Economists finally figured out when you can actually measure how shocks spread through networks

This paper identifies the mathematical conditions under which you can look at economic data and figure out how different firms or regions affect each other when something goes wrong. Previously, economists couldn't tell whether spillover effects came from hidden connections between actors or from something everyone was experiencing at once. The paper shows you need asymmetric patterns in how different parts of the system amplify shocks — when that's present, the network structure becomes visible in the data.

For decades, economists studying production networks, financial contagion, and regional spillovers have had a measurement problem: you observe correlated behavior, but you cannot see the actual connections causing it. This paper gives a precise mathematical answer to when those connections become identifiable from data alone. The constraint is real and specific: spectral heterogeneity must exist. When it doesn't, the network is literally indistinguishable from simpler alternatives. This matters because policy decisions about systemic risk, supply chain resilience, and crisis transmission all depend on knowing which connections actually drive contagion. A government or central bank acting on an inferred network that fails the identification conditions is making decisions based on statistical ghosts.