Macroeconomic storytelling

by Russ Roberts on October 12, 2011

in Podcast, Truth-seeking & ideology

This week’s EconTalk is with Frank Rose about his fascinating book, The Art of Immersion which tells the story of how the web interacts with storytelling, movies, videogames, and advertising. I really enjoyed the conversation.

In the book, Rose describes how technology lets viewers and readers create their own story and affect the narrative. A primitive version of this took place in the 19th century when the readers of Dickens’s novels which were released in serial form, reacted to the story as it unfolded and affected how Dickens worked on his novel. The modern versions of this are much more dramatic and allow readers and viewers to immerse themselves in a narrative in ways that were not possible in the past.

In the podcast, Rose mentioned how we all like to create our own narrative. I realized how much this applies to sports and yes, especially to macroeconomics, a topic we went on to explore a bit in the conversation. Check it out. That part of the conversation reminded me of Ed Leamer.

Be Sociable, Share!



16 comments    Share Share    Print    Email


Greg October 12, 2011 at 12:02 pm

Russ Roberts is some what of an “economic relativist”. He seems to find that economic arguments (at least when it comes to more vigorous and empirical modern economics) do not really convince anyone of anything. This is quite a common theme in EconTalk. For a Chicago PhD (which Roberts is) graduate this is very strange.

Although Roberts has had a few pobcasts on “truth in economics” he has never really clearly and vigorously defined his position. I wish he would do a podcast specifically on this topic.

Note that Roberts often finds empirical/mathematical results which support his position quite convincing (e.g. Roberts basically accepted most of Robert Barrow’s work without really questioning it on a podcast he did on Barrow).

Seth October 12, 2011 at 1:11 pm

“Note that Roberts often finds empirical/mathematical results which support his position quite convincing…”

I believe he also consistently mentions that he is biased into accepting those results and still skeptical of the studies, even when they do support his views.

Russ Roberts October 12, 2011 at 1:34 pm


Martin Brock October 13, 2011 at 11:37 am

You seem to want a precise, rigorous, mathematical proof of the limitations of precise, rigorous, mathematical proofs. Such proofs exist, in Godel’s Incompleteness theorems and in chaos theory for example, but such a proof may not capture the essence of the problem or explain the problem as well as a Hayek’s Nobel lecture, which Russ often cites.

I’ve taught the first Incompleteness Theorem and understand its principles in detail. I also know something about chaos theory, though I’ve never studied it rigorously, and I know something about the indeterminacy that is fundamentally a part of the standard model in Physics.

I’ve studied probability theory in detail and understand the difference between the law of large numbers and the central limit theorem, so I know why fat-tailed distributions often wreak havoc on statistical models and not only in economics. I can present examples from astrophysics.

I spent a lot of time and money learning these things, and I rarely see much return on the investment; however, we could discuss these issues in a more precise, rigorous, mathematical way here. Most readers would be lost, but I would discuss the issues with you this way. I also suggest that you read The Pretense of Knowledge.

You might also consider the Fourth Commandment and the Orthodox Jewish practice of never writing the name of G-d. G-d does not strike anyone down for violating the commandment, but any image of G-d (including a name assigned to G-d) is necessarily a product of a man’s laughably limited imagination and therefore incredibly childish and naive.

Of his own work, Einstein said, ” All our science, measured against reality, is primitive and childlike — and yet it is the most precious thing we have.” That’s also the idea behind the Fourth Commandment in my opinion.

The Torah is a story about G-d and traditional law, and it’s the most precious story a faithful Jew has, but it’s also primitive and childlike. A Jew must continually remind himself of this fact. He must not confuse a traditional name, and all of its associations within a traditional system, with G-d, just as a scientist must not confuse his theories with the reality that scientific theory seeks to describe.

Of course, that’s only my theory of the Fourth Commandment and the Orthodox practice. I’m not Jewish myself and don’t intend to offend anyone.

vikingvista October 16, 2011 at 3:44 pm

” more vigorous and empirical modern economics”

I assume you meant “rigorous”, not “vigorous”. The problem with empirical macroeconomics is that it lacks economic rigor, in spite of its mathematical rigor.

Greg Ransom October 12, 2011 at 12:13 pm

Very good philosophers of science explain how _all_ of science has aspects of narrative explanation built within it.

Science is about explaining things that _humans_ find puzzling and question raising. If we were Gods there would be no patterns for us which raised questions and which required explanations.

The mistake is to imagine that all explanations are “explanations by elimination” — i.e. that explanations have significance in all by themselves in complete isolation. They don’t. Scientific explanations have their significance in the context of puzzles and questions raising patterns. The explanations don’t eliminate that context and those (once) question raising patterns. The problem is that science textbooks focus on the products of science and they neglect the explanatory nexus of the problem situation where the math formula and “models” — the explanations — were produced (Kuhn).

The danger here is becoming Derrida, Nietzsche, McCloskey, or some other Post Modernist — i.e. “everything” become “text” and everything is perspectival, i.e. there is no such thing as knowledge or learning or getting a better understanding things.

And a lot of what you are saying, Russ, is heading off in the direction of Derrida & French literary philosophy.

And the roots of this attack on knowledge & science in many ways go back to the intellectual attack on classic liberal economics.

This is one reason _why_ they sought to undermine the idea of the growth of knowledge and understanding.

Ultimately, at some point this trend of thought is self-destructive to all knowledge, not just the most controversial — and unfounded — claims made by math economists for mathematical and statistical macroeconomics.

BZ October 12, 2011 at 1:00 pm

Hi Greg,

As another econtalk fan, I would say that my impression is that Dr. Roberts is mostly concerned with excessive hubris. He looks around at the economics profession and sees mostly Sophists on the street corners hawking charts of snake oil. That’s not quite the same thing as nihilism.

For example, Dr. Roberts would say that demand curves slope downward. Yet he will doubt your model that maps aggregate demand over the next 5 years to the third decimal place.

At least, that’s how I read him. :)

Greg October 12, 2011 at 1:39 pm

I agree with Greg Ransom. I often feel that Roberts is a bit too pessimistic about economics as a science, for the lack of a better word.

I myself am doing an MSc in economics at a university with an econ. department which is very theoretic (lots of separating hyperplane theorem, fixed point theorems and all that good stuff).

What I found was NOT that a very mathematical approach to economics pushes one towards liberalism or socialism. What I instead have found is that mathematical approach to economics makes things a lot more clearer and the arguments a lot more consistent and generally well informed.

I know that Roberts has a PhD from Chicago (so he is far more qualified in mathematical approach to econ. than me). But I think he often gives a very wrong impression of vigorous economics. He often makes it sound like mathematical economics is just a load of hot air and totally out of touch with reality. But this is quite the opposite of why mathematics is so important in economics. Mathematics strips any argument naked and makes it crystal clear why a particular view maybe wrong or at least what it may be missing.

My favourite example of this is a simple Keynesian approach to consumption versus a life-time budget constraint approach (RBC). In my first and second years of macro in under-grad. my lecturers did most of their explaining with hand waving (like Roberts suggests). This resulted in me not really understanding what the real difference between a simple Keynesian approach and RBC was. It was only after I saw the equations and the optimisation and the consumption smoothing resulting from optimisation and diminishing utility of consumption that I really got why a simple Keynesian approach was wrong.

That is not to deny that Roberts makes an important point that telling a verbal story is important. What I mean is that that verbal story which is told must be based on vigorous mathematical logic, empirical results and consistent epistemology.

PrometheeFeu October 12, 2011 at 2:21 pm

I think the problem is that most people (including many economists) don’t really understand the math. So they look at things such as the Arrow-Debreu theorem miss half the assumptions or misunderstand them and then go on to tell the world how markets always clear and that’s that. Also, let’s be honest. When an economist assumes a particular shape for indifference curves for the first time, they may think about what that means in the real world. But many people will then build upon that model only ever thinking of “convex indifference curves” completely forgetting that it’s not just a math problem and that assuming convex indifference curves has real implications for the real world. Everybody gets excited about relaxing perfect information assumptions or getting rid of cardinal utility, but those things are quite minor assumptions compared with some of the things economists assume just so they can do the math.

Greg October 12, 2011 at 5:44 pm

I think that the basic assumptions that the preference relation is complete, transitive, there is continuity in the weakly preferred set and there is local nonsatiation are very weak assumptions (as in if you don’t assume these things you are basically calling the representative agent an idiot).

the assumptions i outlined (which lead to convex indifference level sets (curves)) i think go a very long way in providing us with understanding of how the market works.

basically what i am saying is that, if one begins with an agent who is not an idiot one can quite easily show that markets work.

in the 1980s behavioural economics and experimental economics people have pointed out that in fact people are rational in more subtle ways. but on the whole, even in behavioural economics economists are finding that people are very good at dealing with their biases and weaknesses and market forces very often provide goods and methods to deal with these weaknesses of character. but this behavioural analysis would have been entirely impossible without mathematics to describe the problem and describe solutions.

vikingvista October 16, 2011 at 4:09 pm

“I often feel that Roberts is a bit too pessimistic about economics as a science,”

I don’t know if you have any background in the hard sciences, but comparing them to macroeconomics, or much of the social sciences, leaves one not just pessimistic, but downright contemptuous. It is a standard feeling in university settings among majors in physics, chemistry, and engineering. It is unfortunate, because the social sciences are an attempt to understand reality. But it is precisely because of the clear misapplication of scientific principles in those fields. It is the use of scientism among so many social scientists that makes them a joke to real scientists.

The nature of the social sciences requires a greater dependence upon fundamental principles and deductive logic, and that leaves much room for deductive error, and ideological conflict, but there simply is no other way. Ignoring that for empiricism (and ignoring the massive flaws in the underlying statistical assumptions), is a worthless endeavor.

Gray Brendle October 13, 2011 at 6:48 am

This is a very interesting discussion. I listen to Econ Talk frequently and I don’t come away with the impressions that have been expressed here. When I listen carefully I hear a thoughtful discussion between experts. I don’t hear relativism.

I really enjoy the way Russ lets the guest express their position and after they are significantly through it he asks a couple of ‘leading’ questions and then tells the guest that he is probably going to disagree with them.

I think the problem that I see in economics when economics are discussed is that not enough of the discussion focuses on the assumptions that the ‘expert’ uses to make their position.

The econometric explanation of observed data has limitations. That is in the assumptions (i.e., recent online discussion about real wages) so is a commentator wrong to question how to incorporate additional facts/data in the model to make the econometric explanation more complete?

All of this said I am an amateur economist who is a salesman in real life, but I think I get Econ Talk, so Russ keep it up!

Greg October 13, 2011 at 8:26 am

I really like EconTalk because Roberts has some really good guests and he does basically let them just say what they want to say. I also like EconTalk because they keep it strictly econ and don’t do any autobiographical stuff like asking the guest to talk about their childhood.

My problem is with Roberts’ views, not the quality of his podcast. My point is that I see great advantages of mathematical economics over more verbal economics (which Roberts advocates). He always makes it sound like mathematical economics is mostly poop. But its not. Its a much better framework within which one can talk about economics.

Mathematics is a language, which is much clearer than english or russian or german. As a language it can be used to express any idea. So i have seen papers on for example analytic Marxism. Because they were written using mathematics it was much easier to understand what they were talking about, at least much easier than reading Das Kapital (which is not easy!!!). Natural languages can not be used to express relationships which are more complicated than 2 variables.

Take this new arguments between Roberts and Krugman. Both of them just quote papers at each other. But in a more academic discussion this would not happen. Academic economists understand that each paper uses different models in different contexts and different data and the arguments become much more subtle. As in its not like it is now Roberts says “the multiplier is less than 1 because i said so” and Krugman says “the multiplier is more than 1 because I said so.”

Jon October 13, 2011 at 4:38 pm

Two points:

First, I find economics, true economics, to be more a philosophical discussion rather than cold, hard models. After all, we are assuming how people react, and no model can predict this. Yes, QD drops as P increases. But to put a number on it is a fallacy. As a great economics professor (Dr. Robert Wallace, Framingham State University) once said to me when I was but an undergrad: There are some aspects of economics best discussed over a pitcher of beer.

Point Two: Personal Narratives play an important part in our interpretation of events. Not dispute this (or ignore it) is willful blindness. This is why (hearkening back to my first point) economics is more a philosophy than a science; 10 economists can look at the same event (Great Depression, for example) and have 11 different opinions, all supported by math.

Greg October 14, 2011 at 7:53 pm

If you look at modern analytical philosophy it mostly uses mathematical logic. This is because mathematics is the best language for philosophical discussions because it is the clearest language you can get. The same applies to economics.

vikingvista October 16, 2011 at 4:18 pm

Symbolic logic most definitely permits a more precise expression of many ideas (in spite of what some anti-symbolism bigots on this site might say). But some people assume the use of symbolism obviates the need for sound concept formation, or simply use it as a tool for passing sophistry past unskeptical or less fluent peers.

Previous post:

Next post: