The Formation of Production Networks: How Supply Chains Arise from Simple Learning with Minimal Information

TL;DR

The paper investigates how production networks can form endogenously when firms learn under limited information, without assuming equilibrium. It introduces an agent-based, reinforcement-learning–type framework where firms set prices, outputs, and inputs to maximize profits using minimal knowledge of production technologies, leading to steady-state networks. Prices and input choices evolve via explicit update rules that depend on observed profits and marginal product signals, demonstrating robust consistent learning across linear and CES technologies and under diverse shocks, with clear upstream and downstream transmission patterns. The framework scales to larger economies and offers insights into resilience and policy implications, while acknowledging empirical calibration challenges and avenues for future empirical validation.

Abstract

We develop a model where firms determine the price at which they sell their differentiable goods, the volume that they produce, and the inputs (types and amounts) that they purchase from other firms. A steady-state production network emerges endogenously without resorting to assumptions such as equilibrium or perfect knowledge about production technologies. Through a simple version of reinforcement learning, firms with heterogeneous technologies cope with uncertainty and maximize profits. Due to this learning process, firms can adapt to shocks such as demand shifts, suppliers/clients closure, productivity changes, and production technology modifications; effectively reshaping the production network. To demonstrate the potential of this model, we analyze the upstream and downstream impact of demand and productivity shocks.

Figures (122)

• Figure 1: Implied productive structure with linear technologies
• Figure 2: Consistent learning under linear technologies
• Figure 3: Dynamics under linear technologies
• Figure 4: Price dynamics
• Figure 5: Output dynamics