[Marian Griffith]
>> - This is not a lower price for a product sold by a female
>> seller. Nor a lower price for the time spent by a female seller.
> Nobody ever said it was.
You said exactly that it was an opportunity to study the devaluation
of a female's labour. You certainly feel that there's a 'devaluation
of female labour' somewhere. And that it has some bearing on
something important.
I thought it might be uncharitable to assume that you thought female
characters selling for less was a devaluation of those characters'
labor - a mistake like assuming that the price of a toy duck
reflects the valuation of the toy duck's breeding potential. A toy
duck is not a duck. It is a real toy duck, but it is not actually a
duck. A female character is not a female. It is a female character,
but it is not actually a girl or woman. But if no one said that it
was about the actual sellers, then it must be about the
character. You think there's something important about the valuation
of a character's labor (poor character!) OK...
>> Rather, it is a lower price for a commodity labeled 'female,'
> Exactly. Hence the parallel that was drawn with the observed fact
> that work done by women is valued less by companies than the same
> work done by men. The author of the article pointed out that here
> the common explanations for that discrepancy did not apply, and
> thus concluded that the difference in selling price for female a-
> vatars was an indication by how much society undervalued female
> labour.
"Here is an orange, and here is a forklift. Both roll along the
ground (albeit in different ways), and while I recognize that the
forklift is not a fruit, it is only natural to assume that people
treat it as if it were a fruit. Now, via the pricing of forklifts, I
have a model system for why people value oranges from different
places differently (useful since it rules out the possibility of
different flavors coming into the preferences). If I see a pricing
difference in Chinese vs. American forklifts, it can't be due to
differences in flavor, freshness, water content, suitability for
fruit salad etc. - these are forklifts for God's sake! - and so it
must be for some reason which basically bears only on them being
Chinese or American forklifts. Because forklift prices are a model
system for orange prices (and, clearly, how highly people think of
different kinds of orange) we arrive at the natural conclusion that
any price differences observed in Chinese vs. American oranges,
which were formerly argued to have something to do with one or the
other being better in some legitimate way, are now revealed as
having no legitimate basis in the attributes which oranges should be
compared on."
I don't mean to be snide or dismissive, but only to point out in the
clearest way possible that there is some kind of non-obvious
auxiliary reasoning which is not being exposed, and which would be
necessary in order to think that drawing the parallel was sane to
begin with. But no one is exposing it, which you'd think they would
if the case were as strong as suggested.
> If you can find that 1) there is no difference between the abili-
> ties of elf clerics as opposed to the clerics of other races and
> 2) there is indeed a price difference then, yes, you have an
> argument that undermines the conclusion as it was picked up by the
> media. You also would have proof that, in the game context, elf
> clerics were discriminated against.
Just to make sure, I wasn't arguing that this would undermine the
conclusion as it was picked up by the media.
Suppose that one thing within the game is being sold, and gets a
lower price than another thing, but that thing happens to be
something for which there is no human 'parallel' in real life. It
naturally follows that we cannot cry foul on behalf of the real
beings who are similar to the items for sale (elf clerics), since
there are no elf clerics to be discriminated against.
That leaves 'female' characters in the same place, albeit sharing a
word with a real life group. But, once again, a toy duck is not a
duck just because it shares a word and both have some kind of tail.
> The author of the original article believes he has found a situ-
> ation in which this effect could be demonstrated where the usual
> arguments in favour of that difference (e.g. real difference in
> skill, difference in seniority) do not apply.
That's just the thing. They don't apply in the real world. The gap
WIDENS when you control for education and hours worked (and any
number of other variables). Women make more for part-time work
(possibly because men are supposed to work full time, and the ones
who work part time are less motivated or educated or what have you).
When I only look at women who have no children the wage gap
continues merrily on. You can get fluctuations with several
variables, but when you put them together they don't wipe out the
gap.
> In a game setting female avatars should be exactly equally capable
> as male ones, and seniority is not an issue either (nor is the
> possibility of maternity leave).
This would be nice if female avatars were really female in any
meaningful sense and we had a sensible model of wage discrimination
within the world, which drew on the same independent
variables. Instead we have female avatars (like toy ducks) selling
for less than male avatars (like stuffed ducks). We have
something-not-female, of which we can say something-not-
wage-discrimination. How do you map the pricing of forklifts onto
prejudices about oranges? It's possible, but not by assuming their
equivalence.
> Because of the relatively large difference observed and the fact
> that it appears (or seems to appear) in a situation that is free
> of gender distinction (if not necessarily bias) this is an im-
> portant observation.
To a man with a low p-value, everything looks like a question. A
mean of the wrong measurement is meaningless, and by every sign I
can see this is not just the wrong measurement, but of the wrong
things even within EQ, to draw the conclusion that is going around
in the media.