The baseline begins before construction

A project can change its setting before it changes the ground. A public announcement, land discussion, procurement signal or proposed route may influence expectations and choices long before construction. Whether that happens, to whom and with what consequence is an empirical question. It should not be assumed away because a formal assessment begins later.

For a board or public institution, the practical problem is timing. A baseline collected after an important decision can describe the present accurately while missing the earlier condition against which change should be understood. The answer is not to claim that every shift was caused by the project. It is to make the sequence of decisions and observations visible.

What a baseline must explain

IFC Performance Standard 1 calls for a risk and impact identification process grounded in recent environmental and social baseline data at an appropriate level of detail. World Bank ESS1 guidance asks for the dates, reliability, gaps and uncertainty of baseline data and for relevant changes anticipated before a project commences. These requirements support disciplined assessment; they do not, by themselves, prove that a particular announcement caused a particular outcome.

A useful starting-condition record therefore answers more than “what do we see today?” It records when the observation was made, what was already known about the project, what people could reasonably have heard, and where evidence is thin. If an earlier record does not exist, that absence belongs in the decision record. A later survey should not be labelled “pre-project” without qualification simply because physical works have not begun.

Record the sequence, not just the snapshot

Four dates matter. First, when the project became a credible prospect. Second, when material decisions or announcements were made. Third, when observations were gathered. Fourth, when the evidence was reviewed by someone with authority to act. The dates may differ across communities, suppliers and institutions.

The record should also identify the source of each observation: administrative data, a household or business survey, a consultation record, a site visit or another documented source. Each has a different reach and limitation. A small early sample may reveal a question worth testing; it does not justify a population-wide conclusion. A grievance log can show what was reported; it cannot, on its own, measure everyone’s experience.

Separate observation from attribution

An observed change, a plausible mechanism and a demonstrated project effect are three different statements. For example, an increase in local rental prices may be observed. Demand linked to a proposed project may be one possible explanation. Establishing how much of the increase is attributable to that project requires a suitable comparison and attention to other forces. A credible executive brief keeps those statements separate.

This discipline also protects positive claims. New registrations, training interest or supplier enquiries may signal opportunity, yet they are not evidence that households or firms have gained durable capability. The question is what those changes enable, for whom, and whether they endure if plans change.

Put an owner beside the evidence

Baseline work becomes useful when it changes a decision. A board paper can identify which assumptions remain untested, who will obtain the next evidence, what threshold triggers a response and when the issue returns for review. Where risks fall within a statutory environmental and social assessment, the early record should inform that process and the responsible specialists. It is not a substitute for required assessment or consultation.

The immediate executive question is simple: what starting condition are we using for this decision, and can we defend its date and limits? If the answer is uncertain, that uncertainty is itself decision-relevant. It can be named, assigned and revisited before the next commitment narrows the available choices.