Morality and Value Neutrality in Economics: A Dualist View

Download full paper

This paper is closed for comments.


In this paper, we propose a dualist view that economics exhibits the properties of both moral science and value-neutral approach, regardless of the normative-positive distinction. Our argumentation is derived from the understanding that, analytically, economics is a broadly-defined rational choice theory. As implied by this claim, on the one hand, economics behaves as a moral science for two main reasons: all economic theories and policy discussions are necessarily based on moral premises about means-end considerations; economics as an analytical approach can be and has been applied to explanations of a wide range of moral phenomena. On the other hand, since economists — without being informed of some ethical presuppositions of higher order — cannot deal with the comparisons among different value criteria, their approach remains neutral regarding judgmental positions, which should be given a priori to make economic enquiries possible. Ultimately, by this view we reconcile morality with value neutrality in economics, without slicing the discipline into two distinctive branches.

Posted for comments on 1 Dec 2016, 1:16 pm.

Comments (3)

  • Ramzi Mabsout says:

    Comments on Cheng Li “morality and value neutrality in economics.

    Although I find the topic of Li highly interesting, I think this is a conceptually ill-conceived paper. I list below a number of misconceptions. In the end, it was not clear to me what the dualist view is and how “economics exhibits the properties of both moral science and value neutral approach, regardless of the normative –positive distinction”. I am still not sure what was achieved even if the thesis is accepted. I am specifically concerned about the definition of economics (as means–ends rationality) used, the meaning of value neutrality, and the lack of coverage of the recent literature on ethics and economics.


    There is an important difference between Lionel Robbins and JN Keynes, namely, Keynes considered ethics a science whereas for Robbins it was philosophy (see discussion in Hands 2001, Reflections Without Rules).

    What is a value neutral approach? A descriptive/explanatory definition of the positive branch of economics can be value laden. Stand point epistemology and certain feminist economics have emphasized that there is no view from nowhere, no neutrality: values, moral and non-moral, are present in all branches of science.

    “.. analytically economics is based on the principle means-ends rationality , taken in its broader sense. It turns out that holding this principle is a common, but often unspoken and thus forgotten property underlying all approaches relying on the economic way of thinking−from the neoclassical school to behavioral economics –regardless of how they differ in the concrete empirical counterparts of the mean sand end.”

    I find this a very reductionist delineation of economics. It is both too broad and too narrow. It is too narrow because it is reducing economics to Robbins’ definition, ie, to a certain type of microeconomics. But it is also too broad a definition since almost any social science can be conceptualized in terms of means–ends rationality. What makes economics different? What is different about economics is its subject matter and not an abstract rationality maxim. Or if the author insists on using this definition then it needs some justification and not merely acceptance because this is Robbins’ or Becker’s definition.

    “it is because without certain ethical which a specific …” long confusing sentence, what is the author trying to say?

    On the use of “value neutral facet of economics” I don’t think there is anything value neutral economics.

    Section 2

    Footnote 5: How can the debate about the definition of economics be beyond the scope of this paper when the paper is in fact using a definition of economics as means-ends rationality?

    “It turns out that economics is a broadly defined rational choice theory” see earlier comments. This cannot be accepted at face value. The author is forcing a definition of economics as means–ends rationality without argument.

    Definition of rationality : rationality is “a form of intuition which makes economic explanation possible and further defines epistemological limitations of the discipline … rational in the above sense means is just a synonym of explainable and according to an economic point of view explain is the equivalent to rationalize” So rationality is form of explanation and explanation is equivalent to rationalizing? Isn’t this circular? And what is the economic point of view if we are trying to use means-ends rationality to determine what kind of questions are valid economic questions?

    To be accurate the prevalent definition of rationality in economics is not the one given by the author (or Robbins or Becker) but the two core axioms of choice (completeness and transitivity; cf Mas-Colell et al.). So the author’s definition seems to have been relegated to textbooks.

    Table 1: The author does not explain why neo-classical economics cannot include psychological needs in its utility function. I would think that neo-classical economics is quite liberal at what is included in the utility function. As Binmore (among others) has argued a utility function can include the preferences of others (for example consider Becker’s family household model where the utility function of other members are included).

    Section 3

    “To put differently from an economic point of view, one cannot explain /understand is only from is.” This I would say is not only from the economic point of view , but for any scientific endeavor: facts are theory laden and the rejection of empiricism as advocated by the early views of logical positivists was undermined by K. Popper, W. Quine, and T. Khun.

    Section 4

    If the end pursued are moral ends, that does not make economics a moral science. Moral ends (say religious beliefs) can be explained by a descriptive science without assuming an ethical stance. Sociology and other social sciences can explain why people hold religious belief without adopting an ethical view on religion whether it is what one ought to do or not.

    The third cause : I would recommend studying the work of H. Gintis and K. Binmore , both of which have attempted to explain ethics away as it where (a critique of their work is offered by Bashour & Mabsout in the Review of Social Economy).


    What’s wrong with arguing that different branches of economics (descriptive/explanatory/positive and normative/prescriptive) complement each other? Besides the fact that I don’t think value neutrality (or the view from nowhere) is possible, it is still not clear to me what has the author achieved by the reconciliation of morality with value neutrality. The fact that economics has many faces (JN Keynes argued that we can divide it into three branches positive, normative, and the arts as did L. Walras) is not necessarily a bad thing, a distinction to be dissolved so we get a dualist view that meshes both without clearly explaining what was gained by this distinction.

    In summary the author has raised a strawman of economics (as means–ends rationality) which fails to capture the richness and diversity of economic thought. This definition fails to capture a very wide diversity of economic projects (from institutionalist, to evolutionary, to neuro-economics and so on). As it stands I don’t think the paper should be published in its current form. A revision would need to reconsider the definition of economics, the meaning of value neutrality (if it is possible at all), review the literature on how economics have attempted to explain away moral norms (Gintis and Binmore), and integrate many of the recent contributions that have been made to the topic of ethics and economics to contrast them with the author’s approach (for example see the work of I. V. Staveren, M. White, J. Whigt, W. Beckerman, and D. McCloskey among many other recent contributions).

    Ramzi Mabsout
    American University of Beirut

  • Vítor Neves says:

    The concern with values, morality and value judgment in economics is centuries old and seems to be, as Terence Hutchison termed it, “a perennial philosophical chestnut”. However, as this author also claimed, even a philosophical problem as controversial and seemingly insoluble as this one should be subject to reconsideration from time to time. In recent years, there have occurred many and significant developments in this regard.

    In the paper considered here, Cheng Li attempts to contribute to that discussion by submitting what he calls a “dualist view” according to which economics is both a moral science and a value-neutral approach (technique) – “neutral regarding judgmental positions and value systems” – and that this is so “regardless of the normative-positive distinction”. According to the author, holding a “rationality principle”, understood as “an a priori common grammar of economic analysis”, and considering that, “analytically, economics is a broadly-defined rational choice theory”, would allow us to reconcile morality and value neutrality in economics “without slicing the discipline into two branches” (positive and normative economics).
    I take issue with several aspects of this paper that, in my view, preclude its publication in this journal in its current form.

    The argument is in general somewhat vague and unclear (section 3 is a case in point), insufficiently elaborated and does not engage, as I think it would be required, with the extensive (and highly relevant) literature published on the subject in the last few years. Ultimately, the reader gets little more than largely unsubstantiated statements. In particular, I see no reasons to accept (or at least to take seriously) the argument provided to support the fundamental conclusions of the paper, that is, that the means-end rationality principle (“broadly” understood) leads to “the reconciliation of morality with value neutrality in economics” and that economists could “remain neutral regarding judgmental positions”.

    The following points illustrate my position.

    1. On the role that the author assigns to the “means-end rationality principle”, we read:
    “It turns out that holding this principle is a common, but often unspoken and thus forgotten property underlying all approaches relying on the economic way of thinking – from the neoclassical school to behavioral economics – regardless of how they differ in the concrete empirical counterparts of the means and the end” (p. 3).

    Two pages below we are told that this economic way of thinking, “by common understanding, is self-evidently founded on the principle of means-end rationality” and that this principle “is shared, explicitly or implicitly, with all economic approaches and schools of thought, as long as they do not abandon the economic way of thinking when enquiring into human choices and trade-offs” (p. 5).

    It is manifest the circularity of the definitions provided for the “rationality principle” and the “economic way of thinking”. Further, with a single phrase all economic approaches and schools of thought that do not share this so-called “economic way of thinking”, based on the means-end rationality principle, are excluded from economics. But what does this “economic way of thinking” mean after all? Apart from linking it to the rationality principle, the author does not say. To make things worse, one is told that the debate on the definition of economics is beyond the scope of the paper (p. 4).

    In addition, the author claims: “theorizing upon the rationality principle remains an integral feature of economic analysis” and adds: “Here, the term ‘rational’ – being a major source of confusion – should be taken in its broader sense. It by no means implies that economic agents always make the right choice that leads to the highest level of material satisfaction. (…) Instead, the term ‘rational’ merely means that from an economic point of view, all human actions and choices are perceived as the outcomes of certain kinds of means-end reasoning, and thus explained in this way.” (ibid.)

    Although in the argument presented it is given a central importance to the rationality principle, this is clearly unelaborated. The author suggests that a broader scope and meaning be ascribed to it. However, I am afraid that, in order to be taken as “a common grammar of economic analysis”, the rationality principle has been converted to a mere rhetorical device emptied of any explanatory capacity (and, as such, unable to fulfill the function the author assigns to it in the paper). Further elaboration on the rationality principle is required. Let me suggest in this regard a reading of Neves (2004) and Kerstenetzky (2009).
    On the long-standing critique of the means-ends reasoning in economics (see, as illustrations, Kapp 2011 and Crespo 2007) unfortunately nothing is said.

    2. In the end of section 3 the author says: “despite its popularity, the normative/positive distinction, even if properly considered, is of no direct relevance with respect to the dual identity of economics [both a moral science and a value-neutral approach]”. It is amazing how easily the author gets rid of the positive-normative distinction and of the difficult problem of how to deal with morality, objectivity and value judgment in economics.

    3. Finally, in the conclusion (p. 15), the author states that the means-end rationality principle enables economics to remain “neutral regarding judgmental positions and value systems” and this “because, without ethical presuppositions, economists cannot judge and evaluate the human choices in which different ultimate value criteria or ends are involved.”

    The fact is that the author recognizes that he cannot dispense with some sort of meta-criterion (he calls it a priori) in order to weigh and choose among different intermediary or instrumental ends. Making judgments and adopting some sort of ethical values are inevitable. However, he seems to believe that these could be “borrowed from other disciplines including ethics, or even be given arbitrarily” allowing economics to remain “a neutral instrument to explore the consequences and implications of certain moral premises and value criteria given beforehand, regardless whether they are morally justified by the economists who draw on them” (p. 13).

    The question is: how can economists adopt moral premises and value criteria provided elsewhere without exercising value judgment themselves?

    To conclude:
    I do not think the paper should be accepted in its current form. Nevertheless, I believe that tackling the issue of the double nature of economics as an ethics-related and an engineering-like discipline (not to be confused, in my view, with value-neutrality) is a worthy, although difficult, endeavor that must continue to be pursued.

    Vítor Neves, University of Coimbra, Portugal

    Crespo, R. (2007). ‘Practical comparability’ and ends in Economics. Journal of Economic Methodology 14(3): 371-393.
    Kapp, K. W. (2011). The Foundations of Institutional Economics. Edited by S. Berger and R. Steppacher. London and New York: Routledge.
    Kerstenetzky, C. (2009). Plural situational logic: the rationa(lisabi)lity principle. Cambridge Journal of Economics 33(2): 193-209.
    Neves, V. (2004). Situational analysis beyond ‘single-exit’ modelling. Cambridge Journal of Economics 28 (6): 921-936.

  • Cheng Li says:

    Reply to Doctor Mabsout,

    I thank the reviewer Doctor Mabsout for his valuable comments and criticisms, which give me some inspirations and opportunities to further clarify my thoughts regarding this topic of great interest. Below I will first reply to the three major concerns of the reviewer, namely the definition of economics, the meaning of value neutrality, and the lack of coverage of the recent literature on the topic, and then I will address some of other comments or questions.

    Definition of economics
    First and foremost, my position on the definition of economics has been deeply (if not entirely) misunderstood. As well known in the literature, there is an extensive debate about whether economics should be defined according to approach or subject matter. However, as I wrote in Footnote 5 (Section 2), the definition of economics is beyond the scope of my paper. That is to say, I am not concerned about whether economics should be defined according to approach or subject-matter, nor what appropriate subject matter economics should focus on. Instead, the topic I am concerned with in this paper is how economists explain or understand whatever their subject matter. My answer, in short, is that economists explain or understand things by organizing the messy empirical materials or data through the means-end rationality principle. The reviewer or many other readers may think this point has little value added, but indeed it has some extremely important implications on the nature of economics and its usefulness and limits (thus responsibilities), which become hot topics especially in the aftermath of the 2008 financial crisis. The duality of economics is just one of these implications. At this juncture, I may digress by adding that my answer to “how economists explain,” contributes to the relevant literature of economic methodology. Indeed, in his book The Methodology of Economics, Mark Blaug (1980, 1992) asked this question but did not directly answer it. However, I would not by surprised by this apparent contradiction after reading his chapter on rationality postulate (Chapter 15, 1992), where Professor Blaug did not make a distinction between the abstract framework of rationality and its empirical counterparts.

    With the above understanding, when claiming that “economics is a broadly defined rational choice theory,” I by no means intend to propose a new definition of economics, nor take those from others (such as Robbins or Keynes). What I want to argue is that economic enquiries of all schools (neoclassical, behavioural, and so on) are necessarily based upon the means-end rationality principle (in its broader sense), and thus it turns out that economic explanations are equivalent to rationalization-By the term “rationalization”, I mean organizing the raw material about human behaviour by a means-end structure. From this angle, one may think that being a rational choice theory is a necessary but insufficient condition to define economics. For reference purpose: defining economics only upon its subject matter, as the reviewer proposed, seems not enough. Say, if economics is defined as a study of wealth, then agricultural science, which obviously is to do with production of wealth, should be viewed as a branch of economics?

    Meaning of Neutrality
    As I wrote in Section 5, first paragraph, by “value neutrality” we mean: economic enquiries should be necessarily based on some premises on means-end considerations. It is obvious that once these premises are given beforehand as explanantia, their moral rightness or wrongness has been put beyond discussion.” I shall stress that I view this as an implication of economics being a rational choice theory.

    Lack of coverage of the recent literature
    The reading recommendations are appreciated. Unfortunately, there seems to be no clear literature line on rationality in economics, nor on ethics–economics relations. Indeed this paper and its early drafts have been read by a few experts, and some of them – including the two reviewers of the ET journal, recommend very different literature/reference. This, I think, demonstrates the complexity of the topic and the diversity of perspectives. Moreover, I should stress that my paper is a rethinking of economics by economist, when the discipline has been troubled in the aftermath of the Global Financial Crisis. It is not an application of recent studies on ethics or philosophy of science to economics.

    Neoclassical economics in Table 1
    I think utility function with psychological needs is not typical of a neo-classical model. I think it is only a matter of name. Some may still call this neo-classical economics, some may call this behavioral economics or by other names. Here the only important thing is that different economic branches / approaches explain the subject-matter of economics in the same manner (means-end consideration or constrained maximization), but with different empirical counterparts.

    Too narrow or too broad
    Regarding the “too narrow” or “too broad” views, I think both are untenable. My claim that economics is a rational choice theory in a broader sense does not imply that only with the abstract means-end framework can economists explain their subject-matter. This framework should be combined with empirical contents, for example maximizing consumption and being subject to budgetary constraint. From this angle, every economic enquire is a mix of induction and deduction.

    Definition of rationality
    With all due respect, there is a serious error in the reviewer’s quote from my paper: In Section 2, page 5, I wrote “to use Kant’s terminology- the a priori form of intuition which makes economic explanations possible…” However, the reviewer quotes it as “a form of intuition which makes economic explanation possible….” The meaning of those two sentences is quite different. Moreover, it is to my great surprise that the reviewer next summarizes my position as “so rationality is form of explanation.” It is not my position at all.

    Regarding Mas-Colell and others’ definition of rationality, I do not see any inconsistency between theirs and mine (although theirs is more convenient for mathematical treatment). For example, if the axiom of transitivity is violated, then the choice / behaviour of the economic agent in question cannot be explained out of means-end consideration, and will be no longer a subject of economic explanation. Think about a simplified example: If an individual prefers A to B, B to C, and C to A, and in reality B is chosen by this individual, without additional information, what do we expect from an economist to explain this choice? In my view, economics ends here.

    Reply to Doctor Neves,

    I also thank the reviewer Doctor Neves for his valuable comments and criticisms. Below is my reply to him:

    1. (1) Regarding the definition of economics which is beyond the scope of my paper, I refer the reviewer to my reply to Doctor Mabsout. (2) Regarding the relevant literature, I refer the reviewer to my reply about the “lack of coverage of the recent literature.” (3) Regarding the explanatory capacity of rationality principle, I do not think this principle alone can explain anything. It should be equipped with empirical counterparts of means and end or utility function and constraints. In this regard, I also refer the reviewer to my reply about “too narrow or too broad.”

    2. My sentence “despite its popularity, … dual identity of economics” should be taken in its context, especially the paragraphs preceding this conclusion in Section 3. It seems to me that the question of the reviewer needs more elaboration.

    3. I think the reviewer asked a very good question, namely “how can economists adopt moral premises and value criteria provided elsewhere without exercising value judgment themselves?” My answer is below: of course, say, Professor Samuelson or Professor Friedman exercised value judgment beforehand and then adopted some moral premises when doing economic analysis. But, they did not do that qua economists. Another example seems helpful. Of course, I can explain to my children the acceleration of a free falling body. But when doing that, I do not explain qua economist (or, I do not provide an explanation from an economic point of view). Instead, I’m just using my knowledge of physics.