Nursing Aides - Chapter 4

Nursing Aides, Home Health Aides, and Related Health Care Occupations -- National and Local Workforce Shortages and Associated Data Needs

Printer Friendly Adobe .pdf (1,970K)

Chapter 4. Existing National Data Sources

This chapter describes the national data sources and includes the following sections:

Introduction
Occupational Employment Statistics
Current Population Survey
Current Population Survey March Supplement
National Compensation Survey
Employment Projections
BLS Survey of Occupational Injuries and Illnesses
Online Survey Certification and Reporting System
Decennial Census

Introduction
An important part of any assessment of data resources related to direct care paraprofessionals is a careful review of existing sources of data. Such a review helps planners and policymakers understand the strengths and limitations of current data resources. It also reveals appropriate ways to use existing data and suggests ways to improve data collection and analysis techniques, with the goal of creating databases that are more useful for workforce planning.

Several national surveys that collect general employment statistics also collect data relating to the direct care paraprofessional workforce. However, the data collection is not exclusive to direct care paraprofessionals, and the terminology and definitions the surveys use are not necessarily consistent from one to the next or with current workforce conditions. This chapter briefly describes the surveys and suggests improvements in data collection and analysis to provide better information for workforce planning.

Table 4-1 lists the surveys, summarizes their primary data characteristics, and notes their respective strengths and limitations. The surveys are:

Occupational Employment Statistics (OES)
Current Population Survey (CPS)
CPS March Supplement
National Compensation Survey (NCS)
Employment Projection
BLS Survey of Occupational Injuries and Illnesses
US Decennial Census
Online Survey Certification and Reporting System (OSCAR)

Subsequent sections describe each survey in more detail.

Table 4-1. Comparison of Direct Care Workforce Data Sources

Occupational Employment Statistics
Overview
The OES program is an annual mail survey that supports estimating employment and wages for over 700 occupations in the United States. It is a cooperative program that includes the BLS and State Employment Security Agencies (SESAs). Its Internet address is http://www.bls.gov/oes/.

OES collects number and wage/salary data on both full-time and part-time wage and salary workers in non-farm establishments. It does not collect data on self-employed, household, or unpaid family workers. The program surveys approximately 400,000 establishments per year for three years. The data it collects fall into two primary categories: geographic area (national, state, metropolitan) and industry. Prior to 1996, OES produced only occupational employment estimates by industry. In 1996, it began collecting both occupational employment and wage data. In 1997, it began estimating cross-industry as well as industry-specific occupational employment and wages.

In 1999, the OES survey began using the new Office of Management and Budget (OMB) 2000 Standard Occupational Classification (SOC) system. Due to the transition to the SOC system, 1999 OES estimates are not directly comparable with previous OES estimates, the classifications of which are compatible with the 1980 SOC and the U.S. Bureau of the Census occupational classifications. OES uses definitions of industries from the Standard Industrial Classification (SIC) system. Chapter 6 provides an overview of these classification systems and definitions of relevant occupations/industries.

See Appendix D for sample OES data.
OES Strengths and Limitations
OES Strengths
OES’s primary strength is its large sample size, which allows developing and comparing estimates by geographic area and industry. It also allows more detailed occupational classifications, which better describe the current direct care workforce.

OES Limitations
Unlike some other surveys, e.g., CPS, OES does not provide data on demographic characteristics and work conditions. In other words, OES tells how many people are in a particular occupation in a particular industry and how much they earn, but it does not describe them beyond their numbers and wages.

As stated earlier, OES does not collect data on self-employed, household, or unpaid family workers. This is a substantial limitation considering the potentially large number of home care workers who don’t work through organizations but through contracts with patients and families.

Definitions of each occupation and industry are also problematic in that they do not reflect current conditions. Also, OES’s data definitions have changed significantly through its history, which makes it difficult to conduct analyses over time.

Current Population Survey
Overview
The CPS is a fifty-year-old monthly survey of about 50,000 to 60,000 households the Bureau of the Census conducts for BLS. CPS is the primary source of information concerning U.S. labor force characteristics. Its sample represents the civilian, non-institutional population aged 15 years and over. Informants provide information about their employment status, earnings, hours of work, occupation, industry, and demographics. Data falls into three geographic areas: national, state, and sub-state. CPS occupational and industrial data classifications are based on the coding systems the 1990 census used.

The CPS Internet address is http://www.bls.census.gov/cps/cpsmain.htm.

See Appendix D for CPS sample data.

CPS Strengths and Limitations

CPS Strengths
Unlike other national surveys, CPS has demographic data on each respondent, which helps to understand which sectors of the population work in which occupation and industry groups. The CPS also includes self-employed workers, which is particularly important for the home care industry given that a number of direct care workers contract directly with individual patients/clients.

Relative to those of other surveys such as OES, CPS data definitions have not changed significantly, which makes it easier to conduct analyses over time.
The monthly survey also has a State variable (not available in the March supplement); however, due to the small sample size of direct care workers, it may be necessary to combine data from several months to conduct meaningful analyses by state.

In a few years, CPS will start using uniform classification systems that are consistent with other survey programs. Those classifications generally reflect current conditions better.

CPS Limitations
The CPS data’s primary limitation relates to occupation and industry definitions. The welfare service aide’s category (Code 465) includes individuals who are not necessarily direct care workers. Some industry codes also contain work settings irrelevant to the direct care workforce, e.g., medical laboratories, youth services, crisis center, food bank, etc. The lack of clear definitions makes it harder to draw accurate pictures of direct care workers.

The change to a uniform classification system will make it harder to conduct analyses of CPS data over time.

Current Population Survey March Supplement
Overview
The CPS March Supplement, also called the Annual Demographic Survey, is the primary source of detailed information on income and work experience in United States. Relative to the monthly survey, the CPS March Supplement contains more detailed data on individuals, including: geographic mobility, income and poverty status, and labor force and work experience. It also includes personal, family, and household data.

The CPS March Supplement’s sample size is slightly larger than monthly surveys. For example, in 1995, it included the basic monthly CPS sample of 60,000 housing units and 2,500 housing units that had at least one Hispanic member the previous November. It also includes members of the U.S. Armed Forces, who are excluded from the monthly surveys. Like the monthly CPS survey, the CPS March Supplement uses occupational and industrial classifications based on the coding systems the 1990 census uses.

The CPS March Supplement’s Internet address is http://www.bls.census.gov/cps/cpsmain.htm.

See Appendix D for CPS March Supplement sample data.

CPS March Supplement Strengths and Limitations

CPS March Supplement Strengths
Like the CPS monthly survey, the CPS March Supplement provides detailed data on each worker. It has even more detailed data such as availability of benefits, e.g., health insurance, pension, and recipients of public assistance, e.g., Medicaid, food stamps.

It has also benefited from consistent definitions of occupations and industries over time.

Like the monthly survey, the CPS March Supplement will start using uniform classification systems that are consistent with other survey programs.

CPS March Supplement Limitations
Unlike the monthly survey, the CPS March Supplement does not have a State variable. Although it contains a region variable, it is of very limited use for researchers who are interested in particular states or who would like to compare different states.

Like the monthly survey, the CPS March Supplement has limitations in occupation and industry category definitions.

Also like the monthly survey, the change to a uniform classification system will make it harder to conduct analyses of CPS data over time.

National Compensation Survey
Overview
NCS is a BLS survey that provides comprehensive measures of occupational earnings, compensation trends, benefit incidences, and detailed benefit provisions. It also includes average weekly work hours. It integrates three BLS programs: the Occupational Compensation Survey, the Employment Cost Index, and the Employee Benefits Survey. Participants respond via personal interviews that are conducted annually.

Like the OES, NCS also excludes self-employed, household, and unpaid family workers. In addition, while the OES includes Federal government employees, NCS includes only State and local government employees. It covers approximately 36,000 establishments per year and compares earnings and weekly work hours using several variables, including: full-time versus part-time, private industry versus government, level of work, and geographic areas (national, regional, and metropolitan).
NCS defines each occupation by using the Occupational Classification System Manual, which is based on the 1990 Census Index. Although NCS has wage data by industry, only major industry divisions are available. Therefore, researchers cannot analyze NCS data by detailed industry setting, e.g., home care, nursing homes, hospitals.

The NCS Internet address is http://www.bls.gov/ncs.

See Appendix D for sample NCS data.

NCS Strengths and Limitations

NCS Strengths
NCS provides detailed wage information for each occupation. Unique to NCS are the wage data by work level. NCS data show that the wages of aide workers differ depending on the worker’s knowledge and responsibilities. NCS data are also consistent with OES data in a sense that the highest wage aide workers can make is about $13 and that the average wage is between $7.50 and $9.00. One can also see in NCS data that, despite the existence of several work levels, even the highest level is 8 out of 15 work levels, suggesting that the aide occupations are at the low end among different occupation groups.

NCS Limitations
Despite the detailed wage data, NCS has several limitations that make it harder to use the data to understand working conditions of direct care workers. Unlike OES data, NCS data do not use a detailed industry classification. Hence, NCS cannot distinguish direct care workers in different settings, e.g., nursing homes, hospitals, home health care, assisted living, etc. In addition, the occupation codes NCS uses do not seem to be consistent with current conditions.

Employment Projections
Overview
The BLS Office of Employment Projections develops ten-year estimates about the national labor market. Their work includes labor force trends by sex, race, national origin, and age; employment trends by industry and occupation; and the implications of these data for employment opportunities for specific groups in the labor force. BLS updates the projections every other year.

BLS develops the National Industry-Occupation Employment Matrix as part of its ongoing Occupational Employment Projection Program. The matrix provides information on the distribution of employment for an occupation across industries. The latest matrix gives information on occupational employment growth in different industries between 1998 and 2008. The 1998 matrix uses the Occupational Employment Statistics (OES), Current Employment Statistics (CES), and CPS surveys. Projections are by labor force, aggregate economy, final demand, industrial activity, employment by industry, and employment by occupation.

The projections use the occupational classification that reflects the OES survey. Data on self-employed workers and unpaid family workers are based on CPS data for equivalent occupations. A crosswalk, based on each survey’s compatibility with the 1980 SOC, attributes CPS data to an equivalent occupation in the industry-occupation matrix. Industries covered in the matrix reflect the 1987 SIC. Self-employed, unpaid family workers, and workers who have a second job in private households are listed as separate industries to derive total employment.

The BLS employment projections Internet address is http://www.bls.gov/empover.htm.

See Appendix D for the latest projections, which show dramatic increases in CNAs, HHAs, and PCAs between 2000 and 2010.

BLS Employment Projections Strengths and Limitations

BLS Employment Projections Strengths
These data provide estimates and projections for each occupation by industry, as well as by state. Unlike the OES data, the projections also include self-employed and household workers, which apply to a number of direct care workers in community settings.

BLS Employment Projections Limitations
The projections make no distinction between PCAs and HHAs. Although those two occupations share a number of elements, some important factors seem to differ, including their wages, employers (industry), and some tasks. Also, like other data sources, the industry definitions seem to be problematic and may not reflect current realities. Chapter 5 discusses the issues regarding occupation and industry classifications in greater detail.

BLS Survey of Occupational Injuries and Illnesses
Overview
The current BLS survey of occupational injuries and illnesses evolved from annual BLS surveys first conducted in the 1940s. The older surveys had several limitations, including voluntary reporting and exclusion of injuries that did not involve lost work time. In 1970, the Occupational Safety and Health Act was enacted, and its implementation required that most private industry employers regularly maintain records and prepare reports on work-related injuries and illnesses. The current survey selects approximately 250,000 private sector organizations that have 11 employees or more. National data, as well as State data to a certain extent, are available on the web site. Data include incidence of occupational injuries and illnesses by industry, occupation, workers’ demographic characteristics, employer size, event or exposure, nature of injury, and part of body affected. The survey uses 1990 census codes for occupations and 1987 standard industrial classifications.
The survey’s Internet address is http://www.bls.gov/iif.

See Appendix D for sample data from the BLS Survey of Occupational Injuries and Illnesses.

BLS Survey Strengths and Limitations

BLS Survey Strengths
This survey provides valuable data on occupational safety. The literature points out a number of injuries (particularly back pain and falls) among direct care workers. The survey data not only confirm the literature but also show the severity of the problem.

BLS Survey Limitations
Although the survey contains both occupation and industry variables, the cross-tabulation of the two variables is not available on its web site. Because each industry contains different occupation groups, e.g., doctors, nurses, administrative staff, etc., this survey may have very limited use for comparing direct care workers in different settings. Also, as with other surveys, definitions of each occupation and industry are problematic because they do not reflect current labor situations and conditions.

Decennial Census
Decennial Census Strengths
The decennial census is an important source of information about the population of the U.S. The one-in-six sample used for the long form of the census questionnaire provides limited information about the employment status of members of households residing in the U.S. Perhaps its greatest strength is related to the fact that the file permits tabulations for small geographic areas (down to census tracts and for some questions down to block groups.

Decennial Census Limitations
The decennial census was not designed to support workforce planning. The several components of the long form of the census questionnaire that deal with occupations and industries are designed primarily to provide very basic information and insights about the kinds of jobs that U.S. residents hold. The key limitations of this file for understanding long term care paraprofessional workers include: the ten-year gap between successive collections, the delay in processing the long form questionnaires, the lack of appropriate detail about the occupational categories, and the fact that the geographic tabulations represent where people live rather than where they work.

Online Survey Certification And Reporting (OSCAR) System
Overview
OSCAR provides staffing data for all U.S. nursing homes that Medicare and/or Medicaid certifies. State survey and certification agencies collect the data, which are part of the annual nursing home certification and recertification process. Each facility completes a standardized form about the facility characteristics, e.g., number of beds, affiliation, etc., resident characteristics, e.g., limitations, chair bound, etc., and staffing levels. State surveyors review the form and enter the data into the OSCAR database. State surveyors also visit each facility and decide whether the facility meets each standard.

OSCAR staffing variables cover a small number of occupations, including registered nurses (RNs), licensed practical nurses (LPNs), and nurse aides. Each occupation breaks down into full-time (35 or more hours per week), part-time (less than 35 hours per week), and contractors. Staffing variables are reported in full time equivalency (FTE) based on a 35-hour workweek. To convert from FTEs to staff-hours per patient-day sum staff types within each staffing category.
Although OSCAR does not have an official web site from which to retrieve data, researchers can purchase raw data from CMS. CMS’s Internet address is http://www.medicare.gov/NHCompare/home.asp. Using information on the site, consumers can compare different aspect of nursing homes, including staffing levels.
Harrington and colleagues [2000] also summarized OSCAR data from 1993 to 1999 by state. Their summary is available online at http://cms.hhs.gov/medicaid/services/nursfac99.pdf.

OSCAR Strengths and Limitations

OSCAR Strengths
OSCAR provides comprehensive information on certified U.S. nursing facilities. Although very limited staffing data are available, one can analyze the data to see the association between staff levels and facility characteristics, resident characteristics, and other quality indicators.

OSCAR Limitations
Validity analyses have shown considerable differences between staffing levels from OSCAR and payroll data for the same time period, suggesting that OSCAR staffing data for some facilities are unreliable. The data were even less consistent for nurse aides than for RNs and LPNs. Also, old OSCAR data were overwritten when a new survey was conducted, which makes it very difficult to conduct historical analyses.

A report by HCFA [2000] points out some data errors and inconsistency over time. A report by Harrington and colleagues [2000] excluded such data to maximize data validity and reliability. If a researcher obtains raw data and conducts analyses, he/she will need to exclude data for facilities with obvious data errors and inconsistencies over time.