>        1st technique      2nd technique       1st technique
> 
>     |<--------------->|<----------------->|<-------
>     0                50%                 100%
> Suppose we compare stationary states producing 1 widget each year. For the
> first technique, the output-labor ratio is 1/7 Widgets per worker. It
> is 1/8 Widget per worker for the second technique. Note that around an
> interest rate of 100%, a lower interest rate is associated with a lower
> output-labor ratio. (Isn't your intuition that a lower interest rate
> implies more capital per head implies greater output per head?)



There isn't any capital in the model, or if there is, its level cannot be 
changed? Anyways, this is again the Leontieff model, and even w/ 2 kinks, I 
find L to be not the appropriate choice for models where the crucial issues 
are inputs and their prices and especially when the model is supposed to 
discuss long run positions.

Anyways, to inject some realism, first, I find interest rates in excess of few 
dozen percentage points unappealing (there's no risks here). We could fool w/ 
the numbers to have cut-off interest rates less than 50 %?

Second, here, the interest rate is important b/c the hiring and manufacturing 
occurs before the item is payed for. So, the model applies only (i) to items 
that take a long to produce and (ii) in mkts where the buyer cannot or will 
not pay before delivery. If production takes days or weeks instead of years, 
the interest charges are next to irrelevant. If the buyer pays ex ante or as 
the costs of the project incur, the interest charges are next to irrelevant. 
In these cases, the first tech would always be preferred, and no switching 
would occur.