Lesson 7 of 7
The attacks on the measure
Knight in the 1930s and Cambridge in the 1960s both attacked the idea that capital has a measurable length. What each established, and what survives.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- The first dispute, in the 1930s, was definitional: whether capital is a permanent homogeneous fund or a set of specific produced goods arranged in time.
- The second, in the 1960s, was technical: reswitching showed that techniques cannot be ranked by how capital intensive they are independently of the interest rate.
- What survives for this track is the qualitative claim only. Production has a time structure and some activities sit further from consumption. There is no defensible single number for its length.
Lesson three said the average period of production is the weakest part of this tradition. This lesson says who attacked it and what they established, because a course that taught the measure and skipped the attacks would not be teaching.
Two disputes, thirty years apart, about the same idea
- 1935Knight'Professor Hayek and the Theory of Investment', Economic Journal, volume 45, attacks the idea of a period of investment.
- 1936Hayek'The Mythology of Capital', Quarterly Journal of Economics, volume 50, replies in defence of treating capital as structured in time.
- 1960SraffaProduction of Commodities by Means of Commodities sets out the framework the later dispute is argued in.
- 1966The symposiumA group of papers in the Quarterly Journal of Economics, volume 80, settles that reswitching of techniques is possible.
The 1930s dispute: what kind of thing capital is
Knight held that capital is best treated as a permanent fund of value: a homogeneous quantity that yields a return, is maintained out of its own earnings, and has no meaningful internal time structure. On that view a period of investment is not a useful concept and asking how long a process is asks nothing.
Hayek held the opposite, in the tradition of the first lesson of this course: capital is specific produced goods, arranged in a structure, and the arrangement is what a theory of investment is about. He replied to Knight in the Quarterly Journal of Economics in 1936, in a paper whose title tells you the temperature of the exchange.
Neither side conceded. What the dispute established is that the disagreement is about the definition and cannot be settled by evidence, because the two definitions make different things visible and hide different things.
The 1960s dispute: reswitching
This one is sharper because it is a mathematical result rather than a preference.
The intuitive picture behind roundaboutness is that a lower interest rate makes longer, more capital-intensive techniques worth adopting, so techniques can be ranked by capital intensity and a falling rate walks you up the ranking.
Reswitching is the demonstration that this can fail. A technique can be the profitable one at a high rate of interest, be replaced by another at middling rates, and become profitable again at low rates. If a technique can be chosen, dropped and chosen again as the rate falls, then techniques cannot be ordered by capital intensity in a way that is independent of the rate, and there is no physical quantity of capital sitting behind the ordering.
The dispute ran between economists at Cambridge in England and Cambridge in Massachusetts, it was argued in the framework Sraffa set out in 1960, and it came to a head in a group of papers in the Quarterly Journal of Economics in 1966.
What this track keeps
Two claims have to be separated, and the honest reading keeps one.
The quantitative claim, that the capital of an economy has a length or an intensity expressible as one number, does not survive. Nothing later in this track uses it.
The qualitative claim does survive, and it is the one the rest of the track runs on: production has a time structure, some activities sit further from final consumption than others, and a change in the cost of waiting changes which activities look worth starting. That claim needs no index number. It needs only that some projects pay off later than others, which no dispute has ever denied.
Check yourself
What was the 1930s dispute about?
- Whether interest exists
- Whether capital is a permanent homogeneous fund or a structure of specific produced goods
- Whether roundabout methods are more productive
Show the answer
Answer: B.
Knight argued for the fund conception and against a period of investment; Hayek replied in 1936 defending capital as structured in time. It is a definitional disagreement.What does reswitching show?
- That longer methods never produce more
- That techniques cannot be ranked by capital intensity independently of the interest rate
- That interest rates cannot fall to zero
Show the answer
Answer: B.
If a technique is chosen, dropped and chosen again as the rate falls, no ordering by capital intensity exists that is independent of the rate.Which claim does this track keep after these disputes?
- That the average period of production can be computed for an economy
- That production has a time structure and some activities sit further from consumption than others
- That capital is a fund with a single rate of return
Show the answer
Answer: B.
The qualitative claim needs no index number and survives. The quantitative one does not, and nothing later in this track uses it.Sources
- Bohm-Bawerk, The Positive Theory of Capital, 1889, English translation by Smart, Book II, chapter II, for the average period of production that both disputes are about.
- Knight, 'Professor Hayek and the Theory of Investment', Economic Journal, volume 45, 1935, and Hayek, 'The Mythology of Capital', Quarterly Journal of Economics, volume 50, 1936, as the two sides of the first dispute.
- Sraffa, Production of Commodities by Means of Commodities, 1960, and the group of papers on the reswitching of techniques in the Quarterly Journal of Economics, volume 80, 1966, for the second.
What this leaves out
- This is a heterodox tradition. The mainstream account of the same phenomenon differs, and nothing here is a forecast or a view on any security.
- The papers by Knight, Hayek, Sraffa and the 1966 symposium are named here with their journals and years so a reader can find them. They were not read in the original for this lesson, and nothing above reports a specific argument from inside any of them beyond the position each was arguing for.
- The reswitching example is an invented sketch of the shape of the result, not a worked numerical case. A reader who wants the arithmetic should go to the 1966 papers.