At places in the foregoing discussion, I have added the current-dollar values of certain variables for various war years, notwithstanding the fact that the purchasing power of the dollar was falling throughout the war. For the purposes I was trying to serve, not much harm was done to my argument by those summations, but for other purposes, such as the determination of the net result of the government’s wartime investment program, the use of constant-dollar values will not suffice, and we must attend to the task of deflation. Unfortunately, wartime price indexes in general are unreliable, if indeed they have any validity at all (Higgs 1992, 49-52).

Matters are even worse in relation to the deflation of durable munitions, which, as we have seen, accounted for the great bulk of the government’s wartime investment, according to the Commerce Department. All economic statisticians seem to have recognized the essential futility of constructing a reliable price index for munitions output during World War II. Having made an attempt to create such an index in his 1945 monograph National Product in Wartime, Simon Kuznets later decided to abandon the effort in his 1961 treatise Capital in the American Economy, remarking: “with the inclusion of additional war and nonwar years, it became exceedingly difficult to adjust the cost of military construction and munitions to levels comparable with normal, peacetime output. Instead, it seemed best to accept the price adjustment used in the Department of Commerce national income accounts” (471). Unfortunately, the Commerce Department accountants themselves had already admitted they could not do the job. In 1954 they had confessed that their “method of deflating munitions expenditures” has “severe limitations,” in large part because “the price information relating to munitions is deficient, largely owing to the fact that there are insurmountable obstacles to the compilation of adequate time series on prices (or quantities) in this area which is characterized by extreme product change” (U.S. Department of Commerce 1954, 157). Therefore, the great bulk of the government’s official capital formation during during the war—its expenditures for durable munitions—cannot be deflated reliably. Because “insurmountable obstacles” cannot be surmounted, we need go no further down this path, except to emphasize the permanent dark cloud that hovers over all data that purport to represent in any way the real value of munitions outputs or stocks over time, especially for periods that include World War II.

Matters may not be so desperate with respect to the deflation of the government’s wartime expenditures for industrial plants and equipment, in particular for the GOPO stock created during the war and, in part, sold to private owners during the immediate postwar years. Table 3 presents the most significant “real” data (expressed here in constant 1958 dollars), those pertaining to the private nonfarm business economy. In the Commerce Department report from which these data are drawn, four different estimates are given, based on two different deflators and two different depreciation formulas, but the four series all show essentially the same profile, and one variant will serve our purposes here well enough.

Table 3: Net Fixed Business Capital (billions of 1958 dollars)

Source: Robert C. Wasson, John C. Musgrave, and Claudia Harkins, “Alternative Estimates of Fixed Business Capital in the United States, 1925-1968,” Survey of Current Business 50 (April 1970): 23, 36.
Note: Variant calculated by using straight-line depreciation and “constant cost 1.” In these data, used assets acquired by business from government are valued at sales prices, not at the government’s original cost of production.

As the table shows, “real” privately owned capital fell from 1941 to 1944, rose slightly in 1945, then grew rapidly in each of the next five years. “Real” GOPO capital mushroomed from nearly zero in 1940 to a peak of $50.2 billion (in 1958 dollars) in 1945, then began to decline rapidly, losing more than 60 percent of its value by 1949. The total privately operated stock of capital, shown in column 3 of the table, grew steadily from 1940 to 1945, fell slightly in 1946—a decline attributable entirely to the estimated decline of nearly 30 percent in the real value of GOPO capital, a substantial part of that decline representing the transfer of assets to private owners—then grew steadily for the remainder of the decade. From 1940 to 1945, the estimated total stock increased by 20 percent; then, from 1945 to 1950, by 16 percent.

It is almost certain that the Commerce Department figures overstate the increase in the privately operated capital stock during the first half of the 1940s, because the standard formulas for computing depreciation fail to take into account certain extraordinary conditions during the war, especially “the accelerated depreciation resulting from intensive plant use and scarcity of replacement parts” (McLaughlin 1943, 113). According to a War Production Board report,

plant utilization in the munitions industries increased sharply after Pearl Harbor . . . . [T]he average utilization of facilities in the metal products industries late in 1944 was about two-thirds above the prewar level, after having reached nearly twice the prewar level in the spring of 1943; the increase in the remaining industries, though smaller, was still substantial. . . . [T]he increased utilization of existing facilities contributed nearly as much to the increase of total industrial output during the war as did the construction of new facilities; though the contribution made by more intensive utilization was much more important in the earlier part of the period, particularly in 1940 and 1941, than it was in 1943 and 1944. (U.S. War Production Board 1945, 7, emphasis added)

Double-shift and even treble-shift operation of plants became much more common during the war (U.S. War Production Board 1945, 31-32).

Recently, Lee Ohanian (1997) also has noted that “capital utilization increased substantially during the war” (33, citing Foss 1963). Foss had reported, among other things, that in a comparison of conditions in 1934-39 and those in 1940-44, hours of usage per year increased by 53 percent for railroad freight cars, by 54 percent for freight locomotives, and by 34 percent for passenger locomotives (1963, 15). Hours of usage per year per spindle in the cotton textile industry increased by about two-thirds during the war years (1963, 9).

At the same time that the capital stock was being used far more intensively, the lack of replacement parts and repair materials kept producers from doing the normal upkeep on their property. In the housing sector, rent controls induced landlords to forgo ordinary repairs. For apartment houses and small structures, index numbers of repair and maintenance expenditures fell some twenty percent during the war (Rockoff 1984, 156). The motor-carrier industry during the three peak war years obtained “only 195,000 new trucks and buses,” which amounted to “less than 10 percent of the number it would normally require for replacements and expansions,” while increasing its loads each year and hastening wear and tear on its fleet of vehicles (Director of War Mobilization and Reconversion 1945, 39). At the beginning of 1945, the Director of War Mobilization and Reconversion confirmed that “wear and tear on [industrial] plants has been far above normal, while repairs and replacements have been below normal” (1945, 50).

We can gain a rough idea of the effect this accelerated wartime depreciation by making some conservative adjustments to the Commerce Department’s capital consumption allowances. From the gross and net investment figures shown in Table 1, we can infer that the official allowances for depreciation of the private capital stock during the five wartime production years were as follows, in current dollars: $9.0 billion in 1941, $10.1 billion in 1942, $10.1 billion in 1943, $10.4 billion in 1944, and $10.7 billion in 1945. If we adjust these estimates by adding just 10 percent in 1941, and 20 percent in each year from 1942 through 1945, the total additional depreciation for the entire wartime period comes to $9.1 billion, a far from trivial sum—consider, for example, that it offsets approximately 53 percent of the amount ($17.2 billion) the government spent on new industrial facilities (structures, equipment, and reconversions) from July 1940 through June 1945 (Smaller War Plants Corporation 1946, 37-38). In addition, we may presume that the official depreciation formulas also failed to capture fully the actual wartime depreciation of government-financed industrial facilities, which were also operated at extraordinary intensity, probably more so than the private capital stock as a whole. Given that more than half of the government’s industrial investment had been completed by the end of 1942 (ibid., 37), then an allowance for extra depreciation of the GOPO industrial capital might easily add another $1 billion to the total adjustment for wartime understatement of actual depreciation, bringing that total to more than $10 billion in current dollars, or approximately $20 billion in 1958 dollars (adjustment made by using the implicit deflator for private nonresidential fixed investment, from U.S. Council of Economic Advisers 1970, 180).

If we deduct just half of this amount— $10 billion in 1958 dollars—from the total net fixed business capital estimate shown in table 3 for 1945 (which represents only part of the nation’s total private capital stock), the total falls to $222.0 billion. One implication is that the total net fixed business capital stock increased between 1940 and 1945 not by the 20 percent implied by the official figures, but by just 15 percent. And given the new figure for 1945 (and a corresponding adjustment for the data for later years), another implication is that the total increased between 1945 and 1950 not by the 16 percent implied by the official figures, but by 17 percent. In my judgment, a correct adjustment—arriving at which would require an enormous research effort, if it is possible at all, given the sorts of data that would be required to carry it out—almost certainly would be greater than my crude adjustment here.

Thus, taking into account the undoubted measurement errors in the wartime depreciation allowed in the Commerce Department figures shown in table 3, caused by the inappropriate application of uniform, standard depreciation schedules, we may conclude that the actual drop in the privately owned net stock of capital was even greater than shown during the war years and that the increase in that stock during the second half of the 1940s was greater than shown. Thus, for the capital stock, as for the economy’s real output (Higgs 1992, 1997, 1999), the official data have misled us by making the wartime expansion appear bigger than it really was and the postwar expansion smaller than it really was.