Abstract:In this paper, we study the allocation of indivisible chores and consider the problem of finding a fair allocation that is approximately efficient. We shift our attention from the multiplicative approximation to the additive one. Our results are twofold, with (1) bounding how the optimal social cost escalates resulting from fairness requirements and (2) presenting the hardness of approximation for the problems of finding fair allocations with the minimum social cost. To quantify the escalation, we introduce cost of fairness (CoF) $\unicode{x2014}$ an alternative to the price of fairness (PoF) $\unicode{x2014}$ to bound the difference (v.s. ratio for PoF) between the optimal social cost with and without fairness constraints in the worst-case instance. We find that CoF is more informative than PoF for chores in the sense that the PoF is infinity regarding all EQX (equitable up to any item), EQ1 (equitable up to one item) and EF1 (envy-free up to one item), while the CoF is $n$ regarding EQX and 1 regarding EQ1 and EF1, where $n$ is the number of agents. For inapproximability, we present a detailed picture of hardness of approximation. We prove that finding the optimal EQX allocation within an additive approximation factor of $n$ is NP-hard for any $n \geq 2$ where $n$ is the number of agents and the cost functions are normalized to 1. For EQ1 and EF1, the problem is NP-hard when the additive factor is a constant and $n \geq 3$. When $n = 2$, we design additive approximation schemes for EQ1 and EF1.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the trade - off between fairness and efficiency in the allocation of indivisible chores. Specifically, the author focuses on how to find an approximately optimal allocation scheme to make the social cost as low as possible while meeting the fairness requirements. To study this problem, the author introduced a new quantitative index - Cost of Fairness (CoF), and made a comparative analysis between it and the existing Price of Fairness (PoF).
### Research Background
The resource allocation problem has always been a fundamental problem in economics and game theory. In multi - agent systems, resource allocation usually involves two orthogonal goals: social welfare and individual fairness. The research trend in recent years is to understand the trade - off between these two goals, especially in the allocation of indivisible tasks, how to minimize the efficiency loss while ensuring fairness.
### Fairness Concepts
The paper discusses two comparison - based fairness concepts:
1. **Equitability (EQ)**: All agents have the same cost.
2. **Envy - freeness (EF)**: No agent wishes to exchange its task package with that of other agents.
Due to the existence of indivisible tasks, a complete EQ or EF allocation may not exist, so researchers introduced relaxed conditions:
- **EQX**: After removing any one task, the costs of all agents are equal.
- **EQ1**: After removing one task, the cost of all agents does not exceed the cost of any other agent.
- **EF1**: After removing one task, the cost of any agent does not exceed its cost for the task packages of other agents.
### Main Contributions
1. **Definition and Analysis of Cost of Fairness (CoF)**:
- CoF is defined as the difference in social cost between the optimal allocation satisfying fairness constraints and the unconstrained optimal allocation in the worst - case scenario.
- The author proved that for EQX, EQ1 and EF1, CoF is n, 1 and 1 respectively (where n is the number of agents). This shows that EQ1 and EF1 have a better trade - off between fairness and efficiency.
2. **Computational Complexity**:
- For EQX allocation, the author proved that it is NP - hard to calculate an approximate solution with a social cost less than n.
- For EQ1 and EF1 allocations, when n ≥ 3, it is also NP - hard to calculate an approximate solution with a social cost less than a constant; but for the case of n = 2, an additive approximation algorithm was designed.
### Conclusion
By introducing CoF, the author provided a more accurate method to quantify the impact of fairness on efficiency, and in the allocation of indivisible tasks, EQ1 and EF1 showed good compatibility. In addition, the author also explored the computational complexity of optimization problems under these fairness criteria, further deepening the understanding of the trade - off between fairness and efficiency.
### Related Formulas
- **Price of Fairness (PoF)**: \[ \text{PoF} = \sup_I \frac{\min_{A \in P(I)} SC(A)}{\min_{A} SC(A)} \]
- **Cost of Fairness (CoF)**: \[ \text{CoF} = \sup_I \left( \min_{A \in P(I)} SC(A) - \min_{A} SC(A) \right) \]
where \( I \) represents an instance, \( P(I) \) represents the set of allocations that satisfy the fairness standard \( P \), and \( SC(A) \) represents the social cost of allocation \( A \).