, which may be pretty small.(II) Profits and Plasticity of Prime Costs. p. 81 seq.
See accompanying letter
I agree that if the equality between marginal costs and marginal revenues is maintained during a fluctuation in output of given magnitude, then the plasticity of prime costs will not affect the relative share of profits. But I do think you make the whole thing seem too simple, and too determinate, when you write (p. 81) "A rise in these rates of pay involving a proportionate rise of marginal costs is associated, ceteris paribus, with a rise of prices". Among things which are definitely not equal are entrepreneurs' expectations of market conditions. For prices to rise proportionately with prime costs it seems to me that two conditions must be satisfied. (i) The entrepreneurs expect market demands to increase so that they can continue to market the previous output at proportionately higher prices. (ii) The entrepreneurs & rentiers decide to increase their own disbursements proportionately, and at once, without staying to see if their first expectation is realised. If either of these conditions fail, the stream of money spent will not rise sufficiently to absorb the whole output at proportionately higher prices, and the equality between marginal costs and marginal receipts breaks down, and/or stocks are accumulated with subsequent effects upon output. I greatly doubt whether either of the required conditions would in fact be realised, and I think readers (if they are like me) will be helped to overcome their "mental resistance" to the "paradox", if you refer to the possibility (at least) of changes in prime costs not being accompanied by such behaviour on the part of the income recipients in the system as will enable the marginal cost, marginal return, equality to be maintained without some alteration of output. I really do not see how anything is gained when discussing this problem by concealing under a "ceteris paribus", the complicated nature of the prevailing "anticipations" which are assumed.
p.
84. line. 9. Is there not either a misprint of
[d] "prime" for "average", or an omission of the word
"average"? [5] When you say
that "the matter only becomes doubtful if the marginal cost is
greatly in excess of the average", it is perhaps worth noticing that
how "great" the excess must be depends on how rapidly the marginal
costs are increasing. If the marginal cost curve has a constant
elasticity, h, then the ratio of marginal to average costs is
constant, whatever the output, at
, which may be pretty small.
{If C = Total costs, x = output, [and] 
then 
, and 
; 
. \ 
.}
2. See Harrod's letter 563 to Robertson, [jump to page] .
3. Harrod added a footnote to p. 45, referring to a further footnote to p. 51 where the cogency of Allen's point is recognised.
4. Harrod, "Imperfect Competition and the Trade Cycle" ( 1936:7 ).
5. The word "average" was eventually inserted in the published version.