The number of married and widowed women has continued to increase substantially since 1950, but the number of single women has remained about the same, according to results of the Current Population Survey of the Bureau of the Census. In the five years between March 1950 and April 1955, the number of women 14 years old and over increased by 3.6 million, or approximately 700,000 per year. The inamong married women amounted to about 550,000 per year; the increase among widows accounted for about 150,000. Changes in the numbers of single and divorced women were small and tended to balance each other.

crease

This pattern of changes differed in several respects from that of the 3-year period between April 1947 and March 1950, when the number of marriages was very large. During this period the number of women 14 years old and over increased by 1.8 million, or about 625,000 per year. Married women, however, increased by 775,000 per year, and there was an increase of about 150,000 widows annually. But the number of single women decreased by 350,000 per year. There was a very small increase in the number of divorced women.

Of the 116.2 million persons 14 years old and over in the civilian population in April 1955, 24.5 million, or 21 percent, were single; 75.1 million, or 65 percent, were married persons living with their spouse; 4.3 million, or 4 percent, were married persons living apart from their spouse; 10.0 million, or 9 percent, were widowed; and 2.4 million, or 2 percent, were divorced at the time of the survey (table 1).

The increase since 1950 in the number of

women "married, husband absent" has arisen in large part from the fact that the number of men in the Armed Forces has doubled during this period and that many of the men in the Armed Forces are married and stationed away from home. One of the subgroups of persons classified as "married, spouse absent" is "separated persons." It is possible that the introduction of the category "Separated" into the Current Population Survey on a national basis in April 1951 may have affected the trend since 1950 in the total number of men and women reported as married but living apart from their spouse.

Differences by residence and color.. According to the most recent survey, the ruralnonfarm population included the smallest proportion of single persons. In 1955, about 22.3 percent of the men 14 years old and over and 15.2 percent of the women 14 years old and over living in rural-nonfarm areas of the United States were single. The proportion single was highest among persons on rural farms--31.5 for men and percent 19.9 percent for women. The proportion single in urban areas was intermediate--23.2 percent and for men and women,

18.8 percent tively (table 2).

respec

For the country as a whole, about 13 percent of the women 14 years old and over, but only 4 percent of the men, were widowed. The proportion widowed was about the same for men in urban and rural areas but varied widely for women. In urban areas, there were about 3 widows for each widower; in rural-nonfarm areas, the ratio was 3 to 1; and in rural-farm areas, the ratio was 2 to 1. This situation arose in part, no doubt, because many farm women move to nonfarm areas when their husbands die.

Only about one-half of the nonwhite women 14 years old and over were married and living with their husbands in April 1955, as compared with about two-thirds of the white women (table 3). The proportion of nonwhite women who were married but living apart from their husbands was especially large, 13.6 percent for nonwhite women as compared with 3.6 percent for white women. Persons with marriages broken by widowhood or divorce were also somewhat more numerous, proportionately, among

The figures shown in table A are approximations based on census and survey data on marital status and age, supplemented by in

formation for men in the Armed Forces obtained from the Department of Defense. Census data by single years of age were used for 1890 to 1940. Sample survey data by 5-year age groups, reduced to comparable intervals by interpolation methods, were used for 1947 to 1955. For 1940, the corresponding medians using 5-year age groupings are 24.8 years for men and 21.5 years for women. (For a discussion of the method used in estimating the medians, see the section below on "Definitions and explanations.") Data on age at first marriage for women based on the 1954 Current Population Survey which included questions on age at first marriage will be presented in the forthcoming Current Population Report on households and families for 1954 and 1955.

100 are heads of households. Starting with about age 65, there is a drop; but even after age 75, three-fourths of the men are heads of households.

head of a

A woman's chances of becoming household or wife of a head are also very closely associated with her age. Thus, about 1 out of every 10 women between 14 and 19 years of age are heads or wives of heads of households, and by the age of 30 to 34 years the ratio is about 9 out of every 10. The proportion of women who are heads or wives of heads remains at a high level until about age 60 when 7 out of 10 are in one of these oategories. After age 75 only about 6 out of 10 are heads or wives of heads of households.

Close to 90 percent of the men who are heads of households are married and with their wives. Women are, of course, much more often wives of heads than heads until the older ages when the death of the spouse marks a change in status from a husband-wife household to some other type. Usually, the wife is the survivor but in many cases she does not maintain a household. Among women 75 years in the

old and over,

about 40 percent live home of others or in an institution.

Detailed age distributions of heads of various types of family units, and of wives of heads, are presented in tables 4, 5, 6, and 7. Some of the family types contain too few cases for separate age analyses but are presented in order to permit users to combine the data into a variety of larger and more reliable groupings.

Family status varies with residence.--Of the 41.9 million families in the United States

in April 1955, two-thirds were located in urban areas, one-fifth in rural-nonfarm areas, and one-eighth on rural farms (table C). A

of the women who (about three-fourths)

relatively large proportion were heads of families were found in urban areas.

About 7 out of every 8 families in the United States as a whole included the head and his wife. A somewhat larger proportion of families living in rural areas than in urban areas were husband-wife families.

Household size and family size cease to decline.--The average size of household has not changed much since 1950, whereas it dropped relatively sharply during the decade of the 1940's (table D). The average size of household was estimated at 3.34 persons in both 1955 and 1954, 3.37 in 1950, and 3.67 in 1940. The relative stability of the average size of household since 1950 can be ascribed to the offsetting effect of several factors that tended to depress household size, including the large expansion in size of the Armed Forces and the sharp decline in the rate of doubling among married couples, and factors that tended to increase household size, in particular the sustained high level of the birth rate.

The trend of the average number of children under 18 years old per family may not follow closely the pattern of concurrent changes in fertility or birth rates. The average number of dependent-age children per family at any time is affected by the fact that large

numbers of families were either formed too recently to have had all their children or were formed so long ago that some or all of the children have grown up and left their parental homes. For this reason, changes in the average number of children per family do not necessarily reflect recent birth experience alone. The number of families may increase at a faster rate than the number of births (as was the case during the 1940's), in which case it is possible to have no change or even a decrease in number of children per family at the same time that the number of births is climbing to record levels.

RELATED REPORTS

Figures for 1955 and earlier years on several types of family units, including the number of married couples, were issued in Series P-20, No. 59, "Households and Families, by Type: 1947 to 1955." Statistics for years prior to 1955 on subjects covered in this report have been published by the Bureau of the Census in the following reports in Series P-20: No. 56 (1954), No. 50 (1953), No. 44 (1952), No. 38 (1951), No. 33 (1950), No. 26 (1949), Nos. 21 and 23 (1948), No. 10 (1947), and in Series P-S, No. 10 (1946), and No. 1 (1944). Additional statistics on marital status of workers for most of the above-mentioned years have been published in reports in Series P-50. Statistics for 1940 and earlier years on marital status are presented in Volume IV of the 1940 Reports on Population. Statistics on marital status are presented in Volume II of the 1950 Census of Population, for standard metropolitan areas, cities, States, regions, and the United States. Detailed statistics on marital status, family status, income, and age based on a tabulation of a sample of the 1950 Census returns are presented in a special report entitled "Marital Status."

DEFINITIONS AND EXPLANATIONS

Coverage.--The figures for 1955 shown in this report relate primarily to the civilian population. Approximately 800,000 menbers of the Armed Forces living off post or with their families on post in 1955 are included, but all other members of the Armed Forces are excluded. For convenience, the population covered is referred to as the ci

and reliability of the estimates."

vilian population.