ISET

ISET Economist Blog

A blog about economics in the South Caucasus.

World of EconoBlogs

This will be the first of weekly briefs - a summary of interesting posts on different economic blogs. Readers are welcome to suggest their favorite economic blogs or send links to interesting posts they have recently read.

1. Art of economic complexity  by Tim Harford on his excellent blog (a highlight) and also on NY Times shows us the new way of visualizing the difference between national economies.

2. Lucy Kellaway at Financial Times calls to ban the Power Point (you might need to register for free in order to read the article) – the horror!

3. ElectEcon  has an interesting piece about why Fed pumping lots of money into the monetary system didn’t quite work as well as it should have.

4. An interesting post on Worthwhile Canadian Initiative discusses why research higher status than teaching. While I might have some reservations “aboot” it (e.g. the quality of research output is measured by that same researchers that produce it, thus creating a certain bias, which is not accounted for in comparing teaching and research), it certainly is worthwhile to read.

5. Econbrowser has a very interesting material on estimating output gaps. Read it if you are going to work on macroforecasting, or if you are going to work on DSGE models. Or even if you aren’t.

6. Variant Perception discusses a mechanisms of a currency breakup  (a particularly interesting subject, given the problems within the Eurozone) and surprisingly finds that it’s not all as bad as it is cracked up to be.

Rate this blog entry:
21 Comments

Related Posts

Comments

 
Guest - Zak on Sunday, 04 March 2012 16:24

Great initiative! Keep it up.

Skipped the second and the sixth pieces. The letter one was simply too lengthy.

Liked (naturally) the first and the fourth posts. The latter one addresses indeed a very acute issue, or rather puts it forward than addressing it.

Lost interest somewhere in the middle of the fifth post (this was also natural, I guess) but to me there are two greater problems with DSGE:
First is the fact that DSGE models still need external shocks to generate an important feature of modern macro dynamics such as crisis. I would not call any methodology where crises cannot happen without exogenous intervention "a cutting edge"!
The second is that main merit of the approach is introducing micro foundations into micro models which it does in a completely unrealistic way.
Anyway, these thoughts had nothing to do with the post itself, of course.

Great initiative! Keep it up. Skipped the second and the sixth pieces. The letter one was simply too lengthy. Liked (naturally) the first and the fourth posts. The latter one addresses indeed a very acute issue, or rather puts it forward than addressing it. Lost interest somewhere in the middle of the fifth post (this was also natural, I guess) but to me there are two greater problems with DSGE: First is the fact that DSGE models still need external shocks to generate an important feature of modern macro dynamics such as crisis. I would not call any methodology where crises cannot happen without exogenous intervention "a cutting edge"! The second is that main merit of the approach is introducing micro foundations into micro models which it does in a completely unrealistic way. Anyway, these thoughts had nothing to do with the post itself, of course.
Guest - Giorgi on Monday, 05 March 2012 16:53

Zak, I think you'll agree that 99% (if not more) of crises are always caused by an external trigger, which is outside of our control. This, in a nutshell, is a definition of an exogenous shock. Hence, I don't see how endogenization of shocks will actually help.

Zak, I think you'll agree that 99% (if not more) of crises are always caused by an external trigger, which is outside of our control. This, in a nutshell, is a definition of an exogenous shock. Hence, I don't see how endogenization of shocks will actually help.
Guest - Zak on Monday, 05 March 2012 18:51

Oh, Giorgi, I do not agree even a bit!

I believe crises CAN be caused by external shocks, but they NEED NOT. Indeed most recent crises have been generated from within the system and had nothing to do with the exogenous shocks.

I am not proposing endogenizing shocks, I am proposing ditching them and doing towards models that can generate crises without shocks.

Oh, Giorgi, I do not agree even a bit! I believe crises CAN be caused by external shocks, but they NEED NOT. Indeed most recent crises have been generated from within the system and had nothing to do with the exogenous shocks. I am not proposing endogenizing shocks, I am proposing ditching them and doing towards models that can generate crises without shocks.
Guest - Giorgi on Monday, 05 March 2012 19:12

No, I don't agree with that. Systems can perfectly exist in an unstable equilibrium, until something pushes them away from it - that would be an external shock. Even the crises that are generated from within the system needed a push to actually become crises. I agree that the roots of recent crises (and most crises, come to that) lay within the (imperfect) systems themselves, but I don't agree with the view that they can evolve on their own, without any triggers. So, no shocks - no crises.

No, I don't agree with that. Systems can perfectly exist in an unstable equilibrium, until something pushes them away from it - that would be an external shock. Even the crises that are generated from within the system needed a push to actually become crises. I agree that the roots of recent crises (and most crises, come to that) lay within the (imperfect) systems themselves, but I don't agree with the view that they can evolve on their own, without any triggers. So, no shocks - no crises.
Guest - Zak on Monday, 05 March 2012 19:22

I agree with the first statement: "systems can perfectly exist in an unstable equilibrium, until something pushes them away from it".

What I am saying is that shock not a necessary condition. Not at all. They can be generated without any triggers. And you, as a physicist know that :)

So, why would you go around and blame some out-of-the-model(world) forces for every crises? I think we should allow for the possibility of internal crises and try to device policies to avoid them. As long as we treat them as external, the best thing we could do is to device policies to make economies more resistant to those extra-model forces, which can strike from any side in any guise.

I agree with the first statement: "systems can perfectly exist in an unstable equilibrium, until something pushes them away from it". What I am saying is that shock not a necessary condition. Not at all. They can be generated without any triggers. And you, as a physicist know that :) So, why would you go around and blame some out-of-the-model(world) forces for every crises? I think we should allow for the possibility of internal crises and try to device policies to avoid them. As long as we treat them as external, the best thing we could do is to device policies to make economies more resistant to those extra-model forces, which can strike from any side in any guise.
Guest - Giorgi on Monday, 05 March 2012 19:32

I still think that even in physics we need a trigger. Take a ball placed safely on the very top of an inverted-U shaped hill. It is an unstable equilibrium - any force applied to the ball will cause its movement away from that equilibrium point. But you need a trigger, blow of wind, earthquake, something, to apply this initial force. Or take overcooling - you can cool distilled water to the temperatures way below zero, but once you introduce the freezing center - the whole system instantly becomes frozen. I can't recall any examples of such phenomena, not needing an external push. If we were to talk of the quantum world, it's even better :) Everything there is probabilistic - and, with a certain degree of simplification, there's your white noise processes, affecting those systems.

My point is that even internal crises need an unanticipated push. Because, if the push is anticipated, it is already part of the model and hence cannot create a crisis (very simplistic, I know, but still).

I still think that even in physics we need a trigger. Take a ball placed safely on the very top of an inverted-U shaped hill. It is an unstable equilibrium - any force applied to the ball will cause its movement away from that equilibrium point. But you need a trigger, blow of wind, earthquake, something, to apply this initial force. Or take overcooling - you can cool distilled water to the temperatures way below zero, but once you introduce the freezing center - the whole system instantly becomes frozen. I can't recall any examples of such phenomena, not needing an external push. If we were to talk of the quantum world, it's even better :) Everything there is probabilistic - and, with a certain degree of simplification, there's your white noise processes, affecting those systems. My point is that even internal crises need an unanticipated push. Because, if the push is anticipated, it is already part of the model and hence cannot create a crisis (very simplistic, I know, but still).
Guest - Zak on Monday, 05 March 2012 20:57

I have not said we do not need a trigger. I said we do not need a shock (as understood in modern economics) to serve as the trigger.

And I agree with the ball example, just as I agreed before that external shocks can indeed generate crises. No doubt about that. What I am saying, again, is that it is not only external shocks that can do this.

There are many models in physics which display drastic changes in aggregate behavior (the crisis) at some point in time without any external intervention. These are, for example systems that display one type of behavior as long as certain values (which are endogenously changing) stay within certain bands, but display sudden phase changes at point when they leave the bounds (clearly the trigger is crossing the band, but its not a shock). For economics applications of few of these you can have a look at early works of Axel Leijonhufvud of UCLA. In general models with self-oganizion criticality from physics fit the bill. There triggers are not external and there are no internal shocks either, in a sense that there is nothing out of ordinary going on with the variable. But, system suddenly collapses.

Now imagine how much more exciting (and useful) would be the economics discipline if we could have models like this. How much easier it would be to design the policy: it would have a clear aim of keeping a certain variable within the defined bands. Instead at the moment economic policy making looks like asking questions: what type shock should we expect? when is it likely for this chock to occur? how can we resist the shock of certain type? how marge of a shock can we actually survive? And, sadly, we usually never have answers to firs two questions out of these four.

I have not said we do not need a trigger. I said we do not need a shock (as understood in modern economics) to serve as the trigger. And I agree with the ball example, just as I agreed before that external shocks can indeed generate crises. No doubt about that. What I am saying, again, is that it is not only external shocks that can do this. There are many models in physics which display drastic changes in aggregate behavior (the crisis) at some point in time without any external intervention. These are, for example systems that display one type of behavior as long as certain values (which are endogenously changing) stay within certain bands, but display sudden phase changes at point when they leave the bounds (clearly the trigger is crossing the band, but its not a shock). For economics applications of few of these you can have a look at early works of Axel Leijonhufvud of UCLA. In general models with self-oganizion criticality from physics fit the bill. There triggers are not external and there are no internal shocks either, in a sense that there is nothing out of ordinary going on with the variable. But, system suddenly collapses. Now imagine how much more exciting (and useful) would be the economics discipline if we could have models like this. How much easier it would be to design the policy: it would have a clear aim of keeping a certain variable within the defined bands. Instead at the moment economic policy making looks like asking questions: what type shock should we expect? when is it likely for this chock to occur? how can we resist the shock of certain type? how marge of a shock can we actually survive? And, sadly, we usually never have answers to firs two questions out of these four.
Guest - Giorgi on Monday, 05 March 2012 21:39

In short, you want to present an economic system in a way similar to thermodynamic one? I can't think of another example from physics at the moment that would fit your description (and I'm not sure of thermodynamics either :)), please tell me if you have anything specific in mind. But if we think of thermodynamics, self-organization there is an aberration with an extremely low probability, right? And the triggers there actually, if I still remember physics' stuff correctly, cause the complete disruption of the system - not a crisis magnitude that would be useful for us. I've heard of this Axel guy (from you, most likely), but haven't actually looked into his research, could you send/link me whatever you think would be interesting to me?

As for the models, answering all the four questions - I think that human factor is too high in economics and it can only be captured by shocks.

In short, you want to present an economic system in a way similar to thermodynamic one? I can't think of another example from physics at the moment that would fit your description (and I'm not sure of thermodynamics either :)), please tell me if you have anything specific in mind. But if we think of thermodynamics, self-organization there is an aberration with an extremely low probability, right? And the triggers there actually, if I still remember physics' stuff correctly, cause the complete disruption of the system - not a crisis magnitude that would be useful for us. I've heard of this Axel guy (from you, most likely), but haven't actually looked into his research, could you send/link me whatever you think would be interesting to me? As for the models, answering all the four questions - I think that human factor is too high in economics and it can only be captured by shocks.
Guest - Zak on Monday, 05 March 2012 22:46

Thermodynamics is one option. There are many more. Think of percolation models, where slight changes in density have crucial consequences in aggregate outcomes. The same is true for network models. Think of work that Thomas Schelling got the nobel prize for. If crucial parameters in these models could become endogenous variables (and I believe they can) we would have something valuable at hand.

I have written a bit more extensively on these issues here.

Leijonhufvud is just one of the guys advocating this approach. Most of his arguments should come through by reading this. It might well have been the "EconBlogs" that you have heard from about Leijonhufvud, rather than me: there was a discussion past october/november that Nobel should have gone to him, rather than to Sargent! Anyway, there are more guys in the camp: another notable one is Peter Howitt of Brown. The most notable european one is Giovanni Dosi.

As about the human factor: it is exactly its "highness" that pushes us to distance ourselves from non-human/extra-model forces in forms of shocks and embrace self-generated crises.

Thermodynamics is one option. There are many more. Think of percolation models, where slight changes in density have crucial consequences in aggregate outcomes. The same is true for network models. Think of work that Thomas Schelling got the nobel prize for. If crucial parameters in these models could become endogenous variables (and I believe they can) we would have something valuable at hand. I have written a bit more extensively on these issues here. Leijonhufvud is just one of the guys advocating this approach. Most of his arguments should come through by reading this. It might well have been the "EconBlogs" that you have heard from about Leijonhufvud, rather than me: there was a discussion past october/november that Nobel should have gone to him, rather than to Sargent! Anyway, there are more guys in the camp: another notable one is Peter Howitt of Brown. The most notable european one is Giovanni Dosi. As about the human factor: it is exactly its "highness" that pushes us to distance ourselves from non-human/extra-model forces in forms of shocks and embrace self-generated crises.
Guest - Zak on Monday, 05 March 2012 22:49

I do not know why it did not work, but in the previous post it should be:

...by reading this.

I do not know why it did not work, but in the previous post it should be: ...by reading this.
Guest - Eric on Monday, 05 March 2012 23:27

Do we have a macro model that incorporates financial pyramids (that eventually exhaust themselves), stock market casinos, excessive leveraging, and other kinds of crisis generating behaviors?

Do we have a macro model that incorporates financial pyramids (that eventually exhaust themselves), stock market casinos, excessive leveraging, and other kinds of crisis generating behaviors?
Guest - Giorgi on Tuesday, 06 March 2012 14:18

Basic macro models actually prevent Ponzi schemes from being created, by imposing a transversality condition, which basically states that you can't borrow more than the present value of your lifetime earnings stream. You can introduce excessive leveraging, as far as i know.

Basic macro models actually prevent Ponzi schemes from being created, by imposing a transversality condition, which basically states that you can't borrow more than the present value of your lifetime earnings stream. You can introduce excessive leveraging, as far as i know.
Guest - Michael on Tuesday, 06 March 2012 14:54

There are sunspot (self-fullfilling prophecies) models in macro, so yes such models do exist, and they have arrived in the mainstream.

There are sunspot (self-fullfilling prophecies) models in macro, so yes such models do exist, and they have arrived in the mainstream.
Guest - Michael on Tuesday, 06 March 2012 14:58

Although stock market casinos? That makes me *shudder*.

Although stock market casinos? That makes me *shudder*.
Guest - Eric on Tuesday, 06 March 2012 16:26

Don't shudder, Michael. Stock markets are worse than casinos (or better, depending on which side of the "table" you sit).

I personally participated (only once, in Israel, many years ago, so won't be prosecuted) in a well-orchestrated effort to drive up the price of an option by 1000%!!! I was told by an uncle (who was close to the epicenter of the conspiracy) to buy at 50 cents and sell at 500 cents, so there was no risk for me. (Unfortunately, I was a poor student at the time and did not have much money to spare/invest). People who were late to join this little pyramid have lost big time. This type of criminal behavior is actually rational (exploiting other people's greed and naivete), but has nothing to do with market valuation of a firm.

In my case, the crime was perpetrated by a few people in a tiny market. Big institutional investors or the Goldman Sachs of the world can and do the same on a much bigger scale, generating crisis after crisis, at fairly regular intervals. These crises and the intervals at which they occur could be modeled. For instance, they must be governed by the people's ability to forget (which depends on the magnitude and impact of the latest crisis) and availability of other investment or saving opportunities (for instance, interest on bank deposits, government bonds, rate of return on investment in own business or education).

How useful are the currently available models, with or without sunspots, in modeling these types of rational behaviors, and the crises they generate?

Don't shudder, Michael. Stock markets are worse than casinos (or better, depending on which side of the "table" you sit). I personally participated (only once, in Israel, many years ago, so won't be prosecuted) in a well-orchestrated effort to drive up the price of an option by 1000%!!! I was told by an uncle (who was close to the epicenter of the conspiracy) to buy at 50 cents and sell at 500 cents, so there was no risk for me. (Unfortunately, I was a poor student at the time and did not have much money to spare/invest). People who were late to join this little pyramid have lost big time. This type of criminal behavior is actually rational (exploiting other people's greed and naivete), but has nothing to do with market valuation of a firm. In my case, the crime was perpetrated by a few people in a tiny market. Big institutional investors or the Goldman Sachs of the world can and do the same on a much bigger scale, generating crisis after crisis, at fairly regular intervals. These crises and the intervals at which they occur could be modeled. For instance, they must be governed by the people's ability to forget (which depends on the magnitude and impact of the latest crisis) and availability of other investment or saving opportunities (for instance, interest on bank deposits, government bonds, rate of return on investment in own business or education). How useful are the currently available models, with or without sunspots, in modeling these types of rational behaviors, and the crises they generate?
Guest - Zak on Tuesday, 06 March 2012 15:00

Indeed, sunspot models are discussed even in grad macro texts, such as Blanchard and Fisher for example.

Indeed, sunspot models are discussed even in grad macro texts, such as Blanchard and Fisher for example.
Guest - Zak on Monday, 05 March 2012 21:00

So, in conclusion, many models of complexity that are well elaborated in physics should be interesting object of study for economists.

That is one of the reasons I was excited about the first blog-post on the list here -- networks: classical complexity stuff! Change small thing in their composition and you get totally different behavior of the system.

So, in conclusion, many models of complexity that are well elaborated in physics should be interesting object of study for economists. That is one of the reasons I was excited about the first blog-post on the list here -- networks: classical complexity stuff! Change small thing in their composition and you get totally different behavior of the system.
Guest - Michael on Tuesday, 06 March 2012 14:57

They are, although to be honest I am very sceptical about the field of econophysics, and its contributions so far.

They are, although to be honest I am very sceptical about the field of econophysics, and its contributions so far.
Guest - Zak on Tuesday, 06 March 2012 15:02

As are majority of mainstream people :) "We" are still probing grounds. I believer there is a potential.

As are majority of mainstream people :) "We" are still probing grounds. I believer there is a potential.
Guest - Giorgi on Tuesday, 06 March 2012 15:45

The idea comes from as early as Georgescu-Roegen, but still hasn't fulfilled it's potential. I'm not sure, whether it's possible at all

The idea comes from as early as Georgescu-Roegen, but still hasn't fulfilled it's potential. I'm not sure, whether it's possible at all
Already Registered? Login Here
Register
Guest
Friday, 03 July 2020

Captcha Image

Our Partners