This paper is a (to me) startling undermining of a bedrock concept for WW2 economics. The authors plausibly argue that much of the famous "learning by doing" (LBD) touted by wartime and postwar economic analysis is at least far overstated, possibly entirely fictional.
The paper is quite dense, with subtle but necessary distinctions between causes of productivity rise. On some levels they distinguish productivity gains from LBD that most of us would subsume under LBD as we use the term in our lay discussions. Here I'm thinking of engineering/investment improvements - the authors focus on labor LBD but we might consider engineers figuring out how better to structure ongoing investment as LBD for our purposes.
Many of the points, however, are indeed a direct challenge to how we (amateur or professional) historians disaggregate LBD from other sources of productivity gain. Here the authors discuss industrial planning:
This passage is potentially of particular relevance to discussion of the German early-war economy. Though the authors reviewed only US-focused data and studies, human behavior in Germany was not radically different (for this discussion). The passage suggests that firms withheld productivity-enhancing planning/investment until they were assured of long/large production runs. GSWW v.5-1 convincingly argues that German firms anticipated imminent return to peacetime production until well into 1941, so it's plausible that these expectations held back German productivity.Firstly, during the pre-production phase and in the early stages of production, there
were often considerable uncertainty about future production rates and about the duration
of the programmes and hence about total production volume. This limited the extent to
which methods and techniques could be planned and designed at the pre-production stage
in order to fit closely with future rates and total volumes of output. But, as this
uncertainty declined over time, more resources were often allocated to change the
production processes used.
Another illustrative passage:
Sticking with the German case, firms and/or even the SS later began training courses for foreign workers, including for Zwangsarbeiter (the SS would presumably get higher "wages" for higher productivity from its camp inmates).In addition, several comments suggest
that a fair amount of the learning resulted from purposeful training rather than being a
costless by-product from doing. Stanley (1949), for example, refers to the important role
of training in improving the performance of workers. Carr (1946) also notes that training
of inexperienced workers was necessary in many situations since ‘experienced aircraft
mechanics were not available’, but these inexperienced operators, ‘once taught, became
highly proficient’.
One more:
This is another important factor to the early-war German economy. The materials allocation chaos of the quota system plausibly hurt productivity by grinding down production to the pace of material bottlenecks.Rapping (1965) also discusses the importance of
improvements in the reliability of supply and in the quality of materials in the
shipbuilding industry.
Finally, the paper makes a point that seems obvious in retrospect: increasing production volumes and productivity rise happened simultaneously (in general). Therefore we have an inescapable temporal correlation between economies of scale and productivity, which the older studies often fail to disaggregate. The older studies often ascribe temporal trends in productivity to LBD superficially, without analyzing whether economies of scale (and the other above-discussed factors) played a role.
-------------------------------------------------------
The paper postdates Tooze's Wages of Destruction and influential studies of German wartime productivity such as "Fixed-price contracts, learning, and outsourcing: explaining the continuous growth of output and labour productivity in the German aircraft industry during the SecondWorldWar" by Budrass, Scherner, and Streb. That paper discusses "learning-by-doing" throughout, emphasizing it as a central plank in a narrative of smooth German productivity growth during most of the war ("the driving forces of productivity growth were primarily learning-by-doing and outsourcing").
What's surprising is that the paper shows that the older studies on which the LBD narrative relies were often themselves dubious about the extent and nature of LBD. These older studies contain numerous caveats regarding investment, training, and supply changes, for example. None of these caveats emerge from the specialist literature into the historiography (nor, in line with the authors' concerns, into literature on industrial policy).