Fair Inference on Outcomes

By Razieh Nabi, Ilya Shpitser. 2017

389 citations recorded by the source corpus.

In this paper, we consider the problem of fair statistical inference involving outcome variables. Examples include classification and regression problems, and estimating treatment effects in randomized trials or observational data. The issue of fairness arises in such problems where some covariates or treatments are “sensitive,” in the sense of having potential of creating discrimination. In this paper, we argue that the presence of discrimination can be formalized in a sensible way as the presence of an effect of a sensitive covariate on the outcome along certain causal pathways, a view which generalizes (Pearl 2009). A fair outcome model can then be learned by solving a constrained optimization problem. We discuss a number of complications that arise in classical statistical inference due to this view and provide workarounds based on recent work in causal and semi-parametric inference.

Subfields: Sociology and Political Science, Statistics and Probability, AI Ethics, Fairness & Society

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Sociology and Political Science

In Statistics and Probability

In AI Ethics, Fairness & Society

Citations in the corpus

A Survey on Bias and Fairness in Machine Learning (2019, methodology): “dual-level discrimination and solving each using their previous methods, in addition to targeting direct and indirect discrimination. By expanding on the previous work and generalizing it, authors in [95] propose a similar pathway approach for fair inference using causal graphs; this would restrict ”
Prediction-Based Decisions and Fairness: A Catalogue of Choices, Assumptions, and Definitions (2018, background): “No Unfair Path-Specific Effects specifies no average effects along unfair (user-specified) directed paths from A to D [91].”
On formalizing fairness in prediction with machine learning (2017, neutral): “These measures model fairness through tools from causal interference and can be considered as a sub-class of causal measures (39; 31).”
Causal Modeling-Based Discrimination Discovery and Removal: Criteria, Bounds, and Algorithms (2019, background): “recent work [25] proposes three principled approaches: (1)”
Learning Optimal Fair Policies (2018, methodology): “We have extended a formalization of fair inference from (Nabi and Shpitser 2018) to the setting of learning optimal policies under fairness constraints.”
Causal Reasoning for Algorithmic Fairness (2018, methodology): “…from A to Y for a particular set of observed variables V , by taking expectations under Y we have that a fair ˆ Y can be simply deﬁned as Like in [28], it is however not obvious which optimally is being achieved for the prediction of Y as it is unclear how such a transformation would translate…”
Delayed Impact of Fair Machine Learning (2018, neutral): “This can be seen as relaxing the requirements compared with recent work on avoiding discrimination through causal reasoning that often required stronger assumptions [Kusner et al., 2017, Nabi and Shpitser, 2017, Kilbertus et al., 2017].”
Path-Specific Counterfactual Fairness (2018, background): “Nabi & Shpitser (2018) justify averaging over M and L due to the need to account for the constraints that are potentially imposed on θ ma and θ , and suggest that this is an inherent limitation of path-speciﬁc fairness.”
Avoiding Discrimination through Causal Reasoning (2017, background): “A true individual-level comparison of the form “What would have happened to me, if I had always belonged to another group” is captured by counterfactuals and discussed in [15, 19].”
On the Fairness of Causal Algorithmic Recourse (2020, methodology): “2017a,b), these are sometimes mutually incompatible (Chouldechova 2017), and it has been argued that discrimination, at its heart, corresponds to a (direct or indirect) causal influence of a protected attribute on the prediction, thus making fairness a fundamentally causal problem (Kilbertus et al. ”
Learning Fair Rule Lists (2019, background): “In a nutshell, three families of fairness notions have emerged in recent years: causal notions of fairness [36, 40, 49], which rely on causal assumptions to estimate the effects of sensitive attributes and build algorithms that ensure a tolerable level of discrimination with respect to these…”
Decolonial AI: Decolonial Theory as Sociotechnical Foresight in Artificial Intelligence (2020, background): “…whose recent efforts, in response, have proposed fairness metrics that attempt to use causality (Chiappa and Isaac 2019; Mitchell et al. 2018; Nabi and Shpitser 2018; Madras et al. 2019) or interactivity (Canetti et al. 2019; Jung et al. 2019) to integrate more contextual awareness of human…”
Algorithmic Fairness (2020, background): “[93] suggest categorizing causal fairness methods according to three dimensions: individual vs. group level [84] causal effects; explicit vs. implicit [105] structural equations; and creating fair predictors vs. explaining and quantifying discrimination [146].”
A survey of algorithmic recourse: definitions, formulations, solutions, and prospects (2020, background): “…above are prevalent not only in finance [20, 133] but also in justice (e.g., pretrial bail) [10], healthcare (e.g., prescription of life-altering medication) [24, 27, 72], and other settings (e.g., hiring) [43, 134, 160] broadly classified as consequential decision-making settings [20, 34, 45, 91].”
Fairness Violations and Mitigation under Covariate Shift (2019, background): “For instance, [44] require that causal effects of”