Chris Blattman on Experiments

Posted on by

In a must-read post, he describes a number of methodological problems with the interpretation of experiments in social science, but says

There’s no problem here if you think that a large number of slightly biased studies are worse than a smaller number of unbiased and more precise studies. But I’m not sure that’s true. My bet is that it’s false. Meanwhile, the momentum of technical advance is pushing us in the direction of fewer studies.

For me, the crux of the issue is this remark from Blattman.

It’s only a slight exaggeration to say that one randomized trial on the shores of Lake Victoria in Kenya led some of the best development economists to argue we need to deworm the world. I make the same mistake all the time.

The way I would put it is that there is no such thing as a study that is so methodologically pure that by itself it can serve as a reliable guide to policy. As I wrote in What Else Would be True?, the results of any study need to be thought about in the context of other knowledge.

Often, one encounters studies with conflicting results. You tend to focus on the methodological flaws only of the studies with results that you do not like. But remember Merle Kling’s third iron law of social science: the methodology is flawed. That law applies to every study, including experiments.

Thoughts on War

Posted on by

Not my area of expertise, of course. But Neerav Kingsland, also not an expert, wrote this post on the Ian Morris book, which I have not read, and he wondered if I had thoughts. First, some excerpts from Neerav’s post:

Morris’ thesis is this:

1. Government is the primary source of the reduction of violence in societies.
2. Wars caused societies to merge, thereby increasing the scope, scale, and efficacy of government.
3. It would have been great if societies had figured out a way to merge without war, but this, unfortunately, has rarely happened.
4. So, like it or not, war has been the driver of government innovation.
5. Therefore, wars have been the primary cause of our long-term decline of violence.

…Generally, massive war breaks out when a superpower declines.

My thoughts:

1. If you’re a libertarian having a hard time getting your mind around this, think of war as a way of achieving open borders. That is, before they fight, country X and country Y have borders. After they fight, the winner takes over all the territory, and the borders no longer matter.

2. If you are a Hobbesian, then you believe that only a strong government can produce peace. You might regard the U.S. Navy as the force that made the last 70 years of globalization possible.

3. In the wake of the attacks in Paris and San Bernadino, you would think that Congress should declare war on somebody.

Instead, we have this vague “authorization to use military force.” The most charitable reading I can give of that is that it allows the President maximum flexibility to wage war in a very ambiguous setting, in which enemies do not wear uniforms and they are embedded with civilians. But I personally do not like this approach. Here are two alternatives that I think are better, although it should be clear on reflection that there are major problems with every possible approach.

a) Get rid of the authorization to use military force and legislate a strict non-interventionist policy. I think that has at least two things going for it. First, it is a clear, unambiguous policy. Second, it does not run all of the risks of flawed execution and unintended consequences that flow from interventions. However, it does mean that whatever advantage there is/was from having the U.S. as hegemonic power gets tossed away. For example, we might go through a period of de-globalization, as various conflicts spin out of control.

b) Get rid of the generic authorization to use military force and instead declare war on the Islamic State. One advantage of this is that it designates a specific enemy and implies a finite objective. We would stop sending drones all over the map and instead focus on taking over the territory that now belongs to barbarians. The disadvantages are that this increases casualties in the short run and it probably means that we would have to undertake a long-term military occupation, which has many pitfalls. It exerts no leverage against Syrian President Assad (it probably helps him). It ignores any barbarism that originates elsewhere.

A Commenter Reviews The Midas Paradox

Posted on by

It’s ‘Handle’:

Any proposed answer to the question of “What really caused the Great Depression” is inevitably pregnant with policy implications, and so is unavoidably incredibly politically charged. That’s one of the reasons so many top economists are so keen on studying it. Big names like Keynes, Hayek, Friedman, and Bernanke, just to name a few.

[more below the fold] Continue reading →

My Chips

Posted on by

Recall that the three choices were:

a) I would be concerned that Muslim intolerance of non-Muslims threatens our way of life.

b) I would be concerned that backlash against Muslims will get worse, empowering radicals and provoking more conflict.

c) I would be concerned that the media will blow the incident out of proportion and that politicians will use it as an excuse to expand surveillance, restrict gun rights, or restrict immigration.

Yes, (a) is meant to reflect the Conservative civilization-barbarism axis, (b) is meant to reflect the Progressive oppressor-oppressed axis, and (c) is meant to reflect the Libertarian freedom-coercion axis. However, note that I use the three-axes model to characterize people’s preferred language. People’s thinking is more complex.

That said, I would allocate my own chips this way:

a) 30 b) 10 c) 59

Explanation:

1. Because there are reasonable people who would favor each of the three choices, I do not think it would be wise for me to put less than 10 chips on any one choice.

2. I find choice (c) most compelling. What do media do other than blow incidents out of proportion? What do politicians do other than take advantage of Fear Of Others’ Liberty (FOOL) to expand their coercive powers? Incentives dictate such behavior.

3. I think that many libertarians would rank (b) higher than (a). I imagine that the “liberaltarian” types would. Those libertarians who are descended from Rothbard might want to put a lot of chips on (b). You will find some who fault the U.S. for the Cold War and perhaps even for its role in WWII.

4. Choosing between (a) and (b) depends on what type of error we might be making about “the other.” Usually, you err on the side of thinking that “the other” is worse than what it really is. The accounting people think that the marketing people are out to screw the company, and the marketing people think the same of the accounting people, when both are just doing their jobs as best they can. So I think that looking at the most probable error, (b) wins over (a).

But what if radical Muslims really are like Nazis? Most Muslims are not radicals, but did it help that most Germans were not Nazis? Even if it is unlikely that we are under-estimating the Muslim threat, the consequences of making that mistake could be quite dire. So my own inclination is to weight (a) more than (b). In terms of the axes, I suppose I am more susceptible to civilization-barbarism language than to oppressor-oppressed language.

5. I can understand someone putting the majority of their chips on (b), but I would not trust a progressive who puts 90 or more of their chips on (b). To go that far toward denying the validity of (a) and (c) strikes me as dogmatic. Ironically, a progressive who did put 90+ chips on (b) probably would think of himself or herself as showing superior nuance and sophistication. And, yes, I worry that the President is one of those. And, yes, I also would worry about a President who puts 90+ chips on (a), or even on (c).

Dusting Off David Brin

Posted on by

Eugene Volokh writes,

If you think you can prove someone is a terrorist, lock him up. If you have probable cause to think he’s a terrorist, and think you can develop proof beyond a reasonable doubt, arrest him. Even if you have only suspicion, follow him, ask people about him, and so on. But if you don’t have enough to prosecute or even arrest someone, you can’t take away his constitutional rights, even if you suspect he’s a terrorist (or if you suspect he’s a drug dealer or a gang member or whatever else).

He is talking about President Obama’s proposal to deny guns to people on the “no-fly” list.

In general, I think that the “no-fly” list and the “terrorism watch list” are of doubtful value. The latter, according to Wikipedia, has one million names. I do not believe that any agency is watching one million people. And, of course, the San Bernadino shooters were on neither list, as far as I know.

It is, once again, time to re-read David Brin’s The Transparent Society. You may also wish to read or re-read my essay on it.

Brin’s vision bothers many people. However, I think it is the most reasonable equilibrium in a world of terrorism and powerful surveillance technology.

The Midas Paradox So Far

Posted on by

That is Scott Sumner’s magnum opus on the Great Depression. I am part way through it. The importance of the book cannot be overstated. Here, I want to mention two problems I am having.

1. The structure of the book. It is difficult to get through. He constantly interrupts his own interpretive narrative to discuss other economists’ narratives. I would have rather seen the discussions of other narratives cordoned off into appendices.

2. Sumner wants to rely on stock market movements as indicators of the effect of monetary policy on the economy. I am never a fan of trying to interpret the stock market, and this time period seems particularly problematic. Robert Shiller’s famous argument against the efficient markets hypothesis is the “excess volatility” that he found in stock prices. The period that Sumner is describing has volatility that goes far beyond even what caused Shiller to reject market efficiency. Reading Sumner’s narrative, there seem to be a lot of days where the market goes up or down by 4 percent, 5 percent, or more. In today’s terms, that would mean the Dow often gaining or losing 700 to 850 points in a day. I am inclined to dismiss 90 percent of these movements as irrational (I would do the same for short-term movements in the market now). I understand that Sumner is looking for indications that the economy is responding to monetary policy, as opposed to some other factor, but I find it more persuasive when he cites medium-term trends in commodity prices than when he cites short-term stock price behavior.

Donald Trump, Progressive

Posted on by

Virginia Postrel writes a must-read review of a forthcoming book on racism in the Progressive movement.

restricting immigration was as central to the progressive agenda as regulating railroads. Indeed, in his five-volume History of the American People, Wilson lumped together in one long paragraph the 1882 Chinese Exclusion Act and the 1887 Interstate Commerce Act as “the first fruits of radical economic changes and the rapid developments of trade, industry, and transportation” — equal harbingers of the modern administrative state. With a literacy test and ban on most other Asian immigrants enacted in 1917 and national quotas established in 1924, the progressives bequeathed to America the concept of illegal immigration.

This would put Trump in the Progressive tradition. Perhaps he does not fit there. But he certainly does not fit in the libertarian tradition.

The book Postrel review in her essay (not the book by Wilson) is “Illiberal Reformers: Race, Eugenics and American Economics in the Progressive Era, by Thomas C. Leonard. Surely there will be a review by Jonah Goldberg in some future issue of the Claremont Review of Books.

Matt Ridley’s Explanation for Mass Shootings

Posted on by

Tyler Cowen asks,

What is the best theory for the rise in mass shootings?

In The Evolution of Everything, Matt Ridley writes,

Surely the explanation for most killing lies in the fact that natural selection has endowed human beings with the sort of instinct that means that (in [Martin] Daly and [Margot] Wilson’s words) ‘any creature that is recognizably on track towards complete reproductive failure must somehow expend effort, often at risk of death, to try to improve its present life trajectory’.

I would note that (a) the proportion of young men without fathers has gone up, and (b) I cannot recall any mass shooters who came from households where their father was present.

The problem with this explanation, from Tyler’s point of view, is that it should point to an increase in violence of all sorts, not just mass shooting. But overall violence is down.

Perhaps the decline in overall violence is due to high rates of incarceration? Not a very libertarian thing to say.

Computers and Go, Continued

Posted on by

MIT Tech review reports,

This is how Tian and Zhu begin. They start with a database of some 250,000 real Go games. They used 220,000 of these as a training database. They used the rest to test the neural network’s ability to predict the next moves that were played in real games.

Pointer from Tyler Cowen. Quite a while ago, I thought that computers could master Go, in part by doing something like this.

On another AI note, this article says,

Tenenbaum and colleagues tested the approach by having both humans and the software draw new characters after seeing one handwritten example, and then asking a group of people to judge whether a character was written by a person or a machine. They found that fewer than 25 percent of judges were able to tell the difference.

To me, the program did not sound like the great breakthrough that the media are touting. But I am not an expert in the field of AI, just an opinionated observer.

Where do you put your chips?

Posted on by

Imagine that there is another San Bernardino. This might raise a number of concerns. Three of these concerns are listed below. Decide which ones matter the most to you. Mentally, take 99 chips and place them on the concerns. You could put all 99 on one concern, or you could spread them in varying amounts among the three. You cannot put less than 0 chips on any option. You cannot add a fourth option or modify the wording of the three options. If you don’t like the game, then you don’t have to play.

a) I would be concerned that Muslim intolerance of non-Muslims threatens our way of life.

b) I would be concerned that backlash against Muslims will get worse, empowering radicals and provoking more conflict.

c) I would be concerned that the media will blow the incident out of proportion and that politicians will use it as an excuse to expand surveillance, restrict gun rights, or restrict immigration.

In the comment section, just say how you allocated your chips and how well your allocation lines up with where you see yourself on the three-axes model.* Please do not add other comments. Please do not leave multiple comments. On Sunday, I will give my own answers in the comment section and perhaps add a post with further remarks.

*If you are not familiar with the three-axes model, just say that. If you want to become familiar with it, you can look at previous posts in the category (scroll down to get to older posts) or get the Kindle edition of the book.