In the 1960 Census information on whether the person was married more than once was determined by a direct question for all persons ever married (see below).

The year of first marriage for all ever-married persons was obtained from answers to the following questions on the Household Questionnaire:


This information was then transcribed, to the sample FOSDIC schedule as follows:


If information on these items had not been entered on the Household Questionnaire for an ever-married person, the enumerator was instructed to ask the questions. The enumerator was instructed to find out the date of .marriage for the husband and enter this also for the wife if both the man and wife had been married only once. If either had been married more than once, the enumerator asked for the date of first marriage of each. If a person had been married before 1890, the enumerator marked 1890 and found out in which calendar quarter the marriage occurred. When the year of first marriage was known and the month was not, the enumerator attempted, to determine the season in which marriage took place. When the year was unknown, the enumerator tried to find out, as accurately as possible, how many years ago the person was first married. Using this information, and month of marriage, the enumerator obtained year of marriage by subtraction; where both month and year were unknown, the enumerator obtained year of first marriage from information on the number of years since first marriage, if available, on the assumption that the month of marriage was between April and December.

All persons who first married in the same period may be regarded as members of the same marriage cohort. For information on Uses and Limitations of the Data in the present report for cohorts of persons man years ago, see the section below on "Uses and Limitations of the Data."

The number of years since the person's first marriage was derived by subtracting the date of first marriage from April 1, 1960, and represents the interval in completed years.

In this report, age at first marriage is shown in completed years for all ever-married persons. In tables showing age at marriage by age at census, the data are based on persons who were 14 to 79 years old at the census date. The upper limit was set at 79 years because it is the upper limit of the oldest 5-year age group in which no persons could have been married before 1890, the earliest year for which year at first marriage was recorded on the sample FOSDIC schedule.

In deriving age at first marriage, the year and quarter of the person's birth was subtracted from the year and quarter of the person's first marriage. Where the result of this subtraction included a fraction (e.g., age at marriage 21 ¼, 21 ½, or 2 ¾), the fraction was dropped, so as to express the result in completed years of age at first marriage. Where the result of the subtraction included no fraction, which was the case when the quarter of birth and first marriage were the same, the result was reduced by 1 year in approximately one-half the cases, namely, those in which the quarter was January to March or October to December. This adjustment was not made, however, where the result of the subtraction, was 14 years.

The adjustment was made in order to eliminate the upward bias in the distribution by age at marriage that would otherwise have occurred. For example, the calculated difference between a birth date of April- June 1925 and a marriage date of April-June 1950 is 25 years. However, on the assumption that births and marriages were evenly distributed throughout their respective quarters, only one-half of the persons with these dates were married after their 25th birthday, and one-half were married before their 25th birthday. Omission of the adjustment would have resulted in classifying about one-half of all persons who had the same quarter of birth and marriage in the next higher year of age at marriage than the one in which they were actually married. It is recognized that in age ranges where the marriage rate increases rapidly, as from 14 to 20 years of age, the adjustment resulted in some overstatement of the number of persons marrying at very young ages. However, the General effect of the adjustment was to increase the accuracy of the overall distribution by age at marriage and, especially, the median age at marriage.

For information on Editing and allocation procedures for year of and age at first marriage, see the section below on "Editing."

Average annual first marriages per 1,000 single persons14 to 44 years old are shown, by age, in tables 16 to 18. The numerator of this rate is the average annual number of persons who first married at a given age in 1958 and 1959. The denominator is the average annual number of persons who were of that age and single on July 1, 1958 and 1959. This figure was obtained by averaging the number of persons who were of the given age and single on July 1, 1958, with the corresponding number for July 1, 1959. The number of single persons on July 1, 1958, is the sum of (1) persons who were single at the census date and (2) persons ever married at the census date who first married after July 1, 1958. The number of single- persons on July 1, 1959, is defined in corresponding fashion, except with reference to July 1, 1959, instead of 1958. Persons who were 14 to 44 years old and single on both dates were counted twice-once in the column for their age on July 1, 1958, and once in the column for their age on July 1, 1959.

T he statistics published in reports of the 1940 Census under the Differential Fertility-1940 and 1910 included data on age at first marriage for women 15 to 74 years old for 1940. The published data on this subject were limited to women who were reported, as married once with husband present, because relatively large proportions of the other ever-married women made no report on this subject. In the 1940 Census, women were asked their age at first marriage, whereas in the 1960 Census, age at first marriage was derived from answer to questions on birthdate and date of first marriage. About 4.4 percent of women who were married once with husband present did not respond to the question Census, as compared with the figure of 7.0 percent of ever-married women14 to 79 years old for whom age at first marriage was allocated in the 1960 Census. Allocations for nonresponse were not made for the data on age at first marriage in 1940, hut were made during the Editing of the 1960 Census data. These differences between the 1940 and the 1960 Censuses with respect to the method of obtaining information on age at marriage, the extent of nonresponse, and the method of dealing with nonresponses may have created son of comparability between the resulting statistics.

The statistics on duration of current marital status for women that were published in 1950 Census Population, Volume IV, Part 2, chapter E, Duration of Current Marital Status, included data on age at first marriage for women married once with husband present. For these women, information on number of years since first marriage was obtained from answers to the following question: "How many years since this person was married?" age at first marriage was derived by subtracting the number of completed years since first marriage from the woman's age in completed years. For example, if a woman reported that she had been married for 3 years and that she was 23 years old at the time of the enumeration, she was classified as 20 years at first marriage.

This basis of classification yields results that differ from the data on age at first marriage shown in the present report. If age in completed years at census is 23 years, the range of exact ages is 23.00 to 23.99 years; if the duration of marriage is 3 years, the range of exact duration is 3.00 to 3.99 years. The limits of the range of derived age at marriage are, therefore, 19.01 to 20.99 years and the midpoint for this interval is 20.00 years. Data in the present report were derived from information on calendar quarter and year of birth and marriage, rather than year alone. As a result, age 20 at first marriage in the present report has the more conventional meaning of 20.00 to 20.99 years, and the midpoint is one-half year higher than the midpoint of age 20 at first marriage in the 1950 Census data.

The median ages at first marriage shown in the 1950 Census data take account of the method by which age at first marriage was derived, and may be compared with corresponding medians in the present report. Comparability of the figures for any given age at marriage is impaired, however, for the reason set forth above and because the derivation of age at first marriage from information in completed years tends to produce a flatter distribution than would be produced if more detailed information on birthdate and date of first marriage were available. As indicated above, the women classified in each single year of age at marriage in the 1950 Census data were drawn from women whose true ages at marriage covered a range of 2 years. Consequently, the figures are subject to an averaging effect, whereby the true frequencies in each age at marriage are averaged with the frequencies for adjoining ages. The resulting distribution tends to understate the true frequencies in the most popular ages at marriage, and to overstate the others.

Data in this report are consistent with corresponding data in 1960 Census of Population, PC(2) -4E, Marital Status. There are some small differences between the statistics in these reports and the statistics on age at first marriage, year of first marriage, and years since first marriage in PC(2)-3A, women by number of children ever born, and PC(2)-4A, Families. These differences are the result of a supplementary edit which was performed to correct faulty dates of first marriage for a relatively small number of persons. The supplementary edit was performed prior to tabulation of the data in this report and the report on marital status, but after tabulation of the other two reports.

Table E shows distributions by age at first marriage for white women who first married in 1900 to 1959, from the present report and from the August 1959 Current Population Survey (CPS), the results of which were published in Current Population Reports, Series P-20, No. 108. Some differences between the two sets of figures are to be expected because of differences in dates and coverage between the census and the CPS, and because of the sampling error associated with estimates based on the CPS. The CPS data were collected in August 1959, as compared with the April 1960 date of the census. The CPS survey was limited to the civilian non-institutional population, whereas the census included military personnel and inmates of institutions; and the CPS did not include Alaska and Hawaii, which were included in the census. However, these differences do not account for differences in the figures for two of the groups shown. The census shows significantly more women first married in 1930 to 1939 and significantly fewer women first married in 1900 to 1909 than does the CPS. The latter difference is to be expected because this group has reached an advanced age, at which the effect of mortality during the 8-month period between the census and the survey would be substantial. However, the inclusion of inmates of institutions in the census, and their exclusion from the CPS, would have a compensating effect, tending to make the numbers of women first married in 1900 to 1909 in the census and the CPS more nearly equal than they are.

The explanation for both these differences appears to lie in differences between the age structure of the population enumerated in the 1960 Census and the age structure of the independent estimates used to inflate the sample data from the CPS in 1959. Thus, the census shows more women of an age to have been married for the first time in 1930 to 1939, and fewer women of an age to have been married for the first time in 1900 to 1909, than were shown in the independent population estimates used to inflate the CPS data.

The proportion of women who first married at 25 years of age and over is generally higher in the census data than in the CPS, and the proportion of women who first married under 22 years of age is correspondingly lower. In the figures for women first married before 1930, these differences produce differences of three- or four-tenths of a year in the median age at first marriage. The aforementioned differences between the census and the CPS cannot account for these differences in the figures; nor can the amount of Editing, which was similar. In-the census, about 6 percent of the white women had a date of first marriage that was allocated for nonresponse and inconsistency. In the CPS, an estimated 5 percent of the white women had a blank decade or year of first marriage, and an additional 1.4 percent had a reported date of first marriage that yielded a derived age under 14 years at first marriage prior to Editing.

In the CPS, a manual edit was performed in which women with blank dates of first marriage were assigned a date of marriage in accordance with distributions, based on the 1950 Census, which took into account date of birth of the first child, where available. The relatively small number of white women whose reported date of first marriage yielded a derived age under 14 years in the tabulations were distributed pro rata among ages 14 years and over at first marriage. Differences between those procedures and the 1960 Census procedures described in the section on "Editing," below, might give rise to some differences in the resulting statistics. However, given the relatively small proportion of persons requiring allocation, it is unlikely that differences in editing procedure could produce differences of the size shown for some cohorts in table E.

In the CPS, the question on date of first marriage was asked at the end of the interview, following a series of questions on the dates of birth of the woman's children. This sequence may have led some women to give a marriage date earlier than they might otherwise have given. In the 1960 Census, date of first marriage was asked in the self-administered Household Questionnaire; this did not include the question on birthdates of children, which had been reported previously in the Advance Census Report.

Table F shows comparisons by color for women who first married in 1950 to 1959. Most of the difference between the CPS and the census with respect to the percent of nonwhite women who first married at 14 to 17 years of age is attributable to differences in the handling of ages under 14 years at first marriage. Whereas white women in the CPS with a reported age under 1.5 years at first marriage were distributed among all ages 15 years and over at first marriage, among non- whites the number of women with a reported marriage age under ill was so large that these cases were assumed to reflect chiefly early marriages and, therefore, were combined with data for the group) married at age 14 to 17.

Table G shows comparisons of the average annual number of persons first married at ages 14 to 44 years in 1958 and 1959 for 26 States from table 18 of this report with the corresponding figures based on registered marriages, published in Vital Statistics of the United States for 1958 and 1959. The 26 States are those for which the number of registered first marriages by age and sex was available for the entire State in 1958 and 1959. The differences shown between the census and vital statistics with respect to total number of first marriages are relatively small. Such differences as there are may be attributable to the fact that in the vital statistics, marriages were classified by the State in which they occurred, whereas in the census, persons who first married in 1958 and 1959 were classified by their State of residence in 1960. Thus, differences in the number of marriages from the two sources may appear because some persons married in States other than their State of residence in 1960, and because some persons changed their residence from one State to another between marriage and the date of the enumeration. Furthermore, differences in reporting of date of first marriage or age could result in the inclusion of a person in the census figures and his exclusion from registered marriages, and vice versa.

Table E. Comparison of 1960 Census Data by age at First Marriage with Data from the August 1959 Current Population Survey, For White Women Who First Married In 1900 To 1959, For the United States

Year of first marriage of women	Total white women (thousands)	Percent by age at first marriage							Median age at first marriage (years)
		Total	14 to 17 years	18 and 19 years	20 and 21 years	22 to 24 years	25 to 29 years	30 years and over
1960 Census
1950 to 19591	10,618	100.0	19.0	26.1	20.6	16.5	9.9	7.9	20.5
1940 to 1949	11,987	100.0	14.6	21.1	20.7	20.4	13.9	9.3	21.4
1930 to 1939	8,745	100.0	16.2	19.6	19.3	21.0	15.5	8.6	21.5
1920 to 1929	7,321	100.0	18.4	20.1	18.8	19.6	15.1	8.0	21.2
1910 to 1919	5,185	100.0	18.1	20.6	19.9	20.7	14.8	5.9	21.1
1900 to 1909	2,862	100.0	22.3	23.0	20.6	19.0	11.5	3.6	20.5
1959 CPS
1950 to 19592	9,823	100.0	19.9	25.7	22.0	16.4	9.1	6.8	20.4
1940 to 1949	11,871	100.0	14.3	21.8	20.7	20.5	14.7	8.0	21.3
1930 to 1939	8,406	100.0	16.1	20.9	19.5	20.8	15.8	6.8	21.3
1920 to 1929	7,482	100.0	20.3	21.1	19.6	19.1	13.6	6.2	20.8
1910 to 1919	5,137	100.0	19.8	22.6	20.4	20.0	12.1	5.1	20.7
1900 to 1909	2,925	100.0	22.6	25.3	20.8	19.4	10.4	1.6	20.2
Deviation of census from CPS
1950 to 19592	795	…	-0.9	0.4	-1.4	0.1	0.8	1.1	0.1
1940 to 1949	116	…	0.3	-0.7	…	-0.1	-0.8	1.3	0.1
1930 to 1939	339	…	0.1	-1.3	-0.2	0.2	-0.3	1.8	0.2
1920 to 1929	-161	…	-1.9	-1.0	-0.8	0.5	1.5	1.8	0.4
1910 to 1919	48	…	-1.7	-2.0	-0.5	0.7	2.7	0.8	0.4
1900 to 1909	-243	…	-0.3	-2.3	-0.2	-0.4	1.1	2.0	0.3

¹Census includes women who first married in January to March 1960.
²CPS excludes women who first married in September to December 1959.

Table F. Comparison of 1960 Census Data by age at First Marriage with Data from the August 1959 Current Population Survey, For White and Nonwhite Women Who First Married In 1950 To 1959, For the United States

age at first marriage	White			Nonwhite
	1960 Census	1959 CPS	Deviation of census from CPS	1960 Census	1959 CPS	Deviation of census from CPS
Women first married in 1950 to 19591 (thousands)	10,618	9,823	795	1,457	1,219	238
Percent	100.0	100.0	…	100.0	100.0	…
14 to 17 years	19.0	19.9	-0.9	21.8	25.5	-3.7
18 and 19 years	26.1	25.7	0.4	20.2	21.7	-1.5
20 and 21 years	20.6	22.0	-1.4	16.7	16.7	…
22 to 24 years	16.5	16.4	0.1	16.1	16.1	…
25 to 29 years	9.9	9.1	0.8	13.3	11.1	2.2
30 years and over	7.9	6.8	1.1	11.9	8.9	3.0
Median age (years)	20.5	20.4	0.1	21.0	20.3	0.7

¹ Census includes women who first married in January to March 1960; CPS excludes women who first married in September to December 1959.

The census data show somewhat fewer first marriages at ages 20 to 24 than are shown by the registration data and more first marriages above and below these ages. These differences may arise from the differences noted above or from a variety of other sources, including the approximations required in order to de-rive some ages at first marriage (see section on "age at first marriage," above), and the fact that age at first marriage was allocated for 7 percent of persons in the census data, whereas the registration data included a negligible proportion of persons with a non- response on age at first marriage.

Table G. Comparison of 1960 Census Data on Average Annual First Marriages at ages 14 To 44 Years in 1958 and 1959 with Data Based On Registered Marriages, By age and Sex, For 26 States

age at first marriage and sex	1960 Census	Registered marriages	Deviation of census from registered marriages
Male
Total (thousands)	481	466	15
Percent	100.0	100.0	…
14 to 19 years	19.3	16.6	2.7
20 to 24 years	50.7	54.6	-3.9
25 to 29 years	19.3	18.9	0.4
30 to 34 years	6.5	6.0	0.5
35 to 44 years	4.2	3.8	0.4
Female
Total (thousands)	493	464	29
Percent	100.0	100.0	…
14 to 19 years	50.1	49.8	0.3
20 to 24 years	35.8	37.7	-1.9
25 to 29 years	8.6	7.4	1.2
30 to 34 years	3.0	2.8	0.2
35 to 44 years	2.5	2.5	0.3

In addition to persons of Negro and of mixed Negro and white descent, this classification includes persons of mixed Indian and Negro descent, unless the Indian ancestry predominates or unless the Individual is regarded as an Indian in the community.

In addition to full-blooded Indians, persons of mixed white and Indian blood are included if the proportion of Indian blood is one-fourth or more, or if they are regarded as Indian In the community.

Separate statistics are given in this report for Japanese, Chinese, and Filipinos. The category "Other races" includes Koreans, Hawaiians, Asian Indians, Eskimos, Aleuts, Malayans, etc.

Persons of mixed racial parentage are classified according to the race of the nonwhite parent, and mixtures of nonwhite races are, classified according to the race of the father, with the special exceptions noted above.

Comprises persons born in the United States, the Commonwealth of Puerto Rico, or a possession of the United States; persons born in a foreign country or at sea who have at least one native American parent; and persons whose place of birth was not reported and whose census report contained no contradictory information, such as an entry of a language spoken prior to coming to the United States.

Includes all persons not classified as native.

Consists of native persons both of whose parents are also natives of the United States.

Includes native persons one or both of whose parents are foreign born.

Includes foreign-born persons and native persons of foreign or mixed parentage.

A household consists of all the persons who occupy a housing unit. A house, an apartment or other group of rooms, or a single room, is regarded as a housing unit when it is occupied or intended for occupancy as separate living quarters; that is, when the occupants do not live and eat with any other persons in the structure and there is either (1) direct access from the outside or through a common hall or (2) a kitchen or cooking equipment for the exclusive use of the occupants. A household includes the related family members and also the unrelated persons, if any, such as lodgers, foster children, wards, or employees who share the housing unit.

For married couples without own household, four categories of relationship of the husband to the head of the household are recognized in this report:

1. A son of the head is a child, stepchild, or adopted child of the head.
2. A don-in-law is the husband of the head's daughter, stepdaughter, or adopted daughter.
3. An other relative of the head is a man related to the head of the household by blood, marriage or adoption, but not the head's own son or son-in-law.
4. A nonrelative of the head is any member of the household who is not related to the household head. This category includes lodgers (roomers and partners, relatives of such persons, and foster children) and resident employees (maids, hired farm hands, etc.)