194 REPORTS ON THE STATE OE SCIENCE. 



persons, with an average of 681. ; and a third — a very important com- 

 pilation kindly handed to us by the Financial Secretary of the 

 National Amalgamated Union of Shop Assistants, relating to 3,171 

 persons, whose average cash income is 74Z. In the first and third 

 cases something would have to be added for about 22 per cent, of 

 the assistants in the one group and 16 per cent, in the other who 

 live in, or have meals in, if we were including such payments. Further 

 information is given in ' Industrial Co-operation ' by Miss C. Webb, 

 1904, where the range of earnings of the branch managers and counter- 

 men in the Co-operative Ketail Distributive Societies are given in some 

 detail relating to over 10,000 employes; the data are not sufficient for 

 a perfect calculation of the average, but it cannot be very far from 68L 

 We have these four different estimates relating to a variety of districts 

 and trades, and all being contained in the limits, 67L ± 11. Neverthe- 

 less, we think that this wage is rather higher than the one we ought to 

 adopt for the 900,000 persons in Class 29, since many of these are 

 probably engaged at very low salaries in shops from which no informa- 

 tion was obtained in any of the returns. We therefore prefer to take 

 65l.+lbl. as the average annual wage. We have practically no informa- 

 tion as to the number of shop assistants who receive over 160L, but it is 

 certainly extremely small, and we take it as about one per cent. 



Class 30. Coster mongers. — This class includes all kinds of street 

 sellers and hawkers. % Probably none of them pay income-tax. We 

 may divide them as lads under twenty, average perhaps 201. a year; 

 men of over twenty receiving somewhat more than the wage of unskilled 

 labour, on an average, say, 101., with a wide margin; and women at 

 perhaps 301. This gives an average of about SOL, and we- will take it 

 as between 40L and 701. 



Class 31. Sweeps (employers). — None need be counted as payiug 

 income-tax, and the average may again be taken as between the wages 

 of totally unskilled and of highly skilled labour. We put it at 

 101 ±302. 



In the table on p. 195 the estimates we have now made are tabu- 

 lated in detail. In the columns 1 and 2 are given the number of persons 

 estimated from the census of 1901 raised by 10 per cent, throughout, 

 except in the case of farmers, whose numbers are left unaltered; 

 including fanners, an increase of 8'6 per cent, is allowed, which is about 

 the increase of the whole population up to the end of 1909. Columns 

 3 and 4 give the number paying and not paying tax respectively, and 

 column 5 the margin assigned. Column 6 gives the average income of 

 those not paying, column 7 its margin. Column 8 gives the aggregate 

 of the estimates assumed correct. 



The total of the numbers we have found subject to income-tax is 

 900,000, while, if all the margins are added together and all taken 

 as positive, they amount to 212,000. Compiled, however, by the 

 theory of error, as explained below, the margin suggested is about 

 75,000, and is almost entirely due to that of Classes 16, 25, and 27. If 

 the estimate of 800,000 income-tax payers be correct, we have put 

 rather too many in each of these classes, and they should be transferred 

 to the intermediate class. 



