Charles Perrow described the 1979 Three Mile Island nuclear reactor meltdown as a Normal Accident. Normal accidents are a class of events produced when subsystems in risky, complex and tightly coupled systems fail. Tight linkages between subsystems propagate failure, and local breakdowns cascade into systemic collapse. Diane Vaughn applied the normal accident metaphor to the 1986 Challenger Disaster. A gasket failed because it was cold outside, on blastoff rocket fuel leaked and then exploded, destroying the rocket booster. This in turn destroyed the space shuttle, and seven crewmembers died.
The COVID-19 pandemic, I think, is reasonably described as a normal accident, complete with cascading social and economic failures. Virus mutations and viruses jumping across species, including into humans, are a normal part of natural selection. This has happened in the past. It will happen again in the future.
The key difference is that the human race is now socially and economically tightly coupled on a global scale. Tight social coupling permits rapid transmission, multiplying the impact and speed of contagion, and tight economic coupling propagates cascading failures across multiple sub-systems.
In the case of the Challenger disaster, Vaughn found that some engineers were aware that catastrophic failure was possible, but organizational leaders ignored or suppressed their warnings. Similarly, for many years public health professionals have known that a COVID-19 type pandemic was coming, and the U.S. Pentagon had even gone so far as to produce a playbook for government response. As in other normal accidents, we see some leaders ignoring or suppressing expert warnings and advice both before and as system failures cascade.
As in other normal accidents there is a tension between the advice of technical experts and the needs of managers and politicians to keep things running. Federally in the U.S., we see a consistent contrast between the scientific and military establishments, both of which urged quick action, proactive planning, and social distancing, and a president who urged people to go to church and sowed distrust of experts. Some states and countries quickly instituted contact tracing, closed all non-essential businesses, and forbade congregating in groups, while others urged people to stay at work and go shopping.
When managers and politicians win and scientific experts are silenced, the pain of normal accidents expands as system failures propagate. The U.S. is particularly at risk because its political system is decoupled both ideologically and organizationally. Ideologically, the major party in power has endorsed distrust of both government and science in its pursuit of election success. Organizationally, the federal system of government, which relies on local decision makers for public health action, is unusually ill-suited to a coordinated response. Germany, a country which has been more effective in reducing infections and preventing economic cascades, also has a federal system of government, illustrating that policy effectiveness is never mono-causal, but reflects configurations of policy and political responses.
Economically, we have long had a tightly coupled economy. Economic recessions happen because overproduction or financial malfeasance spreads from one corner of the economy to another. Old school recessions of over-production cascade relatively slowly because sectors and economic geographies, while tightly coupled, fail at a relatively slow pace – months not days. Financial crises happen rapidly because the financial sector is both central to much economic activity and is itself internally tightly coupled—when one big bank fails they all do. Shutting down much of the economy in order to slow disease spread propagates recession even faster.
But things are even worse. Today’s pandemic is happening in the wake of two earlier economic crises that left the United States more vulnerable to economic system failure. These vulnerabilities have lain latent, like herpes, in our social body; their activation by the COVID-19 recession will magnify its destructive impact.
The first is a slow moving crisis that began in the 1980s, as the U.S. pursued public and private policies that weakened the bargaining power of low-skilled labor, shrank union influence, shredded an already weak safety net, and accelerated economic globalization. Some of the results are that today 40% of jobs pay poverty-level wages and 45% of households have no or inadequate health insurance. The U.S. has produced an economy that leaves its people more vulnerable to both the virus and the COVID-19 recession than perhaps any other high-income country in the world. While other high-income countries have weakened their institutional protections, none are as institutionally weak as is the United States.
The second systemic economic crisis happened in 2008, when the world economy shut down as a result of cascading defaults of collateralized mortgage securities, propagated through a tightly connected world financial system. That tightly coupled financialized economy with its risky financial practices remains intact and risky behaviors have changed in form but not levels. Since 2008, corporate debt, repackaged as securitized bonds, has proliferated. Many companies in the real economy now use debt instead of earnings to fund investment. The COVID-19 pandemic could very well lead to systemic collapses of the corporate bond market, leading to a longer 1930s style great depression. Of course, the mortgage market may fail a second time as well.
In the two previous crises the U.S. government responded by securing the well-being of large firms and their stockholders, leaving the systemic risks to households and workers. Today’s crisis threatens both households and the production economy. The key question is whether the political response, like the response to the two earlier crises, will again be socially destructive, preserving the wealth and “freedom” of the wealthy and the largest corporations, while laying the groundwork for yet another round of citizen pain when the next normal accident unfolds.
It could go the other way, of course. This could be the wake-up call to strengthen institutional protections and transcend the United States’ low-wage, precarious social system. Systemic transitions to new institutional configurations have happened before, with the 1930s “New Deal” being the most obvious. We know how to build more resilient institutions to support the well-being of citizens; now is the time for a new “new deal”.