United States 1986 - documentation

Table of Contents

• A. GENERAL INFORMATION
• B. POPULATION AND SAMPLE SIZE, SAMPLING METHODS
• C. MEASURES OF DATA QUALITY
• D. DATA COLLECTION AND ACQUISITION
• E. WEIGHTING PROCEDURES
• F. DETERMINATION OF SURVEY UNIT MEMBERSHIP
• G. CHILDREN AND SPOUSES
• H. AVAILABILITY OF BASIC SOCIAL AND DEMOGRAPHIC INFORMATION
• I. AVAILABILITY OF LABOUR MARKET INFORMATION
• J. AVAILABILITY OF GEOGRAPHIC INFORMATION
• K. SOURCES AND AMOUNTS OF CASH INCOME
• L. TAXES
• M. BIBLIOGRAPHY OF MAIN PUBLICATIONS

 

 A. GENERAL INFORMATION back

Official name of the survey:
The March Current Population Survey (CPS)

Administrative unit responsible for survey:

The Bureau of the Census is an agency of the US federal government, and operates as the central statistical office for the United States.
Funding for the survey is provided by the budgets of the Bureau of Labor Statistics (Federal Agencies), and the Bureau of the Census.
The data are stored at the Demographics Survey Division at the Bureau of the Census at the address above.
Copies of the original codebooks and other documentation can be obtained from the address above.
For additional information to that provided here on various aspects of the survey, the following people may be contacted:

 Sample Design:

Data collection process:

Demographic Information:

Income Information:

Editing and Survey Processing:

The principle users of the data collected in this survey are government agencies, universities, public and private research institutes, businesses, and any other organizations or individuals who wish to use it.

The March CPS is conducted every year. The first year in which the CPU was conducted was 1947, and was first made available as microdata in 1964.

The main purpose of the survey is to provided estimates of employment, unemployment, and other characteristics of the US labor force.

The process of data collection (i.e. field work) for the first wave of LIS data was begun in March 1980 and completed in March 1980, and for the second wave of LIS data it was begun in March 1987 and completed in March 1987.

A bibliography of the most important publications based on the CPS data is provided in Section N. Also included in the bibliography in Section N is a list of the user documentation for the CPS data.

 B POPULATION AND SAMPLE SIZE, SAMPLING METHODS back

1. The Sample Design and Sampling Frames

The sampling frame for the survey consists of a list of all housing units compiled from the most recent decennial census of population and housing, supplemented by lists of newly constructed housing units that did not exist when the census was conducted.

People living in institutions (such as hospitals, nursing homes and longterm care facilities, homes for the aged, penitentiaries, housing for members of the military not living with their families, dormitories at schools, and schools) were not included in the survey.

All housing units containing one or more permanent residents are eligible for interview.

A household is defined as a group of persons or a person living alone in a housing unit. Family relationships are not taken into consideration in this definition.

The CPS universe is comprised of households and persons living in households.

Once the sample was selected from the sampling frame, all sampling units were eligible for interview.

The sample design of the CPS based on this sampling frame is a multistage stratified probability sample of the population. It is a self weighting sample.

The survey unit was comprised of housing units which were defined as a house, apartment, flat, or other group of rooms, or a single room, intended for occupancy as separate living quarters (i.e. when the persons living there do not live and eat with any persons in the structure and there is direct access from the outside or through a common hall).

There were different selection probabilities for each group. Households with Hispanic household members had a lower probability of selection.

There were a total of approximately 89,000,000 occupied units, and approximately 10,000,000 vacant housing units in the population from which the sample was selected.

Unoccupied units were not excluded.

The total number of respondents who were actually surveyed were drawn from approximately 62,000 interviewed households. When weighted this amounts to 89,473 cases.

The sample that was made available for the LIS database was comprised of 62,000 cases, which when weighted amounts to 89,473 cases. LIS was able to make available to its users 11,614 unweighted cases.

There were 2,500 sample units in the sampling frame which were not interviewed because they were found to be ineligible at the time of the interview.

The entire geographic area of the United States, including Hawaii and Alaska, is included in the CPS survey design. Puerto Rico and other territories are not included.

Standard errors are used to determine the reliability of the survey estimates and to evaluate the statistical validity of conclusions made about the data.

C. MEASURES OF DATA QUALITY back

1. Item and Group Nonresponse Rates

Those households or individuals who fell into the survey frame but which were not interviewed because they could not be located or because they refused to participate in the survey were between 4.5 - 5% of households. Moreover, 8% to 9% refused to answer any of the income questions, even though they had provided responses to the labor force activity questions.

If these cases are combined with others for which responses to some but not all income questions occurred the "item" non- response rate for income amounts averages about 15%.

Higher rates of missing responses can be found for self-employment income (33 percent) and property income (25 percent).

These rates have shown a slight upward trend over the past 10 years.

A breakdown of the non-response rates for amounts of income from the CPS is reported in Table .1.

The rates in Table .1 reflect the percent of persons receiving the income source, but not answering the question concerning the amount received.

Table .1:
Non-response rates for amounts of income from the March Current Population Survey (CPS) 

2. Reporting and Under-Reporting and Income Data Quality

A comparison has been made of CPS estimates and national income account estimates of the total amount of income received by the population for selected income sources.

Internal edits and checks for consistency errors or for data items have been performed. The processing system contains extensive consistency edits that are designed to assure internal consistency with regard to survey concepts.

The estimates, based on the household sector of the national income accounts, have been adjusted to be consistent with the survey's noninstitutional universe.

3. Data Corrections and Adjustments

Some attempt is made to adjust the CPS for two of the three main sources of nonsampling error: noninterviews and nonresponse.
All missing survey responses have been imputed using "Hot Deck" methodology.

No attempt is made to adjust for errors present in the responses.

In the case of noninterviews, the weights of interviewed households are adjusted upward by characteristics that were measurable for both noninterviewed and interviewed units. Since very little information is available for the noninterviewed households (location and race of head), this adjustment has limited effect.

For nonresponses, imputation procedures are used to replace the nonresponse to the question with an answer that was characteristic of other households with similar characteristics.

Recent research has shown that the current imputation procedures correct only slightly for the bias due to the fact that nonrespondents have, on average, higher levels of income than respondents.

4. General Assessment of Income Data Quality

Users of this data should be aware of the following problems regarding the quality of the income data collected in this survey:

1. Poor quality of the property income (interest, dividends, rent).
2. Poor quality of means-tested transfer income.
3. It appears that due to general nonsampling errors at the upper tail of the income distribution, very rich people are not well represented.

D. DATA COLLECTION AND ADMINISTRATION back

1. Method of Data Collection

The data was collected by telephone or personal visit by an interviewer. About 65% of the interviews are made by telephone and 35% by personal visit to the household.

Response to all of the questions on the survey is voluntary, and households may refuse to participate without penalty.

2. Structure of the Survey Instrument

In the CPS a two part questionnaire was used. The first part contained questions on current labor force participation (employment and unemployment). The second part covered questions about income and work experience for the preceding calendar year.

Questions, for the most part, are on an individual (person) basis for all persons age 15 and over. All persons age 15 and over are asked the same group of questions.

3. Survey Respondent Rules

All responses for the household members in the CPS are provided by the same "household" respondent.

Questions may be answered by either the individual him/herself or by some other responsible person (proxy) who has knowledge of the income and work experience of other household members.

An introductory letter to each household suggests that records be consulted. Over 50 percent of the responses are given by proxy respondents. While respondents are encouraged to consult their records to help provide accurate answers, only a very small proportion (approximately 5% - 10% ) actually use records when responding.

E. WEIGHTING PROCEDURE back

In the CPS a two stage weighting procedure is used. First, the "Base" weights of interviewer households are adjusted to account for noninterviews. Next these weights are adjusted to take into account independent estimates of the total population by age, sex, and racial composition.

The sum of the weights reflects the number of households (families) existing at the time of the survey.

F. DETERMINATION OF SURVEY UNIT MEMBERSHIP back

A person is considered a member of a household (survey unit) if that housing unit is their usual place of residence, or if at the time of the interview they have no usual place of residence.

Usual place of residence is the place at which a person usually eats and sleeps, and to which that person is free to return to at any time.

This definition includes persons who usually live in the sample unit but who are temporarily absent, such as unmarried children away at school who live there when not at school, persons traveling on business, seamen, railroad workers, persons away on vacation, or temporarily in hospital.

Also included as household members are lodgers, servants, farm workers, and other employees who live in the unit and consider it their usual residence.

Citizens of foreign countries living in the United States are defined as household members using these same definitions.

Members of the military who are away on assignment, and who do not eat or sleep in the housing unit are not included as household members, even though they may be married to one of the current household members.

Individuals were the basic unit of aggregation which was used in this survey.

Families are defined as groups of people within the household who are related by blood, marriage, or adoption. In addition, subgroups within the household are established based solely on relationship. Persons are considered to be related if they are bound by blood, marriage, or adoption. All related persons living within the same housing unit are grouped together and considered a family, even though the family may include several generations, or several separate married couples, such as a husband and wife and the wife's parent.

If a housing unit contains two unrelated family groups, each family group is defined separately, as are the relationships of the members of each group to the family head.

G. CHILDREN AND SPOUSES back

In the CPS children are defined in terms of their relationship to household and family heads. They are included in the household in which they reside, even if their parents are separated or divorced, or the children are supported mainly by the absent spouse's income (dependents for tax purposes).

A person of any age can be considered an "own" child if their parent is the household or family head. Other persons of any age who are related to the household or family head, but are not an own child, are defined as "other relatives."

Foster children are not considered children of the head, but as unrelated individuals.

In the CPS spouses must be legally married. Thus, "married" means legally married. Cohabiting men and women are not defined as related in the survey even if they have children.

Persons who are not related to any other person in the household are not members of any family and are considered to be unrelated individuals.

H. AVAILABILITY OF BASIC SOCIAL AND DEMOGRAPHIC INFORMATION back

In Table US86.2 are summarized the basic social and demographic information which is available in the CPS.

Table US86.2
Basic Social and Demographic Information 

I. AVAILABILITY OF LABOUR MARKET INFORMATION back

Labor force status information is available for the CPS. The definition of work for the CPS includes work for wages, salaries, tips, commissions, or pay in-kind. It includes work of the self-employed operating their own business or farm, and work of unpaid family workers who are working in a business owned by the family. Weeks worked includes weeks spent on vacation if pay was received. It does not include volunteer work, work around the house, or work in other areas of home production.

Persons are classified as unemployed if they have no employment during the reference period, but were available for work and had made specific efforts to look for work. Also counted as unemployed are persons who are waiting to return to a job from which they had been laid off, and persons waiting to report to a new job within 30 days.

Weeks worked full time refers to the number of weeks in which the person worked for 35 hours or more.

Weeks worked part time refers to the number of weeks in which the person worked for less than 35 hours. This and other labor market information is summarized in Table .3.

Table US86.3
Labor Market Information 

 J. AVAILABILITY OF GEOGRAPHIC INFORMATION back

Information on the state in which the household is located is available. In the case of large metropolitan areas, the name of the city is available. For metropolitan areas, whether the location inside or outside the "central city" is also available.

Farm and nonfarm designations are made based on gross farm sales. Any households outside metropolitan areas with gross sales of $1,000 per year are designated as farm residences.

 K. SOURCES AND AMOUNTS OF CASH INCOME back

Sources and amounts of income are recorded for a calendar year period (January 1 through December 31). The amounts recorded are gross amounts, before any taxes or other deductions were made. All sources of cash income are covered, both taxable and nontaxable, except capital gains (or losses) from the sale of real property.

Exchange of income between members of the same household is not defined as income.

Sources of other noncash income, such as net income (rental value) from housing, health, or education are not collected directly in the survey but have been estimated using statistical models.

L. TAXES back

For the CPS statistical models have been developed to estimate federal income taxes, state income taxes, and payroll taxes. Tax paying units and the amount of taxes paid are simulated using data reported in the survey. The amounts of taxes paid by each tax paying unit are then added to obtain family and household tax amounts.

M. BIBLIOGRAPHY OF IMPORTANT PUBLICATIONS  back

U. S. Department of Commerce; 1986: March 1986 CPS Interview`s Instructions Section I; CPS Interviewers Memorandum no. 86-03, Section I- March 1986.

U. S. Department of Commerce; 1986: Household After-Tax Income: 1986; U. S. Department of Commerce; Bureau of Census.

U. S. Department of Commerce; 1986: Money Income of Households, Families, and Persons in the United States: 1986; Consumer Income; Series P-60, No. 159.