Freedom, Counterfactuals and Economic Laws: Further Comments on Machaj and Hülsmann

[Full issue of the Quarterly Journal of Austrian Economics 20, no. 4 (2017)]

KEYWORDS: free choice, counterfactuals, economic laws

JEL CLASSIFICATION: B00, B40, B53

INTRODUCTION

In a series of articles written around the turn of the century, Guido Hülsmann has tried to answer one simple question: “How can we reconcile the idea that there are laws of human action, that manifest themselves in market prices and the structure of production, with the idea that there is also freedom of choice?” (Hülsmann, 2000, p. 48) He has addressed the question most extensively in his “Facts and Counterfactuals in Economic Law” (Hülsmann, 2003), but his distinctive approach is present in several other articles as well (Hülsmann, 1998, 1999, 2000, 2004). Moreover, the first explicit development of this insight is in Hülsmann (1999), in response to Caplan’s elaborate critique of the Austrian methodology, thereby indicating how crucial the issue is to praxeology as an intellectual enterprise, and to that extent to Austrian economics. Mateusz Machaj has commented on the very core of Hülsmann’s proposal (Machaj, 2012). As an epigraph, he chose a quote from Morpheus in The Matrix: “What happened, happened, and couldn’t have happened any other way”—pun intended or not against Hülsmann’s (metaphysical) libertarianism. This paper will briefly present Hülsmann’s main claim, Machaj’s comments, and offer a reply to those comments, further clarifying Hülsmann’s point.

HÜLSMANN’S CLAIM

Hülsmann considers the essence of scientific explanation to give “a law-based account of facts in terms of other facts, so scientists search for and study laws that exist among the things observed in our world. A thing X is scientifically “explained” if we can show that there exists a constant (e.g., causal) relationship between X and another thing Y.” (Hülsmann, 2003, p. 67) However, given the existence and nature of free human choice, no constant relationships seem to exist between a particular human choice or action, and, quite literally, anything else in the universe. There are no necessary constant relationships between anything a person does at a certain point in time, and anything preceding that choice in time—including all the past choices of that person—or anything in the world at that instant, including all facts about that person, that could explain the choice made. Sciences such as psychology, sociology, or first-hand acquaintance with a person can mitigate the strictness of that fact, but metaphysically it remains the case—unless one adopts some version of determinism. Economics, likewise, can mitigate the implications of (metaphysical) free choice by adopting some stylized ‘homo oeconomicus’ that works in most cases or is sufficiently useful for purposes of prediction and modelling. As Hülsmann puts it in his critique of neoclassical economics: “They want to analyze how people act as a corollary or sequel of given circumstances; that is, they want to explain human behavior in terms of other observable and introspectively knowable facts.” (Hülsmann, 1999, p. 5) The significance of Hülsmann’s proposal is that it grounds the existence of economic laws precisely on the metaphysical irreducibility of free choice with alternative possibilities, instead of trying to mitigate its implications:

I will argue that the bulk of economic laws are based on relationships that are contained within choice. The visible part of a choice, the realised alternative, brings an observable fact into being, for example, a walk in a park. This fact stands in certain essential relationships to the unrealised alternatives of the same choice, for example, staying home to watch TV, staying home to eat ice cream, etc. These unrealised alternatives are the other side of choice, its invisible part. They have no actual existence for the very reason that they are unrealised alternatives.” (Hülsmann 2003, p. 70, emphasis in original)

This proposal is a philosophical treasure trove—or hornet’s nest—but the main implication for economic methodology would be that economics can therefore provide us with strict counterfactual laws that do not need a ceteris paribus (CP) clause. If the essential relationships discovered by economists are between facts within choice, it is unconditioned by what is or is not happening ‘outside’ the choice. Economists can therefore not merely predict what will happen other things being equal, but what will happen regardless of other things—but compared to counterfactual, unrealized possibilities. That is, the seemingly problematic metaphysical status of human choice, as being unrelated to anything outside of that choice, has become the very foundation for the epistemological robustness of economic laws.

MACHAJ’S COMMENTS

Machaj critically engaged the very core of Hülsmann’s proposal (Machaj, 2012), ultimately defending a modified version of the ceteris paribus approach. He gives a stylized reconstruction of Hülsmann’s argument by introducing a simplified equation for capital accumulation, for illustrative purposes only, whereby K stands for the amount of accumulated capital, T for the influence of taxation, and the letters A, B, C, D for other factors affecting the amount of capital:

K = f(A, B, C, D, T)

(Machaj, 2012, pp. 445–446)

In the CP approach as understood by Hülsmann and Machaj, assessing the influence of taxation on capital accumulation would look thus:

K↓ = f(A, B, C, D, T↑)

(Machaj, 2012, p. 446)

That is, only if the other factors remain constant can we know that an increase in taxation will lead to a decrease in capital accumulation.

Machaj grants Hülsmann’s basic point that we can weaken this strict ceteris paribus rule towards a counterfactual rule, because even with other factors (A, B, C, D) influencing K, we can still know that the increase in T led to a lower level of K than otherwise would have been the case. Hence, even if all the other factors contributed towards a higher level of K, and a higher level of K than in the previous period was indeed obtained, still the level would have been even higher without the increase in T. But here is his worry:

There are numerous possible worlds in the counterfactual ladder—which possible world does Hülsmann advise us to hide behind the phrase “otherwise would have been”? Certainly he cannot have in mind the whole set of all the possible worlds that could have existed. (Machaj, 2012, p. 448, emphasis in the original)

He makes this point more explicit with the capital accumulation equation used above, asking us first to consider a case where the other factors changed as well:

K = f(A↑, B↓, C↑, D↑, T↑)

(Machaj, 2012, p. 448)

For such a case, the counterfactual law would be that the capital level is lower than in the counterfactual state of affairs in which taxation was not increased. But as Machaj points out, we simply do not know what that other scenario would have looked like—except for the level of taxation—so we do not know whether or not capital would have been higher. If we do not know what’s behind the question marks, we do not know what happens:

K′ = f(A?, B?, C?, D?, T)

(Machaj, 2012, p. 449)

Hence, he concludes:

The suggested answer would be that in the alternate scenario other factors have to change in the same way as in the actualized scenario of increased taxation: K′ = F(A↑, B↓, C↑, D↑, T) [...] Factors do not have to stay the same, but in counterfactual scenarios they have to change in the same way as in the factual scenario. That is why the counterfactual approach can be seen as a broader ceteris paribus assumption. (Machaj, 2012, p. 449, emphasis in the original)

To sum up, Machaj grants to Hülsmann that we can go beyond a strict CP rule claiming that economic laws only hold between different points in time where all other factors remain the same, but that Hülsmann’s counterfactual approach only holds if we compare the factual scenario to a counterfactual scenario in which the other factors changed in the same way as in the factual scenario. That is, the CP rule still holds, but between two possible scenarios at one point in time (one with, one without a tax increase) instead of between two scenarios at different points in time.

A COMMENT ON MACHAJ’S COMMENT

However, Hülsmann’s point is that the counterfactual law quite literally does apply to “the whole set of all the possible worlds that could have existed.” It does not matter at all what is behind these question marks in the counterfactual case, and hence these counterfactual laws are indeed unconditioned by evolutions in these other factors. Hülsmann’s point is precisely that no matter what one fills in for these question marks, in each and every counterfactual state of affairs, K’>K.

To continue on the toy-model, let

(K1, K2, K3, …, Kn)

stand for all possible states of affairs where taxes were increased, with the four other variables varying in all possible ways. In each and every one of these cases, there is a corresponding scenario

(K’1, K’2, K’3,…, K’n)

in which taxes would not have been increased and where capital accumulation therefore would have been higher. Hence, no matter in which of the possible increased-taxation scenarios we end up, the counterfactual scenarios in which taxes would not have been increased are ones with a higher level of capital accumulation

(K’1>K1, K’2>K2, K’3>K3, ... , K’n>Kn),

or:

∀ i ∈ {1, 2, 3, …, n) : K’i > Ki

The CP rule still holds between any of these pairs, with the only changed variable being T, whereas all others remain constant. Hence, looking backwards in time, one can indeed point at the one factual scenario (e.g. K3) and say that in the counterfactual case (K3’), capital would have been higher.

However, when choosing for a higher level of taxation, it is not only not yet known in which scenario one is, it is strictly speaking not yet settled—it is still metaphysically ‘open’—which counterfactual scenario will obtain, given the countless free choices of other persons affecting the course of events. The validity of the law therefore does not depend upon that one CP pair of the factual and counterfactual scenario (K’3 > K3), but upon the entire range of possible scenarios for which (K’i > Ki) holds at the very moment of choosing. The validity of the purported economic law holds regardless of—unconditioned by the fact—which scenario eventually obtains, and the case for Hülsmann’s strong claim still stands.

CONCLUSION

Hülsmann’s proposal is rife with philosophical assumptions, and implications for economic methodology. Carefully unpacking them all will require a lot more work and cooperation between economists and philosophers, but the stakes are high—both for praxeology and economics if they claim to be a science of (free) human action, and for philosophy for seeing how a science of free action is at all possible. After the stimulating challenge by Machaj, this paper offered a further contribution to clarify and strengthen that project.