Superior productivity of American industrial workers

[TheMarcksPlan]

Wartime America, however, lost part of its comparative advantage when it switched to relatively sophisticated military/producer goods like airplanes and ships, for which mass production is not ideal.

This is another reason why, btw, America's air-heavy, small-army strategy didn't play to its strengths. The War Department (AAF and ground forces) spent only 4.1% of its procurement budget on artillery ammo in 1943 and 6.4% in '44. Global Logistics and Strategy v.2, app. C-1. As Navy Department was ~1/3 of war spending, probably 2.5%-4% of total went to army arty ammo. Germany spent 25-30%. Arty ammo is ideally suited to mass production and to transforming US steel production directly into combat power.

Had the US played to its strengths and spent 35% of its budget on arty ammo, it would produce ~12mil tons in '43 and '44 (vs. 2.6mil and 3.3mil German) . It could have buried the German army under an avalanche of shells, perhaps regardless of what happened in Russia.

That would of course assume a strategic orientation towards land warfare, which for political reasons was likely infeasible.

[historygeek2021]

I wasn't claiming that Tooze's article challenges Broadberry's. Rather, Tooze's article points out that Broadberry's thesis sparked a long-running debate on the topic. Your 1993 article is only the first in a 15-year debate. Again, no idea how that debate plays out. I might get around to reading it; will report back if I do.

The "debate" wasn't over whether Germany was close to the USA in labor, capital or total factor productivity. It was over the nature of capital investment in both countries and understanding West Germany's post-war economic boom, which Tooze sheds further light on by examining the type of machine tools that both countries invested in. There is no doubt in anyone's mind outside of this thread that Germany never came close to matching the industrial productivity (labor, capital and TFP) of the United States during the early and mid 20th Century.

A data set that doesn't cover WW2 - i.e. Broadberry's - will not tell us anything that happened during WW2. As the Broadberry article compares the late '30's and mid '40's, it misses intrawar patterns entirely.

The data set covers the entire 20th century. At no point does Germany come close to U.S. productivity. The idea that German productivity briefly spiked during WW2 and then plummeted for the rest of the century is preposterous. Find me a single scholar anywhere who supports such an assertion. Also, the Tooze article fills in the missing time period in Table 1, showing the gap between US and German labor productivity widened during the war, as did the gap in capital productivity (Table 8) and total factor productivity (Table 9).

KDF33 has produced compelling evidence that Germany:US labor productivity varied enormously during the war; his data show Germany:US productivity ratio varying between .90 (Feb '43) and .33 (Feb '44). If KDF33's trend is correct, as well as Broadberry's post-WW2 ratio of 2.42, then midwar German productivity was up to ~88% of American ( .33 / .9 * 2.42 ). KDF's data endpoints align very neatly with Broadberry's immediate postwar numbers.

No, Broadberry shows a U.S. advantage of more than 2:1 immediately before and after the war. KDF has analyzed labor productivity of a single item (aircraft frames sans engines and propellers), and found that German labor productivity may have reached as high as 70% of U.S. productivity at one point in time (1943). The other data in this thread support the finding that German labor productivity in this one area may have approached the 60-70% range at this one point in time, but that fails to take into account the longer German work-week, and in any event, a single data point doesn't tell us much about what both countries were capable of achieving over time.

Is there a plausible story that explains KDF33's picture of German productivity converging towards American during midwar, then rapidly plummeting in ratio to US? Yes, absolutely there is:

Pre- and postwar America did what it did better than anyone: it mass-produced consumer goods like cars and refrigerators that were relatively simple, large, and expensive. Wartime America, however, lost part of its comparative advantage when it switched to relatively sophisticated military/producer goods like airplanes and ships, for which mass production is not ideal. This caused US-Germany productivity convergence that is apparent in mid-war stats. From early '43 onwards, however, bombing effects and labor force degradation caused German productivity to plummet, relative to American (and in the aircraft case, a German shift to fighters in '44 further complicates the data). By war's end, these dynamics had caused Germany to lose her relative wartime productivity gains.

Do I think 88-90% of American productivity is the "true" German norm? No, that's probably a bit high. German factories were farther along the learning curve at the early '43 relative peak than were American. But very clearly 30% or 40% German "true" wartime productivity is too low.

Or Germany spent an entire decade overinvesting in military airframe production (even going so far as to produce more airframes than available engines) while America was just ramping up production in 1943, leaving Germany in the dust by 1944 even without taking into account German industry dispersal.

In any event, the data is clear that for manufacturing as a whole, the United States was 2-3 times ahead of Germany during the war.

[KDF33]

In any event, the data is clear that for manufacturing as a whole, the United States was 2-3 times ahead of Germany during the war.

It's not clear at all. Using this source for U.S. labor, and the well-known figures for munitions production, I get:

U.S. / German munitions $ value, per worker:

1942: 3,137 / $2,110 = 149%
1943: 4,636 / $2,949 = 157%
1944: 5,355 / $3,497 = 153%

With average yearly labor forces in the munitions-producing sectors of 6,376,000 / 8,196,000 / 7,843,000 for the U.S. and 4,029,000 / 4,578,000 / 4,862,000 for Germany.

Munitions output value is the widely-reproduced data from Goldsmith, as seen for instance here.

[historygeek2021]

In any event, the data is clear that for manufacturing as a whole, the United States was 2-3 times ahead of Germany during the war.

It's not clear at all. Using this source for U.S. labor, and the well-known figures for munitions production, I get:

U.S. / German munitions $ value, per worker:

1942: 3,137 / $2,110 = 149%
1943: 4,636 / $2,949 = 157%
1944: 5,355 / $3,497 = 153%

With average yearly labor forces in the munitions-producing sectors of 6,376,000 / 8,196,000 / 7,843,000 for the U.S. and 4,029,000 / 4,578,000 / 4,862,000 for Germany.

Munitions output value is the widely-reproduced data from Goldsmith, as seen for instance here.

Page cites and showing your work would help.

And yes, the data from Broadberry and Tooze I presented above is clear. No one has refuted that.

[KDF33]

Page cites and showing your work would help.

I'll detail it later tonight - was a quick post, being in a hurry.

And yes, the data from Broadberry and Tooze I presented above is clear. No one has refuted that.

The data from Broadberry doesn't track wartime inputs and outputs and, frankly, I'm curious as to how he calculated it. Ultimately, whether or not it is correct in the abstract (and I have my doubts), it doesn't tell us much that is relevant to this discussion.

As for Tooze, he uses Goldsmith's data for U.S. and German overall armaments output.

[historygeek2021]

Page cites and showing your work would help.

I'll detail it later tonight - was a quick post, being in a hurry.

And yes, the data from Broadberry and Tooze I presented above is clear. No one has refuted that.

Ok, take your time, no rush.

For what it's worth, your data seem to show that German munitions labor was more productive than the United States during the war, a result that seems absurd.

WW2 Labor Productivity KDF Munitions Data.png

The data from Broadberry doesn't track wartime inputs and outputs and, frankly, I'm curious as to how he calculated it. Ultimately, whether or not it is correct in the abstract (and I have my doubts), it doesn't tell us much that is relevant to this discussion.

As for Tooze, he uses Goldsmith's data for U.S. and German overall armaments output.

The relative industrial productivity of Germany and the United States is the topic of this thread, and it's exactly what Broadberry analyzes. No one has produced any substantive rebuttal. Tooze confirms Broadberry's findings. The United States was far more productive than Germany throughout the first three quarters of the 20th century. It's a simple fact.

[TheMarcksPlan]

The data set covers the entire 20th century

Except WW2.

Tooze has a data point for 1944 but only compared to 1929; the entire point in dispute regards the nature and magnitude of intrawar trends.

[historygeek2021]

The data set covers the entire 20th century

Except WW2.

Tooze has a data point for 1944 but only compared to 1929; the entire point in dispute regards the nature and magnitude of intrawar trends.

No, Tooze shows the change in each country from 1939 to 1944 for labor productivity, capital productivity and TFP:

WW2 Labor Productivity Tooze Table 1.png

WW2 Labor Productivity Tooze Table 8.png

WW2 Labor Productivity Table 9.png

[TheMarcksPlan]

The data set covers the entire 20th century

Except WW2.

Tooze has a data point for 1944 but only compared to 1929; the entire point in dispute regards the nature and magnitude of intrawar trends.

No, Tooze shows the change in each country from 1939 to 1944 for labor productivity, capital productivity and TFP:

WW2 Labor Productivity Tooze Table 1.png
WW2 Labor Productivity Tooze Table 8.png
WW2 Labor Productivity Table 9.png

Between 39 and 44 (I'm not reading 39 as year in which wartime productivity dynamics showed significantly). KDF, as I read him, doesn't dispute massive productivity differential in 44. I certainly don't.

The entire point of the thread revolves on the midwar productivity gap - we argue it was significantly lower than 44's and that 44's stems from bombing, labor quality, etc.

Tooze - all modern scholars AFAIK - has nothing directly bearing on those points.

You're right though that Tooze compares 44 to 39 and 29, I was incorrect in saying only 29.

[KDF33]

Ok, take your time, no rush.

For what it's worth, your data seem to show that German munitions labor was more productive than the United States during the war, a result that seems absurd.

Here's the data:



Source, U.S. munitions output: Mark Harrison, reproducing data from Goldsmith
Source, U.S. workforce: American Industry in War and Transition, Table 12 (p. 35), reproduced here:



The relevant categories are 'Iron & Steel & Products', 'Machinery (excl. electrical)', 'Transportation Equipment (excl. Auto)', 'Automobiles', 'Nonferrous Metals & Products' and 'Arsenals and Navy Yards'.

These categories correspond to Germany's 'Foundries', 'Machinery, transportation and other industrial equipment', 'Iron, steel and plated products', 'Other iron and steel products' and 'Miscellaneous metal products', as well as their corresponding 'Basic materials' categories. These five categories accounted in 3Q43 for 94.0% of Germany's armaments and ammunition production, and 97.7% of that production accomplished by the metalworking industries.

***

Source, German munitions output: Mark Harrison, reproducing data from Goldsmith
Source, German workforce: USSBS, The Effects of Strategic Bombing on the German War Economy, Appendix Table 9

To construct yearly averages, I used the data for 05/31 and 11/30, as well as for 11/30 of the preceding year. Thus, the average for 1942 is 11/30/1941, 05/31/1942 and 11/30/1942. 1943 follows the same pattern. For 1944, for lack of data it is just the average of 11/30/1943 and 05/31/1944.

The categories used for Germany are:

2. Blast furnaces, steel works and rolling mills
3. Nonferrous metals
8. Foundries
9. Machinery, transportation and other industrial equipment
12. Iron, steel and plated products
13. Other iron and steel products
14. Miscellaneous metal products

They correspond to the categories used for the U.S.

The relative industrial productivity of Germany and the United States is the topic of this thread, and it's exactly what Broadberry analyzes. No one has produced any substantive rebuttal. Tooze confirms Broadberry's findings. The United States was far more productive than Germany throughout the first three quarters of the 20th century. It's a simple fact.

More productive, yes. 'Far more' is just a superlative. The data indicate that Germany maintained roughly 2/3 of the U.S. productivity in the years 1942-44, and this with a significant share of its workforce being constituted of slave labor, to say nothing of the adverse effects of strategic bombing that began in 1943.

Note that this 2/3 overall ratio closely tracks that calculated on a weight basis for airframes.

[KDF33]

I've added the U.K. to the table:

[historygeek2021]

Ok, take your time, no rush.

For what it's worth, your data seem to show that German munitions labor was more productive than the United States during the war, a result that seems absurd.

Here's the data:

Did you take into account the proportion of each workforce producing for the armed forces?

Here is what I get if we take the numbers for USSBS Appendix Tables 11 and 13 and compare them with the manufacturing number in Table 13 of the US War Manufacturing Board (p. 36):

WW2 Labor Productivity USSBS Tables 11 13.png

https://babel.hathitrust.org/cgi/pt?id= ... &skin=2021
https://books.google.ca/books?id=jAc_Qg ... &q&f=false

These figures show an even narrower productivity gap than yours. Note that we also agreed on a narrower productivity gap in munitions production upthread: viewtopic.php?p=2352550#p2352550

This suggests that the overall US productivity advantage in manufacturing established by Broadberry did not entirely carry over to the production of munitions. One possible reason is the relatively quick timeframe in which workers in the US entered and exited the armaments industry (by 1944 the number of armaments workers was already declining). Another possible explanation is differences in the exchange rates used by Goldsmith and Broadberry. Finally, Broadberry does not have to allocate his worker pool between war and civilian production, nor does he have to differentiate between military and civilian outputs, allowing him to give more accurate estimates of worker productivity in each country.

[KDF33]

Did you take into account the proportion of each workforce producing for the armed forces?

No. I am unaware that any breakdown between war and non-war exists for the categories in question.

Here is what I get if we take the numbers for USSBS Appendix Tables 11 and 13 and compare them with the manufacturing number in Table 13 of the US War Manufacturing Board (p. 36):

I don't understand what the figures for German labor represent. How did you derive them?

This suggests that the overall US productivity advantage in manufacturing established by Broadberry did not entirely carry over to the production of munitions.

If that is so, then Broadberry's data tell us very little about the U.S. and Germany's respective war-industrial potential. It also raises questions about the inevitably of the U.S. outproducing Germany's Grossraum.

[historygeek2021]

I don't understand what the figures for German labor represent. How did you derive them?

They are the yearly averages of the USSBS numbers I've boxed in yellow below:

USSBS Table 11.png

USSBS Table 13.png

[TheMarcksPlan]

This suggests that the overall US productivity advantage in manufacturing established by Broadberry did not entirely carry over to the production of munitions.

In addition to the reasons you list, again we have the US moving from exceptionally efficient mass production of simple consumer towards "bespoke line" production of sophisticated military/producer goods, where its efficiency isn't superlative. Meanwhile Germany moves from low-volume benchwork to "bespoke line" production, reaping economies of scale impossible within the smaller German peacetime market.

This dynamic only arises during the war and largely disappears with its end, completely missing from Broadberry's data and (due to bombing etc) from Tooze's 1944 data.

Broadberry seems to accept the thesis that the USA's productivity advantage lay in its use of manufacturing processes that, by European standards, were "wasteful" of raw materials.

Returning to this earlier point to say... US rationed steel during WW2 and controlled its price (I can track down a cite but assuming it's common knowledge). Under that condition, firms' profit maximization strategy would have changed from being liberal with steel to conserving it. Again, it's a war-specific shift away from practices that maximize US labor productivity.

Another possible explanation is differences in the exchange rates used by Goldsmith and Broadberry.

A massive potential discrepancy. Tooze uses 3.8 for wartime; Goldsmith assumes <2.5. Using Tooze's figure increases the national productivity disparity by >50% but obviously doesn't change the physical armaments outputs, leaving us to assume enormously higher output in the US civilian sector.

As most economists recognize, however, US real civilian wartime consumption is overestimated by official price indices that don't adjust inflation for, e.g., qualitative decline of controlled goods. https://www.independent.org/publication ... asp?id=138 Consumption was either constant or slightly declined after '41, contrary to official statistics.

If that is so, then Broadberry's data tell us very little about the U.S. and Germany's respective war-industrial potential.

At least it is far from the whole story. Weird things happen in war; a forced temporary convergence of US and German manufacturing practices appears to have been one of them.