>In another thread (Straight and Moral) questions
>arose concerning the manner in which fixed capital
>depreciates. Implicit in that discussion is the
>difficulty of separating depreciation from profit.
>Further problems arise in estimating the economic
>life of fixed capital. Given this, how can or is
>fixed capital be measured in empirical work
There's a bit of work on this, mostly connected with measuring the
parameters of the circuit of capital. Andrew Senchak wrote a 1981 thesis at
Columbia U. on this, and Peter Matthews at Middlebury College also worked
on this problem in his thesis work. Unfortunately neither of these efforts
have been published.
Basically you have to make an a priori assumption about the shape of the
depreciation schedule (such as exponential, or a Koyck lag, or pure time
delay), and then you can use econometric time series techniques to estimate