Austrian Economics and Compatibilist Freedom

Abstract

The present paper probes the relation between the metaphysics of human freedom and the Rothbardian branch of Austrian economics. It transpires that Rothbard and his followers embrace metaphysical libertarianism, which holds that free will is incompatible with determinism and that the thesis of determinism is false as pertaining to human action. However, as we demonstrate, their economics with its reliance on value scales requires for its tenability compatibilist freedom. Moreover, we attempt to show that the notion of value scales (or preferences) postulated by them implies that value scales are determinative of choices people make. We contend that it is for this reason that the said Austrians should jettison their metaphysical libertarianism.

1 Introduction

Whether human agents are endowed with metaphysical libertarian freedom¹ seems to be no petty matter. Although the present paper focuses on the type of human freedom purportedly cohering with the overall conceptual framework of Austrian economics, we cannot but notice that the metaphysics of human freedom matters most in moral philosophy, and especially in assigning moral responsibility. Before Frankfurt’s (1969) seminal paper, it was almost universally accepted in the philosophical literature that moral responsibility requires metaphysical libertarian freedom (hereinafter referred to as MLF); that is, the ability to do otherwise, ceteris paribus. And it must be granted that this dominant pre-Frankfurt view has some intuitive appeal. After all, at least prima facie, it seems that, given determinism, an ought-implies-can principle² would be violated. For, under determinism, if a person fails to discharge his duty and thus violates somebody’s right, the former does so necessarily. But, if he could not have done otherwise, then a fortiori he could not have discharged his duty, everything else equal. In other words, there is no possible world sharing the history of the actual world (up to the moment of him actually failing to discharge his duty) and its laws of nature in which he discharges his duty. And if so, it appears to be odd to hold people responsible for the actions that were unavoidable, keeping everything else fixed.

And indeed, even after Frankfurt (1969), there is a strand of thought associated for example with Pereboom (2001; 2014) or Caruso (2021) arguing that the truth of determinism rules out moral responsibility in a basic-desert sense in which “the agent would deserve to be blamed or praised just because she has performed the action, given an understanding of its moral status, and not, for example, merely by virtue of consequentialist or contractualist considerations” (Pereboom 2014, 2). It is also Parfit (2011, 265) who claims that, given the absence of MLF, we cannot be responsible for our acts in a way that could make us deserve to suffer. However, the controversy around the notion of basic desert aside, we cannot but stress that, generally speaking, compatibilism about moral responsibility and determinism is the dominant view nowadays.³

It might seem that, compared to the problem of compatibility of moral responsibility with determinism, settling the question of whether an economic choice is compatible with determinism is of lesser practical importance. However, the problem of metaphysical freedom is no less pressing in economics than it is in moral philosophy, for the obvious worry (whether well-warranted or not) is that there would be no economics were determinism to hold true.⁴ After all, economics is a science of human choices and it appears that if certain choices (i.e. the ones that will actually and necessarily be made under determinism) were inevitable, this ipso facto would nullify their existence. In other words, there is a concern that the concept of choice would not apply under determinism, which would in turn entail the inapplicability of economics.⁵

However, we hold that the worry that the truth of determinism would drain economics and Austrian economics in particular of its significance is unfounded. Even more, we argue that adopting MLF, as some Austrian economists do, does not cohere with subscribing to the conceptual framework of Austrian economics and it is precisely the compatibilist freedom (i.e. the one compatible with determinism) that Austrian economics requires.⁶ More specifically, our point is–contra quite a few Austrians with a notable exception of Ludwig von Mises ([1949] 1998) or Hayek ([1960] 2011)–that Austrian economics with its reliance on value scales and the relation of strict preference, which economic agents act on, coheres only with the compatibilist sort of freedom. However, let us not precipitate things as the technicalities of our solution will play out in Sect. 2. For the time being, let us provide some textual support for Austrians’ avowed sympathies with MLF.

By far the most eloquent expression of the adherence to MLF among Austrian economists is to be found in Rothbard (2011b, 5–6):

Surely we can, at the very least, tell the determinists to keep quiet until they can offer their determinations—including, of course, their advance determinations of each of our reactions to their determining theory. But there is far more that can be said. For determinism, as applied to man, is a self-contradictory thesis, since the man who employs it relies implicitly on the existence of free will.⁷

The last sentence of the above excerpt is the most telling: Rothbard makes it explicit that one cannot coherently state the thesis of determinism since the very statement thereof putatively presupposes the existence of libertarian free will.⁸ But if so, we can easily attribute to Rothbard the belief in MLF as this author clearly subscribes to the existence of free will which simultaneously rules out determinism, the position being nothing short of metaphysical libertarianism.

Hülsmann (2003, 7) is another prominent Austrian economist who is most explicitly sympathetic to MLF:

Irrespective of how small the scope for choice is, within this scope, choice is to some extent undetermined and undeterminable; here, the human being is self-determining. To sum up, because of the existence of choice, it is impossible to give a full explanation of human behaviour in the mere light of any other event, or a combination of other events.

If we are to take Hülsmann’s view at its face value, we are to conclude that since human behavior cannot be explained “in the mere light of any other event”, then a fortiori human (purposeful) behavior cannot be explained by such mental events (or states) as the occurrences of certain desires and beliefs, the view we are about to deem irreconcilable with Austrian economics in the forthcoming section.

It is also Block (2015, 6) who unwaveringly subscribes to MLF, while claiming that it is only the libertarian free will that is compatible with Austrian economics, whereas determinism is not:

Another difficulty for the determinist philosophy is that it runs counter to Austrian economics. One of the key elements of this school of the dismal science is methodological dualism: that a different perspective must be brought to bear in the study of human action, on the one hand, and the physical world, on the other. Since under determinism man is merely in effect a machine, contrary to Austrianism, there is a need for a separate method to study these two very different aspects of reality.

Elsewhere, van Schoelandt et al. (2016, 91) maintain that:

[T]he Austrian tradition, in contrast to theories in the grip of scientism and adopting an overly mechanistic view of human behaviour, sees humans as purposeful choice makers, responding not merely in instructive ways to external stimuli, but through their own desires and preferences in ways that are largely unpredictable.

But then again, as we are going to claim, holding that humans respond “in ways that are largely unpredictable” to their respective “desires and preferences” is to distort the Austrian-spirited view on the relation between choice and preference.

Eventually, most recently, Wiśniewski (2009) took a stand on the problem of free will vis-à-vis determinism. This author (2009, 1) argues that “the nomenclatural affinity” of metaphysical libertarianism “with political and economic libertarianism […] is by no means accidental”. Moreover, he (2009, 1) decidedly states his case against compatibility of determinism with laissez-faire in the following manner:

I remain convinced that as soon as one grants that every human decision can be traced back to factors beyond one’s control (e.g., genetic makeup, environmental influences, personal upbringing etc.), the notions of sovereign choice and personal liberty become empty.

The above statement coupled with the fact that Wiśniewski is a declared Austrian economist leaves no doubt as to the way we should interpret his conception of “sovereign choice”: that it is incompatible with determinism.

By contrast, and as promised, we are about to argue–contra some prominent Austrians–that Austrian economics with its overall conceptual framework and especially with its reliance on value scales and with its conception of action as reflecting strict preference⁹ should embrace compatibilist freedom (i.e. the one compatible with the truth of determinism) and it is precisely the exposition of key features of Austrianism and the statement of our argument that we are now turning to.

2 Why Austrians’ Belief in MLF Is Incompatible with Their Economics

The purpose of the present section is to establish the incompatibility between Austrians’ avowed belief in MLF and the sort of economics (i.e. Austrian economics) they subscribe to. In order to do this, we should have a closer look at the notions of value scales and strict preference employed by Austrians, for it is there that the key to our compatibilist solution lies. First and foremost, what needs illuminating is the Austrian view on the relation between value scales (or preferences for that matter) and actual choices. This view is best expressed in Rothbard ([1962] 2009, 6):

Thus, suppose that Jones ranked his alternative ends for the use of an hour of time as follows:

• (First) 1. Continuing to watch the baseball game.

• (Second) 2. Going for a drive.

• (Third) 3. Playing bridge.

This was his scale of values or scale of preferences. […] Suppose now that he is allocating two hours of his time and can spend an hour on each pursuit. If he spends one hour on the game and then a second hour on the drive, this indicates that his ranking of preferences is as above. The lowest-ranking end–playing bridge–goes unfulfilled.

And elsewhere, Rothbard ([1956] 2011a, 290) expresses the very same point explicitly in terms of demonstrated preference:

Man has means, or resources, which he uses to arrive at various ends; these resources may be time, money, labor energy, land, capital goods, and so on. He uses these resources to attain his most preferred ends. From his action, we can deduce that he has acted so as to satisfy his most highly valued desires or preferences. […] The concept of demonstrated preference is simply this: that actual choice reveals, or demonstrates, a man’s preference; that is, that his preferences are deducible from what he has chosen in action. Thus, if a man chooses to spend an hour at a concert rather than a movie, we deduce that the former was preferred, or ranked higher on his value scale.

Now, how should we conceive of the above fragments? First, let us note that Rothbard maintains that actual choices people make are indicative of their respective value scales. Moreover, he seems to hold (however implicitly) that value scales exist independently of choices; after all, he concludes that once the economic actor picks up option 1, that is, continuing to watch the baseball game, his option 3 (playing bridge) “goes unfulfilled”. So, he seemingly grants the possibility of the existence of some preferences which the actor does not actually choose to act on. And with preferences being independent of choices, we can infer the former from the latter.¹⁰ Furthermore, the said inference seems to be all the more possible (if not apodictically valid) when we assume along with Austrians that people choose the most preferred option, all things considered.¹¹ And it must be borne in mind that Austrian economics by virtue of its nature strips the human action down to its bare essentials (e.g. Mises [1949] 1998; Rothbard [1962] 2009). That is to say, Austrians are not interested in which specific reasons were relevant to the actor’s final resolution as to which course of action to undertake or indeed which specific reason tipped the balance in favor of the actual action performed. For instance, suppose that subject S buys a blue shirt instead of a white one. Are Austrians entitled to infer that S bought the blue shirt just because he preferred this very shirt to its white equivalent? Not quite. What they can at most infer is that S preferred to buy the blue shirt rather than the white one, all things considered. Perhaps, S knew that his wife would rather have him dressed in blue. S can even sincerely declare that he prefers white shirts to blue ones himself.¹² However, this declaration is mute on S’s wife’s preferences as to the color of S’s shirts, which in turn might count as a reason when it comes to S’s choosing between shirts. So, whatever the reason (S’s wife preferences or anything else for that matter) that tipped the balance in favor of S’s buying the blue shirt, the very fact that S did so indicates that S preferred to buy the blue shirt rather than the white one (or indeed, rather than to do anything else), all things considered.

Still, there is another aspect of Austrian economics around which our argument revolves. The aspect in question is the Austrian view on indifference. The dominant view within the Austrian school is that man acts on strict preference¹³ but never on indifference.¹⁴ Or, positively speaking, man’s actual choice reflects the fact that he strictly prefers the option actually chosen to other actions he deemed feasible (including inaction). The pertinent fragment from Rothbard¹⁵ ([1956] 2011a, 304–305) runs as follows:

Indifference can never be demonstrated by action. Quite the contrary. Every action necessarily signifies a choice, and every choice signifies a definite preference. Action specifically implies the contrary of indifference. The indifference concept is a particularly unfortunate example of the psychologizing error […]. If a person is really indifferent between two alternatives, then he cannot and will not choose between them.

To put it more formally, if the economic actor S is indifferent between x and y, then he or she does not choose between these two options. And conversely, if S chooses x over y, then he or she has not been indifferent between x and y.

Now, we believe that this exposition of the two crucial aspects (i.e. actors’ value scales being aligned with their respective choices and man acting on strict preference, with the latter relegating indifference out of the realm of human action) of Austrian economics suffices to allow us to make our compatibilist point. So, let us stipulate that in the actual world W a certain economic actor S has the following value scale and that he has 2 hours to economize, with each end (1–3) taking an hour to be satisfied.

Value scale v₁:

• 1. Resting on the couch.

• 2. Walking in the forest.

• 3. Reading a book.

Note that the above postulated value scale is just a slight variation on the scenario adduced earlier. So, let us mutatis mutandis apply Rothbard’s reasoning to S’s preference scale. Then, we can say that if S spends an hour on resting on the couch and subsequently an hour on walking in the forest, this indicates that S’s value scale is as above and that the lowest-ranking end–reading a book–goes unfulfilled. At this point, it would be also worth noting that the inference running from actual actions (or choices) to underlying value scales is of epistemic nature. However, genetically speaking, it is only a combination of desires (represented on value scales) and beliefs (but not desires alone) that issue in action. Just to illustrate the point further, if our S in W did not believe that there was a forest nearby he or she could access, he would never attempt to satisfy his end 2 (i.e. walking in the forest). That is to say, his desire to walk in the forest in and of itself was powerless to cause him to act. However, that very desire coupled with relevant beliefs already can issue in S’s action.

Having elucidated the causal role of beliefs, let us consider a slight variation on S’s behaviour in W described above. Suppose now that in a close possible world (W*) S economizes his two hours differently. Instead of prioritizing resting, S spends his first hour on walking in the forest and only subsequently he or she is resting on the couch for an hour. On Austrian grounds, this differential action would be aptly accommodated by a different value scale informing S’s action in W*. Specifically, S’s action in W* would indicate that his value scale therein is as follows¹⁶:

Value scale v₂:

• 1. Walking in the forest.

• 2. Resting on the couch.

• 3. Reading a book.

It therefore follows that for any world W and any agent S it is impossible for S to choose anything but the highest ranked option in W on Austrian grounds. For on this view any actual choice indicates the fact that the actor picks up the most preferred option available, thus maximizing his welfare at least in expectation. So, to come back to our example, given that the value scale v₁ operates in W (i.e. resting on the couch is the most valued end for S), Austrian economics does not allow us to envisage a possible world in which S chooses anything but the most preferred option,¹⁷ for if S were to choose in W* to spend his first hour walking in the forest (i.e. a less preferred option in W), this fact would automatically imply that walking in the forest is the most valued end for S in W*. But if so, then preferences appear to be determinative¹⁸ of choices under Austrian economics. And if preferences are determinative of choices, then there seems to be no room for MLF. In other words, given the actor’s actual choice at a given time, he could not have done otherwise at that time, with his or her value scale fixed.

As it seems, probably the most obvious escape route open to MLF-inclined Austrians is to claim that human preferences are indeed unstable (Rothbard 2011b, 295). Let us consider two possible interpretations of this claim. The most straightforward one (interpretation 1) amounts to the repudiation of the so-called assumption of constancy, which has it that people’s preferences are stable over time.¹⁹ So, technically speaking, Austrians might argue that at time t₁ one’s choice may be informed by the value scale v₁, whereas at time t₂ it might be informed by the value scale v₂, where v₁ and v₂ are not identical. So certainly, a human actor is not stuck with one and the same value scale over time. However, this does not affect our compatibilist point at all. Granted, the value scales might be different but this may simply mean that different value scales guide the actor’s choices at different times; that is, it is easily conceivable that a value scale v₁ guides the actor’s action at t₁, whereas at t₂, v₂ does so, where v₁ ≠ v₂. Yet, this feature of Austrian economics goes no distance towards establishing MLF. Value scales may be different in the sense that they may vary over time. Still, upon the reading under consideration, what their instability over time implies at most is that a particular value scale determines a choice at a particular time. But this implication is simply indistinguishable from the compatibilist position. Hence, interpretation 1 does not pose any threat to the kind of compatibilism between (free) choice in Austrian economics and the truth of determinism we advance.

So perhaps there is a way of reinterpreting the instability of value scales so that MLF can prove to be reconcilable with Austrianism. More interestingly, Austrians may try to hold that for each time (t), an economic actor is free to choose a value scale guiding him at t. But note that this position, most crucially, invokes the notion of choice. The fact that this time choice is presumed to operate among value scales rather than within them is irrelevant. Remember, for Austrians, any actual choice is evidential of the underlying preference. So, in the end, an actual choice among first-order value scales must imply the existence of a second-order value scale (or a second-order preference) informing that very choice (for an interesting objection suggesting a possibility of the free creation of value scales, which would allegedly not involve a choice behavior, see footnote 23). For instance, if an economic actor was indeed choosing between v₁ and v₂ and eventually opted for, say, v₁, this would ipso facto indicate that his or her second-order value scale ranked v₁ higher than v₂.

It is clear to see that the appeal to second-order preferences brings about the element of recursiveness here. For Austrians might now claim that there can obviously be various second-order preferences, which would in turn open the following two logical possibilities:

• 1) A second-order preference might be determined itself. If so, it would in turn determine a first-order value scale, which would eventually determine the pursuit of a particular highest valued end figuring in the said first-order value scale. It is to be noted that this possibility does not pose any threat to our compatibilist contention that preferences determine choices. In the end, it was a second-order preference that ultimately determined (via a first-order preference) a particular choice.

• 2) An economic actor might be free to choose among second-order preferences. But then again, once the actor chooses a given second-order preference, this fact indicates that there was a certain third-order preference accounting for that choice.

• 3) And so on and so forth ad infinitum. Hence, Austrians would have to either accept the fact that preferences of whatever order are determined at a given time (interpretation 1 expressing the gist of compatibilism), which eventually determines the choice of pursuing a particular material end through the direct influence of a first-order value scale or face an explanatory regressus ad infinitum, with either way falling short of establishing MLF on Austrian grounds.²⁰

To bolster our point further, let us now take the idea of MLF seriously and study what sort of implications we can draw therefrom. If any of them proves to be inconsistent with Austrian economics, this would provide us with another reason why Austrians should adopt compatibilist freedom. So, consider again the following value scale taken from Rothbard ([1962] 2009, 6).

Value Scale v₃:

• 1. Continuing to watch the baseball game.

• 2. Going for a drive.

• 3. Playing bridge.

Now, we know from the earlier analysis that if S does actually spend his first hour on watching the baseball game and only the second hour on going for a drive, his value scale is indeed the one right above. So, for all we know so far, in W the agent S in fact chooses to spend his first hour on watching the baseball game and his second hour on going for a drive, from which actions we inferred the value scale v₃. That much can be granted just for the sake of argument. But now we should investigate what are the demands of MLF in the scenario under scrutiny. Remember, metaphysical libertarianism envisages the freedom of an action in a characteristically demanding manner; that is, for agent S to perform the act x freely implies that S can do otherwise, everything else equal. Or to put it in the vernacular of possible worlds, an act x is performed freely by the agent S at time t only if there is a possible world where everything is the same up until t and S does something other than x at t. Given this, let us spell out what MLF predicts in the Austrian scenario considering the actor whose actual choices are indicative of value scale v₃.

Metaphysical libertarians would have it that if S in W does indeed freely choose to spend his first hour on watching the baseball game, he ipso facto could have chosen to spend his first hour on going for a drive, everything else equal. However, and crucially, the condition of everything else equal implies–among other things–the value scale equal.

Therefore, if S could have chosen to go for a drive first, everything else equal, then a fortiori, S could have decided to do so, with S’s value scale fixed. In other words, MLF commits us to the view that if S freely chooses to first watch the baseball game and only then to go for a drive (which is informed by value scale v₃), this very fact implies that there is a possible world W* in which the very same agent chooses to first go for a drive and only then to watch the baseball game, with, most crucially, S’s choices in W* being still indicative of value scale v₃– after all, everything but choices are equal across the two worlds under scrutiny. But this in turn would mean that the situation in W* is as follows:

Value Scale v₃ still applies:²¹

• 1. Continuing to watch the baseball game.

• 2. Going for a drive.

• 3. Playing bridge.

However, in this scenario S chooses to first go for a drive and only then to watch the baseball game. But then, no longer do choices indicate preferences. Even more, in W*, S’s choices seem to run counter to his preferences, a sheer impossibility on Austrian grounds. And since MLF implies by definition an ability to do otherwise, ceteris paribus, and because an ability to do otherwise, ceteris paribus, implies that there must be a possible world in which actor’s choices run counter to his value scales therein, which is impossible on Austrian grounds, we end up with the conclusion that MLF is incompatible with Austrian economics. We cannot retain the two at the same time: either actors’ choices are aligned with their respective value scales, in which case we give up MLF, or we adopt MLF but disavow the characteristically Austrian commitment to a harmony between actions and preferences.

This in turn gives rise to yet another problem. As we remember, taking MLF seriously led to the possibility that S acts differentially between the worlds (W and W*) without any accompanying alteration in S’s value scale. Specifically, what we discovered is that, according to metaphysical libertarianism, S can spend his first hour either on watching the baseball game (the actual world W) or on going for a drive (a close possible world W*), with one and the same value scale (v₃) applying across these two worlds. But if so, then what explains S’s differential action across these worlds? Ex hypothesi, W and W* are identical in all respects but in S’s choice. So, there seems to be nothing that can explain why during his first hour, S watches the baseball game in W and goes for a drive in W*. Hence, it seems that whether S does one thing in W and another thing in W* ultimately depends on luck.²² However, as we believe, economists of any persuasion would like to have at their disposal a contrastive explanation of why a given actor chooses one thing rather than another. In fact, it is precisely for that reason that economists appeal to actors’ preferences or value scales for that matter. Hence, if there is nothing that can account for S’s watching the baseball game in W and S’s going for a drive in W*, this must count as a major embarrassment for Austrians, which is another reason that speaks against embracing MLF.

To put it in still another way, if we take MLF seriously, then during his first hour, S acts differently in W* from the way he acts in W. And what is it that explains the differential action of S across these two worlds? It seems that MLF-inclined Austrians are caught in a particularly thorny dilemma: either (a) they have to concede that there is literally nothing that accounts for S’s cross-world differential action or (b) they have to admit that W and W* after all differ with respect to some external circumstances since S’s preferences (value scale) are held fixed ex hypothesi. If they decided on option (a), they would be deprived of a contrastive explanation of the cross-world difference in S’s action. If, on the other hand, they picked up option (b), they would indeed have a contrastive explanation (i.e. W and W* differ in some external respect and that is why S’s action varies accordingly across the two worlds) but at the cost of leaving their metaphysical libertarianism unsupported. After all, it would then be some external factor that would be doing the explanatory work regarding S’s differential action rather than S’s free will. In other words, appealing to some difference between W and W* to explain why S acts in W differently to the way he does in W* can be in no way supportive of metaphysical libertarianism since, as we remember, this doctrine has it that at least sometimes an agent (S) acts freely, which, technically speaking, implies that there is such a time t at which there are two possible worlds (one actual and the other merely possible) which share literally everything apart from S’s action in them at t.²³

Faced with this dilemma, MLF-inclined Austrians might try to save their commitment to Austrianism by arguing that what explains S’s differential action in the above scenario is the fact that S was in fact genuinely indifferent between watching the baseball game and going for a drive. In other words, they might claim that there is such a value scale (v₄) that can account for the fact that S watches the baseball game in W and goes for a drive instead in W*.

Value Scale v₄:

• 1. Continuing to Watch the Baseball game or Going for a Drive²⁴

• 2. Playing Bridge

And indeed, the dominant Austrian view on indifference²⁵ has it that if the agent S is indifferent between doing one thing or another (or between employing one means or another, given his end), the agent may do one thing in the actual world (W) and another thing in a close possible world (W*), with the same value scale (v₄) being operative. Given this, Austrians might hold that it is precisely v₄ that does explain S’s cross-world differential action.

For while on v₃ S’s differential action would indeed be impossible to account for in terms of some contrastive explanation, on v₄ it is relatively easy to explain; that is, in both worlds S chose option 1 over playing bridge but since he was indifferent between the disjuncts within his most preferred option, it transpired that in W he watches the baseball game whereas in W* he goes for a drive, with the condition of everything else being equal up to the moment of choice preserved. Hence, it seems that S might have exercised his libertarian freedom without running into any problems mentioned above.

Granted, on Austrian grounds, if S is indifferent between watching the baseball game and going for a drive, it would indeed follow that it might as well be the case that S spends his first hour on watching the baseball game in W and on going for a drive in W*, with S’s value scale (v₄) held fixed. Unfortunately, this would in turn imply that neither does S choose to watch the baseball game in W nor does he or she choose to go for a drive in W* since there is no choice under indifference.²⁶ Hence, even if Austrians were able to explain different behaviours of S in W and W* by resorting to the concept of indifference, they would still be unable to explain S’s different choices. For in order to say that S chooses differentially in W and W*, Austrians would have to admit that S’s behavior is informed by strict preference between watching the baseball game and going for a drive. Since indifference excludes this possibility, Austrians are barred from appealing to this concept insofar as they are interested in explaining economic actor’s choices and actions. Indeed, it would be a very strange sort of economics that would have it that economic actors do not value their ends (and derivatively: their actions) differentially; that is, for any agent S and any action x, S might as well do something other than x and be equally well off–the most absurd conclusion. Given that, on Austrian grounds, once we admit that agents do choose at times, we cannot appeal to indifference any more. So, technically, if S chooses option x over y in W and option y over x in W*, this differential action must be explained by a corresponding shift in relative evaluations of these two options; that is, S prefers x to y in W but when it comes to W*, S ranks these two options in the opposite order.

Moreover, resorting to indifference suffers from infinite regress. Note that metaphysical libertarianism predicts that, given v₄, S could have still chosen the less preferred option (2), that is playing bridge, everything else being equal up to the moment of decision. However, S’s choosing in W* to play bridge rather than to watch the baseball game or go for a drive would immediately imply that this option is strictly preferred to anything else, which would in turn render v₄ inoperative in W*. Most certainly, Austrians do have resources to accommodate the fact that there is a possible world in which S plays bridge instead of continuing to watch the baseball game or going for a drive. So, in order to account for the libertarian possibility of S’s playing bridge instead of continuing to watch the baseball game or going for a drive, Austrians would have to postulate S’s indifference between all these three options.

Value Scale v₅:

• 1. Continuing to Watch the Baseball game or Going for a Drive, or Playing Bridge

• 2. Playing Chess

But now, metaphysical libertarianism would predict that, given v_5,the economic actor S could still have played chess. In other words, there was a possible world in which S plays chess, everything else equal. However, this statement could only²⁷ be saved by postulating that in fact it was v₆ that could account for this possibility.

Value Scale v₆:

• 1. Continuing to watch the baseball game or going for a drive, or playing bridge  or playing chess.

• 2. Watching a film (or whatever else).

But now we can appreciate the gravity of the problem. It is clear to see that appealing to indifference cannot explain differential choices across worlds. Granted, resorting to indifference is able to explain why S does one thing in W, another thing in W* and still some other thing in W** but at the cost of S being barred from choosing between the said options. And more critically, it seems that once we employ indifference to accommodate S’s cross-world differential behavior, the implication is that whatever S happens to do across worlds, he is always equally well off. In other words, even though his behavior may vary considerably across worlds, his welfare would be fixed²⁸–the most unwelcome consequence for economics.

Therefore, it seems that Austrians have little (if anything) to appeal to in order to save their proclaimed metaphysical libertarianism. The characteristically Austrian rejection of the assumption of constancy of preferences poses no problems to the compatibilist view, whereas citing the existence of second-order preferences catches MLF-oriented Austrians in a predicament, for then they either have to face an explanatory infinite regress or admit that n-order preferences are determined and they eventually (regardless of their order) issue in an action, with neither option being palatable to metaphysical libertarians. Moreover, once we take MLF seriously and spell out its implications, we–rather unsurprisingly–discover that the adoption of MLF runs into the problem of luck. That is to say, if Austrians accept that sometimes an economic actor chooses freely (in the metaphysical libertarian’s sense), they are at a loss to explain the cross-world difference in the actor’s behavior, for they can neither resort to the actor’s value scale varying across the worlds or to a cross-world difference in some external conditions.²⁹ And as long as we speak of two distinct choices made by S (i.e. one option being chosen in W and the other in W*), Austrians are barred from explaining this sort of cross-world difference in terms of S’s indifference between the said two options.

Having thus made a preliminary case for compatibilism as a metaphysical stance that Rothbardians seem to be logically committed to adopting, let us now proceed to consider whether there are other prominent Austrians who already embrace compatibilism or at least have a good reason to do so.

3 Compatibilist Freedom vis-à-vis Other Austrians

If the arguments made thus far count for something, it seems that the branch of Austrian economics which relies on value scales, with the value scales being evidenced by actual choices, indeed has a rather compelling reason to embrace compatibilist freedom. However, as our reasoning targeted Rothbardians, the question arises whether there are other prominent Austrians for whom a compatibilist notion of freedom might well prove to be attractive.

Starting our investigations with Mises would be much in order. As aptly put by Linsbichler (2017, 14–15), Mises “denies individuals a free will in the metaphysical sense”, while “explicitly acknowledge[ing] the possibility of absolute determinism on a purely physical or physiological basis.” However, there are places in which Mises makes an even stronger case for the sway of determinism over human action. Says this author: “To be sure, even the valuations of individuals are causally determined” (Mises [1933] 2003, 129).³⁰ Elsewhere, Mises ([1957] 2007, 77) avers that “[t]he determinist is right in asserting that everything that happens is the necessary sequel of the preceding state of things.” Given the above citations, it appears as though it is no stretch to classify Mises as a determinist.³¹ Still, being a determinist does not by itself makes one a compatibilist, for one might as well believe that the truth of determinism is indeed incompatible with free will. Therefore, we should make sense of Linsbichler’s assertion to the effect that Mises “denies individuals a free will in the metaphysical sense”, for if he denies individuals a libertarian free will, he might as well be a compatibilist. Specifically, what would make Mises a compatibilist is his allowing the possibility of the economic actor acting on his or her reasons, desires, preferences, beliefs or what have you. As we stressed time and time again, compatibilist freedom reduces to man’s having power to do what they happen to want to do. That is to say, compatibilist freedom requires one’s utmost desires or preferences (coupled with relevant beliefs) to be causally efficacious even if they were to be determined themselves. There is indeed ample evidence that Mises might be read along compatibilist lines. For instance, says Mises ([1949] 1998, 97):

Action is an attempt to substitute a more satisfactory state of affairs for a less satisfactory one […]. A less desirable condition is bartered for a more desirable. What gratifies less is abandoned in order to attain something that pleases more. That which is abandoned is called the price paid for the attainment of the end sought.

Now, crucially, Mises does not say that action might involve substituting a less satisfactory state of affairs for a more satisfactory one, something MLF would permit by all means. Rather, he seems to imply that actors via their action attempt to bring about the most preferred state of affairs. But if so, this position is nothing short of compatibilism avowed here. To put it succinctly, actors strive to attain their highest ranked ends by dint of their causally efficacious desires (and beliefs) even if the latter are causally determined.

Hence, metaphysically speaking, one can rather unproblematically conceive of Mises as a compatibilist. However, one caveat is important at this point. Although, metaphysically speaking, Mises can be undoubtedly ranked as a determinist, he remains an epistemological indeterminist. That is to say, even though all the events (including human action) might indeed be causally determined, future events are unknowable in principle. Mises ([1957] 2007, 77) indeed acknowledges that “[th]e choices a man makes are determined by the ideas that he adopts”. However, he also believes that the thesis of determinism is “considerably weakened by the fact that nothing is known about the way in which ideas arise.” Allegedly then, metaphysical determinism does not automatically translate into epistemological determinism since we do not know the way in which ideas governing our actions are formed.³² In other words, because we do not know which ideas are going to determine our actions, our future remains epistemically indeterminate.

We submit that Hayek might be equally plausibly ranked as a compatibilist.³³ It is especially his Constitution of Liberty that contains most pertinent and telling assertions. For instance, Hayek ([1960] 2011, 136) says that “the whole suggestion that “free” in any relevant or meaningful sense precludes the idea that action is necessarily determined by some factors proves on examination to be entirely unfounded.” Incidentally, it is worth noting that whereas Mises allowed for “the possibility of absolute determinism”, Hayek’s thesis is even stronger. For the latter author claims that “the conception of responsibility rests, in fact, on a determinist view.” In other words, while Mises apparently regards free will (i.e. having one’s actions guided by one’s ideas) as simply compatible with the truth of determinism, Hayek conceives of responsibility assignments as requiring determinism.³⁴ Generally, Hayek’s idea of responsibility assignments is forward-looking. That is, he views the institution of responsibility as serving to influence the prospective behaviour on the part of human agents who are eligible for responsibility ascriptions in the first place. To this effect, says Hayek ([1960] 2011, 137): “It is just because there is no separate “self” that stands outside the chain of causation that there is also no “self” that we could not reasonably try to influence by reward or punishment.” Hayek’s compatibilism is probably most plainly expressed in the following fragment:

If we say that a person is responsible for the consequences of an action, this is not a statement of fact or an assertion about causation. The statement would, of course, not be justifiable if nothing he “might” have done or omitted could have altered the result. But when we use words like “might” or “could” in this connection, we do not mean that at the moment of his decision something in him acted otherwise than was the necessary effect of causal laws in the given circumstances. Rather, the statement that a person is responsible for what he does aims at making his actions different from what they would be if he did not believe it to be true. We assign responsibility to a man, not in order to say that as he was he might have acted differently, but in order to make him different. (Hayek [1960] 2011, 137)

The above fragment clearly contains a compatibilist (and not libertarian) reading of “can”. Hayek most definitely does not view the ability to do otherwise as a contra-causal power. Rather, human agents can act otherwise, given the institution of responsibility assignment. Or to put the point more precisely, it is the said institution that makes human agents act differently from the way they would have acted in the absence of the institution. By contrast, believers in MLF would have it that, first of all, human agents can still act counter to institutions’ causal influence³⁵ and, second of all, would claim that the ascription of responsibility depends on agents’ ability to do otherwise, everything else equal, to start with.

Moreover, as in the case of Mises, Hayek makes room for indeterminism in the realm of epistemology. Elsewhere in his Constitution of Liberty, he submits that when we call our actions free, we imply that “we do not know what has determined it, and not that it has not been determined by something”. (Hayek [1960] 2011, 138) So then again, as Mises, Hayek appears to be a determinist (compatibilist) metaphysically and an indeterminist epistemologically. Therefore, it seems that the Misesian and Hayekian position scrutinized above is unproblematically reconcilable with our view that human action is determined by a combination of the actor’s ordinally ranked desires (represented by value scales) and his or her fallible knowledge.

Concluding, it appears as though Mises and Hayek already embrace compatibilist metaphysics. We submit that this fact alone serves to support our contention that Austrians in general should adopt this very sort of metaphysics. And yet, there are other (non-Rothbardian) prominent Austrians who explicitly advocate MLF. The following authors who we are going to subject to critical scrutiny are, metaphysically speaking, pretty much like Rothbardians. Still, we contend that they might find our compatibilist account attractive as it would enable them to avoid otherwise problematic agent causation, while keeping everything they hold dear (e.g. Kirznernian (2000) entrepreneurship or Lachmannian (1977) radical subjectivism) intact.

Moreover, it is the very distinction between metaphysical and epistemological (in)determinism drawn above that can help us demonstrate that it is precisely compatibilist metaphysics that may prove to be friendly to the said MLF-inclined Austrians notwithstanding their pretentions to the contrary. For example, O’Driscoll and Rizzo (1996, 257) have it that

[t]he approaches of both Kirzner and Schumpeter remove entrepreneurial discovery from the confines of maximization models. The latter imply that, given the data, the outcome is inevitable. It is in this sense that these models are deterministic. Entrepreneurial discovery or creativity is not, however, a determinate outcome of the data. If it were, it would not be entrepreneurship!

However, the above-quoted “determinate outcome of the data” is underdescribed, given the bifurcation of determinism into metaphysical or epistemological one. When we employ the epistemological reading of the cited phrase, then O’Driscoll and Rizzo’s point amounts to the assertion of a conceptual truth. Indeed, if a certain discovery were to be known in advance, it would never be a discovery in the first place. Still, metaphysically speaking, discoveries might as well be determined. Hence, as it seems, Kirznerian or Schumpeterian entrepreneurship are in principle reconcilable with the compatibilist metaphysics we are advancing here.

On the other hand, elsewhere, Kirzner (2000, 56) advanced a stronger thesis, while laying bare his metaphysical indeterminism within the purview of human action. Says our author: “For the subjectivist, human action is, in this sense, “uncaused”; it is not determined by circumstances (even by the agent’s own preferences).”³⁶ However, to try to conceive of human action as “uncaused” is to call forth otherwise controversial agent causation.³⁷ And as Hodgson (2004, 187) aptly put it, “[w]e may not find the cause, but in which case it is better to leave the matter open, rather than proclaiming an uncaused cause”. Moreover, it seems to us that there is indeed no need for Kirzner to postulate agent causation. The author, while approvingly drawing on Lachmann’s (1977) subjectivism, admits that “economic outcomes are not determined by any objective physical phenomena whatever” and that “[a]ll the powerful influences exercised upon human affair by external phenomena are exercised strictly through the intermediation of active human minds.” (Kirzner 2000, 55) If we take this citation at its face value, it tallies well with compatibilist metaphysics. That is to say, actors effectuate certain states of affairs via their respective preferences even though these preferences are determined themselves.

However, a few passages later, Kirzner (2000, 56) cautiously adds what follows: “What people believe, expect and know is not determined by outside phenomena. What they do is therefore, indeed likely to have been significantly affected by, even influenced by, physical phenomena, but is never determined by them.” However, this seems to amount to yet another invocation of the problematic doctrine of agent causation. Moreover, it is not readily noticeable what is purchased at this cost. As observed above, Kirznerian entrepreneurship is not threatened by the truth of determinism. Rather, what the phenomenon of entrepreneurship presupposes is the present ignorance of a future discovery, the ignorance that the truth of determinism leaves unaffected.

Concluding, we submit that trying to save indeterminism by resorting to agent causation is more trouble than it’s worth. Additionally, we claim that for the notion of entrepreneurship as well as for the entire Austrian subjectivism to fare well it is enough to posit causal efficacy of certain mental states (most notably, preferences and beliefs). In other words, it is compatibilist metaphysics that appears to adequately (and most parsimoniously) account for many phenomena Austrians firmly adhere to such as entrepreneurship or their programmatic subjectivism, to name but a few. Hence, even if Kirzner (2000)³⁸ and Lachmann (1977) explicitly subscribed to indeterminism, we claimed that (less expensive) compatibilism could constitute an adequate metaphysics grounding their respective conceptual frameworks.

4 Conclusion

The aim of this paper was to demonstrate that MFL and some essential features of Austrian economics (i.e. its reliance on value scales and the view of action as indicating strict preference) are incompatible. As we observed, the metaphysical libertarian view of freedom of action has it that once an actor has acted (or chosen) freely, this fact implies that the actor could have done otherwise, everything else, including value scales, equal. However, as demonstrated, this fatally collides with the Austrian view to the effect that economic actors’ actual choices are indicative of their most (and strictly) preferred ends. Moreover, it appeared as though citing the lack of constancy of preferences and appealing to second-order value scales fall short of saving MLF within Austrian economics. If anything, it is the compatibilist freedom that accommodates the above two phenomena easily, while holding firmly to the spirit of Austrianism. Eventually, it transpired that resorting to indifference, as conceived of by Austrians, would go no distance towards supporting MLF either.

On a positive note, it was clear to see that we can easily make sense of value scales (coupled with relevant beliefs) being determinative of choices if we submit to the compatibilist notion of freedom. As opposed to MLF, which is haunted by the problem of luck, the compatibilist freedom is a power to do what we happen to want to do, with the latter being most certainly a “variety of free will worth wanting”. But even more critically, it appears to be only the compatibilist freedom that is compatible with the conceptual edifice of Austrian economics. Additionally, it is worth remembering that even though compatibilism per se does not assert the truth of determinism, the very essence of the former doctrine is that free will is indeed compatible with determinism. And since determinism seems to govern the natural world at least on a macroscale, a welcome corollary of our investigations is that Austrian economics (with its reconstructed notion of free choice along the compatibilist lines) smoothly coheres with the general scientific world-view. In other words, if Austrian economics proves to be compatible with determinism, then so much the better for Austrian economics, for it is only then that this sort of economics is internally consistent and that the gap between sciences of human action and nature gets at least partially narrowed.