SECTIONAL TRANSACTIONS.— F. 455 



tion of the migrants. Using three different types of material available at 

 the Oxford Employment Exchange, the immigration into Oxford from 

 various counties of Great Britain was analysed. It was then possible to 

 calculate mobility indices, with respect to Oxford, for various counties (for 

 men and women separately). 



These indices were found to be closely correlated with distance and to 

 vary more than proportionately with distance : the ' elasticity of migration 

 on distance ' was found, with various data, always to lie between i • 6 and 

 2'0. The residual variations in mobility, i.e. those which could not be 

 accounted for by distance, were then examined in the light of the industrial 

 structure of the counties of origin. It was found, for instance, that the 

 Oxford motor industry seems to present little attraction to textile workers 

 but relatively strong attraction to insured persons working in agricultural 

 districts (this does not include agricultural workers, for whom no figures 

 were obtainable). Further, mobility varies according to the type of 

 unemployment in the district of origin : the chronically unemployed appear 

 to be less mobile than the short-period unemploved. 



The working hypotheses underlying the mobility measures used should 

 be tried out with more general material. It might then be possible (i) to 

 explain the actual movements of labour between any two counties (or 

 industries), and (2) to measure changes in mobility due to causes other than 

 distance and industrial structure (e.g. administrative measures of training 

 and transfer). 



Mr. R. C. Tress. — Local diversity of industry and the rate of unemployment 

 (12.0). 



An area in which a large percentage of workers are occupied in a single 

 industry is liable to be faced at some time or other with a heavy unemploy- 

 ment problem. The occasion will depend upon the industry, but certain 

 factors involved in its concentration are likely to intensify the problem 

 whenever it should occur. An accurate policy of diversification would 

 need to consider the industrial structure of a locality as an organic whole, 

 as a series of relationships between different types of industry, viewed in 

 terms of employment prospects. Determination of the optimum size of 

 such an area would also consider the economies of localisation and of the 

 scales of plants. 



For certain purposes a cruder method is permissible. That suggested 

 defines optimum diversity in terms of an even distribution of the occupied 

 population amongst a number of arbitrarily defined industries, the relation- 

 ships between which, as distinct from their size, are ignored. It is here 

 used to examine recent changes in the industrial structure of a number of 

 towns where there was heavy unemployment in 1931, and where a high 

 percentage of the population were occupied in one or two industries. 



Afternoon. 

 Sir William Beveridge, K.C.B. — Unemployment in relation to the trade 

 cycle (2.15). 

 The paper gives the results of an analysis of the Unemployment percentages 

 recorded for about 100 insured industries from 1927-37, with a view to 

 comparing the position of the different industries in respect of the pro- 

 portionate increase of unemployment in the depression and the times at 

 which downward and upward movements took place. 



Broad comparisons are made between industries of different types 



