One of the issues Böhm-Bawerk had to deal with in his theory of roundabout production was measuring the period of production. He noted that technically all production that uses any bit of previously produced capital may be traced back to primitive times when the first bit of capital was built up to facilitate longer, but more productive, production. This means that the length of the structure of production would span millennia. Böhm-Bawerk resorted to thinking in terms of the average period of production and even dabbled in the (what I think is more tractable) investment period.
Klaus Hennings, in his short, but content-dense book, The Austrian Theory of Value and Capital, decides that even Böhm-Bawerk’s average period of production concept is deficient.
At this point Böhm realized that the existence of durable capital goods causes difficulties for the determination of the period of production, and therefore proposed to reckon with an average period of production instead. This is then used as an index for the roundaboutness of capital, so that the production process is described in terms of the average period of production.
It is easy to see that this argument must run into difficulties. Two main objections may be singled out from among the host of those that can be levelled against it. First, the presence of durable goods and the possibility of further usage of the materials from scrapped capital goods leads to an almost infinite historical regress […]
Second, Böhm’s device of using an average does not solve this difficulty. It is only too easy to show that rather restrictive assumptions must be made about the sequence of non-produced inputs in order to get a consistent ordering of these averages — quite apart from the fact that it may be impossible to identify the non-produced inputs which were used for the production of some particular unit of consumable output even if one disregards the possibility of joint production.
Hennings offers a way to conceive of the length of production that does not suffer from these problems:
Consider first of all a production process which is well synchronized and stationary; even if there is a lag in production it will not show up.
By “lag”, he means there is some unrealized change in output. Something has changed in an earlier stage that has not manifested itself as a change in output, yet. Hennings is abstracting away from these sorts of production hiccups.
Technically the output obtained in this period may be due to the non-produced inputs of earlier periods, but it is possible to consider the production process as if it were direct rather than roundabout.
Read: the roundaboutness of production is beside the point. Capital-intensity, complexity, roundaboutness, whatever.
The lag will show up only when the production process for one reason or another is not well synchronized (as Böhm himself noted at one stage). [Footnote: “Böhm (1895b), 126-7.”] This can be used to ‘measure’ the lag involved.
So, something happens, then later on output changes. If production does not have any unrealized changes, it may be called “synchronized”. This “synchronization time”, or “adjustment time”, may be measured. Or at least theorized.
Consider a production process in an economy with full employment, and assume that a ‘marginal’ (in the Austrian sense) amount of all non-produced resources becomes available in addition to those already used. Assume further that these additional inputs are immediately made us of, that they are used to increase the scale of operations at an unchanged method of production, and that there are constant returns to scale. Then if production does take time, some period will elapse before the production process as a whole is adjusted to the higher level of activity (as shown in a higher level of output) made possible by the increase in non-produced resources. It is this period of adjustment which I suggest is the period of production germane to the production process in operation, an which is the technical concept Böhm should have defined.
Instead of Friedman’s helicopter drop of money, imagine a helicopter drop of laborers or some other non-produced factor, immediately ready to be integrated into production. The time between their employment and the full increase in output is the measure Hennings is talking about here.
My initial response is that this is a good way to think about the length of production, especially compared to Böhm-Bawerk’s treatment, but I’d like to give it some more thought. There are a lot of assumptions thrown in, which isn’t necessarily a bad thing, but they should be considered carefully to make sure they don’t confound or assume the result.