I come in peace, fellow rationalists.
I know that some truths can be discovered through the application of pure reason, without appealing to empirical data. These truths are limited in number and tend to be very abstract, but they still exist and are foundational to our other beliefs. I give an argument for them in my book Square One: The Foundations of Knowledge.
However, rationalism is easily abused and often turns dogmatic. Our perennial critics – empiricists – correctly point out the many dogmas common to rationalism. In turn, we rationalists correctly point out the many unspoken assumptions of empiricism. The purpose of this article is to point out where my fellow rationalists are indeed being dogmatic, in particular, with regard to Austrian Economics.
I am an apriorist when it comes to economics. There is indeed a non-empirical, purely conceptual framework that we all bring to our analyses of human action. However, apriorist arguments are frequently over-extended by extreme apriorists, whose ideas are best represented by the philosopher Hans-Hermann Hoppe.
So for the sake of apriorism, we need to tighten up our arguments and specify exactly what can and cannot be claimed by appealing to pure logical analysis. We can make axiomatic deductive claims in economics, but they don’t tell you almost anything about the world. They are important claims – even fundamental – but they are so abstract that most people won’t find them relevant.
Take the common question:
“Does increasing the minimum wage cause a disemployment effect?”
I’ll be analyzing two different answers: “Yes, it certainly does,” and “Yes, it probably does, given reasonable assumptions.” The former is an apriorist claim: on purely logical grounds, we can know that increasing the minimum wage causes disemployment. The latter is an empirical claim: given what we know about the world, it’s most likely true that increasing the minimum wage causes disemployment, though it’s not logically necessary.
For all practical purposes, the latter position is correct, and the former position is dogmatic.
Other Things Equal…
The careful reader might have thought, “Ah, we have to clarify the proposition further. It’s not simply that an increase in the minimum wage causes disemployment. It’s that an increase in the minimum wage causes disemployment, ceteris paribus. You have to hold every other variable constant.”
I concede this point, and this new, more precise proposition is indeed true.
It’s true and neutered.
There is a world’s worth of difference between claiming,
“X causes Y,” and
“X causes Y, everything else constant.”
The claim that “X causes Y” is a regular claim about the world. The claim that “X causes Y, everything else constant” is not a regular claim about the world. In fact, it’s not really saying something about the world; it’s talking about a hypothetical world where only one variable changes at a time, which is not the world we inhabit. I fully recognize that the ceteris paribus condition can be helpful as a thought experiment to clarify our concepts, but it often gets abused by the dogmatic rationalist who ends up claiming:
“X causes Y, ceteris paribus.”
“Therefore, X causes Y.”
To state the error more concretely, the dogmatic rationalist says:
1) An increase in the minimum wage causes disemployment effects, ceteris paribus.
2) Therefore, an increase in the minimum wage causes disemployment effects.
This might seem like a subtle error, but in fact, it’s a catastrophic one. It’s partly the reason why aprioristic reasoning is seen as being dogmatic. Extreme apriorists try to claim that “X is a matter of logical necessity,” when in fact, it’s actually an empirical matter, and because of the axiomatic-deductive nature of their argument, they are not open to being convinced otherwise. I’ve had to change my own beliefs on this subject, having been more on the dogmatic side before realizing my ideas were true yet neutered.
Let’s take a simple example. Say I ask, “Will increasing the minimum wage in Seattle cause disemployment effects?” An extreme apriorist answers, “Yes, certainly, because of these particular causal connections…”
Now imagine I follow up with the question, “But what if nobody follows the minimum wage law? If nobody follows the law, then surely it won’t cause disemployment.”
I’ve actually asked this questions several times to economists in person, and I’ve heard some interesting answers. One prominent apriorist told me, “Well, they’ve got to follow the law!”
But surely, in the real world, people don’t have to follow the law. And if they don’t follow the law, then the standard economic story simply doesn’t apply. I recognize that in the thought experiment they must follow the law – otherwise the analysis won’t work – but that’s not the world we live in. Nobody is asking economic questions about your thought experiment. They want to know what will happen in the real world if Seattle actually raises the minimum wage. Whether or not people follow the law is an empirical question; you cannot logically deduce the answer. Therefore, questions about the minimum wage in the real world require empirical assumptions in order to answer. Whether or not people follow the law is only one example; there are innumerable other empirical assumptions that get packed into economic claims. These assumptions might be perfectly reasonable, but they’re still empirical in nature.
As I like to say in metaphysics, it might be the case that the minimum wage in your head is not the same as the minimum wage in the world.
Imagine that the following were true:
“When the minimum wage increases, it changes the self-image of employees. They view themselves as being higher-quality workers and raise their productivity levels accordingly.”
If that were true, then an increase in the minimum wage could, in fact, increase employment. The increased productivity of workers could make their employers more money, which means the employers could afford to hire more people.
Notice that this is not a ceteris paribus scenario. The minimum wage would change, causing another variable to change: the ideas of employees.
So the question is this: do we live in a world where increasing the minimum wage changes the ideas of workers so that they are more productive?
It’s an empirical question.
Now, I personally don’t think we live in such a world. (Or if we do, the gains in productivity are not sufficient enough to offset the additional costs of employment.) However, I didn’t arrive at those conclusions through a series of logical deductions. I’ve observed the world, and I don’t think that’s the one we live in.
Imagine that somebody were making an explicitly psychological case for raising the minimum wage. They wouldn’t claim, “We can increase employment by changing one-and-only-one variable: the minimum wage.” They’d claim, “We should increase the minimum wage because it changes other variables in the real world that increase worker productivity.” This is a coherent, empirical claim that you cannot refute by responding, “But ceteris paribus an increase in the minimum wage causes disemployment!”
If you think about it, the ceteris paribus parameter is odd in the first place. On the one hand, it’s important to help us understand cause and effect relationships in economics. On the other hand, it can lead to extreme myopia. Take an example outside of economics to see just how odd it is.
Imagine I’m a martial arts instructor, and I tell you, “When kids get promoted to a new colored belt, they perform at slightly higher levels because they think of themselves as being higher ranked.” This is something akin to the “winner effect” – the idea that winners gain confidence because they’re winners, which makes them more likely to win in the future.
Now imagine an apriorist walks in and says, “Bah! That’s logically impossible! Simply awarding a belt to a kid will not improve his martial arts skills because, ceteris paribus, there is no causal connection between wearing a different colored belt and gaining greater skill!”
It would be a bizarre, myopic argument indeed. Technically, the apriorist would be right. It’s not actually the different colored belt which makes kids better; it’s because their self-image changed. But this entirely misses the relevant phenomena we observe in the world.
(Personal anecdote: I actually experienced this in reverse when receiving my first black belt in karate. I was so focused on getting it that once I finally did, I found my skill level dropped. My body and mind got lazy because I wasn’t as focused on improvement anymore. The belt actually made me worse!)
It’s not hard to imagine a child improving his skill level because of a new belt. So why would it be hard to imagine an employee improving their skill level because of a raise? Yes, technically it’s not directly because of the belt or raise, but that’s a claim nobody is making. It’s a clear empirical possibility.
In What World?
Let’s take a step back to see the limitations of ceteris paribus reasoning.
Say you make a claim that “X is true.” Your claim should not change if you add the phrase, “in the real world.” So for example, if I say, “X is true,” I should be able to say that “X is true in the real world.”
What good is a proposition that claims “X is true, but not in the real world”?
For example, the claim “The minimum wage causes a disemployment effect” should be the same as “The minimum wage causes a disemployment effect in the real world.” Yet, the extreme apriorist’s claim is actually, “The minimum wage causes a disemployment effect ceteris paribus, but I can’t tell you what happens in the real world.”
Again, I think the claim is true, but it’s also neutered. If adding “in the real world” changes the validity of your claims, it should be a red flag. This is especially true for the many abuses of mathematics throughout various disciplines, but that’s an article for another time.
One more example of the abuse of apriorism, taken from a lecture I attended in 2011 at Mises University, delivered by Hans-Hermann Hoppe about Praxeology. He gives many examples of what he considers to be certainly-true apriorist claims about how the world works.
Take this one, for example:
“If we increase the amount of money without increasing the quantity of non-money goods, social wealth will not be higher, but only prices will rise.”
This is a great example of dropping the ceteris paribus condition. His claim is true, but only if we’re holding every other variable constant. In the real world, where multiple variables change – including variables that we don’t know are causally connected – it might be the case that an increase in the amount of money could increase the amount of social wealth. It just takes a bit of imagination.
Imagine that we live in a world where the most incompetent people are the wealthiest, and the most capable entrepreneurs are all stuck in poverty. Now imagine there’s a large monetary inflation – a bunch of new money is printed and given to the poor, competent entrepreneurs. Suddenly, they have new means available to them. They start undertaking projects, employing people, and end up creating wealth for society. Without this inflation, they wouldn’t have had enough capital to undertake the new projects. This scenario is essentially a redistribution of wealth from the incompetent to the competent.
So, by increasing the amount of money, social wealth could indeed increase. This is possible because the increase of money isn’t ceteris paribus. When the entrepreneurs received the new money, their behavior also changed, which caused an increase in total social wealth.
I’m not advocating for such an inflation. I’m simply giving an example of where dropping the ceteris paribus condition turns a true-but-neutered claim into a false-and-dogmatic one.
In Defense of Apriorism
Alright, after criticizing the abuse of apriorism, I want to defend the methodology, because I do think it plays a fundamental role in economic reasoning.
From my perspective, sound apriorist claims are rarely about states of the world. They are about our concepts. A careful rationalist will correctly point out that everybody brings pre-empirical concepts to the table before analyzing any data. These concepts are themselves not the subject of empirical inquiry; they are the lens through which we make sense of empirical data.
Every discipline has these presuppositions – including physics, mathematics, biology, etc. – though most people simply aren’t aware of them because they tend to be very abstract and philosophic rather than concrete and scientific.
For example, take the claim that “Humans act purposefully.” It might sound like an empirical claim – that we could go out and test whether or not humans are acting purposefully. But that’s not really correct. What careful thinkers like Ludwig Von Mises point out is that we interpret data about humans through the lens of purposeful action. When we observe humans, we presuppose a fundamentally different lens than when we observe billiard balls – that of purposeful action. It doesn’t make sense to say, “The billiard balls intended to go into their pockets after getting struck.” It does make sense to say, “Johnny intended to go to work at 3pm,” or “Johnny chose to take the train instead of the bus,” or “Johnny values classical music higher than electronica.”
The apriorist is examining the concepts within our own mental framework. When Johnny chooses Beethoven, we can also meaningfully say, “And he could have chosen otherwise, which means Johnny has a preference scale for music.” This preference scale isn’t measured or observed – it’s not really a thing-in-the-world. It’s an analytical construct that’s implied by our concepts about human action.
Take a fundamental economic law: the law of diminishing marginal utility. The first unit of a good gets employed to satisfy the most highly valued end; each additional unit will satisfy a lower-valued end. Is this an empirical claim? Not really. What we mean by “highest valued end” is precisely that it gets satisfied before other ends. It’s part of the definition – the pre-empirical conceptual lens. To say, “Johnny satisfied the lower-valued X prior to the higher-valued Y” is really to say that “Johnny actually valued X higher than Y.”
When I worked for FEE, I remember listening to a lecture by Israel Kirzner, who was a student of Mises. He and some fellow students apparently asked Mises, “But how do we know that humans act?” To which Mises replied, “We observe it.”
I think this is key. To say, “We observe it,” is to say, “Well, we’re guessing, based on the behavior of the objects we observe and based our own internal introspection, since we have special access to knowledge about what humans are. We’re interpreting human phenomena through the lens of purposeful action. It might very well be the case that humans do not act, but then you’ve got a great deal of explaining to do.”
In other words, the concept of “purposeful human action” is extraordinarily powerful – so powerful that it’s fused to the lens of virtually anybody analyzing human behavior. If the concept correlates to the world, we can use pure logical deduction to come up with a kind of aprioristic framework for analyzing human action. If the concept is false, then… perhaps everything is a great hallucination or something else wild and bizarre.
If you accept that humans act purposefully, then you are bound by a particular logical framework that rationalists have discovered. The framework is abstract and limited in scope, but it still exists and is fundamental.
There are other places where apriorism is fundamental to economic reasoning, though I will not spend much time covering them. For example, take the proposition that “Scarcity exists” – i.e. there aren’t enough goods for everybody to have all their ends satisfied. If this is true, it implies other aprioristic truths that we don’t have to observe in the world. If scarcity exists, then humans must choose which ends to satisfy with their scarce means. And if humans choose “this” over “that,” we can meaningfully talk about preference scales, the law of diminishing marginal utility, and if we’re careful, we can even deduce the general framework of the laws of supply and demand. Also note, the particular claim that “scarcity exists” is actually a claim about the world. It’s not just purely about our concepts.
Even in the earlier example about the minimum wage, apriorism and ceteris paribus reasoning can serve a valuable purpose. You could say, “To the extent that employment increases after a minimum wage hike, it is certainly not the case that the additional employment was solely caused by the cost of employment increasing.”
Again, it’s not a particularly relevant claim, since innumerable variables are always changing, but it is still true. Careful ceteris paribus reasoning allows us to hyper-focus on cause and effect relationships. What exactly causes what, and for what reasons? If my martial arts skill improved when I get a new belt, it must be the case that some other variable changed. Holding “everything else constant” in our minds greatly improves our ability to identify cause and effect in a complex world.
So, I think the most accurate approach to economic reasoning is a mixture of rationalism and empiricism. We all bring non-empirical conceptual frameworks to the table whenever we analyze a particular phenomenon. It’s valuable – even essential – to examine, explain, and flesh out the implications of these pure concepts. However, conceptual frameworks and extremely distant abstract truths do not tell you almost anything about what happens in the world. Economics is supposed to be about the world, not about our own minds, and the world is extremely complex. Multiple variables are changing every instant. The ceteris paribus parameter tells you essentially nothing about a world where more than one variable changes at the same time. Since that’s the world we live in, I suggest we rationalists stop abusing apriorism in economics.