📚 Provision of Social Welfare: Mathematical Optimisation and Decision Models | 社会福利提供：数学优化与决策模型
Mathematics may not seem an obvious partner to social welfare, but behind every efficient food bank, every fair allocation of housing, and every rapid disaster response lies a network of mathematical models. A-Level Mathematics, especially the Decision Mathematics strand, equips students with tools to analyse constraints, optimise resources, and evaluate fairness – all fundamental to the provision of social welfare. This article explores how linear programming, network flows, game theory, and other mathematical techniques transform abstract welfare goals into actionable, equitable strategies.
数学似乎与社会福利风马牛不相及，但在每一次高效的食品援助、每一次公平的住房分配以及每一次快速的灾害应对背后，都隐藏着数学模型的网络。A-Level 数学，尤其是决策数学方向，赋予学生分析约束、优化资源和评估公平性的工具——这些都是社会福利提供的基石。本文探讨线性规划、网络流、博弈论等数学方法如何将抽象的福利目标转化为可操作、公平的策略。
1. Linear Programming for Budget Allocation | 线性规划与福利预算分配
provides a systematic way to maximise the total social benefit. Decision variables x₁, x₂, … onstraints include the overall budget Σ aᵢxᵢ ≤ B and minimum service requirements xᵢ ≥ mᵢ.
地方议会经常面临将固定预算分配到多个福利项目（如送餐服务、青年中心和心理健康支持）的任务。线性规划（LP）提供了系统方法来最大化社会总收益。决策变量 x₁, x₂, … , xₙ 表示分配给每个项目的金额，目标函数 Z = Σ cᵢxᵢ 用权重系数 cᵢ 捕捉总收益。约束条件包括总预算 Σ aᵢxᵢ ≤ B 和最低服务要求 xᵢ ≥ mᵢ。
Graphically, the feasible region is bounded by resource lines, and the optimal solution lies at a vertex. Using the simplex method or simple graphical methods for two-variable cases, policymakers can determine the precise mix that delivers the greatest welfare impact without overspending. Sensitivity analysis further reveals how changes in unit costs or benefit weights alter the optimal allocation, providing a robust framework for responsive governance.
从图形上看，可行区域由资源线界定，最优解位于一个顶点。通过单纯形法或简单的两变量图解法，决策者可以确定在不超过预算的情况下实现最大福利影响的最优组合。敏感性分析进一步揭示单位成本或收益权重的变化如何改变最优分配，为响应式治理提供了稳健的框架。
2. The Transportation Problem in Food Aid | 运输问题与食品援助
Food banks and disaster relief agencies must ship supplies from several warehouses to multiple distribution centres at minimum cost. The transportation problem, a special case of linear programming, models supply points with capacities Sᵢ and demand points with requirements Dⱼ, where total supply equals total demand. The objective is to minimise Σ Σ cᵢⱼxᵢⱼ, where xᵢⱼ is the quantity shipped from i to j and cᵢⱼ is the unit transport cost.
食品银行和救灾机构必须以最低成本将物资从多个仓库运送到多个分发中心。运输问题是线性规划的一个特例，它模拟了具备容量 Sᵢ 的供应点和需求量 Dⱼ 的需求点，其中总供应等于总需求。目标是最小化 Σ Σ cᵢⱼxᵢⱼ，其中 xᵢⱼ 是从 i 到 j 的运输量，cᵢⱼ 是单位运输成本。
The stepping-stone method or the MODI (modified distribution) method allows logistic planners to iteratively improve an initial feasible solution. In welfare contexts, this ensures that perishable goods reach the most vulnerable communities quickly and cost-effectively. Moreover, unbalanced problems – where supply exceeds demand – can be resolved by introducing dummy demand points, mirroring surplus being diverted to long-term storage or neighbouring regions.
踏脚石法或 MODI（修正分配法）使物流规划者能够迭代改进初始可行解。在社会福利情境下，这确保易腐物品快速、经济地到达最脆弱的社区。此外，不平衡问题——供应超过需求时——可通过引入虚拟需求点来解决，反映剩余物资转入长期储存或邻近区域。
3. Network Flow and Community Centre Siting | 网络流与社区服务中心选址
Deciding where to build community health centres so that the maximum number of residents are within a 15-minute walk involves maximum flow and shortest path algorithms. A network is constructed with nodes representing neighbourhoods and potential sites, and arcs weighted by capacity (pedestrian pathways or road links). The Ford-Fulkerson algorithm can determine the maximum flow of people from residential zones to a proposed site under capacity constraints.
决定在哪里建造社区卫生中心，以使尽可能多的居民在步行15分钟范围内，涉及最大流和最短路径算法。构建一个网络，节点代表街区和潜在地点，弧线按容量（人行通道或道路连接）加权。Ford-Fulkerson 算法可以在容量限制下，确定从住宅区到拟选址的最大人流量。
Dijkstra’s algorithm, on the other hand, pinpoints the shortest routes from all homes to the nearest facility, aiding equitable placement. By layering socio-economic data on the network nodes, planners can prioritise deprived areas. This blend of graph theory and welfare policy ensures that ‘accessibility’ is not just a slogan but a measurable, optimisable quantity.
另一方面，Dijkstra 算法找到了从所有家庭到最近设施的最短路径，有助于实现公平布局。通过在网络节点上叠加社会经济数据，规划者可以优先考虑贫困区域。图论与福利政策的这种结合确保“可达性”不只是一个口号，而是可衡量、可优化的量。
4. Critical Path Analysis for Emergency Shelters | 关键路径分析与应急避难所
When a natural disaster strikes, setting up emergency shelters involves a sequence of interdependent tasks – securing land, installing sanitation, setting up tents, registering evacuees. Critical Path Analysis (CPA) identifies the longest sequence of dependent activities that determines the project’s minimum completion time. Each task has a duration, and earliest start times (EST) and latest finish times (LFT) are calculated via forward and backward scans.
自然灾害发生时，建立应急避难所涉及一系列相互依赖的任务——确定用地、安装卫生设施、搭建帐篷、登记撤离人员。关键路径分析（CPA）识别决定项目最短完成时间的最长依赖活动序列。每项任务都有持续时间，通过前向和后向扫描计算最早开始时间（EST）和最迟完成时间（LFT）。
Float times reveal which activities can be delayed without affecting the overall shelter readiness. In welfare provision, CPA helps coordinate volunteers and supplies so that vulnerable groups are not left exposed for longer than necessary. A cascade chart (Gantt chart) then visualises the schedule, enabling aid coordinators to react dynamically when a late delivery threatens the critical path.
浮动时间揭示了哪些活动可以推迟而不影响避难所的整体就绪。在福利提供中，CPA 有助于协调志愿者和物资，以免弱势群体暴露过久。随后，级联图（甘特图）将进度可视化，使援助协调员能够在交付延迟威胁关键路径时动态做出反应。
5. Fair Division and Game Theory | 公平分配与博弈论
Allocating a fixed stock of social housing or limited childcare vouchers among claimants with different needs introduces questions of fairness. Cooperative game theory provides solution concepts such as the Shapley value, which distributes the total ‘value’ of a resource pool according to each participant’s marginal contribution. In welfare, this translates to weighting allocations by vulnerability indices, so that an individual’s share reflects their relative need.
在需求各异的申请者之间分配固定的社会住房存量或有限的托儿券，引发了公平性问题。合作博弈论提供了诸如沙普利值等解概念，它根据每个参与者对整体的边际贡献来分配资源池的总“价值”。在福利中，这转化为按脆弱性指数加权分配，使个体的份额反映其相对需求。
The divide-and-choose method and the adjusted winner procedure are practical fair-division protocols that ensure envy-freeness: no claimant feels they would have been better off with someone else’s bundle. By encoding these protocols into allocation software, housing officers can demonstrate procedural equity and reduce appeals. Game theory thus operationalises lofty ideals of distributive justice into transparent algorithms.
分割-选择法和调整赢家程序是实用的公平分配协议，确保无嫉妒性：没有申请者会觉得拥有他人的包会更好。通过将这些协议编码到分配软件中，住房官员可以展示程序公平并减少申诉。博弈论由此将分配正义的崇高理想转化为透明的算法。
6. Matching Algorithms for Social Housing | 社会住房的匹配算法
Many welfare systems let applicants rank their preferred housing options while landlords rank applicants. The resulting two-sided matching problem calls for stable assignments where no landlord-applicant pair would rather be matched with each other than their current match. The Gale-Shapley deferred acceptance algorithm achieves stability, though which side proposes affects the outcome’s bias.
许多福利系统让申请者对心仪住房排序，同时房东对申请者进行排序。这形成的双边匹配问题需要稳定分配，即没有房东-申请者对双方来说宁肯彼此匹配而不是现有匹配关系。Gale-Shapley 延迟接受算法可实现稳定性，尽管哪一方发起提议会影响结果的偏向。
In welfare contexts, the applicant-proposing version often yields more favourable results for vulnerable groups. School choice and organ donation matching apply the same logic. By studying these algorithms, students learn that mathematical abstraction can directly impact hundreds of thousands of lives, guiding reforms in how society allocates scarce, life- changing resources.
在福利背景下，申请者发起版本的算法通常对弱势群体更有利。学校选择和器官捐献匹配也运用了相同的逻辑。通过研究这些算法，学生们认识到数学抽象可以直接影响数十万人的生活，指导社会在如何分配稀缺、改变生活的资源方面进行改革。
7. Queueing Theory in Healthcare Delivery | 排队论与医疗服务递送
Walk-in clinics and social care hotlines experience random arrival patterns. Queueing theory models the system as M/M/1 if arrivals follow a Poisson process and service times are exponential with a single server. Performance measures – average queue length Lq, waiting time Wq – is to determine the staffing levels needed to keep waiting times below a legislated target.
门诊诊所和社会关怀热线面临随机到达模式。如果到达遵循泊松过程且服务时间服从指数分布的单服务台，排队论将系统建模为 M/M/1。性能指标——平均队列长度 Lq、等待时间 Wq——由交通强度 ρ = λ/μ 导出。福利管理者可以利用这一点确定在法定目标下保持等待时间所需的人员配置水平。
More complex models with priority classes mirror triage systems, where urgent cases jump the queue. Simulation software then tests various configurations, ensuring that elderly and disabled clients are not systematically disadvantaged by average-focused policies. Mathematics thus directly shapes the human experience of waiting for care.
带有优先级的更复杂模型反映了分诊系统，其中紧急案例可以插队。模拟软件随后测试各种配置，以确保老年和残障客户不会被以平均值为中心的政策系统性地置于不利地位。数学因此直接塑造了等待照护的人类体验。
8. Cost-Benefit Analysis with Decision Trees | 决策树与成本效益分析
Should a government invest in a new universal credit scheme or extend an existing child benefit programme? Decision trees frame uncertain outcomes with probabilities and monetary payoffs. At each chance node, expected monetary value (EMV) is calculated. Sensitivity analysis on the probabilities reveals the thresholds at which a policy becomes viable.
政府应该投资新的通用福利金计划还是扩展现有的儿童福利项目？决策树用概率和货币收益来框定不确定结果。在每个机会节点，计算期望货币值（EMV）。对概率的敏感性分析揭示了政策变得可行的阈值。
Utility functions can replace pure monetary payoff to capture risk aversion and social preferences. In welfare CBA, intangible benefits like improved mental well-being are monetised through willingness-to-pay studies. Mathematics provides the structure to weigh these profoundly human costs and benefits in a transparent, replicable manner.
效用函数可以取代纯货币收益，以捕捉风险厌恶和社会偏好。在福利成本效益分析中，通过支付意愿研究将改善的心理健康等无形收益货币化。数学提供了一种结构，以透明、可复制的方式权衡这些深刻的人类成本与收益。
9. Simulation for Policy Stress-testing | 政策压力测试模拟
Before rolling out a nationwide welfare reform, governments use Monte Carlo simulation to stress-test the system. Random numbers generate thousands of possible economic scenarios – varying unemployment rates, inflation, uptake rates – and the model outputs the distribution of budget deficits and poverty headcounts. This stochastic approach acknowledges that deterministic models cannot capture real-world volatility.
在全国推行福利改革之前，政府使用蒙特卡罗模拟对系统进行压力测试。随机数生成数千种可能的经济情景——变化的失业率、通胀、参与率——模型输出预算赤字和贫困人口数的分布。这种随机方法承认确定性模型无法捕捉现实世界的波动。
Cumulative frequency graphs and confidence intervals then inform policymakers of the risk of breaching a fiscal rule or missing a deprivation target. A-Level Statistics topics like the normal distribution and random sampling are the building blocks of such simulations. The lesson is powerful: mathematical prudence can prevent real welfare cuts from backfiring.
累积频率图和置信区间随后告知政策制定者违反财政规则或未能达到减贫目标的风险。A-Level 统计学主题如正态分布和随机抽样是这类模拟的基石。其启示是强大的：数学审慎可以防止真实的福利削减适得其反。
10. Ethical Algorithms and the Limits of Models | 伦理算法与模型的局限
Mathematical models in welfare are not ethically neutral. An LP solution that maximises total benefit might ignore minority groups whose needs are costlier to meet. Incorporating additional constraints, such as a minimum proportion of resources for each group, translates legal equality duties into mathematical form. However, these constraints are choices made by humans, not derived from the data.
福利中的数学模型在伦理上并非中性。一个最大化总收益的线性规划解可能会忽视那些满足其需求成本更高的少数群体。纳入额外约束，例如为每个群体分配最低资源比例，将法律平等义务转化为数学形式。然而，这些约束是人类的选择，并非来自数据。
Students are encouraged to question the assumptions behind any algorithm: Are the weightings fair? Who is excluded by the modelling? By engaging with the limitations, they develop a critical mathematical literacy essential for modern citizenship. Provision of social welfare, when illuminated by mathematics, becomes a discipline of both rigour and compassion.
我们鼓励学生质疑任何算法背后的假设：权重是否公平？建模排除了谁？通过面对这些局限，他们培养了现代公民至关重要的批判性数学素养。数学照亮下的社会福利提供成为了一门兼具严谨与同理心的学科。
Published by TutorHao | Mathematics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图