Tipping Point Model – An Appeal

  by Georges T. Roos
Director European Futurists Conference Lucerne
November 2010

--> Download as PDF here.
--> Call for papers.

Let’s admit it without being envious: Meteorologists have gotten a lot better at making weather forecasts over the past decades. Their predictions are more accurate, geographically differentiated, and above all more long-term. They can even issue weather alerts with good probabilities. No question about it: They have become more and more useful. But what of futures studies? We consider it disreputable to even talk about predictions. Counter to the expectations of our readers, listeners, and customers we always emphasize that exact predictions are impossible. Instead we diagnose the present and try to find patterns that we call trends or megatrends. And we collect signals that may indicate a change or confirm our trends. Apart from that we systematically describe possible futures based on modeled fundamental assumptions. I am by no means disparaging this approach. Those who listen to us are grateful for our expertise in opening new horizons to people.

But this is what really counts: The forecast of discontinuities
At the end of the day, trend forecasting will remain our main business. With a little practice in unconventional thinking, we can describe developments and their consequences for many trends or megatrends. Far be it from me to say that this is of no value. However, what would certainly be even more important to know for decision makers in economy, society, and politics is whether and why a development is no longer a trend but will fundamentally change its direction and/or momentum. It’s just as important to know when something will turn up that is  new, unexpected and that could lead to  economic or social paradigm shifts. We call these developments discontinuities or wild cards, and it is at the same time the most difficult and the most important task of those in charge to prepare for them. But to prepare one would have to know at least the outlines of those game-changers and wild cards. And not just that. One should also be able to place them on a time line - at least roughly. I am convinced that it is necessary for us as futures researchers to face this challenge head-on.

There are many good reasons to reject predictions.  But might it be that we are more capable of doing something in that direction than we let on? I met John Casti for the first time at the European Futurists Conference Lucerne 2006.  He is an American mathematician who now lives in Vienna, Austria. In his presentation "The Decline and Fall of Globalisation" Casti predicted that the boom would soon come to an end and give way to a deep recession. In addition, Casti was very concerned about the future of the EU and of the Euro. Think back: In 2006 we were in the midst of a long-lasting boom. His predictions were met with vehement skepticism by many of the futures researchers present. John Casti presented as basis for his forecasts  a theory heretofore unknown to me that he called Socionomics. It is based on the counterintuitive assumption that social moods are not the result of events but rather their cause. The mood of the stock brokers is therefore not negative because the markets dip dramatically, but the markets dip because of a pessimistic collective mood of the brokers. The father of Socionomics is Robert R. Prechter Jr. His theory is itself deeply connected with ElliottWaveTheory, developed by Ralph Nelson  Elliott (1871-1948) for the prediction of trends in the financial markets.  Yet, the social mood paradigm is not necessarily linked to the ElliottWaveTheory, as John Casti states.  Put simply, the social mood fluctuates between optimism and pessimism following a pattern, which may or may not be characterized by Elliott waves. Depending on whether the social mood is swinging one way or the other, certain types of social behavior and events are simply more likely than others (Casti, John L.: Mood Matters. Copernicus, New York. 2010).

I have to admit that I was astounded at Casti’s accuracy when the 2009 financial and economic crisis emerged. In the following year, one of our academic advisory board members, Prof. Markku Wilenius from Finland, started making arguments with another wave theory that is also supposed to show what is to come: He referenced the Kondratjev cycles, a theory of long-wave cyclical economic developments. Nikolai Kondratjev (1892-1938) was a Russian economist who was able to follow two and one-half such long waves at the time of the publication of his model in 1926. According to Kondratjev, the starting points for long waves are paradigm shifts and innovation-inducing investments in new technologies associated with them. This creates an upturn. A cycle lasts several decades and is initially accompanied by phases of mostly positive economic trends, and then towards the end by phases of recession. New cycles usually begin with (or after) deep economic crises. Joseph Schumpeter helped establish this model when he popularized the theory as Kondratjev-cycles. What interests Markku Wilenius and the current proponents of the theory of long waves is the question: Where in the wave are we currently situated and what will follow next? Markku noted in an internal paper that the most important characteristics for the end of a cycle are currently in place: The potential for basic innovations is exhausted; there is an excess of financial compared to physical capital; a period of severe recession has begun; and social and institutional transformations are taking place. According to Markku, all these indicators show that a new economic and social cycle is beginning.

Black Swans vs. Kondratjev Cycles and ElliottWaves
Both wave models fulfil an important objective: They identify discontinuities and make predictions. However, both models are also controversial. The appearance of "The Black Swan: The Impact of The Highly Improbable" by Nassim Nicholas Taleb (2007) claimed the impossibility of predicting rare, disruptive events. All important events in history, economy, and politics thus appear without warning. Even after-the-fact genealogy (as reverse forecasting) only gives the impression that it explains anything. Or as one historian from Lucerne put it: After-the-fact explanations are always oversimplified. In essence, Taleb says that nobody saw the really important events coming precisely because they are so unusual, rare and escape any notion of predictability. Taleb's argument is mostly epistemological: Humans have a blind spot for the immense role that rare events play.

Taleb’s book struck a responsive, contemporary nerve. The market value of the largest banks in the world had evaporated. Many had to be propped up by governments. Many millions lost a lot of money, and worse, their jobs. In the end, the crisis drove some countries over the edge and into the abyss. For many of the protagonists in charge of the preceding boom phase Taleb’s theses provided consolation and an excuse: Nobody could have seen this coming! However, others such as John Casti could argue   just the opposite. They had seen the crisis coming.. (Interestingly, Nikolai Konratjev  found confirmation for his model in the Great Depression.)

But predictions are much less convincing if they cannot identify the causal chains that lead to the disruptions – for the simple reason that they could have also been right by accident, or at least for the wrong reasons. In other words, the next prediction of an apparent guru or clairvoyant is as good and as bad as any roll of the dice.  In order to appreciate the predictions of John Casti, we have to follow him to the point to believe that today's known “causes” of the financial crisis were at best triggers but not causes. Thus the politically motivated, over-financing of the private real estate market was not the root of all evil; rather, it was the social mood that caused the US government to support private real estate ownership and devil-may-care attitude to risk.

Even Taleb does not claim that events arise out of a vacuum. He merely states that we cannot recognize their origin and thus are surprised when they occur. Are we really that helpless? Is early recognition or forecasting of disruptive events impossible, in principle? Or  can we sometimes see the point at which something fundamentally new will develop or a system will behave completely differently from before? In the search for the possibility of being able to make robust predictions that go beyond the current state of the art of my futurist brethren, I was struck by a headline in the January-February 2010 issue of "The Futurist", the journal of the World Future Society:  "The Science of Tipping Points – Researchers look for warning signals of major shifts" (The Science of „Tipping Points“. In: The Futurist, January-February 2010, p. 6).  The article summarized a report in "Nature" (Nature, Vol 461/3 September 2009) magazine in which European and American researchers from various disciplines described how complex, dynamic systems provided certain early warning signals before they changed their state. The systems described in the article "Early-warning signals for critical transitions" ranged from eco-systems to financial markets and climate. The researchers acknowledged that it was extremely difficult to predict critical tipping points. At the same time they felt that their belief was well-founded that there might be generic early warning signals for such state transitions that would apply to a wide variety of systems. Among these signals, the most important was what they termed system slow-down and fluctuations.

Tipping Points – Approaches to Early Warning
I sent the article to some of my colleagues in the advisory board, together with a plan to align our organization to conduct research on a theory of tipping points. The journalist and business consultant Malcolm Gladwell had popularized the term "tipping points", which confused my colleague at first. They thought I was talking about continuing the work of Gladwell who in 2000 published the bestseller "The tipping point – How Little Things Can Make A Big Difference". In essence, he stated that even small causes can have great consequences. While Gladwell became famous with the term "tipping point", he is not the creator of the term. The Nobel Prize winner Thomas Schelling used the term twenty years before in connection with his studies of racial segregation (in: Micromotives and Macrobehavior. 1978). According to Schelling’s observations, the "white flight" from neighborhoods that are being desegregated occurs when a tipping point is passed. Like the famous drop that makes the barrel run over, suddenly even those white people flee the neighborhood that generally spoke out in favor of desegregation and did not have any racist feelings. Schelling's argument is similar to earlier work  by Professor Morton M. Grodzins of the University Chicago, who in the same connection (white flight) spoke of tipping points already in 1958 (in: The Metropolitan Area as a Racial Problem). Gladwell's main contribution thus lies in that he took  the term out of  the context of race research and introduced it into general trend analysis studies and research.

Despite the initial skepticism of some of my colleagues, I continued my search for robust early recognition and early warning models and theories. I came across a study that Google had published together with the Center for Disease Control and Prevention, also in the journal “Nature" making use of the analysis of queries in the Google search engine. The researchers succeeded in identifying the spread of flu epidemics one to two weeks earlier than was possible using traditional methods. The authors reached the conclusion that that under certain conditions, an analysis of search queries or web mining can make possible recognition of the spread of a flu epidemic early on.

The HP lab in Palo Alto have also developed algorithms that allow very accurate predictions of box-office receipts of films based on Twitter entries. Peter A. Gloor at the Massachusetts Institute of Technology MIT does research in a similar direction. The increasing popularity of Twitter made it possible for Gloor to analyze positive and negative moods of the masses and compare them to stock market indicators. He discovered that after emotional expressions on Twitter (either hopes or fears) the stock market dips the following day. If the analysis shows that very little emotion is at play, stock prices climb the following day.

I am convinced that all over the world there is research on further models, theories, and approaches that deal with the possibility of robust predictions on various time scales regarding  a variety of questions. What would happen if we compared the most promising approaches side-by-side or even combined them? Could we with time perhaps even find an umbrella model that would help us identify areas that are approaching a tipping point based on these different approaches? Is it possible to combine this with approaches to early warning or early recognition signals, which would make it possible to place tipping points on a time line? These questions drive me forward in these days. There is no guarantee of success, but if this work were to  succeed we would clearly enhance the value of futures studies. I was able to convince the board of the European Futurists Club, being the carrier of the European Futurists Conference Lucerne, of the potential value of this idea. We then received seed money to invite researchers with the best forecasting approaches to a workshop in order to elucidate the possibilities of a tipping point model.

Call for Papers
For this reason we are launching a call for papers on our website and in the newsletter of the European Futurists Conference Lucerne. We are looking for rational prediction models, theories, and approaches. Based on the submitted papers, a committee of the advisory board of the European Futurists Conference will decide those to be invited to a three-day workshop in Switzerland in the spring of 2011. We will also directly invite researchers and scientists who have already published promising work in forecasting, early warning, or early recognition. The goal of this workshop will be to explore the possibilities of coming up with a theory of tipping points by combining various different approaches.

Our long-term goal is an international research institute for tipping points. The European Futurists Club cannot accomplish this using its own means alone. So based on the results of the workshops, we will look for foundations and other supporters that would fund our initiative. We are planning to use the European Futurists Conference Lucerne for the development of the tipping point model. We believe that a broad conference is a suitable forum to explore the idea of the research institute and discuss applicable areas for further exploration.

Will it be possible to move futures studies a significant step forward in this way? Currently, the meteorologists have a clear advantage over us. Let’s use them as an example of success and let them serve to motivate us to improve our own models and tools. The world depends on robust statements.