Fairness Criterion Research Articles

The Price of Fairness of Scheduling a Scarce Resource

Balancing Fairness and Efficiency: The Price of Fairness in Scheduling Scarce Resources “The Price of Fairness of Scheduling a Scarce Resource” investigates the trade-offs between fairness and utility when scheduling scarce resources among multiple stakeholders. The research focuses on envy-freeness, a widely recognized fairness criterion, and explores how fairness constraints impact overall utility, quantified by the price of fairness—the ratio of attainable utility with and without fairness considerations. By analyzing various factors, including the number of agents, time horizon, discount factors, and switching costs, the authors derive tight bounds on the price of fairness. Interestingly, they discover that, whereas fairness often increases utility loss, certain parameters—such as reducing the time horizon or increasing the number of agents—can counterintuitively reduce this loss. These findings offer strategic insights for decision makers facing fairness constraints in scheduling, helping them navigate complex trade-offs between equitable resource allocation and maximizing efficiency.

Absolute Fairness and Weighted Lotteries

Abstract Weighted lottery proposals give guidance in rescue dilemma situations by balancing the demands of comparative and absolute fairness. While they do not advocate for saving the greater number outright, they are responsive to absolute fairness insofar as they show a certain sensitivity to the numbers involved. In this paper we investigate what criterion of absolute fairness we should demand weighted lotteries to fulfill. We do so by way of critically examining what is probably the most sophisticated weighted lottery on the market: Gerard Vong's Exclusive Composition-Sensitive (EXCS) lottery. We find that both the standard that seems most common in the debate, and a different standard Vong uses to criticize Jens Timmermann's Individualist Lottery are in contradiction to basic demands placed upon weighted lotteries and are therefore unsuitable as necessary conditions for absolute fairness. We instead propose a purely gradual understanding of absolute fairness.

STUDENTS’ ABILITY IN WRITING DESCRIPTIVE TEXT AT THE ELEVENTH GRADE OF SMA NEGERI 1 TELUK DALAM

This study dealt with how students' ability in writing descriptive text at the eleventh grade of SMA Negeri 1 Teluk Dalam. The purpose of this study was to describe students' ability in writing descriptive text. This research used qualitative research. The data were taken from 26 descriptive text that made by students. The data source was eleventh students at SMA Negeri 1 Teluk Dalam Academic Year 2023/2024. The researcher used documentation technique to collect the students’ writing of descriptive text. The finding shows that students’ abilities in writing descriptive text were based on five assessment components in writing. First, the content component with an average score 20.1 or including fair criteria, second, the organization component with an average score 15.8 or including good criteria, third, the vocabulary component with an average score 13.3 or including fair criteria, fourth, the language use component with an average score 15.6 or including fair criteria, and lastly, mechanics component with an average score 4.3 or including good criteria. From the result of the data, only 2 components were classified into good criteria and there are 3 components under the good criteria. This is proven based on the students’ level of difficulty in writing descriptive text. Lastly, the result of data analysis of students’ ability in writing descriptive text at the eleventh grade of SMA Negeri 1 Teluk Dalam was categorized in fair.

What's Distributive Justice Got to Do with It? Rethinking Algorithmic Fairness from a Perspective of Approximate Justice

In the field of algorithmic fairness, many fairness criteria have been proposed. Oftentimes, their proposal is only accompanied by a loose link to ideas from moral philosophy -- which makes it difficult to understand when the proposed criteria should be used to evaluate the fairness of a decision-making system. More recently, researchers have thus retroactively tried to tie existing fairness criteria to philosophical concepts. Group fairness criteria have typically been linked to egalitarianism, a theory of distributive justice. This makes it tempting to believe that fairness criteria mathematically represent ideals of distributive justice and this is indeed how they are typically portrayed. In this paper, we will discuss why the current approach of linking algorithmic fairness and distributive justice is too simplistic and, hence, insufficient. We argue that in the context of imperfect decision-making systems -- which is what we deal with in algorithmic fairness -- we should not only care about what the ideal distribution of benefits/harms among individuals would look like but also about how deviations from said ideal are distributed. Our claim is that algorithmic fairness is concerned with unfairness in these deviations. This requires us to rethink the way in which we, as algorithmic fairness researchers, view distributive justice and use fairness criteria.

Tipping the balance between fairness and efficiency through temporoparietal stimulation

Maximizing the welfare of society requires distributing goods between groups of people with different preferences. Such decisions are difficult because different moral principles impose irreconcilable solutions. For example, utilitarian efficiency (maximize overall outcome across individuals) may need trade-off against Rawlsian fairness norms (maximize the outcome for the worst-off individual). We identify a brain mechanism enabling decision-makers to solve such trade-offs between efficiency and fairness using separate neuroimaging and sham-controlled brain stimulation experiments. As activity in the temporoparietal junction (TPJ) increases, people are more likely to implement the Rawlsian fairness criterion rather than efficiency or inequality concerns. Strikingly, reducing TPJ excitability with brain stimulation reduces the concern for fairness in fairness-efficiency trade-offs. Moreover, the reduced fairness concerns statistically relate to stimulation-induced reductions in perspective-taking skills as measured in a separate task. Together, our findings not only reveal the neural underpinning of efficiency-fairness trade-offs but also recast the role of TPJ in social decision-making by showing that its perspective-taking function serves to promote fairness for the worst-off rather than efficiency or equality.

Multiple fairness criteria in decision tree learning

The use of algorithmic decision-making systems based on machine learning models has led to a need for fair (unbiased) and explainable classification outcomes. In particular, machine learning algorithms can encode biases, which might result in discriminatory decisions for certain groups such as gender, race, or age. Although a number of works on decision tree learning have been proposed to decrease the chance of discrimination, they usually focus on the use of a single fairness metric. In general, creating a model based on a single fairness metric is not a sufficient way to mitigate discrimination since bias can originate from various sources—e.g., the data itself or the optimization process. In this paper, we propose a novel decision tree learning process that utilizes multiple fairness metrics to address both group and individual discrimination. This is achieved by extending the attribute selection procedure to consider not only information gain but also gain in fairness. Computational experiments on fourteen different datasets with various sensitive features demonstrate that the proposed Fair-C4.5 models improve fairness without a loss in predictive accuracy when compared to the well-known C4.5 and the fairness-aware FFTree algorithms.

New Possibilities for Fair Algorithms

We introduce a fairness criterion that we call Spanning. Spanning i) is implied by Calibration, ii) retains interesting properties of Calibration that some other ways of relaxing that criterion do not, and iii) unlike Calibration and other prominent ways of weakening it, is consistent with Equalized Odds outside of trivial cases.

From efficiency to equity: A multi-user paradigm in mobile route optimization

This study addresses the challenge of optimizing vehicle mobility in urban environments, which is significant for the advancement of smart city initiatives and spatial data analysis. We introduce a novel mobile recommendation system designed for multi-user scenarios, aiming to achieve a balance between effectiveness and fairness. The system prioritizes maximizing the profitability of vehicle service providers while ensuring an equitable distribution of recommended routes among users. Our approach features a redefined objective function that integrates a fairness criterion alongside path quality optimization. We further propose PSA-DLMA (Parallel Simulated Annealing with Deep Learning-Guided Move Adaptation), a stochastic path search method that leverages deep learning to guide move and strategy selection, alongside a dynamic termination mechanism and a parallel processing strategy. We validate our methodology using recent yellow taxi data from New York City and its surroundings, conducting comprehensive experiments to assess the performance of the system. The results demonstrate the superiority of PSA-DLMA over existing state-of-the-art solutions, offering significant contributions to improving urban vehicle mobility within the smart city framework.

Fair and optimal prediction via post‐processing

AbstractWith the development of machine learning algorithms and the increasing computational resources available, artificial intelligence has achieved great success in many application domains. However, the success of machine learning has also raised concerns about the fairness of the learned models. For instance, the learned models can perpetuate and even exacerbate the potential bias and discrimination in the training data. This issue has become a major obstacle to the deployment of machine learning systems in high‐stakes domains, for example, criminal judgment, medical testing, online advertising, hiring process, and so forth. To mitigate the potential bias exhibited by machine learning models, fairness criteria can be integrated into the training process to ensure fair treatment across all demographics, but it often comes at the expense of model performance. Understanding such tradeoffs, therefore, is crucial to the design of optimal and fair algorithms. My research focuses on characterizing the inherent tradeoff between fairness and accuracy in machine learning, and developing algorithms that can achieve both fairness and optimality. In this article, I will discuss our recent work on designing post‐processing algorithms for fair classification, which can be applied to a wide range of fairness criteria, including statistical parity, equal opportunity, and equalized odds, under both attribute‐aware and attribute‐blind settings, and is particularly suited to large‐scale foundation models where retraining is expensive or even infeasible. I will also discuss the connections between our work and other related research on trustworthy machine learning, including the connections between algorithmic fairness and differential privacy as well as adversarial robustness.

An Open Information Model-Based Repository for Sustainable Re-Use of Heterogeneous Pandemics Research Data.

The COVID-19 Research Network Lower Saxony (COFONI) is a German state network of experts in Coronavirus research and development of strategies for future pandemics. One of the pillars of the COFONI technology platform is its established research data repository (Available at https://forschungsdb.cofoni.de/), which enables provision of pseudonymised data and cross-location data retrieval for heterogeneous datasets. The platform consistently uses open standards (openEHR) and open source components (EHRbase) for its data repository, taking into account the FAIR criteria. Available data include both clinical and socio-demographic patient information. A comprehensive AQL query builder interface and an integrated research request process enable new research approaches, rapid cohort assembly and customized data export for researchers from participating institutions. Our flexible and scalable platform approach can be regarded as a blueprint. It contributes, to pandemic preparedness by providing easily accessible cross-location research data in a fully standardised and open representation.

Incentive-Aware Recommender Systems in Two-Sided Markets

Online platforms in the Internet Economy commonly incorporate recommender systems that recommend products (or “arms”) to users (or “agents”). A key challenge in this domain arises from myopic agents who are naturally incentivized to exploit by choosing the optimal arm based on current information, rather than exploring various alternatives to gather information that benefits the collective. We propose a new recommender system that aligns with agents’ incentives while achieving asymptotically optimal performance, as measured by regret in repeated interactions. Our framework models this incentive-aware system as a multi-agent bandit problem in two-sided markets, where the interactions of agents and arms are facilitated by recommender systems on online platforms. This model incorporates incentive constraints induced by agents’ opportunity costs. In scenarios where opportunity costs are known to the platform, we show the existence of an incentive-compatible recommendation algorithm. This algorithm pools recommendations between a genuinely good arm and an unknown arm using a randomized and adaptive strategy. Moreover, when these opportunity costs are unknown, we introduce an algorithm that randomly pools recommendations across all arms, utilizing the cumulative loss from each arm as feedback for strategic exploration. We demonstrate that both algorithms satisfy an ex-post fairness criterion, which protects agents from over-exploitation. All code for using the proposed algorithms and reproducing results is made available on GitHub.

Distributive Justice Analysis in the Context of Workers' Leave Time Regulation: A Legal Perspective

The controversy surrounding Law No. 6/2023 revolves around the loss of the two-day weekend holiday entitlement for workers, the provision of only one day off in a six-day workweek, changes in working hours, which also affect rest time to only one day per week for a 6-day work schedule. The existence of a law should inherently provide fairness to society and contribute to the happiness of the people, but on the other hand, the government must also strive consistently to foster the economic growth of the population as a means to achieve social welfare. This research aims to investigate the regulation of workers' rest time under Law No. 6/2023 and to assess whether the provisions for rest time under Law No. 6/2023 have met the criteria of fairness for workers. To address these objectives, a normative juridical method with legislative and conceptual approaches is employed. Based on distributive justice theory, Law No. 13/2003 better reflects justice in labor relations in Indonesia as it considers basic needs, equality, and workers' contributions more comprehensively. Conversely, Law No. 6/2023, despite providing benefits to employers and potentially stimulating economic growth, appears to reduce the protection and welfare of workers, which can be deemed less fair according to distributive justice principles. Therefore, strict monitoring and improvements are necessary in the implementation of Law No. 6/2023 to ensure fairness for workers is maintained.

The system of horizontal fiscal equalization in the context of ensuring the implementation of the principles of unity and fairness of the budget system of Ukraine

The article is dedicated to defining the peculiarities of the system of horizontal fiscal equalization and substantiating the proposal for its improvement in terms of ensuring the implementation of the principles of equality and fairness of the budget system of Ukraine. The tax capacity index and reverse and basic grants were calculated for 1439 municipalities of Ukraine for 2022, taking into account the possible inclusion of the Kyiv city budget in the horizontal fiscal equalization system. This index is also computed for 2023 based on the population of territorial communities as of the beginning of 2022. It has been established that the exclusion from the equalization system of municipalities with the highest tax potential artificially increases the tax capacity of all other municipalities, thereby reducing the amount of the basic grant for grant recipients, and reduces their ability to fulfill responsibilities without attracting debt financing, and exacerbates the already significant development disparities between the capital and the regions. The involvement of the Kyiv city budget in the system of equalizing horizontal fiscal imbalances will satisfy the criteria of fairness and completeness in accordance with the principles of the budget system of Ukraine, significantly reduce the burden on the donors of the reverse grant and at the same time create a minimal (3% of the total need for basic grants) additional burden on the state budget. The comparison of the scheduled data for 2023 and the estimated data has shown that the amount of the estimated reverse grant is significantly higher than the scheduled amount of the grant for certain municipality, in particular for Kharkiv UTC (2.1 times) and Zaporizhzhia UTC (2.5 times), which suggests that the Ministry of Finance did not comply with the approved calculation methodology when planning interbudgetary transfers for 2023. The reasons for significant differences in the annual growth rates of PIT revenues (from 14.8% to 45%) to the budgets of different municipalities require additional research, as this makes the tax potential of a municipalities disproportionate to its grant need or grant potential. It is suggested to improve the system of horizontal equalization of fiscal imbalances in Ukraine by implementing EU legal norms into Ukrainian legislation, as a result of which personal income tax should be transferred to the local budget at the place of residence of the taxpayer, not at the place of registration of the tax agent.

Development of a Higher-Order Thinking Skills (HOTS) Test Instrument on Electrochemical Material for High School Students

This research aims to develop a HOTS test instrument on electrochemistry topics that meets valid and reliable criteria. Apart from that, the level of difficulty, distinguishing power, ability level, and detection of respondent bias were also analyzed. This research is development research using the Tessmer model, which includes preliminary and formative evaluation stages. The HOTS questions consist of 15 multiple-choice questions. The validation data from experts is measured using a Likert scale. Coefficient Aikens’ V is then used to measure the validation data from reviewers. Experts and reviewer assessments indicated that the HOTS question instrument developed was very good. Analysis of questions using the Rasch model in a field testing of 33 high school students using a purposive sampling technique resulted in 14 valid questions and 1 invalid question with a reliability value of 0,69. There were 3 questions in the easy category, 7 in the moderate category, 1 in the hard category, and 4 in the extremely hard category. The analysis of the distinguishing power of the questions obtained 9 questions with very good criteria, 3 questions with good criteria, and 2 with fair criteria. The results of the validity of the students` responses showed that there were 4 invalid responses. The ability level shows that 5 students have high ability, 7 have moderate ability, and 21 have low ability. In conclusion, the developed HOTS test instrument for electrochemical material has a good quality and is suitable for evaluation.

What is fair? Proxy discrimination vs. demographic disparities in insurance pricing

Discrimination and fairness are major concerns in algorithmic models. This is particularly true in insurance, where protected policyholder attributes are not allowed to be used for insurance pricing. Simply disregarding protected policyholder attributes is not an appropriate solution as this still allows for the possibility of inferring protected attributes from non-protected covariates, leading to the phenomenon of proxy discrimination. Although proxy discrimination is qualitatively different from the group fairness concepts discussed in the machine learning and actuarial literature, group fairness criteria have been proposed to control the impact of protected attributes on the calculation of insurance prices. The purpose of this paper is to discuss the relationship between direct and proxy discrimination in insurance and the most popular group fairness axioms. We provide a technical definition of proxy discrimination and derive incompatibility results, showing that avoiding proxy discrimination does not imply satisfying group fairness and vice versa. This shows that the two concepts are materially different. Furthermore, we discuss input data pre-processing and model post-processing methods that achieve group fairness in the sense of demographic parity. As these methods induce transformations that explicitly depend on policyholders' protected attributes, it becomes ambiguous whether direct and proxy discrimination is, in fact, avoided.

Best of Both Worlds: Agents with Entitlements

Fair division of indivisible goods is a central challenge in artificial intelligence. For many prominent fairness criteria including envy-freeness (EF) or proportionality (PROP), no allocations satisfying these criteria might exist. Two popular remedies to this problem are randomization or relaxation of fairness concepts. A timely research direction is to combine the advantages of both, commonly referred to as Best of Both Worlds (BoBW). We consider fair division with entitlements, which allows to adjust notions of fairness to heterogeneous priorities among agents. This is an important generalization to standard fair division models and is not well-understood in terms of BoBW results. Our main result is a lottery for additive valuations and different entitlements that is ex-ante weighted envy-free (WEF), as well as ex-post weighted proportional up to one good (WPROP1) and weighted transfer envy-free up to one good (WEF(1, 1)). We show that this result is tight – ex-ante WEF is incompatible with any stronger ex-post WEF relaxation. In addition, we extend BoBW results on group fairness to entitlements and explore generalizations of our results to instances with more expressive valuation functions.

Reinforcement Learning in Personalized Medicine: Tailoring Treatment Protocols to Individual Patients

Personalized medicine goal is to provide treatment plans that are unique to each patient; however, barriers like fairness, interpretability, and heterogeneity hinder the development of the approach. To overcome these drawbacks, this work proposes a fairness-aware hybrid RL model to enhance the clinical decision-making process. The proposed framework integrates the model-based and model-free RL to achieve short-term patient stabilization and long-term patient recovery. With the help of the fairness-aware reward adjustments, the model minimizes demographic gaps, obtaining a low bias index of 0.05 while keeping an action accuracy of 88.5%. To achieve this, SHAP (SHapley Additive exPlanations) is used to make the model more interpretable and explain to the clinicians what features are important in the treatment process like heart rate and lactate levels to build trust. The framework generates a higher cumulative reward than the standalone RL models (+1500 vs. +1350 and +1200 for model-free and model-based RL, respectively) and can be fine-tuned for different patient populations. The results confirm the model’s ability to prioritize fair and explainable healthcare solutions when tested with the MIMIC-III dataset. This work closes the gap between theory and practice, and opens the path towards large-scale, responsible, and learning from data AI-Powered Precision Medicine. The future work is to integrate with clinical processes in real-time and broaden the fairness criteria to account for multiple forms of discrimination.

Monotonicity anomalies in Scottish local government elections

The single transferable vote (STV) voting method is used to elect multiple candidates in ranked-choice elections. One weakness of STV is that it fails multiple fairness criteria related to monotonicity and no-show paradoxes. We analyze 1079 local government STV elections in Scotland to estimate the frequency of such monotonicity anomalies in real-world elections, and compare our results with prior empirical and theoretical research about the rates at which such anomalies occur. In 62 of the 1079 elections we found some kind of monotonicity anomaly. We generally find that anomaly rates are similar to prior empirical research and much lower than what most theoretical research has found. Most STV anomalies we find are the first of their kind to be documented in real-world multiwinner elections.

The Complexity of Finding Fair Many-to-One Matchings

We analyze the (parameterized) computational complexity of “fair” variants of bipartite many-to-one matching, where each vertex from the “left” side is matched to exactly one vertex and each vertex from the “right” side may be matched to multiple vertices. We want to find a “fair” matching, in which each vertex from the right side is matched to a “fair” set of vertices. Assuming that each vertex from the left side has one color modeling its “attribute”, we study two fairness criteria. For instance, in one of them, we deem a vertex set fair if for any two colors, the difference between the numbers of their occurrences does not exceed a given threshold. Fairness is, for instance, relevant when finding many-to-one matchings between students and colleges, voters and constituencies, and applicants and firms. Here colors may model sociodemographic attributes, party memberships, and qualifications, respectively. We show that finding a fair many-to-one matching is NP-hard even for three colors and maximum degree five. Our main contribution is the design of fixed-parameter tractable algorithms with respect to the number of vertices on the right side. Our algorithms make use of a variety of techniques including color coding. At the core lie integer linear programs encoding Hall like conditions. We establish the correctness of our integer programs, based on Frank’s separation theorem [Frank, Discrete Math. 1982]. We further obtain complete complexity dichotomies regarding the number of colors and the maximum degree of each side.

Almost proportional allocations of indivisible chores: Computation, approximation and efficiency

Proportionality (PROP) is one of the simplest and most intuitive fairness criteria used for allocating items among agents with additive utilities. However, when the items are indivisible, ensuring PROP becomes unattainable, leading to increased focus on its relaxations. In this paper, we focus on the relaxation of proportionality up to any item (PROPX), where proportionality is satisfied if an arbitrary item is removed from every agent's allocation. We show that PROPX is an appealing fairness notion for the allocation of indivisible chores, which approximately implies some share-based notions, such as maximin share (MMS) and AnyPrice share (APS). We further provide a comprehensive understanding of PROPX allocations, regarding the computation, approximation, and compatibility with efficiency. On top of these, we extend the study to scenarios where agents do not share equal liability towards the chores, and approximate PROPX allocations using partial information about agents' utilities.