With reference to the character of the offenses it was found that the working children inclined to the more serious kinds. Recidivists were found to be far more numerous among the workers than among the non-workers. Summing up the results of the discussion to this point the report says: "It is found that the working children contribute to the ranks of delinquency a slightly larger number and a much larger proportion than do the non-workers, that this excess appears in offenses of every kind, whether trivial or serious, and among recidivists even more markedly than among first offenders."[113] With reference to the connection between recidivism and street work the report says: "The proportion of recidivism is also large among those who are working while attending school, and the numbers here are very much larger than one would wish to see. Some part of the recidivism here is undoubtedly due to the kind of occupations which a child can carry on while attending school. Selling newspapers and blacking shoes, acting as errand or delivery boy, peddling and working about amusement resorts account for over two-thirds of these boys (478 of the 664 are in one or another of these pursuits). These are all occupations in which the chances of going wrong are numerous, involving as they usually do night work, irregular hours, dubious or actively harmful associations and frequent temptations to dishonesty. In addition, something may perhaps be attributed to the overstrain due to the attempt to combine school and work. When a child of 13, a bootblack, is 'often on the street to 12 P.M.,' or when a boy one year older works six hours daily outside of school time, 'often at night,' as a telegraph messenger, it is evident that his school work is not the only thing which is likely to suffer from the excessive strain upon the immature strength, and from the character of his occupation."[114] While reflecting on the excess of working children among the delinquents, one may be inclined to attribute this to bad home influences; but the report shows that only one-fifth of the workers as opposed to nearly one-third of the non-workers come from distinctly bad homes, while from fair and good homes the proportion is approximately 76 per cent to 65 per cent. Consequently, the working child goes wrong more frequently than the non-working child in spite of his more favorable home surroundings.[115]