This chapter describes the occupation and industry classification systems used to differentiate and categorize different components of the direct care paraprofessional workforce, and includes the following sections:

• Introduction
• Occupation Categories
• Industry Categories
• Bridging Different Data Sources

Introduction

The national surveys that collect and describe data related to long-term care paraprofessionals and other workers use several different occupation and industry classifications. This chapter describes them in detail. Appendix E lists detailed definitions of different occupation and industry codes in each data source.

Occupation Categories

Standard Occupational Classification
The SOC system was introduced in 1970 as a response to a growing need for a universal occupational classification system. It was revised in 1980 and in 1998. It covers all for-pay or for-profit occupations in the U.S. and reflects the current occupational structure.

While the original SOC consisted of 22 divisions in a 4-digit hierarchical structure, the latest SOC uses a 6-digit structure for its 822 occupational categories. The occupational categories are across 23 major groups that are also called "job families." The latest SOC also classifies workers at four levels of aggregation as follows in Table 6-1.

Table 6-1. SOC Classifications

The next major review and revision of the SOC will most likely begin in 2005 in preparation for use in the 2010 Decennial Census. Because the latest revision rearranged the entire classification structure, analysis of SOC data across time will be very challenging.

Major classification changes occurred in the latest SOC division. First, there is now a separate, new code for home health aides. Previously, home health aides were part of the classification that included nursing aides, orderlies, and attendants (1980 SOC code 5233). The latest SOC also separates personal and home care aides from other welfare service aides such as case aides and outreach workers. Those new classifications will help identify direct care workers in the community settings more accurately.

However, nursing aides, orderlies, and attendants (1998 SOC code 31-1012) are still in one group. Orderlies tend to have different demographic characteristics, e.g., more male workers, and job responsibilities from direct care workers. Therefore, putting them in a separate category would allow describing direct care workers more accurately. The latest SOC also combined health aides (1980 SOC code 5236) and nursing aides, orderlies and attendants (1980 SOC code 5233) in one category (nursing aides, orderlies and attendants: 1998 SOC code 31-1012). As the definitions in Appendix C indicate, health aides seem to have more technical tasks, which would justify putting them in a separate occupation group.

Although the occupational classifications used to categorize the health workforce have differed by dataset and varied over time, an announcement in the Federal Register Notice of September 30, 1999, indicated that all Federal agencies that collect occupational data are now required to use the 1998 Standard Occupational Classification. In addition, all State and local government agencies, as well as private sector organizations that gather occupational data are strongly encouraged to use the 1998 SOC. In the words of the announcement, "This national system ... provides a common language for categorizing occupations in the field of work."
The SOC Internet address is http://www.bls.gov/soc/.

Census Occupation Classification
The latest census occupational classification system was developed to be consistent with the 1998 SOC. It has 509 separate categories across the 23 major groups of SOC. Since the census codes are consistent with the SOC, it is also difficult to analyze census occupation data over time. Crosswalk between the census occupation codes and the latest SOC is available on the following web site. Compared to the SOC, the census occupation codes for direct care workers are not as detailed. For example, one code (2000 census occupation code 360) covers home health aides, nursing aides/orderlies/attendants, and psychiatric aides.
The census occupation classification’s Internet address is http://factfinder.census.gov/maetadoc/occupation.pdf.

Occupational Classification System Manual
The Occupational Classification System Manual (OCSM) is based on the 1990 Census of Population and Housing Classified Index of Industries and Occupations. The census index classified occupations into about 500 occupation classifications within 13 major group categories, whereas the OCSM has 11 major occupation groups. Currently, the NCS, which uses the OCSM, uses 9 of the 11 groups. The OCSM uses nearly all census occupations. In addition, the OCSM includes at least one not elsewhere classified (NEC) occupation within each group. NCS also adds the corresponding major occupation group alpha code to a 3-digit occupation code to establish a 4-character occupation code. The numeric codes correspond to the census code. Because the OCSM codes are consistent with the census occupation classification, they share common problems, including the inability to separate nurse aides, orderlies, and home health aides.

The OCSM Internet address is http://www.bls.gov/ncs/ocs/ocsm/comuseocsm.htm.

Industry Categories

Standard Industrial Classification
The U.S. government established the SIC system in the 1930s to promote uniformity and comparability of data various levels of government, trade associations, and research organizations collected and published. Although the overall structure of the SIC remained essentially unchanged since the establishment, the government has revised the SIC periodically to reflect changes in the U.S. economic structure. Such revisions include adding new industries and deleting or combining small or declining industries. As of the last revision in 1987, the SIC had 1,004 industries, of which 416 were service-related.

SIC is a 4-digit system that is structured as follows. The OES and occupation projections use the 3-digit SIC to classify industries as follows in Table 6-2.

Table 6-2. SIC Classifications

Although the SIC provides more detailed industry classifications than 1990 census codes, it still has several limitations, particularly in residential and community-based services. For instance, the SIC has a separate code for home health services while 1990 census does not. But for residential settings, the SIC only has one code (8361: residential care). This code includes not only residential care service providers for the people who need long-term care, e.g., assisted living, retirement homes, group homes for disabled, etc., but also places like boot camps, halfway group homes for juveniles, orphanages, and homes for unwed mothers. As for community-based care, home care of the elderly (SIC 8322) is mixed with other senior services, as well as completely different fields. For example, it is included with senior centers and adult day care, as well as adoption agencies, youth services, counseling services, food banks, and soup kitchens. Although further classifications by occupation may prevent misclassification of direct care workers in each industry category, inclusion of different industries in one group will make it harder to provide accurate pictures of workers.

The SIC Internet address is http://www.osha.gov/cgi-bin/sic/sicser5.

North American Industry Classification System
On April 9, 1997, the OMB announced its decision to adopt the North American Industry Classification System (NAICS) as the industry classification system U.S. statistical agencies will use. The NAICS replaced the 1987 SIC, which data users and analysts had criticized as being outmoded and unreflective of the U.S. economy. The NAICS accommodates such new industries as information services, health care services, and high-tech manufacturing. It includes 1,170 industries, of which 565 are service-based industries. Although few government agencies currently use the NAICS, it will become the uniform industry classification system across the Federal government. It also allows government and business analysis to compare industrial production statistics collected and published in the U.S., Canada, and Mexico. Each participating country can individualize the system to meet its own needs by using the 6th digit, as long as data can be aggregated to standard NAICS industries (5-digit).

While the SIC has a 4-digit system, the NAICS uses a 6-digit system for greater flexibility and international comparability. The NAICS structure is shown in Table 6-3.

Table 6-3. NAICS Classifications

Compared to the SIC and the census industry classification, the NAICS has more detailed categories, particularly for residential and community settings. For residential settings, the NAICS has separate classifications by whether or not nursing care is involved, as well as by resident population groups, e.g., the elderly, people with mental retardation, psychiatric and substance abuse. For community settings, the NAICS also separates services for the elderly and disabled from other population groups such as children and substance abuse patients. The detailed classifications in the NAICS give a potential for accurate understanding of workers in particular industries. However, depending on how detailed each survey program wants to be, i.e., what digit the program uses for classification, the detailed NAICS classifications may not be effectively implemented. For instance, OES starts implementing the NAICS in 2002. If OES decides to use the 3-digit classification, home health care services (621610) will be put together with medical laboratories (621510) and other outpatient care centers (621490) as ambulatory health care services (621). For those who study workforce issues in home health industries, this could become a problem. The NAICS Internet address is http://www.census.gov/epcd/www/naics.html.

Census Industry Classification
The 2000 census industrial classification system uses the NAICS structure. It consists of 265 categories in 20 sectors, which are the same as those in NAICS. The 1990 census industry classification uses the SIC structure.

A comparison of census industry classifications (1990 and 2000) and NAICS is available at http://www.census.gov/hhes/www/ioindex.html

In the latest census classification, there is a separate code for the home health care industry (2000 census code 817) that was not available in the 1990 classification.

However, definitions of residential and other community based programs are still problematic, because they include irrelevant industry settings, e.g., child guidance agencies, food banks, boot camps, and juvenile halfway homes.

Bridging Different Data Sources

Bridging Implementation Plan
The existence of different occupation and industry data collection systems in different government organizations presents a serious problem for policy analysts. Comparisons across programs are limited due to different definitions and classifications. In response to this problem, Federal government agencies are now shifting to uniform occupation and industry classifications. For occupation classifications, all Federal government agencies will adopt the SOC over the next few years. For industry classifications, Federal government agencies, including the Census Bureau and BLS, will start using the NAICS. The implementation schedule for some relevant programs is as follows in Tables 6-4.

By using uniform classification systems, it will be much easier to obtain workforce data from different sources. For example, one can find detailed wage data for nurse aides in skilled nursing facilities from OES; meanwhile one can also obtain demographic characteristics and work conditions for workers in the same occupation and industry groups from CPS data without having difficulty identifying corresponding occupations and industries. However, until the uniform classifications are implemented, bridging different data sources and definitions will still be necessary, as it will when working with historical data.

Table 6-4. Bridging Schedule

Bridging Definitions
As mentioned before, different occupation and industry classification systems have different definitions, and they do not always correspond to each other. The question becomes which occupation and industry codes should be used in each classification system to identify direct care workers most accurately?

Since most surveys will start using the SOC for occupation classifications and the NAICS for industrial classifications, it seems logical to use them as starting points. Tables 5-5 through 5-10 show occupation and industry codes in different classifications that correspond to the SOC and NAICS, although the match is not perfect. Depending on a researcher’s interest, he/she can use these bridging tables differently. For instance, if researchers want to study nursing aides, regardless of settings, they can focus on codes that correspond to 1998 SOC 31-1012, e.g., 2000 census code 360, 1980 SOC code 5233, 1990 Census code 447, ignoring any industry codes. If they want to focus on nursing aides in nursing facilities, they can further narrow the data by industry codes that correspond to NAICS 623110, e.g., 1987 SIC code 805, 2000 census code 827, 1990 census code 832. In any case, researchers must be aware of irrelevant settings and occupation groups that are currently included in each classification system.

Ideally, it would be possible to adjust the detailed definitions so that the employment estimates do not include irrelevant components.

Tables 6-5 through 6-10 point clearly to the need for standardizing the terminology, definitions, and taxonomies used to collect, maintain, and share data on direct care paraprofessional workers and the organizations and settings in which they work. The inconsistencies in and across the current data systems make systematic comparisons and analyses impossible. Even obtaining reliable estimates of the numbers of these workers is difficult at best.

Table 6-5. Bridging Definitions of Different Data Sources: Occupation

Table 6-6. Bridging Definitions of Different Data Sources in Hospital Settings

Center for Health Workforce Studies, 10/01

Table 6-7. Bridging Definitions of Different Data Sources in Residential Settings

Center for Health Workforce Studies, 10/01

Table 6-8. Bridging Definitions of Different Data Sources in Community Settings