouraging content creators to produce videos that users would finish watching. This was designed to increase viewer satisfaction and the time they spent on Youtube. This change, coupled with the advertising rule changes — where a channel doesn’t need to be vetted by youtube to earn ad revenue, had an unintentional effect of boosting political and pseudoscientific videos, allowing any channel regardless of content to earn money from ad revenue.

YouTube has continued to try and improve and train their recommendation system, over the continuing scrutiny over policing hateful, harmful and conspiratorial content. In the wake of the 2017 Las Vegas shooting, Youtube search results in addition to reputable news stories also recommended a variety of shooting conspiracy video such as “ISIS claims responsibility for Las Vegas Shooting” and “The Truth About the Las Vegas Shooting” (Warzel, 2017). At the peak of the measles outbreaks, advertisers expressed concern over their ads being displayed on anti-vaccination content on Youtube, this resulted in Youtube demonetising a range of anti-vaccination videos and banning accounts (“YouTube takes ads off ‘anti-vax’ videos,” 2019).

Youtube Ads and Earning Money

YouTubes revenue system is complicated, the system has undergone many revisions of the platforms lifetime, many channels receive income only partially from YouTube ads, some get regular donations from audiences on platforms like patreon, sell audience merchandise and can be sponsored by businesses to advertise products or services (Booth, 2013). Analysing the entire monetary network of a single channel is a large undertaking and beyond this paper, due to the availability of channel and video statistics and the lack thereof of alternative channel revenue sources, the primary target for this research is an estimate of the potential revenue from advertising alone.

In order to host an ad on youtube, a advertiser bids on the highest amount they would pay for a video view of their ad, click on the call-to-action buttons and banner clicks. The actual amount advertisers pay for their ads called a cost-per-view (CPV)is based on two factors: Quality Score, which is a measure of how relevant an ad is to the watcher, this score also includes performance indicators like view rates, and an Ad Rank, which is your Quality Score multiplied by your max bid (“About cost-per-view (CPV) bidding — Google Ads Help,” 2019).

To calculate the CPV the two factors: Quality Rank and Ad Rank, are then used by YouTube to consider the ad rank of bidders below the advertiser and the advertiser quality score, from this the highest rank is then paid the CPV of the bid of the most competing ad (“About cost-per-view (CPV) bidding — Google Ads Help,” 2019).

In order for a video to be able to earn revenue from ads, they first must meet Youtubes advertiser-friendly content guidelines and be approved based on the video metadata (thumbnails, titles, tags and description). From there, video revenue, watch-time, ad clicks and total views, as well as a ads bid are all factors that influence the amount a channel receives or it’s Cost per thousand (CPM) value, an industry term for revenue per thousand engagements (Rosenberg, 2019).

2. Method

It should be acknowledged that there are many factors that go into how much a channel receives from ad revenue and that we can’t, of course, know all the details about revenue or production costs (see method below for further caveats). Video click, views, ad clicks, sponsors, partnerships and external merchandise are all variables that come into play. We hope that rough ballpark figures may start painting a more detailed picture of the current economy of pseudoscience.

Finding and Categorizing Youtube Channels.

In order to gather channel information, we need to be able to identify a pseudoscience channel. As mentioned earlier, there is no single authority on what categories are classified as pseudoscience. However, there are a range of labels and subjects that are most commonly used and referenced by scientists and the public that is categorised as a pseudoscience these include labels such as “Homeopaths”, “Astrologers”, “Flat Earthers” and “Vaccine Skeptic” or “Anti-vaxxer”(“Advocates,” 2019; Lockett, 2019; Mukherjee, 2019) as well as subjects like “Astrology” (“Pseudoscience Category,” 2019 ;“Examples of Pseudoscience,” 2019).

Channels were sourced through three primary search methods: 1. Direct search terms through Youtube’s search base using identified categories, 2. Direct google video search and 3. Curated listicle articles on Youtube Channels.

A large number of channels contain videos on multiple pseudoscience categories like ufology, mysticism and conspiracies, or were hyper-specific and only promoted one branch of pseudoscience like “Homeopathy”, “Astrology” or ”Vaccine Denial”. As a result, channels were categorized either as a broad-based category like Conspiracy or with a specific label like Chiropractic, this was determined from the quantity of total content for all channels or their strong social relevance like “Anti-vaccination” and “Climate Change”. This categorization method was used in order to give a contextually broad and socially relevant view of the current pseudoscience economy.

Search Terms

Search terms were used from the Wikipedia category page on pseudoscience (“Pseudoscience Category,” 2019). Each category and its corresponding wiki page was assessed to find search terms that would most commonly be used such as “Earth is flat” and “Vaccines cause autism”. Question Phrasing like “Are vaccines dangerous?”, direct category terms like “Astrology” and a category or it’s colloquial name appended with a variety of superstitious and negative suffixes such as “false”, “fake”, “bad”, “vaccine conspiracy”, “hoax” were also used to locate potential channels. Search sessions were conducted with caution, ensuring that a blank history and cookies browser was used, with no accounts logged-in in-order to avoid potential bias from Youtube and Google algorithms or echo-chambers that could develop whilst searching.

Channel Data Gathering

Pseudoscience channels were logged with their affiliated name, URL and category when they contained a majority of videos that advocated for the searched pseudoscience topic. Furthermore, the related channels bar was used on channel profiles to find additional channel networks that an identified channel either recommended to their native audience or that its content was similar to. Additionally, two listacles articles were used to identify popular channels (“Top 50 Astrology YouTube Channels for Horoscope & Zodiac Sign Videos,” 2017) (“Top 25 Chiropractor Youtube Channels To Follow in 2019,” 2018).

In total 109 channels were identified. Each channels statistics included channel publish date, the total amount of videos with each videos’ view, likes, publication date and comments were extracted using Python and YouTubes channel and video API. Gathering data took around 2 weeks due to the quota on free Youtube API requests and occurred between the 9th of May and the 23rd of May 2019. Over the two weeks period, one channel was excluded from the list due to the videos of the channel being erased.

CPM and Revenue Values

YouTube doesn’t officially publish an average CPM statistics. However, over time CPM amounts from digital media benchmark reports and existing Youtubers have been reported, giving a CPM value ranging from a value of $2, upwards of $7.50(Rosenberg, 2019; Green, 2015). On top of this, Youtube takes an estimated 45% of the CPM revenue, for this paper we’ll be using the more conservative estimate of $2 CPM for calculating revenue estimates.

3. Results

In total 14 categories encompassed 106 channels, combined they represent 64,314 videos with over 5.2 Billion views. These channels don’t represent all pseudoscience channels but instead, are indicative of the more successful and popular channels.’ Figure 1.0 illustrates the distribution of the categories across all logged channels, this represents the prevalence of different categories on Youtube. “Chiropractic” and “Conspiracy” channels were the most found channels with “Creation Science” and “Telepathy” channels the least prevalent.

Figure 1.0 — Category Distribution

The total video count of all channels had a mean of 607 with a median of 253. A channels total video views have a mean of 49,667,221 with a median of 12,026,969. On average a single video was viewed 81,860 times.

How much money?

Each channel had its total video views converted into a revenue value using the conservative $2 CPM value, this combined with the total channel videos quantity and the channels earliest and latest video publish dates, allowed us to put an average “revenue per video” and overall “revenue per year” figure to each channel. The histogram in figure 2.0 depicts the total revenue per year, that a channel has earned over the duration of their lifetime. In total 28 Channels potentially earned under $1,000 dollars per year, 34 earned between $1,000-$10,000. 27 channels between $10,000-$50,000, 10 channels earned between $50,000 — $100,000, 5 channels had a yearly revenue of between $100,000-$200,000 and only 1 channel had a yearly potential revenue of over $200,000.

Figures 2 — Yearly Revenue

The scatter plot chart in figure 3 shows the ad revenue per video estimate. 66 channels earned under the 100$ revenue mark, with a further 22 earning between 100–500$, in the minority are 9 channels earning $500-$1000 and 9 channels earning $1000 or more per video produced.

Figure 3.0 — Ad Revenue Per Video

Figure 4.0 — Video Count to Views

Discussion

Interestingly, when we compare a category’s total video quantity with their corresponding views in figure 4.0, we can see that the channels’ videos quantity does not correspond to a larger community. Astrology has the greatest quantity of videos but accounts for only 6.55% of total views whilst the superstition category has only 2.06% of total videos yet accounts for 20.97% of total views.

Based on the distribution of channels across categories and the total viewer engagement, From figure 1.0, we can consider “Chiropractic”, “Conspiracy”, “Ufology” and “Superstition” to be the most prevalent communities on the YouTube platform garnering a greater amount of channels and views. This could suggest that the less pervasive communities operate primarily through different social media platforms.

Whilst obvious harmful pseudosciences such as anti-vaccination might not be the most prevalent on Youtube, being misinformed is an indicator of a person’s ability and motivation to spot falsehoods (Scheufele and Krause, 2019). Being exposed to lesser harmful pseudoscience may act as a gateway pseudoscience to being more receptive for other kinds of misinformation.

Comparing the mean channel revenue of $24,054 against the median yearly salary of $47,060 in the US (Doyle, 2019. A conservative estimate was used to estimate the yearly channel revenue, even with this lowest value it was surprising to see that 16 of channels logged, could potentially have earned over $50,000 exceeding the median yearly salary with 6 Channels surpassing the $100,000 revenue.

Conclusion

The results of this study reveal that a significant amount of profit could potentially be gained from promoting pseudoscience just purely through ad revenue alone. Despite Youtube’ attempts to limit the spread of pseudoscience on the platform, there still exists avenues of pseudoscience businesses may be supporting through ad revenue that might not be outright labelled as dangerous but may still have an impact on the economy of misinformation.

Youtube has made progress on improving its recommendations systems to decrease the spread of misinformation. However scientific misinformation still has an impactable presence on the youtube platform, as audiences who are vulnerable to misinformation grow this presents a complex and difficult problem for the company to solve. Although there have been studies on the monetary profit of non-scientific misinformation, further analysis on the financial network of pseudoscience communities may be a worthwhile avenue of investigation.

This study provides an early insight into the potential role advertising revenue plays in sustaining misinformation in social media. It would be important to extend this work to explore the extent of pseudoscience revenue networks and their role in supporting pseudoscience. Potential areas for future research could include an analysis of the complete revenue network of top pseudoscience communities, media content investigation, the role that other social media platforms revenue systems play or the extent of alternative revenue sources like Patreon donations or merchandise sales.

In conclusion, it is hoped that in this new media age of the 21st century, that this and further research can uncover, and bring to attention, major supportive sources that play a role in the growth of pseudoscience.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

References

1. About cost-per-view (CPV) bidding — Google Ads Help (2019). Available at: https://support.google.com/google-ads/answer/2472735?hl=en-AU (accessed 5 June 2019).
2. Amarasekara I and Grant WJ (2019) Exploring the YouTube science communication gender gap: A sentiment analysis. Public Understanding of Science 28(1): 68–84. DOI: 10.1177/0963662518786654.
3. Booth A (2013) How to Make Money on YouTube: 14 Creative and Effective Ways. Available at: https://www.lifehack.org/articles/money/11-creative-ways-make-money-youtube.html (accessed 10 June 2019).
4. Boyd DM and Ellison NB (2007) Social Network Sites: Definition, History, and Scholarship. Journal of Computer-Mediated Communication 13(1): 210–230. DOI: 10.1111/j.1083–6101.2007.00393.x.
5. Brown P (2012) The evolution of online communities and the social web. Available at: https://culttt.com/2012/02/06/the-evolution-of-online-communities-and-the-socialweb/ (accessed 30 May 2019).
6. Burgess J (2012) YouTube and the Formalisation of Amateur Media. DOI: 10.4324/9780203112021–12.
7. Cases and Outbreaks (2019) Measles CDC. Available at: https://www.cdc.gov/measles/cases-outbreaks.html (accessed 28 May 2019).
8. Category:Pseudoscience (2019) Category:Pseudoscience. Wikipedia. Available at: https://en.wikipedia.org/w/index.php?title=Category:Pseudoscience&oldid=883572259 (accessed 12 June 2019).
9. Center for Mobile Communication Studies (2017) Printing Press, Digital Age, and Social Movements. Available at: https://sites.bu.edu/cmcs/2017/11/16/printing-press-digital-age-and-social-movements/ (accessed 29 May 2019).
10. Doyle A (2019) How Much Is the Average Salary for US Workers? Available at: https://www.thebalancecareers.com/average-salary-information-for-us-workers-2060808 (accessed 14 June 2019).
11. Duprex P (2019) Measles: Why it’s so deadly, and why vaccination is so vital. Available at: http://theconversation.com/measles-why-its-so-deadly-and-why-vaccination-is-so-vital-110779 (accessed 7 June 2019).
12. Feedspot Blog (2017) Top 50 Astrology YouTube Channels for Horoscope & Zodiac Sign Videos. Available at: http://blog.feedspot.com/astrology_youtube_channels/ (accessed 9 June 2019).
13. Feedspot Blog (2018) Top 25 Chiropractor Youtube Channels To Follow in 2019. Available at: http://blog.feedspot.com/chiropractor_youtube_channels/ (accessed 9 June 2019).
14. Friedman T (1999) The Lexus and the Olive Tree: Understanding Globalization.
15. Glen Fuller, Caroline Fisher (2019) Digital news report: Australia 2019. News and Media Research Centre (UC). Available at: https://apo.org.au/node/240786.
16. Green H (2015) The $1,000 CPM. In: Hank Green. Available at: https://medium.com/@hankgreen/the-1-000-cpm-f92717506a4b (accessed 5 June 2019).
17. Guess A and Brendan Nyhan (2018) Selective Exposure to Misinformation: Evidence from the consumption of fake news during the 2016 U.S. presidential campaign.: 49.
18. Hansson SO (2017) Science and Pseudo-Science. In: Zalta EN (ed.) The Stanford Encyclopedia of Philosophy. Summer 2017. Metaphysics Research Lab, Stanford University. Available at: https://plato.stanford.edu/archives/sum2017/entries/pseudo-science/.
19. Kaminska I (2017) A lesson in fake news from the info-wars of ancient Rome. Available at: https://www.ft.com/content/aaf2bb08-dca2-11e6-86ac-f253db7791c6 (accessed 29 May 2019).
20. Kaplan AM and Haenlein M (2010) Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons 53(1): 59–68. DOI: 10.1016/j.bushor.2009.09.003.
21. Kluver R (2000) Globalization, Informatization, and Intercultural Communication. Available at: http://unpan1.un.org/intradoc/groups/public/documents/apcity/unpan002006.htm (accessed 30 May 2019).
22. Korsunska A (2019) The Spread and Mutation of Science Misinformation. In: Information in Contemporary Society (eds NG Taylor, C Christian-Lamb, MH Martin, et al.), 2019, pp. 162–169. Lecture Notes in Computer Science. Springer International Publishing.
23. Kottke J (2019) Flat-Earther Proves in Simple Experiment that the Earth Is Round. Available at: https://kottke.org/19/02/flat-earther-proves-in-simple-experiment-that-the-earth-is-round (accessed 14 June 2019).
24. Lakatos I (1978) The methodology of scientific research Programmes.: 257.
25. Lockett J (2019) Antarctica: End-of-the-world trip to prove Earth is flat. Available at: https://www.news.com.au/travel/world-travel/antarctica/flat-earthers-plan-incredible-trip-to-antarctica-to-reach-end-of-the-world/news-story/0a2e9079e5dba11d56c95590e18dafa3#.fsy4c (accessed 12 June 2019).
26. Mason GC (n.d.) Measles outbreak: Why are anti-vaxxers risking a public health crisis? Available at: http://theconversation.com/measles-outbreak-why-are-anti-vaxxers-risking-a-public-health-crisis-116334 (accessed 28 May 2019).
27. Measles (2019). Available at: https://www.who.int/news-room/fact-sheets/detail/measles (accessed 29 May 2019).
28. Mitchell D (2017) The Earth may not be flat, but it just might be doomed | David Mitchell. The Guardian, 19 November. Available at: https://www.theguardian.com/commentisfree/2017/nov/19/the-earth-may-not-be-flat-but-it-is-possibly-doomed-fake-news (accessed 12 June 2019).
29. Mukherjee S (2019) The Anatomy of an Anti-Vaxxer. Available at: http://fortune.com/2019/05/13/measles-outbreak-anti-vaxxers/ (accessed 12 June 2019).
30. Novela S (2013) Is There a Pseudoscience Event Horizon? In: NeuroLogica Blog. Available at: https://theness.com/neurologicablog/index.php/is-there-a-pseudoscience-event-horizon/ (accessed 2 June 2019).
31. O’donovan C, Warzel C, McDonald L, et al. (2019) We Followed YouTube’s Recommendation Algorithm Down The Rabbit Hole. Available at: https://www.buzzfeednews.com/article/carolineodonovan/down-youtubes-recommendation-rabbithole (accessed 13 June 2019).
32. Pigliucci M and Boudry M (eds) (2013) Philosophy of Pseudoscience: Reconsidering the Demarcation Problem. Chicago: The University of Chicago Press.
33. Pinterest help (2019) Health misinformation. Available at: https://help.pinterest.com/en/article/health-misinformation (accessed 10 June 2019).
34. Powell A (2018) Harvard researchers study how mindfulness may change the brain in depressed patients. In: Harvard Gazette. Available at: https://news.harvard.edu/gazette/story/2018/04/harvard-researchers-study-how-mindfulness-may-change-the-brain-in-depressed-patients/ (accessed 3 June 2019).
35. Press — YouTube (2019). Available at: https://www.youtube.com/intl/en-GB/yt/about/press/ (accessed 29 May 2019).
36. Rapoza K (2019) Can ‘Fake News’ Impact The Stock Market? Available at: https://www.forbes.com/sites/kenrapoza/2017/02/26/can-fake-news-impact-the-stock-market/ (accessed 29 May 2019).
37. Resnik DB (2000) A pragmatic approach to the demarcation problem. Studies in History and Philosophy of Science Part A 31(2): 249–267. DOI: 10.1016/S0039–3681(00)00004–2.
38. Roose K (2019) The Making of a YouTube Radical. The New York Times, 8 June. Available at: https://www.nytimes.com/interactive/2019/06/08/technology/youtube-radical.html, https://www.nytimes.com/interactive/2019/06/08/technology/youtube-radical.html (accessed 10 June 2019).
39. Rosenberg E (2019) How Youtube Ad Revenue Works. Available at: https://www.investopedia.com/articles/personal-finance/032615/how-youtube-ad-revenue-works.asp (accessed 5 June 2019).
40. Schalkwyk F van and Joubert M (2019) Why anti-vaccine beliefs and ideas spread so fast on the internet. Available at: http://theconversation.com/why-anti-vaccine-beliefs-and-ideas-spread-so-fast-on-the-internet-111431 (accessed 28 May 2019).
41. Scheufele DA and Krause NM (2019) Science audiences, misinformation, and fake news. Proceedings of the National Academy of Sciences 116(16): 7662–7669. DOI: 10.1073/pnas.1805871115.
42. Shapiro MA and Park HW (2015) More than entertainment: YouTube and public responses to the science of global warming and climate change. Social Science Information 54(1): 115–145. DOI: 10.1177/0539018414554730.
43. Shermer M (1997) Why people believe weird things: Pseudoscience, superstition, and other confusions of our time.
44. Smith K (2019) 48 Fascinating and Incredible YouTube Statistics. Available at: https://www.brandwatch.com/blog/youtube-stats/ (accessed 30 May 2019).
45. Su LY-F, Akin H, Brossard D, et al. (2015) Science News Consumption Patterns and Their Implications for Public Understanding of Science. Journalism & Mass Communication Quarterly. DOI: 10.1177/1077699015586415.
46. W. Lack C and MA JR (2016) Critical Thinking, Science, and Pseudoscience: Why We Can’t Trust Our Brains. Springer Publishing Company.
47. Weill K (2019) Facebook Removes Conspiracy Site Natural News. 10 June. Available at: https://www.thedailybeast.com/facebook-removes-conspiracy-site-natural-news (accessed 10 June 2019).
48. Welbourne DJ and Grant WJ (2016) Science communication on YouTube: Factors that affect channel and video popularity. Public Understanding of Science 25(6): 706–718. DOI: 10.1177/0963662515572068.
49. Wikipedia (2019a) Category:Advocates of pseudoscience. Available at: https://en.wikipedia.org/w/index.php?title=Category:Advocates_of_pseudoscience&oldid=896198898 (accessed 12 June 2019).
50. Wikipedia (2019b) Information Age. Available at: https://en.wikipedia.org/w/index.php?title=Information_Age&oldid=898504349 (accessed 30 May 2019).
51. Wikipedia (2019c) Modern flat Earth societies. Available at: https://en.wikipedia.org/w/index.php?title=Modern_flat_Earth_societies&oldid=899610538 (accessed 12 June 2019).
52. Wolchover N, May 30 LSS| and ET 2017 10:24am (2017) Are Flat-Earthers Being Serious? Available at: https://www.livescience.com/24310-flat-earth-belief.html (accessed 12 June 2019).
53. Wong JC (2019) Facebook to ban anti-vaxx ads in new push against ‘vaccine hoaxes’. The Guardian, 7 March. Available at: https://www.theguardian.com/technology/2019/mar/07/facebook-anti-vaxx-vaccine-hoax-ads (accessed 10 June 2019).
54. World Health Organisation (n.d.) Ten health issues WHO will tackle this year. Available at: https://www.who.int/emergencies/ten-threats-to-global-health-in-2019 (accessed 28 May 2019).
55. YourDictionary (2019) Examples of Pseudoscience. Available at: https://examples.yourdictionary.com/examples-of-pseudoscience.html (accessed 5 June 2019).
56. YouTube (2019) Search and discovery on YouTube. Available at: https://creatoracademy.youtube.com/page/lesson/discovery (accessed 30 May 2019).
57. YouTube takes ads off ‘anti-vax’ videos (2019) 25 February. Available at: https://www.bbc.com/news/technology-47357252 (accessed 14 June 2019).