RoR: Forum

Thanks, again, Joseph. As usual, my 6 are among the 30 Atlas Points.

We believe that our monetary media allow us to measure economic events to two decimal places, at least. As you note - and as an axiom to the Austrian economists - qualitative predictions may be the best we can do. In fact, they are not deficient for not being quantitative. Qualitative predictions are sufficient.

Passive description can be good science – and good description is not passive. We bring our expectations to our observations. A scientist knows to be ware of preconceived notions and to be open to new perceptions. At the same time, the scientific observer often has good reason to seek exactly the phenomenon under investigation. It is not really passive. If only implicitly, certain factors are held constant, while others change. Aristotle’s description of the chick embryo is a paradigmatic example cited as one of the greatest scientific experiments in history.

Since the Renaissance, the term experiment has been used in diverse ways to describe a variety of procedures such as a trial, a diagnosis, or a dissection …
"What Counts as an Experiment?: A Transdisciplinary Analysis of Textbooks, 1930-1970," Andrew S. Winston and Daniel J. Blais. The American Journal of Psychology, Vol. 109, No. 4 (Winter, 1996), pp. 599-616.

Post 1

Thursday, March 7, 2013 - 4:32am

Nice article.

Qualitative predictions usually have a sign associated with them ("Better/worse...increase/decrease....love/hate")

The 'quality' of something improves or not, though it is still often left to political argument what represents an improvement.

Quantitative predictions have both a sign and magnitude.

Quantitative predictions have an associated uncertainty expressed in magnitude.

Qualitative predictions have an associated uncertainty expressed as percentage confidence of getting the sign right, and the discipline of uncertainty analysis is applicable to each. By applying a sign to a quantitative analysis, it has already implicitly included a qualitative analysis.

Both are calibratable in many though not all instances: "In a thousand applications of this predictive qualitative hypothesis, we got the sign right (23%/78%) of the time, statistically (worse/better) than pure chance." For some domains, it is possible to run a thousand experiments. For others, like the economy, a million such experiments are often being run in parallel, muddying up each others ceterus paribus.

Economics is an example of something that in practice is largely qualitative though not always in nature but seldom calibrated; when an economic theory is applied on a broad scale, as in "a few trillion in borrowed/printed stimulus will prime the pump of something called 'the' economy, coupled with low interest rates, moving unemployment below 6%", and both the qualitative and quantitative predictions are wrong, there is only additional political wrangling to explain away the bad experiment, and the science does not move forward.

When my son attended UVA, he took a course titled "Calibrative Economics" or some such. There are some who recognize this deficiency in the field.

regards,
Fred

Post 2

Thursday, March 7, 2013 - 4:48am

Michael;

"We believe that our monetary media allow us to measure economic events to two decimal places, at least."

I believe that our monetary theories are grossly incomplete, based largely on politics, and have yet to demonstrate a pronounced ability to get the sign right beyond chance.

This defect in the science goes all the way back to Economic Marxism 101, and the simple assertion

PRICE = COST + PROFIT

vs

PROFIT = PRICE - COST

Mathematically equivalent, but political opposites. Which terms in the above simple equations are independent, and which are dependent?

Marxist economics emphasizes that PROFIT is largely independent, COSTs are impressed, and PRICE is dependent.

Libertarian economics emphasizes that PRICE is largely independent, COSTS are controllable, and PROFIT is dependent.

(And economic argument uses weasel words like 'emphasizes' and 'largely.')

And that political wrangling via math goes all the way through Paul Krugman and his "topsy turvy economics...paradox of toil...liquidity trap...nonsense."

Imagine we were a pure endowment economy indeed...and then come up with a model of aggregate supply and demand both with positive slopes instead of opposite signed slopes.

Topsey-turvey indeed. Fortunately, uncalibrated and without uncertainty analysis because none is required for Krugman's political purposes.

regards,
Fred

Post 3

Thursday, March 7, 2013 - 6:17am

Right, Fred. My intention was only that our monetary media (dollar and cents) allow us the mistaken belief that ...

Economics is pretty easy for us here to understand. Joseph's claims apply to all sciences, really. I do not want to divert this thread based on nuance, but it seemed to me that he uncritically accepted that quantitative predictions are "better" (more "scientific") than qualitative. Indeed, science is taught that way, that the scientific method demands quantitative predictions that are tested to our epistemological limits of tolerance and confidence.

However, as with Aristotle and the Chick (or economics), such quantities may not exist - and as in economics - pursuing them may be misleading or even destructive.

Certainly, quantitative predictions are important. Take the mundane melting of table sugar. From simple syrup through creme broule to carmel fudge, the amount of heat makes a difference in the outcome. Plastics, metals, ... we would live in a world of useless trashes from brittles to gooeys if we could not control the quantities of heat, pressure, mass, volume, time...

As you noted, with love, you either have it or you do not. We do not go from 45.3% to 72.1% (+/- 5%) on Valentine's Day.

Interestingly, that most quantitative of sciences, computing, actually rests on quality: the bit is either on or off.

(Edited by Michael E. Marotta on 3/07, 6:20am)

Post 4

Thursday, March 7, 2013 - 8:41pm

Michael:

Interestingly, that most quantitative of sciences, computing, actually rests on quality: the bit is either on or off.

Yes, and at one level of detail, digital is really analog...but if we look deeper still, eventually the universe gets discrete yet again, with only finite states found.

Stuff is weird. Molecules are mostly space, and atoms. And then, atoms are mostly space, and then protons, neutrons, electrons. And then each of them is mostly space...

The whole thing is a facade, mostly space, like that fake village in "Blazing Saddles" which was ... mostly space.

And still, as you well know, if you drop Wolfram's book on your toes, they are going to hurt like a bitch, mostly space or not.

I didn't detect much bias in Joe's examination of quantitative vs. qualitative; it didn't come across to me as a better/worse comparison. I think the hard sciences look purely quantitative the farther away one is from them.

There are some domains where uncertainty is so high and information so incomplete that it would be redundant to every day declare "uncertainty far exceeds relevance." People need to plow ahead anyway; sometimes progress is made just by showing up, you know?

That isn't really my long running gripe with the field of economics; my gripe is based on the blatant politicization of economics, with Paul Krugman being one of many poster children for that.

A pure politician, to me, is one who is driven solely by what he wants from others; all other considerations fall away; economics becomes a voodoo priest's robe.

To me, it is clearly a qualitative spectrum:

Asking...trading.....begging .... politics .... crime .... war.

A sliding spectrum of how we get what we want from our peers on this planet.

Life in the tribe as peers is easier and more peaceful when tribe members interact at the 'asking' end of the spectrum...unless one is driven purely by what they want from others. Because after all, when asked, someone might say 'no.'

I don't confuse politicians with government staff and public employees. They're just folks with jobs, and not all of them paid. (My wife the school board president comes to mind; I can't believe that in 500 school districts, PA convinces 4500 people to serve on those boards for free...)

No, a politician is somewhere between beggar and criminal.

regards,
Fred

Post 5

Tuesday, March 12, 2013 - 4:36am

When my son attended UVA, he took a course titled "Calibrative Economics" or some such. There are some who recognize this deficiency in the field.

Correction: it was "Experimental Economics" and calibration was a part of the course.

Post 6

Saturday, April 13, 2013 - 11:41pm