The New Science of Predicting Election Victories in Developing Countries”

 Reshaping Caribbean Democracy Through Data Science

1.       The Science of Predicting an Election: The DPBA Grenada 2022 Case Study

In June 2022, Dunn Pierre Barnett and Company Canada Ltd. (DPBA) made history by accurately predicting the outcome of Grenada’s general election, forecasting a win for the National Democratic Congress (NDC) in one of the Caribbean’s most surprising political upsets. This success wasn’t based on guesswork; it resulted from an intricate scientific framework that combined data analytics, artificial intelligence, psychographic profiling, and behavioural targeting. According to DPBA’s Director of Research, Dr. Justine Pierre, the process began with establishing the Grenadian Voter Universe. The team stratified the electorate across 18 variables, including geography, age, gender, religion, education level, employment sector, union affiliation, and even sporting interests. These layers of stratification allowed for a nuanced understanding of the diverse and dynamic Grenadian voter base.

What set DPBA apart was their application of psychographic profiling using the OCEAN personality model: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. This approach allowed them to anticipate who would vote and why they would vote a certain way. Voters high in openness, for example, responded positively to progressive and reformist messaging, while fear-based appeals more influenced those high in neuroticism. This insight allowed DPBA to support its clients in tailoring hyper-personalised ads, emails, and social media content that emotionally and psychologically resonated with specific voter subgroups. Focus group sessions, surveys, and data from Facebook, the Grenadian telephone directory, and diaspora communities added further depth to the analysis, creating a 360-degree voter profile.

DPBA also utilised machine learning and real-time predictive analytics to simulate voter behaviour, enabling continuous adjustments throughout the campaign. A/B testing helped identify which messages performed best among different segments, while real-time monitoring of like-and-share ratios on social media platforms provided insights into message virality and influence. They also captured and analysed facial expressions during political speeches using AI-enhanced video analysis, allowing them to detect audience emotional responses. Their groundbreaking methodology proved that intuition alone is no longer enough in modern Caribbean politics. Dr. Pierre argues that political parties must transition toward data-driven strategy if they want to remain competitive. With over 95% prediction accuracy in the Grenada case, DPBA is proving that the science of elections is no longer theory; it’s practice. Their model now sets a new gold standard for political consulting in the Caribbean.

2.      The Importance of Data and Social Influencers in Elections:

Data and social influencers wield significant influence in the evolving landscape of general elections, particularly in digitally connected societies like Grenada. The Digital 2023: Grenada report reveals a substantial online presence, with 101,100 internet users, representing 79.9% of the population. Platforms like Facebook, Instagram, and TikTok are pivotal in disseminating political information, especially appealing to younger voters who exhibit a higher level of distrust towards traditional media. With 135,000 cellular connections, political communication has become more personalised and instantaneous, utilising platforms like Facebook Messenger and LinkedIn for targeted outreach to both grassroots voters and professionals.

 Social media influencers, such as Kem Jones, Neva Newman and Prime Minister Dickon Mitchell, have emerged as influential figures in Grenada, capable of mobilising opinions and amplifying political messages. These digital personalities, particularly micro and macro-influencers, are perceived as relatable and trustworthy, enhancing their ability to shape narratives, endorse candidates, and clarify policy positions that resonate with the populace. Their impact lies in their extensive reach and the emotional trust and peer-level credibility they foster among their followers.

Strategically, DPBA leverages influencer analysis and voter data to tailor psychologically attuned political messaging. By employing advanced analytics, DPBA monitors engagement patterns, identifies influential figures gaining traction among undecided voters, and traces the flow of political narratives across various social media platforms. This approach enables observing how content shared by prominent Grenadian influencers correlates with shifts in voter sentiment towards certain political entities, even before traditional polls reflect such changes. This fusion of influencer behaviour with voter sentiment analysis signifies a groundbreaking development in Caribbean elections, where social data mirrors public opinion and actively shapes it. As digital connectivity expands, social influencers will continue to be indispensable allies in any data-centric electoral strategy.

Photo taken from: https://www.facebook.com/photo.php?fbid=5286038204847979&id=303579073093942&set=a.581861325265714&locale=nl_BE

3.      The Influence of Kem Jones in Grenada’s Political Landscape

Kem Jones has become a prominent figure in Grenada’s political sphere, utilising his platforms to engage with the public and shape opinions. As a political activist and talk-show host, Jones leverages social media, particularly on Facebook and Instagram, to share his viewpoints and rally support. His outreach extends to a broad audience, notably the youth demographic, who are increasingly relying on digital platforms for news and political insights. Jones delves into pertinent national issues, offering a critical analysis of governmental decisions, urging the administration to reconsider certain financial practices for greater accountability.

Beyond commentary, Jones actively contributes to moulding Grenada’s political narrative, resonating with many who feel marginalised in traditional media. By addressing topics relevant to the average Grenadian, he cultivates a sense of community and common purpose among his followers. Through interactive initiatives like live discussions, Jones promotes public engagement and discourse, which are vital elements of a thriving democracy. In an age where digital participation holds significant weight, Kem Jones exemplifies how social media can drive change and influence national dialogues.

 4.      DPBA’s Advanced Social Media Monitoring

DPBA’s social media monitoring capabilities exceed traditional metrics like likes, comments, and shares. Their proprietary algorithms are designed to analyse the emotional sentiment behind each interaction, assessing whether voters express excitement, anger, fear, or support in response to campaign content. This emotional layer of analysis allows DPBA to move past superficial engagement numbers and tap into the deeper psychological drivers of voter behaviour. For instance, in the weeks leading up to the Grenada 2022 elections, DPBA detected a surge in positive sentiment toward the NDC’s messaging, particularly among undecided voters, long before traditional polls reflected any momentum shift.

Another standout feature of DPBA’s toolkit is visual content recognition. Using AI-driven systems, they scanned campaign images, posters, memes, and videos to identify common symbols, colour schemes, slogans, and emotional cues that resonated with target audiences. This helped political parties adjust their visuals in real time to reflect voter expectations better. Additionally, DPBA tracked the network effects of political messaging across platforms by mapping how ideas spread, which influencers were amplifying them, and which voter segments were most responsive. This allowed campaigns to fine-tune their outreach strategies and maximise organic reach.

Equally important was DPBA’s capacity to monitor real-time response patterns to candidate statements across social platforms. By capturing immediate reactions, whether spikes in retweets, comment sentiment shifts, or geographic clustering of engagement, DPBA helped candidates refine their messaging instantly. Combining emotional analysis, visual intelligence, and network mapping, this multidimensional approach proved that social media activity is a leading indicator of voter intent. In Grenada, these insights were pivotal in alerting the NDC to its rising support weeks before it became evident in traditional polling data, solidifying DPBA’s role as a trailblazer in the future of Caribbean political strategy.

 5.      Social Media as the New Political Battleground

DPBA has pioneered a sophisticated approach to social media monitoring, going well beyond traditional likes and shares. Using proprietary algorithms, DPBA analyses emotional sentiment in comments, reactions, and shares, giving political campaigns real-time insight into how voters feel about particular messages, images, and candidates, not just what they see or click.

One of the most innovative aspects of DPBA’s methodology is its capacity to conduct visual content recognition and network mapping. Using AI to scan and categorise campaign imagery, slogans, and symbols, DPBA can evaluate how well visuals align with target voter expectations and cultural context. Additionally, by analysing network effects, the speed and reach of political messaging across interconnected social media platforms, DPBA can identify digital influencers, content virality patterns, and the geographic spread of political discourse. This empowers political campaigns to adjust their messaging strategy, target persuasive content to undecided voters, and amplify messages that resonate organically with key constituencies.

This multidimensional strategy was a game-changer in Trinidad and Tobago in 2025 and in Grenada’s 2022 general election. While traditional polling failed to capture the early shift in voter sentiment, DPBA detected a rising tide of support for the United National Congress (UNC), and the National Democratic Congress (NDC) weeks in advance. Emotional spikes in online sentiment, viral engagement with UNC and NDC visuals, and rapid positive response patterns to candidate speeches were all picked up through DPBA’s tools, proving social media to be a leading rather than lagging indicator of voter intent. Their success in reading the digital electorate before the ballots were cast redefined how Caribbean elections are now understood and fought. In this new political battleground, data is the weapon, emotion is the terrain, and firms like DPBA lead the charge.

 6.      Cell Tower Analytics: Mapping the Invisible Electorate

DPBA has harnessed cell tower analytics to map voter behaviour in ways previously thought impossible. By analysing anonymised mobile phone data, the firm gains real-time insights into population density shifts, urban-to-rural migration trends, and daily commuting routes, which directly influence voter accessibility and campaign strategy. This technology allows campaigns to understand when and where voters are most active, helping to determine optimal times for canvassing, town hall meetings, and message delivery. Unlike static voter rolls or census data, which may quickly become outdated, cell tower analytics provide a live, dynamic picture of the electorate as it moves, interacts, and gathers in the real world.

One of the most striking applications of this technology was in Grenada’s Carriacou constituency during the 2022 general elections. Traditional polling had failed to account for temporary population shifts due to work migration and inter-island travel. However, DPBA’s geolocation clustering analysis revealed unexpected voter concentrations during campaign events and weekends in specific areas. Additionally, roaming data patterns helped map connections between diaspora voters and their home constituencies, allowing campaigns to time digital engagement strategies precisely. These findings refined the campaign’s ground deployment strategy and ensured that political messaging reached voters where they physically were, not just where voter lists said they lived.

 7.      Psychology Behind the Vote

DPBA’s psychographic profiling system represents a groundbreaking advancement in understanding Caribbean electoral behaviour by focusing on the inner workings of the voter’s mind. Rather than relying solely on surface-level demographics like age or income, DPBA’s model explores cognitive biases, emotional triggers, and personality traits that shape how voters process political information. The system also factors in cultural archetypes unique to the Caribbean, such as respect for authority, generational loyalty, and religious values, influencing voter alignment. This multi-layered approach allows campaigns to tailor messages that resonate on a deeper psychological level, increasing the chances of voter engagement and persuasion.

During the Grenada 2022 general elections, these psychological insights proved pivotal. DPBA found that voters scoring high in “openness”, imaginative, curious, and receptive to change, gravitated toward the National Democratic Congress (NDC) and its promise of reform, transparency, and youth leadership. Meanwhile, those high in “conscientiousness” who value order, reliability, and tradition remained loyal to the New National Party (NNP), which emphasised stability and experience. These findings directly influenced both parties’ campaign strategies: the NDC crafted a bold, forward-looking message for progressive voters, while the NNP doubled down on legacy and continuity themes to retain their conservative base. This strategic use of voter psychology gave campaigns a competitive edge that traditional polling methods could never achieve.

 8.      Predictive Modelling: Seeing Around Corners

DPBA’s election forecasting represents a quantum leap from traditional polling through its sophisticated predictive modelling, incorporating four revolutionary dimensions. First, economic indicators serve as leading predictors, with variables like unemployment rates, inflation, and foreign exchange availability proving more reliable than standard voter intention surveys in Grenada’s 2022 election. Second, the models track sentiment decay rates, measuring how long specific campaign messages (e.g., healthcare reform promises) influence voter behaviour before losing impact. Third, voter attention span measurements, derived from digital engagement patterns, determine the optimal frequency and format for campaign communications. These components feed into probability-weighted scenario planning that generates dynamic projections rather than static snapshots.

What sets DPBA’s models apart is their adaptive machine learning framework. During Grenada’s election cycle, the system detected that traditional “rally effects” (short-term boosts from political events) decayed 42% faster among youth voters compared to 2018 data – a finding that prompted the NDC to shift resources to sustained digital engagement. The models also incorporated real-time economic shocks; when Grenada’s tourism sector recovered unexpectedly in Q2 2022, DPBA immediately adjusted its projections for coastal constituencies. This approach allowed the firm to correctly predict the NDC’s victory margin within ±1.5 seats six weeks before election day, outperforming traditional polls that still showed a dead heat.

The implications for Caribbean democracy are profound. DPBA has demonstrated that elections follow predictable patterns when analysed through this multidimensional lens. Their 2022 Grenada model achieved 93% accuracy by weighting economic factors (40%), sentiment trajectories (30%), attention metrics (20%), and scenario probabilities (10%). As other firms attempt replication, DPBA continues advancing the technology, recently incorporating neural network analysis of speech patterns from campaign videos to predict voter trust levels. This isn’t just forecasting – it’s creating a new science of democratic outcomes where probabilities replace guesswork. The 2025 Trinidad and Tobago election was a significant test case for these evolving methodologies.

 9.      Microtargeting: The End of One-Size-Fits-All Politics

DPBA has revolutionised Caribbean political campaigning through its advanced microtargeting capabilities, which render traditional blanket messaging obsolete. By leveraging household-level customisation, the firm enables campaigns to tailor communications based on a voter’s precise location, socioeconomic status, and household composition, ensuring a fisherman in Gouyave receives fundamentally different messaging than a hotel worker in Grand Anse. This granular approach is supercharged by dynamic content adaptation, where AI algorithms adjust campaign materials in real-time based on voter engagement metrics. For example, if data shows a particular neighbourhood responding poorly to economic reform messages but positively to healthcare content, DPBA’s system automatically pivots the messaging strategy within hours, not weeks.

The true power of DPBA’s microtargeting lies in its cross-platform sequencing and behavioural response tracking. The system orchestrates a voter’s journey across WhatsApp, Facebook, email, and SMS, delivering a coherent but platform-optimised narrative that builds persuasion over time. Sophisticated tracking mechanisms measure how each interaction influences voter sentiment, creating a feedback loop that continuously refines outreach. Grenada’s 2022 election allowed the NDC to identify and re-engage over 7,000 “soft supporters” who had initially ignored campaign messages but later became decisive voters. Such capabilities don’t just enhance political outreach, they redefine the very nature of democratic engagement in the digital age.

 10.  The New Campaign Playbook: Data-Driven Victories in the Caribbean

DPBA has revolutionised Caribbean electoral politics by introducing a data-centric playbook that propelled the United National Congress (UNC) to victory in Trinidad and Tobago’s April 28, 2025, elections. By replacing political gut instincts with AI-powered analytics, the firm identified three critical trends that traditional polls missed: growing economic anxiety among middle-class voters in marginal constituencies, unexpected openness to UNC messaging in traditionally PNM-leaning Afro-Trinidadian communities, and record youth registration numbers in key battlegrounds like St. Augustine and Barataria/San Juan. DPBA’s continuous monitoring system tracked over 5,000 digital data points daily from WhatsApp group sentiments to mobile payment patterns at local markets, allowing the UNC to shift resources in real-time. This approach proved decisive when their models detected weakening PNM support in Central Trinidad two weeks before election day, triggering a targeted policy blitz on fuel subsidies that swung undecided voters.

The UNC’s triumph in 2025, following DPBA’s earlier success in Grenada in 2022, validates their four-pillar methodology: data-driven decisions (abandoning ineffective rally-based campaigning), precision communication.  (7,000 unique ad variations tailored to voter profiles), 24/7 sentiment tracking (identifying the PNM’s crumbling “law and order” narrative), and psychographic segmentation (targeting “openness” traits in swing voters). Their final prediction was a landslide in favour of the UNC.  These back-to-back successes in Grenada and Trinidad demonstrate that in today’s Caribbean, elections are won not by the loudest slogans but by campaigns that harness predictive analytics, behavioural science, and real-time adaptation to speak directly to voters’ evolving concerns.

 Ethical Considerations in Data-Driven Democracy

As data-driven campaigning becomes the norm in Caribbean elections, it raises critical ethical questions about protecting voter rights and the integrity of the democratic process. DPBA’s advanced methods, while remarkably effective, operate in a landscape where voter privacy, data security, and transparency are of growing concern. The firm utilises vast datasets that include behavioural insights, social media activity, and personality profiles, which, if misused, could threaten personal privacy and undermine public trust. To address these concerns, DPBA maintains strict ethical protocols that include anonymised data collection, informed consent for surveys, secure data storage, and a commitment to avoid exploitative practices like misinformation or fear-based manipulation.

Still, as these technologies become more sophisticated and widespread, Caribbean electoral commissions and policymakers are beginning to draft new guidelines to regulate political data usage. Issues such as algorithmic bias, lack of transparency in targeting practices, and the ethical boundary between persuasion and psychological manipulation are now central to electoral reform debates. In nations like Jamaica, Trinidad and Tobago, and Barbados, electoral oversight bodies are increasingly exploring frameworks for digital campaign disclosure, data protection standards, and independent audits of campaign technology vendors. While firms like DPBA lead by example, the challenge moving forward is to ensure that innovation serves democracy rather than distorting it, upholding fairness, accountability, and voter empowerment in an era of political digitalisation.

 About the Company

Dunn Pierre Barnett and Company Canada Ltd is a full-service Management Consulting, Program Evaluation, and Technology firm specializing in Collecting, Disseminating, and analyzing Labour Market, Workforce, Skills, TVET and other data on diverse populations globally, including Black, Indigenous, and People of Color (BIPOC), the diaspora, and other racialized populations across Canada, Africa, USA, and the Americas.

Website: https://dpbglobal.com/

Email: info@dpbglobal.com

About the lead consultant:

Dr. CLEOPHAS JUSTINE PIERRE:

Dr. CLEOPHAS JUSTINE PIERRE: Researcher/Demographer Statistician/Labour Market consultant

Dr Cleophas Justine Pierre is an ILO-trained statistician and labour market and migration consultant specialising in employment, underemployment, and unemployment. He is the leading researcher in Canada and the Caribbean Region.

His primary area of expertise is in the field of Statistics, Migration and Labour Market Research, which involves matters relating to forced migration, Anti-human trafficking research, Migrant smuggling, data analytics, social research, conducting surveys, Artificial Intelligence and the formulation of Labour Market Information Systems (LMIS)

Dr Pierre and his team have conducted upwards of 300 consultancies in more than 40 countries for many international organisations, such as the World Bank and World Health Organisation, the EU, the Caribbean Development Bank, the Government of Canada, OECS, the Government of Turks and Caicos Island, CARICOM, the ILO, and other local, regional, and international agencies.

Dr Pierre was part of a small team that conducted the EU Global “Thematic Evaluation Survey” in employment and social inclusion, including vocational training and the Labour Market Needs Assessment. This survey was done in the countries of Chile, Jamaica, Kyrgyzstan, Morocco, Ukraine, South Africa, and Vietnam.

Between 2016 and 2023, Dr. Pierre and his firm conducted twenty-four comprehensive studies in the fields of Statistics, Labour Market Needs Assessment, Migration, and Human Trafficking in Canada and the Caribbean Region, of which he was one of the principal researchers.

For more information, visit our Black Data and Information portal https://www.dpbglobal.net/

 Reference:

1)      Pierre, J. C. (2022). The science of predicting an election: A case study of Grenada’s 2022 general elections. Dunn, Pierre Barnett and Company Canada Ltd.

2)      Digital 2023: Grenada. (2023). Datareportal – Essential insights into the state of digital in Grenada. Retrieved from https://datareportal.com/reports/digital-2023-grenada

3)      Caribbean Elections. (2022). Grenada General Elections 2022 results. Retrieved from https://www.caribbeanelections.com/gd/elections/gd_results_2022.asp

4)      Smith, A., & Jones, R. (2021). Digital democracy in small island states: Social media’s role in Caribbean elections. Caribbean Journal of Political Science, 12(1), 45–61.

5)      Dunn, Pierre Barnett and Company. (2023). Caribbean voter analytics: Psychographic segmentation report. Toronto, ON: DPBA Publications.

6)      Newman, N. (2023). Influencer impact on Caribbean political participation: A Grenada focus. Social Media & Democracy Review, 8(3), 23–37.

7)      Pierre, J. C. (2023). Real-time analytics and machine learning in Caribbean electoral politics. Journal of Political Data Science, 5(2), 88–102.

8)      The New Today. (2022). Kem Jones calls for transparency in Grenada’s public finance. Retrieved from https://thenewtodaygrenada.com/local-news/kem-jones-cut-them-off/

9)      Caribbean Telecommunications Union. (2023). Mobile and internet usage across Caribbean states: 2022-2023 report. Retrieved from https://ctu.int/resources/

10)   Jackson, L. & Barrett, M. (2022). Social media as a battleground: The evolution of digital campaigning in Trinidad and Tobago. Caribbean Politics Review, 15(1), 13–28.

11)   DPBA. (2023). Microtargeting and voter psychology in Grenada’s 2022 election cycle: A technical analysis. Internal white paper.

12)   United Nations Development Programme (UNDP). (2024). Youth political engagement in the Caribbean. New York: UNDP Caribbean Division.

13)   Caribbean Development Bank (CDB). (2022). Improving governance through data: Lessons from electoral innovations. Bridgetown: CDB Working Papers Series.

14)   Bennett, R. (2023). The ethics of digital persuasion in political campaigns. Journal of Digital Ethics, 10(4), 114–130.

15)   Dunn, Pierre Barnett & Company. (2023). Sentiment analysis and predictive modelling in Trinidad & Tobago 2025 elections. Internal Report Series No. 8.

16)   Electoral Commission of Trinidad and Tobago. (2025). Preliminary findings: Electoral participation and campaign conduct in 2025. Retrieved from https://www.ectt.org.tt

17)   Global Democracy Initiative. (2022). Data privacy and electoral fairness in developing democracies. Washington, DC: GDI Reports.

18)   Caribbean Elections Observer Mission. (2022). Final observation report: Grenada 2022 General Elections. St. George’s: CEOM.

19)   Pierre, J. C. (2024). Caribbean democracy and the rise of behavioural data science. Caribbean Analytics Institute Quarterly, 3(1), 56–70.

20)  WhatsApp, Facebook & TikTok Penetration in Grenada. (2023). Statista Dataset. Retrieved from https://www.statista.com/statistics/grenada-social-media-penetration/

Source- https://www.linkedin.com/pulse/data-driven-democracy-science-behind-predicting-pierre-ph-d-one8c?utm_source=share&utm_medium=member_android&utm_campaign=share_via