Karen Jeffrey

Thesis title:

The Political Economy of AI-driven Automation and Technological Change: Effects on Income Inequality and Policy Responses


In recent years, leaps in the field of artificial intelligence (Silver et al., 2016) fuelled by fierce competition between powerful players such as Google, Apple and Amazon (Pratt, 2015), have produced intelligent machines and algorithms capable of completing non-routine tasks. Coupled with the falling cost of robotics (Graetz and Michaels, 2015), these developments are creating profound potential for human labour to be automated (Benzell et al., 2015; Brynjolfsson and McAfee, 2014; Citi GPS, 2016; Ford, 2015; Pratt, 2015; Sachs et al., 2015). Indeed, one influential study has forecast that such AI-driven automation could displace as much as 47% of workers in the USA in the next two decades (Frey and Osborne, 2013).Commentators from academia, the business world and wider civil society have suggested a range of policies to mitigate any potential negative effects of AI-driven automation; however, it is not immediately clear which policy responses will most effectively share the benefits associated with AI-driven automation throughout society, or which are capable of winning popular support. Karen proposes a five-stage programme of research to explain the factors most likely to affect the level of adoption of AI-driven automation, and to compare the economic and political viability of a range of possible progressive policy responses: Stage 1: Investigating the influence of political factors during periods of technological change Stage 2: Identifying potential progressive policy responses to the increasing automation of work Stage 3: Explaining variation in the economic effects of the different policy responses in the context of AI-driven automation Stage 4: Explaining variation in the popularity of the different policy responses Stage 5: Development of policy recommendations.

First supervisor:

Konstantinos Matakos


5 – Work, Organisations & Business Management