conventional 8 years. For the sake of comparability, persons who had progressed beyond a 7-year elementary school system were treated as though they had progressed beyond the usual 8-year system. Junior high school grades were translated into their elementary or high school equivalent.

In the case of persons whose highest grade of attendance was in a foreign school system, the enumerator was instructed to obtain the approximate equivalent grade in the American school system or, if that were too difficult to determine, the number of years the person had attended school. Persons whose highest level of attendance was in an ungraded school were treated in similar fashion. Persons whose highest level of training was by a tutor and whose training was regarded as qualifying under the "regular" school definition were also given the approximate equivalent in the regular school system.

Completion of highest grade.-The second question on educational attainment was to be answered "Yes" if the person had completed the full grade. If the person was still attending school in that grade, had completed only a half grade, or had dropped out of or failed to pass the last grade attended, the required answer was "No." In the case of persons who failed to report on completion of the grade, those classified as enrolled were assumed not to have finished and those not enrolled were assumed to have finished.

Comparability

Year of school in which enrolled and years of school completed. In the present report, the year of school in which enrolled is shown for enrolled persons 5 to 29 years old, and the years of school completed are shown for all persons 5 years old and over. For 1950, statistics on educational attainment for persons

they were enrolled, whereas in the 1940 reports statistics were shown in terms of the highest grade they had completed. The present procedure was adopted because it provides statistics in a form that should be generally more useful to those interested in school systems.

Generally, the grade in which persons are enrolled is one grade higher than the highest grade completed. Data from a preliminary sample of the 1950 Census, however, indicate that, by the time of the census enumeration, about 15 percent of the "enrolled" population 5 to 29 years old had completed the same grade in which they had been enrolled. This apparent contradiction occurs because the question on enrollment referred to "any time since February 1" whereas the completion question referred to the date of enumeration. Thus, highest grades of school completed for the enrolled population obtained by subtracting one grade from the grade in which enrolled must be considered only approximately correct.

The 1940 Census reports included data on highest grade of school completed for the population 5 to 24 years old not enrolled in school. As a result of the facts noted above, similar data for 1950 could be only approximated. Two steps would be involved: First, one grade should be deducted from the grade in which enrolled (as given in table 112) in order to approximate the highest grade completed for persons enrolled in school; second, the number of enrolled persons who have completed a given grade should be subtracted from the total number of persons who have completed the grade (as given in table 114).

Quality of Data

In 1940 a single question was asked on highest grade of school completed. Analysis of data from the 1940 Census returns and from surveys conducted by the Bureau of the Census using the

quently reported the year or grade in which they are enrolled, or had last been enrolled, instead of the one completed. There is evidence that, as a result of the change in the questions in 1950, there was relatively less exaggeration in reporting educational attainment than in 1940. For example, data from a preliminary sample of the 1950 Census for persons of elementary and high school ages show larger proportions in 1950 than in 1940 in both the modal grade and the next lower grade for a particular age, and smaller proportions in each of the first two grades above the mode. Hence, the indicated increases in attainment between 1940 and 1950 tend to understate the true increases.

For 80 percent of the persons enumerated in the census, the original report on educational attainment and the Post-Enumeration report were the same or only one grade apart. The net effect of misreporting educational attainment on statistics for specific grades varied from grade to grade but was minor for many grades. In addition, 1950 Census data on median lever of school attainment are but slightly affected by errors in reporting. The median number of school years completed by persons 25 years old and over derived from census data which have been adjusted for this misreporting of grade of school attained differs by only about onetenth of one year from the median originally obtained from the

census.

For persons 25 years old and over, the amount of overreporting of years of school completed somewhat exceeded the amount of underreporting. Apparently, when reporting educational attainment, respondents had some tendency to round upward to a "terminal" grade such as the 8th or 12th grade.

Median School Years

Median educational attainment (i. e., either median year of school in which enrolled or median school years completed) is expressed in terms of a continuous series of numbers. For example, the fourth year of high school is indicated by 12 and the first year of college by 13. For the sake of comparability, the first year of high school is uniformly represented by 9, although, as previously noted, there are some areas with only 7 years of elementary school.

The procedure used in both 1950 and 1940 for calculating the median year of school completed makes allowance for the fact that many persons reported as having completed a given full school year have also completed a part of the next higher grade. It is assumed, for example, that persons who reported 6 full years of school completed had actually completed 6.5 grades. At the time of enumeration, persons enrolled in school had probably completed somewhat more than one-half grade beyond their last full year, on the average, whereas persons who had left school had probably completed less than one-half year beyond their last full year, on the average. A similar procedure was followed in the computation of the median school year in which enrolled.

Illiteracy

For

The last decennial census in which information was collected on Miteracy was that of 1930. This question was dropped because of the relatively small number of illiterates remaining even at that date and because of the more useful information obtainable from questions on educational attainment. The transition from literacy to literacy is obviously a gradual process and is not reflected in the completion of a particular grade of school. example, data from the Current Population Survey of October 1947 indicate that among persons 14 years old and over the proportion reported as illiterate ranged from 80.1 percent of those who had not completed a year of school to 4.7 percent of those who had completed 4 years. From this survey and a similar one conducted in October 1952, it is estimated that the number of illiterates 14 years old and over in 1950 was about 3,600,000, or

RESIDENCE IN 1949

Definitions

The data on residence in 1949 were derived from answers to several questions asked for a 20-percent sample of persons 1 year old and over. The first question was, "Was he living in this same house a year ago?" Each person who was not living in the same house was asked, "Was he living in this same county a year ago?" and if not, "What county and State was he living in a year ago?"

Residence in 1949 is the usual place of residence one year prior to the date of enumeration. As indicated by the categories of table 48, residence in 1949 was used in conjunction with residence in 1950 to determine the extent of mobility of the population. Persons who changed residence from 1949 to 1950 were classified by the type of move, viz., "Different house, same county," and "Different county or abroad." Residence abroad includes residence in all foreign countries and all Territories and possessions, etc., of the United States. The group whose residence was "same house as in 1950" includes all persons 1 year old and over who were living in the same house on the date of enumeration in 1950 as on the date one year prior to enumeration. Included in this group are persons who had never moved during the 12 months as well as persons who had moved but by 1950 had returned to their 1949 residence. Persons 1 year old and over for whom complete and consistent information regarding residence in 1949 was not available are included in the group "residence not reported."

The number of persons who were living in different houses in 1950 and 1949 is somewhat less than the total number of moves during the year. Some persons in the same house at the two dates had moved during the year but by the time of enumeration had returned to their 1949 residence. Other persons made two or more progressive moves. Furthermore, persons in a different house in the same county may actually have moved between counties during the year but by 1950 had returned to the same county of residence as that in 1949.

The mobility statistics in this volume relate to one particular 12-month period. Annual data from the Current Population Survey indicate that in the period of slight economic recession from 1949 to 1950, mobility was at a relatively low level as compared with that of other postwar years. Therefore, the directions of net flow and the patterns of mobility shown by the 1950 Census data may not be typical in some respects of the period since the end of World War II.

Comparability

A similar set of questions on mobility was first asked in the 1940 Census. These questions, however, applied to residence five years earlier rather than one year earlier.

Quality of Data

According to the Post-Enumeration Survey, the population covered by the mobility inquiry (persons 1 year old and over) was underenumerated by the same proportion as the total population, i. e., 1.4 percent. For the enumerated population, differences in reporting of mobility status between the 1950 Census and the Post-Enumeration Survey amounted to 8.2 percent. Most of these errors, however, were offsetting, so that the net error as measured by the Post-Enumeration Survey was negligible.

Another measure of the consistency of response in the mobility questions is provided by a comparison of the 1950 Census data with Current Population Survey data for roughly the same period. The questions on migration were carried in the Current Population Survey for March 1950 and also related to residence a year

A higher proportion of the population was reported as living in the same house a year ago in the 1950 Census than in the Current Population Survey. On the other hand, the proportion of those who had moved within the same county was lower in the census. The proportion of migrants, both within the State of residence in 1950 and between States, was the same in both. Sampling variability and the slight difference in dates cannot account for the differences entirely. They seem to be further evidence that the Current Population Survey enumerators tend to be more successful than the census enumerators at identifying marginal groups with respect to a given phenomenon. In the case of mobility, the marginal group is represented by persons who moved only

The statistics on mobility in Chapter B of this report are slightly affected by tabulation errors, according to evidence provided by later tabulations. Mobility data are also presented in Mobility of the Population: State of Residence in 1949 and 1950, Series PC-14, No. 17. The total number of persons in the United States who were not reported as living in the same house in 1950 as in 1949 based on table 48 of the present report is 27,972,895. The number indicated in the Series PC-14 report is 27,813,260. The difference, 159,635, represents 0.1 percent of the total population 1 year old and over.

A comparison of the statistics for States which appear in table 70 of this volume with those presented in the Series PC-14 report indicates that, in all but five States, the differences did not exceed 0.2 percent of the population 1 year old and over. The largest difference 1.2 percent-was in the statistics for Rhode Island. In the remaining four States-Alabama, Iowa, Louisiana, ang Virginia-the differences ranged from 0.7 percent for Alabama to 0.5 percent for Louisiana. These five States account for more than half of the difference between the national figures in the two reports. The data for the five States which are presented in Volume II appear to be at fault, but the errors were discovered after it was no longer feasible to make corrections. The major part of the difference in Rhode Island is attributable to mechanical failures in the tabulation of the data for Volume II, which resulted in the tabulation of some cards as "residence not reported" which should have been classified as "same house as in 1950." In the tabulations of the data for Volume II for Alabama, Iowa, Louisians, and Virginia, special equipment was used; and here a procedure was followed whereby some persons not in the sample were counted

EMPLOYMENT STATUS

Definitions

Census week.-The 1950 data on employment status pertain to the calendar week preceding the enumerator's visit. This week, defined as the "census week," is not the same for all respondents, because not all persons were enumerated during the same week. (A majority of the population was enumerated during the first half of April.) The 1940 data refer to a fixed week for all persons, March 24 to 30, 1940, regardless of the date of enumeration.

Employed.-Employed persons comprise all civilians 14 years old and over who, during the census week, were either (a) “at work"-those who did any work for pay or profit, or worked without pay for 15 hours or more on a family farm or in a family business; or (b) "with a job but not at work"-those who did not work and were not looking for work but had a job or business from which they were temporarily absent because of vacation, illness, industrial dispute, bad weather, or layoff with definite instructions to return to work within 30 days of layoff. Also included as "with a job" are persons who had new jobs to which they were scheduled to report within 30 days.

Unemployed.-Persons 14 years old and over are classified as unemployed if they were not at work during the census week but were either looking for work or would have been looking for work except that (a) they were temporarily ill, (b) they expected to return to a job from which they had been laid off for an indefinite period, or (c) they believed no work was available in their community or in their line of work. Since no specific questions identifying persons in these last three groups were included on the census schedule, it is likely that some persons in these groups were not returned by the census enumerators as unemployed. For some purposes, unemployed persons with previous work experience are classified separately. When information on the schedule was insufficient for this distinction to be made the unemployed person was classified as an experienced worker, since the great majority of persons seeking work have had previous work experience.

Labor force. The labor force includes all persons classified as employed or unemployed, as described above, and also members of the armed forces (persons on active duty with the United States Army, Air Force, Navy, Marine Corps, or Coast Guard). The "civilian labor force," comprises the employed and unemployed components of the labor force. The "experienced labor force," consists of the armed forces, employed workers, and the unemployed with previous work experience. The "experienced civilian labor force" comprises the two latter groups.

Not in labor force.-Persons not in the labor force comprise all civilians 14 years of age and over who are not classified as employed or unemployed, including persons doing only incidental unpaid family work (less than 15 hours during the census week). Persons not in the labor force are further classified in this report into the following categories:

1. Keeping house.-Persons primarily occupied with their own home housework.

2. Unable to work.-Persons who cannot work because of a long-term physical or mental illness or disability. There is some evidence, however, that some persons were reported as "unable to work" who were only temporarily ill or who, although elderly, were not permanently disabled.

3. Inmates of institutions.-Persons, other than staff members and their families, living in institutions. (See definition of institutional population on p. 43.) Staff members of institutions and their families are classified into employment status categories on the same basis as are persons living outside of institutions.

4. Other and not reported.-Persons in this general category include the following two groups which are combined for the purpose of this report:

a. Persons not in the labor force other than those keeping house, unable to work, or in institutions. This group includes

idle, and seasonal workers for whom the census week fell in an "off" season and who were not reported as unemployed.

b. Persons for whom information on employment status was not reported. Although the number of persons classified as "not reported" was not tabulated separately for this report, it is estimated on the basis of preliminary data that the number in this group is approximately 1.2 million for the United States as a whole, or only about 1 percent of the total United States population 14 years old and over. Analysis of the characteristics (sex, age, color, marital status, school enrollment, and urban-rural residence) of persons in this group suggests that approximately half a million might have been added to the labor force had the necessary information been obtained. There may be considerable variation from State to State, however, in the proportion of persons classified as "not reported" and within this group, in the number who might have been labor force members.

Basis for Classification

The employment status classification is based primarily on a series of interrelated "sorter" questions designed to identify, in this sequence: (a) Persons who worked at all during the Census week; (b) those who did not work but were looking for work; and (c) those who neither worked nor looked for work but had a job or business from which they were temporarily absent. The four questions used for this purpose are described below:

1. "What was this person doing most of last week-working, keeping house, or something else?" This question was designed to classify persons according to their major activity and to identify the large number of full-time workers. Persons unable to work at all because of physical or mental disabilities were also identified here.

2. "Did this person do any work at all last week, not counting work around the house?" This question was asked of all persons except those reported in the previous question as working or as unable to work. It was designed to identify persons working part time or intermittently in addition to their major activity.

3. "Was this person looking for work?" Asked of persons replying in the previous question that they did not work at all, this question served to obtain a count of the unemployed.

4. "Even though he didn't work last week, does he have a job or business?" Persons temporarily absent from their job or business were identified by means of this question, which was asked of persons neither working nor looking for work.

Problems in Classification

Classification of the population by employment status is usually correct for most regular full-time workers but is subject to error in marginal cases. Some of the concepts are difficult to apply; more important, for certain groups, the complete information needed is not always obtained. For example, housewives, students, and semiretired persons, who are in the labor force on only a part-time or intermittent basis, may fail to report that they are employed or looking for work unless carefully questioned. In many cases, enumerators may assume that such persons could not be in the labor force and will omit the necessary questions. As a result of these failures, the statistics will understate the size of the labor force and overstate the number of persons not in the labor force. (See also paragraph below on "Current Population Survey.")

Comparability

Statistics on gainful workers.-The data on the labor force for 1940 and 1950 are not exactly comparable with the statistics for gainful workers presented in this report for 1920 and 1930 because of differences in definition. "Gainful workers" were persons reported as having a gainful occupation, that is, an occupation in which they earned money or a money equivalent, or in which they assisted in the production of marketable goods, regardless of whether they were working or seeking work at the time of the census. A person was not considered to have had a gainful occupation if his activity was of limited extent. The labor force is defined on the basis of activity during the census week only and includes all persons who were employed, unemployed, or in the armed forces in that week. Certain classes of persons, such