Public Engagement for a Long-Range Transportation Plan

The two-audience problem

Every long-range plan has to satisfy two audiences whose engagement looks nothing alike. Organized stakeholders — committee members, municipal staff, advocacy groups, adjacent agencies — are a known list, will engage at length, and expect to be asked substantive questions. The general public is an unknown population, will give you ninety seconds, and mostly won't come looking.

Agencies usually solve this by running two separate processes, which produces two separate datasets that have to be reconciled by hand in the appendix — and which makes the obvious question ("how many people did you actually hear from?") surprisingly hard to answer. Running one survey through two channels removes the reconciliation entirely.

Reporting through a months-long window

Don't wait for the close. The Report page is live, and the useful discipline is to check the respondent-density map and the demographic composition every two weeks. A gap found in week three is an outreach task; the same gap found at the close is a paragraph in the appendix explaining a limitation. Same data, completely different document.

At the close, the branded PDF gives you the appendix package. Keep the CSV and GeoJSON exports with it — a plan that gets challenged years later is much easier to defend when the underlying responses can still be produced. Title VI documentation →

Common questions

How do we run an invited process and an open process without double-counting?

Run them as one survey with two distribution channels. Email invitations to your committee and stakeholder lists produce identity-verified responses; a public link and posted QR codes produce open responses. Where an email address is collected on either path, TIPS deduplicates across both — so a technical advisory committee member who also scans the flyer at the library is one respondent, not two.

A plan window runs for months. Does the survey have to stay frozen?

No, but understand what changes. TIPS lets you edit a live survey, and it snapshots each question's wording onto responses already collected, so historical answers keep meaning what they meant. Structural changes bump the survey version, which is visible in reporting. The practical guidance is to add questions freely and rewrite existing ones sparingly — a mid-window rewrite of a core question makes the aggregate harder to defend even though the data stays intact.

How do we handle multiple phases — visioning, alternatives, draft plan?

Copy the survey per phase rather than reusing one instrument. Each phase asks genuinely different questions, and a deep copy keeps the demographic and profile sections identical across phases so you can compare who participated at each stage. That comparison — did the same population show up in phase three as in phase one? — is a finding worth reporting.

What do we do with hundreds of open-ended comments?

Design so you have fewer of them and they are better targeted. One open-ended box placed after a structured question about the same topic produces comments that are already sorted by subject; three open boxes scattered through an instrument produce an undifferentiated pile. Everything exports to CSV for whatever coding process your team uses.

Can respondents prioritize investments rather than just react to them?

Yes — a Route Ranker question puts candidate projects or corridors in preference order, which is far more useful for a constrained plan than approve/disapprove on each one. Ranking forces the same trade-off the plan itself has to make, and the results tend to hold up better in front of a board because they acknowledge scarcity.

What goes in the plan appendix?

The branded PDF export covers the standard package — counts, per-question charts, density map, demographic cross-tabs, and the verified/anonymous split. Keep the raw CSV and GeoJSON alongside it; if the plan is challenged later, the ability to produce the underlying data is worth more than any summary.