F.— ECONOMIC SCIENCE AND STATISTICS. 123 



difficulties which prevent large numbers of firms from installing improve- 

 ments, which they would like to install, all obscure the issue. Without, 

 however, attempting to decide the larger question, we may note two 

 things. 



The first is that the contemporary experience of America shows that 

 collective bargaining is not a necessary condition or the only means of 

 stimulating an increase in efficiency. Since the pre-war period, while 

 British industry with its new equipment of universal collective bargaining 

 has at most increased output sufficiently to compensate for the reduction 

 in normal hours, manufacturing output per head in America increased 

 in the ratio of 105 to 147 between 1919 and 1925. In the same period 

 the extent and strength of trade unionism in America declined, except 

 in certain industries, in which the unions have departed from the old 

 policy of leaving the employers to find ways and means of meeting their 

 claims, and have assumed a direct responsibility on behalf of their members 

 for reducing labour-costs as a condition of maintaining or increasing 

 wage-rates. 



The second note is this. In so far as the extension of collective 

 bargaining does stimulate or compel economy in labour — and, if it has 

 not done so on any large scale at present, it may do so in the near future 

 — it may maintain wages at the expense of increasing unemployment. 

 In the great export industries of coal and cotton, for example, demand 

 for British production appears to be inelastic, and considerable reductions 

 in cost have not resulted in any substantial increase in employment. 

 Moreover, much of the employment at present given is given at a loss. 

 A reorganisation that made it possible to maintain present wage-rates 

 without loss would probably, therefore, involve a reduction in the numbers 

 to whom employment could be given. Such an extrusion of unwanted 

 labour, as a result of improvements in the technical processes or organisa- 

 tion of industry, is a normal incident of economic progress ; and the 

 hardship it may involve need be only temporary, provided that the 

 expansion of industry as a whole is great enough and rapid enough to 

 absorb the extruded labour. When all the industries of the country 

 adopt collective bargaining, and all begin to adopt the policy of 

 holding up wage-rates, leaving it to the employers to tune up industry 

 to the pitch at which such rates can be paid, the numbers of extruded 

 workers for whom the new and expanding industries have to find employ- 

 ment is likely to be increased, and the rate of expansion of industry as 

 a whole becomes a factor of much wider and more pressing interest in 

 wage-negotiations than before. 



III. 



Mr. and Mrs. Webb in their rationalisation of trade union policy 

 distinguished sharply, as we have seen, between the effects of a policy 

 of restriction and those of the policy of regulation. The benefits of trade 

 union organisation to society at large, as distinct from the sectional 

 interest of the members, accrued only if union policy eschewed restrictive 

 practices, and, by concentrating on maintaining and advancing wages 

 and conditions, brought about an improvement in industrial efficiency. 



It is obvious that an advance in wages secured by any one class of 



