Start with the change the policy is meant to make
A policy goal is a statement of the change a government or public program hopes to bring about. “Improve access to public transportation” expresses a direction, but it leaves important questions open. Access for whom? Access to which destinations? Does improvement mean shorter journeys, lower costs, more reliable service, or some combination? Answering these questions makes the goal easier to discuss and assess.
It helps to distinguish a broad aim from a more specific objective. A broad aim might be to make essential trips easier for residents without cars. An objective could focus on reducing the time it takes those residents to reach a clinic by transit. Specificity does not require pretending that every value can be reduced to one number. It gives policymakers, service providers, and the public a clearer basis for deciding what progress would look like.
Connect activities to outcomes
Before choosing indicators, sketch the steps through which a policy is expected to work. A transit agency might add a route, publish a clearer timetable, or extend operating hours. These are activities. The services delivered, such as additional trips, are outputs. Outcomes are changes people experience, such as more dependable journeys or better access to appointments. A longer-term goal may be broader still, such as improved access to essential services.
This distinction prevents a common mistake: treating evidence that work was done as proof that the intended benefit occurred. Counting new bus trips can show that service expanded, but it cannot by itself show whether people can use those trips to reach a clinic when needed. Conversely, an outcome may change for reasons outside the policy. Identifying the expected links between activities and outcomes helps reveal what needs to be measured and which assumptions need checking.
Choose a small set of useful indicators
An indicator is a defined measure used to observe some part of a goal. For transit access, possible indicators include the share of residents within walking distance of a stop, the frequency of service, the time required to reach a clinic, and riders’ reports of missed trips. Each answers a different question. A useful set usually includes measures close to the policy’s activities as well as measures of the outcomes it seeks.
For each candidate indicator, ask whether it is relevant to the goal, understandable, feasible to collect consistently, and sensitive enough to show meaningful change. Define its population, unit, and time period. “Average journey time,” for example, is ambiguous unless readers know whose journeys count, which destinations are included, and when travel is measured. Also consider what the measure leaves out. An average may improve while journeys remain difficult for people with mobility needs or residents in less-served areas.
More indicators do not automatically produce a better assessment. A long list can be costly to maintain and difficult to interpret. Select measures that answer distinct questions, and pair numbers with carefully gathered accounts of how a service works in practice. If survey responses or administrative records are used, explain their limits, including who may be missing from the data.
Set a baseline and interpret targets carefully
A baseline describes conditions before a policy begins or at a clearly identified starting point. It allows later measurements to be compared with an initial value, provided the definitions and collection methods remain reasonably consistent. If a method changes, that change should be documented; otherwise, an apparent improvement may reflect a new way of counting rather than a change in people’s experiences.
A target states the level of change sought within a chosen period. Targets can help direct effort, but they should be understood as planning tools rather than guarantees. An ambitious target may be useful, while a target chosen only because it is easy to meet may reveal little. Public reports are more informative when they explain why a target was chosen, whether it was met, and what else might have affected the result.
Use indicators to learn, not just to score
Indicators need interpretation. If clinic journey times fall, expanded service may be one explanation, but changes in clinic locations, traffic, or travel patterns could matter too. Comparing results across neighborhoods or groups may reveal uneven effects, though small groups can require care to protect privacy and avoid drawing strong conclusions from limited data. A measure can show that something changed without establishing that a particular policy caused the change.
Measures can also shape behavior. If an agency is judged only on the number of trips it runs, it may have little incentive to examine whether routes match residents’ needs. Reviewing indicators alongside feedback and the policy’s original goals makes it easier to notice such gaps. The aim is not to find one perfect metric. It is to build a transparent, manageable set of measures that supports better questions, honest reporting, and adjustments when the evidence calls for them.
