Harvard Data Science Review Podcast: Polling for 2024 U.S. Election – What Should Voters Look for and Trust?

Release Date: March 28, 2024

In the March 28, 2024 episode of the Harvard Data Science Review Podcast, hosts Liberty Vittert and Shaolin Meng engage in an in-depth discussion with two prominent experts: Kai Chen, a pollster and partner at Echelon Insights, and Scott Tranter, Head of Data Science at Decision Desk HQ. The conversation centers around the upcoming 2024 U.S. elections, exploring the reliability of polls, shifting voter demographics, and the evolving landscape of election data science.

Introduction: Setting the Stage

The episode opens with host Liberty Vittert introducing the topic and guests. With another presidential election on the horizon, concerns about the integrity of polling and voting processes are paramount. Liberty sets the tone by highlighting the importance of understanding how data science influences political outcomes.

Biden’s Declining Approval vs. Trump’s Steady Popularity

Shaolin Meng begins the discussion by reflecting on recent primary milestones, notably Nikki Haley becoming the first woman to win a Republican primary. He poses a critical question to Kai Chen about the distinctive data points that make the 2024 election cycle stand out.

Kai Chen [01:39]:
“The biggest consistent factor throughout this election cycle isn't exactly a new insight. Biden favorability has dropped by 14 points since 2020. That puts his popularity numbers lower than Trump's.”

Chen emphasizes the significant decline in President Joe Biden’s approval and favorability, noting that Biden’s favorability has decreased by 14 points since 2020, while former President Donald Trump’s favorability has remained relatively stable. This shift potentially marks the first time since 2004 that a Republican presidential candidate leads consistently in both state and national polls.

Shifting Voter Demographics: A Potential Game-Changer

The conversation shifts to voter demographics, with Kai highlighting troubling trends among traditionally Democratic-leaning groups.

Kai Chen [03:18]:
“We’re seeing very clear weaknesses when it comes to young voters, the youngest voters, approval of Joe Biden, and as well as preferences for left-leaning positions really eroding among non-white voters.”

Chen points out that declining support among young and non-white voters could significantly impact not only the presidential race but also down-ballot races. Scott Tranter adds to this by discussing historical voting patterns and recent shifts.

Scott Tranter [10:46]:
“There is some evidence here that there is some shifts among the Hispanic and African American voters. Whether that's enough to tilt the election towards the Republicans this fall is an open question.”

Tranter references Miami-Dade County’s recent shift from a Democratic stronghold to a Republican one as a potential indicator of broader demographic changes that could influence electoral outcomes.

Understanding Polling Reliability and Data Collection

The reliability of polling methods is a central theme of the discussion. Scott Tranter explains Decision Desk HQ’s role in election night data aggregation.

Scott Tranter [05:04]:
“We’re one of the three organizations in the United States that collects election results on election night, aggregates them up and then packages them and gives them to news organizations.”

Tranter details their non-partisan approach, emphasizing direct data collection from over 3,000 counties through in-person tallying, scraping county websites, and direct fax transmissions in certain states. He reassures listeners of the transparency and accuracy of their data collection methods.

On the other hand, Kai Chen discusses Echelon Insights’ strategies to maintain polling accuracy.

Kai Chen [08:49]:
“We were ranked one of the top five most accurate pollsters by FiveThirtyEight’s recent analysis. We’re building probabilistic models of the likely electorate in every state and diversifying our sampling methods to include text messaging, which we found pretty effective for response rates.”

Chen outlines Echelon’s use of probabilistic models and diversified sampling techniques to enhance the reliability of their polling data, adapting lessons learned since the 2016 election to address current challenges.

The Impact of New Technology: AI and Deepfakes

The rise of AI and deepfakes presents new challenges for election integrity. Kai Chen provides his perspective on the potential impact of these technologies.

Kai Chen [16:19]:
“The vast majority of voters aren’t going to have their decision made up based on whether or not they viewed or did not view a potential deepfake.”

Chen suggests that while deepfakes are a concern, most voters prioritize tangible issues like the economy, immigration, and public safety over manipulated media content. He believes that increased skepticism and media literacy will mitigate the influence of deepfakes.

Scott Tranter adds a different dimension, addressing the broader issue of misinformation and data accuracy.

Scott Tranter [17:39]:
“We worry about all these different election interference type things... It is very important to be very transparent and open with people about how we collect this stuff and when there’s a mistake, why there’s a mistake and why we corrected it.”

Tranter emphasizes the importance of transparency in maintaining trust in the electoral process, highlighting the challenges of ensuring data integrity across thousands of localities.

Voter Information Literacy: Navigating the Information Landscape

A significant portion of the discussion focuses on how voters can critically assess information amidst a saturated and often misleading media environment.

Scott Tranter [20:17]:
“Consume a lot and average it all out. And don’t rely completely on the outliers. Look at the data and look at the articles that are more transparent in how they talk about their polling or their modeling or their analysis.”

Tranter advises voters to engage with multiple reputable sources and to be wary of sensational headlines that confirm personal biases. He advocates for a data-driven approach to understanding election forecasts.

Kai Chen [26:03]:
“Try to look past the noise, look at reputable numbers... Don’t believe everything you see on social media.”

Chen reinforces the need for media literacy, encouraging voters to seek out original source data and avoid being swayed by partisan headlines or unverified information on social platforms.

Building Trust in the Electoral Process

Towards the end of the episode, both experts address the importance of trust in the electoral system.

Kai Chen [26:55]:
“Keep looking for reputable original source data and look past those partisan headlines.”

Scott Tranter [26:57]:
“I would just tell people before they vote, trust in the system a little bit more than they do. If you’re worried about it, you can go talk to your local election officials.”

Tranter highlights the robustness and transparency of the U.S. electoral system, urging voters to build trust through direct engagement with election processes and officials.

Conclusion: The Critical Role of Data Science

The episode concludes with Shaolin Meng acknowledging the complex interplay between data science and electoral integrity. He underscores the essential role data scientists play in interpreting election data and ensuring accurate, transparent information reaches the public.

Shaolin Meng [29:08]:
“This is a data science podcast, and I want to talk a little bit about all the technologies, particularly with the rise of AIs and all the deep fakes globally. Have these technologies significantly affected 2024 election or will they affect the 2024 elections?”

Meng appreciates the insights shared by Kai and Scott, emphasizing the importance of continued dialogue and research in the realm of data science and electoral processes.

Key Takeaways

1. Biden’s Declining Favorability: President Biden’s approval ratings have significantly dropped since 2020, making his favorability lower than Trump’s for the first time since 2004.

2. Shifting Demographics: There are notable shifts among young and non-white voters, with some traditionally Democratic-leaning groups showing weakening support.

3. Polling Accuracy: Both Echelon Insights and Decision Desk HQ employ rigorous, non-partisan methods to ensure accurate polling and data collection, adapting strategies to maintain reliability.

4. Technological Impact: While AI and deepfakes pose potential risks, most voters prioritize substantive issues over manipulated media, and increased media literacy can mitigate these challenges.

5. Voter Information Literacy: Voters are encouraged to engage with multiple reputable sources, verify original data, and avoid being influenced by partisan or sensationalist headlines.

6. Trust in the Electoral System: Building and maintaining trust through transparency and direct engagement with election processes is crucial for electoral integrity.

Notable Quotes

• Kai Chen [01:39]:
  “Biden favorability has dropped by 14 points since 2020. That puts his popularity numbers lower than Trump's.”

• Scott Tranter [05:04]:
  “We’re one of the three organizations in the United States that collects election results on election night, aggregates them up and then packages them and gives them to news organizations.”

• Kai Chen [08:49]:
  “We were ranked one of the top five most accurate pollsters by FiveThirtyEight’s recent analysis.”

• Scott Tranter [10:46]:
  “There is some evidence here that there is some shifts among the Hispanic and African American voters.”

• Kai Chen [16:19]:
  “The vast majority of voters aren’t going to have their decision made up based on whether or not they viewed or did not view a potential deepfake.”

• Scott Tranter [20:17]:
  “Consume a lot and average it all out. And don’t rely completely on the outliers.”

• Kai Chen [26:03]:
  “Try to look past the noise, look at reputable numbers... Don’t believe everything you see on social media.”

• Scott Tranter [26:57]:
  “I would just tell people before they vote, trust in the system a little bit more than they do.”

This episode of the Harvard Data Science Review Podcast provides a comprehensive analysis of the current polling landscape, voter behavior, and the critical role of data science in navigating the complexities of the 2024 U.S. elections. By leveraging expert insights and emphasizing the importance of reliable data and media literacy, the podcast equips listeners with the knowledge to make informed voting decisions amidst a rapidly evolving political and technological environment.