Amazon's Algorithm Does What Taxi Medallions Did

Barbara Kopple's Union Town documents a structural problem masquerading as a motivation problem — Amazon delivery workers attempting to organize in New York.

The film follows deliveristas, many undocumented, fighting for union representation with conviction and exhaustion. But defiance alone has never unionized a workforce whose substitutability is the entire business model.

The closest historical echo is New York City's taxi medallion crisis. Drivers organized successfully enough to win collective bargaining agreements. Medallion prices decoupled from driver earnings as owners treated them as speculative assets — and by the 2000s, medallion holders were lending drivers the right to drive in exchange for nearly all the money drivers earned.

Amazon's algorithm does the work medallions did for taxis. Deliveristas have no formal ownership stake. The substitutability is more naked. Assignment algorithms automatically route work to whoever will accept the lowest rate, the fastest delivery time, the least complaint. And app-based platforms require only that enough drivers remain available on the other side of one policy change. A wage agreement doesn't constrain something that can be reset with an update.

This is not an argument against organizing, it's an argument about which structural variable actually constrains wages. In taxi medallion systems, the constraint was ownership concentration and regulatory scarcity. In app-based delivery, the constraint is constant availability of a replacement labor pool. Kopple's film shows workers fighting the former fight against the latter problem. The organizing strategy that matters is one that solves the substitutability question itself. Not the one that assumes clarity and commitment can defend a position the structure was designed to make irrelevant.