Economic anomaly?

[From Bruce Gregory (971126.1325 EST)]

Bruce Abbott (971126.1245 EST)

Samuelson said:

  Which will be more powerful, the substitution effect or the income effect?
  Or will they just balance each other and cancel each other out--so that
  the supply curve neither rises forward nor bends backward, but rises
  perfectly vertically and inelastically? There is no single correct
  answer. It depends upon the individual and the situation faced.

Samuelson gets an A.

You said:

Apparently the African workers of the earlier example were operating in the
upper, negatively sloped region of the curve.

You get an F :wink:

Bruce

[From Bruce Abbott (971126.1245 EST)]

Bill Powers (971125.0706 MST) --

Even if we manipulated one of the parameters, wages, I think it would come
as a shock to many people to find that effort is actually a _declining_
function of wage rates, and that received reward hardly changes with wage
rates.

(Of course that is not a very realistic example, in a society where most
people work fixed hours for a fixed wage and can't increase or decrease
their earnings so easily. I suspect that the economic structure has evolved
so that workers have lost this ability to control their income by varying
their efforts. The only reference to this possibility I ever saw in an
economics text was in an early edition of Samuelson, in a footnote where he
described this effect among African workers building a road. When their
wage rate was doubled in a desperate attempt to get them to build the road
faster, they started working half as many hours, keeping their earnings at
a level that evidently satisfied them. Samuelson noted this, with scholarly
reponsibility, as an "anomaly."

It would appear that Samuelson found a way to account for his anomaly, as
the following quotation shows:

  In analyzing the labor supply, one of the most important issues is how
  labor responds to higher wages. What will be the effect of higher wages
  on the number of hours worked per lifetime? Figure 28-2 illustrates the
  issues; it shows how the supply curve of total hours that a group of
  people will want to work at each wage rate. Note how the supply curve
  rises at first in a northeasterly direction; then at the critical point C,
  it begins to bend back in a northwesterly direction. How may we explain
  why higher wages may either increase or decrease the quantity of labor
  supplied?

        >
        > X
        > X
        > X
        > X
        > X
  WAGE | X C
  RATE | X

···

                 X

        > X
        > X
        > X
        +---------------------------
             QUANTITY OF LABOR
        Figure 28-2.

  Put yourself in the shoes of a worker who has just been offered higher
  hourly rates and is free to choose the number of hours worked. You are
  torn two different ways: on the one hand, you are tempted to work some
  extra hours because now each hour of work is better paid. Each hour of
  leisure has become more expensive--hence you are tempted to substitute
  extra work for leisure.

  But acting against this "substitution effect" is an opposing "income
  effect." With the wage higher, you are, in effect, richer. Being
  richer, you can buy more clothes, better food, more of other consumer
  goods. But, most important for the present problem, you'll also want
  more leisure time! Now you can afford to take Saturday off, have a
  week's vacation in the winter or an extra week in the summer, or take
  your retirement a couple of years earlier.

  Which will be more powerful, the substitution effect or the income effect?
  Or will they just balance each other and cancel each other out--so that
  the supply curve neither rises forward nor bends backward, but rises
  perfectly vertically and inelastically? There is no single correct
  answer. It depends upon the individual and the situation faced. In the
  case shown in Figure 28-2, for all wage rates up to point C, labor supplied
  increases with a higher wage; the substitution effect outweighs the
  income effect. But from point C upward, the income effect outweighs the
  substitution effect and labor supplied declines with higher wages.
    Samuelson, P. A., & Nordhous, W. D. (1985). _Economics_ (12th Ed).
      New York: McGraw-Hill, Pp. 616-617.

Apparently the African workers of the earlier example were operating in the
upper, negatively sloped region of the curve.

Regards,

Bruce

[From Bruce Gregory (9711261610 EST)]

Bruce Abbott (971126.1600 EST)]

> Bruce Gregory (971126.1325 EST) re: Bruce Abbott (971126.1245 EST)

>You get an F :wink:

I don't think so, but you've left it just a tad unclear why you hold this
opinion. Assertions aren't worth much in my book unless they come with support.

For your part, you still have an incomplete for failing to respond to the
issue I raised for you concerning incentives. I think you're stonewalling. (;->

I'm a little puzzled. I sent two responses. The first was my
usual pithy self. The second slightly more detailed. If you did
not receive them, let me know and I'll try again.

Bruc

[From Bruce Abbott (971126.1600 EST)]

Bruce Gregory (971126.1325 EST) re: Bruce Abbott (971126.1245 EST)

You get an F :wink:

I don't think so, but you've left it just a tad unclear why you hold this
opinion. Assertions aren't worth much in my book unless they come with support.

For your part, you still have an incomplete for failing to respond to the
issue I raised for you concerning incentives. I think you're stonewalling. (;->

Regards,

Bruce

[From Bruce Gregory (97111720 EST)]

Bruce Abbott (971126.1700 EST)

>Bruce Gregory (9711261610 EST) --

>>Bruce Abbott (971126.1600 EST)

>> For your part, you still have an incomplete for failing to respond to the
>> issue I raised for you concerning incentives. I think you're
stonewalling. >> (;->

>I'm a little puzzled. I sent two responses. The first was my
>usual pithy self. The second slightly more detailed. If you did
>not receive them, let me know and I'll try again.

Yes, I did receive them, but found my question unanswered. Instead, you
asked a series of leading questions back. I appreciate your attempt at the
Socratic method and would be glad to answer presently, but I asked first and
would like the favor of an answer before we change the subject and deal with
your questions. My question was, if incentives don't work, why were all
those miserable people puting themselves through hell for a Corvette?

[Aside: I know what you think I think, but what you think I think is not
what I think, even if you do think so. (;-> I'll explain after I've finally
gotten a straight answer from you to my simple question, which you have thus
far evaded.]

I'm still puzzled. Here is what I responded to your earlier
request:

Some people respond to some incentives some of the time. Depends on what their
intentions are, whether something is or is not, at this moment, an incentive,
a counter incentive, or something to be ignored.

If this is _not_ an answer to the question, I have no idea what
is. If I want a car badly enough to go through the gyrations
then offering to provide me with one is an incentive that will
work. If I don't, your incentive will not work. Do your students
have difficulty figuring out what you are looking for? Or have
you trained them better than you have trained me?

Bruce

[From Bruce Abbott (971126.1700 EST)]

Bruce Gregory (9711261610 EST) --

Bruce Abbott (971126.1600 EST)

For your part, you still have an incomplete for failing to respond to the
issue I raised for you concerning incentives. I think you're

stonewalling. >> (;->

I'm a little puzzled. I sent two responses. The first was my
usual pithy self. The second slightly more detailed. If you did
not receive them, let me know and I'll try again.

Yes, I did receive them, but found my question unanswered. Instead, you
asked a series of leading questions back. I appreciate your attempt at the
Socratic method and would be glad to answer presently, but I asked first and
would like the favor of an answer before we change the subject and deal with
your questions. My question was, if incentives don't work, why were all
those miserable people puting themselves through hell for a Corvette?

[Aside: I know what you think I think, but what you think I think is not
what I think, even if you do think so. (;-> I'll explain after I've finally
gotten a straight answer from you to my simple question, which you have thus
far evaded.]

Regards,

Bruce

[From Bruce Abbott (971127.0950 EST)]

Bruce Gregory (97111720 EST) --

I'm still puzzled. Here is what I responded to your earlier
request:

Some people respond to some incentives some of the time. Depends on what their
intentions are, whether something is or is not, at this moment, an incentive,
a counter incentive, or something to be ignored.

If this is _not_ an answer to the question, I have no idea what
is. If I want a car badly enough to go through the gyrations
then offering to provide me with one is an incentive that will
work. If I don't, your incentive will not work. Do your students
have difficulty figuring out what you are looking for? Or have
you trained them better than you have trained me?

Not usually, and no. I'm sorry if I overlooked what I was looking for in
your reply. Let's review. I gathered from your first post on economics
that you do not believe that incentives "work." So I provided an example in
which a specific incentive seemed to be working very well, thank you, for a
specific group of individuals, and asked you to explain their behavior, if
incentives don't work. Your reply was to make some remark about how "some
of us" are more interested in the cases in which incentives do not work. I
replied that this did not answer the question, and you responded by noting
that whether an incentive works will depend on a number of factors,
including whether it is an incentive. I guess I was looking for a clear
statement that this particular incentive (the Corvette) did indeed work, for
those who participated in the contest. Then the question becomes why. Of
course, we can't really answer that for any particular contestant without
first learning quite a bit more about that individual, and the reasons may
differ from case to case. But we could make some general statements about
the kinds of circumstances that would lead to the Corvette functioning as an
incentive for an individual.

For your part, you have been assuming a fact not in evidence about my
position on the matter. You would have me taking the position that
incentives attract people as a powerful magnet attracts an iron bar,
irresistibly pulled by a force outside themselves. Is this your view of my
position, of do I have you wrong?

Regards,

Bruce

[From Hank Folson 971127]

(Bruce Gregory (97111720 EST)) (sic)

Some people respond to some incentives some of the time. Depends on >>>what

their intentions are, whether something is or is not, at this >>>moment, an
incentive, a counter incentive, or something to be ignored.
...

If I want a car badly enough to go through the gyrations,
then offering to provide me with one is an incentive that will
work. If I don't, your incentive will not work...

I would say it differently, because I believe there is no incentive involved.

I agree with all your earlier comments, and with the idea that the desire to own
a Corvette _already_ exists in this individual. Normally the system output
(behavior) that will put a Corvette in this person's garage, and thus reduce his
error signals for Corvette ownership to zero, is to buy one. If he doesn't have
the money, the means to reduce his error is unavailable. His system simply
cycles through the Corvette-lust loop with the error signal stuck at some
painful level because there is no output possible. Awareness of the promotion
changes the next circuit around the Corvette-lust loop. Now output to reduce the
error is possible: Enter contest -> win Corvette -> bliss of zero error signal.
He enters the contest, unless some internal conflict with other goals
interferes.

There is no incentive at all. All that has happened is that he now can produce
an output to reduce his error signal. Incentive is an illusion, at least, it is
to a PCTer.

Now I wonder if in all cases apparent incentive is simply the result of the
recognition of opportunity for potentially error reducing output?

Sincerely, Hank Folson

From Bill Powers (971127.0941 MST)]

Bruce Abbott (971127.0950 EST)--

For your part, you have been assuming a fact not in evidence about my
position on the matter. You would have me taking the position that
incentives attract people as a powerful magnet attracts an iron bar,
irresistibly pulled by a force outside themselves. Is this your view of my
position, of do I have you wrong?

I think that is the view of most of us of the position you've described.
Outside the realms of insider jargon, most of us would assume the
dictionary definition of incentive: "something that incites to action." And
"incite" comes out this way: 'to urge on; stimulate or prompt to action."
These definitions clearly come from a tradition in which it is believed
that things or events outside a person act upon the person to produce
action. Of course the magnet you cite exerts a pull on the person rather
than a push, but the underlying model is the same.

The PCT model would imply quite a different explanation. The person who
offers an "incentive" is offering an opportunity for someone to obtain
something that is desired by using the means now made available. At the
same time, the offerer of the incentive is making a deal with those who
accept the offer: If you give me what I want to get, I will give you what
you want to get. This "deal" doesn't have to be explicitly announced;
whether made explicit or not, each party to the deal gets what is desired
by doing what satisfies the other party's desire. Needless to say, this
"deal" is actually consummated by the offerer only with those people who
want what is offered and are willing to do what is asked. With respect to
all others, the offer falls on deaf ears, and doesn't work.

You will recognize, I hope, that this analysis of incentive constitutes a
scientifically valid proposition, although not cast in any special
scientific language. This analysis is based on interactions between
autonomous control systems, so use of a misleading term like incentive is
to be avoided.

You ask whether this isn't a proof that incentive "works." This depends on
what you mean by "working." If I offer the incentive of $1 to anyone who
will lick my shoes, it is possible that some people will take me up on my
offer and lick my shoes, after which I will have to pay them $1 (or renege
on my offer). But suppose I publicize this offer so that 100,000 people
read and hear about it, and that of that 100,000, 100 show up at the
ceremony where they find me standing in a pigpen, and of that 100, 1
actually licks my shoes and gets $1. You appear to be saying that I can
point to that one person and say "See? Incentives work!"

Coincidentally, Mary came across a newspaper story about a police
department that held a raffle apparently imitating the incident you cite.
An offer was made to all 16 and 17 year olds in the community: they could
pay $20 to participate in the "laying on of hands" process, with the last
one still touching the Bronco to receive it. One 17-year-old girl
registered and showed up; she touched the Bronco and it was hers for $20.
This incentive was accompanied by a relatively modest cost, the $20
registration fee, but that's only a matter of degree -- taking advantage of
an incentive always involves some sort of cost.

When you ask whether incentives work, you have to consider all the data,
not just the positive instances. And to define working, you have to specify
the context. Does an incentive work for the person who fails to win the
contest? Obviously not. Does it work for the person who paid for an
advertising campaign and got one taker? Again, obviously not. Working has
to be judged in terms of the goal of the person who offered the incentive,
as well as on the basis of the number of times it might have worked in
comparison with the number of cases in which it did work. One can hardly
claim that an incentive works when in 999 cases of a thousand (as in the
mail-order industry) it fails to produce any action.

It seems to me that we have had exactly this discussion about the word
"reinforcement."

Best,

Bill P.

[From Bruce Abbott (971127.1400 EST)]

Hank Folson 971127 --

(Bruce Gregory (97111720 EST)) (sic)

Some people respond to some incentives some of the time. Depends on
what their intentions are, whether something is or is not, at this
moment, an incentive, a counter incentive, or something to be ignored.
...
If I want a car badly enough to go through the gyrations,
then offering to provide me with one is an incentive that will
work. If I don't, your incentive will not work...

I would say it differently, because I believe there is no incentive involved.

I agree with all your earlier comments, and with the idea that the desire
to own a Corvette _already_ exists in this individual. Normally the system
output (behavior) that will put a Corvette in this person's garage, and
thus reduce his error signals for Corvette ownership to zero, is to buy
one. If he doesn't have the money, the means to reduce his error is
unavailable. His system simply cycles through the Corvette-lust loop with
the error signal stuck at some painful level because there is no output
possible. Awareness of the promotion changes the next circuit around the
Corvette-lust loop. Now output to reduce the error is possible: Enter
contest -> win Corvette -> bliss of zero error signal. He enters the
contest, unless some internal conflict with other goals interferes.

There is no incentive at all. All that has happened is that he now can
produce an output to reduce his error signal. Incentive is an illusion, at
least, it is to a PCTer.

Now I wonder if in all cases apparent incentive is simply the result of
the recognition of opportunity for potentially error reducing output?

Hank, I agree completely with your analysis (indeed, it is identical with my
own analysis), but not with your conclusion that "there is no incentive at
all."

Our disagreement here, I suspect, is completely one of definition. I would
say that an incentive to a person is something the person wants, which is
offered contingent on the performance of some act. The person will perform
that act only if certain conditions hold, e.g., the act is not too costly,
it does not lead to serious conflict with other goals, and more satisfactory
avenues for obtaining the incentive are not available. (I can envision other
possible constraints as well.) A detailed analysis of that individual case
would allow one to predict whether a given offer would function as an
incentive for that person, under given conditions.

Regards,

Bruce

[Martin Taylor 971127 18:00]

Judging from the dialogue, I'm wondering whether my earlier post on
incentives [Martin Taylor 971124 10:30] ever got out. I repeat it here:

···

--------------

As I understand an "incentive," it is a change in an environmental
feedback function set up by one person and made known to another. What
the other does about it depends entirely on whether the perceptions
affected by actions through the changed function are controlled or
not, and whether the changed environmental feedback function provides
a path that gives better control of those perceptions without also
introducing conflict with respect to other controlled perceptions.

If someone offers me an incentive of a million dollars to kill you, I
would do it if it didn't thereby introduce horrible conflicts with other
perceptions, such as, to name one, self-image as a law-abiding
nice guy. I have a lot of controlled perceptions for which the control
would be a lot easier if I had a million dollars:-)

Martin

[From Hank Folson (971129)]

Bruce Abbott (971127.1400 EST)]

There is no incentive at all. All that has happened is that he now can
produce an output to reduce his error signal. Incentive is an illusion, >>at

least, it is to a PCTer.

Now I wonder if in all cases apparent incentive is simply the result of
the recognition of opportunity for potentially error reducing output?

Hank, I agree completely with your analysis (indeed, it is identical with >my

own analysis), but not with your conclusion that "there is no >incentive at
all."

Our disagreement here, I suspect, is completely one of definition.

Of words or psychology? I think what is blinding mainstream psychology is the
giving of unwarranted importance to observable behaviors. "offer an incentive"
is a behavior. "performance of some act" is another behavior. In mainstream
psychology, behaviors like these are mistakenly thought to have important
meaning, as they are thought to be the end results of some process(es). Under
PCT, behaviors are simply the outputs of a control system, used solely as an
_expedient means_ to reduce error signals, and are chosen not for their
consistency or appearance to an outside observer, but according to their
perceived likelihood of reducing some internal error signal.

I think of the rubber band experiment in which, in addition to the standard
target dot on the page, there is a circle drawn nearby. The experimenter wants
the person to place their hand over the circle. The person has a goal to keep
the knot in the rubber bands over the dot. The experimenter then moves his end
of the rubber band so the person's hand must be over the circle in order to keep
the knot over the dot. The person doesn't care where his hand ends up, so he
does not resist his hand being there. He also makes absolutely no attempt to put
it there. His 'behavior' is an unintended consequence of controlling his
perceptions. Why would a PCTer use the word incentive here?

Sincerely, Hank Folson