Praxeology and Prediction
[UPDATE: It was not until I pushed this post out on Twitter that I realized I should have called it, “WHY DID THE ECONOMIST CROSS THE ROAD?” Ah, life is full of regrets.
But seriously, that is a clever title for this content, beyond the obvious.]
I am in the middle of a lot of “day job” projects and can’t devote the full time to this post that it deserves. The context is that Milton Friedman famously said in economics, models need to make testable predictions, otherwise we are not being scientific but instead are engaged in spinning out mere tautologies.
Mises and many of his followers reject this strongly, thinking that economic laws should be deduced logically from the axiom of human action.
OK now to my anecdote: To walk from my apartment to my office at Texas Tech, I have to cross a busy street that has 4 total lanes, with a divider in the middle. So in the morning when it’s busy, what usually happens is that I wait for a break on my side of the street, I walk to the middle and hang out on the divider, and wait for a break in the traffic going the other way to get totally across the street.
The rub is that there’s a gap in the divider, so that cars going one way on the street are allowed to do a U-turn through the gap, so that they can effectively get onto the other side of the big street, going the other way.
Anyway, yesterday it was really busy. I made my way across the first two lanes of traffic, and I was hanging out on the divider. I saw a car turned right onto the busy street (behind me), drove up past me on my right, and then was waiting “in” the gap in the divider (about 20 feet in front of me), trying to turn left.
The driver of that car and I were *both* waiting for the oncoming traffic to give us an opening, so that I could cross and the driver could make the turn. Now normally in this situation I have to be careful, because if a car is pulling a U-turn, then obviously it will be heading straight for me if I’ve also entered the road (jaywalking across the remaining two lanes).
But I was very confident that the car would NOT pull a U-turn, but instead would make a 90-degree left turn and go straight through to the side street that intersected the busy 4-lane road.
How did I “know” this? Because I realized that if the driver pulled a U-turn, it would simply bring him back to where he had been a few minutes ago. In other words, had the driver wanted to get to where the U-turn would place him, it would have been dumb for him to turn right onto my busy street in the first place. He should’ve kept going straight.
Now granted, this wasn’t a prediction on the order of astronomers telling us where to find Halley’s comet. But I *was* right. The driver *didn’t* pull a U-turn, and I was able to cross the street. (By the way, I made sure the driver wasn’t turning. I didn’t stake my life on my prediction.)
One last thing: Notice that my prediction, based on “getting inside the head” of the other acting being, was way more tractable than if I had asked a bunch of neurobiologists to analyze the state of the driver’s brain stem and predict the physical movements of his feet and hands. My knowledge of praxeology allowed me to navigate in the real world a lot better than the natural sciences alone would have allowed.
So if the driver had done a U-turn would this post have been about Donald Trump ?
Yes, Transformer, I would’ve come up with excuses for the driver running over an immigrant.
This depends if the driver was a licensed Cabby or an Uber driver doing the U-turn. One turn might be deliberate while the other might run up the meter and if English was a second language it might not matter what was said exactly. Bets are still on Praexology.
Did you win anything in the Liberty Classroom contest? Or were you a complete and utter #loser?
I was not eligible for the car. I won the satisfaction of knowing that more people are learning about liberty.
You had a model of the driver’s mind, and you used the model to make a prediction. By getting your prediction correct you have collected one empirical data point in favour of your model (mind you, we need to be careful with survivor bias in the sample here; if you had died out on the road, you would not be writing up a report).
I would argue that Milton Friedman’s criteria are satisfied.
Also, IMHO from a very pure and clean praxeological point of view, you cannot reliably deduce the internal objective of the driver from a history of their route. For example, maybe the driver is simply lost and just realized they are going the wrong way, or sometimes communication companies want to check mobile coverage and signal strength in a city, so they just get people to drive around and try to collect measurements all around the area. Or another example, maybe the driver is an under cover cop who just got a blip on the radio asking them to change route and check out something that might be happening down town (but don’t make any obvious noise like sirens). I knew a guy who used to tune engines with a small laptop discretely on the passenger seat beside him and an adjuster knob on the dash so he can tweak it up or down a bit as he drives around. What I’m saying is there’s a lot of possible scenarios where you won’t be able to deduce the internal preferences of the driver.
However, that said… you also have a working knowledge of all the likely scenarios so you are not using praxeology in the pure sense, you are using a combination of praxeology plus background “common sense” information. For convenience you automatically rule out a lot of possible but rare situations. The praxeology provides the “glue” that holds all those little bits of knowledge together so that combination is what builds a model.
“The context is that Milton Friedman famously said in economics, models need to make testable predictions, otherwise we are not being scientific but instead are engaged in spinning out mere tautologies.”
We know this well-known and oft-repeated argument is not true, because the argument is itself a “mere tautology” by its own terms, since it is not a hypothesis subject to falsification. There is no room in the argument for any testing. It doesn’t leave any possibility of being wrong and the counter-argument being true. It is an a priori argument that absolutely denies the validity of a priori knowledge.
It is a self-defeating argument.
Are you saying that tautologies are not true?
Only untrue tautologies are untrue
No that wasn’t what I was saying.
In short what I am saying is that tautologies that purport to deny or reject or invalidate tautologies as not really knowledge, cannot possibly be true because they deny, reject and invalidate themselves.
If you’re going to refute a tautology as untrue, you cannot invoke an tautology to do it.
Moreoever, there are true tautological statements. The law of marginal utility for example is a grand complex tautology. It is nevertheless true in that it really does describe a logical reality of our actions. It cannot be undone. While not a prediction of content, it is a description of the form of our action.
Finally, the empiricist or positivist methodology, rests on tacit tautologies. Positivism nullifies itself when applied to the positivist him or herself.
M-F, because something isn’t science doesn’t mean it has no meaning or value. Dr. Murphy’s Christianity and philosophical thought would be examples. There are lots of things meaningful and valuable that are not science. I read Karl Popper’s book over fifty years ago when I was in college and he made the case; as it appears Milton Friedman made it, that if it is not falseifiable it is not science. Economics for the most part would be classified as a pseudo-scientific along with Climate Science and praxeology though that does not mean they do not have meaning and value. (Full disclosure, I assume the book I read was “Conjectures and Refutations” published in 1963. I don’t remember any of it except the requirement for falsifiability. I have ordered a copy from my local library to read it again.)
This brings me to the latest Contra-Krugman podcast. I don’t believe that Krugman IS contradicting himself when he recommended President Obama do stimulus spending but now is recommending against it for President Elect Trump. The reason is as Jack Kelly of the Pittsburgh Post-Gazette pointed out seven years ago; this administration’s (speaking of the Democrats) goals are to support their core constituencies, at the head of the line are crony capitalists who receive subsidies and bailouts. Next are faculty and staff at colleges and non-profits that depend on government funding. Then come the unions. This administration works solely for their own personal gain and to retain their power and prestige and have done nothing for the economy. Obviously Krugman believes that President elect Trump’s stimulus will reward the wrong (deplorable?) people and that is why he is against it.
1. OK. Set up an experiment to falsify the assertion that people know what to produce and what skills they should learn based upon observing what prices the great mass of strangers pay for, offer to sell and offer to purchase those goods and skills.
2. Krugman isn’t a hypocrite. He’s just a relentless but bad liar. Of course Krugman only supports “stimulus” to aid his political favorites. The market does not have or lack “momentum” and does not require “stimulus” in any event. Keynesianism is just a scam to justify the theft and transfer of wealth.
“We know this well-known and oft-repeated argument is not true, because the argument is itself a “mere tautology”” (emphasis mine).
Yet you say there are true tautologies, so you cannot say it is not true because it is a tautology.
If I were a psychologist that had never heard of Praxeology I might come to exactly the same conclusion you did. Would you say I was using Praxeology without being aware of it, or are you using psychology and attributing it to Praxeology, or are they the same thing?