Balance fairness and capacity
6 min read
Lesson objectives
- Apply balance fairness and capacity to a fictional humanitarian case and justify the decision, limits and observable output.
Please sign in to use all the tools
Once you sign in, the reading tools are yours: larger or smaller text, typeface, line spacing, focus mode, highlighting and bold, and notes saved to your account and gathered in My study.
Accountability does not disappear when every option is imperfect. In humanitarian work, people may ask for a service point farther from the centre while staff fear slower decisions on existing applications or weaker quality oversight. A professional decision begins by naming each option's possible losses instead of presenting management's preferred choice as the only realistic one. Separate a non-negotiable condition, such as protecting sensitive information, from a criterion open to trade-off, such as waiting time. Then identify who bears each effect. The average wait can fall even while a distant group remains unable to use the service. A first-come, first-served rule is not neutral when access to the queue is unequal. Choose within actual capacity, explain who has not benefited, and set out how that exclusion will be reviewed.
Use a compact trade-off matrix, but do not let a number conceal judgement. Put options in rows and access, equity, safety, deliverability and reversibility in columns. Each cell should carry evidence, not just a score. Moving two staff members would delay existing application review by a day is more informative than capacity equals two out of five without explanation. If weights are used, disclose who chose them, how they affect the ranking, and whether a plausible change reverses it. Two options may tie numerically while harming different groups. An aggregate score should not settle that conflict. Set a minimum condition for safe access and data protection, then compare only options that meet it. Invite affected people to discuss what criteria mean in practice rather than asking them to accept unexplained weights.
Distinguish three layers of information. A fact is the number of pending files and the wait recorded last week. A forecast is what might happen to that wait if one staff member moves. A value judgement is how much delay is tolerable in exchange for access by people currently excluded. Do not present a forecast as a measured fact; use a short trial to test it. Do not present a value judgement as the result of one survey; discuss it with affected groups and record objections to the team's recommendation. Where data conflict, ask if sources measure the same thing. Visits are not households, and the number of messages is not a measure of harm severity. Recording evidence limits makes a decision easier to correct later; manufactured certainty resists correction.
Option A may appear faster because no worker is redeployed, but it shifts access costs to households. A household unable to use shared transport because of timing, disability or care duties may still have no usable service. Option B may appear more equitable, yet could slow decisions needed urgently by applicants already waiting. The recommendation should therefore be conditional: trial a mobile day with a monitoring boundary for older applications, and retain targeted transport for people the trial cannot reach. Name who can halt the trial, what data they will see, and how people will be informed before a further timetable change. A reasoned refusal to open a permanent office today must not become a vague promise that one will open later. Attach a funding review date and a specific decision question.
Review the trade-off after the trial rather than only when announcing it. Eastern households may travel less while applications from others accumulate. The average wait may stay level because easy files are completed quickly while older cases stall. Compare the distribution of waits, not only the mean, and report effects on people absent from the first consultation. If B fails, that does not prove the access barrier was imaginary. The announcement may have been weak or the day poorly chosen. Investigate the failure mode before declaring the conclusion. If B works, two successful weeks still do not establish that a permanent site is affordable and sustainable. Record which assumptions changed and which remain unsettled so the next decision improves.
Reverse the decisive assumptions before sending the memo. If a distant household cannot reach supported transport either, A may fail the minimum access condition. If the redeployed worker can clear existing files remotely without delay, B's time objection weakens. These are tests, not known facts. Name who will run them and when. A stop trigger should also watch the oldest files and reopened cases, so staff are not rewarded for closing easy files quickly while complex applicants wait.
If you select a time-based stop trigger, check that it does not reward the wrong behaviour. Staff might rush to close easy files to protect an average while complex files age. Monitor the age of the oldest files and the share reopened after premature closure. Do not repair the performance report by exposing household data; groups that do not identify individuals are enough. Good professional judgement ties each measure to the harm it is meant to detect and specifies an action when the trigger is crossed, rather than placing a decorative number on a dashboard.
Lesson summary
Accountability begins with a decision that people can understand and challenge, not with a feedback box detached from management.