« PreviousContinue »
THE STATISTICAL STUDY OF CAUSES OF
By Gustav KLEENE, PROFESSOR OF ECONOMICS, TRINITY COLLEGE.
In regard to the causes of poverty, and of the degree of poverty that may be designated as destitution and leads to dependency, there are found in the literature of the social sciences a few formal discussions, but more frequently only remarks made in passing, or more or less obvious implications. The usual manner of treatment is descriptive rather than analytic, and is confined to particular causes and conditions. Professor Warner, one of the few writers who have aimed at a comprehensive treatment of the subject, points out that three tolerably distinct methods of investigation have been tried. First, there is the method of writers like Malthus, Marx, and Henry George, “who from the well-known facts of social organization have sought to deduce the causes tending to poverty." Secondly, there is the study of "the classes not yet pauperized to determine by induction what forces are tending to crowd individuals downward across the pauper line. . . . The best example of such work is probably that of Mr. Charles Booth in his 'Labor and Life of the People.' Almost all of the reports of our labor statisticians, the works on occupational mortality and morbidity, and in fact everything of a descriptive nature that has been written about modern industrial society, can be used in this second method of seeking for the causes of poverty.” Thirdly, there is the .“inductive study of concrete masses of pauperism, usually separating the mass into its individual units, seeking to ascertain in a large number of particular cases what causes have operated to bring about destitution."
* American Charities, pp. 22-117. See also articles by Professor Warner in Volumes I and IV of the Publications of the American Statistical Association.
It is proposed in the following to examine the third method only. This method, that of case-counting, has been applied by Warner to the records of American charity organization societies and by Charles Booth to English paupers. The National Conference of Charities some years ago appointed a committee to prepare a statistical blank for the use of charity organization societies in collecting data for studies of this kind. Mr. Booth sought for each case the “principal or obvious” and one “contributing" cause, and has tabulated his data so as to show the number of cases of pauperism attributable to each cause as principal, and also the number of cases attributable to each combination of principal and contributory cause. The following is taken from his tabulation of almshouse paupers at Stepney:
Professor Warner gives only the number and percentage of cases attributable to each cause as principal (see Table . IV, American Charities). He admits that this is unsatisfactory because there are few cases in which destitution has resulted from a single cause, and for many cases “to pick out one cause and call it the most important is a purely arbitrary proceeding.” He proposed at one of the meetings of the Na
tional Conference of Charities "to consider the influences resulting in destitution in each case as making up ten units, and indicate the relative force of each cause by a proportionate number of units.” The method was rejected as too complicated, but has been used by Mr. A. M. Simons in a study of cases treated by the Chicago Bureau of Associated Charities.* The tabulated results are as follows:
In Professor Lindsay's report to the National Conferencet the method employed by Warner is used. A condensed statement is here given:I
A. Causes indicating misconduct:
Drink, 5–23; immorality, none; shiftlessness and inefficiency,
4.93–14; crime and dishonesty, 0-1.5; roving disposition,
0-3.26. B. Causes indicating misfortune:
a. Lack of normal support.
Imprisonment of breadwinner, 0–2; orphans and aban
doned children, 0-1.5; neglect by relatives, 0–2.3; no male support, 3.6–7.22.
* American Journal of Sociology, vol. iii.
| Report National Conference of Charities, 1899, p. 369.
The numbers in this statement are percentages of cases. The range from the lowest percentage in any city for one year to the highest is given. Thus the lowest percentage assigned to drink is five for the year 1894–95 in Baltimore, the highest is twenty-three per centum for the year 1889–90, and also for 1892-93 in Boston. The cities studied are: New York, 1889–98; Boston, 1889–93; and Baltimore, 1889–95.