“Robots, not immigrants, are taking American jobs,” reads the headline from a March 21, 2017 Los Angeles times article. Two months later, a TechRepublic article ran with the headline, “Why robots won’t replace most jobs any time soon.” Which of these conflicting narratives is the concerned reader to believe? And how influential are such articles in the overall media coverage of the future of work?
The “future of work” is a complex concept that involves the increasing role of robotics and automation technology in the American economy, the mapping of worker skills to employer needs in the digital economy, and the challenges and opportunities present therein. The concept has also been referred to in industry and media as the “gig economy,” “fourth industrial revolution,” and “second machine age.”
Media Cloud was engaged by the Ford Foundation to assess how future of work narratives have developed over the past year and who is driving these narratives. The foundation was particularly interested in the prevalence and prominence of fear-driven narratives, such as “robots are coming for our jobs,” versus narratives that focus on the future of work as an opportunity to address inequities in the capitalist system. This is reflective of the foundation’s framing of the future of work within the goal of challenging inequity.
To answer these questions, we created a future of work topic in our Topic Mapper tool, using a boolean query that captured the intersection of key technology terms (robot(s), robotics, machine learning, artificial intelligence/AI, automation, algorithm) with key workforce terms (labor, jobs, employment, workplace, workforce, employees) to discover relevant content. We removed stories that included the phrase “Google employees” due to a swarm of coverage around Google employees protesting the company’s participation in projects using AI in weapons development, which hit our keywords but was tangential to our topic. We searched into US news sources for the year 2018 and used spidering to collect any content these news sources linked to that also matched our query terms. Overall, we identified 30,167 stories from 3,094 news sources.
We found that overall, coverage of this topic is more thematic rather than event-driven. The majority of the most influential stories, as measured by in-links from other media, had a strong emphasis on automation and robots and their impact on jobs.
Top articles as measured by media in-links
- “What the future of work will mean for jobs, skills, and wages,” McKinseyQuarterly.com, 11/27/17
- “Amazon scraps secret AI recruiting tool that showed bias against women,” Reuters, 10/10/18
- “Gartner Says By 2020, Artificial Intelligence Will Create More Jobs Than It Eliminates,” blogs.gartner.com, 12/13/17
- “The robot that takes your job should pay taxes, says Bill Gates,” Quartz (Qz.com), 2/17/17
- “The Fourth Industrial Revolution: what it means and how to respond,” WeForum.org, 4/5/17
- “Flippy the Burger Flipping Robot is Now Cooking at the Caliburger Fast Food Chain,” KTLA.com, 3/5/18
- “Public Predictions for the Future of Workforce Automation,” PewResearch.org, 3/10/16
- “Robots and Jobs: Evidence from US Labor Markets,” NBER.org, 3/23/17
- “Amazon wins a pair of patents for wireless wristbands that track warehouse workers,” GeekWire.com, 1/30/18
- “Short of Workers, Fast-Food Restaurants Turn to Robots,” Wall Street Journal, 6/24/18
Despite the influence of the “robot taking jobs” narrative in the media ecosystem, less than 2% of the all stories collected for the topic contained the words “robots” and “jobs” with “take,” “steal,” or “destroy” in the same sentence. It is interesting to see a narrative that represents a small percentage of the coverage accounting for the majority proportion of influential articles. This is perhaps because this narrative intersects with other social issues, such as immigration, economic inequality, and the role of the government in guaranteeing livable income streams. It is also perhaps due to that it involves technology, fear, and nebulous powers influencing human lives, which together produce a fascinating subtopic for media consumers. Overall, the coverage is quite conflicting surrounding this narrative; readers would be left confused about the level of threat and its immediacy.
Conversely, while we don’t see stories explicitly focused on the issue of inequity in the top ten influential articles, overall this narrative appears in 17% of the coverage on the topic. However, less than 2% of these articles had equity keywords in the headline, indicating that it was a mention but not a primary focus. Income and class inequality are the primary types of inequality mentioned; gender inequality and racial inequality are also included, though less frequently. Discussion and debate about universal basic income (UBI) are prominent in this coverage. Many articles argue that automation will lead to greater inequality and social and political polarization, and that UBI may be a solution.
When we explored who was making the news on this topic, we see that slightly more than half of the top 20 mentioned individuals are current or former politicians/political leaders, including international leaders, and nearly half are current or former technology CEOs. There is no clear leading advocate voice rising to the level of coverage of the political leaders and CEOs.
Top mentioned individuals (above) and organizations (below)
Similarly, when looking at the most mentioned organizations, tech companies are by far the most prominent. Government entities/agencies also make the list, along with two universities (MIT, Harvard).
We evaluated which news outlets are the most influential in this issue area by looking at number of media in-links received, number of Facebook shares on related content, and number of stories published on the topic. Five media sources ranked highly across multiple factors: The Guardian US, The New York Times, The Atlantic, CNBC, and Quartz.com.
Finally, we mapped the conversation using network analysis. The bipartite network graph below depicts the top words found in coverage and the media sources and connects sources that use the keywords. We then used a community detection algorithm and found four distinct sub-communities within the coverage ecosystem.
The community marked in blue, which we identified as the economic research community, used terms such as “labor” and “industries,” and contained sources like Bureau of Labor and Statistics and the National Bureau of Economic Research. The community marked in orange, identified as the business logistics community, used terms such as “deliver,” “solutions,” “retail,” and “users,” and contained sources like Business Wire. The community marked in green, identified as the mainstream media community, used terms like “Facebook,” “election,” “robots,” “gender,” and “computer,” and contained sources like The Washington Post, The New York Times, Politico, CNN, and The Atlantic. And the community marked in purple, identified as the tech industry and associated reporting community, used terms such as “AI,” “privacy,” “algorithms,” “Amazon,” and “platforms,” and contained sources like Wired, Geekwire, McKinsey.com, Technology Review, Fast Company, WSJ, Forbes, and the typically mainstream NBC. A map like this allows those interested in promulgating certain narratives in the future of work space to see which media outlets to target, as well as understand the different sub-conversations taking place within this topic.
This topic will likely continue to gain attention in the media, and it will be of particular interest to see if this issue and any of the narratives explored are surfaced in 2020 elections discourse. We welcome your outreach and questions about this work!