Bryan Caplan  

How Normative Should Economics Be?

PRINT
No Conspiracy... Tom Sowell's Latest Book...
Almost all economists take one of the following three positions on normative economics:

1. Economics should never be normative.  Economics is about what is, not what should be.

2. Economics should be normative about government policy, but not individual behavior.  Economics is about identifying sound policies, not saving souls.

3. Economics should be normative about both government policy and individual behavior.  Economics is about identifying sound decisions in every walk of life.

(Of course, there is a fourth logically possible position - economics should be normative for individuals but not for government policy - but as far as I know no one holds this position).

What's your position - and why?


Comments and Sharing





TRACKBACKS (2 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/3048
The author at It Don't Mean Much, These Seats are Cheap. in a related article titled The Proper Stance writes:
    Over at EconLog, Bryan Caplan asks, “How normative should economics be?” Since it is an impossibility to separate economics from normative judgement, I would have to say that economics must seek to be carefully normative about both individu... [Tracked on February 14, 2010 5:17 PM]
    Bryan Caplan asks "How Normative Should Economics Be?": Almost all economists take one of the following three positions on normative economics: 1. Economics should never be normative. Economics is about what is, not what should be. 2. Economics should ... [Tracked on February 17, 2010 1:27 PM]
COMMENTS (73 to date)
Steve writes:

5. Economics should be positive as it supports the conclusions that I would come to were I to do a normative analysis based on my own preconceived notions. Otherwise, economics should be normative.


Or do I just read too many econblogs?

Norman writes:

I take the position that economics should be normative only when the objective function is externally provided; that is, if government says "we want economic growth," economics should tell government what to do. If on the other hand government says "we want to balance growth and stability," economics should only tell government what to do once the relative weights have been decided by the government. The same logic would apply to guiding individual decisions: it can help individuals maximize expected utility given a list of inputs and functional form, but it can't provide any meaningful ideas about what a sensible utility function looks like. So if government says "we want to maximize consumer utility," I'd say economics can't be normative, and shouldn't try.

Not sure where such a conditional position falls on your scale.

Travis writes:

Heretofore, economists have only attempted to understand the world.

The point, however, is to change it.

(Modified Theses on Feuerbach).

Hume writes:

Is it fair to argue that an economist making normative judgments regarding whether certain ends are appropriate isnt really engaging in economic theory at all? Its analogous to an engineer criticizing a bridge not because of its design, but because (s)he believes the bridge would be better utilized in a different location. (I am out of my league here. I am not familiar with the arguments regarding normative economics and what it means to be taking part in "economic debate." This is just an intuitive response.)

Doc Merlin writes:

IMO:

The distinction between positive and normative here is a false one.

zefreak writes:

Economics is value-free. Economics can inform normative arguments, but when economists engage in normative debate they are doing so outside of their capacity as economists.

Norman:

"if you want x, then you should do y" is not normative analysis. Normative economics would be "y is a goal we should pursue, and we will use means x to attain it"

Your position properly understood lies in the first category.

Lauren writes:

Doc Merlin:

But _why_ is it a false distinction in your opinion?

It's fine to have an opinion that the "distinction between positive and normative here [on EconLog] is a false one," but it's not helpful to other readers to not say why.

It would not be helpful for me merely to comment on a blog that in my opinion, the distinction between the moon being made of green cheese or oxidized copper ore is a false one. What would help readers would be for me to explain why the distinction is or is not false. Assertions of opinion without explanations are neither helpful nor convincing.

The distinction between positive and normative economics in Bryan's post is very clear, quite correct, and consistent in every way with standard economics usage about positive verus normative economics. What about the distinction between these terms do you think is false? If you think the distinction is false here, do you you also think it is false everywhere in economics? Does some other person or website distinguish the two terms correctly in your opinion?

Randy writes:

I think of economics as the philosophy of the real. Its value is in its ability to explain to the philosophers of the unreal why their utopian visions will not work. And yes, many who now call themselves "economists" are actually in the latter group.

Robert Johnson writes:

Lauren,

You could say that engineering is normative (because it seeks to prescribe solutions) while physics is positive. But I agree with what I think Doc Merlin is saying, that "normative" is the wrong word.

Normative is typically used to mean value-based, moral, something like that. Few people would argue that engineers are seeking to impose their values on society while physicists are merely describing the natural world.

So maybe Prof. Caplan should be asking "When should economics be descriptive, and when should it be prescriptive?" or "When is economics more of a science and when is it more of an engineering discipline?"

ajb writes:

I think economists should always be neutral when speaking as experts and promote advocacy as individuals only.

Bryan often oversteps this and weakens his expertise when he argues about e.g.: child-rearing, religion, and immigration as if they flow naturally from economics, when in fact they require many beliefs and values others/most don't share.

The fact that most economists act like Bryan while also benefiting from the mantle of "scientific expertise" is a sad fact of the human condition.

Hence I'm caught in normatively indicating to Bryan that he should speak more "positively," but realize such advocacy is futile.

I also acknowledge that there's a prisoner's dilemma where if those you disagree with insist on speaking normatively it seems foolish to hold back as well.

Daniel Kuehn writes:

I suppose a mix of 1 and 3 actually. Economics itself is a social science and therefore economics proper should be strictly positive.

But the point of positive science is to improve the world. So economists should absolutely express opinions on what is good in whatever field of life they feel comfortable.

The trick isn't to do normative economics or not to do normative economics. The trick is to make sure you distinguish between your normative and your positive economics. Never make the claim that positive economics alone backs your normative claims - say that positive economics, when in combination with your particular set of values, leads to the normative claims you make.

david writes:

@zefreak

Economics isn't value free, even in that manner, sadly: you still have to make judgment calls on which models approximate reality the best, and that invariably entails dismissing certain phenomena as unimportant.

(Take, for instance, the standard assumption that mutually-agreed upon trade is mutually beneficial. We know already that people can be influenced - advertising is a very big industry! Nonetheless, it is mathematically much simpler to just assume equivalence).

Caplan provides another example by asserting that optimizing behavior is material self-interest. Which immediately attaches normative implications.

Mathematical modeling also predisposes us towards certain conclusions. Pareto optimality is very attractive because it is simple, despite the absence of another, presumably non-zero-desirable, goal of equality of opportunity (the sadist torturing a victim is Pareto). General equilibrium effects are known to be indeterminate, so we focus on the partial equilibrium. Network effects, strategic behavior, and imperfect information are difficult to model intuitively, so in a pinch we flee back towards well-known models. You may have noticed this back when the financial crisis started: many bloggers reverted back to prewar Say's law classicism while others reverted back to old Keynesianism, with everybody forgetting all the progress made since 1950.

Daniel Kuehn writes:

Randy -
RE: "I think of economics as the philosophy of the real. Its value is in its ability to explain to the philosophers of the unreal why their utopian visions will not work. And yes, many who now call themselves "economists" are actually in the latter group."

Many people who talk like this look a lot like utopians themselves to the rest of us (I assume you're taking this from a Hayekian/libertarian angle).

Tobi writes:

I agree with Merlin, I don't think there is a distinction between normative and positive economics. Economics is fundamentally about man and that's what also makes it a moral science.

The comparison with engineering is an invalid one. Physicist, engineers and chemist do not impose their values on society for they're not describing society. Their object of study, how relevant to society it may be, has nothing to do with society.

The only thing economics can do is show you what the likely or possible outcome of your actions/decisions might be. This then can give you feedback on the goals you'd want to achieve.
In other words it can tell you the price, but not whether or not you should buy it.

Doesn't this make it a solely positive science then? I'd say no, asking the question is in my opinion a normative act. Seperating the normative from the positive would be seperating the question from the answer. That would reduce both to gibberish.

I mean look at all the economists from Adam Smith, including Karl Marx, to the present day, haven't they all been motivated by a better tommorow? For instance both Keynes and Hayek were both motivated in their study to preserve liberty in the face of an economic downturn. The former to fix it and the latter to prevent it. Their research question were fundamentally normative, ie. how to preserve liberty?

This also brings me back to the natural sciences, physicists have not divided their science in ideological cleavages. There is no saltwater school or freshwater school of physics, they might disagree but their truth is an objective one, it doesn't tell you how to run society.

liberty writes:

"Their research question were fundamentally normative, ie. how to preserve liberty?"

That is why people *see* economists as normative. But if you a mathematician and you are asked to do a calculation for a priest, does that make you religious?

If the people want to preserve liberty, and they ask the economists how it can be done, it does not matter whether the economist also desires liberty or not.

Now, here is the distinction as I see it:

If an economist goes on television and says "Health care bill HR 3200 will likely not result in lower health care costs, and this is why..." he is doing his job. He had been hired (by society if not by an individual) to analyze the policy - so long as he has done so with intellectual honesty, this is positive work.

If an economist goes on television and says "Heath care bill 3200 is the wrong direction for America. It spends too much, it will put our children in debt, and it will restrict our freedom..." then he is no longer acting as an economist. At this point he is advocating policy - he is arguing for certain ends, not analyzing the means that have been chosen in order to achieve agreed-upon ends.

liberty writes:

Actually, come to think of - that contingency is more important than the primary distinction: "so long as he has done so with intellectual honesty"

There is not much difference between an economist that says "If you want growth, this policy is no good" and one that says "this policy is no good, it will destroy growth" - but there is a world of difference between a good and a bad economist.

Neal W. writes:

(1) Economics should not be normative.

Economics just tells us what is, and when an economist gets normative, (s)he puts on the hat of an ethicist.

Daniel Klein writes:

The wise man expunges "positive v. normative" from his vocabulary. Ises and oughts are easily and naturally translated into one another, based on the purposes of the interlocutors and the discourse situation.

The words "positive" and "normative" do not mean nothing, but what they mean can always be expressed in better terms. "Normative" often means outspoken, unconventional, strident, etc. It can also mean loose, vague, and indeterminate.

Tell me "positive" or "normative" for each of the following:

(1) The minimum wage ought to be repealed.

(2) I think the minimum wage ought to be repealed.

(3) The minimum wage reduces social welfare.

(4) Wise people oppose the minimum wage.

The primary verb of (1) is an ought, while the primary verbs of (2), (3), and (4) are ises. But all four statements are really the same.

Coase used the term "affectation" for posing as "positive" and not "normative."

Taras Smereka writes:

I take the fourth position. There is no way for me to affect the government's decision making in a way that there is a net benefit to me, so I just move on.

Randy writes:

Daniel: "Many people who talk like this look a lot like utopians themselves to the rest of us."

Liberty is worth fighting for. Society is not.

Lauren writes:

Thanks for those clarifications, Robert Johnson and Tobi. You are both on target. And the terms themselves are not the best. But the concepts are meaningful, so let me give this a shot from a different direction.

I think of normative statements as any statements or questions that involve words such as "should," "ought to," "good," or "bad." All of those involve some sort of value judgment. Positive statements or questions involve factual descriptions or claims. (Note that the word "positive" in this sense is not the opposite of "negative." The statement that the sun does not rise at night is "negative" in daily use because of the word "not," but in terms of the positive/normative distinction it's a positive statement because it's about facts, not value judgments.)

A positive statement can be false. "The sun rises every day in the West" is a positive statement, but it's false. A positive statement or question is simply one that can potentially be addressed objectively--if one had enough facts and could perform the appropriate experiment.

Most sciences are about positive rather than normative questions. No one tries to assert in physics that the sun should rise in the West or that it would be better if electrons really did behave like planets instead of doing their cloud thing.

But when it comes to economics, people--especially newcomers to economics--often get positive and normative issues mixed up. People have very strong emotional reactions to statements like "Minimum wages increase unemployment," hearing it as if the speaker is saying "The government should not raise the minimum wage." They hear--and speak--normative statements where economists, at least when speaking to each other, sort out the positive from the normative. Newcomers often immediately leap to the normative implications of a positive statement and slip into making normative judgments where only positive ones are intended.

No one thinks economics is a science in the same sense as physics. We all agree that economists' ability to run controlled experiments is deeply limited. Thus, its ability to distinguish truthful positive statements from false positive statements is limited.

Economists mix it up, too, but in a more subtle way. We fight like the dickens in academic conferences and over lunch about the results of empirical tests and theoretical results precisely because there is so much room for doubt. Just like the newcomer from two paragraphs ago, economists, too, immediately see the normative implications of whatever positive results are involved. We tend to study things that lead to positive conclusions that support our deeply-held normative values. Economists who believe the government can actively help people be better off tend to become Keynesian. Economists who believe the government is more likely to cause harm than good when it acts tend to ally themselves with neo-classical lines of thought.

But that doesn't mean there is no distinction. It is still good practice to distinguish between normative statements--such as "Economics should be normative about government policy," and positive statements, such as "Economics is partly about the effects of government policy." The former involves getting other people to agree with you about something that ultimately is a matter outside of science or facts--it's about values. The latter is potentially answerable by sitting 10,000 economists in a room and asking them if the work they actually do involves asking about the effects of government policies.

Only by making the distinction can you have a rational conversation. Or pose the interesting question Bryan has posed!

Kurbla writes:

I think it must be normative: without at least some value system, one cannot distinguish important from irrelevant.

Mario Rizzo writes:

It depends on what you mean by "economics." But I believe that the science itself is non-normative. The normativity comes in only after you introduce value judgments that come from elsewhere. It is true, however, that economics can detect inconsistencies in goals or preferences. That would seem relevant to both individuals and policy-makers. But in the final analysis I say #1.

I recommend Chap. 2 of John Neville Keynes, The Scope and Method of Political Economy for an excellent treatment of the positive-normative distinction.

Tom West writes:

"If you want growth, this policy is no good" and one that says "this policy is no good, it will destroy growth"

What? There's an enormous difference. The former allows for evaluation of the policy on the basis of other considerations. The latter states that the only acceptable metric for evaluation of policy is how it affects growth.

Tobi writes:

@Liberty: "That is why people *see* economists as normative. But if you a mathematician and you are asked to do a calculation for a priest, does that make you religious? "

Well no, for two reasons:
1. Mathematics is not a science about man, it is an analytical apriori system. The knowledge exists independent of the observer.
2. The questions you ask cannot have a normative character. There is nothing moral about real or imaginary numbers.

I think your healthcare example actually shows it the best. The very fact that we discuss costs and benefits already implies a value judgement. Minimizing dollar costs or maximizing efficiency depends on you choosing certain constraints.

Not to mention that maximizing efficiency is not entirely value free either. It is an implicit judgement of efficiency vs other considerations.

So when you start building your models to answer certain questions. Your questions will have certain constraints in them, including or not including certain constraints is a value judgement. It is inherently normative.

@Lauren: I do not disagree with you that there is no distinction. What I meant when I said there was no distinction, is that the distinction turns it into gibberish.

You could very well call the (mathematical) techniques used positive, that's the form of in which economics is practiced. The questions asked however are inherently normative. The fact that we largely agree on the norms does not make it less normative. Nor does the application of these techniques to these questions make them or their answers valuefree.

Now the problem we have here is one of domain, what do you consider to be the scope of economics? I consider the questions asked and the answers given to be part of the field. That's why I think there is no positive/normative distinction and I don't think you can ask questions without being normative.

I agree with you that we should be well aware of the normative elements. To state however that there is a distinction is where we part ways. I don't think you can remove the normative part any more than I belief you can remove the beating heart out of living human being without consequence. It ceases to be meaningful.

If you'd say well I disagree I only think that technique is part of the field, then you're absolutly right, it is purely positive. However I doubt that would be your stance. As that would make economics indistinguishable from mathematics.

The problem I have with the question as formulated by Prof. Caplan is "should" and that it takes a policy perspective. If you take economics as a policy science there is no distinction. Stating that it is merely positive is like randomizing your starting point for a random number sequence, there is still a starting point, it is however not explicit.

Jayson Virissimo writes:

"the sadist torturing a victim is Pareto"

No it isn't. A Pareto improvement is where you can move from one state to another state while making at least one person better off without making any other person worse off. Moving from a state without torture to a state with torture makes at least one person worse off and so is not a Pareto improvement. You are confusing Pareto efficiency with some kind of utilitarianism.

Kevin Dick writes:

As it happens, I have been contemplating my own blog post on this topic for about a month.

The key distinction in my mind is between instrumental and terminal values. Economics shouldn't be normative about terminal values. Economists have no more right than anyone else to tell other people what ultimate goals they should strive for.

But it should be normative about instrumental values: "Oh, your terminal goal is A. Economics research shows that instrumental values X, Y, and Z lead to A. Therefore, you should pursue X, Y, and Z. Research further shows that policies M, N, and O are the best ways to achieve X, Y, and Z so you should devote the most resources to M, N, and O."

Moreover, given a hypothetical set of _societal_ terminal values, economics should (to the extent they are able) present the list of actions that will lead to the greatest production of those values.

I'm puzzled by the way Bryan phrases the three positions on his list. The first sentence of each position is normative ("Economics should be ..."), the second positive ("Economics is ..."). So when he asks me which position is mine, I don't know whether he's asking me for my normative position (on how economics ought to be practiced) or my positive position. And if the latter, is it for my position on what the term "economics" denotes, or instead on how economists actually behave?

In making these distinctions I agree with Mario Rizzo's comment above, and with Lionel Robbins. There are propositions of positive economics, of economic theory and economic history that are uncolored by the researcher's ethical judgments. (Alternative test: they are accepted by economists across the spectrum of ethical positions.) And there are propositions of normative economics, of political economy (advice to policy-makers) and of home economics (advice to consumers) that combine positive economics with value judgments. The value judgments come from "outside" in that they (obviously) do not derive from the propositions of positive economics.

Dan Klein has a point: an economist who makes an affectation out of avoiding "ought" statements, while still wanting to get across his policy judgments, will often end up using awkward circumlocutions. Or he will smuggle norms into his language under sheep's clothing, for example talk about "social welfare" as if that were a concept of positive economic theory. Still, as Robbins taught us, a wise man recognizes and judiciously applies the distinction between positive and normative. He does not abstain from normative statements, but he recognizes them for what they are.

Tobi writes:

Prof. White, isn't that simply limiting economics to catallaxy? I won't disagree that the latter is positive, it 'merely' studies the mechanism. However I'd think that 'economics' as distinghuished from catallaxy is broader.

Daniel Klein writes:

Thanks Larry.

I don't think "uncolored by the researcher's ethical judgments" makes much sense. Aren't judgments (presumably "ethical judgments") -- and the purposes they serve-- involved in deciding the terms, interpretations, formulations one practices? The problems taken up? The alternatives considered? The arguments put on the table? The audiences to address? The literature to relate to?

We may identify a kind of scholarly discourse practiced by ethicists, such as Peter Singer. Nonetheless, something much broader, ethical judgment, is not some specialist area of science. It is inherent in being human. And every economist is a human. "Economist" is a subset of "human."

You write: A wise man "does not abstain from normative statements, but he recognizes them for what they are."

Well, all statements involve ethical judgments, so all statements are "normative." Once we recognize that, what's the point of saying a statement is "normative"?

Again, the term does mean something to people. Better terms are available, so junk "normative."

Practicing "pos/norm" ways, you imply that "social welfare" is not "a concept of positive economic theory." Why not? Because it is loose and vague? Because there is much disagreement about how to express or characterize it? If so, will these qualities hold up as definition of "normative"? Is a thing's being vague a necessary condition for its being "normative"? A sufficient condition? Is a thing's being controversial or outside the 40-yard lines a necessary condition for its being "normative"? A sufficient condition?

Again, I say scrap "normative" and just say "vague," "controversial," "libertarian," etc.

Eapen Thampy writes:

I think the work of economics is in positive description of reality, but the meat of economics is in identifying sound decisions in every part of life.

zefreak writes:

Tell me "positive" or "normative" for each of the following:

(1) The minimum wage ought to be repealed.

(2) I think the minimum wage ought to be repealed.

(3) The minimum wage reduces social welfare.

(4) Wise people oppose the minimum wage.

The primary verb of (1) is an ought, while the primary verbs of (2), (3), and (4) are ises. But all four statements are really the same."


Daniel, I think you are confused.
1. Normative
2. Positive (if it is indeed a fact that you believe so)
3. Positive (if it is indeed a fact that minimum wage reduces welfare)
4. Positive (although its truth value can be debated.

The 4 are not the same. The first is universally prescriptive and from a non-cognitivist standpoint is meaningless. The other three are statements of fact (whether valid or not) and thus have truth values.

I think your confusion stems from the colloquial usage of the word 'should', as most people equivocate between position 1 and 2, or the universally prescriptive and 'I feel that' uses of 'should'.

"Ises and oughts are easily and naturally translated into one another, based on the purposes of the interlocutors and the discourse situation."

If you mean that by using an 'if-then' framework you can translate an ought to an is, then yes. But such an analysis is still not normative, as you are assuming the truth of the ought to derive further oughts, not asserting the truth of the ought.

As an example, if I were to say "for the sake of argument, IF Yahweh exists THEN praying to him would be a good idea" does not assert that Yahweh exists nor does it assert that anyone should pray to him. It is a positive analysis contingent on the validity of its assumptions. This is what economics does.

David: "(Take, for instance, the standard assumption that mutually-agreed upon trade is mutually beneficial. We know already that people can be influenced - advertising is a very big industry! Nonetheless, it is mathematically much simpler to just assume equivalence)."

The assumption that voluntary trade is an example of slipping value statements into a positive science. Sure, most people do believe that such trade is mutually beneficial, but they are applying economic theory to their value system in order to derive such a conclusion. Such a conclusion (or any conclusion of value) cannot be derived from economic theory alone, but must be applied to a system of values or teleological end.

I will give you that certain epistemological values such as Occam's Razor apply, so it is not strictly value free. But that is a long way from asserting objective value or moral realism, which normative economists seem to assume.

Les writes:

The comments show one thing for sure. Few people understand the terms "positive" and "normative" even though Bryan neatly defined them at the outset as "what is" and "what should be" respectively.

Dan Klein writes:

I don't think "uncolored by the researcher's ethical judgments" makes much sense.

How about this statement: Caesar crossed the Rubicon in 49BC. The statement is positive. It as true to those who admire Caesar as to those who despise him.

Dan continues:

Aren't judgments (presumably "ethical judgments") -- and the purposes they serve-- involved in deciding the terms, interpretations, formulations one practices? The problems taken up? The alternatives considered? The arguments put on the table? The audiences to address? The literature to relate to?

Granted. But it doesn't mean that every particular statement the researcher makes is normative.

Dan:

Well, all statements involve ethical judgments, so all statements are "normative." Once we recognize that, what's the point of saying a statement is "normative"?

I can't agree. "Caesar crossed the Rubicon" isn't normative. "1+1=2" isn't normative. "The demand curve for bananas slopes downward" is not normative.

Practicing "pos/norm" ways, you imply that "social welfare" is not "a concept of positive economic theory." Why not? Because it is loose and vague?

No, because a "social welfare function" has no analytical use, as a demand curve does, but only an ethical use.

The first position is self-contradictory:

"Economics should never be normative."

Here, the speaker is using the word "should" in a normative sense, so it is making a normative statement. And this statement is about economics. So he is doing a normative statement about economics, the very thing it is arguing against.

Tracy W writes:

zefreak: Economics can inform normative arguments, but when economists engage in normative debate they are doing so outside of their capacity as economists.

I don't see the distinction. If I engage in normative debate that is, at least on my side, informed by economics, how am I doing outside of my capacity as an economist?
Of course I might be incompetent at applying the economics, but that is also possible if I engage in purely descriptive economics.

Robert: Few people would argue that engineers are seeking to impose their values on society while physicists are merely describing the natural world.

I don't know about "few". But many people have been opposed to inventions like trains, cars, radios, cellphones, because they change societies' values. And obviously many people were opposed to nuclear weapons (which required physicists as well as engineers to build).

ajb: I think economists should always be neutral when speaking as experts and promote advocacy as individuals only.

But how can you separate expertise from advocacy? An economist can't switch off all the bits of their brain that know about economics when they decide to promote advocacy, nor can I think of any reason why they should. (And when were experts not individuals?)

Tobi: Physicist, engineers and chemist do not impose their values on society for they're not describing society. Their object of study, how relevant to society it may be, has nothing to do with society.

This is false. Things like cellphones or the Internet have changed society, and the engineers and scientists who introduced the Internet were expecting it to change society. In my engineering course we were advised to think about the impact of our work on society (a pointless bit of advice, as they provided no training in working out what the impact would be, but it was there).

This also brings me back to the natural sciences, physicists have not divided their science in ideological cleavages.

Again, this is wrong. Consider the debate about to what extent physicists should engage in weapons research.

Liberty: If an economist goes on television and says "Heath care bill 3200 is the wrong direction for America. It spends too much, it will put our children in debt, and it will restrict our freedom..." then he is no longer acting as an economist.

He might well be acting as an economist. Whether he is acting as an economist or not depends on whether he has done the economic work behind his statements. Eg is he making up statements like "it will restrict our freedom" out of thin air, then he is not an economist.

...not analyzing the means that have been chosen in order to achieve agreed-upon ends.

What agreed-on ends? What country/planet are you living in that has agreed-on ends? And, even if these ends have been agreed on, why shouldn't an economist (or anyone else) try to change them? Our view of what ends are desirable depend in part on what ends are achievable and at what cost. I don't see any reason why we should ignore economics expertise in setting the ends. (Which is not to say that economics is the only relevant expertise).

zefreak writes:

Maurizio: "The first position is self-contradictory:

"Economics should never be normative."

Here, the speaker is using the word "should" in a normative sense, so it is making a normative statement. And this statement is about economics. So he is doing a normative statement about economics, the very thing it is arguing against."

This is patently false. "Economics should never be normative" is a statement 'about' economics, not a statement 'of' economics. If anything, it would be considered a meta-economic proposition.

zefreak writes:

Tracy: "I don't see the distinction. If I engage in normative debate that is, at least on my side, informed by economics, how am I doing outside of my capacity as an economist?"

I'm sorry if you don't see the distinction between policy informed by economics and policy 'being' economics.

Economics deals with how humans act and what institutions help or prohibit coordination between participants' varying ends. Which ends ought to be pursued belongs to a very different field of study.

Notice I am not stating that Economists cannot make policy advice or express normative positions. However, they do so as human beings and not in their capacity as economists (just as I express preference in music and film but do not do so in my capacity as a computer technician).

"What agreed-on ends? What country/planet are you living in that has agreed-on ends? And, even if these ends have been agreed on, why shouldn't an economist (or anyone else) try to change them?"

You misunderstood me. I did not imply universal ends nor did I state that economists shouldn't try to change them.

"Our view of what ends are desirable depend in part on what ends are achievable and at what cost. I don't see any reason why we should ignore economics expertise in setting the ends."

This is false, as economic expertise cannot help in deciding ends. Perhaps you do not understand the means/ends dichotomy. Ends are not chosen, means are chosen to satisfy given ends.

Tobi writes:

@Tracy, cellphones may be relevant to society, however cellphones have nothing do to with the study of society. Hence I made the distinction.

Physicist may or may not decide to do weapons research, that's a person's ethical judgement. However molecules, atoms or the laws of physics still look the same to you whatever your ethical judgement may be.

Society however looks quite a bit different when you're a socialist than when you're a capitalist, to put it crudely. Take child labour for example.

Doc Merlin writes:

@ Neal W.

"(1) Economics should not be normative.

Economics just tells us what is, and when an economist gets normative, (s)he puts on the hat of an ethicist."

Economics has its roots in the moral ethicists of the Scottish enlightenment. I think its pretty impossible to separate the ethics from the behavioral description.

And Daniel Klein says my thoughts here very clearly.
Imo: The normative/positive distinction is the result of us trying to be more like physics, a field where the distinction makes sense. In the social sciences it doesn't make sense, its just one of wording.

zefreak writes:

Can you give me one example of a true statement of economics that is 'by necessity' normative?

For example, stating that "price ceilings cause shortages" may be true, but the truth of the normative proposition "we should not have price ceilings because it will cause shortages" is contingent on the truth of its value assumption, namely that shortages are bad.

I realize that epistemology and meta-ethics is not everyone's forte, but I don't see how anyone who has read either Hume or Mises (whom I imagine to be popular among this blog's readers) can believe in objective value which economics as normative science requires (remember that conditional statements like 'if x then y' is not normative).

Jeremy, Alabama writes:

I choose (1).

This understands economics as how people react to forces.

Option (2) creates forces for people to react to.
Option (3) is naked force.

Daniel Klein writes:

Larry,

Thanks for the reply.

I listed four sentences to show that looking at an individual sentence apart from the rest of the speaker's statement, and assessing the individual sentence as "positive" or "normative," isn't sensible.

Zefreak above says that my statements (2), (3), and (4) are positive. I can write a whole piece using such phrasing. Would my piece then be less "normative"?

Likewise, someone can build a model or report an empirical investigation and write a string of sentences each of which you would call "positive" (or perhaps "analytical", a term you introduce). But still there are purposes and (ethical) judgments in the endeavor. You would have to say that the author writes only "positive" sentences. Does that mean the author isn't being "normative"?

Doc Merlin says the distinction doesn't make sense in economics, whereas it does in physics. I think Doc is mistaken. Not that the distinction makes no sense, but even in physics I don't think it is worthwhile. There are better words for what people are trying to say when they talk "pos/norm".

Michael Polanyi sheds light:

"The words I have spoken and am yet to speak mean nothing: it is only I who mean something by them."

(Personal Knowledge, 252)

We have to ask what the speaker means, and to do that we read his sentences as an expression of his purposes, intentions, character, judgments. With every statement there is an I, just as with every pitch there is a pitcher, with every shot there is a shooter.

Tobi writes:

"Shortages" are people substituting time (standing in line) for money. The market still clears though not necessarily due to money transactions.

You could say well not everybody with the willingness to pay gets what he or she actually wants, but that's merely looking at the money at their disposition, but not their time. Money is the common means of exchange, it is not the sole means of exchange.

Yes the gains from trade are lower, because you cannot pay somebody through standing in line (unless the particular shopkeeper derives some gain from it). However the mechanism does affect the allocation/distribution of goods and this trade-off (lower gains of trade vs allocation) is a normative choice.

Calling that "shortages" does seem to be a normative judgement prefering one form of allocation over another based on future gains that benefit society at large. Or would you say that calling something "shortages" is valuefree?

Don't forget corporations (Apple or fashion designers) create artificial shortages just to create advertisiment through the people standing in line. We however only condemn institutionaly created shortages (by government) as 'inefficient' based on the fact that it is not voluntary on the part of 'the producers' in the alternative.

That however is an implicit value judgement in favour of liberty on utilitarian grounds as providing the greatest benefit to the most people by putting an equal value on each persons gain.

You could however also say that you value the gain of those that have the least higher, because you believe that narrowing the gap creates the greatest happiness for society.


Mario Rizzo writes:

I side with what Galileo said, at least according to legend: "But it still moves..." Dan Klein is making confusions where there are none (in this way he is like Amartya Sen [ouch!]).

Where is the normativity in the Law of Demand? I may be interested in applying it to apples because I think people should -- or should not -- eat apples. But so what? People may cause the price of pork to fall because they suddenly think pork is unclean to eat. But that is their value-judgment. I may decide to study a particular problem, like AIDS transmission and such, because I care about people dying of the disease. But that is scientific motivation, not science itself.

Telling the truth about your research is an ethical-normative rule but it won't turn stones into bread. It has to do with the norms of human (scientists') interaction.

The dedication to science instead of magic is a value judgment but again this goes to why I engage in a certain action and not the content of the action itself.

If Dan Klein wants to enter the world of mixing all this up, he will lose the game. There are many expert obfuscators out there (usually leftists) who will wrap him around their little fingers.

Buck Farmer writes:

@Daniel Klein

Vagueness, controversialness, and libertarianess are neither necessary nor sufficient conditions for normative statements.

A vague positive statement:
"Someone did something on some day"

There is nothing immediately normative about this statement. It is purely positive. Like most positive statements it can be used as the hypothetical of a normative statement and so can have normative implications when combined with a normative statement, but by itself it is positive albeit vague.

How it could lead to normative implications:

(1) Someone did something on some day
(2) If "Someone did something on some day" is true, then I should vote for Pat Buchanan
(3) I should vote for Pat Buchanan by (1) and (2)

Strictly speaking, the normative statement (2) is independent of the truth value of the positive (but vague) statement (1). The normative implication (3) is not independent of that the truth value of (3).

For a second, more serious example, see Amartya Sen's paper on the ambiguity of revealed preferences (the title escapes me and google is not being helpful, does anyone remember this?).

The gist is that just because we see someone choose something it is ambiguous what they are not choosing and so specifying some sort of robust preference function is difficult. There're no implications of this--alone--on people's choices.

Greg Ransom writes:

I'm with Larry and Mario on this.

Sometimes it is helpful to make it clear what part is the science and what part are the purposes and values.

If you are building "cargo cult" airport towers and you want to improve things, it can help to get clear about what your aims are and what your means are for achieving these aims is.

If your have a false causal picture of how the world work, this matters, and it won't help to double down on your value judgments.

Similarly, it help to seperate out these things when trying to determine how ethically perverse the globle warming folks are when evaluating their scientific claims about the Medieval warming period or the elimination of all Himalayan glaciers by 2035. We need to seperate out the factual and causal claims to discover how corrupt and scientifically shady these people are -- faking data, hiding the decline, using hack work by political activists as if it were peer reviewed science, and on and on (if you've been informing yourself about the U. Of East Anglia, IPCC and weather station fraud stories.)

Greg Ransom writes:

In short, we need to look at our purposes in separating out the causal claims and fact claims from the goals and values --otherwise you are doing 5th wheel "philosophy" of the kind that good philosophers have no use for (but the journals editors love).

Gene Callahan writes:

Daniel Kuehn wrote: "But the point of positive science is to improve the world."

You do realize that this itself is a controversial, normative claim, right?

Gene Callahan writes:

'Here, the speaker is using the word "should" in a normative sense, so it is making a normative statement. And this statement is about economics. So he is doing a normative statement about economics, the very thing it is arguing against.'

This argument fails because it confuses making a statement *about* economics with making a statement *within* economics. (That doesn't mean the conclusion is (necessarily] wrong -- just that it is a bad a[gument.)

Buck Farmer writes:

@Gene Callahan

I don't know the quotation, but my reading of it is:

"Positive scientists should improve the world (improving the world defined implicitly or elsewhere)."

This could be further stretched to:

"Positive scientists should improve the world by doing positive science."

Or differently:

"Positive scientists should choose to do only that positive science that improves the world."

Regardless of it's controversy, it is definitely a normative claim...but the same type of statement could be made regarding people engaged in any human activity. This doesn't make the activity itself normative.

"Children should eat their vegetables."

This doesn't make being a child normative or vegetables normative. It's a meta-statement establishing a relationship between a bundle of ideas that make up "Children" and a bundle of ideas that make up "eating their vegetables."

Tobi writes:

@Greg, I doubt that anyone disagrees with you on that it is helpful. There is however a world of difference between what is helpful, what is possible and what is.

@Mario, that the law of demand is not normative in itself does not mean that economics is not inherently normative. That is generalizing from one example.
A more interesting question though what meaning does the law of demand have outside of reality? What meaning does it have outside of its context?

Greg Ransom writes:

The "normative" and the "factual" are inter-related in the social process which is science.

So what.

Some talented people have teased out some of the inter-relations -- e.g. Thomas Kuhn, Ludwig Mises, etc.

A lot of academics (typicallypolitically motivated leftists) looking to create publications and who don't know the history of the literature have made a conceptual mess of the the matter.

Beyond what Larry and Mario said, there isn't much worth saying about the matter in blog comment posting.

Is there?

Gunthter40 writes:

Economics qua economics is a science and must be devoid of normativity. As a science, economics attempts to understand the the manifold mechanisms for the rationing of resources and their related trade-offs.

This does not mean that economists (or any other scientist for that matter) must refrain from discussing the moral implications of their field of study. However, economists should be clear about distinguishing their science from their value judgements when they engage in ideological debate.

By objectively and clearly portraying the trade-offs and consequences of economic actions, economists can assist decision makers, whether they be centralized or decentralized.

Daniel Klein writes:

Some of the discussion here makes me think of Gunnar Myrdal.

In The Political Element in the Development of Economic Theory, a very significant book when it appeared in Swedish in 1929, he criticized liberal economics for tacitly mixing values and analysis, and called for separating them carefully.

But in his 1953 preface to the English edition he said his view of the matter has changed. He no longer believes in meaningful separation of values and analysis. He says that values are formative all the way down.

What he then calls for, much elaborated in his 1969 Objectivity in Social Research, is openness about who one is, where one is coming from. The point is not to separate values and analysis, but to make clear one's purposes in the whole so that it can be more meaningfully interpreted and challenged.

He speaks of planes of valuations -- not some single demarcation between values and analysis/fact. The key thing to understand, in my view, is that there is no upper-most plane and no lower-most plane. We are enmeshed within webs of discourse, and there is no getting outside of valuations. We should be more candid about where we are coming from and where we mean to go. Thus, he rejects the idea of discourse free of values. He says lets openly discourse about the comparative worthiness of different values. As I said early, all discourse is "normative" (and hence it does not advance matters by talking "pos/norm").

I've collected a bunch of Myrdal's pertinent 1969 passages on pp. 182-4:

http://econjwatch.org/file_download/97/2006-01-klein-char_issue.pdf

Loof writes:

How normative should economics be?

Besides begging to know what’s now normatively organized as apposed to positively mechanized-mathematized in the vast field and various disciplines about Economics, L believe the answer to the question depends on the purpose of any science done in economics.

In principle: positively, facts are better uncovered about Right; normatively; values are better revealed about Good.

Jay Jeffers writes:

Daniel Klein @ 2:09,

I think I agree with everything you said in your set-up, but I just don't see why you have to draw the conclusion you do, because I don't see that it follows that it does not advance matters by talking "pos/norm."

Hilary Putnam said it best in his book Collapse of the Fact/Value Dichotomy, when he said that making distinctions between facts and values is still perfectly useful and legitimate. It's the dichotomy that has traditionally been set up that is not useful or warranted. That we are enmeshed in a web of discourse seems undeniable, but it does not follow that everything is normative and no distinction can be made between facts and values. "It is 48 degrees outside" is just not the same thing as "murder is wrong."

But if there's no distinction between facts and values, as you seem to want to assert, then there truly is no interesting distinction between the two statements. And if there is no distinction between those statements, then we might as well throw in "ice cream is better than cake." And once we do that, then I have no reason to take anything you say seriously based on the content of what you say, rather, I'll go ahead and use something like your charisma, social authority, or some such characteristic like that.

I mean, at the very least, the argument you're engaged in relies on there being a distinction between the positive and the normative (even if there's no dichotomy). If it doesn't, then anything goes. Seeing the normative nature of even our descriptive endeavors can only be illuminating, but it does not follow that there is *no* distinction between the positive and the normative. There need not be any sort of grand metaphysical or eternal separation for there to be real and principled distinctions between the positive and normative.


Jay Jeffers writes:

To clarify a bit, there need not be any sort of metaphysical dichotomy discovered in order to make the distinctions, because it's sensible enough to say that some statements are instrumentally useful but still normative, while others are merely normative. And while there is normativity present throughout our discourse, some statements are merely positive.

"There is an external world" is merely positive (in spite of the normative course we take to build a language) because of what it aims at, even if it turns out to be unwarranted (I won't hold my breath).

"Murder is wrong" is merely normative because it's not instrumentally loaded at all unlike a statement like "all other things equal, the most parsimonious theory ought to be accepted."

I don't pretend to have highlighted all the relevant ways to make the distinction, or the most important, I'm just sayin', right off the top of my head, it seems like there's good reason not to throw the baby (the positive/normative distinction) out with the bathwater (our forgetfulness of-obliviousness to the normative structure of our web of discourse).

I have been thinking about this issue a fair bit as I oversee academic research that might be helpful to the Financial Accounting Standards Board.

In my view, the traditional positive-normative distinction is very useful. The FASB wants to answer normative questions about what they should do. But assertions about what should be done invariably rest on assumptions about positive relationships of what is or will be true.

My job is to marshal academics to provide evidence on the positive assumptions that inform normative recommendations.

I've written a bit about this topic a blog devoted to financial reporting research and standard setting here .

zefreak writes:

Tobi: You focused too much on the incompleteness and poor wording of my example (which I simplified for good reason) and missed my point. Replace 'shortage' with 'decrease supply relative to demand' if you think shortage is a value-laden term.

Gene, Mario and Larry are right on.

Daniel Klein writes:

To Jay Jeffers,

Thanks for commenting.

Again, I don't say that people mean or say nothing when they talk "pos/norm." I say that they could do better.

You speak of there being "a real and principled distinction between the positive and normative," but you do not say what that distinction is.

With the case we may infer varying distinctions from "pos/norm" talk. I can't offer up a single, all-purpose replacement.

One reason to learn to disentangle the cluster of connotations of "pos/norm" is that "pos/norm" is a club used often to do evil -- to shut down challenge, deeper inquiry, expansion of considerations.

Jay Jeffers writes:

Daniel Klein,

Thank you for the reply. I think this is a fruitful discussion.

OK, I didn't say what the distinction is, true enough. So as you know, a positive statement is a statement used to express a state of the world, regardless of whether anyone likes it. I personally do not like how cold it is right now, but that's too bad for me.

A normative statement, on the other hand, is a statement about how the world ought to be. I know I'm not telling you anything new, but I say this to illustrate the point that we can already tell the difference between a statement about the way the world is and one about the way the world ought to be. The fact that our ways of telling how the world is involves a value-laden process does not necessarily make our venture less likely to grasp the real world.

It's true that our whole web of discourse (and maybe even our "field of discursivity" ;) is normatively policed, but it still is used to express different things. Some of those different things are statements about the way the world is on the one hand, and the way it ought to be, on the other.

When I say it's such and such temperature outside, part of what allows me to do that is that we've already done the normative work, and we can get to describing the world (almost like a basketball team makes the lessons learned in practice manifest in the real game, this doesn't make the game any less real). As for how understanding the environment via science if concerned, any normative struggles we have in deciding how to proceed scientifically are instrumental, as opposed to arguing over intrinsic goods (say, equality versus efficiency).

In your earlier post, when you say,

"The words 'positive' and 'normative' do not mean nothing, but what they mean can always be expressed in better terms. 'Normative' often means outspoken, unconventional, strident, etc. It can also mean loose, vague, and indeterminate."

I haven't seen any better way to distinguish between the way the world is and the way the world ought to be. Something could be unconventional, controversial, etc and still be either positive or normative. Now if we have no confidence that we can apprehend the world except through a normatively colored lens, that's fine, so long as we believe the image isn't very distorted. But if we have no confidence about how distorted our image is, then there's no reason not to toss epistemological standards out the window and start using pure rhetoric and forms of emotional pressure that are currently frowned upon (even if sometimes employed) in the academy.

An important kind of rhetorical pressure we aim at in the contemporary academy (ideally) is showing our colleagues and interlocutors how the world actually is. True enough, the fact that they're supposed to care about that sets the stage for a form of normative discourse, but that doesn't necessarily mean we've distorted the world. For example, peer review, and the justification behind it, is value-laden, but there's no reason to think that processes which serve to weed out bias and error make our ventures less objective.

it's important to us that we're apprehending the world the way it is. I'm not trying to say that you dispute that, rather, I'm trying to appeal to your sense of realism, to express that I have heard no other way to insure that we're talking about apprehending the world the way it is, other than through the word "positive."

On the other hand if a person says that equality is the best state of affairs for humanity, and someone else contradicts that, and asserts that it's better to make the pie bigger so everyone has a bigger piece, even if that means more inequality, we're really just at a loss to adjudicate this dispute. The word "normative" seems like a good way to describe the field in which disputes like this take place. So fighting over the axioms that undergird our systems of thought seems like a normative process.

Now I know there's an obvious middle ground, and that's the instrumental way we use normative thinking to asses (or perhaps, normative policing to enforce) the proper ways of speaking and using concepts, both moral and ostensibly non-moral ones. But I think paying attention to the difference between moral arguments that involve disputes over instrumental means on the one hand, and those over intrinsic goods (if such a thing can be argued over) on the other, can keep us in line. Aside from the poles, and the middle, I feel I should stress that I imagine you're aware of all this already, but I make a point to express it the way I did, because I don't think any of the stand-in terms you've offered do the important work the the pos/norm distinction does (even if it is an overused weapon).

if there's a better way to signal to people that we're talking in some contexts about the way the world is, and in others about how a speaker would like the world to be, then I'm open to it.

As for learning to disentangle "the cluster of connotations of 'pos/norm'" I'm with ya 100%; I think this is great stuff. I had, however, thought you were saying something more radical than that earlier in the thread.

(sorry for the long post)


Greg Ransom writes:

David writes:

"We are enmeshed within webs of discourse, and there is no getting outside of valuations."

Really?

The atom bomb either goes off at the test site in New Mexico or it doesn't.

Sure there are lots of enforced norms in the process of producing models and equations making it all possible, but again, so what.

It's pretty easy to separate all those different shared norms or whatever you want to call them in the community finally managing to tape together the bomb -- from the fact that all of that either did or didn't give us a big flash in the sky.

All sorts of purposes were behind the big flash in the sky -- and there is ultimately no getting perfectly clear about them, many of them never articulated and only reconstructed after the fact.

But we have no such problem about whether or not there was a big flash in the sky, of if none.

Myrdal is making a hash of things, putting the hard stuff before the cart.

Greg Ransom writes:

Daniel writes:

"We are enmeshed within webs of discourse, and there is no getting outside of valuations."

Isn't it actually the case that those who usually emphasize this point do so to throw sand in the practice of good science -- i.e. they invent "ambiguity" purposely in order to delegitimitize knowledge and scientific achievement itself, for political purposes?

This is the legacy of Marx and historicism and "Western Marxism" and "Post Modernism" and all the rest isn't it?

Daniel writes:

"We should be more candid about where we are coming from and where we mean to go."

Exactly. And the folks who make a hash of the human purposes / good causal explanation distinction need to be nailed for where they are coming from and where they want to take us -- most of them are coming from a politically motivated leftist political starting point and they want to take us in the direction of delegitimizing sound knowledge, to be replaced by a fully politicized "science" or "knowledge".

Think of the same effort in jurisprudence, where leftists (e.g. legal realists or critical legal theorists, etc.) aim to advance the idea that all legal interpretation is merely an act of power politics -- it's all non-shared normativity untied to any shared factual reality of any sort.

They come from a desire to overturn the current order and they aim to get to a completely politicized leftist legal regime, the facts of the matter be damned.

The analogy is less than perfect -- and you can intentionally misunderstand the point if you wish -- but the leftist agenda and its aim at a political war on achieved knowledge and shared understanding is made plain.

Tobi writes:

zefreak, does it really matter what you label it? I used "shortages" in my post because I mostly focused on the content of the label and not the label itself.

It is in your description of something that you're passing normative judgement. Saying that you're objective simply means that you don't know your bias yet.

I have a strong bias in favour of classical liberalism and that's why I liked the austrian school and the chicago school. Their picture of reality heavily corresponds with my bias.

This doesn't mean that there is no such thing as truth or meaningful statements, however I do think I agree with Myrdal then that we should be honest about what our values are.

The social sciences are a distinct science for a reason and I'd even say far more difficult than the natural science for the very simple reason that we're describing other humans as humans ourselves.

@Greg, what you're saying sounds an awful lot like a conspiracy theory. If it isn't a conspiracy theory, isn't the very fact that 'leftist' have a different picture of reality support for inherently normative social sciences?

You could say that they made it normative, but that would be rather unfair as before 'them' there wasn't really a plurality of perspectives. The fact that there was a single perspective doesn't make that perspective any less normative.

If it is a wrong perspective and there is a single objective perspective then we should be able to tease that perspective out through argumentation on why it is wrong. The fact however that the battlelines are pretty much drawn along ideological lines doesn't really help the case of a merely positive social science.

Daniel Klein writes:

Hi Jay Jeffers,

Thanks for your latest. Yes, fruitful stuff. However, I will need to leave the conversation. This will be my last entry, I expect.

Your distinction doesn't wow me.

Are the following not descriptions of "a state of the world"?:

(1) The world is such that if min wage were repealed, social welfare would increase.

(2) I think we should abolish the min wage.

(3) Wise men think we should abolish the min wage.

As for your characterization of "normative," isn't it the case that saying "1 + 1 = 2" can be read as "You should believe that 1 + 1 = 2"?

My view is not to deny the moral realm, but to see it in every utterance. With better practice we may better discriminate between the moral judgments people make.

Kind regards,

/Dan

Jay Jeffers writes:

Well Dan,

Now that you're leaving, maybe it's a bit ungentlemanly to sneak in the last word, but I'll just say that no, I do not think "1+1=2" can be read as "You should believe that 1+1=2."

Those are different statements. Assuming a realist view of mathematics (for the sake of argumentative convenience) "1+1=2" is a fact, while whether we should believe it or not is a value. Now, we haven't yet gotten into whether values are real or not, such a conversation would be neither here nor there.

Whether one thinks that facts are the only real things, while values are arbitrary human constructs, or whether one thinks there are simply different kind of truths (the descriptive and normative) the issue remains the same. Nothing is what 1+1=2 means, has anything to do with whether I should believe it to be so. Now if the normative realm is real, and its truths include epistemological duties, then the fact that 1+1=2 *does* have bearing on what I should believe, so in such a scenario there would be no dichotomy. But it's important to say that nothing in the bare meaning of statement "1+1=2" implies what I should believe. And that's what the is/ought distinction is really about.

Anyway, good talk.

Jay

Tracy W writes:

zefreak: Economics deals with how humans act and what institutions help or prohibit coordination between participants' varying ends. Which ends ought to be pursued belongs to a very different field of study.

I don't think the two are neatly separable. For example take a fully centrally planned economy in which all decisions are made democratically. Some people might favour this as an end in itself, say they place a high value on democracy and think all decisions, including decisions about which resources to utilise and how, should be made democratically. Now, if it's not possible for there to be a centrally planned economy (democratic or otherwise) then doesn't that make that end less desirable? (At least to people who found central-planning a good idea, anyone who values market economies for non-economic reasons is clearly not affected by increased economic evidence in favour of market economies).

Or take the end of eliminating all deaths due to accidents. An economist would say that "all" is only achievable at too high a cost, this can change opinions about the desirability of the end. I've managed to convince people of this in arguments, at least briefly.

Notice I am not stating that Economists cannot make policy advice or express normative positions. However, they do so as human beings and not in their capacity as economists ... (just as I express preference in music and film but do not do so in my capacity as a computer technician).

But this is different. A computer technician doesn't have any expertise relevant to music and film in their role as a computer technician (of course there is nothing stopping someone being simultaneously a computer technician and an impressive musician). But an economist can have expertise relevant to making policy advice and expressing normative positions, for example the desirability of making decisions democratically versus making them via a market.

This is is false, as economic expertise cannot help in deciding ends.

This statement would be rather more convincing if you gave reasons for this assertion.

I did not imply universal ends nor did I state that economists shouldn't try to change them.

So what did you mean to imply when you said "agreed-upon ends"?

Perhaps you do not understand the means/ends dichotomy. Ends are not chosen, means are chosen to satisfy given ends.

Perhaps I do not understand the means/ends dichotomy, perhaps you are wrong to think that there is dichotomy at all. I think this assertion "ends are not chosen" is only true if you define ends very broadly, eg utility, or for particular definitions of "chosen". Take for example that one might feel a conflict between the ends of happiness, and of valuing autonomy for its own sake. Now if it turns out as an empirical matter that increased autonomy increases happiness, no need for any such conflict. (This is admittedly a matter for psychologists more than economists, but let's say that an economist applies their econometric skills to this question).

Tobi: Society however looks quite a bit different when you're a socialist than when you're a capitalist, to put it crudely. Take child labour for example.

But on the other hand, empirical evidence can and has moved people away from being socialists. Socialism, in the sense of centrally-planned markets is a lot less respectable since the collapse of the Soviet Union and in particular the differing experiences of West Germany and East Germany. People can fool themselves more about living standards than about molecules and atoms, but in the end the difference in wealth became undeniable.

Jay Jeffers writes:

And, not to put too fine a point on it, but most of us don't have any trouble believing in the reality of the external world. In other words, though refuting the skeptic has been a preoccupation of philosophy, the more relevant dispute is between people who believe in such a thing as moral truths ("murder is wrong" is true) and those that don't ("murder is wrong" either has no truth value or is false on account of the lack of moral facts).

Now one cannot deny the role that normative thinking plays in science, in terms of its *instrumental* use. Nevertheless we're still left with no way to settle things up between moral realists and anti-realists. If someone denies the external world, however, we feel pretty confident that they hold an unwarranted position.

So all this goes toward confidence in science and hesitation over morality. BUT, the reverse is also a problem, insofar as we may believe we have no way to apprehend the world as it is, but rather only through the imposition of our (possibly skewed) perspective.

In such a case, *everything* would be normative, which could (in my view, would) have a counterintuitive, perhaps ironic, result: nothing would be sacred in science anymore. Whether or not we believed things would no longer be constrained by good old fashioned truth about the world, but would only be a function of some amount of social pressure we're under or social alliance we hold. Now, maybe you already believe that, it's just that I would be surprised if you did, because presumably, your own work represents ways you've illuminated certain facts about the world, facts that won't be changed when one or another of your colleagues sneers and jeers at them.

This disagreement has some of the same form of one in the philosophy of mind between dualists and everyone else. What someone like Donald Davidson has done is assert his commitment to materialist monism, and say that mind is nevertheless something that can't be lawfully reduced to matter (a position called "anomalous monism"). Some other materialists are emergentists, in that they believe in the primacy of matter, but believe that mind "emerges" in a sort of 'sum is more than the whole of its parts' view. What I'm driving at is that other than the materialists on one end that claim that mind is either reducible to matter or eliminate consciousness all together, it doesn't rid of us of the concept of mind just to say everything is material. The flip side is idealism, which denies matter, asserting everything is mind. But this view doesn't say that we can walk through walls, the world has all the same laws we believe it to, it's just that its fundamental nature is mental, not material, according to this view. But see even here the concept of matter, and the way we use it, hasn't been discarded.

So, in a similar spirit, perhaps moral truths exist, in which case "murder is wrong" is a fact like "It is 48 degrees outside." And perhaps we can't escape our web of discourse, so we can't really see the world in the pure, naive, unadulterated way we normally take to be the case. Now, no matter what the outcome of these questions, the positive/normative distinction will still do real work (aside from whether it's tossed around too hastily) just like the mental/physical distinction does real work even among those that deny the existence of one or the other (or at least deny the fundamental ontological separation).

We could just call positive and normative "black" and "white," but that wouldn't change the work being done. Your position, is that the work I point to isn't really being done, or no longer needs to be done, I think.

Well anyway, in either case, I'll give it a rest.

zefreak writes:

Tobi:

"It is in your description of something that you're passing normative judgement. Saying that you're objective simply means that you don't know your bias yet."

So if I describe my cat as having two ears and three legs am I being normative? How do you define normative (serious question)?

I will concede that the epistemological framework from which I draw conclusions and judge truth and falsity are normative, in that they delineate what should or shouldn't be believed. However, I am not a universal prescriptivist and don't pretend that everyone should share it. If this is what you mean by everything being normative, you might want to rethink your position.

Regardless, epistemic normativity is a far cry from the sort of normative statements that find their way into economics and science.

I do not think much progress can be made in this discussion without getting into meta-ethics and epistemology, which the comments section might not be the best place to do so. So I'll leave with this:

Tracy: I was using 'end' in the broadest sense.
Tracy, Dan and Toby: I disagree with you because I am a moral non-cognitivist and anti-realist. I think all values should be stated as such, and perhaps if I had stated my own epistemological values at the beginning this would have been less confusing.

Tracy W writes:

Tracy: I was using 'end' in the broadest sense.

In which case it is fine for an economist to argue that values should be something in particular, as they are merely expressing a (possibly informed) opinion about what values are best for achieving the very broadly defined end, of something like "utility" or "happiness".

I think all values should be stated as such,

I don't think so, as it would take too long and no one would read/listen to the statement of all the values anyway.

zefreak writes:

Tracy: "In which case it is fine for an economist to argue that values should be something in particular, as they are merely expressing a (possibly informed) opinion about what values are best for achieving the very broadly defined end, of something like "utility" or "happiness"."

Expressing opinions is fine, but opinions and normative assertions are very different beasts. 'x is bad' and 'I don't like x' are not the same; the first is meaningless (except to the extent that it is cultural shorthand for 'I don't like x', the emotivist argument), while the second can actually be considered true or false.

There are no universal valued ends (other than happiness/utility, which is by necessity valued) and there are no universally valid means to achieve happiness. I recommend reading behavioral economic research regarding values and frames which supports the Austrian thesis of radical value subjectivity.

"I don't think so, as it would take too long and no one would read/listen to the statement of all the values anyway."

Fair enough. Of course, when I come across an unjustified value statement, or non-contextual normative statement I automatically stop reading so perhaps its a damned if you do, damned if you don't situation. (This is not always true, if it is something I think is worth reading ahead of time, but it is always in the back of my mind)

Troy Camplin writes:

Economics should be both about what is and, knowing what is, make recommendations about ought.

For example,

What is: the best, most detailed knowledge is local, and becomes increasingly diluted the further away one gets from the source.

What ought: therefore, decisions should be kept in the hands of those who have the most, and most accurate, knowledge - meaning, those who have the most diluted knowledge (those in government) should not be allowed to make decisions for those with the most knowledge.

Comments for this entry have been closed
Return to top