Illustration: Chris Sharp

Clifford G. Holderness is a professor of finance at the Carroll School of Management, and his particular interest is the behavior of large shareholders in public corporations. He is also a World War II buff.

Seven years ago, he was browsing through the National Archives’ online World War II Prisoners of War Data File. As Holderness explained recently in an interview, the usual scholarly method of inquiry is to shape questions first, and then seek data for illumination. Joined by Jeffrey Pontiff, holder of the James F. Cleary Chair in Finance, Clifford took the opposite tack. With the POW data in front of them (augmented by the Archives’ World War II Army Enlistment Records), the two settled on a question: “Is a hierarchy that is optimal in one environment [the battlefield] still optimal in a related but different environment [a prison camp]?”

In an article forthcoming in the journal Management Science—”Hierarchies and the Survival of Prisoners of War during World War II”—Holderness and Pontiff examine the fate of approximately 123,000 Americans held by the Axis powers in 280 camps. One out of 10 prisoners died in these camps overall, most from starvation or disease, some by execution.

In some camps the battlefield chain of command remained intact—with the familiar distribution of captains, colonels, and other officers, along with a relatively high number of enlisted men. In other camps, the chain was broken, with a disproportionate number of brass or of privates and corporals. In facilities where the ranks were disrupted, POWs did not fare worse than their counterparts in well-ordered, hierarchical camps. In fact, they fared better. “Prisoners in the most hierarchical groups were 20 percent less likely to survive than those in the least hierarchical groups,” the professors report. This statistic held fairly constant both for German camps (where the overall death rate was 3 percent) and for Japanese camps (where the death rate was 36 percent). And it held steady when the authors controlled for factors such as how long the POWs had been imprisoned, their age, the climate of their locale, and the severity of the labor forced upon them.

Did survival rates suffer in camps with intact chains of command because the brass provided for themselves at the expense of the men below them? The authors say this scenario is unlikely: In fact, officers survived at a lower rate than enlisted men. Holderness and Pontiff found that survival generally increased with the population of the camps. Borrowing an idea from Adam Smith, the authors speculate that in POW camps, as in markets, “larger cohorts offered greater possibilities for voluntary exchange”—in this instance, the trading of food and creature comforts—provided that trading was allowed to take place. Holderness and Pontiff cite survivor accounts to suggest that, to the detriment of prisoners, “some officers sought to suppress‚” the informal trade in rations, at times out of concern for the POWs’ well-being, and ‚”at times on the grounds that [trading] subverted military protocol.”

In an interview, a survivor of a camp in Japan described for the authors a kind of futures market, whereby a POW who was too sick to eat might trade his ration of rice to a fellow detainee for a future quantity. Another former POW recalled a well-intentioned medical officer’s efforts to control such “sales” an all-out ban on trading is described by another.

Economists going back to the 1930s “have recognized that markets are an alternative to hierarchies for allocating resources,” say Holderness and Pontiff. “It is difficult to imagine markets playing any meaningful role on a battlefield, but there is considerable evidence,” the professors say, “that markets were important in many POW camps during World War II” and that officers trained in centralized command and control “did not adapt well.”

Read more by William Bole