Back   OpEdNews
Original Content at
(Note: You can view every article as one long page if you sign up as an Advocate Member, or higher).

January 10, 2009

Cascading Failures

By James Brett

Cascading failure within an emergent process is hard to understand and more difficult yet to fix, but fix it we must,and this is not the time for compromise.


Jeffrey D. Sachs, whose monthly essay in Scientific American is one good reason among many to subscribe, writes in the current issue about the similarity between an electricity blackout, such as we have seen in the American northeast several times, and the collapse of the financial markets. The phenomenon of emergent properties and processes is not well understood by most Americans, but is recognized in economics as the difference between microeconomics, the kind where individuals make nearly rational decisions, and macroeconomics, where emergent phenomena happen and high-rollers lose their grip. In an electrical blackout one thing leads to another, and we humans are already used to the idea that there are chains of causation, so the blackout is understandable in principle, if not in detail. What happens in the blackout, though, is a cascading effect where the vulnerabilities of a system are tested and fail as the chain of causation not only moves from one locality to others, but spreads out fan-wise so that each new failure adds a multiplier effect, a factor that was not present before, in this case a huge surge of electricity that "wants" to go somewhere, but is too much for any individual part of the system. Soon the entire system is engulfed in failure and safeties all over the place are locked into place to prevent further damage, and so it is with markets today. One of the points of Sachs's essay is that cascading emergent phenomena are a class of phenomena which we often have a psychological resistance to understanding in foresight. We see them only in the rubble and anguish of hindsight. To put this in almost concrete terms: we have to make the point better than they did in the 1930's that cascading emergent market phenomena are real, dangerous, and recurring. Public policy must take them into account, even when we are not at all certain what the emergent properties are or the systems they will crash. One of the ways to see into the future is to examine the past and see what went wrong. In every case of failure there are "assumptions" that collapsed under the inexorable weight of facts. One of the meta-assumptions of the current financial disaster was that "because market forces are essentially rational, they are less likely to need regulation for irrational activity." This notion allowed sane (but greedy) men on Wall Street to pretend that packaging low-grade mortgages along with good ones somehow diluted the risk, if someone else decided that buying that risk could be profitable. Greed is not rational, so the meta-assumption was false at the outset. And, of course, there were many similar assumptions about markets and trends and regulations that proved to be false or inadequate to the possible (even foreseeable) loads that would be put on the system. What our Congress and White House must do now is to explicate the failures down to the assumptions that went into them and then devise failsafe mechanisms, many of which will clearly impede free-wheeling financial practices that have been practiced for a generation or more. There will be strong and powerful resistance, and since this is a political process, there will be compromises made in the structure. They will fail, inevitably. If you read Paul Krugman yesterday in the NYT you understand that compromise is already apparent in the Obama proposal. Psychological barriers like the magic trillion dollar threshold or worries over the total debt. Yes, these worries and in-for-dime-in-for-a-dollar thresholds are real, but they pale in the face of the larger problem, the emergent property of crashed markets (and circuit breakers) to seize up and remain seized up indefinitely. Re-establishing an electricity grid after a total blackout is an instructive task too. One does not just flick a switch and power on everything at once. It has to be done by stages that conform to the load limits of each part, otherwise it cascades back into failure again. The task our government has right now is exactly analogous to that of a major electrical grid provider knowing that the rest of the grid is going down. The job is to protect equipment from burning out, so that the job can be restored with minimal distruptions. There are no compromises on the electrical side and there should be few, if any, on the finance side. Here's hoping the Barney Franks and Mitch McConnells in Washington can get over themselves and do it right this time. The consequences if they do not will be dire indeed ... a cascading failure of the entire system, government included! JB

Authors Bio:

James R. Brett, Ph.D. taught Russian History before (and during) a long stint as an academic administrator in faculty research administration. His academic interests are the modern period of Russian History since Peter the Great, Chinese History, the history of science, and the history of ideas, including psychology and consciousness studies. He is retired and living on the Left Coast.