Avalancheon's source (hereinafter "Katkoff": https://www.jstor.org/stable/3159589?re ... b_contents
) gives us some contemporary regional breakdowns of production and land yield:
The total grain production in all the eastern regions could be estimated at 25.0 percent of the total Soviet production, which was produced on 30.3 percent of the total Soviet acreage. [data for 1940] @209
Arithmetic shows that the Eastern yields per hectare sown were 76.7% of western yields. Katkoff also gives 23.0% for Ukraine's share of grain production (p.208).
My OP estimated 75% east-west land yield ratio and 23.37% production share for Ukraine. Gotta say- not bad Mr. Marcks.
That said, my postulated productivity difference between Chernozem and non-Chernozem western land appears to have been too low: Given that the east-west ratio is ~.77 and that only ~ half of western land is Chernozem, it's likely that Chernozem land was at least 50% more productive than average eastern land.
What about the trend in land yield during the war?
Avalancheon already quoted the 1940 grain harvest at 118.8mil t; from a diagram on page 208 we can infer total Soviet grain-sown area at 110.6mil ha: yield was 1.074 t/ha.
Page 213 gives the 1942 stats as 53.1mil t grain production on 70.8mil hectares: yield of .75 t/ha.
Grain yield declined by 30% between 1940 and 1942.
As Katkoff and many other sources attest, much of this decline is attributable to shortages of labor, fuel, transport, and fertilizer. Indeed, as the yield erosion is greater than the 1940 East-West yield delta, it's obvious that this would be so.
Just as my OP was able to come within a few points of estimating east-west yield ratios and Ukraine's share of production, likewise I think we can glean some top-line projections about the impact of losing the entire western region:
- Between '40 and '42, the east-west ratio of sown grain land moved from 30:70 to nearly 50:50.
- Were the SU to retreat to/past the Urals, that ratio would be 0:100 - i.e. all eastern.
- Had the 1942 east-west yield ratio been equal to 1940's ~77:100, a shift of sown-land ratio from 50:50 to 0:100 would cause a 12% decline in grain yield per hectare.
...the foregoing assumes it's possible for the SU to increase sown land in the east more than OTL - something that is likely but, as Kotkoff states, only "with some improvement." (p.208).
Could the SU survive a 12% decline in grain per capita? Certainly not. As Hunger and War
discusses, the SU's population was running a caloric deficit in '42-'43 (perhaps '44 as well) and the accumulation of that deficit is what caused widespread starvation and/or malnutrition-related illnesses like tuberculosis. A 12% decline in per capita grain supplies would drastically accelerate caloric-deficit accumulation, sending millions to their deaths.
The only route to survival would be to sow even more eastern land, which would require more labor. The simple fact that such land was not sown, in preference to other eastern land, shows that it was less productive and/or more remote. In either case, cultivating such land would require more labor per grain output (either due to more plowing/sowing for less yield or more building of improvements before cultivation).
Say the Soviets needed 14% more agricultural labor (1/.88) to compensate for a 12% decline in land yield.
In 1942 the SU's labor force (non-military) of 41mil was ~60% agricultural (See Harrison's Accounting for War
A 14% increase moves Ag's labor share to ~68%, which is a ~20% decline in the economy's non-Ag labor share.
The '42 Soviet workforce was stripped to the absolute bare-minimum of civilian-oriented production. Even so, essential civilian services (government, transport, health, etc.) occupied at least 20% of the labor force. Assuming that's a floor for civilian-oriented production, the extra agricultural workers - 8% of labor force - have to come from the direct industry-weapons production complex's ~20% of the labor force. That 8% labor shift means:
40% reduction of the per-capita industry-weapons complex's labor share.
Note again that this is a 40% reduction in "pie-slice," independent of reducing "pie size."
If the SU retreats behind the Urals it's unlikely to have even half of its OTL 1942 population, but even with half the military's productive labor slice would be only 30% of its OTL slice ( 0.6 * 0.5 ).
Hopefully it's clear that a Red Army with 30% of OTL resources is a different ballgame.
I'd anticipate an intuitive response from many: how can a 14% increase in agricultural labor demand cause a 40% decrease in munitions production?
My first answer is in the numbers above but I'd also challenge folks to reimagine the context of the wartime SU. A society that still employs over half of its population in agriculture is so far removed from our own that it's difficult for intuition to track. This was true even for contemporary Westerners such as the German soldiers who were shocked by the conditions of the Soviet peasantry (e.g. dirt-floored wood/thatch huts, lack of candles/oil for lighting let alone electricity). As such a society ekes out its industrial/services surplus from the agricultural base, any change in agricultural productivity has effects far greater than in our times or in the contemporary West.
Czarist Russia collapsed in WW1 because its agrarian economy could not withstand the stresses of modern war.
Stalinist Russia layered more industry atop the old economy but it's an error, IMO, to exaggerate the distance between 1941 and 1917.