Why Longer Processes Are More Productive
Thursday, April 10th, 2008The term “roundabout” does not mean “requiring more capital and intermediate steps before final production” by rather simply “requiring more time.” According to Gene Callahan, “Eugen von Böhm-Bawerk… attributed the bulk of increases in productivity to the adoption of more time-consuming, or roundabout, methods of production.” (Economics for Real People, 131ff) Adds Mises: “As acting man prefers those processes which, other things being equal, produce the products in the shortest time, only such processes are left for further action which consume more time. People embark upon these more time-consuming processes because they value the increment in satisfaction expected more highly than the disadvantage of waiting longer for their fruits.” (Human Action, 481) It is time that costs money, along with labor and land; the actual number of intermediate steps is next to irrelevant. If it takes 10 years to carry good A through 10 transformations down to the final consumer good, and it takes 1 year similarly to carry good B through 100 transformation, then the process involving A is more roundabout than the process involving B. “Roundabout” may be synonymous with “circuitous,” but its technical economic meaning is “requiring more time.”
Four points:
- Callahan continues that “It is not always true that the only available method of increasing production is to adopt more roundabout methods. Perhaps a shorter route to some goal just has not been imagined yet. … But historically, the constant agitation of humans to improve their circumstances is such that most of the opportunities to increase productivity lie in the adoption of more roundabout processes. Humans are adept at spotting the direct route to a goal in the first place.” (134)
Reply: Suppose Crusoe uses sticks to collect berries and replaces his stick once it wears out, a process which takes him 1 day. But then he finds a cache of sticks on a steep hill, such that to go there, grab a stick, and come back takes 1 hour. Unfortunately, climbing the hill is too much for Crusoe. So, we see that not all shorter processes are in use.
- Isn’t it often possible to initiate a more productive and at the same time shorter process with the help of new technology? What if it’s the same process yet with lower costs of, say, labor (e.g., Crusoe gets healthier due to all the manual work he had to do and capable of climbing the hill readily)? Won’t it be more productive yet take the same amount of time?
Reply: Yes, but reducing the time it takes to build a thing is not the only way of cutting costs. One can also rely on less labor or less land. Finding an equally good shorter process will mean disinvestment with the same revenues and therefore greater profit. The problem arises if no way of profitably disinvesting can be found. Then only longer processes remain to be considered.
- Time is just one of the traditional three factors of production. Why not say that if a particular process has expensive labor, it will be undertaken only if it is sufficiently productive to offset the extra costs? What’s so special about time? Why do we not say: “more labor-costly processes tend to be more productive”? Or: “processes which use more and more expensive natural resources tend to be more productive”? Or don’t we?
- Suppose process A takes 1 week to produce its consumer goods. Process B is just like A but is preceded by a 3 week diversion of digging ditches and filling them back up. Surely, B is less productive than A, and yet it is longer. What’s going on here?
See below for answers to (3) and (4).
Now it is true both that longer processes tend to be more productive, and that more productive processes tend to take longer.
I. Longer processes tend to be more productive, if we assume that entrepreneurs tend to profit despite higher interest outlays. Again, mixing labor with natural resources or higher-order capital goods in order to advance them down to the final consumer goods or having the goods mature on their own (e.g., wine) takes time. The more time you have, (1) the more labor productivity-increasing these goods must needs be; and (2) the greater the amount of labor that can be mixed in a step-by-step fashion which makes it possible to produce such goods (which will probably be more complex and sophisticated purely from the engineering perspective). Otherwise, no entrepreneur will bother investing. Profits due to higher productivity of labor have to outweigh the extra interest payments (in terms of money; or, in Crusoe’s case, the psychic profit of more berries a week from now and for as long as the stick lasts have to outweigh the psychic cost of lessening present consumption while the stick is being built.) It is true that digging and refilling ditches will make a process longer, but that’s why we say that longer processes merely tend to be more productive: we assume the human desire not to waste production time.
To put it simply, longer processes are more expensive to set up. Hence they had better be more productive. And the same thing can be said about the other original factors: a process requiring more labor or land is more expensive, so, analogously, it’d better be more productive, as well.
We have to see why time is a factor of production. There is a general relation that the more labor-saving you want your machine to be, the more time you have to spend constructing it. You can accomplish more with more time, just as you can accomplish more with more labor or land. There is, of course, no precise equation connecting physical productivity of a machine and the time invested into building it, but a general proportion still holds. In other words, you can hire several laborers to work on things simultaneously; or you can hire a single laborer and give him more time, depending on your production needs. In both cases you invest more and hope to gain more. (Or you can do both, if all stages of production exist alongside one another at any moment in time.) Otherwise there is no explanation for our phenomenon.
II. More productive processes (producing either (a) much more of the same thing or (b) things that were completely impossible to produce in shorter time intervals) tend to take longer, because there is an incentive to utilize all the feasible shorter processes first, other things being equal: this way the waiting time will be as low as possible. I add “feasible,” because there can always be a short process that, for example, is hugely expensive in terms of the original factors other than time, such as labor (e.g., Crusoe’s climbing a hill), and for that reason is eliminated from consideration despite its shortness. However, in the case of (b) the process will be not longer but, on the contrary, the shortest possible one to a particular want-satisfaction. Even in the case of (a) the allegedly longer process is nothing of the sort; it is rather the most efficient way of satisfying a desire for more goods or services. From the point of view of technology, the new process is longer, though physically more productive; from the point of view of economics, it is the shortest one, given that we aim at a certain end and disregarding entrepreneurial error and ignorance.
Again, simply put, we want to get results as soon as possible, so it is likely that most the shorter processes have been perfected, and any increase in business efficiency must therefore require more time-consuming (and therefore more productive for reasons outlined in (I)) production processes. Of course, that’s just one of the ways of boosting efficiency.

The answer to that important question is that what we are concerned with is not so much the structure itself, but (1) the fact that advancing capital goods of higher order into capital goods of lower orders and finally into consumer goods takes time; and (2) that that time costs money, an income called interest paid by the demanders of present goods (and suppliers of future goods = borrowers) to the suppliers of present goods (and the demanders of future goods = lenders). Thus, if interest is paid once a year, and a particular production process takes 5 years, then we can consider that process’s structure to have 5 layers, each layer representing payments to the higher-order capitalists and being shorter than the previous one as we move up towards earlier stages, because of payments to factors and interest. If interest is paid once a month, then the structure will have 60 stages. That the real particular structure of any process involves numerous transformations of intermediate goods each taking various amounts of time is true but irrelevant for our analysis. There thus comes into consideration a general production structure reflecting the timing of interest payments and summarizing the payments to factors in between.