Housing Need and Equity Diagnosis · Lesson 4

Framing the Problem and Testing Equity

5 min read

Lesson objectives

  • Turn a diagnosis into a bounded problem statement and equity test.
Reading tools

After calculating cost ratios and counting dwellings, do not open a brief with 'the city needs more housing.' That is too broad to identify who faces harm, where, on what evidence, and over which period. Distinguish total supply from outcomes for the intended households. Unit production can rise while low-income households remain far from suitable locations, or usable stock can shrink because maintenance is delayed. A good problem statement names the affected population, the kind of harm, the spatial and time boundary, direct evidence, and what remains unmeasured. It is a revisable working proposition, not a slogan or a decision already made to build a project.

Organise conflicting evidence rather than hiding it. A household survey describes the sampled households; a permit register records building approvals; a service record describes connections or complaints. None alone measures every aspect of need. Ask how the sample was selected and how large it is, when a register was updated, and how complete its coverage is. If a developer reports vacant flats while households say they cannot find homes, both accounts may be true: the vacant units might be unaffordable or far from jobs. Test that possibility by linking price, location and unit size with household circumstances before announcing either a general shortage or general surplus. UN-Habitat's guide for Lebanon illustrates a sector diagnosis followed by policy priorities and an action plan; its country-specific recommendations are not universal law.

Test equity after stating the intended outcome. A general average can hide a heavier burden among renters, older people, disabled residents or households with variable earnings. Choose distinctions that matter to the decision and can be measured ethically; do not publish small cells that identify people. Ask whether a programme rule excludes someone without a formal lease or someone unable to apply online despite real need. Do not assume every member of a group has the same experience. In practice, calculate the indicator for groups whose differences would alter the intervention, then ask who is missing from the data and why.

Use a data-gap register with three columns: unanswered question, evidence needed, and how the answer could change the choice. For example, are the 40 high-rent flats near the surveyed households' jobs? The team needs work locations, journey times and unit prices. If the units are both distant and expensive, they will not meet the target need without another instrument. Another question is whether the 90 applications are unique and current. A privacy-preserving clean-up and update dates are needed; duplicates would alter the programme demand estimate. Rank evidence collection by the risk of making the wrong decision, not by which existing spreadsheet is easiest to obtain.

Lesson summary

A large quantity of data is not enough if it describes different things. Make the policy question specific about what the team knows and does not know; compare household evidence with units, locations and a matching time period. An equity test asks whom the measure represents, whom it misses, and whether that changes the instrument chosen. Every data gap in a brief should have a reason why filling it could affect the decision.