Grant: Discovery Grants Program - Individual

If, in some population, being educated at a particular institution is positively correlated with having future wages above a certain level, is it because the degree leads to better job prospects, or is it because a third factor, such as aptitude, leads to both a greater chance of being admitted to the institution and greater success in one's career? This is the sort of question that is the purview of the field of causal inference, a general-purpose science of cause and effect, applicable in domains ranging from epidemiology to economics. Researchers in this field seek to find techniques for extracting causal conclusions from statistical data. Meanwhile, one of the most significant results in the foundations of quantum theory, Bell's theorem, can also be understood as an attempt to disentangle correlation and causation. It has recently become apparent that the insights derived from almost 60 years of intensive research on Bell's result can supplement and improve upon state-of-the-art causal inference techniques. In the other direction, the conceptual framework developed by causal inference researchers provides a fruitful new perspective on how to provide a satisfactory causal explanation of the correlations observed in Bell experiments. Specifically, it suggests that to do so, one must develop intrinsically quantum generalizations of the notions of causal influence and probabilistic inference. My research program aims to leverage the flow of techniques and insights in both directions. Drawing causal conclusions from observational data, without the need for randomized controlled trials, has a tremendous scope for applications. The approach proposed here is innovative insofar as causal discovery techniques based on inequality constraints (such as Bell inequalities) have received little attention in the causal inference community but constitute a very powerful supplement to the usual techniques. Furthermore, developing the quantum science of cause and effect is important because it provides the means for determining what multipartite states, channels, and other input-output processes are compatible with a given hypothesis about causal structure. This structure might arise as an imposed constraint, such as a shallow circuit, or the nature of interactions in a many-body system, or from relativistic constraints in a quantum gravity context. It also provides a means for proving novel quantum advantages for information processing tasks and for witnessing quantumness in experimental implementations thereof, without requiring precise characterizations of the components. Through training of students, I will lay the groundwork for a growing local research community in causal inference, with the goal of establishing Canada as a world-leader in this area. The expertise they acquire will prepare them for positions in data science and quantum information, both of which are growth areas in Canada's knowledge economy.