Is PCT value neutral?

CSGnet Archive CSG2011
21

AM:

How do you mean that economic theories don’t get used for prediction?

For example, if a theory says that a minimum wage law would increase average pay, it gets accepted, and lo and behold - average pay increased. Theory confirmed.

The problems are 1) that the theory fails to explain the mechanism that allowed this or a wrong but plausible-sounding mechanism is provided. Average pay increase might result from things unrelated to minimum wage laws. 2) That there are (might be) non-obvious consequences.

Average pay increase might result from lowering owners share of profit, but also by simply firing people who were payed less than minimum wage and making others work more; or perhaps from lowering reinvestment in business; or by inflating the money supply so it’s just the amount of pay rising, not the buying power; or by a number of other ways. The unseen consequences might be increased unemployment, inflation, and others.

Still, the theory is considered sound since it “obviously resulted in an average pay increase”.

Best,

Adam

···

On Wed, Apr 20, 2011 at 8:35 PM, Bill Powers powers_w@frontier.net wrote:

[From Bill Powers (2011.04.20.10455 MDT)]

BP: Rick, I think you have put your finger on a basic problem with
conventional economic (and other) theories. They are almost always used
to explain what has already happened, not to predict what is going to
happen as a function of current and past variables.

22

[From Bill Powers (2011.04.20.1730 MDT)]

[From Bill Powers (2011.04.20.10455 MDT)]

BP: Rick, I think you have put your finger on a basic problem with
conventional economic (and other) theories. They are almost always used
to explain what has already happened, not to predict what is going to
happen as a function of current and past variables.

AM: How do you mean that economic theories don’t get used for
prediction?

For example, if a theory says that a minimum wage law would increase
average pay, it gets accepted, and lo and behold - average pay increased.
Theory confirmed.

BP: Really? How did they compute that outcome? Actually, it’s a
tautology, isn’t it? If you increase the lowest pay, how can the average
pay do anything but increase? And how closely did the observed increase
match the predicted increase? A theory can get the sign of change right
and still be useless for predicting anything important. What if the
average pay went up 1 cent per day? That’s an increase, isn’t
it?

The problems are 1) that the
theory fails to explain the mechanism that allowed this or a wrong but
plausible-sounding mechanism is provided.

If no mechanism is described, how can you call this a “theory?”
If the mechanism isn’t verified, how can you believe any predictions even
if the guess is right?

Average pay increase might
result from things unrelated to minimum wage laws. 2) That there are
(might be) non-obvious consequences.

Average pay increase might result from lowering owners share of profit,
but also by simply firing people who were payed less than minimum wage
and making others work more; or perhaps from lowering reinvestment in
business; or by inflating the money supply so it’s just the amount of pay
rising, not the buying power; or by a number of other ways. The unseen
consequences might be increased unemployment, inflation, and
others.Still, the theory is considered sound since it “obviously
resulted in an average pay increase”.

From your discussion I wouldn’t call it much of a theory at all if none
of these obvious possibilities was taken into account. The point isn’t
just to guess right. It’s to show that if the prediction isn’t right,
something is wrong with some really basic premises. A model isn’t just
made out of thin air. Each of its premises has to have some reasonably
sound basis, unless it’s labeled clearly as a premise being tested. The
whole model is always being tested, not just the behavior of one or
another variable. Try doing the tracking experiment with your eyes
closed, and see if vision is really required.

···

At 12:32 AM 4/21/2011 +0200, Adam, Matic wrote:

On Wed, Apr 20, 2011 at 8:35 PM, > Bill Powers > powers_w@frontier.net > wrote:

Best,

Bill P.

23

AM:

I don’t mean to say that those theories are any good, but that people use them for prediction**. As you surely know, there are whole university-level textbooks calling statements like that “scientific theories”, in economics as well as psychology. And as you said, a theory like that would never even get published in a field such as physics, but in some fields it might be considered a theory and presented with statistical evidence as a scientific model.

**Perhaps ‘betting’ more accurately describes the procedure than ‘prediction’. The fact that it’s called prediction and that there are more times when it comes out right than times when it comes out wrong gives people the impression they’re doing science.

OK, thanks for the clarification, I understand you meant accurate prediction of change in a variable and not guessing or betting.

Best

Adam

···

On Thu, Apr 21, 2011 at 1:52 AM, Bill Powers powers_w@frontier.net wrote:

[From Bill Powers (2011.04.20.1730 MDT)]

BP: Really? How did they compute that outcome? Actually, it’s a
tautology, isn’t it? If you increase the lowest pay, how can the average
pay do anything but increase? And how closely did the observed increase
match the predicted increase? A theory can get the sign of change right
and still be useless for predicting anything important. What if the
average pay went up 1 cent per day? That’s an increase, isn’t
it?

If no mechanism is described, how can you call this a “theory?”
If the mechanism isn’t verified, how can you believe any predictions even
if the guess is right?

24

[From Bill Powers (2011.04.20.1900 MDT)]

···

At 02:36 AM 4/21/2011 +0200, Adam Matic wrote:

AM: OK, thanks for the clarification, I understand you meant accurate prediction of change in a variable and not guessing or betting.

Ah, a man of peace. Nice to have you here.

Best,

Bill P.

25

AM:

Thank you.

It’s great to be here and talk to people whose work and

intellect I admire so much. A man rarely has such an opportunity.

Best

Adam

···

On Thu, Apr 21, 2011 at 3:04 AM, Bill Powers powers_w@frontier.net wrote:

[From Bill Powers (2011.04.20.1900 MDT)]

Ah, a man of peace. Nice to have you here.