The 2027
Oxford Essay Prize

Economists study heterogeneity in many ways, but economic models necessarily abstract from much of the extraordinary complexity of the global economy. Modern computing power and big data make it increasingly conceivable to model economic agents individually: what each person consumes and how much they work, what each firm produces and invests, and how each government taxes, spends and borrows. Such a model could serve as a digital laboratory for studying how financial crises, trade disruptions, technological change or new policies propagate through the global economy. Essays should consider how far this approach could realistically go, whether greater granularity necessarily produces better answers, and what might remain fundamentally impossible to simulate.

Policy makers make decisions that significantly affect the lives of many. Yet in making those decisions they must contend with a great deal of uncertainty: measurements contain errors, the future can be forecast but never known, and our understanding of structures and institutions is imperfect at best. How should policy makers deal with the uncertainty inherent in their decisions? Think carefully about the different considerations one could prioritise and how to manage the trade-offs between them.

From the Arab Spring to the storming of the U.S. Capitol, digital technologies have been celebrated as tools of liberation and condemned as weapons of repression. The same platforms that enable mass mobilisation and amplify marginalised voices can also spread disinformation, enable surveillance and deepen polarisation. As AI-generated content and algorithmic governance become ordinary features of political life, we must ask what they do to the demos itself. Is the net effect of our digital world to empower citizens or to erode the foundations of democratic self-rule?

Many of the problems we face today involve trade-offs between wellbeing now and wellbeing later. Growing national debt, climate change and other environmental harms, and rising existential risks from AI and other sources are just a few examples. Given the uncertainty involved, how far should our window of concern spread: a decade, a generation, a century, a millennium?

Technology is often seen as neutral, with consequences depending only on how it is used. But design involves choices about whose needs matter and what trade-offs are acceptable, from AI to medical technology to everyday infrastructure. Does this mean true neutrality is impossible, and that every invention carries the values of the society that built it?

Artificial intelligence may appear to be entirely digital, but it ultimately depends on physical infrastructure: data centres that consume electricity, generate large amounts of heat, require cooling and, in some cases, use significant quantities of water. As AI grows, these facilities may increasingly compete with homes, industries and communities for limited resources, while also producing potentially significant scientific, economic and social benefits. How should we decide whether those benefits justify their environmental and infrastructural costs? And what principles should determine where data centres are built and how efficiently they must operate?