EMPLOYMENT STATUS

Employed. Employed persons comprise those who, during the survey week, were either (a) "at work"-those who did any civilian work for pay or profit, or worked without pay for 15 hours or more on a family farm or business; or (b) "with a job but not at work"those who did not work and were not looking for work but had a civilian 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 are persons who had new jobs to which they were scheduled to report within 30 days.

Also

Unemployed.-Unemployed persons include those who did not work at all during the survey week, and who were looking for work. included as unemployed are persons who 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 line of work or in the community.

Labor force.-Persons are classified as in the civilian labor force if they were employed as civilians, or unemployed during the survey week. Persons in the Armed Forces are considered part of the total labor force, but in this report are combined with persons not in the labor force.

Age. The age classification is based on the age of the person at his last birthday.

SOURCE AND RELIABILITY OF THE ESTIMATES

Source of data. The estimates presented in this report are based on data obtained in connection with the monthly population sample survey of the Bureau of the Census. The 1954 income statistics, collected in April 1955, are based on a new sample design instituted in January 1954. This sample is spread over 230 sample areas, comprising 453 counties and independent cities, in 47 States and the District of Columbia.

Data on income of families were collected from approximately 14,000 households, or about 75 percent of the households included in the April 1955 survey. Persons in the following categories were not included:

1. Members of the Armed Forces living in barracks, etc., on military reservations. (Members of the Armed Forces living off post or with their families on military reservations were included.)

2. Inmates of penal and mental institutions and homes for the aged, infirm, and needy.

On approximately 5 percent of the 14,000 schedules, no information was recorded because no interview could be obtained during the week in which the enumeration was conducted. In order to account for these schedules, the weights assigned to other schedules for households of similar characteristics residing in the same sample areas were increased accordingly. In addition, complete income information was not reported for about 10 percent of the households. Substitutions were not made for these schedules. Punchcards, which were prepared from these schedules, were included in the tabulations which provided

the base numbers for persons shown in the published tables. The distributions by income levels for each group, however, are based only on those cases which reported complete income information.

The estimating procedure used in this survey involved the inflation of weighted sample results to independent estimates of the civilian noninstitutional population of the United States by age, color, and sex for April 1955, and by age, sex, and veteran status (for males, for earlier years. The independent estimates for April 1955 were based on statistics from the 1950 Census of Population; statistics of births, deaths, immigration, and emigration; and statistics on the strength of the Armed Forces. To these totals were added the population in the Armed Forces living off post or with their families on post. For April 1949 the independent estimates of the population were based on the 1940 census data brought forward to the survey month to take account of births, deaths, net immigration, and aging of the population.

Reliability of the estimates.-Since the estimates of income distributions are based on a sample survey, they are subject to sampling variability. Table A presents approximate standard errors of estimates of selected sizes. The reliability of an estimated percentage depends upon both the size of the percentage and the size of the total on which it is based. Table B presents the approximate standard errors of estimated percentages computed by using data from the report for both numerator and denominator.

The standard error is a measure of sampling variability. The chances are about 2 out of 3 that the difference due to sampling variability between an estimate and the figure that would have been obtained from a complete count of the population is less than the standard error. The amount by which the standard error must be multiplied to obtain other odds deemed more appropriate can be found in most statistical textbooks.

The estimates of sampling variability shown above are not to be applied to estimates of changes between 1948 and 1954. The standard error of differences between the 2 years is approximately the square root of the sum of squares of standard error of each estimate taken separately.

In addition to sampling variation, the figures are subject to errors of response and nonreporting, but the possible effect of such errors is not included in the above measures of reliability. In most cases the schedule entries for income are based on memory rather than on records, and in the majority of instances on the memory or knowledge of some one person, usually the wife of the family head. The memory factor in data derived from field surveys of income probably produces underestimates, because the tendency is to forget minor or irregular sources of income. Other errors of reporting are due to misrepresentation or to misunderstanding as to the scope of the income concept. The figures on aggregate income are subject to errors of estimation in addition to those noted above.

SECTION 2. CHARACTERISTICS OF LOW-INCOME FAMILIES, 1948, 1953, AND 1954

Prepared by the Board of Governors of the Federal Reserve System 1

Source of data.-These data regarding low-income families and unattached individuals are based on the surveys of consumer finances conducted each year since 1946 by the Federal Reserve System in cooperation with the Survey Research Center of the University of Michigan.

Survey data are obtained by personal interviews with consumers living at a randomly selected sample of addresses in the 12 largest metropolitan areas and 54 additional counties and groups of counties throughout the country. Separate interviews are taken with each spending unit in the dwelling. The spending unit is defined as all persons living in the same dwelling and related by blood, marriage, or adoption who pool their incomes for major items of expenses. their incomes are not pooled, related individuals in the same dwelling unit belong to separate spending units. The units which do not contain the heads of households are called related secondary spending units.

If

Family data can also be derived from the surveys. To provide. family data, information obtained in interviews with related secondary spending units is integrated with that from the spending unit containing the family head. Previously published survey distributions have usually combined families and unattached persons living alone or

1 The Director of the Division of Research and Statistics indicated that the materials were prepared by John Frechtling of the Division's staff.

with nonrelatives. The data presented here, however, include separate tabulations for families and for unattached individuals.2

Data obtained from sample surveys are subject to response and sampling errors. Response errors are known to exist for certain data from the Surveys of Consumer Finances but, because the same questions are used for all respondents and because the questions have not been changed significantly in the areas considered here, response errors are believed to be of relatively minor importance in intergroup and year-to-year comparisons.

Data based on samples are affected by chance variations in the distribution of the characteristics of units interviewed from the distribution of characteristics in the entire population. Sampling errors indicate the expected magnitudes of these variations. Tables 7 and 8 contain recent estimates of sampling errors for Survey of Consumer Finances data. (Tabular material presented on pp. 23-30.)

Little change occurred between 1953 and 1954 in the level and distribution of the income of families and unattached individuals. Therefore, the following discussion will be based on averages of the data presented for each of the 2 years in the accompanying tables. There are no statistically significant shifts in the characteristics of the low-income families from 1953 to 1954. These averages probably give a fairly accurate picture of conditions generally prevailing during these 2 years.

THE ESTIMATED NUMBER OF LOW INCOME UNITS

In 1953 and 1954 ten million out of about forty-nine million families and unattached individuals had money income prior to taxes of less than $2,000. About one-fourth of these low-income units were families living in urban areas and almost two-fifths were rural families. One-fourth of the units were unattached individuals living in urban areas and one-tenth were unattached individuals in rural areas.

The choice of $2,000 as a dividing line between low and moderate income families and of a population concentration of 2,500 as a division between urban and rural areas is necessarily arbitrary. A money income of $2,000 can support a retired couple in more comfortable circumstances than those which can be attained by a family with several children. Similarly, the use of only two locational classifications ignores differences in levels of income, and of expenses, between metropolitan areas and small cities.

COMPARISON OF LOW AND MODERATE INCOME UNITS

Units with incomes of less than $2,000 differ in many respects from units with moderately higher incomes. Units with incomes of $2,000 to $4,999 have been chosen as a comparison group to focus attention on problems of moderate improvement in the status of the low-income group.

Comparisons will be made for low and moderate income units of three types: Urban families, rural families, and unattached individuals living in urban areas. The number of cases of unattached individuals living in rural areas found in the surveys is not large enough to merit further statistical treatment. Table 2 indicates that

2 For additional information regarding procedures of the survey of consumer finances, see Methods of the Surveys of Consumer Finances, Federal Reserve Bulletin, July 1950.

they constitute a problem of limited scope relative to the other three low-income groups. The salient points of the comparison of the low and moderate income groups are listed below.

Low-income families tend to be smaller than those with moderate incomes. About half of the low-income urban and slightly fewer of the low-income rural families included only two adults. In contrast, one-third of the urban and one-fourth of the rural families with moderate incomes included only two adults.

Broken families are more common in the low-income group.—Oneseventh of the low-income urban families included 1 adult and 1 or more children but only one-twentieth of the middle income families were of similar structure. Broken families were relatively infrequent in both low and middle income rural groups.

The heads of low-income units tend to be older.-More than one-third of the urban and more than one-fourth of the rural low income families were headed by persons 65 years or more of age. In contrast less than one-tenth of the middle income urban and rural groups were headed by older persons. More than half of the low-income unattached individuals in urban areas were 65 or over in contrast to one-eighth of those with moderate incomes.

The heads of low-income units tend to have had less education.-Seven-tenths of the heads of urban low-income families and threequarters of those in rural areas had not had any formal education beyond grammar school. Among families with moderate incomes, less than two-fifths in the urban areas and about half in rural areas were headed by persons of such limited educational attainment. A similar pattern was shown for unattached urban individuals.

A majority of urban low-income units were headed by persons engaged in very limited or no productive effort.-More than half of the lowincome urban families were headed by retired or unemployed persons, students, housewives or protective service workers, but only one-sixth of the moderate income families were headed by persons in these groups. Two-thirds of the unattached urban individuals having low incomes and one-fifth of those with incomes of $2,000 to $4,999 were in the above-mentioned occupational groups.

In rural areas, farm operators and retired persons were more important in the low than in the moderate income group.-Two-fifths of the rural families with money incomes of less than $2,000 were headed by farm operators and another fifth by retired persons. Farm operators headed less than one-fourth and retired persons less than 5 percent of the moderate income rural units.

A disproportionate number of low-income units live in the South.Among the low-income groups, almost half of the urban families, three-fifths of the rural families, and one-fourth of the unattached urban individuals live in the South. For family units with incomes of $2,000 to $4,999, the corresponding proportions were about onefifth, two-fifths, and one-fifth.

Low-income units in general do not have as large accumulations of liquid assets as middle income units.-Less than one-fourth of the lowincome families and unattached individuals reported accumulated liquid asset holdings of $500 or more. (Liquid assets include demand deposits, savings accounts, shares in savings and loan associations and credit unions and U. S. Government bonds.) More than one-third of