I have for some time found a common human behaviour very disturbing. Michael Shermer describes it in his book The Believing Brain:
Dr. Shermer also provides the neuroscience behind our beliefs. The brain is a belief engine. From sensory data flowing in through the senses the brain naturally begins to look for and find patterns, and then infuses those patterns with meaning. The first process Dr. Shermer calls patternicity: the tendency to find meaningful patterns in both meaningful and meaningless data. The second process he calls agenticity: the tendency to infuse patterns with meaning, intention, and agency.
This innate tendency to identify patterns and infuse meaning, which developed during our evolution, was critical to our intelligence. However, it may now be our downfall as a civilization.
The problem is that while patternicity and agenticity were effective with simple concepts, when faced with complex systems the tendency to see patterns in meaningless data or incomplete data becomes a very serious problem. We perceive a pattern and subsequently must determine the cause. But as a species we appear to have a tendency to jump to conclusions due to oversimplification. This can be seen in a number of issues in todays society.
My first example is environmentalist. I do not disagree with the idea that protecting the natural environment is a good idea. I even think that some of the advancements in industrial activity with regard to environmental protection over the last fifty years have been generally positive (with some exceptions, such as blanket bans on DDT and other pesticides). However, recent activity in environmentalism has rested very heavily on the problem of “patternicity and agenticity”, coupled with confirmation bias that only predetermined causes and effects are considered. Fourier and Arrhenius posited over a century ago that some gases in the atmosphere help trap heat. This was largely ignored for much of the 20th century. Much of the arguments against the linkage of CO2 in the atmosphere and temperature in the early days were that the model was overly simple and did not consider all the input factors.
The evidence in the modern instrumentation period already shows that there isn’t a simple linkage between atmospheric CO2 and temperature. The Mauna Loa CO2 measurement shows a monotonic increase, yet the temperature data shows warming to 1940, cooling from 1940-70, warming from 1970-98 and a static temperature since 1998. It seems obvious to me (and others) that there are numerous other factors – the most significant being the activity of the sun (whether it be solar irradiance or solar wind intensity and the effect on incident cosmic radiation).
The problem is that some people, many of them educated, observed a pattern in some data, performed an analysis that did not consider all of the inputs appropriately, or the analysis suffered from confirmation bias. And lay people accepted the results and politicians saw an opportunity to act, in an attempt to fix the perceived problem. Even though it is all based on an oversimplification.
The next example is economics, and the belief of many economists and politicians that the economy can be managed and targeted. Mises and Hayek clearly defined during the first half of the 20th century that the central planning of socialism was doomed due to a lack of information. It is impossible for anyone to have enough data to accurately direct the whole economy. A recent blog post covers this very nicely:
When SocGen’s Dylan Grice was asked if he was a fan of the idea of nominal GDP targets! He snapped he is not and thought it “a terrible idea”. As he opines, today’s various issues – the euro, China’s economy, over-indebtedness – are the cumulative unintended consequences of such past targets, and the naïve presumption that complexity can be commanded.Even mildly complex systems, any outcome is the wrong thing to target, with the processbeing where the focus should be. Expressing how little time he has for macroeconomics, reasoning that it’s obsessed with the targeting of interest rates, GDP, inflation, unemployment, exchange rates, et cetera, as though such a thing was possible without unintended consequences; Grice notes that Austrian economists understood this too. Ludwig von Mises distilled social phenomena to the simple observation that “man acts purposefully”.
And from Dylan Grice directly:
Today’s various issues – the euro, China’s economy, over-indebtedness – are the cumulative unintended consequences of such past targets, and the naïve presumption that complexity can be commanded.
All outcomes are caused by an underlying process.
But I’d argue that for even mildly complex systems, any outcome is the wrong thing to target. As we just saw, targeting one outcome of such a process changes that process, and changing the process subsequently changes all the other outcomes. In any kind of complex system where the underlying outcome generating processes aren’t well understood – whether a company, or a society – the effects of changing the process won’t be well understood either. Unintended consequences must ensue.
Yet even a cursory glance at the news shows ‘outcome targeting’ to be endemic: in response to the damage caused by Basle II, we’re given the ‘new and improved’ targets of Basle III (now already being traduced); the insurance industry now faces Solvency II targets; investors fret that banks won’t be able to hit their RoE targets; investors wonder if China will be able to hit its 8% GDP growth target; most major central banks target some sort of CPI inflation rate.
This is lunacy! How much damage has already been caused by banks that overreached themselves in trying to meet their RoE targets? How lopsided and capital destructive has China’s insistence on hitting its breakneck GDP growth targets at all costs been? How much of today’s painful credit deflation was caused by the credit inflation central banks pumped up while aiming for their CPI inflation target? In targeting these outcomes, underlying processes were distorted. Unforeseen outcomes resulted. But regulators continue to prescribe capital targets, banks continue to target RoE, China continues to target a growth rate, and central banks continue with ever more experimental methods in defence of their inflation targets. Indeed, today in Europe we’re seeing the unintended consequences of imposing outcomes (i.e. an exchange rate) on the eurozone economies.
The whole basis of attempting to target specific outcomes from such a complex system using blunt instruments like monetary and fiscal policy on a large scale is ridiculous. In my opinion, part of how we got here was through the improper targeting of inflation in the period following the collapse of the Soviet Union up to the financial crisis of 2008. The problem was that globalizing trade and rapid advancements in technology were inherently deflationary. Yet the central banks continued to try to manage inflation with interest rates, using a target of 2%. But in a deflationary environment, this resulted in loose monetary policy to drive up inflation. And it worked. Right up until the unintended consequence – which was that individuals, businesses and especially government took on way to much debt.
Similarly, the actions of the central banks since the collapse to use monetary policy to drive the economic recovery. The problem is that pouring new money into the system is inherently inflationary. The weak economy also masks the general inflationary nature, but the inflation will come. In some respects it already has. Much of the debt racked up over the past 20 years, and much of the new stimulus money has had little impact in the wider economy but has propped up banks and stock prices. Stock prices are still overvalued when using historical measures such as P/E ratio or return on equity. Stock prices rose exponentially since the early 1990s, with a small upset during the dot-com collapse.
Now, we see central banks and politicians trying to measure the overall economy with the stock market and unemployment rate. But targeting these will also result in bizarre consequences.
Public Health Care
Publicly funded healthcare is another area where attempts to target specific outcomes results in unintended consequences. All of the provinces in Canada are targets wait times in six key areas, including cardiac surgery, radiation therapy and orthopaedic surgeries such as hip and knee replacements. There are also other less well publicized targets evident in the system, such as emergency room (ER) wait times in Children’s Hospitals (high wait times for children make for angry and vocal parents…)
The problem with trying to manage something as complicated as a health care system with a small number of targets will inevitably result in unintended consequences. Focusing on many treatments that are primarily used by people over the age of 50 means that the system will focus on older patients, potentially to the detriment of younger patients. We hear news items about the problems of mental health treatment – because this is an area that doesn’t have “wait times” and isn’t a focus area.
The problem is that we aren’t focusing on the right things. This goes back to the problem of it being a complex system, like the economy, where the central planner lacks information. This is made worse by the failure to measure things in the health care system, such as how much individual treatments actually cost…
A better solution would be to set up the system such that individuals can make decisions on their merit for those individuals.
The solution to all of these problems is to train people to constructively utilize the patternicity without the inherent agenticity to which we are prone. Our public schools and universities should set this as part of the curriculum. It is coupled to teaching people to be inquisitive and skeptical of rapid conclusions. We need to teach the hazards of oversimplification and the impossibility of commanding a complex system. One of the things that early (Enlightenment era) universities did well was to establish this inquisitiveness. It is only in the 20th and 21st centuries that our centres of higher learning became corrupted by politics and causes. We need to cleanse the universities such that inquisitiveness, skepticism and independent thought are treasured. Can you imagine any of our universities generating someone who shakes the foundations of thought like Newton or Spinoza did? I have trouble seeing it.