The Undercover Economist Strikes Back: How to Run or Ruin an Economy




There's a community radio station, 3CR, way out in 'Stralia which hosts Renegade Economist. The host, Karl Fitzgerald, proposes startlingly refreshing ideas about economic regimens. Here in North America and in Western Europe, for just about any economic discussion, the only, ahem, 'ideas' that are presented are the increasingly obviously failed Chicago School's. And that's is a dry well.

The RE podcasts are short - only half an hour - and very much worth a listen. The next time you have to argue with a Hayekian, you can propose replacing income tax with land value tax.


It's very difficult to make any statement about the failure of any school of economics since most research has no empirically verifiable predictions. What we are seeing are not failures of the economic profession. We are just observing pundits with PhDs. There is nothing inherently wrong with the Chicago school. They are ideas that exist in a vacuum with hardly any confrontation with any data at all.

And let's not forget that there is no way at all to summarise the complexity of a single economy with a handful of headline variables such as inflation, unemployment and so on.


For a subject that claims to be about reality, not making predictions you can look at is a complete failure. But I think we are well past the point where the Chicago school has indicated certain trends should happen, and they don't.


I find I often get annoyed when the economic status of any organization is described in terms of 'growth'. For national purposes, I fail to see why the economy needs to grow beyond the rate of population increase and any inflation (or decline) in the value of the currency that was counted? If that is the base measure, and performance is seen as deviation in either direction, then it's entirely simple - we're generally doing better or worse as a gross measure. (This measure says nothing about distribution, true.) But any other measure seems focused on profits and confuses 'growth' with 'profits', which is a pretty crappy way to decide how we're ALL doing in general. Corporate profits don't equal jobs, or the health of small businesses, or whether people are eating regularly - they only equal wealth delivered to a very small group of people - all of whom are already well off enough to have interests in corporations anyway.

Sorry if I'm simplistic. I'm definitely NOT an economist. I'm a systems and data brain. And that's why economic forecasts and reports drive me a wee little bit nuts. It always seems as if they have no real context beyond the stock market and money market.


Forgive me, but isn't this a contradiction? Unless you are using a very strict definition of "inherent," creating an economic theory without any data seems to be something very wrong indeed.


There is a lot wrong with the discipline of economics. Let me be more precise. Most economists with big opinions are theorists. In economics, theories do not start by observing facts and then constructing a model that generates new insights. Models start with the personal and political biases of the researcher. Every theorist of every school of thought works like this. The major problem with economic theory is that economic mathematical models depend on parameters that are not directly measurable through experimentation of from the data. (Physicists on the other have can pin down the parameters of their models and discard models that fail to explain other data).
That is why Keynesian economists can work with basically the same models as a Chicago school economist and come up with completely contradictory conclusions. Change the primitive parameters, you get different results. Ariel Rubinstein has written about this in his paper "Dilemmas of an Economic Theorist".

It gets worse. Because theorists know that their models can be manipulated based on parameter changes, the next step is to ask: why even bother writing a model that is realistic. Make as many simplifying assumptions so that your paper can get published quickly and does not require any heavy lifting. Models that attempt to understand complex dynamic relationships are turned into models with two time periods (today and tomorrow). They are so simple there is no way to map the underlying structure to data.

The Chicago school (and by that I mean people who have worked on rational expectations type models) has provided us with a direction that allows us to map data into models and to actually test models rigorously. These researchers are known as structural econometricians. The beauty of this research agenda is that it is probably the most apolitical branch of economics. No one talks about this research agenda because, it does not involve sumo wrestlers or other sexy freakonomicsy things and does not fit the pop economists message. Bringing data and economics is messy, difficult and not conducive to a "Naked Economics" type analysis. Rant done.

By the way, I am an economist.


Is there any field involving humans that is amenable to physics-like analysis? There's simply too many degrees of freedom in human behavior to assume that economic analysis is going to be more than "generally correct".

As for blaming economists for bad predictions, I do have some sympathy. Most organizations that face the public are not interested in an economist whose predictions come with error bars, which all predictions involving humans do. So, an economist who actually wants to keep their job has to, in effect, lie.

Or as one mentor told me a long time ago - they'll forgive you for lying and then getting caught out, but they'll never forgive you for telling the truth straight out. (This was about trying to inform management about the uncertainty in software project deliverables, but it was the same idea.)


Well, one problem is that any company or nation that grows significantly more slowly than its competitors will be destroyed, either economically or politically. For companies, it's pretty obvious how you die - you simply become insignificant compared to the competition. For communities and countries, you grow poorer and poorer relative to your higher growth counter-parts.

For communities, it means the industrious and ambitious leave, and you stagnate as your "middle-class" becomes "abject poverty" in relative terms, along with the negative social consequences that brings.

For countries, it's a little more complex. The US has not destroyed all the other countries in the world, but the simple fact that their economic growth has dwarfed every other country has essentially forced the entire world to change their systems and culture to resemble the US system to one degree or another. Those governments that attempted to prevent that change have often been overthrown by their own people. A people that, I will add, often did not want to change their culture, but found the idea living with less growth too awful to contemplate.

If the US chose to accept a low-growth strategy in the long term, it would simply be replaced by another country that surpassed it, and once the gap grew large enough, the US would be forced to adopt whatever strategy it had chosen to eschew.

That's the thing about a globalized world. No one gets to opt out, and being a comparative species, rich countries conquer other cultures simply by existing and making inhabitants of those cultures feel poor by comparison.


I completely agree. I often scream at my MP3 player, whenever I'm listening to economics podcasts, to an economist making some assertion or other: "Where are the freakin' paired studies, eh?"

The Hayekians and the Chicago School sure don't seem to have a problem 'summarizing the complexity of a single (or even the entire universe's) economy' by ignoring history or by what works in other locales. They need to be called out. Regularly, and often. In other places, they use different rules and these work (too?).


The Keynesians also have no problem doing that.

There is another issue with data analysis in economics that someone from the Chicago school pointed out. When dealing with policy shifts there is no value in using historical correlations. Those measurements are from one equilibrium, change the policy environment, people adapt and your policy can be rendered ineffective. This is known as the Lucas critique ( This is what has pushed a number of empirical economists to focus data analysis on estimating primitive parameters from observed choices and to then use those to simulate alternative regimes. It is also the reason why macroeconomics has not pushed the empirics as much as it should have.


Forgive me if I'm wrong but wasn't it proponents of the Chicago School that got us into this mess?
It strikes me that economics is not much help to actual policy decisions. The goal of governments should be keeping their citizens happy and safe. The goal of companies should be delivering goods and services to people willing to trade money for them. But it seems that the ways of measuring these results have been abstracted to the point of meaninglessness.
(By the way, I'm a mathematician. I get your idea about theory vs. real life. I wouldn't use the theory of anti-comutative geometry that I'm studying right now to explain real life phenomena other than at the quantum level, and then only maybe.)


Thanks for the reply. I appreciate!

But, you used that word 'grow' as a generic term again. My problem is that it doesn't actually mean anything in a generic sense, although, from the way you followed up, I assume you meant GNP. The very idea of GNP as the ultimate measure of anything assumes that all societies are competitive by nature. Example: I live on my island. My island produces all my basic needs, therefore, I could really care less whether your island produces thing-a-ma-jiggers or not, much less how you price them or whether Joe Capitalist's island makes more or less than you. That is to say, you still used capitalism , per se, as the basis for your whole model, and gross measures such as nations as units. That kind of negates the process having any use to individuals at all, since they are relegated to the position of unnamed points in your data collection.

I started life as a manufacturing systems analyst, back when almost none of those systems had been automated. (These days, you can get a degree in the subject area, but there was no such thing then.) What we later saw, was that you can define and automate any actual processes you like - because they are physical circumstances that produce physical things, and can therefore be measured and modeled. What you CAN'T model, is production forecasting. Oh, you can build systems to automate the calcs and check capabilities and capacities and such - but you cannot, no matter what you do, make it forecast accurately. Because that, by its very nature, is guesswork. Some forecasters are better than others, true. But we don't actually know why and they can't explain why, and so it's never been successfully automated. What economic measures try to do is SO much fatter and lives in such a hugely expanded context, that I think of it as the same kind of problem.

I guess, I'm not arguing whether it's useful at all. I'm just having a problem with it holding any actual human relevance...unless you own a nation, in which case, game on?


Ah, I misinterpreted your post. I do agree that "growth" is meaningless in and of itself.

However, it's generally used as a stand-in for the huge number of factors that comprise the economic welfare of the populace. It's a bit like conflating weight and health. They're not the same thing, at it is possible to be overweight and healthy, but given the general correlation, doctors will encourage you to lose weight in the assumption that if you do that, it's probably by making yourself healthier in a variety of other ways.

If you can increase GNP significantly, it probably means your decreasing unemployment, increasing wages, etc.

As for comparing between nations, then my post above has some application. If your nation consistently falls behind others in GNP, it probably means their individual wealth is falling behind other nations, and then your citizens will eventually become less happy because in their own eyes they've become poorer, even if their living standard have improved.


That's true - assuming they think like westerners about such things. It depends on what the population does or does not value, and what distribution is actually like. For example, India - huge country, has some incredibly wealthy individuals and a rapidly growing economy, but people still die of starvation in the streets because distribution doesn't take care of those problems - even when many people are hugely over-employed, by western standards. So, GNP still has no particular value where individuals are concerned.

I find the health vs weight thing fascinating/disgusting, because I am very familiar with what started that. The assumptions are highly questionable to begin with, but we see that, very like economic forecasts, the theories it produced are now being used (and quite purposely, in the health/weight case) to scare the daylights out of the general population on yet another theory - that they will lose weight because they are afraid they will die. So now, we see issues such as huge amounts of additional stress being heaped on people who weight more, and truly crazy stuff, such as a Chicago area school district banning all lunches brought from home - even though some students have particular health needs and allergies the school is not prepared to address.

So, there's a case where such studies and estimates actually bring harm to individuals. I think of that as being very like economic forecasts that create havoc in the stock market. It sounds like everyone is getting more information from which they can make better decisions...but then, there are the individuals who get their accounts creamed on the basis of nothing more than a guessimate. We don't hear from them, so it's easy to assume we can judge their financial well-being by whether the NASDQ rose a few points and hit a new high, even though that figure does not account for volumes of stocks available, nor the amounts of trades happening. It only says each aggregate stock is worth a little more than last year.

Meh - I fully admit it. My problem is that, as much as I do love systems (and I do!), humans are not and cannot be statistics - and really bad stuff usually happens to some of them every time their lives are impacted by such systems. Humans just don't 'count' very neatly.


I'd agree about distribution, but one thing that's become increasingly clear to me as I age is that "values" don't trump economics. It seems almost inevitable that when one's culture falls too far behind, there's a popular push to discard those cultural values, despite paying a dreadful price. Having been part of the dominant culture, I don't think I can appreciate why that price is so high, but the despair, resentment, and anger at the loss of those values is huge.

Yet the wealth of others also seems to render meaningless the sacrifices that those values required, which engenders a whole different sort of despair and resentment. (Few resist trading in a nomadic lifestyle for grey apartment buildings that are consistently dry and heated, but it's still infinitely worse than being cold and wet before one realized there was a choice. At least then you had pride and shared your travails.)

It's an awful thing, but seems an almost universal tendency with only small groups or individuals managing to escape.

Also, one thing we can easily miss is how significant a transformation life from dirt poor to poor is. Huge numbers of Chinese and somewhat fewer Indians are having their life radically changed for the better, although we rarely hear about it, and that is reflected in GNP. The last few decades have seen the biggest alleviation in poverty the work has ever seen.

With the weight/health thing, it's tricky. Everybody wants a single parameter function and once one is found "f(weight) = health" they go completely to town with it regardless of applicability. Likewise with metrics - soon a value that should correlate with objective X is mistaken for X.

I've long since concluded that average human beings aren't built for statistics. Daniel Kahneman's book "Thinking Fast and Slow" simply provided proof for what I've observed over the last 40 years. (I'm still enough of a geek to think statistics can help, but we don't want trade-offs and we don't want error bars, so the list of policies we can actually enact to help people is very limited.)


I do see your point. Although, I tend to think statistics are better for the simple-minded. That's what we see in the news. Whole realms of human realities reduced to statistics, which are then further edited down to some dramatic head-line number. I also agree that statistics can be useful. (I actually own an antique newspaper announcing and explaining the first time the federal government ever hired a statistician - it was done to try to predict imports vs. exports so they could control the budgeting process a little better.) But, other than some specific measure of some interesting thingie? Lies, damned lies, and statistics.

I also agree that it's a lousy way to try to help people. All I can conclude at this point is that the best thing we could probably ever do is to offer opportunities. Some people will take them, and some won't, but hammering the government with numbers and then hammering the people half to death with statistics isn't way. Humans are still animals. Food, clothing, shelter, and access to information so as to see to one's own survival. After that, you can invent or build whatever you want - and motivated people always do, given the chance. Anything more is so completely subject to corruption that...we see its evidence all around us, every single day, don't we?


This topic was automatically closed after 5 days. New replies are no longer allowed.