Re-organization of the U.S. economy after the great recession

The Great Recession had deep and lasting effects on the American economy, with GDP cumulatively falling by $4 trillion below potential and nonfarm payroll employment continuing to decline even after the recession ended in June 2009, by an unprecedented 6%. In this paper we explore formally the idea that an economic system evolves dynamically over time, guided by Adam Smith’s invisible hand and it moves from one growth path to another as it is perturbed by both internal and external shocks. Key to analyzing such a complex system is that it is not necessary to know the behavior of individual components in order to understand the behavior of the entire system. Here we examine the U.S. economy as a self-organizing, adaptive complex system, by focusing on the network of firms linked through their respective industries, as represented in the national input-output table (IO table). This is an applied study and our data was gathered from the World Input Output database.

In order to measure changes or distance in the IO table over time we apply a cosine similarity measure to the network. Results show that a relatively large amount of reorganization occurred in the depth of the Great Recession, from 2008 to 2009. Moreover, some amount of rewiring occurs in each period, suggesting this is a normal phenomenon as the economy adjusts and evolves over time in response to basic changes in demand and supply, including productivity adjustments. In addition, the amount of rewiring declined over the period 2001 and 2006, and then again from 2012 onward; this is worth further exploration as a possible cause or effect of the lagging recovery.

Another important tool for examining networks is the distribution of transactions. The distribution of weights in the IO table follows a power law distribution, where a few inter-industry transactions are very large, and many such transactions are small, so that this represents a scale-free network. Our calculations show that intermediate sector transactions within the U.S. economy have become more concentrated over time into fewer industries. This also means that the national economy has become more vulnerable over time to targeted attacks such as on the real estate sector, and less vulnerable to random attacks or failures of industries.