AI Economics

Concept

Transaction costs

Transaction costs are the costs of arranging and governing an exchange beyond the price of the good or service itself.

Explanation

Coase used transaction costs to explain why firms exist. When repeatedly buying an input on the market is costly, an organization can replace many separate contracts with an employment relationship and internal direction. A firm expands until organizing another activity internally becomes more costly than obtaining it through the market.

AI can lower both sides of this boundary. Agents may reduce the cost of search, comparison, contracting, and monitoring across firms, while internal AI systems may also make management, planning, and communication cheaper. The theory therefore does not predict that AI must make firms either larger or smaller.

Technical capability is only part of the transaction cost. Identity, delegated authority, liability, platform access, privacy, and trust determine whether an agent is actually allowed to transact and who bears the cost when it makes a mistake.

Examples

  • Booking a flight

    An agent may understand the itinerary but still be blocked by authentication, anti-bot systems, payment authority, and uncertainty over responsibility for a bad purchase.

  • Make or buy

    A company may outsource a task when agents make suppliers easy to find and supervise, or bring it inside when the same technology makes internal coordination cheaper.

Sources