# Why denominators matter in policing bias | Lisa Tompson | TEDxUniversity of Waikato Source: https://www.youtube.com/watch?v=FNotdw--TcA Recap page: https://rapidrecap.app/video/FNotdw--TcA Generated: 2026-01-22T21:04:30.533+00:00 --- ## Quick Overview Lisa Tompson argues that misinterpreting or selecting the wrong denominator when calculating statistics about policing bias, such as using the general population instead of the population at risk, creates misleading narratives that exacerbate political conflict and undermine efforts for genuine reform. **Key Points:** - The most dangerous errors in public debate about policing bias stem from using incorrect mathematical denominators rather than false numbers. - When calculating police stop or search rates, using the general population as the denominator (e.g., census data) instead of the population truly at risk of being stopped yields misleading disparities. - Data showing Māori and Pacific people are stopped nearly four times as likely as white drivers is stark, but the denominator (the population at risk) drastically alters the narrative. - In the Northern Ireland Troubles, police service statistics comparing Catholic versus Protestant arrests skewed public perception because the denominator failed to account for the actual populations involved in conflict. - Researchers have mapped disparities in over 500 studies across fields like healthcare, climate change, and policing, often finding the same denominator error. - When the denominator used is the general population (e.g., census data) instead of the population actually exposed to the risk (e.g., people driving in high-risk areas), disparities appear much larger than they truly are. ![Screenshot at 00:32: Tompson on stage at TEDx University of Waikato, setting up her argument that public trust in institutions is declining due to erroneous mathematical conclusions in public debate.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-00-32.jpg) **Context:** Lisa Tompson presents this TEDx talk to expose how fundamental mathematical errors, specifically in selecting the correct denominator when calculating statistics related to policing bias, drive misleading public controversy and hinder constructive policy reform. She emphasizes that while disparities in police stops affecting Māori and Pacific people are real, the way these statistics are framed—often using the entire census population as the baseline—distorts the true level of unfair treatment and fuels political agendas. ## Detailed Analysis Lisa Tompson delivers a presentation arguing that the most damaging errors in public debate concerning policing bias are not lies, but mathematical mistakes, specifically the choice of the denominator when calculating rates. She warns that when disparities are calculated using the general population (like census data) instead of the population actually exposed to risk (the true denominator), the resulting statistics wildly exaggerate the problem, fueling political agendas and distrust in institutions. Tompson cites examples from policing in New Zealand concerning Māori and Pacific people being stopped four times more often than white drivers, and historical examples from Northern Ireland during the Troubles involving religious divisions. She notes that her team has mapped over 500 studies across various fields (healthcare, climate change) that suffer from this denominator error. The core issue is that using the total population as the denominator for police stops or searches inflates the apparent disparity, making groups seem disproportionately targeted when they may simply be overrepresented in the specific context being measured (like driving in high-risk areas). Tompson concludes that if we get the denominator wrong, we argue about the wrong problems, reinforcing harmful stereotypes instead of achieving genuine equality-focused reform. ### Introduction to Policing Bias - The most dangerous errors in public debate are caused by math, not lies - Misunderstanding denominators fuels political agendas and destroys trust in institutions. ### The Problem of Denominators - Disparities in police stops against Māori and Pacific people are stark, but using the general population (census) as the denominator exaggerates the true risk compared to using the population at risk (e.g., those driving in high-risk areas). ### Historical Context - Cites the Northern Ireland Troubles where religious data skewed perceptions, showing how denominator choice shapes the narrative (Catholics vs. Protestants). ### The Solution - Researchers must be rigorous about using the correct denominator—the population genuinely exposed to the risk—to move past misleading statistics and address the actual policy issues related to fairness and equality. ![Screenshot at 00:00: Title slide for Lisa Tompson's TEDx talk at the University of Waikato: 'The hidden maths behind Policing bias: Why denominators matter'.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-00-00.jpg) ![Screenshot at 00:32: Tompson on stage using hand gestures to emphasize her point about declining public trust in institutions due to mathematical errors in public debate.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-00-32.jpg) ![Screenshot at 00:50: Close-up of Tompson speaking passionately, illustrating how using incorrect denominators \(like general population\) stacks the odds against certain groups in policing interactions.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-00-50.jpg) ![Screenshot at 05:08: Screen displays headlines confirming data shows ongoing racial bias in police warrant-less searches, setting up the need to examine the underlying math.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-05-08.jpg) ![Screenshot at 11:36: The speaker bows her head slightly as the audience applauds following her concluding remarks.](https://ss.rapidrecap.app/screens/FNotdw--TcA/00-11-36.jpg)