James Broughel calls on libertarians to reevaluate the influence of James Buchanan, arguing several of his ideas have led them astray. An undercurrent of his argument is that libertarians have gravitated toward Buchanan because he confirms their generally negative views of government. But Broughel’s claims are based on mere assertions, deep misunderstandings of social choice theory, lack of familiarity with the public finance literature, and factual mistakes.
Confused about Cost
Broughel argues that Buchanan’s radical subjectivism about choice is fundamentally misguided:
The linked article that purports to show the dangers of thinking in terms of subjective costs is by Murray Rothbard. It does not cite or even mention Buchanan. Moreover, to call Rothbard a disciple of Buchanan is an simply incorrect. Rothbard is as much a disciple of Buchanan as Buchanan is a disciple of John C. Calhoun. The most charitable interpretation that can be given to citing Rothbard here is that Broughel picked an article that takes a radically subjectivist stance to try to illustrate his point. Admittedly, one has to get all the way to the very first volume of Buchanan’s collected works to find Buchanan’s actual critique of Ronald Coase (“Rights, Efficiency, and Exchange: The Irrelevance of Transaction Costs”).
Buchanan’s own use of radical subjectivism is very distinct from Rothbard’s. Rothbard wants to discredit a social cost approach to make way for his normative theory of property rights. Buchanan uses subjectivism to emphasize the importance of agreement in both markets and politics. He objects to Pigovian analysis because such analysis confuses preferences that are inferred by the analyst with the actual agreement of others. Standard welfare economics, for Buchanan, confuses the map for the terrain.
Moving on to the second claim above, it is hard to know which libertarians Broughel is referring to since he does not cite any. The idea that radical subjectivism commits one to a rosy view of government policies is silly. Yes, the radical subjectivist—if consistent—would take umbrage at using the term opportunity cost to refer to the consequences of policies. Opportunity costs are what individuals forego when they make a choice. No choice, no opportunity cost. But this does not stop a subjectivist from recognizing that policies can impose losses in terms of real income or wealth. Gains and losses are perfectly comprehensible without tying them to the act of choice. Hailstorms and taxes both impose losses on individuals.
Even if one dislikes radical subjectivism, Buchanan is the strangest target to go after in the subjectivist camp. Buchanan in Cost and Choice is careful to distinguish “cost in the predictive theory” vs. “cost in a theory of choice” (pp. 40-41). Cost in a theory of choice is radically subjective, which how Buchanan thinks the world really is. But at most, such a theory can only offer generalized explanations of social phenomena, or what I have called elsewhere “origin stories.” To do the predictive work of economic science, one must abstract from human subjectivity and creativity and treat humans as reacted to objectively measurable costs.
Buchanan thinks it is vital that economists pursue both of these projects, a point he reemphasizes in “The Domain of Subjective Economics” (again, in Volume 1 of the collected works). Of all the major radical subjectivist authors, he is the friendliest to the sort of work that Broughel seems to think economists should do. Buchanan recognizes the distinct contribution of subjectivist economics but counsels against staring too long into the Shackelian abyss.
Zombie Welfare Functions
Broughel also wants to raise the zombie idea of social welfare functions, both in his critique of Buchanan and in an earlier Econlib piece. I’ll consider where these arguments go awry in a follow-up post.
Adam Martin is Political Economy Research Fellow at the Free Market Institute and an assistant professor of agricultural and applied economics in the College of Agricultural Sciences and Natural Resources at Texas Tech University.
For more articles by Adam Martin, see the Archive.
READER COMMENTS
Thomas Hutcheson
Sep 26 2023 at 5:37pm
I’d like to see a test drive of both models over some familiar ground, say, taxation of net CO2 emissions or income vs consumption taxation.
Don Boudreaux
Sep 27 2023 at 6:43pm
Mr. Hutcheson: While Jim Buchanan fully embraced the utility of cost-benefit reasoning, he also understood that the world’s complexities often (although not always) make the performance of cost-benefit measurements dangerous – as in leading to unwarranted conclusions. He also understood the importance of choosing rules in a cost-benefit manner: those rules that are likely over time to produce positive results on net are acceptable while those that are likely over time to produce negative results on net are unacceptable.
One rule that I’m confident Buchanan would have endorsed is a presumption against active government intervention if insufficient knowledge is available to allow a well-grounded conclusion that that intervention is likely to produce benefits on net.
So in this spirit I ask: How does anyone today know that the current level of taxation of carbon-fuel producers and carbon-fuel users isn’t already ‘optimal’ – or that it isn’t even excessive? Oil companies, for example, have been paying lots of taxes for decades, as have buyers of gasoline and diesel. Where’s the evidence that whatever discouragement these taxes have worked, and continue to work, on carbon-fuel production and consumption is suboptimal? It might be, in the eyes of God. But God stubbornly refuses to share with us his full knowledge on this matter. So in the eyes of God we might already be taxing carbon fuels (and emissions) optimally or even excessively.
The case for still-higher carbon taxation gets yet weaker in light of this observation from the science writer Matt Ridley:
Would it be reasonable to conclude from the fact that cold kills many more people than does heat that we should subsidize the production and consumption of carbon fuels? Such a conclusion would be no less justified than is the conclusion that we should tax these activities.
Pro-carbon-tax people point to rising global temperatures (and the fact that at least some of this rise is indeed caused by human-induced carbon emissions), and they point also to unsurprising consequences of these rising temperatures (such as rising sea levels). They then leap to the conclusion that we need to tax carbon more heavily. The upsides of global warming (see above) are largely ignored or discounted, while unsupportable assumptions are made about how humans will respond or not to the consequences of global warming.
In all of this the amount of genuine, detailed knowledge of the sort that is necessary to justify government intervention is minuscule.
My argument here is not that costs are so subjective that they cannot be measured. My argument here, instead, is that we don’t know – and practically can’t know – the full consequences of existing taxes and other regulations. We can’t gather this knowledge even if we limit our investigation only to ‘objective’ consequences of the taxes and regulations. When we add in the fact – and a fact it certainly is – that costs, and benefits, and risk preferences are indeed all subjective, the epistemic challenge becomes too great for humans to deal with. Given this reality, we should default to having no government intervention on this front.
Thomas L Hutcheson
Sep 29 2023 at 7:42am
I really appreciate your point about using cost benefit analysis to make rules. I think we are closer to thinking about taxing and spending in cost-benefit terms than in using it to make rules. Which it sort of the wrong way around in that taxing and spending decisions ought, in my opinion take account of income distribution effects which are not easy (possible?) to incorporate into C -B analysis, I think distribution ought to be excluded from most regulatory decisions.
As for the discussion of the optimal level of taxation of net CO2 emissions, I fully agree that if current policies are modeled as remaining, they would affect the estimate possibly making it zero, as it is at present. But I’d like to see that as the result of an analysis, subject to all the uncertainties about economic-climate models, rather than a presumption that it ought to be zero and implicitly that the existing set of policies is optimal, which frankly just seems absurd to me, especially after IRA.
Pierre Lemieux
Sep 30 2023 at 9:45am
Thomas: One important thing to understand about Buchanan is that the only “cost-benefit analysis” he agreed with required that the benefit of the system of rules must be higher than its cost for each individual. Otherwise, the individual cannot be presumed to have given his consent. This is obvious in his first major work, with Gordon Tullock: The Calculus of Consent. Except at the constitutional stage, where there is “actual” exchange about rules (and, I think, bribes or side-payments in the form of favorable rules, but this may be controversial), there is no weighing of the benefits of some individuals against the costs of others. And there is never any legitimate weighing of the costs and benefits by an external observer such as a government or its consultants. The only exception to this latter sentence is that at the post-constitutional stage (day-to-day politics), politicians and majorities are given some limited leeway in order to prevent free-riding and strategic blocking in the production of public goods; but these activities must abide by the constitutional rules.
Pierre Lemieux
Sep 30 2023 at 9:53am
Thomas: PS: My review of The Calculus of Consent may provide some keys to understanding Buchanan’s theory. It marked a fundamental break with the standard, Musgrave-type theory of public finance.
Jon Murphy
Sep 30 2023 at 10:35am
To Pierre’s point, I think he highlights the main analytical difference between the subjective cost theory of the LSE school (Buchanan, Coase, Hayek, Robbins, Wiseman, etc) and the Cambridge School’s theory of objective cost (Marshall, Keynes, Edgeworth, etc).
In the subjective theory of cost, cost benefit analysis only makes sense as a normative prescription at the individual level. It is only at the individual level where genuine alternatives are faced, so it is only at the individual level where genuine costs are calculated. Consequently, it is only at the individual level where genuine choice is made. Thus, it is only at the individual level where one can say that one should act where costs are less than benefits if one wants to improve their lot.
Since costs are subjective (they depend on the realizable alternatives faced by each individual), aggregating them into an objective number is incoherent. Price is a useful signal, but it is a noisy signal. Price is part of the cost, but it is not the total cost. In the Marshallan sense where price encompasses both monetary and opportunity cost, aggregation makes sense. But in the subjectivist sense, where price includes just one element of cost, aggregation does not make sense.
So, at the individual level, cost-benefit analysis is a coherent normative descriptor. But at the collective level, cost-benefit analysis does not contain the same normative prescription because there is no true sense of the true cost. At the collective level, one cannot say with any reasonable level of precision that an activity should be undertaken if costs are less than benefits since the measure of cost (and benefit, for that matter) are wholly different from the individual level.
None of this is to say that cost-benefit analysis has no role at the collective level. Rather, I am just emphasizing that it is not the same as at the individual level and there is not the same normative power. In the subjectivist cost perspective, elements like jurisprudence matter a whole lot.
Pierre Lemieux
Sep 30 2023 at 12:07pm
Jon: Interesting thoughts to advance the conversation! You write:
I am not sure of this. Did Buchanan ever use the expression “cost-benefit analysis”? I don’t think so, but perhaps I missed something.
This being said, I am not totally sure I correctly understand the relation between the subjective costs of Buchanan’s Cost and Choice and the straight idea that interpersonal comparisons of utility are impossible; but they must be two ways of roughly saying the same thing. (Perhaps I have to reread Cost and Choice.) In my mind, price is objective because it represents the ratio (number of apples and number of oranges) at which we observe that mutually profitable bargains are made. With these market or shadow prices (and costs), CBA tries to estimate consumer surpluses (and perhaps producer surpluses, but this is more controversial) as a proxy for utility, but they are not comparable from one individual to another. Ultimately, in a general equilibrium setup, CBA tries to reach the bliss point on the PPF at which the social welfare function is maximized. But Buchanan rightly believed that a social welfare function is not conceptually definable. (Personally, it took me some time to discover the total vacuity of CBA.)
Jon Murphy
Sep 30 2023 at 1:15pm
I should have been more clear: that statement you’re quoting from me is my POV, not Buchanan’s. I think looking at monetary prices and making projections about benefits should be used as part of the political decision-making process, but it should not be seen as the same as what happens when an individual chooses. A lot of business/public finance texts have some sort of normative claim like “calculate the NPV and, if it is positive, the project should be undertaken” (my emphasis). I am saying that normative claim (if NPV > 0, then take action) is invalid and not comparable to the individual decision-making CBA.
Don Boudreaux
Sep 26 2023 at 6:56pm
Adam: Spot on. Thanks for this excellent post.
Don
Pierre Lemieux
Sep 27 2023 at 1:01pm
Adam: An enlightening post!
Sam Branthoover
Sep 29 2023 at 3:01pm
This post was very necessary, thanks for doing this.
Comments are closed.