Andrew here, replying to Costas (ope-l 1250):
Again, I think Costas has hit the nail on the head.  Changes in the amount
 of labor extracted matter, benefit/hurt the capitalists, even if the
 material substance of the extra labor remains unchanged.  
Costas notes correctly that this is what Marx's value theory implies.
  This is exactly my point.  Only a temporal conception of the DETERMIN-
ATION of value can capture this.  
Let me try another simple example.  Assume 5 qtrs. of corn used as 
seed and 3 qtrs. paid as wages, and output of 10 qtrs.  Assume 5 units
 of living labor are extracted.  Assume either that the input value of
 corn = 1 or compute the unit value of corn simultaneously.  In either
 case, the unit output value also = 1.  In this case, then, both 
simultaneous and temporal valuation give a profit rate of 20%.
Now assume the same input and outputs, and same wages in the next
 period, but assume that only 3 units of living labor are extracted.
  If one's THEORY is that valuation is simultaneous, i.e,
  10v = 5v + 3
then the unit value falls to 0.6, but the profit rate remains 20%!
  This is because the input value is also 0.6, somehow, although
 corn entered with an input value of 1 (the output value of the first
 period).  What is really going on is just the Dmitriev-Sraffa theory
 that value is just a veil; profitability is determined solely by
 physical input/output relations.  Labor extraction per se is
 irrelevant.
But in the temporal conception, the unit INPUT value of the 2d period
 remains = 1 (the output value of period 1), and this enters into the
 DETERMINATION of the output value (actually, the unit input value
 figures in here because it is the same as the unit input price), so
 that we have
  10v(output) = 5*1 + 3,
and thus v(output) = 0.8.
The profit rate is thus 
0.8*10 - 1*5 - 1*3
------------------
1*5 + 1*3
i.e., zero.  Here, the reduced extraction of living labor has lowered the
 profit rate, just as Marx's law of the falling profit rate implies.
  Here, value really is being DETERMINED by labor-time, whereas in 
the simultaneist theory, value is really being DETERMINED by physical
 input/output relations, and labor-time just acts as an irrelevant
 numeraire.
Andrew Kliman