March 2017

Allied Bombing


When I think about strategic bombing in WWII, I try to imagine my team changing our job from building economic forecasts for our firm to trying to build prediction models and narratives for where we should bomb, and why. There is a banal practicality behind having to make choices under uncertainty that doesn’t change form just because you’re deciding who to bomb.

This makes WWII strategic bombing very strange. What if the choices that resulted in millions of civilian deaths were based on the same low-information fuzzy decision making that basically all other choices are made from? It’s sad, and I think we all sort of hope that a group of brilliant British and American men sat around in 1943, with full information, and decided they had no choice but to burn hundreds of thousands of civilians to death in order to win the war. That would mean they had a clear philosophy and model of the world, which we can then try to supplant with our own refined philosophy and model. Wouldn’t it be more likely that they were instead a mix of CEO types and military bureaucrats, who were making choices without historical precedent in an existential environment? And if that were the case, doesn’t spending decades analyzing what they ought to have done seem… kind of silly?

If need be, as the war went on, we hoped to shatter almost every dwelling in almost every German city.  ”  (Official transcript of the meeting at the Kremlin between Churchill and Stalin on Wednesday, August 12, 1942, at 7 P.M.)

“The destruction of German cities, the killing of German workers, and the disruption of civilized community life throughout Germany [is the goal]. … It should be emphasized that the destruction of houses, public utilities, transport and lives; the creation of a refugee problem on an unprecedented scale; and the breakdown of morale both at home and at the battle fronts by fear of extended and intensified bombing are accepted and intended aims of our bombing policy.  They are not by-products of attempts to hit factories.” — “Air Marshal Arthur Harris to Sir Arthur Street, Under Secretary of State, Air Ministry, October 25, 1943”

August 7th, 1943, NYtimes

If you do want to analyze it, there is the economic statistical measurement of strategic bombing. What was the investment? What was the return? How do we measure this? In some aspects it’s incredibly open to economic analysis: There were inputs into planes and bombs, and we had an outcome of measured civilian deaths, shut down factories, and less economic efficiency. The civilian deaths of the axis are estimated at 305,000-600,000 German civilians, 330,000-500,000 Japanese civilians, and 60,00-100,000 Italian civilians. The allied forces also suffered, but not as much, excluding the Soviet Union which lost more than 500,000.

In Jurgen Brauer’s book, Castles, Battles, Bombs, which is a book by an Economic Historian, the author wrote that the British Air Chief Marshal Sir Arthur Harris hoped bombing would be a method to win a war without the abject horror of WWI, which removed a generation of young men from Europe. Harris’ strategy was to measure success in square miles burned. His view was that surgical bombing was ineffective, and to win a war you had to burn or blow up everything in an area that was conducive to the enemy’s war effort, including enemy morale. The author’s interpretation of strategic bombing, which he took from analyzing the official military documents from the United States Strategic Bombing Survey, is that the goal is to “first destroy the opponents ability to defend itself against air power; then, second, attack those targets inside the opponent’s territory that support it was war-making on the front; and third, wait until the enemy collapses from within.”

The problem though is that we have no counterfactual framework to even consider the question. Although we can observe that the war didn’t end until there was a ground invasion of Germany. In that case was strategic bombing useless? Obviously not, since the invasion might have gone faster or easier as a result of the bombing. The author goes on to try and measure the marginal returns from incremental inputs in bombings, which he argues were too low to justify the massive investment into each marginal bomb.

He uses the bombing in Hamburg as an example of a poor outcome on bombing returns. In Hamburg, Air Chief Marshal Harris created the first air-generated firestorm. From July 24th to July 30th the Brits and Americans dropped over 9,000 tons of bombs from thousands of planes. Between 35,000 and 50,000 civilians died in the inferno. Although it appeared that within days the bombed railroads and industries were back up running. That’s the type of stuff that I don’t think anyone could have predicted. It’s too complex. If a world has no real experience with this type of massive bombing, it would seem like a reasonable prediction that if thousands of planes create an inferno in a city, it will be permanently out of commission. That ended up not being the case, strangely enough. In that case we could say with new information it was a bad choice, but at the time it was a rational strategy.

It’s hard to know what percent of efficiency the city returned to, but let’s imagine 95%. WWII was a close war, what if 95% was enough? And what about the tens of thousands of civilians that were burned to death? At the company I work at now, if we randomly lost a few buildings and a few percent of our workforce we would have a tough time continuing at the same efficiency, even more so if we all knew people who had died. And could this have damaged the morale of the country? As far as I’m aware, measuring how slight changes in the morale of a country propagate through towards defeat or victory is essentially impossible. These things are impossible to measure.


I have told my sons that they are not under any circumstances to take part in massacres, and that the news of massacres of enemies is not to fill them with satisfaction or glee. I have also told them not to work for companies which make massacre machinery, and to express contempt for people who think we need machinery like that. -Slaughterhouse 5

There is a pattern. These people who lived through the bombings are willing to make it an inalienable principle to never repeat this violence. When I was a little younger my rationality was more naive. I took a sense of pride in the fact that I’d have dropped the bombs. Why would I even think that? Well, dropping the bombs was a hard choice, it was a horrific thing to do. Yet we had to do it, for the good of mankind and our survival. My role model, George Orwell, wrote that pacifists secretly know that “Those who ‘abjure’ violence can only do so because others are committing violence on their behalf.” I took that to mean that while the weak sit around, it takes an iron will and strength to make these hard choices.

May 30th, 1945, NYtimes

Both Vonnegut and Orwell had the same other worldly response to the bombs. The protagonist Billy in Slaughterhouse-5 is warped between the bombed German town Dresden, where he was kept as a prisoner of war, and an alien biodome run by the Tralfamadorians. His book juxtaposed the equal absurdity of mass incineration of civilians and alien abductions, as something we can’t accept understand or accept as true.

Orwell wrote:

A world in which it is wrong to murder an individual civilian and right to drop a thousand tons of high explosive on a residential area does sometimes make me wonder whether this earth of ours is not a loony bin made use of by some other planet.

What worries me is that there may be no way for us to evaluate whether it was the right or wrong choice without going through their experience. They try to write it for us and reconstruct the information, but when we read it and run it through our heads it’s pointless. If I could spin off simulations of myself who experience strategic bombing, then have them come back and merge, would my view change? Would I experience a surge of pacifism like Vonnegut?

An abstraction of the cost vs. benefit of a grand historical experiment is great when there is a clear signal, since we can then carry that knowledge with us into the future. When the signal is so volatile that reasonable people can’t come to a conclusion, it might mean the system is too noisy and has too much information loss for us to take a clear lesson. Even for those of us today who read and think about these things, the amount of information available in historical archives is too great for us to fully know, and the more crucial information is probably lost to the sands of time.

When scholars disagree, and 80 years after the fact people are still debating and arguing over the right answer, it might mean our estimation methods aren’t going to converge. It is possible that one side is totally right, and they will eventually convince the other side of the argument that they are wrong, but I tend to think that is less likely as time moves on from the event.

Why then are the modern debates so important to political identities? Different political groups take a stance on the strategic bombing question, since they trace their identity or roots back to this time. Depending on how they evaluate it, it counts as for or against their cosmic scorecard. There are clearly lessons to be learned from WWII and the existing political systems at the time, but how granular do these lessons get? Can we really add or subtract points to democracies or fascist governments based on the death tolls and bombing strategies during a period of total war?

While writing this I stumbled upon subreddits dedicated to making fun of the internet neo-Nazis, who often write about so-called alleged allied war crimes. These neo-Nazi groups that use the strategic bombings as evidence that the Nazis weren’t uniquely bad. Neoreactionary literature uses the WWII strategic bombing of the allies as evidence that the US and the allied forces don’t have a strong claim as the ‘good guy.’

And, of course, (c): the Allies positively reveled in the aerial mass incineration of German and Japanese civilians. They did not kill six million, but they killed one or two. There was a military excuse for this, but it was quite strained. It was better than the Nazis’ excuse for murdering the Jews (who they saw, of course, as enemy civilians). In fact, it was a lot better. But was it a lot lot better? I’m not sure. -Moldbug

There is probably some use in making that argument. I don’t think it’s related to political systems as tightly as many might like to believe. It’s instead related to the ability of humans to justify and commit massive atrocities, even when we think they are the good guys.

I think the lesson behind studying this topic isn’t that there is a clear answer, but that it’s a complex problem that can’t be clearly classified. I see this as an optimistic lesson, because if we are able collectively to study history without trying to tally up the winners and losers, and fix past injustices, and as a result focus less on our classifications, identities, and historical records, we can lower the chance we end up in the same situation.

Bias Correction

[epistemic warning: I was recovering from surgery and wanted to document my strategy for reading the news and correcting for bias. This is a very boring post, read at your own peril.]

Contemporary media seems to be growing increasingly outrage driven. Trying to explain the media as a causal reason for increased polarization, or increased polarization as a causal reason for outrage driven media, is a hopeless exercise. In my own mind at least, I think about it as a complex simulation of interacting humans that is running a democratic algorithm, which due to our stronger information technology, has more interconnections every year. When Democracy is running as an algorithm, the optimal classification would separate two sides as efficiently as possible.

The equilibrium here will be whatever information sources most efficiently can divide an individual into one of two political tribes. Truth has no reason to replicate better than fiction. No one person can absorb all the truth of a given day, but the way they absorb information can be wrong in different ways.

I recently read this Yale Cultural Cognition piece on alternative facts. The author did an incredible job filtering through the dynamics of how Trump engages with segments of the mainstream media. The piece builds largely off this other equally great previous piece on motivated reasoning.

From this simple model, we can see how identity-protective reasoning can profoundly divide opposing cultural groups.  Yet no one was being misled about the relevant information. Instead, the subjects were misleading themselves—to avoid the dissonance of reaching a conclusion contrary to their political identifies.

Nor was the effect a result of credulity or any like weakness in critical reasoning.

On the contrary, the very best reasoners—the ones best situated to make sense of the evidence—were the ones who displayed the strongest tendency toward identity-protective reasoning.

Such coverage, in turn, impels those who want to defend the truth to attack Trump in order to try to undo the influence his lies could have on public opinion.

But because the ascendency of Trump is itself a symbol of the status of the cultural groups that propelled him to the White House, any attack on him for lying is likely to invest his position with the form of symbolic significance that generates identity-protective cognition: the fight communicates a social meaning—this is what our group believes, and that what our enemies believe—that drowns out the facts (Nyhan et al 2010, 2013).

A problem I see here is that motivated reasoning isn’t always wrong in a clear way. It is one of the trickiest ways you can be wrong, and probably the hardest to identify.

When making motivated reasoning errors, I like to use heuristics to try and estimate how and why my model of the world might be biased.

Once you receive an information set, there are different ways to build your model of the world incorrectly. In this case your model of the world is simply how you build a structure to transform raw information into knowledge of the world around you. Most of the time the information you’re receiving is output from someone else model. While there are exceptions for primary sources, even this is often only a partial-exception, since most of the time someone else had to build the feed to capture primary information and send it someplace where you can access it e.g. youtube videos of primary sources still required another person to build a model of the world where capturing and transmitting that information was beneficial.

There are many different ways you can build an incorrect model. Less Wrong does a great job of documenting these, and going over how almost everyone has a biased view of the world due to simple cognitive biases such as confirmation bias. While there are entire books documenting how to evaluate these models of the world, the goal should be to make sure they are probabilistic, unbiased, and systematically are not over or under certain (i.e. an 80% chance of something happening, happens about 80% of the time).

As I mentioned, I want to focus on motivated reasoning. When we use our brains we are able to think about and ‘scan’ all the information we know on a topic. This also includes references we have, sources we know to investigate, and even general processing skills, such as Causal Inference, that help us organize, evaluate, and structure our information. This is high-dimensional information from a variety of sources, including (primarily) output from other peoples models. While we can control what we read and hear to an extent, we can’t remove information from our brain (unless we simply forget about it over time). If we want to model the world correctly we need to do a few things:

1.) Ensure our information set spans the true outcome set. All this means is that there has to be complete information within our brain such that there exists some model  that is able to map our information set to a true model of reality. For example, if you only ever read the NYtimes, you will have trouble developing any bias correction, since the information you feed into your model is based only on the distribution of the output of their model.

The set of all relevant information is a purely theoretical concept, which is a record of all information possibly related to a given event, which is far too much and far too complicated for the human brain to absorb. We can imagine any given source (whether it’s news, primary sources, videos, twitter etc) as representing a subset of the set of all relevant information. For raw information you’re simply receiving an arbitrary slice of recorded or documented information on an event. Sometimes news outlets attempt to aggregate this information, and then reduce its dimensionality to give you a small model that represents reality. The reality is all information we receive is a tiny slice of the total information set, and typically suffers from model bias. However, if we are able to aggregate enough information from different sources, such that we have the fidelity that there does exist some bias correction and weighting that could allow us to create a true unbiased model of the event, then our information set will span the outcome set.

2.) Have an existing estimate of how the unbiased distribution should look. The only way to do this, properly, is to study all past events where a given information provider (e.g. NYtimes), view their prediction or explanation of the world, then find systematic biases or misses. This is challenging in part because we aren’t used to thinking or reading news articles as predictions, which we generally think of as clear statements with assigned probability values. We know that lots of news sites writes lots of articles on differences in outcomes between white and black people, and reports on how these differences are due to systematic racial discrimination at an individual and institutional level. This constitutes a prediction of the world. It is comparing the mean between two populations, based on a chosen variable, and then interviewing or asking for comments by journalists or academics who share an underlying causal view on why this is the case.

Since no two events are the same the goal is to identify a set of latent unobserved dimensions that map to salient policy dimensions. What this means is that while there might be thousands of articles on race relations, we can explain the bias behind the model in each article by using a Bayesian filtering algorithm. Having an idea of what it means to detect and filter out bias correction using modern methodological research design helps us frame the correction in a more scientific way. I can try to get my brain to simulate how I know those models would work if we could actually run them.

Unfortunately, estimating this is still very far out of the realm of data science, as it requires robust evaluation of reports vs. reality, both at a deep level, and across decades. The closest our best Political Scientists can get now is to extract the single left-right latent dimension from US congress, and use textual analysis to match the text base in a newspaper to the estimated dimensional points. While this doesn’t let us identify latent dimensions behind the entire set of news reports, we can use it as a first-order bias correction. You already know this though, Fox is ‘too far to the right’ and the NYtimes is ‘too far to the left.’ (or there is the more common view that one of them is perfectly unbiased and correct, and the other is hopelessly wrong.)

3.) The first two are problems for all scientific inference as applied to inference of the news: Ensuring you have the proper information set, and correcting for bias. The third is similarly a challenge across disciplines, but is the most egregious in political analysis. Motivated reasoning is when you have an attachment to a specific model of the world being the correct outcome. This seems to be a conflict between how humans evolved to build and form tribal political relationships, and what it means to accurately perceive a political reality.

Using motivated reasoning we are more likely to engage in biased search to seek out information or sources that confirm our views, accept evidence using biased assimilation, and in general across the board seek out bias confirming information. It’s silly when you think about it, if you are confident that your view of the world is correct, you would want to build that confidence by pursuing an unbiased look at reality. As far as I can understand, the only reason this isn’t what happens is due to our evolutionary preferences to belong to tribes. We literally get high when we follow political meme pages, watch Hannity, or share Jon Stewart clips. That tribalistic feeling of belonging helps us run our democratic algorithm to gain more political power for our side.

Politics and power is fun. I remember when I was 17 and in my first year of university, I knew that I wanted to be an intellectual, and I liked being edgy. I remember going to the public library and renting books with titles like “The Anti-Corporate America Reader.” This was when Obama was campaigning, and everyone knew it was subversive and cool to be progressive. When I couldn’t reconcile these very far-left views with my Microeconomics courses, I resigned to be a Paul Krugman liberal.

With motivated reasoning it’s too easy to become and stay a progressive or a conservative. Since these represent the only two rallying points of political power in our democracy, almost every argument or model of the world attaches itself to one of the two. If you imagine these two points as being circles embedded in n-dimensional space, where every dimension is an abstracted political issue, no matter where you are, you must be close to one than the other (or equidistant). Here I think of motivated reasoning as not simply the way we tether ourselves to the point of closest political-power, but also the way everyone else works to keep you tethered, since your membership improves their power.