FECUNDITY OF IMMIGRANT WOMEN.

INTRODUCTORY.

Valuable information regarding the relative fecundity of different foreign races and the native American stock in the United States was collected by the Twelfth Census (1900), in answer to questions on the population schedule asking of every married, widowed, and divorced woman how many children she had borne, and how many of these were still living. Recognizing the great value of the material thus collected, the Immigration Commission obtained permission to use the original data which the census had collected. The sections of the United States selected for study were as follows: The State of Rhode Island, the city of Cleveland and 48 largely rural counties in the State of Ohio, and the city of Minneapolis and 21 largely rural counties in the State of Minnesota. Rhode Island was chosen as being a compact eastern State, with a population largely urban and manufacturing in character, nearly four-fifths of the total population being in cities of at least 10,000 inhabitants. Ohio and Minnesota are middle-western types, Cleveland and Minneapolis presenting urban and manufacturing conditions, and the selected counties in each State rural and agricultural conditions.

The data tabulated for the purpose of this inquiry relate to the number of children borne by married women, that is, women living in the married state at the time the census, was taken. Widowed and divorced women were not included, mainly because the number of years that they had been married could not be ascertained from census returns. In the case of married women this information was available, the census schedule stating the number of years married, or duration of present marriage, up to the time that the census was taken; in the case of widowed and divorced women the number of years married was not stated.

At the same time it was considered inadvisable to include all classes of married women.

Classes of married women omitted.-Women married less than one year were omitted because a large proportion of them more than three-fourths of the total number-had not been married long enough to have borne children.

Women over 45 years of age were omitted. In analyzing and discussing the results it seemed best to consider only those women who were under 45 years of age and whose entire married life, therefore, would fall within the childbearing period.

Every woman was omitted whose parents were born in different countries. The reason for this omission arises from the fact that the main object of the inquiry was to arrive at a comparison as regards fecundity between different nationalities or races, such as the Irish, the Germans, the Scandinavians, and the native Americans. The classification by nationality was based upon the country in which the parents were born. A woman was classified as Irish if both of her parents were born in Ireland, and as American if both parents were born in the United States. Cases where the parents were born in different countries, as father in Italy and mother in Ireland, or father in Germany and mother in the United States, would, of necessity, be classified either with respect to the birthplace of one parent only or with respect to the parentage combination, as Italian-Irish, or German-American. In a tabulation covering the entire United States or a large number of States such a classification might be of some significance; but in a population no larger than that of Rhode Island, or the selected areas of Ohio and Minnesota, the number of cases under each of the many combinations of parentage would be too small to afford any basis for generalizations. It has been thought best, therefore, to omit the cases of mixed parentage altogether and confine the comparison to classes representing what might be termed "pure stock," so far as could be determined by the nativity of the parents.

Prior marriages. The number of children given in the census schedule comprises the total number borne by the woman to whom the return refers. It includes therefore the children born of prior as well as of present marriages. The number of years married, however, represents the duration of the present or existing marriage only. The children that may have been born of a prior marriage can not be distinguished, and are therefore credited to the present marriage, except in cases where the number is too large to have been born during the present marriage. Such cases have been eliminated. For instance, if six children were reported for a woman married two years, it is practically certain that some of these children must have been born during a former marriage or else out of wedlock. It being impossible to ascertain the actual facts, these cases have been left out of the tabulation altogether. There remain, of course, a considerable number of cases in which children born of a prior marriage may have been credited to the present marriage and retained in the tabulation, because the fact of a former marriage was not made apparent either by a disproportion between the number of children. and the duration of the present marriage or by any other evidence. The inclusion of these children of a prior marriage introduces an element of uncertainty or of error, the effect of which is to exaggerate in some degree the fecundity of present marriages by crediting them with the children of prior marriages. How serious this element of error is depends, of course, upon the relative frequency of second marriages. It is obvious that the probability of a prior marriage depends upon. the age at which the present marriage was contracted. As the census returns give the age and number of years married, it is possible in analyzing the results of our tabulation to eliminate the women whose present marriage took place late in life, or after they had reached an age where the probability of a prior marriage is too great

to be ignored. For instance, by considering only women under 45 years of age who had been married at least 15 years comparisons would be restricted to marriages that took place below the age of 30, so that the percentage of prior marriages would be small.

RHODE ISLAND.

The total number of married women in the State of Rhode Island at the census of 1900 was 79,800. The number for whom the data in regard to children were tabulated for the present study was 45,445, the difference between the two figures representing, of course, the classes omitted for the reasons above stated. The following tabulation classifies these 45,445 married women by color, nationality, and place of residence:

TABLE 1.-Women under 45 years of age married more than one year, classified by parentage: Total number tabulated, number in cities of over 10,000 inhabitants, and number in smaller places.

The white women of foreign parentage may be subdivided as regards birthplace into two classes, those born in the United States and those born in foreign countries, the latter, who represent immigrants, being designated as the first generation, and the former, who represent the children of immigrants, being designated as the second generation.

82401°-VOL 2-11-30

TABLE 2.-White women of foreign parentage under 45 years of age married more than one year, classified by parentage and nativity: Total number tabulated, number in first generation (born abroad), and number in second generation (born in United States).

For some nationalities the numbers included in the second generation are so small as to afford no adequate basis for statistical analysis. This of course results from the fact that the immigration of these nationalities is of such recent origin that comparatively few of the children born in this country had reached maturity at the time of the Twelfth Census. The Irish are the only nationality in Rhode Island in which the number of married women in the second generation is more than two-thirds as large as it is in the first. In the case of the Germans it is one-half as large.

The facts regarding number of children, as has been pointed out, must be considered in connection with the duration of the marriage. Theoretically, perhaps, the most satisfactory comparison would be that between women married the same number of years, a result which could be secured by classifying them with respect to the exact number of years married, as married 1 year, 2 years, 3 years, 4 years, and so forth. Our total, however, is not large enough to warrant such detailed treatment, and it becomes necessary to substitute a classification by periods of years.

A study was made first of the data for those women under 45 years of age who had been married from 10 to 20 years. The lower limit of 10 years was adopted with a view to limiting the comparison to women who had been married long enough to have borne at least three or four children, and also with a view to excluding women who married after the age of 35, and for whom therefore a prior marriage was rather probable. The upper limit of 20 years was selected somewhat arbitrarily, but with the idea of not having the limits any further apart than was necessary to include enough women to make the figures significant. In more general terms the women considered may be defined as married women of childbearing age who are in the second decade of their married life.

Proportion bearing no children.-The following table shows the proportion of the group of Rhode Island married women under consideration that have borne no children:

TABLE 3.--Women under 45 years of age married ten to twenty years, classified by parentage and nativity: Total number tabulated, and number and per cent bearing no children.

The table requires little comment or explanation. The proportion of the native white women of native parentage who had borne no children is relatively large, being 17.5 per cent, or more than 1 woman in 6. The proportion of white women of foreign parentage who had borne no children is but 8 per cent, or less than 1 woman in 12. The smallest percentages are those for the French Canadian and Italian women, of whom only 1 in 20, or about 5 per cent, have had no children. The largest percentage is that for native negro women, of whom 22.5 per cent have had no children.

The table brings out the fact that among married white women of foreign parentage the proportion bearing no children is greater in the second generation than in the first, being 10.5 per cent in the second and 7.2 per cent in the first generation. For every foreign nationality shown except the French Canadian, the percentage in the second generation is also greater than in the first. In no instance, however, is the percentage, even in the second generation, as high as it is for the native white of native parentage (17.5 per cent).