Media Portrayal and Public Opinions of Trump’s Crypto Agenda during 2024 US Presidential Election

Bruce Yuan | School of Journalism and Communication, University of Oregon

Abstract

Cryptocurrency has been increasingly addressed in political context, which was widely used by Trump as a political agenda during the 2024 US Presidential Election. This study uses a lexicon-based sentiment analysis approach on 28140 comments from 44 videos covering Trump’s engagement in crypto-related events from 15 unique major news accounts on YouTube. The results show that despite most news titles being framed as a neutral emotion of surprise, the overall public opinions are more positive (53.86% ) than negative (46.14%). Three themes are detected: 1) the crypto-specific theme that encompasses recognition of the cryptocurrency-related events by iteration, 2) the campaign-specific theme that encompasses a dominant positive voice expressing trust on Trump, and 3) the attitude/action-specific theme that encompasses concerns on cryptocurrency as scam, manipulation, and financial crimes and higher willingness to vote for Trump than Harris. This study advances understanding of public opinions on cryptocurrency being a political agenda in political elections. Methodologically, not only it contributes to the understanding of why AI and LLM is not appropriate in textual analysis with emotions, but also suggests using multimodal domain-specific lexicon when evaluating public opinions and effectiveness of political campaigns related to specific technologies.

Keywords: Cryptocurrency, Public Opinion, Opinion Mining, Political Communication

Cryptocurrency has been at the center of discussion in political contexts across the world. For example, Bitcoin is designated as a legal tender and fiat currency in El Salvador and the Central African Republic (Alvarez et al., 2023). The user base of crypto in the US is estimated to reach 96.67 million by 2025 (Statista, 2024). Almost 44% of all corporate election funds, totaling $274 million, are received from crypto backers in the 2024 US Presidential Election (Claypool, 2024). In the most recent 2024 US Presidential Election, one of the presidential candidates, President Trump, has repositioned himself from criticizing crypto as “based out of thin air” on social media in 2019 to openly advocate crypto at a major crypto-themed conference by making the statement of “making US the crypto capital of the planet” (Liptak & Vazquez, 2024).

Cryptocurrency has always been a controversial topic since its inception. On the one hand, it is celebrated for its potential in promoting modern democracy (Weiss et al., 2022) and financial inclusion (di Prisco & Strangio, 2021). On the other hand, it has been notorious for encouraging financial crimes such as money laundering (Leuprecht et al., 2023), wash trading (Lin et al., 2022), and frauds (Kerr et al., 2023). Because cryptocurrencies are widely associated with ideologies such as anarchism (Malabou, 2020; Brekhov, 2022), libertarianism (Allon, 2018), communism (Lee, 2023) and democracy (Schneider, 2024), past studies have proposed new voting schemes using cryptocurrency to enhance democratic practice in political elections (Buterin et al., 2018; Jara-Vera, 2022; Kew Yen Nee & Eng Thye, 2024).

Meanwhile, media effects of political campaigns from traditional media (Druckman, 2004) to social media (Gilardi et al., 2022) have been well studied in the past. Although public opinions on cryptocurrency have extant literature in finance for price prediction purpose (Huang et al., 2021; Parekh et al., 2022; Raheman et al., 2022; Nair et al., 2024), it is the first time that a political campaign on cryptocurrency launched during the US Presidential

Election, none of previous studies have addressed the media effects and public opinions on cryptocurrency during national-level presidential election. Provided how influential and controversial cryptocurrency is in the political context, how is crypto portrayed in the titles of these news? What are the public opinions on these portrayals of crypto in the media?

This paper draws an exploratory perspective from 28140 comments made by the public from 44 videos on President Trump’s engagement in crypto-related events covered by 15 unique major news accounts on YouTube. A lexicon-based sentiment analysis approach is conducted on the titles of sampled videos and the comments; a comparative sentiment analysis using Natural Language Processing with ChatGPT is also conducted on the news titles to illustrate the framing of news. This study not only advances knowledge of understanding public perception of new technology that carries democratic ideology in political campaigns, but also shed lights on the significance of methodological development on domain-specific lexicons for sentiment analysis regarding technology policies in political campaign (e.g., in crypto, bearish, bullish, moon, hodl, diamond hand, whale, etc.).

Literature Review Public Opinions on Cryptocurrency and Opinion Mining

Cryptocurrency is broadly defined as a decentralized payment method empowered by blockchain, establishing low-cost facilities to ensure pseudo-anonymity, independence from central authority, and instantaneous processing (Hossain, 2021). Cryptocurrencies encompass mainstream coins (e.g., Bitcoin, Ethereum, and Ripple, etc.), Central Bank Digital Currencies (CBDCs), altcoins (e.g.,meme coins), and the latest innovation, non-fungible tokens (NFTs) (Ong et al., 2024).

Sentiment analysis is a common way to evaluate people’s perceptions on a variety of topics. For example, Ainapure et al. (2023) used a lexicon-based sentiment analysis from tweets to “identify and analyze people’s perceptions about disease and vaccination during the pandemic period”. Computational methods, such as lexicon-based (Zhang et al., 2011) and machine-learning-based methods (Rajput, 2020; Dang et al., 2020; Qian et al. 2022; Ainapure et al., 2023), are commonly used in extracting public opinions (also known as opinion mining) for the capabilities in capturing entity-level sentiment in various contexts, and the methodology has evolved rigorously over the years. Twitter, among which is the most popular destination for opinion mining because of its abundance of texts as data in various contexts (Britzolakis et al., 2020; Qian et al. 2022).

Most sentiment studies of cryptocurrency using computational methods aimed for price prediction of cryptocurrencies. Huang et al. (2021) developed a a long short-term memory (LSTM) as a crypto sentiment analyzer based recurrent neural network, which outperforms autoregressive model in sentiment detection by 18.5% in precision and 15.4% in recall along with the historical cryptocurrency price movement to predict future price trend. Parekh et al. (2022) implemented a deep-learning-based method, DL-GuesS, to predict the price of Bitcoin Cash, integrating the interdependency of different types of cryptocurrencies. Rouhani and Abedin (2020) reported that over 50% positive sentiment on Twitter in their sample using supervised learning with a lexicon-based approach. Combining lexicon-based and

AI-based method, Qian et al. (2022) conducted a sentiment analysis of 290,282 tweets composed from October 15, 2021 to June 22, 2022, with an overall 72% positive sentiment and 28% negative sentiment depicted, the results also revealed that public actions towards NFT are more prone to opinions given by Twitter accounts with large follower base. Previous studies of sentiment analysis on cryptocurrency fall largely into the discipline of finance.

possible prediction patterns from public sentiments aiming at valuation of certain types of cryptocurrency to predict price of cryptocurrencies (Huang et al., 2021; Parekh et al., 2022) and reinforce marketing efforts of cryptocurrency projects (Rouhani & Abedin, 2020; Raheman et al., 2022), while very few address cryptocurrency as general-purpose existence to ordinary people or in political contexts to voters.

Cryptocurrency has become a synonym for anarchism (Malabou, 2020; Brekhov, 2022), libertarianism (Allon, 2018), communism (Lee, 2023) and democracy (Schneider, 2024). However, to associate cryptocurrency with anarchist, or libertarian, or any other ideologies is simplistic because cryptocurrencies are a social construct, representing a form of membership which members enter the community with diverse values, expectations, and attitudes (Obreja, 2024). By studying 38 Romanian cryptocurrency investors on Facebook, four major themes of public opinions on cryptocurrencies are revealed: 1) positive attitude on the unbreakable power of code, 2) positive attitude on being a member of the community with peers, 3) negative attitude towards the risks (e.g., bankruptcy, collapsed value), and 4) negative attitude on safety as preference on safety above all (e.g., safety, risks, losses, bankruptcy) (Obreja, 2024).

The Debate on Cryptocurrency

Cryptocurrency has been controversial in its political association and contribution to democracy and governance (Johnson, 2023) because of its technical and financial nature (Malabou, 2020). Travis (2020) argued that cryptocurrency organizes collective freedom and embodies democracy because it promotes information sharing, open-source software, offering solutions to slow and costly exchange of value controlled by corporations and introducing new socio-economic and political schemes via decentralization. Because of the

neutrality of blockchain network, cryptocurrency related promotion should be exempt from civil liability and pre-registration requirements unless false statements are used, therefore, a human right protected by statute and under the Constitution (Travis, 2020). Many have believed that cryptocurrency empowers higher level of financial inclusion to people who are excluded from by formal financial institutions (Chung et al., 2023), especially in emerging countries (Schuetz & Vankatesh, 2020; Setyawan et al., 2024).

On the contrary, cryptocurrencies are also blamed for its association with undue conducts and criminal actions, such as scams, frauds, money laundering, and wash trades. On an individual level, 5 types of fraudulent activities, namely Ponzi, malware, advance fee, fake services, and blackmail are detected (Bartoletti et al., 2021). According to Kerr et al. (2023), specific fraud and scams include investment scams, rug pulls, romance scams, exchange scams, scam wallets, mining scams, COVID-19 related scams, fraudulent investment apps, and pump-and-dump schemes. The collapse of FTX with confirmed misappropriation of funds and the default risk (Fu et al., 2023) raised with the death-spiral of Terra/LUNA (Liu et al., 2023) and reveals the instability of crypto institutions even with credible founders and

big-name capital firms, posing unprecedented challenges on cryptocurrency regulation in light of the nature of decentralization (Jain, 2023; Aibai et al., 2024). Money laundering (Teichmann & Falker, 2021) and tax evasion (Rohanifa et al., 2021) using cryptocurrency are emerging as national interest because they pose threats to national security and economy. By the definition of the US Department of Treasury (n.d.), money laundering is defined as:

“Money laundering generally refers to financial transactions in which criminals, including terrorist organizations, attempt to disguise the proceeds, sources or nature of their illicit activities. Money laundering facilitates a broad range of serious underlying criminal offenses and ultimately threatens the integrity of the financial system.”

Using Canada as a critical case study, Leuprecht et al. (2023) pointed out that third-party exchanges are mostly likely to conceal the proceeds of crime and illicit funds in the form of Bitcoin and alt-coins. They argued that although new regulations on third-party exchanges positively reduced money laundering, it could still be insufficient taking into consideration the decentralized nature of cryptocurrency and the myriad types of alt-coins; the power of blockchain may provide shelter to various money laundering and cybercrimes for endowing anonymity in transactions, aggravating crimes that pose threats to national security and immigration system, such as terrorism financing (Teichmann, 2018; Fletcher et al., 2020), drug control, arms trafficking (Escobar et al., 2022) and illegal immigration (Haq et al., 2024). Yet, Trump proposed the launch of World Liberty Financial, a third-party exchange with the announcement of “ We're leading a financial revolution by dismantling the stranglehold of traditional financial institutions and putting the power back where it belongs: in your hands.” (World Liberty Financial, n.d.). On a global scale, Ba and Sen (2024) indicated that regime type, exchange rate mechanism, and capital account openness countries determine if a country is more inclined to issue a crypto ban. Countries with less flexibility in exchange rate, less capital account openness, and higher levels of financial development are more inclined to issue a crypto ban (Ba & Sen, 2024). These issues emerging from cryptocurrencies are of paramount significance to national security, international affairs, and economic development, which suited the political agenda of Trump’s political campaign of Making America Great Again (Restad, 2020).

Media effects and Political Agenda-setting

According to the agenda-setting theory, “the mass media set the agenda for each political campaign, influencing the salience of attitudes toward the political issues" (McCombs & Shaw, 1972. The study of political news coverage has shifted from focusing on formal political actors to diversified sources of news (Moy & Bosch, 2013) because political news appears in many forms (Entman, 2005), among which social media has been frequently addressed as an important source of political news than traditional news media (Shehata & Strömbäck, 2021). Druckman (2004) provided evidence in the 2000 US Senate that political campaigns covered in the media primed voters who were exposed and attentive to based their voting decisions on specific issues (e.g., health care and social security) and candidate images (e.g., integrity) highlighted during the campaign. Interpersonal discussion was found to augment the campaign effect on television and newspaper (Druckman, 2024), the affordances of social media for connective action (Vast et al., 2017) and interpersonal discussion in groups (Malhotra, 2024) may create a perfect soil for political campaigns, especially using Facebook and YouTube for political expression (Halpern & Gibbs, 2013).

While massive fake news was curated towards Trump in the 2016 US Presidential Election (Allcott & Gentzcow, 2017), Trump won the election nonetheless. Guo and Vargo (2020) found that political events on front pages of partisan media have also boosted subscriptions of other legacy media, such as the New York Times and the Washington Post since election, giving the legacy media the incentive to massively cover event-specific political news during election. Shehata & Strömbäck (2021) indicated that although digital and social media are constantly replacing traditional media use, traditional news media is conducive to political and current affairs learning, which is consistent with the findings of Langer and Gruber (2021) on the importance of legacy media in political agenda setting, whereas contrasting to the findings that following political news on social media does not contribute to political and

current affairs learning despite of political interest and existing knowledge. Gilard et al. (2022) examined the role of social media in political agenda setting within the Swiss context. Three agendas are identified in this study: 1) the traditional media agenda, 2) the social media agenda of political parties, and 3) the social media agenda of politicians, where the three agendas significantly counterbalanced each other, but none of them consistently lead others. An exception is found on environmental issues that the social media agenda of parties is more predictive than the traditional media agenda, highlighting issues with high public interests or advocacy campaigns may lead traditional media coverage (Gilard et al., 2022).

Methodology

Sampling

This study employs a purposive sampling by first searching with the combination of keywords, “Trump crypto Bitcoin 2024” on YouTube, which aims to receive a return of all videos covering Trump’s participation in major crypto-related events in the year of 2024. Videos and sources are selected based on 1) public perception of trustworthiness on media sources (Pew Research Center, 2014), 2) multiple channels for news coverage on social media and mass media that at least one of the channels existed over 10 years, and 3) relevance to Trump’s participation in crypto-related events in 2024. Videos and news sources are selected if 1) they are at least labeled more trusted than distrusted, 2) if not found on the trusted sources list, it is selected if one of the channels of this program has operated for over 10 years, and 3) the title contains “Trump” and either “crypto”, “Bitcoin”, or ”coin” the same time, and the videos are posted in 2024, which covers President Trump’s engagement in crypto-related events in 2024, including but not limited to his attendance of the Bitcoin Conference 2024 at Nashville, TN, the launch of his crypto exchange (World Liberty Financial), and his attendance in a crypto-themed pub (PubKey) in New York, etc. 44 videos

from 15 unique media sources are selected. The comments left under the videos as of September 22, 2024 are scraped down, reaching 28140 comments in total, converting to a total of 26,554 unique comments that contain 291,829 words.

Data Analysis and Results

Sentiment Analysis: Titles

To understand the framing of 44 titles to an average American audience, a prompt was asked in Perplexity.ai Pro using GPT 4o from OpenAI, to categorize the 44 titles to positive, neutral, or negative (Taherdoost & Madanchian, 2023; Tsai et al., 2024). Validity of response is checked by human evaluation of the researcher and a second person, an American college student. Below is a sample request and response:

Prompt:

Consider you are an average American, categorize the 44 news titles into positive, neutral, or negative, only return ID and emotion. State your reasoning.

1. Trump to speak at Bitcoin conference
2. Donald Trump launches new cryptocurrency business, World Liberty Financial
3. Donald Trump Pledges to Fire SEC's Gensler, Make US 'Crypto Capital'
4. Trump Unveils Crypto Project | Bloomberg Crypto 09/17/2024
5. Trump to Address Bitcoin Event in Show of Crypto Support

…

Response:

ID Rating Reasoning

The AI returns a result of 11 positives, 20 neutrals, and 13 negatives.

By human evaluation, the result returned from the prompt is somewhat biased. For example, for title 2 and 3, “Donald Trump launches new cryptocurrency business, World Liberty Financial” and “Donald Trump Pledges to Fire SEC's Gensler, Make US 'Crypto Capital'” are merely stating the event or words as fact, but it is labeled as 7/9 and positive. Moreover, for title 41, “Trump, the most trustworthy businessman, is teaming up with crypto, the most trustworthy business”, sarcasm is not detected.

After cross validation, the result is adjusted to 12 negatives, and 32 neutrals. While most videos are neutral, reporting Trump’s engagement and words towards crypto-related events as facts, titles of videos from David Pakman Show and The Daily Show, and Forbes Breaking News are labeled as negative.

As a comparison, a sentiment analysis is conducted following text analysis using R (Welbers et al., 2017) . Assuming that emotions can cross over contexts, titles are re-organized as one-word-per-row, and each word is referred to a lexicon of human emotions, Word-Emotion Association Lexicon derived from Plutchik’s wheel of emotions using crowdsourcing on Amazon Turk (Saif & Turney, 2013), and categorized to corresponding emotion in the lexicon. (Figure 1):

Figure 1: Sentiment of words from titles, sorted by sentiment (one-word-per-row)

While all titles demonstrate the sentiment of surprise, a neutral sentiment that is consistent with previous results, the titles of The Daily Show, David Pakman Show, and Forbes Breaking News are not interpreted as negative as the adjusted results in the other way; a portion of fear is captured for all 3, and anger is captured for Forbes Breaking News.

The top five words used in the title are “Trump” (n = 42),”crypto” (n = 26), “bitcoin” (n = 22), “Donald” (n = 11), and “conference” (n = 7). The top five sentiments detected in the titles are surprise (n = 47), positive (n = 18), trust (n = 12), anticipation (n = 12), and joy (n = 8). Surprise is detected in all 15 news sources, and more positive emotions are detected than negative, which is consistent with previous validation by humans and AI.

Sentiment Analysis: Comments

Using the same method, a text analysis on 291,829 words derived from 26,554 unique comments (i.e., duplicate comments eliminated) is conducted. The removal of stopwords and non-annotated words leaves a total of 145,292 words in total carrying sentiment. Figure 2 illustrates the overall sentiment of comments as slightly more positive (n = 78250, 53.86% ) than negative (n = 67042 , 46.14%)¹. The top five words mentioned in all comments (Figure 3) are “Trump” ( n = 7108, 3.05% ), “crypto” (n = 2442, 1.05%), “bitcoin” (n = 2201, 0.94%), “people” (n = 2201, 0.94%), and “money” (n = 1661, 0.71%). The number of positive words ( n = 26612, 11.41%,) is greater than that of negative words (n = 19747, 8.47%). The top 3 sentiments detected are trust (n = 18430, 7.9%), surprise (n = 13770, 5.91%), and anticipation (n = 12969, 5.56%).

Figure 2 : Words mentioned over 500 times in comments

¹ Positive words include words labeled as joy, trust, positive, surprise, and anticipation; negative words include words labeled as anger, fear, negative, disgust, sadness.

Figure 3. Word cloud of all words from comments

Thematic Analysis by Network of Words

As the one-word-per-row method may not properly capture negation, which are comments that contain negation are “not” (n = 105), “no” (n = 124), “never” (n = 14), and “without” (n = 2). The dataset is further arranged to bigrams in order to capture the direction and context of important words. 3 themes are categorized: campaign-specific (i.e., “trump”, “people”, “money”,”president”), crypto-specific(i.e., “crypto”, “bitcoin”), and action/attitude-specific (i.e., “vote”, “scam”). Figure 4 illustrates top 5 words precedes and follows the top 8 appeared words, and Figure 5 provides a holistic view of the network of words in bigrams. Campaign-specific. For the most appeared word “trump” (Figure 6), the top 5 words preceding it, “president” (n = 293) is the second most mentioned word, which represents positive sentiment, and “fail” (n = 31) is the third most mentioned word, although not annotated in the lexicon nrc as negative, it is associate the word “failure”, which is annotated as negative. The word “vote” (n = 86) appears as the fourth most mentioned word.

The word “president” is perceived as trust and positive in the nrc lexicon. While the word that appears most frequently after “president” is “trump”, the word that appears most frequently before “president” is “vice” (n = 22) , which in this context, points to current Vice President of the US and the opponent party, Kamala Harris. It is consistent with the other two words that follow “president”, which is “biden” (n = 21) and “harris” (n = 16). 2 of the top 5 words preceding “people” are perceived negatively, “scam” (n = 24) and “stupid” (n = 14). “scam” will be discussed under the Action/attitude-specific theme. When it comes to “money”, 4 out of the top 5 words preceding it are negative, which are “launder” ( n = 20), “fake” (n = 19), “monopoly” (n = 17), and “lost” (n = 15). The word that appeared most after “money” is “laundering” (n = 97).

Crypto-specific. While most words associated with “crypto” and “bitcoin” are related to the context and event, such as “bitcoin conference”, “bitcoin bar”, “trump’s crypto”, “scam” (n = 60) is the second most appeared word after “crypto”.

Action/attitude-specific. “vote” is annotated across all sentiment with the nrc lexicon according to specific context. The top 4 words following “vote” are , “blue” (n = 144), “trump” (n = 86), “harris” (n = 31), and “red” (n = 18). While “scam” is annotated as a negative word, 3 out of the top 5 words appeared before it can make it contextually negative, namely “crypto” (n = 73 ), “trump” (n = 39 ) and “ponzi” (n = 15 ), whereas the word “total” (n = 21 ) and “scam” (n = 28 ) make the negativity more intense.

Figure 4: Top 8 key words and words appear before and after

Figure 5: Network of words with high frequency of appearance

Figure 6 : Most appeared node and edges from the word “trump”

Domain-specific words. Domain-specific words in crypto are not properly captured by the nrc lexicon. For example, 212 of 241 words that are “bull” or “bullish” is not annotated, when it actually refers to a positive expectation of the market condition. 57 words of “moon” and 24 “hodl” are not counted as well, where they represent long-term optimistic attitude towards crypto in this specific domain of crypto. Moreover, emojis like the small rocket (i.e., 🚀) also demonstrates positive expectations of crypto, which the text-oriented nrc lexicon cannot properly capture, resulting in a dismissal of 82 comments that contain this symbol in the sentiment. 18 comments containing Chinese characters are also not properly captured by the nrc lexicon made in English. Sample comments contain these domain-specific words and symbols as well as in languages other than English are as the following:

“Bitcoin to the moon - BTTM sol meme coin 🚀🚀 jump on board. The man said it”

“NEVER PANIC SELL. Learn the concept of HODL. Welcome to the space, trump

fans!”

“我局外人！一个具有独到见解的灵性道路上的修行者！宇宙没有秘密！一切预

言都是天赐予！美国总统特朗普！路遇不平拔刀相助！”

“没有特朗普美国就没有明天！宇宙高维天意！顺天者昌 逆天者亡！”

Sentiment by Sources

Comments on videos from MSNBC have the most words that are annotated as anger (n = 2142, 5.7%) and sadness (n =1876, 5.0%). Comments on videos from Forbes have the most words that demonstrate surprise (n = 1575, 8.8%) and trust (n = 1924, 10.8%). Comments on videos from Bloomberg Television exhibits the highest level of anticipation (n = 109, 8.8%) and the highest level of joy for USA Today (n = 88, 7.7%) , while comments on videos from David Pakman Show and WSJ News illustrate the highest level of disgust (n = 1895, 3.5%) and fear (n = 120, 5.7%), respectively.

Figure 7: Sentiment of words from comments, sorted by source (one-word-per-row)

Figure 8. Top 20% mentioned words in each source by sentiment (1)

Figure 9. Top 20% mentioned words in each source by sentiment (2)

Polarity

Polarity of opinions are calculated based on the relativity between positive and negative words in each source (number of positive words / number of negative words). While David Pakman Show is the only source that is relatively negative regarding polarity of opinions, others have relatively positive feedback, with USA Today and CNBC Television having 2 times more positive words, 181 than negative words in comments. USA Today (n = 181, 15.8%), Bloomberg Television (n = 175, 14.2%), and LiveNOW from FOX (n = 2814, 13.4%) are the top 3 sources that contain most positive words in their comments, while David Pakman Show (n = 5622, 10.3%), MSNBC (n = 3520, 9.3%), and The Daily Show (n = 2640, 8.7%) are the top 3 sources that contain most negative words in comments.

Figure 10. Polarity in comments (sorted by source)

Discussion and Conclusion

Despite the framing of Trump’s engagement in crypto-related events in the news titles demonstrates an overall sentiment of surprise, which is neutral-negative, the overall public opinions are more positive (53.86% ) than negative (46.14%) on Trump’s engagement with cryptocurrency, which is consistent with the results of positive sentiments on cryptocurrency from previous studies Although surprise can be either positive or negative, in this context, it may represent “an element of anticipating the unexpected and unknown” (Noordewier & Breugalmans, 2013), which encounters the attitude of news outlets on cryptocurrency as an unknown and unexpected topic but with an anticipation of new and emerging technologies being an inevitable topic to address in the coming political campaigns.

Prevalent public opinions are governed in three main themes: crypto-specific,

campaign-specific, and attitude/action-specific. While the crypto-specific theme emphasizes the specific events covered in the news and the existence of cryptocurrency per se, implying no specific meaning, concerns on crypto as scams are detected. In terms of the

campaign-specific theme, positive sentiments are detected toward the presidential candidate Trump, showing a significantly higher level of trust by constantly addressing him as “president trump”. For the attitude/action-specific theme, while negative sentiments are detected toward Trump and crypto, people who left a comment are more willing to vote for Trump than Harris.

This study yields both theoretical and practical implications. While priming campaign issues and candidate characteristics may augment call-to-action of voters, a campaign issue that is on new technology but controversial at the same time does not necessarily yield bad results despite the negative framing. Miskolczy and Gherghina (2024) offered an explanation that male politicians are more likely to survive from character assassination (Samoilenko, 2021) when they are attacked by peers of similar power, adapt their responses to match the complexity of the attack, and when the media does not heavily report on the incident. In 2016, factors such as income inequality, the size of the Black male and female populations, the size of the Hispanic male population, and the percentage of the population with a college degree had significantly larger coefficients than in 2012, resulting in a greater marginal effect boosting the Democratic candidate's vote share. Conversely, counties experiencing higher poverty rates and those situated on the outskirts of urban centers showed a significantly larger marginal effect favoring the Republican candidate's vote share in 2016 compared to 2012 (Kahane, 2020). Practically, since the Harris campaign does not yet develop a specific campaign curated towards cryptocurrency (Christian, 2024), to launch a political campaign of

cryptocurrency that the counterparty does not yet have may result in a favorable position by incorporating both critical voters groups by virtue of the ideological association of cryptocurrency with financial inclusion and democracy.

This study is not without limitations. First, the time at which this research is conducted is too close to the time when these videos were posted, which may not be sufficient to accumulate enough comments, affecting the generalizability of the result. Second, the identity of the people who made the comments cannot be verified, resulting in a selection bias of the sample. Third, people who leave a comment under these videos may have confirmation bias, which means that they are already Trumpers that they leave a comment as a means to support Trump. Last but not least, the method assumes emotion can be carried over contexts, but the lexicon used in this study may not be sufficient enough to capture domain-specific sentiment in the scope of cryptocurrency. Future study on cryptocurrency in election can replicate the study design in Druckman (2004) to combine a content analysis of the news and titles first, and conduct a subsequent research on verified voters to validate the campaign effect and eliminate confirmation bias and selection bias. Methodologically, future research can make an attempt to develop a domain-specific lexicon in cryptocurrency in order to increase the accuracy of capturing public sentiment. Moreover, as the techniques for textual analysis advances rapidly with the development of Large Language Model and Natural Language Processing, multimodal sentiment analysis (Zhu et al., 2022) and multilingual bidirectional encoder representations from transformers(mBERT) (Ho & Chan, 2023) can be utilized for more methodologically rigorous analysis so that domain-specific multimodal data and texts in other languages than English can be appropriately evaluated.

References

Aibai, A., Julaiti, J., & Gou, S. (2024). The asymmetric effects of upside and downside risks in cryptocurrency markets: Insights from the LUNA and FTX crises. Finance Research Letters, 67, 105750.

Ainapure, B. S., Pise, R. N., Reddy, P., Appasani, B., Srinivasulu, A., Khan, M. S., & Bizon,

N. (2023). Sentiment Analysis of COVID-19 Tweets Using Deep Learning and

Lexicon-Based Approaches. Sustainability, 15(3), 2573-. https://doi.org/10.3390/su15032573

Allon, F. (2018). Money after Blockchain: Gold, Decentralised Politics and the New Libertarianism. Australian Feminist Studies, 33(96), 223–243. https://doi.org/10.1080/08164649.2018.1517245

Al-Tawil, T. N. (2023). Anti-money laundering regulation of cryptocurrency: UAE and global approaches. Journal of Money Laundering Control, 26(6), 1150–1164. https://doi.org/10.1108/JMLC-07-2022-0109

Alvarez, F., Argente, D., & Van Patten, D. (2023). Are cryptocurrencies currencies? Bitcoin as legal tender in El Salvador. Science (American Association for the Advancement of Science), 382(6677), eadd2844–eadd2844. https://doi.org/10.1126/science.add2844

Ba, H.-L., & Şen, Ö. F. (2024). Explaining variation in national cryptocurrency regulation: implications for the global political economy. Review of International Political Economy : RIPE, 31(5), 1472–1495. https://doi.org/10.1080/09692290.2024.2325403

Bartoletti, M., Lande, S., Loddo, A., Pompianu, L., & Serusi, S. (2021). Cryptocurrency Scams: Analysis and Perspectives. IEEE Access, 9, 148353–148373. https://doi.org/10.1109/ACCESS.2021.3123894

Brekhov, G. S. (2022). Crypto-Anarchism: The Ideology of Blockchain Technologies. RUDN Journal Of Political Science, 24(3), 393-407. doi: 10.22363/2313-1438-2022-24-3-393-407

Britzolakis, A., Kondylakis, H., & Papadakis, N. (2020). A Review on Lexicon-Based and Machine Learning Political Sentiment Analysis Using Tweets. International Journal of Semantic Computing, 14(4), 517–563. https://doi.org/10.1142/S1793351X20300010

Buterin, V, Hitzig, Z, & Weyl, E. G (2018). Liberal Radicalism: A Flexible Design For Philanthropic Matching Funds. SSRN. https://dx.doi.org/10.2139/ssrn.3243656

Christian. R. (2024, September 24). Election 2024: Where Trump and Harris stand on crypto. In Kennedy. E. (Ed.), Bankrate. Retrieved October 19, 2024 from: https://www.bankrate.com/investing/where-trump-and-harris-stand-on-crypto/?tpt=a

Chung, S., Kim, K., Lee, C. H., & Oh, W. (2023). Interdependence between online peer‐to‐peer lending and cryptocurrency markets and its effects on financial inclusion. Production and Operations Management, 32(6), 1939-1957. https://doi.org/10.1111/poms.13950

Claypool, R. (2024). Big crypto, big spending: Crypto corporations spend an unprecedented

$119 million influencing elections (August 21, 2024). Retrieved from https://www.citizen.org/article/big-crypto-big-spending-2024/

Dang, N. C., Moreno-García, M. N., & De la Prieta, F. (2020). Sentiment analysis based on deep learning: A comparative study. Electronics, 9(3), 483.

di Prisco, D., & Strangio, D. (2021). Technology and financial inclusion: a case study to evaluate potential and limitations of Blockchain in emerging countries. Technology Analysis & Strategic Management, ahead-of-print(ahead-of-print), 1–14. https://doi.org/10.1080/09537325.2021.1944617

Druckman, J. N. (2004). Priming the Vote: Campaign Effects in a U.S. Senate Election.

Political Psychology, 25(4), 577–594. https://doi.org/10.1111/j.1467-9221.2004.00388.x

Entman, Robert M. (2005). The nature and sources of news. In: Kathleen H. Jamieson and Geneva Overholser (eds.), The Press, 48-65. New York, NY: Oxford University Press,

Escobar, K. E. S., Narváez, P. A. A., & Escobar, C. P. S. (2022). Criminal Risks of Virtual Currencies: New Challenges for Criminal Law. NeuroQuantology, 20(13), 791.

Fletcher, E., Larkin, C. J., & Corbet, S. (2020). Cryptocurrency Regulation: Countering Money Laundering and Terrorist Financing. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.3704279

Fu, S., Wang, Q., Yu, J., & Chen, S. (2023, May). FTX collapse: a Ponzi story. In International Conference on Financial Cryptography and Data Security, pp. 208-215. Cham: Springer Nature Switzerland.

Gilardi, F., Gessler, T., Kubli, M., & Müller, S. (2022). Social Media and Political Agenda Setting. Political Communication, 39(1), 39–60. https://doi.org/10.1080/10584609.2021.1910390

Guo, L., & Vargo, C. (2020). “Fake News” and Emerging Online Media Ecosystem: An Integrated Intermedia Agenda-Setting Analysis of the 2016 U.S. Presidential Election. Communication Research, 47(2), 178-200. https://doi.org/10.1177/0093650218777177

Halpern, D., & Gibbs, J. (2013). Social media as a catalyst for online deliberation? Exploring the affordances of Facebook and YouTube for political expression. Computers in human behavior, 29(3), 1159-1168.

Haq, I., El Shahat, I., Abatemarco, M., & Bates, C. (2024). The Impact of 9/11 on Money Laundering. American Behavioral Scientist, 68(2), 261-273.

Ho, J. C.- ting, & Chan, C.- hong. (2023). Evaluating Transferability in Multilingual Text Analyses. Computational Communication Research, 5(2). https://doi.org/10.5117/CCR2023.2.3.HO

Hossain, M.S. (2021), "What do we know about cryptocurrency? Past, present, future", China Finance Review International, 11(4), pp. 552-572.

https://doi.org/10.1108/CFRI-03-2020-0026

Huang, X., Zhang, W., Tang, X., Zhang, M., Surbiryala, J., Iosifidis, V., ... & Zhang, J. (2021). Lstm based sentiment analysis for cryptocurrency prediction. In Database Systems for Advanced Applications: 26th International Conference, DASFAA 2021, Taipei, Taiwan, April 11–14, 2021, Proceedings, Part III 26 , pp. 617-621. Springer International Publishing.

Jain, P. (2023). From Crypto Wild West to Regulated Frontier: Unleashing the Potential of Blockchain Technology. W. Va. L. Rev., 126, 239.

Jara-Vera, V. (2022). New directions in crypto-politics. Journal of Libertarian Studies, 25(1), 272-306.

Johnson, D. (2023). Cryptocurrency and public policy: implications for democracy and governance. Routledge. https://doi.org/10.4324/9781003308195

Julia, S. and Robinson, D. (2016). Tidytext: Text Mining and Analysis Using Tidy Data Principles in r. Journal of Open Source Software, 1(3). https://doi.org/10.21105/joss.00037.

Kahane, L. H. (2020). Determinants of County-Level Voting Patterns In the 2012 and 2016 Presidential Elections. Applied Economics, 52(33), 3574–3587. https://doi.org/10.1080/00036846.2020.1713985

Kerr, D. S., Loveland, K. A., Smith, K. T., & Smith, L. M. (2023). Cryptocurrency Risks, Fraud Cases, and Financial Performance. Risks (Basel), 11(3), 51-. https://doi.org/10.3390/risks11030051

Kew Yen Nee, S., & Eng Thye, K. (2024). Decentralized Electronic Voting With Ethereum Blockchain In Democratic And Political Elections. Journal of Computational Innovation and Analytics (JCIA), 3(2), 111–127. https://doi.org/10.32890/jcia2024.3.2.5

Lee, H. (2023). The Digital Democracy of Blockchain: Beyond Libertarianism to Commonism. In: Affective Capitalism. Palgrave Macmillan, Singapore. https://doi.org/10.1007/978-981-99-8174-8_8

Leuprecht, C., Jenkins, C., & Hamilton, R. (2023). Virtual money laundering: policy implications of the proliferation in the illicit use of cryptocurrency. Journal of Financial Crime, 30(4), 1036-1054.

Lin, W. C., Li, X., Tang, K., & Yang, Y. (2022). Crypto Wash Trading. NBER Working Paper Series. https://doi.org/10.3386/w30783

Liptak, K., & Vazquez, M. (2024, July 27). Trump endorses Bitcoin and cryptocurrency in surprise campaign shift. CNN.

https://www.cnn.com/2024/07/27/politics/donald-trump-bitcoin-cryptocurrency/index.html

Liu, J., Makarov, I., & Schoar, A. (2023). Anatomy of a run: The terra luna crash (No. w31160). National Bureau of Economic Research.

Malabou, C. (2020). Cryptocurrencies: Anarchist Turn or Strengthening of Surveillance Capitalism? From Bitcoin to Libra. Australian Humanities Review, 66, 1–1.

Malhotra, P. (2024). “What You Post in the Group Stays in the Group”: Examining the Affordances of Bounded Social Media Places. Social Media + Society, 10(3), 20563051241285777.

Miskolczy, I., & Gherghina, S. (2024). Making words harmless: Why politicians survive character assassination attacks. The British Journal of Politics and International Relations,

0(0). https://doi.org/10.1177/13691481241264980

Moy, P. & Bosch, B. (2013). 16 Theories of public opinion. In P. Cobley & P. Schulz (Ed.), Theories and Models of Communication (pp. 289-308). Berlin, Boston: De Gruyter Mouton. https://doi.org/10.1515/9783110240450.289

Nair, M., Abd-Elmegid, L. A., & Marie, M. I. (2024). Sentiment analysis model for cryptocurrency tweets using different deep learning techniques. Journal of Intelligent Systems, 33(1), 20230085.

Noordewier, M. K., & Breugelmans, S. M. (2013). On the valence of surprise. Cognition and Emotion, 27(7), 1326–1334. https://doi.org/10.1080/02699931.2013.777660

Obreja, D. M. (2024). The social side of cryptocurrency: Exploring the investors’ ideological realities from Romanian Facebook groups. New Media & Society, 26(5), 2748–2765. https://doi.org/10.1177/14614448221092028

Ong, B., Lee, T. M., Li, G., & Chuen, D. L. K. (2024). Evaluating the potential of alternative cryptocurrencies. In Handbook of digital currency (pp. 79-132). Academic Press.

Parekh, R., Patel, N. P., Thakkar, N., Gupta, R., Tanwar, S., Sharma, G., ... & Sharma, R. (2022). DL-GuesS: Deep learning and sentiment analysis-based cryptocurrency price prediction. IEEE Access, 10, 35398-35409.

Pew Research Center. (2014, October 21). Political polarization & media habits. Pew Research Center. Retrieved October 18, 2024, from https://www.pewresearch.org/journalism/2014/10/21/political-polarization-media-habits/

Qian, C., Mathur, N., Zakaria, N. H., Arora, R., Gupta, V., & Ali, M. (2022). Understanding public opinions on social media for financial sentiment analysis using AI-based techniques. Information Processing & Management, 59(6), 103098-. https://doi.org/10.1016/j.ipm.2022.103098

Raheman, A., Kolonin, A., Fridkins, I., Ansari, I., & Vishwas, M. (2022). Social media sentiment analysis for cryptocurrency market prediction. arXiv preprint arXiv:2204.10185.

Rajput, A. (2020). Natural language processing, sentiment analysis, and clinical analytics. In

Innovation in health informatics (pp. 79-97). Academic Press.

Restad, H. E. (2020). What makes America great? Donald Trump, national identity, and U.S. foreign policy. Global Affairs (Abingdon, Oxfordshire, UK), 6(1), 21–36. https://doi.org/10.1080/23340460.2020.1734955

Rinker, T. W. (2021). sentimentr: Calculate Text Polarity Sentiment. Buffalo, New York. https://github.com/trinker/sentimentr.

Rohanifar, Y., Chandra, P., Rahman, M. A., & Ahmed, S. I. (2021, May). Money whispers: Informality, international politics, and immigration in transnational finance. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-14).

Rouhani, S. & Abedin, E. (2020), "Crypto-currencies narrated on tweets: a sentiment analysis approach", International Journal of Ethics and Systems, Vol. 36 No. 1, pp. 58-72. https://doi.org/10.1108/IJOES-12-2018-0185

Saif,M., and Turney,P. (2013). Crowdsourcing a Word-Emotion Association Lexicon.

Computational Intelligence, 29 (3): 436–65.

Samoilenko, S.A. (2021). Character assassination: The sociocultural perspective. Journal of Applied Social Theory, 1(3): 186–205.

Schneider, N. (2024), Web3 is the Opportunity We Have Had All Along: Innovation Amnesia and Economic Democracy", DuPont, Q., Kavanagh, D. and Dylan-Ennis, P. (Ed.) Defining Web3: A Guide to the New Cultural Economy (Research in the Sociology of Organizations, Vol. 89), Emerald Publishing Limited, Leeds, pp. 13-25.

https://doi.org/10.1108/S0733-558X20240000089002

Schuetz, S., & Venkatesh, V. (2020). Blockchain, adoption, and financial inclusion in India: Research opportunities. International Journal of Information Management, 52, 101936–101938. https://doi.org/10.1016/j.ijinfomgt.2019.04.009

Setyawan, A. A., Winotoatmojo, H. P., & Ananda, N. K. P. (2024). Cryptocurrency And Financial Inclusion: Bridging The Gap In Emerging Countries. Dinasti International Journal of Economics, Finance & Accounting, 5(2), 510-516.

Shehata, A., & Strömbäck, J. (2021). Learning Political News From Social Media: Network Media Logic and Current Affairs News Learning in a High-Choice Media Environment.

Communication Research, 48(1), 125–147. https://doi.org/10.1177/0093650217749354

Statista. (2024). Cryptocurrencies - United States. Statista. Retrieved October 20, 2024, from https://www.statista.com/outlook/fmo/digital-assets/cryptocurrencies/united-states

Taherdoost, H., & Madanchian, M. (2023). Artificial intelligence and sentiment analysis: A review in competitive research. Computers, 12(2), 37.

Teichmann, F. M. J., & Falker, M.-C. (2021). Money laundering via cryptocurrencies – potential solutions from Liechtenstein. Journal of Money Laundering Control, 24(1), 91–101. https://doi.org/10.1108/JMLC-04-2020-0041

Teichmann, F. M. J. (2018). Financing terrorism through cryptocurrencies – a danger for Europe? Journal of Money Laundering Control, 21(4), 513–519. https://doi.org/10.1108/JMLC-06-2017-0024

Travis, H. (2020), Crypto Coin Offerings and the Freedom of Expression. Chapman Law Review, 23 (2). http://dx.doi.org/10.2139/ssrn.3670993

Tsai, C., Nandy, G., House,D. and Carroll, J. (2024). Ensuring Transparency in Using ChatGPT for Public Sentiment Analysis. In Proceedings of the 25th Annual International Conference on Digital Government Research (dg.o '24). Association for Computing Machinery, New York, NY, USA, 627–636. https://doi.org/10.1145/3657054.3657128

U.S. Department of the Treasury. (n.d.). Money laundering. U.S. Department of the Treasury.

Retrieved October 18, 2024, from

https://home.treasury.gov/policy-issues/terrorism-and-illicit-finance/money-laundering

Vaast, E., Safadi, H., Lapointe, L., & Negoita, B. (2017). Social media affordances for connective action. MIS quarterly, 41(4), 1179-1206.

Weiss,D. , Wolmer, J. and Vatsa, A. (2022). Blockchain-based Electronic Voting System for Modern Democracy: A Review, 2022 IEEE Integrated STEM Education Conference (ISEC), Princeton, NJ, USA, pp. 162-166, doi: 10.1109/ISEC54952.2022.10025096.

Welbers, K., Van Atteveldt, W., & Benoit, K. (2017). Text Analysis in R. Communication Methods and Measures, 11 (4), 245-265. http://www.tandfonline.com/doi/10.1080/

19312458.2017.1387238

World Liberty Financial. (n.d.). Earn and borrow crypto. Retrieved October 20, 2024, from https://www.worldlibertyfinancial.com

Zhang, L., Ghosh, R., Dekhil, M., Hsu, M., & Liu, B. (2011). Combining lexicon-based and learning-based methods for Twitter sentiment analysis. HP Laboratories, Technical Report HPL-2011, 89, 1-8.

Zhu, T., Li, L., Yang, J., Zhao, S., Liu, H., & Qian, J. (2022). Multimodal sentiment analysis with image-text interaction network. IEEE transactions on multimedia, 25, 3375-3385.