Mill declares that the method a priori is the true method of the science,and that "it has been so understood and taught by allits most distinguished teachers,"he yet himself in the treatment of production followed an inductive method (or at least oneessentially different from the deductive),obtaining his results by "merely analysing and systematising our common empiricalknowledge of the facts of industry."To explain this characteristic inconsistency,Mr.Sidgwick suggests that Mill,in makinghis general statement as to method,had in contemplation only the statics of distribution and exchange.And in this latter fieldMr.Sidgwick holds that the a priori method,if it be pursued with caution,if the simplified premises be well devised and theconclusions "modified by a rough conjectural allowance"for the elements omitted in the premises,is not,for the case of adeveloped industrial society,"essentially false or misleading."Its conclusions are hypothetically valid,though "its utility as ameans of interpreting and explaining concrete facts depends on its being used with as full a knowledge as possible of theresults of observation and induction."We do not think this statement need be objected to,though we should prefer to regarddeduction from hypothesis as a useful occasional logical artifice,and,as such,perfectly legitimate in this as in other fields ofinquiry,rather than as the main form of method in any department of economics.Mr.Sidgwick,by his limitation ofdeduction in distributional questions to "a state of things taken as the type to which civilized society generallyapproximates,"seems to agree with Bagehot that for times and places which do not correspond to this type the historicalmethod must be used a method which,be it observed,does not exclude,but positively implies,"reflective analysis "of thefacts,and their interpretation from "the motives of human agents"as well as from other determining conditions.In thedynamical study of wealthof the changes in its distribution no less than its productionMr.Sidgwick admits that the method a priori "can occupy but a very subordinate place."We should say that here also,though to a less extent,as a logical artificeit may sometimes be useful,though the hypotheses assumed ought not to be the same that are adapted to a mature industrialstage.But the essential organ must be the historical method,studying comparatively the different phases of social evolution.
Connected with the theory of modern industry is one subject which Bagehot treated,though only in an incidental way,muchmore satisfactorily than his predecessors,namely,the function of the entrepreneur,who in Mill and Cairnes is scarcelyrecognized except as the owner of capital.It is quite singular how little,in the Leading Principles of the latter,his activeco-operation is taken into account.Bagehot objects to the phrase "wages of superintendence,"commonly used to expresshis "reward,"as suggesting altogether erroneous ideas of the nature of his work,and well describes the large and variedrange of his activity and usefulness,and the rare combination of gifts and acquirements which go to make up the perfectionof his equipment.It can scarcely be doubted that a foregone conclusion in favour of the system of (so-called)co-operationhas sometimes led economists to keep these important considerations in the background.They have been brought into dueprominence of late in the treatises of Profs.Marshall and F.A.Walker,who,however,have scarcely made clear,andcertainly have not justified,the principle on which the amount of the remuneration of the entrepreneur is determined.
