A school district has children who need rides and a transportation system that cannot serve them all. The district still has to get those children to school safely and reliably. Platform companies sees something else: an opportunity to match demand with a flexible supply of drivers, routes, vehicles, and scheduling software.

The service may help the district meet an urgent need. I want to know what happens to the district's underlying obligation once the gap in its own capacity becomes a commercial market.

I've been looking at companies that provied student transportation because school transportation is a particularly revealing case. It combines safety requirements, labor, family schedules, geography, disability access, school attendance, and public budgets. Software can coordinate parts of that work. It cannot settle all of those responsibilities by itself.

A contract might solve an urgent transportation problem for a district. It might also create a longer-term dependence on a third party to perform work the district once could do, or never developed the capacity to do. The relevant questions extend past the app and the trip.

Who decides which routes are viable? Who handles a last-minute absence? Who carries responsibility when the system cannot find a driver? What happens to the working conditions of the people giving rides? Can the district see enough of the underlying operation to evaluate performance? If the vendor exits, what does the district still know how to do?

These are questions I would ask of any arrangement, whether it uses a contractor or public employees. They are especially important when technology makes the arrangement appear self-contained.

A platform can make the trip look like a self-contained transaction. The child needs a ride. A driver accepts it. A vehicle appears. The trip is complete. Meanwhile, the school district remains responsible for the conditions that made transportation necessary and for the students the platform cannot serve.

This pattern appears in elder care, food access, health services, housing, and countless other places where public or communal capacity has been allowed to weaken. Software steps into the vacancy. Investors and customers may call it innovation. The people relying on the service have to live with the arrangement after the launch announcement.

I would want a district to ask whether the new system helps it fulfill its obligation, and whether it retains the knowledge and capacity to serve students when a contract ends. If the only lasting change is a new intermediary between a person and a service, we should count that as a cost.