Remarks of Christopher K. Northup

The U.S. Equal Employment Opportunity Commission

Public Hearing of October 29, 2003 on Proposed Revised Employer Information Report (EEO-1)

Thank you for the opportunity to speak at today’s hearing on the revised EEO-1 form. I hope that my comments will be useful as the Commission moves ahead with implementation of the revised form.

The remarks that I offer today focus on two main issues:

  • the three new Officials and Managers categories; and
  • the resurveying employee race and ethnicity

Three New Officials and Managers Categories

With the proposal that the Officials and Managers reporting category be divided into three subcategories: Executive/Senior level, Mid-level and Lower-level, there are three major issues that I would like to raise:

  • First, employers completing EEO-1 forms will face problems in classifying and reporting employees properly and consistently especially in these three new Officials and Managers categories which will diminish the accuracy and usefulness of reported data.
  • Second, the Executive/Senior Level category if implemented will have relatively few numbers that will limit its use beyond general trend analysis at the national level. In addition there could be limited data for small and medium size organizations such that it may become necessary to ignore the data or re-aggregate to a broader category.
  • Third and last, the 2000 Census detailed occupational categories, provided to guide the respondents, are less useful and potentially confusing for some of the three Officials and Managers categories but helpful and useful for the other eight categories.

Classifying and reporting employees properly and consistently

I would like to share with you my observations in working with many employers both large and small over the last twenty years. In order for employers to report the current EEO-1 data it is necessary for someone in the organization to first classify each job or position into one of the nine occupational categories found on the current EEO-1 form and thus allowing the employee data to be summarized and reported out for use on the form. The person responsible for this coding could be an HR professional involved in staffing, recruiting or EEO but just as likely could be an administrative assistant or clerical type.

Many classifiers, especially the clerical workers, have never looked at the EEO-1 instruction booklet nor do they have an understanding of how the EEO-1 form information is used. Even for those familiar with the instructions because there is very limited information as to how to classify many jobs is provided one must be able to exercise their judgment when classifying jobs into EEO-1 categories. Based on my review of many organizations’ detail EEO-1 coding I have observed numerous errors either due to misunderstanding, lack of concern, or perhaps keying errors that are never corrected.

The point I would like to make is that the existing structure within organizations to classify jobs into EEO-1 categories is not always reliable, and if you now require these same persons to have to further distinguish within the officials and managers category, I have serious concerns about how accurately and reliably the coding efforts will be. Classifying of jobs into the executive/senior level is probably not as much of a concern but properly classifying mid-level and lower-level officials and managers will be especially challenging to many employers. Some of the challenges will be:

  • In many larger organizations you may have managers in similar positions or functions, yet their level of responsibility could easily differ depending on whether they are at the corporate headquarters, a large establishment or a smaller establishment.
  • In smaller organizations, employees may have responsibilities that place them in both the lower and mid-level categories as they tend to have multiple roles and duties.
  • Therefore, given the concern for the person doing the coding and that process, as well as the potential difficulties in distinguishing between lower and mid-level categories, I seriously question the validity of the reported data for these categories upon which the Commission and other agencies such as the Office of Federal Contract Compliance Programs will be relying. However, if the Commission does move ahead with separate Officials and Managers categories, consideration should be given to:

  • requiring respondents to provide documentation and explanations, as appropriate in order to support the categorization of their job titles into the EEO-1 occupational categories.
  • providing brief training sessions (for example, via the web) and more instructional material (for example, questions and answers) to educate persons responsible for classification.

In fact even without any changes to the EEO-1 form the above action would certainly improve the reliability of the currently reported data.

Executive/Senior-level category will lack sufficient numbers

With regard to the second issue of the numbers reported in the Executive/Senior-Level category, I raise the concern that there are relatively few employees in most organizations that fall into the upper category. Certainly at larger organizations there are more such employees but at smaller organizations there will be many fewer.

If you assume the reporting universe for the EEO-1 forms is approximately 40,000 employers and on average they have between 5 and 10 employees classified in the Executive category, this would result in only 200,000 and 400,000 employees. While the Executive category data may be useful for analysis on a national level, with the other Officials and Managers categories for small and medium size employers where one would compare employers using sub-national geographic areas, the numbers will become more fragmented. Further fragmentation would occur when segmenting the data by industry or by employer size. All of this potential fragmentation may in fact require re-aggregating the data back into one or two Officials and Managers categories for any meaningful analysis.

While employers that I have worked with will be willing to comply with whatever the Commission ultimately decides, it still raises the question of whether the additional effort to collect this information is offset by the potential benefit of these data. 2000 Census occupational categories less useful and potentially confusing for some categories My comments regarding the third issue that I raise concerning the suggested ranges of 2000 Census categories are based on my thoughts that for some of the EEO-1 categories they are less useful and potentially confusing for these three O&M categories. The inclusion of the 2000 Census codes suggests that one can use these to clearly sort the organization’s jobs into one of the categories. I think one encounters the same problems that I raised with the accurate classification of jobs that are similar but located in different parts of the organization. It would be helpful to consider the following example:

  • The Director of Information Technology at Corporate might be assigned to the 2000 Census category 011 Computer and Information Systems Managers.
  • At one of the organization’s smaller plants an Information Technology Manager with lesser responsibility and pay might also be coded as 011 Computer and Information Systems Managers.

Both of these jobs might require many of the same skills and though the Director might be paid more, it may not be significantly different. The corporate position might be considered a mid-level O&M based on the suggested description. The plant IT Manager might be considered either a mid-level or a lower-level O&M type job given the responsibilities. However, the EEO-1 guidelines for the revised form suggest that both of these positions should be reported in the mid-level Officials and Managers category. This ambiguity will make it difficult for the employer and will result in inconsistent data reported to and used by the Commission. There is also another concern with regards to the presentation of the census categories. For the Lower-level Officials and Managers category census occupational codes such as 081-095 are associated with this category. While the description of this category states that these employees are responsible for day-to-day operations and supervising other personnel, in fact some of persons in the census occupational categories referenced would in many cases be more properly considered “professionals” in that they require a college degree and do not always have primary supervisory responsibilities. For example, 082 Budget Analysts, 083 Credit Analysts or 084 Financial Analysts could very well be non-supervisory positions requiring a bachelor’s degree or perhaps an MBA. In such cases they would belong to the Professionals EEO-1 category rather than the Lower-level Officials and Managers category. Again this is another area of ambiguity that could confuse employers and contribute to reporting inconsistencies. While I have raised concerns about how the census occupational categories are used as guidance for the three Officials and Managers categories, I do believe that for the remaining eight categories they are helpful and useful to employers. I have often used the census occupation codes and their corresponding cross-referenced EEO-1 categories to validate classification work performed by my staff or the employer.

Resurveying Employee Race and Ethnicity

When employers begin making the necessary changes and collect data to report on the revised EEO-1 form in the next year or two, it is important to consider whether or not the Commission should require employers to resurvey employees and provide them with a new opportunity to self-identify their race and ethnicity. There are a number of issues that I would like to bring to your attention to be considered when finalizing the Commission’s decisions in this area:

  • 2000 Census results indicate that relatively few people chose to identify two or more races, and are classified as Native Hawaiians.
  • Additional efforts will be required by employers to determine the race and/or ethnicity for those employees who decline to self-identify or otherwise fail to respond.

Relatively few multi-race and Native Hawaiians reported

As many of you are already aware in the 2000 Census only 2.4% of the U.S. population chose to identify themselves as two or more races. If persons who identified themselves as of Hispanic origin, the two or more races category drops to only 1.6% of the population. The Native Hawaiian and Other Pacific Islander category account for only 0.1% of the population. These statistics are significant in that they show with the addition of the two new race and ethnicity categories on the revised EEO-1 form the impact of such a change affects relatively very few people, potentially only 1.7% (1.6% of the non-Hispanic two or more races, and 0.1% of the Native Hawaiians segment.)

To illustrate what this means for employers of different sizes, an employer with

  • 500 employees might expect on average that less than 9 employees might self-identify as two or more races, or as Native Hawaiian.
  • 1,000 employees might expect on average only 17 employees would self-identify as two or more races, or as Native Hawaiians.
  • An even larger workforce of 50,000 employees only 850 employees might self-identify as such.

Even for those employees who are reported in the multi-race or Native Hawaiian category, their presence in the employee population would be so small that it is unlikely that any meaningful statistical analysis could be conducted for these categories.

Follow-up efforts for employees who do not self identify

If employers elect to resurvey or are required to resurvey, they will need to offer their employees the opportunity to self-identify one or more races and Hispanic origin. While it is likely that the majority of employees will comply with such a request from their employer, it is also likely that there will be some who fail or decline such an opportunity. Since the employer is required to categorize all employees with regard to their race and ethnicity for reporting purposes, the employer will be faced with having to make the determination without the assistance of the employee. If the non-response rate were only 10%:

  • an employer with 1,000 employees would be faced with having to expend additional effort to categorize 100 employees; and
  • an employer of 10,000 employees would have 1,000 employees with which to follow up Given the small HR staff at many organizations especially in smaller companies, the initial effort to plan and rollout a resurveying effort and then address the non-respondents will take significant time and resources. At many organizations the HR staff is already overloaded with day-to-day responsibilities and numerous projects outside of the EEO/AA area, so that the resurvey would in fact be a significant burden to many of them.
  • Therefore, I would ask the Commission to consider with regard to the resurveying issue is:

  • Allow employers to optionally resurvey their employee population. If the employer chose not to resurvey the employer would use the current race and ethnicity of each employee for purposes of the initial revised EEO-1 filing. As the employer updates employee profiles or offers other opportunities to employees to make changes to their personal data, employers can still offer the employee an opportunity to self identify including the choice of multiple race categories.
  • If resurveying of race and ethnicity is mandatory then allow employers the option of using the current race and ethnicity information for those employees failing to self identify as maintained in their Human Resource Information System or other such personnel recordkeeping system.

In either case, newly hired employees would receive the opportunity to self identify under the full range of race and ethnicity options.

Again I would like to express my gratitude for allowing me to be present at today’s hearing.


This page was last modified on October 29, 2003.