Monday, 29 February 2016

Separated by a border (2): electoral divisions

In my first Separated by a border blog (that was three years ago, but it garnered significant attention apparently) I drew institutional implications of how some countries on the same geographical area differ so much with respect to their institutional environments, living standards, income etc. Inspired by Nogales, the infamous US-Mexican example from Acemoglu and Robinson's Why Nations Fail, I drew similar cross-country comparisons between the two Koreas, the Dominican Republic and Haiti, and between Rwanda and DR Congo. The common thing with all these examples is the same geography, same culture, same (similar) history, same diseases, but different outcomes. Why? Institutions, stupid!

Today we will look at the major differences within countries, starting with voting patterns. This time it's not so much the institutions that make the difference (at least not directly), but historical patterns of development. 

We start with the two most famous political science examples of electoral divisions: the Two Polands and the Two Ukraines.
Poland's electoral results in 2007 and the superimposed border of
Imperial Germany in 1914
The results of the 2007 Polish parliamentary elections exhibit a clear non-random pattern across an imaginative line of where the former Imperial Germany used to stand (the patterns are the same for the later presidential elections). The North-West part of the country is usually dominated by the PO party (Civil Platform), which is pro-European and pro-market (liberal conservatives), while the South-East part of the country (except for the capital Warsaw) tends to be dominated by the PIS party (Law and Justice; anti-communist, anti-EU, but social conservative, more politically right-wing but also more economically left-wing and protectionist). So the pro-EU, liberal party tends to win in areas that were once controlled by Germany, while the anti-EU, social conservative party tends to win in areas once controlled by Russia. Split between two large empires - that was Poland's destiny for a very long time. It shows even today. There is obviously no German or Russian influence in Polish votes, but it is very interesting to notice how events that ended 100 years ago still yield stark electoral differences. These electoral divisions are surely embedded in today's Poland, and have a very low chance of ever receding and changing. 

Ukraine is a similar story, even though it doesn't have a superimposed imaginative line of a historical division between two empires, but it does exhibit stark differences in cultural influence (e.g. the language division; the percentage of Russian speakers in the Eastern parts of Ukraine). There are also theories linking the electoral division to the 16th and 17th century rule of the Polish-Lithuanian Commonwealth over the Western parts of Ukraine. 
The results of the 2010 Ukrainian Presidential elections. Very similar
voting patterns were obvious in the 2004 Presidential elections. 
Again the difference is between the pro-EU (Western-orineted) part of the country, and the anti-EU (Russian-oriented part of the country). Much is today known about Ukraine's separation; the war and political deadlock that started in the beginning of 2014 are still no closer to being resolved. This is an example how stark electoral differences can even dissolve an entire country. 

An obvious example here is also Germany. However, since I've already discussed the electoral divisions in 2013 along the Berlin Wall, as well as the stark demographic and economic differences between the former East and West Germany, I'll leave Germany out in these sets of examples, and focus instead on a country with a long democratic tradition, but one that still exhibits stark electoral divisions, many of which can be attributed to history.

In the United States, the main electoral division is North vs South (sort of), with the underlying historical division primarily attributed to slavery and the 1861-1865 Civil War. Most of the seceding states from that time (except for Florida and Virginia), favored Republicans in recent elections (ever since Bush vs Gore), up to the point of becoming Republican strongholds (e.g. South Carolina, Georgia, Alabama, Mississippi, Louisiana, Tennessee, Arkansas, and Texas). However little do most people know that this electoral division is actually not that old, and more importantly, these states weren't Republican strongholds for at least a hundred years after the Civil war and Lincoln's Gettysburg Address.

Until the Civil Rights Movement in the 1960s these Southern states were actually Democratic strongholds. The reason? Many landowners who before the Civil War owned a lot of slave labor hardly favored the abolition of slavery. Ideology had nothing to do with it. It was always a monetary issue. Slave labor meant free labor, so being required to pay money to former slaves and give them a decent living standard was unacceptable for the landowners. However after losing the Civil War and accepting the Union, they still refused to treat African Americans, their former slaves, as equal citizens with equal rights. In fact black labor was still very cheap in the South, while the people were still living in extreme inequality, not to mention being subject to segregation and racist conduct, which was, at least in the South, completely legal!

Source: Allen Gathman
It is not therefore surprising to notice the pattern shown on the map above. The electoral results of the 2008 Obama vs McCain elections by county (blue for Obama, red for McCain), compared with cotton production in 1860, where dots represent the size of production (each dot 2000 bales). The pattern implies that African Americans, who were the crucial part of the cotton economy in the 19th century, even after a lot of them emigrated after the Civil War, still held the same residential patterns in the South, and carried the main votes in these areas for the nation's "first Black President". 

However the South remained dominantly Republican in those elections. But as I've said earlier, it wasn't always like this (for contemporary audiences which don't like to read history very much, I recommend the movie 'Lincoln' staring Oscar-winner Daniel Day Lewis - in the movie you can see how the Democratic party of Lincoln's time was extremely racist, and how it was the Republican party that was progressive). 

It was not until after Kennedy that there was a complete reversal of electoral patterns in the South (and the rest of the country). Consider the following set of maps, each featuring a national US election from 1952 to 1968: 

1952 Presidential election. Eisenhower (Rep) won. 

1956 Presidential election. Eisenhower (Rep) reelected. 
In these two elections you can clearly see how the Southern states that are today Republican strongholds voted for Democrats back in the 1950s. The pattern continued with Kennedy (map below; notice California and New England voting Republican all this time), even though he had "opposition" in the South from Harry Byrd, a Virginia Democrat and racial segregationist who wasn't even a candidate at these elections but still received electoral delegates in Alabama and Mississippi. So some voters in the South already started disliking Kennedy and what his administration would later bring.
1960 Presidential election. Kennedy (Dem) won. 
During the Kennedy administration is when the tides have completely changed (as you can see from the map of the 1964 elections below). Kennedy was strongly opposed to racial segregation in schools and how some Southern states have ignored the Supreme Court ruling from 1954 that segregation was unconstitutional. He promoted legislation that supported the Civil Rights Movement (the March on Washington happened while he was in office), the key domestic issue of his term in office, and, along with his brother, the Attorney General, was involved in preventing some major racist outbursts in Mississippi and Alabama (one of which included the 1968 third party candidate, segregationist George Wallace who won five Southern states in those elections). All this started the turn of the Southern states against the Democrats for good (there was an exception in the 1976 elections when they went to Carter, and in Clinton's elections in states adjacent to his home state of Arkansas). Democrats were still holding majorities in state legislatures in the South until the 1990s (some schools in Alabama remained segregationist until the 90s actually), but it was the events of the Civil Rights Movement during the Kennedy administration that started turning Democrats more liberal, and Republicans more conservative.
1964 Presidential election. Johnson (Dem) won. It is here
where we first notice the reversal of the traditional
Democratic and Republican states. The South turned
1968 Presidential election. Nixon (Rep) won. 

Wednesday, 24 February 2016

What I've been reading (vol. 3)

The third volume of "What I've been reading" is centered around bubbles, animal spirits, crises, and its predictors (real and self-proclaimed). I'll start chronologically:

Shiller, Robert (2005) Irrational Exuberance. Second edition. Princeton University Press

In the sequel of his highly acclaimed first edition that predicted the 2000/2001 crisis a few months before it happened, Shiller continues to debunk the rationality assumption in market behavior as he draws out a specific set of psychological, cultural and structural factors that can be held responsible for bubbles and wide-range volatility on financial markets. His price-earnings (PE) ratio (what eventually landed him a Nobel prize) as a method to predict when the market is due for correction is what got him the well-deserved critical acclaim. The PE ratio (pictured below, real time data available here) measures "how expensive the market is relative to an objective measure of the ability of corporations to earn profit". What Shiller did is that he compared 20-year annualized returns with the 10 year price-earnings (P/E) ratio using almost a 100 years of data and concluded that long term investors did well when prices were low relative to earnings. When P/E is high, risks are high, so one should lower his or her exposure in the market, but get into it when P/E is low. 
Rober Shiller's S&P 500 price-earnings ratio
Notice how the spike in the PE ratio in the year 2000 is exactly what led him to conclude that a market correction was inevitable. The dot-com bust that happened a few months after the book got published rose Shiller to prominence among the media and the general public. Interestingly enough, it seems that the correction that happened back in 2000 had its consequences postponed for a whole decade (possibly due to the Fed's response and its fueling of another bubble), when the PE ratio was still high despite the decline in 2001. But the problem then wasn't so much the PE ratio as was the housing market. 

The difference with the earlier edition is that this one includes a chapter on the housing market, where Shiller presents his home price index (the S&P/Case-Shiller Home Price Index, pictured below). It is in this chapter that he draws on his prediction on the inevitability of a correction on the housing market, just like he did 5 years earlier. Despite the media attention Shiller remained down to earth, which can be seen from his careful assessment in chapter 2: "It is difficult to judge whether the trend is building up or slowing down and when it might eventually stop completely and then reverse itself. Moreover if one extends a forecast to five or ten years out, one is likely to feel that one has virtually no idea where prices will be". It is obvious that Shiller is no Hedgehog when it comes to forecasting - he is a true Fox (see clarification below, in the third book review).  
S&P/Case-Shiller Home Price Index
The housing market index garnered a lot of (positive) attention, which is certainly what contributed to the success of the book as well. Even thought the book offers much more than that. It offers a complete pictures of what drives bubble behavior. The irrational exuberance, a quote borrowed from Alan Greenspan's 1996 speech, is what makes markets overvalued which leads to bubbles and finally bubble bursts and crises shocks. It includes precipitating factors like demographic trends or political decisions (there's 12 of them, surveyed in chapter 3) and all of its amplification mechanisms (such as fraud and a host of other factors that inflate speculative bubbles). Once the bubble starts it basically enters into a state of a reinforcing feedback loop where prices just keep going up. Until, of course, it bursts. The best part of the book is arguably part 4 which describes the underlying psychological factors that help fuel bubbles. Things like anchors, investor overconfidence, hubris and heard behavior, availability bias, there's a narrative fallacy in there as well (although he doesn't call it that), are all responsible for inflating the bubble and keeping it in expansion mode. I won't go too deep in explaining each of these, as I'll leave this for when I review Kahneman's Thinking Fast and Slow (I'm currently finishing it). 

The newest edition (third) of the book was published last year, and it includes a new section on the bond markets. I'm sure it's equally interesting as the stock market and housing market chapters. 

Akerlof, George and Shiller, Robert (2009) Animal Spirits. How Human Psychology Drives the Economy, and Why It Matters for Global Capitalism. Princeton University Press

In this hit book from the beginning of the financial crisis the authors attempt to attribute the crisis to human psychology, in particular the vibrant part of our psychology – our animal spirits (they borrow this exciting term from Keynes). Things like overconfidence, temptations, envy, illusions, biases, etc. tend to be the driving events behind all financial crises. Their effort to place an emphasis on the downsides of human psychology as opposed to the standard rationality assumption in economics is laudable and welcomed, so it's easy to see why the book garnered much critical acclaim back when it was published in the midst of the storm. 

However my main objection to their argument, even though I applaud the re-emphasis on Keynes’ long forgotten idea of animal spirits and the importance of psychology, is that these heuristics have been present, well, all the time. I understand and agree with the idea that the efficient market hypothesis is wrong, that economic agents are not rational (at least not all the time) and that they are prone to biases in judgement - this part isn't problematic at all. The problematic assumption laid out in this book is that animal spirits are responsible for turning the burst of the housing bubble into a recession. This is hardly the only reason. In fact, there are a number of factors that led to the accumulation of risk in the financial sector, psychology being just one of the underlying, indirect causes. Let's not forget the regulatory decisions that made the artificial demand particularly buoyant in the pre-crisis decade, or the CRA and GSE target-oriented policies. All of these were triggered to make the system better, more robust, and in fact more fair, but they've instead made it worse, increasing systemic risk. We can't say that the regulators were being led by heuristics when they've decided to increase the exposure of banks to, what was considered safe, assets. We can blame the house-flippers, the investment bank traders, the hedge fund managers, the rating agencies and the like for being greedy (again, this is something that's always present) and for acting irrationally, but the problem was systemic.
Which is why animal spirits in themselves are not enough of an explanation. Why haven’t they turned the 1987 stock market crash into a recession? Or the dot-com bubble? Or any bubble that subsequently hasn’t resulted in a painful recession? Or if they cause bubbles, they why aren’t they more frequent? The thing is that animal spirits don’t come and go, they are here to stay. We are constantly subject to the same biases, the same heuristics. Sometimes they drive us to good outcomes, sometimes they force us to make rash decisions with long-lasting consequences. Allowing the entire explanation on pure psychological errors which are present all the time is an understatement and oversimplification to say the least.

I understand the notion of a necessity of a Keynesian revival in lieu of the crisis, and I applaud the push of animal spirits back at the helm after they were marginalized during the neoclassical revolution of economic thought, but reintroducing them as one of the most important counterfactuals is one thing – overemphasizing them is something else – it is borderline platonification (vast oversimplification of the problem, most likely as an attempt to reach a broader audience).

And then there are the solutions – a Keynesian stimulus. Lately I vary from the austerity vs stimulus debate primarily since neither of the two can actually be proven to have worked, or more importantly neither can be disproven (falsified as Popper would say). Furthermore parts of the analysis seem to be subject to historicism, selection bias (not mentioning the 70s stagflation that was the key reason as to why there was a neo-classical revolution and abandoning Keynesianism in the first place), and a mistaken correlation for causality. In an effort to explain to a wider audience what happened and how the economy works the authors oversimplified, resorted to biases, unscientific reasoning (which is surprising given some of their earlier work I’ve read – e.g. Akerlof’s Lemon’s market is brilliant! As is Shiller’s P/E ratio theory, not to mention his Irrational Exuberance book, reviewed above), and have thus produced a book that, although promising in its attempt to bring back psychology at the fore, missed its target completely.

The paradoxical thing is, the best parts of the book are those that have little to do with economics. It’s not so much the analysis as the interpretation and faulty explanations of certain economic phenomena. And finally, the agenda of the book was to change the way we interpret and understand economic events. It has done no such thing. Perhaps it would seem that way to an average layman, but for an academic audience (I know, the academia wasn't their target in this case), it fails to live up to the hype.  

Overall, it was good to place a re-emphasis on psychology (Kahneman, Thaler, Taleb all do a better job of this however), but the conclusions are very biased. I’ve read much better books about the financial crisis, not to mention about 'how the economy works'. I had high expectations of the book (given the earlier work of both authors), perhaps that’s why I was a bit disappointed.

Roubini, Nouriel and Mihm, Stephen (2010) Crisis Economics. A Crash Course in the Future of Finance. The Penguin Press

"One man saw it coming. As far back as 2006, Professor Nouriel Roubini - aka 'Dr Doom' - warned that the US housing bubble was set to crash, and what would begin as a national disease would soon spread overseas resulting in a deep recession."

Well, as nice as all this sounds, and as true as his doomsday predictions turned out to be back then ("a once-in-a-lifetime housing bust, a brutal oil shock, sharply declining consumer confidence, and inevitably a deep recession"), I can't help at not being as impressed as much as I should (should I be impressed?). Particularly after having read Taleb and after learning to recognize (1) hindsight bias, (2) overconfidence in predictions, and (3) distinguishing luck from skills. I'm not saying Roubini is a bad forecaster, nor that he is not a good expert at what he does. He is certainly an expert in financial markets (I also enjoy and recommend his textbook on macroeconomic political cycles written with Alesina) and can make some very good predictions from time to time. It's just that I'm not as impressed with his 2006 predictions about the crisis after hearing him shockingly announce a whole set of other crises every year after the Great Recession. For example in 2011 he gave a doomsday prediction that stocks would fall by 20% that year. They didn't (on the same link check out the story from 2009 when he claimed that the US government would have to nationalize its banks as the market contraction would be too big). In 2012 he called another stock market crash by December. This one didn't materialize either. In 2014 he called the mother of all asset bubbles in 2016 (ok, we still have to wait and see whether this actually happens). Just google "Roubini crash 20xx", and see for yourself, there are many other examples of his doomsday, headline-grabbing and vastly pessimistic forecasts.

It seems that Dr Doom is a Hedgehog (see Tetlock's distinction between Foxes and Hedgehogs, two types of forecasters). A hedgehog forecaster sees one big thing, tip-of-the-nose perspective, and interprets all information he or she receives to fit the pattern of his one Big Idea. It seems that Roubini's one Big Idea is - crises. Capitalism is prone to crises. That pretty much sums it up. And another thing, hedgehogs never evaluate the precision of their forecasts. "Maybe it didn't happen now, but it's coming", is what a typical hedgehog would say. After it happens they will be the first to say "I've told you so!" And then comes the media attention that takes every next forecast as a sure thing. 

Now none of this has much to do with the book, I know. But I feel it is important to keep this in mind when evaluating the "crash course in the future of finance", as the book's subtitle reads. So, where to begin? I'll focus on the predictability part, and mainly disregard the oversimplifications, the dumbed-down writing style, as well as his "radical remedies".
After stating how Roubini correctly predicted the crisis, the book goes on to talk about other financial crises and the central idea seems to be that crises are both probable (true) and predictable (certainly not true). He (I'm deliberately emphasizing only Roubini as the book is in some passages written as an ego booster) emphasizes the fragility of financial systems. From this he draws a conclusion that crises are probable. The analysis, although simplified to fit the regular reader, is fine, however, it's the second big emphasis that's worrying for me. Roubini claims that crises are White Swans (as opposed to Taleb's Black Swan analogy) - predictable events. With this I can't agree. And not because I find Taleb's arguments more plausible and much more convincing, it's because I find an enormous hindsight bias in Roubini's analysis. He draws a lot of similarities of today's crisis to those in the past, and while there are similarities to be found, his cherrypicking of ex post predictable causes for each of these events is - naive at best. He does it as to show that he wasn't only correct in predicting this crisis, he would have been correct in predicting every single crisis from the tulip mania in 17th century Holland, to each and every banking crisis in 19th century USA, to the Great Depression, to the Great Recession. Hindsight bias at its best (worst?). 

He offers his own grand view of the global economy, a grand view that helps him formulate a framework for forecasting (a framework that he doesn't actually give to the reader - it is after all his comparative advantage over the rest of us). The grand view is that crises are "hardwierd into the capitalism genome", which must, by that logic, make them inherently predictable. Maybe not the date per se, but in general surely. The same logic can actually be applied to Marxists - they too say that capitalism is prone to crises and that its demise is certain. It's just a matter of time. And just like every Marxist rejoiced in 2008 claiming that this was the end they saw coming since 1876, Roubini too suffers from the same problem. "Economists have successfuly predicted 9 out of the last 5 recessions", Paul Samuelson once said. After having read Roubini's "grand theory", I can't help at feeling he too falls in the same category. On a positive note he does recognize the systemic imbalances that caused the current crisis (chapter 3, parts of chapter 4) as he correctly identifies a myriad of factors that each contributed to the spread of systemic risk. Whenever he focused on the specifics, it was informative, but whenever he drew generalities, it was unconvincing to say the least. 

To conclude, I've read a lot of financial crises books since 2008 (including Johnson, Rajan, Reinhart and Rogoff, Krugman, Rinholtz, Lewis - I actually prefer him to all the economists, Friedman, Kling, etc.). I'm sorry to say this one does not classify as one of the best.

Wednesday, 17 February 2016

What I've been reading (vol. 2)

The second volume of what I've been reading is all about, as I mentioned last time, the brilliant philosophy of Nassim Nicholas Taleb and a large part of his Incerto, a four volume "philosophical and practical essay on uncertainty". It contains Fooled by Randomness, Black Swan, Antifragile, and the Bed of Procrusets, plus there is another book on the way to make it five, Skin in the Game. I read the first three, each is brilliant, so I'll definitely find some time in the future to complete the Incerto. What is interesting about the three of his books I read so far is the re-occurrence of some topics in different scenaries. Already in his first book, Fooled by Randomness, Taleb introduces us to uncertainty, the role of luck, the Black Swan problem, his skin in the game philosophy (to have skin in the game means being able to put your money where your mouth is), and most importantly his underlying life philosophy of an empirical skeptic. Empirical skepticism and outright popperianism take the center stage in all of his books. Perhaps that's why I liked them so much. That's the thing with Taleb, you either love him or you hate him. I'm guessing he's deliberately doing that. 

Taleb, Nassim Nicholas (2004) Fooled by Randomness. The Hidden Role of Chance in Life and in the Markets. Penguin Books

How do we define success? Is it down to skills, ability, and strategy, or is it down to something much more unpredictable – pure luck!?

The book is about how we tend to misinterpret our perception of luck and randomness into something tangible such as skills and knowledge. Taleb uses an autobiographical element of his career as a Wall Street trader to unveil the hoax that successful and rich traders are successful due to their inherent ability, competence, intelligence and/or skills. This kind of simplistic causal thinking is what it means to be fooled by randomness. People genuinely have a lack of understanding of probability and uncertainty, making them believe that events are non-random, whereas in most cases they are really triggered purely by chance.

The biggest mistake we make in not applying a probabilistic reasoning is not taking into consideration the whole set of alternative outcomes. What would have happened if an event carried on differently than it actually did? There are many fallacies arising from this. For example, books on the successes of millionaires usually focus on a particular string of outcomes, as if there is a particular embedded reason as to why someone succeeded in life. So they inform their readers of the optimal formula, the necessary ingredients, to become successful. Work long hours, read and educate yourself, get to know the ‘right’ people, defer your consumption decisions, don’t live an extravagant lifestyle, apply a particular way of thinking about problems, etc. The basic problem with this kind of thinking is that it falls under selection bias (the law of selection – pick a specific, non-representative sample and make general assumptions of the population), or as Taleb calls it survivorship bias. When observing how the rich got rich, we tend to look at a very limited sample. More precisely, we only look at the winners. There is no mention of those who did exactly the same, who applied the exact same strategies (a lot of people work hard, come home late, know the ‘right’ people) but failed. There is no way to test which of these characteristics actually do hold true – whether there is a “success gene” that makes the successful, in any field, – successful. In order to do that we need to assemble a huge sample of winners and losers to see what distinguished the winners from the losers. Taleb has no doubt – it was luck!

“Mild success can be explainable by skills and labor. Wild success is attributable to variance.”

One of the strongest points of the book is how randomness fools us into thinking that certain outcomes we tend to observe are down to skills. In the prologue there is a Table of Confusion where Taleb precisely identifies the central distinctions and confusions later analyzed in the book. These include confusing luck with skill, probability with certainty, belief with knowledge, theory with reality, anecdote(coincidence) with causality, forecast with prophecy, a lucky idiot for a skilled investor, survivorship bias for market outperformance, volatility for return, noise for signal (!), etc.

The problem with randomness (variance) is that no one attributes it to his or her successes, but only to failures. Intelligence, know-how and skills are responsible for success, bad luck for failure. However, this purely depends on the type of profession. In many professions luck plays a minor role, and skills and knowledge dominate. Surgeons, dentists, car mechanics, plumbers, etc. all have a non-random element determining their performance – the level of their own practice, i.e. skill. This of course doesn’t mean that occupations like business CEOs, university professors, or scientists have their fates purely in the hands of chance. On the contrary. Each person’s skills are important. But the extent to which they’re important depends primarily on the volatility of the person’s job/environment. In other words how exposed is he to randomness. The lower the exposure, the more role is attributed to skills, and less to luck.

The book is therefore a defense of science, and an attack of scientists straying from their course (lack of self-critisism and incapability of dealing with randomness). It battles against heuristics and warns of the dangers of simplification. It discusses visible and invisible histories of rare events, probability biases, and what it takes to become an empirical skeptic. The point Taleb is making is that he too can be fooled by randomness, only he has the audacity to accept this fact. There are tricks to deal with this and to try and avoid being fooled by randomness too often. One is to constantly revisit your own opinions, i.e. update your prior beliefs (Bayesian updating). Avoid being “married to your position” (trading lingo). In light of new evidence the least you can do is to update your beliefs.

Taleb, Nassim Nicholas (2007) Black Swan. The Impact of the Highly Improbable. Penguin Books.

In what is very much a sequel to Fooled by Randomness, Taleb further exposes and details his philosophy of an empirical skeptic. The emphasis is once again on the faulty of human perception, however this time it is related to the interpretation and perception of really rare events, the so-called Black Swans.

A Black Swan is an event that has three distinct characteristics: it is completely unpredictable; it bears long-lasting consequences that make the world as it is (it has a high impact); and it has a retrospective explaninability (after it happens people tend to rationalize it and fool themselves into thinking that it, in fact, was predictable - typical example of hindsight bias). Taleb teaches us not to worry too much about predictions and how to be good at them, but to embrace uncertainty instead. The book is an excellent philosophical and psychological treatise of human rationality, the limits and overestimation of our knowledge, our primal heuristics driven by Platonicity, and our chronic lack of abstract thinking. 

The Black Swan logic teaches us that things we don’t know are far more relevant than things we do know. Black Swans are unexpected; if they weren’t we would have prevented them. Since we are generally blind to randomness and large deviations we fail to recognize the importance of Black Swan events in shaping the way the world around us looks (just ask yourself how many things in your life came as planned, and how many were a result of an abstract shock, purely random events that steered you into one direction or another – and don’t fool yourself into thinking you planned to have everything exactly as it is; if you belong to this category, you suffer from some serious hindsight bias).

In fact, no modern day breakthrough (be it technological, scientific, economic or political) came as a result of careful design and planning. It was all trial and error, where some realities turned out to be more (or less) likely than others and have manifested themselves into the social order we have today (think of any great historical event – the world after WWII e.g. – or any technological innovation that has affected how we live today). Markets also work thanks to trial and error. So does democracy in my opinion (this is further developed in Taleb’s next book: Antifragile. My own take on this was the success of democracies over autocracies due to their trial and error process, or what Taleb accounts to their antifragility).

Whenever we observe a Black Swan, apart from fooling ourselves into thinking that it was rather predictable, we also tend to learn from the specifics of the event, not the general existence of such events. For example, after 9/11 people learned the threat of terrorist attacks through airplanes, which is why the reaction was to make air force security tight, in addition to creating a Big Brother society to make us all feel safe. It was a typical example of learning from a very specific case. It is the same way that regulators tend to react to financial crises. They react ex-post, without realizing that the next financial crisis, when it happens, which it surely will, won’t necessarily affect the system through the same mechanisms. Regulating the mortgage market and its financial derivatives (MBSs, CDOs) after the 2008/09 crisis is an expected response, but the next crisis is very unlikely to come from the same financial instruments. It is likely to arise as a consequence of something completely different. Just like the 2008/09 crisis came from the super-safe mortgage market, that “no one saw coming”, the next big crisis is very likely to be a consequence of something else. The reason for this is that we tend to learn from history, and the way it has unfolded, without considering the alternative scenarios. Combined with hindsight bias, in that we should have seen it coming as it was so obvious (!), we tend to draw the wrong implications from historical events. And in that perspective we tend to derive wrong conclusions and wrong solutions.

The bottom line is that in having these types of reactions is not strengthening the system – on the contrary, the more complex the system (of regulations, rules, laws, etc.) the more likely for it to combust. It is still vulnerable to Black Swans, since we never change our way of thinking about them and we never change our way of thinking about uncertainty. We don’t learn “rules” and the general idea of risk and uncertainty, we only learn the facts. We get preoccupied with anecdotes and the interesting and specific cases and stories. We run away from the abstract, since anything that cannot be explained is scary. This is undoubtedly an artifact of our primitive behavioral trails, our evolutionary fear from the unknown, the uncertain. We don’t think as much as we should. We simply react. This too is deeply embedded in our gene pool.

In addition to all of that the book features a brilliant part on our failures of prediction, and the great scam that is the bell curve. 

Taleb, Nassim Nicholas (2012) Antifragile. Things That Gain from Disorder. Penguin Books.

Just like the Black Swan was a sequel to Fooled by Randomness, so is Antifragile a sequal to the Black Swan. Many ideas touched upon in the first two books are further developed in this one. Essentially in Antifragile Taleb offers to solve problems raised in the Black Swan. It is, in his own words (and I agree) his best work! 

The underlying reasoning is a split of the world and everything in it into three distinct categories: fragile, robust, and antifragile. Fragile "things" (which includes people as well as businesses, countries or institutions) avoid disruption from fear of change. By doing so it offers a false sense of security and is making the system even more vulnerable to a shock. The longer it's postponed, the worse its consequences. Robust means that one can bear the burden of the shocks without implying the necessity to change one's behavior. But the best of all three is to be - antifragile. This means that shocks, adversities, and disruptions make you stronger (what doesn't kill you...), you adapt to them and change, improve. In other words this is typical trial and error behavior, you make mistakes, get hit by disruptions, but you gain from that and become more robust to future shocks. You become better by falling down and learning how to pick yourself up again. Only when we're antifragile can we actually avoid being surprised and congested by the Black Swan. That's what the subtitle "gaining from disorder" stands for. Uncertainty here is desirable. We should be well aware of it, embrace it and, more importantly, we should design our institutions to support uncertainty, to become antifragile to it. In fact we should all strive to be antifragile.  

The best way to illustrate this is with the package example used in the beginning of the book. If you pack glasses in a box, even very carefully (pseudostability), they would still be sensitive to being dropped and thus broken, since they are, in their essence, fragile. On the other hand if you pack a steel cube in the same box you can drop it around all you want, the cube won't break. It is robust to disturbances. However an even better alternative is to pack something antifragile. This is an artifact which would become stronger from every drop and every disturbance. In real life Taleb recognizes evolution as the prime example of antifragility. Whenever there was a stressful event evolution made us stronger by adapting. A similar example is entrepreneurial ventures which strive on trial and error. Something that is defined to be anti-fragile, the complete opposite of fragile, is something that benefits from shocks, from volatility, from randomness and disorder, from risk and uncertainty.
When faced with a shock, a fragile thing will break, but an antifragile thing gets better. 

What does a system that benefits from errors look like? Consider the airline industry – each plane crash makes a system safer as we learn what took one plane down and make sure this error is never repeated. The reason why this is possible is because errors are uncorrelated (i.e. a fall of one plane does not cause all the others to go down as well). The economy is different – a major shock in one industry does transfer very quickly onto others (e.g. housing, finance, etc.). This is because in fragile systems errors are compounding and spreading. So while every plane crash makes the next plane less likely to fall, every bank crash makes the next one more likely. 

A few helpful examples of the triad - fragile, robust, antifragile: the mythical creature Hydra is antifragile, while a Phoenix is robust. The banking system is fragile, while the Silicon Valley (characterized by entrepreneurship & innovation) is antifragile. Fragile systems hate mistakes, antifragile systems love them and learn based on trial & error. Debt is fragile, VC capital is antifragile, centralized systems are fragile, decentralized are antifragile, etc. 

Antifragility is the property of all systems that have survived, meaning that depriving them of stressors will hurt them and make them more exposed to hidden risks. By surpressing randomness and volatility we are fragilizing the system (such as increasing moral hazard). Such changes usually come top-down and block antifragility, therefore constraining growth. Changes need to arrive bottom-up through the process of trial and error and discovery. 

Taleb is very sensitive to what he calls fragilistas, people who make us engage into artificial policies and naïve interventions which carry small benefits to society (and potentially large to themselves), however with potential negative side effects to the system which could be huge. This is how systemic risk is created: building up hidden risks by preventing fragile systems to change and blow up before they become too-big-to-fail. The role of the fragilista is to lend credibility to such a systemic error through academic and policy-based advocacy. Systems need to be open to small but vital stressors, small errors which we learn from and improve. When a system is juxtaposed this way it will become antifragile and grow stronger with each shock. Fragilistas (policy-makers, economists, foreign policy hawks, politicians, bureaucrats, bankers, and many many more) actively prevent that. 

One of the most interesting concepts he develops is skin in the game as a way to understand ethics (I'm happy to see there's a new book carrying this exact title). Skin in the game means assigning direct exposure to one's opinions/predictions/goals, so that if you're mistaken you should lose money (or in another way be exposed to the negative event). He stresses out that the lack of skin in the game is one of the major causes of fragility. This is equally applicable to pundits/experts ("empty suits") unaffected by their predictions, managers (CEOs) unaffected by how their company performs or what impact it has on society (e.g. the environment), or politicians personally unaffected by the policies they implement. In other words “learn from people who do what they teach! Ignore others” (this is from Black Swan actually, but I feel it fits more here). 
"To me, every opinion maker needs to have “skin in the game” in the event of harm caused by reliance on his information or opinion ... Further, anyone producing a forecast or making an economic analysis needs to have something to lose from it, given that others rely on those forecasts..."
In addition to all of this the book also offers some very deep and interesting thoughts about debt, innovation, health (especially), politics, foreign policy and wars, urban planning and city states, personal finance, etc. And all of that done through the same framework of popperianism and empirical skepticism. Like all his books, it is fascinating to read as Taleb jumps from topic to topic, from anecdotes to analyses, from order to disorder (sic!). In my opinion, this specific style is exactly what makes him such a compelling author. If you have to read just one of Taleb's books (even though they're all brilliant), read this one. 

Saturday, 13 February 2016

Graphs of the week: Tax evasion and votes for Syriza

Over the past year many commentators on the Greek political situation have claimed that Syriza, the radical left party that surprisingly won the Greek elections in 2015 (twice!), drew much of its support from tax-evasion areas. In other words the tax-evading upper-middle class voted for Syriza as they wanted to continue with their tax evasion (an endemic problem in Greece) and didn't want to accept the EU-imposed bailout deal which necessitates that Greek authorities clamp down on tax evasion. 

In a long overdue post I will briefly look into whether or not there is any evidence to back this claim. Below I show two sets of maps. The upper map shows results of the 2015 general elections held in January (right), and the bailout referendum (left) held in July, while the lower map shows the share of tax evasion per municipality (I've used this map before from a paper by Artavanis et al.). 

Before we make the visual comparison, a quick reminder on the tax evasion part from my earlier post
"...Their main finding is that contrary to popular belief that only the super-rich are avoiding to pay their taxes, it is actually the middle classes (the upper middle classes to be precise) which are dodging their taxes. This makes a lot of sense, as 30bn euros is too much to link it only to the super rich. The problem must be more widespread. And it is. Basically across the Greek society. In the map ... you can see that tax evasion is almost evenly distributed geographically...
The authors note that the primary tax evaders in Greece are doctors, engineers, private tutors, accountants, financial service agents and lawyers ... "Testing the industry distribution against a number of redistribution and incentive theories, our evidence suggests that industries with low paper trail and industries supported by parliamentarians have more tax evasion."
Going back to the 'tax-evaders voted for Syriza' argument. Let's compare the two maps visually (this is how such arguments are usually formed). If we look at, for example the island of Crete and the greater Athens region, the story clearly holds (even as a simple correlation). Even in the circled area (Larissa, an area that has the largest number of Porsche Cayennes in Europe), the voters supported Syriza and rejected the bailout deal. However, one can look at the map the other way around - in the Peloponnese peninsula, where tax evasion also tends to be high, Syriza didn't do so well, nor was the bailout deal rejected. Also I can easily spot a number of areas where tax evasion is low, and the votes for Syriza and against the bailout were high. This still doesn't prove anything.

Producing arguments based on a simple visual comparison of maps like these gets you nowhere. This is pure cherry-picking. You see what you want to see. The best way to actually test this is to take the tax-evasion dataset produced in the paper by Artavanis et al. and regress it against both election results, in addition to a number of country-specific controls (demographic, socio-economic, political). Using surveys is a bit harder to uncover this relationship as no one in their right mind will willingly confess to be a tax evader (even when people are arrested for tax evasion they still very often find ways to justify themselves by saying they didn't actually broke any law). 

To me the only thing obvious from the comparison of these two maps is that tax-evaders have NO causal effect on votes for Syriza or the votes for or against the bailout. As I've noticed in the earlier post, the tax evaders are evenly distributed geographically. Votes for Syriza not as much. The whole country has a problem of tax evasion (perhaps the paper even underestimated the total amount by imperfect measurement - it was an estimate after all (a good estimate but an estimate nonetheless), there's no actual data to measure this), so linking this to voting patterns in naive without a serious, in-depth analysis of actual voter choices in Greece.