Counterfactual history

Counterfactual history is the process of attempting to demonstrate a historical development that did not, in fact, occur. While in some respects "counterfactual history" is oxymoronic, there have been some important works in the field that have resolved some historical problems. The most famous of these problems was by Robert William Fogel in his book Railroads and American Economic Growth. The book attempted to demonstrate that railroads were not indispensable to American economic growth as had generally been assumed by showing that the same level of economic activity could have been achieved through other means of transportation.

However, a major criticism of counterfactual history is the complete lack of historical evidence supporting counterfactual conclusions. Since history did not occur in the manner that a counterfactual argument would suggest, there is simply no historical evidence to support it. Counterfactual arguments therefore must rely upon analogy to carry their points.

The identification problem
No objective assessment of the effects of any historical phenomenon is logically possible without a solution to the identification problem that arises from "the unobservability of counterfactual outcomes." . In counterfactual argumentation, most of the available assessments embody subjective judgments about what would have happened had the historical event not occurred.

For example, a major problem among historians of the New Deal is just how significant was the New Deal: would the economy have recovered as quickly if the New Deal programs had not been in place? At one extreme are those who judge that, without the New Deal, there would have been a rapid economic recovery, and at the other extreme are those who consider the counterfactual outcome to be a total breakdown of the economy. Such differences are clearly brought out in the use of rival versions of the unemployment outcome. The 14.3% version of the 1937 unemployment level that is usually quoted by members of the former group counts those working on New Deal projects among the unemployed; on the assumption that all were engaged on "boondoggles" that did not merit classification as employment, whereas the 9.2% version favored by the latter group excludes New Deal employees from the unemployment figures on the assumption that all were engaged on fully productive activities. .