Using History to Improve Forecasting

“There is no better teacher than history in determining the future.”
– Charlie Munger

To better predict the world’s future, one must understand it’s present. To better understand its present, one must understand its past. Understanding the past is best done by history books. In Susan Wise Bauer’s History of the Ancient World, we learn of environmental problems causing grain yields to decrease throughout the Fertile Crescent. Decreasing grain yields caused social schisms. Social schisms caused uprisings, which developed into civil war. In the modern Middle East, rising grain prices led to…social schisms and civil war.

History does not always repeat itself. There are differences between the ancient environmental problems and those that led to the Arab Spring and the Syrian Civil War. However, history provides certain patterns that make for useful knowledge. If you can identify how modern events are similar and different from past events, prediction becomes easier. Nobody can become a Nostradamus from these techniques. Complete accuracy is not possible, nor should it be the goal. Rather, one should look at long chains of cause and effect, and attempt to estimate where the trends are going, as well as the changes in the trends. A developing economy does not have high growth rates forever. The trend must stop at some point. History books tell us an enormous amount about the chains of cause and effect that led to the present.

As useful as this information is, I have not met a single person who reads history books. After trying to find good history books to learn from, I can see why. Most are terribly written, and terribly specific. Biographies of Napoleon cannot get anyone to understand the cause and effect patterns of the world. Details can be important, but knowing about World War I will only help identify issues similar to those that caused World War I. Much of the time, these problems will not play out as expected. If you see current events from the lens of WWI history, you will always assume that the world is on the brink of destruction and war. To avoid getting lost in the details, we must focus on general histories.

The other requirement is to have a multicultural understanding of history. Some readers may find multiculturalism a worthy goal in and of itself. Some may feel the opposite way. I have no cultural preference as such. I’m not trying to learn multiculturalism because of any political upbringing. Remember, the goal is to forecast more accurately. My liberal readers will be willing to learn multiculturally for no utilitarian reason. Conservatives may not see the usefulness of the multicultural approach.

Skeptics may want to know what is possibly the most important example of multicultural learning. A French translation of Sun Tzu’s Art of War became available in 1772. A minor French officer began to read it. His name was Napoleon, and he used Sun Tzu’s tactics to create an Empire. Be like Napoleon. Learn multiculturally.

Multicultural history is also important for the old adage of an increasingly intersecting world. Culture clash and transnational movements will both take place. Political maneuvering and war between different nations will occur on a scale never seen before. This makes the history of Africa important even to a Chinese person, and the history of China more relevant to a European.

Besides giving a general overview, and providing histories from different parts of the world, one’s history bookshelf should contain books that are easy to read. Reading Toynbee is not something most people do for fun. Good history books are clear to understand and easy to read. The clearer the better. You can read Hegel, Toynbee, and Spengler after you’ve gotten the basics. Philosophy of history books may not actually help with prediction anyways.

If you learn history through reading books that are both general and multicultural, you are (hopefully) able to make better predictions with the time-tested mental models they provide. There is no way for you to do that if you’re not committed to finding and actually reading through these books. I’m not sure how well this actually works, but I have an account at the Good Judgment website, so hopefully I will be able to update this every so often to see my prediction scores.


Understanding the Red Tribe with Music

I used to think I understood American culture. How wrong I was! I was familiar with TV shows of all genres, I listened to different podcasts, and, most of all, I listened to different kinds of music. After some time, I came to the conclusion that I was really just following Blue Tribe media. There was nothing Red Tribe-related in my knowledge. Understanding a culture’s music is a good proxy of understanding the culture itself. During the Rise Of Donald Trump, people on my side of the now-apparent cultural spectrum began to speculate that Trump’s sway in politics was due to cultural, not economic, factors.

So I sought out to understand the culture Trump’s supporters lived in. The musical angle hit me completely by mistake. Before that, I tried to learn about Red Tribe culture by reading books about the modern-day troubles of “poor whites in America”. That didn’t work. I then researched the Civil War era, which helped even less. (I don’t know what I expected.) So I took the next rational step: giving up and going on Reddit.

While browsing one of my favorite subreddits, Let’s Talk Music, I saw a post about country music. Realizing I’ve listened to almost no country music ever, I tried it out. Despite my broad tastes, I have skipped a genre outside the Blue Tribe domain. I didn’t expect too much from country, so Jason Isbell’s Something More Than Free caught me off guard. Not only was it Red Tribe, but it was good! A story of a manual laborer who… Actually, just listen to it. Any of my explanations will sound boring. Try Genius if you want something better.

Like the Blues, the Red Tribe has different people of different viewpoints. Isbell does not speak for his culture any more than I speak for mine. Regardless, it seems to provide a shard of perspective on the cultural beliefs and attitudes of Reds. As a bonus, it even sheds some light on the ground-level view of rural America’s economy.

Most of my posts have some type of intellectual understanding of something. Culture, as far as I know, does not work like that. This is usually the part in the blog post where I try to refactor something so it makes more sense. But most of us Blue Tribe people can’t even see what Red culture looks like from the inside. Before any refactoring is possible, we have to try factoring. Music is a glimpse into this alternate dimension. The real fun is figuring out how to deal with the other side of the looking glass once you’ve seen it, but that will have to come later.

Nanotech Terrorists and Data Proliferation

The Wrong Solution

Nanotech is still far from being dangerous, but solutions to proliferation should be thought of now, rather than later. Currently, the U.S. government must approve of transfers of information that could have military use. This law is broad, far reaching, and inefficient. This doesn’t just apply to nanotech, but other technologies that could have potential military applications as well. New biotechnology, innovative drone blueprints, and anything else that could be used by the military falls under the purview.

The law seems tyrannical. It could prevent useful research from getting to places that need it. The military will err on the side of preventing information from getting through. If someone dies because they didn’t have access to biomedical research, nobody will blame the military. If they are killed by a U.S. military-designed weapon, someone could get fired. Overconservative policy would prevent useful international research, and stifle innovation in the long run. At least this would be the case if the law worked. It does not.

There is no decent enforcement mechanism-nothing to prevent actual data proliferation. Trying to prevent terrorists or other nations from getting this data by preventing information from leaving the country is like trying to prevent a computer hack by banning others from reviewing your source code. When you get hacked anyways, you’ll have no idea what hit you. Here’s Jeffrey Matsuura, in the book Nanotechnology Regulation and Policy Worldwide:

Indeed, a great deal of the information and technology associated with nanoscience that has already been transferred to individuals outside of the United States has most likely been transferred in violation of U.S. technology export control laws.

Stopping information flow is nearly impossible. Government defense policy should assume that enemies already have the information required to create most weapons – they do. The real stopgap is the lack of materials or skills to actually produce anything destructive. Nanotechnology will be different. If the enemy knows how to create these weapons, they will. The specific implications of nanotech are unknowns at this point, but we can look at 3D-printed guns for a fuzzy picture of the future. Anyone can create a weapon. Not a good weapon (yet), but improvements will be made, and useful ideas will spread.


The Less-Wrong Solution?

Information proliferation is one of the only scenarios where prevention doesn’t work; you need cure. Preventing access to information may be impossible, so administrations should approach terrorism directly. The focus should be on the physical acts of terrorism themselves, not the spread of information that could lead to terrorist acts. Cutting off physical access to raw materials is important. Cutting off virtual access to digital information is inconceivable. To move towards physical security rather than virtual security, defense agencies must focus on physical surveillance rather than virtual surveillance. Virtual surveillance is difficult, resource intensive, and has never worked. According to Bruce Schneier:

Whenever we learn about an NSA success, it invariably comes from targeted surveillance rather than from mass surveillance. One analysis showed that the FBI identifies potential terrorist plots from reports of suspicious activity, reports of plots, and investigations of other, unrelated, crimes.

He spends most of his time berating mass surveillance, but Schneier also mentions the success of targeted methods. Preventing the flow of data doesn’t work, but checking out high risk individuals may help. This leads us to another puzzle. How should sort through the pool of all humans to target high risk individuals without resorting to mass surveillance. One way is to target individuals that we have heard are dangerous. This is obviously insufficient, but it is a start. Umar AbdulMutallab, who was behind the failed Christmas terror attack, initially got on the authorities radar when his father sent warnings about him.

Authorities effectively ignored these warnings. Perhaps they didn’t consider him a threat for valid reasons. Perhaps they received many such warnings, most of which turn out to be false positives. In any case, they would have been better off spending more resources towards trying to sort through warnings than trying to sort through the entire internet.


Warnings come prefiltered by the warners. The internet and phone lines could have anything, but going by warnings can serve to shrink the pool by a significant amount. Some terrorists will fall through the cracks, but not as many as are falling though the cracks of mass surveillance. Digital surveillance in general may become less easy, but targeting a small number of high-risk individuals is simpler than targeting everyone all at once. For a more robust antiterror apparatus, I recommend improving physical surveillance. There is no magic bullet to prevent all attacks. There are only underutilized people.

The Problem With Knee-Jerk Coordination Against Moloch

If society responds to its Moloch-related issues by coordinating with each other, they may be able to kill Moloch. If someone overfishes/pollutes/etc. all other members take the offender’s money and distribute it towards fixing the problem. Or they kill the offender. Or they put the offender in jail. If everyone coordinates to have rules to end Moloch-inducing behaviors, we will have killed Moloch. But even if these rules are successfully enforced, we may not have ended our problems.

For those of you that are unfamiliar with the idea of Moloch, here’s the overfishing example:

Let’s say we have a bunch of fishers in a lake. One of them overfishes. The fish population won’t sustain itself at this rate. Usually, the rest of them would begin to fish more and more, afraid that the fish will run out. If they want to get any fish at all, they better start fishing as much as possible. This accelerates the process of overfishing. Everyone does what’s best for themselves, which ironically makes the entire group worse off. This is an oft-cited example of the coordination problem. The blog Slate Star Codex calls it Moloch, after an evil demon from antiquity. Because coordination problems are the closest thing to an evil demon the world has faced, I’ll call them Moloch as well.

To prevent this from happening, imagine the fishers all agree to certain rules. If someone overfishes, the others kill the offender and distribute their stuff amongst themselves. What have they solved? The next person to think of overfishing would be dissuaded. Sounds good, but this has problems as well.

Let’s say that society has anti-Moloch rules in place, but something else changes. Say the fish population adapts to outswim the birds that eat them, and this spreads quickly. They are so ahead of the birds that the predators die out altogether. This would be great; the people have more fish for themselves. They can catch 50 fish a day instead of 40. But the social institutions are solid; they won’t just change because a few people noticed there’s more fish available. Even if everybody noticed that there were more fish in the sea, you couldn’t do anything about it, because the laws are set in stone. If someone fishes again, there’s a good chance they will still be killed. This problem becomes worse the better societies are at fighting Moloch.

An amazingly well-organized (in terms of fighting Moloch) society would be terrible at adapting to changes. The most well-coordinated societies would have static laws. These static systems of actions constrain them. If Moloch becomes less powerful (e.g. more fish are available) we should be able to adapt and update our beliefs and our political coordination structure. Before updating is possible, we need to be able to measure how powerful Moloch is for any given coordination problem.

At any given time, we should know what’s going on with the fish. How many there are, how big they are, as well as if there is anything that could change the population. If something changes the amount of fish we get, we should know about it. Even if something changes the quality of the fish, we should know. Ideally, we should be able to predict the amount of fish as well as meteorologists predict the weather. From there, the rules about how many fish can be caught is dynamic. Each week or each month, fishers should be able to catch the updated amount for that time period.

If there is a change in fish population, then a change in fishing is needed. This becomes more complicated when there are two problems which constrain our actions. If the fish population decreases, we must decrease the amount we catch. But let’s say there is a crop failure at the same time, and we need more fish to make up for the lack of grain. This is where it gets interesting. In today’s world, this would be solved by economics: people would fish more until the next crops came, and then catch less when demand for fish decreased back to normal levels. As long as the famine was temporary, this should allow the fish population to recover. The action that results in fewer people dying is usually taken through the market system. However, with our policy of dynamic changes in fishing based on the changes of fish supply, we can’t solve this problem. Assuming fish supply was stable for this season, a terrible potato famine could cause a huge spike in global fish demand. People should overfish somewhat in the short run, and fish at lower than equilibrium levels when the problem subsides. Hopefully, there won’t be another famine during the fish population’s recovery.

So the amount we fish will have to be dynamic with respect to both fish supply and fish demand. What is the difference between this policy and the current market economy? The new policy would be coordinated in a way to minimize starvation, including situations where Moloch appears. With an uncoordinated free market, we will have Moloch, but as long as we are coordinated, we can fight Moloch to a great degree. This does not just require a legal system that solves current problems, but one that predicts and accounts for other issues as well. There are always constraints placed by the supply and demand curves, and we still have an economy, but it is somewhat resistant to Moloch / multipolar traps / tragedy of the commons.

The other requirement is a constantly updating legal system, that is fast enough to create laws that deal with sudden problems and data-driven enough to understand their causes. Not only that, but the laws themselves need to be updated to face new threats. When the law only focused on fish population, it stopped being effective when crop failure became an issue. The legal system must be able to factor in new threats as quickly as possible, lest the society starve itself of both crops and fish.

More Privacy Online, Less Privacy In Public

Most people believe that there are high levels of privacy in the real world, unlike the online world. Many understand that CCTV cameras and the like diminish real world privacy somewhat, but not to a great degree. The recordings are not stored indefinitely, unlike online data stored by the NSA and private companies. The real world has a level of privacy not found online. That said, the tides are changing. The physical world is becoming less private by the day. At the same time, online privacy is slowly beginning to increase.

In the real world, police and private companies have used camera systems to monitor thefts in shops. With the decrease in camera cost, and new developments in analysis software, they can stitch together multiple public camera feeds into an overview of large areas. They can track individuals from one city block to another, given there are cameras in both areas. With less expensive storage technology, police agencies can probably keep recordings from these cameras for longer periods of time. This would be bad for privacy, but may not actually help police much either. Crimes are usually reported within a day. More than a day’s worth of storage would probably not help much, unless police want to take down a drug dealer that works in public. Even then, no drug dealer would sell in front of a CCTV camera rather than go indoors.

Public and private drones are becoming increasingly common, and may also diminish real world privacy. Drones have capabilities which CCTV cameras do not. They are often used to track and follow particular individuals, even in locations where CCTV cameras aren’t available. A single high-altitude drone can monitor a large area, and governments do not have to spend money installing a large surveillance system on the ground. Drone use is especially common in the war against ISIS, but governments could deploy similar technologies for local policing.

Private drones may be more problematic. Anyone could potentially track anybody else. People complain about online stalking today. Drones may move the concern back to physical stalking. However, personal drones themselves can be easily tracked. Most consumers will probably not be able to afford a high-altitude drone used by governments. They will only have access to smaller ones that need to be much closer in order to track people. If a small quadrocopter is following someone around, they will probably notice. Cyber stalkers can stalk people without their knowledge. Drone users won’t have this ability.

On the other hand, the physical presence of real world stalking makes drones potentially worse. Cyber stalkers don’t necessarily know your location at all times, but drone users will. Like online NSA data collection and cyber stalking, both government and private drone use will likely have their own downsides.

The virtual realm may be doing much better in the future, at least in terms of privacy. The most widely used chat apps now contain end-to-end encryption. WhatsApp automatically encrypts all messages, Facebook is releasing optional encryption for their Messenger app, and Google’s Allo has an optional encrypted chat feature. Google’s new video calling app also comes with encryption built-in by default.

Websites have a long way to go, but web browsing is becoming more secure as well. LetsEncrypt, the organization providing free SSL (Secure Sockets Layer) certification to websites, has added over 7 million domains. SSL enables encrypted connection to all of these websites. There are still numerous setbacks towards becoming more secure online. For example, Google’s servers may be handing data to the NSA, which makes its use of SSL pointless. The data is encrypted between you and Google, so at least nobody other than Google and the government can access it. Most online security issues will not go away anytime soon, but we are at least moving towards increased security.

While one’s public actions are under more surveillance, virtual actions are increasingly secure. If both trends continue, we may end up with amazing online privacy and absolutely none in public.

A real-life surveillance state would be far more useful than a virtual one. People do commit crimes through the internet, but it’s impossible to kill someone through their computer. NSA data gathering can’t solve murders, and they don’t seem capable of preventing terrorist attacks either. CCTV cameras and drones may actually be able to make public spaces safer. This form of surveillance also wouldn’t be as intrusive. If someone is out in public, everyone can see them. There is no reasonable expectation of privacy. CCTV cameras and drones should not make any difference to people’s behavior. This is not the case for private communications. I would rather live in a safer world with public surveillance than the current system that seems to lack public safety, public privacy, online safety, and online privacy. We can at least get safety in public and privacy online, which is a big step forward.

Negative Bond Rates May Cause A Second Housing Bubble

[None of this is any kind of advice. At all.]

Joshua Brown has a new post on his blog, The Reformed Broker, which can be read here. If you don’t want to read the whole thing, Brown basically suggests that Real Estate Investment Trusts have performed better than any other asset class since 2000. I don’t doubt that he’s right. Included in his post is a graph with Vanguard’s REIT ETF. The ETF is seen skyrocketing. It’s “performance against the yield on the 10-year Treasury bond” shows a mirror image. While bond yields go negative, investors pour their money into REITs.

He gives more information on the REIT asset class itself, but I want to discuss the effects of negative bond yields on the greater economy. Policymakers wanted investments to increase, and they have. Lowering the interest rates on Treasury bonds until they are negative will make investors focus on stocks and REITs. Put bluntly, if you loan the government money, you will get back less than you gave them. By making interest rates go negative, the Federal Reserve can be sure that people will invest in other assets. But driving people towards REITs should not have been their goal, at least not to this extent. People looking for a safe place to put their money have lost their best option, and are pressured to buy riskier assets instead. After a financial collapse largely caused by overinvestment in housing, policymakers have created negative yield bonds which lead investors to… overinvestment in housing.

Some caveats apply. REITs are not highly leveraged, toxic assets. They don’t have all the problems associated with the housing crisis. REITs include commercial properties, not just mortgages. A housing bubble would not bode well for them, but it wouldn’t necessarily destroy the value from commercial properties. I’m not forecasting a recession, at least not on the scale that we witnessed in 2007-08.

The bigger problem is that the Federal Reserve is not acting wisely. By making interest rates go negative, the Fed effectively made safe treasury bonds useless to any long-term investor. The price of REITs has increased in the short run, but this is not because they have improved as financial assets. REITs aren’t better, their competition is worse. At these rates, bonds are no longer worth holding. More people investing in riskier REITs rather than in safe bonds means the total risk in the economy has increased. As REITs continue to attract investors escaping from negative bonds, the problem gets worse. The infamous quote rings true, that the Federal Reserve’s job is “to take away the punch bowl just as the party gets going.” The punch bowl is still here. Eventually, we may realize we drank too much.

As with anything I write, feel free to let me know how wrong I am.

Update (July 16, 2016): On July 11th, the 10-Year inflation adjusted Treasury bond rate rose back above zero. It is currently hovering at 0.08%, barely above the zero-bound. Although the more extreme consequences of negative yields will not occur in these cicumstances, low yields still result in many investors choosing higher-yielding REITs over Treasuries.

Population-Adjusted Freedom Index

Freedom House publishes a yearly report with quantified statistics of the level of freedom in each country, according to certain metrics. Their 2016 report states that the world’s freedoms in 2016 were lower than in 2015. Countries are increasingly unfree.

However, these estimates do not adjust for population. If China has 10% more freedom this year than in 2015, but Yemen has 10% less, the total freedom per capita is still
increased due to China’s larger population. I adjusted by adding the points for Political Rights and Civil Liberties for each country, to get a total freedom value for all countries. I then multiplied the total by each country’s population, and added the total adjusted score of all countries in 2015 and 2016.

AND THE RESULTS ARE: Still slightly lower freedom this year. Assuming the measure of freedom is somewhat accurate, and I believe Freedom House is, the world isn’t doing too well. Reactionary attitudes in much of the world, and the continuing terrorist threat in the
Middle East and North Africa may be likely for the decline. Yet even large developing economies were worse off in terms of freedom from 2015 to 2016. The problem isn’t limited to small pockets of radicalization and reactionary politics. The world as a whole seems to be moving away from political freedoms.