From Universal to Individualized Actionability: Revisiting Personalization in Algorithmic Recourse

Lena Marie Budde; Ayan Majumdar; Richard Uth; Markus Langer; Isabel Valera

cs.LGApr 9, 2026

From Universal to Individualized Actionability: Revisiting Personalization in Algorithmic Recourse

Authors: Lena Marie Budde, Ayan Majumdar, Richard Uth, Markus Langer, Isabel Valera

Organizations: Saarland Informatics Campus, Saarland University, Saarbrücken, Germany · Max Planck Institute for Software Systems, Saarbrücken, Germany · Institute for Psychology, University of Freiburg, Freiburg, Germany

Abstract

Algorithmic recourse aims to provide actionable recommendations that enable individuals to change unfavorable model outcomes, and prior work has extensively studied properties such as efficiency, robustness, and fairness. However, the role of personalization in recourse remains largely implicit and underexplored. While existing approaches incorporate elements of personalization through user interactions, they typically lack an explicit definition of personalization and do not systematically analyze its downstream effects on other recourse desiderata. In this paper, we formalize personalization as individual actionability, characterized along two dimensions: hard constraints that specify which features are individually actionable, and soft, individualized constraints that capture preferences over action values and costs. We operationalize these dimensions within the causal algorithmic recourse framework, adopting a pre-hoc user-prompting approach in which individuals express preferences via rankings or scores prior to the generation of any recourse recommendation. Through extensive empirical evaluation, we investigate how personalization interacts with key recourse desiderata, including validity, cost, and plausibility. Our results highlight important trade-offs: individual actionability constraints, particularly hard ones, can substantially degrade the plausibility and validity of recourse recommendations across amortized and non-amortized approaches. Notably, we also find that incorporating individual actionability can reveal disparities in the cost and plausibility of recourse actions across socio-demographic groups. These findings underscore the need for principled definitions, careful operationalization, and rigorous evaluation of personalization in algorithmic recourse.

Figures & tables

Figure 1: iCARMA architecture (adapted from [ 34 ] ); red parts indicate modifications to enable individualized recommendations.

Figure 2: Actionability weights and cost profiles (

k=4

w^{\max}=7

dataset, no. of individually actionable features	method	validity	plausibility	cost
Loan, all (4)	oracle	0.959	1.0	$0.057_{\pm 0.064}$
Loan, all (4)	iCARMA	0.999	0.886	$0.069_{\pm 0.077}$
Loan, 3	oracle	0.796	1.0	$0.07_{\pm 0.085}$
Loan, 3	iCARMA	0.995	0.689	$0.073_{\pm 0.084}$
Loan, 2	oracle	0.539	1.0	$0.078_{\pm 0.096}$
Loan, 2	iCARMA	0.991	0.481	$0.09_{\pm 0.111}$

Table 1: Impact of hard actionability constraints on recourse metrics for amortized iCARMA and non-amortized Oracle.

Figure 3: Distribution of the log densities for the factual and counterfactual (cf) under different numbers of actionable features (indicated in brackets).

dataset+constraints	individual cost profile	method	validity	plausibility	cost	hap
Loan, soft	constant	oracle	0.959	1.0	$0.057_{\pm 0.064}$	0.614
	constant	iCARMA	0.999	0.886	$0.069_{\pm 0.077}$	0.852
	concave	oracle	0.959	1.0	$0.074_{\pm 0.088}$	0.492
	concave	iCARMA	0.998	0.631	$0.079_{\pm 0.088}$	0.044
	linear	oracle	0.959	1.0	$0.072_{\pm 0.085}$	0.443
	linear	iCARMA	0.996	0.532	$0.077_{\pm 0.086}$	0.029

Table 2: Impact of preference consideration as rankings on recourse metrics across four cost profiles: concave, linear, convex, and constant (no cost individualization) for amortized iCARMA and non-amortized causal Oracle.

constraint	individual cost profile	validity	plausibility	cost	hap
soft	constant	0.996	0.841	$0.067_{\pm 0.07}$	0.829
	concave	0.995	0.713	$0.073_{\pm 0.078}$	0.225
	linear	0.991	0.646	$0.074_{\pm 0.086}$	0.178
	convex	0.982	0.5	$0.081_{\pm 0.082}$	0.194
hard (random) + soft	constant	0.991	0.712	$0.073_{\pm 0.083}$	0.838
	concave	0.99	0.618	$0.076_{\pm 0.087}$	0.415

Table 3: Impact of preference consideration as scores on recourse metrics across cost profiles for iCARMA on the Loan Dataset.

Figure 4: Cost (left) and distributional log-density (right) results of recourse recommendations across intersectional gender and race groups for four scenarios: (1) non-personalized: all globally actionable features are equally actionable, (2) random-assigned: privileged preference profiles assigned randomly, (3) race-correlated: 90% of white, 5% of non-white individuals assigned the privileged profile, (4) income-dependent: top-earning 25% of individuals assigned the privileged profile.

Appendix figures & tables7 assets

Supplementary material from the paper’s appendix.

Appendix

$pref_{u,i}=1$	$pref_{u,i}=2$	$pref_{u,i}=3$	$pref_{u,i}=4$	uniform	co	rco	privileged pref profile	non-privileged pref profile
LA	Dur	Inc	Sav	$1/24$	$1/24$	0	0	$1/6$
LA	Dur	Sav	Inc	$1/24$	$1/24$	0	0	$1/6$
LA	Inc	Dur	Sav	$1/24$	$1/6$	0	0	$1/24$
LA	Sav	Dur	Inc	$1/24$	$1/24$	0	0	$1/24$
LA	Inc	Sav	Dur	$1/24$	$1/6$	0	0	$1/24$
LA	Sav	Inc	Dur	$1/24$	$1/24$	0	0	$1/24$

Appendix

Table 4: Different distributions used for preference sampling on the Loan dataset. Each row reports the probability of sampling a particular assignment of preference scores

pref_{u,i}

to the globally actionable features LA, Dur, Inc, and Sav under a given configuration. The distribution biased according to causal order is denoted by co, while the distribution biased according to reverse causal order is denoted by rco. The distributions used to sample the privileged and non-privileged preference profiles for the fairness case study are reported in the last two columns.

Figure 5: Causal Graph of Loan dataset

Figure 6: Causal graph for Give Me Some Credit. Solid gray nodes indicate globally immutable, light gray shaded nodes indicate mutable but non-actionable, and white nodes indicate globally actionable nodes.

Figure 7: Feature reconstruction for Give Me Some Credit by the best performing CNF model.

bias	number of indivdually actionable features	method	validity	plausibility	cost
causal order aligned	all (4)	oracle	0.959	1.0	$0.057_{\text{$ \pm 0.064 $}}$
	all (4)	iCARMA	0.998	0.864	$0.065_{\text{$ \pm 0.075 $}}$
	3	oracle	0.821	1.0	$0.066_{\text{$ \pm 0.078 $}}$
	3	iCARMA	0.997	0.713	$0.062_{\text{$ \pm 0.073 $}}$
	2	oracle	0.597	1.0	$0.072_{\text{$ \pm 0.088 $}}$
	2	iCARMA	0.997	0.564	$0.066_{\text{$ \pm 0.09 $}}$

Appendix

Table 5: Impact on recourse metrics for an increasing number of individually actionable features for the hard-constrained case, assuming non-uniform sampling of individually actionable features : we consider a distribution favouring features high in causal order and a distribution favouring features low in causal order (reverse causal order, rco). For the second case, we additionally compare iCARMA trained with individually actionable features sampled uniformly at random with iCARMA trained with individually actionable features sampled from the distribution biased towards features low in causal order. Oracle knows true SCM and uses grid search with hard constraint for plausibility.

distribution	individually actionable	plausibility constraint	validity	plausibility	cost
uniform random	1	No	0.876	0.25	$0.135_{\text{$ \pm 0.151 $}}$
	1	Yes	0.22	1.0	$0.077_{\text{$ \pm 0.096 $}}$
	2	No	0.994	0.25	$0.096_{\text{$ \pm 0.133 $}}$
	2	Yes	0.539	1.0	$0.078_{\text{$ \pm 0.096 $}}$
causal order	1	No	0.923	0.28	$0.119_{\text{$ \pm 0.139 $}}$
	1	Yes	0.259	1.0	$0.07_{\text{$ \pm 0.087 $}}$

Appendix

Table 6: Ablation of the plausibility constraint in causal-oracle recourse optimization. Removing the plausibility constraint increases validity but produces less distributionally plausible counterfactuals. The effects of individual actionability preferences on other recourse criteria persist even without plausibility constraints.

Figure 8: Log-density distribution (iCARMA) for Give Me Some Credit and different personalization setups.

Algorithmic Recourse

Algorithmic recourse studies actionable changes that can help people obtain favorable outcomes from automated decisions.

See all 26 papers

Algorithmic Recourse Under Competition
Sep 30, 2026Shahin JabbariAlgorithmic Recourse
Drexel University
Diverse and Plausible Algorithmic Recourse via Tractable Recourse Distributions
Aug 5, 2026Anagha Sabu, Hrithik Suresh, Narayanan C. KrishnanProbabilistic Circuits Algorithmic Recourse
Mehta Family School of Data Science and Artificial Intelligence Indian Institute of Technology Palakkad Palakkad, Kerala, India
Personalized Causal Recourse: A Human-In-The-Loop Approach
Jul 3, 2026Denise Tampieri, Giovanni De Toni, Paolo GiudiciStructural Causal Models Causal Structure Learning
The University of Edinburgh, Edinburgh, UK · ⋆Work conducted during the author’s MSc at the University of Trento, Trento, Italy. · Fondazione Bruno Kessler, Trento, Italy +1

Sep 30, 2026J/K move · Enter open · S save

Algorithmic Recourse Under Competition

Shahin Jabbari

Drexel University

Algorithmic Recourse

Open paper AskPDF arXiv

Algorithmic recourse provides individuals who have received undesirable outcomes from machine learning models with suggestions for minimum-cost improvements to achieve the desired outcome. A central assumption when computing recourse is that the decision rule remains fixed throughout the recourse implementation phase. We challenge this assumption in settings where individuals compete for limited resources. In such settings, widespread recourse implementation can change the acceptance threshold even when the scoring model that is used to evaluate individuals remains the same. This change in acceptance threshold can, in turn, invalidate the original recourse recommendations (i.e., following the recourse may not lead to the desired outcome). To address this problem, we introduce a framework called recourse under competition that jointly optimizes for recommendation recipients and the recommended score target they need to satisfy to balance the recourse cost and post-shift validity among initially rejected individuals. We develop an algorithm based on the Implicit Function Theorem and empirically analyze its performance. Experiments on synthetic and real datasets show that personalized score targets can achieve higher validity, albeit at a higher cost. In contrast, common score targets generally offer favorable cost-validity trade-offs for lower to medium validity values.

Abstract

Algorithmic Recourse

Explore similar work

Algorithmic Recourse Under Competition

Diverse and Plausible Algorithmic Recourse via Tractable Recourse Distributions

Personalized Causal Recourse: A Human-In-The-Loop Approach