The Advantage

April 2004

The Census 2000 Special EEO File is Finally Here

Anyone wishing to have access to the new file may go to the Census Bureau’s website at www.census.gov. The direct link to their 2000 Special EEO File is located at www.census.gov/hhes/www.eeoindex.html. There are data search tools on the Census web site to help you locate the information you need. Source files are also available on the Web at www.eeoc.gov/stats/census/availability.html. Occupational codes and industry code references are available at www.census.gov/hhes/www/ioindex.html.

In some cases it is reasonable to use EEO-1 data rather than specific occupational classification data. If you wish to access EEO-1 industry profiles that have been updated through 2001, you will find them for the United States total population, by states, by major metropolitan areas, and by industry. Industrial data comes as a national representation only. Go to www.eeoc.gov/stats/jobpat/2001/index.html. This information is updated after all the current year EEO-1s have been processed.

------------------------------------
David Rainwater is Vice President of Employment Advisory Services, Inc., a consulting firm associated with the Washington, DC law firm of McGuiness, Norris & Williams, LLP, and the Equal Employment Advisory Council (EEAC). He can be reached at drainwater@easiconsultants.com.

Now We Know…"Who is an Applicant?"

It has been several years since the project began, but it has ended with the publication of a 12-page Federal Register posting. We now know what the Federal agencies believe is an applicant...at least when that applicant comes to an employer through the Internet or related technologies.

You are invited to submit your comments about this new proposal. Deadline for your written response is May 3, 2004. You will find the entire posting in the Federal Register, March 4, 2004, Volume 69, Number 43, pages 10152-10158. Comments should be submitted to Frances M. Hart, Executive Officer, Executive Secretariat, Equal Employment Opportunity Commission, 10th Floor, 1801 L Street, NW, Washington, DC 20507. Comments of six or fewer pages may be FAXed to 202-663-4114.

You all know about the Uniform Guidelines on Employee Selection Procedures (UGESP or Uniform Guidelines). They have been around since 1978 and spell out the requirements for all employers who are subject to the Civil Rights Act of 1964 to assure their employment screening devices are non-discriminatory. They specify the statistical testing employers must undertake to assure their written employment tests, interviews and other such screening tools do not have an adverse impact on a protected group of people. The Uniform Guidelines are codified as 41 CFR 60-3. UGESP states the employers should maintain "records or other information which will disclose the impact which its tests and other selection procedures have upon employment opportunities of persons by identifiable race, sex, or ethnic group."

Because the Uniform Guidelines raised so many questions, the government decided to publish a group of Questions and Answers to help explain what was meant by these new validation requirements. You can get a copy of the current Q & A's at www.management-advantage.com/media/UniformGuidelinesQ_A.pdf If you would like a FREE copy of the current Uniform Guidelines go to www.management-advantage.com/media/UniformGuidelines.pdf

Since 1978, the employment world has changed a great deal. It is rare these days for employers to have physical locations called "employment offices." And, it is more and more common to find employers soliciting potential employees from sources on the Internet. In 1978, the World Wide Web had yet to be invented.

The question has arisen, "Who is an applicant?" because when people had to appear in person and fill out a job application form, it was fairly clear who an applicant was. Now, with Internet tools like auto responders, broadcast capability and so forth, individuals are able to submit their names to companies even if they don't look at a job requisition announcement. So, there is a substantial increase in unsolicited resumes today. Companies are receiving thousands of submissions today compared to a small fraction of that amount in the past.

The Uniform Guidelines make it clear that "applicants" as a group are the foundation of all statistical analysis in disparate impact testing. If you aren't sure how to identify an applicant, it makes the process of statistical analysis difficult to impossible.

Enter employers again, telling the government that the old way of identifying an applicant would no longer work in today's reality. Then, the agencies responsible for the Uniform Guidelines agreed to redefine "job applicant" so everyone would know what pool of people to base the statistical computations upon. Those agencies were the Equal Employment Opportunity Commission (EEOC), the U.S. Department of Labor, Office of Federal Contract Compliance Programs (OFCCP), the U.S. Department of Justice (DOJ) and the Office of Personnel Management (OPM). They had originally produced and put their signatures to the 1978 regulations, so they were the agencies charged with creating any updates required.

The original charge came several years ago. And, the agencies in their conversations with one another were unable to agree on the definition of job applicant in the new employment environment. So, discussions went on. And, on. And, on. Deadlines were set by the Office of Management and Budget (OMB) for the group to submit its work. Deadlines were extended by OMB again, and again, and again. Until months turned into years, and the task force still had not reached agreement.

After all that, we now have an agreement among the agencies. Rather than change any of the Uniform Guidelines or existing Questions and Answers, the Task Force has elected to add another set of Questions and Answers as a supplement to the original.

The problem for employers has been trying to identify race and sex for job applicants that come through the Internet. Shear volume of submissions has overwhelmed employers. According to the new posting, "...a major health care employer received 300,000 online resumes in one year. A smaller Pennsylvania employer reported that it received 6,000 to 8,000 resumes a year before going online, but began receiving about 24,000 resumes a year since it went online...After commencing recruitment on the Web, another employer began receiving 20,000 to 40,000 resumes annually, many of which were unsolicited."

Questions that employers have been wrestling with include:

• Do unsolicited resumes count as applicants?
• Do we have to invite people who send unsolicited resumes to identify their race and sex?
• Do we have to count people who send resumes, applications or statements of interest if we don't have a job opening?
• Do we have to include people who aren't interviewed?
• Do we have to include people who aren't qualified?

These and many other questions have been answered by employers based on their own procedures as directed in the original Question and Answer #15. The question was: "What is meant by the terms 'applicant' and 'candidate' as they are used in the Uniform Guidelines?" The answer was, "The precise definition of the term 'applicant' depends upon the user's recruitment and selection procedures. The concept of an applicant is that of a person who has indicated an interest in being considered for hiring, promotion, or other employment opportunities. This interest might be expressed by completing an application form, or might be expressed orally, depending upon the employer's practice."

"The term 'candidate' has been included to cover those situations where the initial step by the user involves consideration of current employees for promotion, or training, or other employment opportunities, without inviting applications. The procedure by which persons are identified as candidates is itself a selection procedure under the Guidelines."

"A person who voluntarily withdraws formally or informally at any stage of the selection process is no longer an applicant or candidate for purposes of computing adverse impact. Employment standards imposed by the user which discourage disproportionately applicants of a race, sex, or ethnic group may, however, require justification. Records should be kept for persons who were applicants or candidates at any stage of the process."

Remember, this was written before the advent of email, resume databases, job banks, electronic scanning technology, applicant tracking systems and applicant service providers. One other thing to remember is that there is a difference between "recruiting" and "selection procedures." There are five additional questions and answers proposed by this filing. (Questions 94-98). In Question 95, the proposal says: "As a business practice, recruitment involves identifying and attracting potential recruits to apply for jobs. Under USESP, 'recruitment practices are not considered ...to be selection procedures,' and the UGESP requirements geared to monitoring selection procedures do not apply. Just as recruiters traditionally researched paper copies of professional and employer publications and listings to identify potential recruits, so recruiters now search huge bodies of information online -- which include new resources such as personal Web sites and a variety of resume databases -- for the same purpose. Online recruitment also involves organizing the search results into usable formats."

Clearly, recruits are not job applicants ... just yet anyway.

Then we get to the heart of the matter. The new Question 96 asks, "For recordkeeping purposes, what is meant by the term 'applicant' in the context of the Internet and related electronic data processing technologies?"

And, the answer is ...

"...In order for an individual to be an applicant in the context of the Internet and related electronic data processing technologies, the following must have occurred:

1. The employer has acted to fill a particular position;
2. The individual has followed the employer's standard procedures for submitting applications; and
3. The individual has indicated an interest in the particular position.

"To elaborate on the three prongs of this test:

1. The employer has acted to fill a particular position.

   An example under the first prong is:

   Example A: Individuals who register online for Customer Service Representative positions with an Internet and cable television service provider are asked to complete online personal profiles for the employer's resume database. The company acts to fill two vacancies at its Greater New York Service Center, and identifies 200 recruits from the database who have indicated that they are available to work in the New York area. One hundred of these people respond affirmatively and timely to the employer's inquiry about current interest in the particular New York vacancies. Even if the employer chooses to interview only 25 people for the position, all 100 are UGESP 'applicants.'

2. The individual has followed the employer's standard procedures for submitting applications.

   If everyone who applies online must complete an online personal profile, only those individuals who do so can be UGESP applicants. If job seekers must use an electronic kiosk or contact a store manager to apply for a sales position, only those who do so can be UGESP applicants. If an employer e-mails online job seekers to ask if they are currently interested in a particular vacancy, only those who meet the employer's deadline can be UGESP applicants. These procedures and directions must be nondiscriminatory because recruitment and the application processes are subject to Title VII and Executive Order 11246.

3. The individual has indicated an interest in the particular position.

   The core of being an 'applicant' is asking to be hired to do a particular job for a specific employer. An individual can only accurately assess her interest in an employment opportunity of which she is aware.

   With respect to Internet recruiting, this means that people who post resumes in third party resume banks or on personal Web sites are not UGESP 'applicants' for all employers who search those sites. By posting a resume, the individual is advertising her credentials to the world and indicating a willingness to consider applying for new positions that may be brought to her attention. The individual is not indicating an interest in a particular position with a specific employer. If an employer contacts this individual about a particular position after finding her resume or personal profile online, and the individual indicates an interest in that position, then the individual becomes a UGESP 'applicant,' if she also meets the second prong of the test set forth above. Similarly, if an employer contacts an individual about a particular position in response to an unsolicited resume submitted online, and the individual indicates an interest in that position, then the individual becomes a UGESP 'applicant' if she also meets the second prong test.

   Furthermore, even if an individual expresses an interest in a whole category of positions in response to an employer's solicitation -- for example, marketing opportunities -- the individual is not an applicant but is identifying the kinds of positions in which she may be interested. She is not indicating an interest in a particular position with a specific employer. It is only with respect to a particular position that an individual can assess her interest and choose whether or not to apply."

Question 97 asks whether or not search criteria employers use to pull people out of a database are subject to disparate impact analysis. The answer is, "Yes." Search criteria are a selection process.

Question 98 asks if online tests required for individuals to be considered for a job opening are a selection process. The answer again is, "Yes."

So, there you have it. We now know what an applicant is when it comes through one of the new technology sources. All employers who are subject to the Uniform Guidelines are obliged to request race and sex identification from job applicants. How an employer does that is left to the employer. The only requirement is that it be a non- discriminatory procedure and the race/sex information is not made available to anyone involved in the selection process.

Watch Out for the New OFCCP Approach to Compensation Analysis

For several years, federal contractors have been at odds with the Office of Federal Contract Compliance Programs (OFCCP) over the question of accurate compensation program analysis. OFCCP proposed what became known as the Dubray approach, testing each individual against every other individual who was considered a cohort. Averages were compared within salary grades, and if the contractor happened to be using broadband salary grades, wide disparities usually surfaced. OFCCP attempted to obtain settlement agreements and back pay awards on each such analysis result.

For just as long, federal contractors have been saying that the only legitimate statistical test for compensation analysis could be found in regression analysis.

Well, Director Charles James has apparently agreed that the agency needed to take a different approach in its attempt to uncover discriminatory practices involving compensation. So, he hired a Ph.D. statistician at OFCCP headquarters to begin the process of guiding the agency into legitimate means of discrimination testing using statistical data.

OFCCP is also in the process of hiring a similarly qualified statistician for each of its six regional offices. These folks are going to develop and implement some rather sophisticated analysis tools and then train the Compliance Officers how to use them.

We can already see the influence of this effort by the agency. In a recent Conciliation Agreement between OFCCP and one of its federal contractors, the issue focused on compensation disparity. Among other things, the contractor was asked to agree to terms such as:

Re-analyze its compensation system and regroup jobs into "reasonable salary or compensation grades or groupings based on relevant and appropriate factors."
• Make one-time salary adjustments once the regrouping of jobs has concluded and disparities identified.

As part of this analysis, the contractor was required to develop Job Groupings and to consolidate job groupings when the total number of employees fell below 30 in any group. Those were then called Consolidated Groupings.

"For each single Job Grouping of 30 or more employees or for each Consolidated Grouping, as applicable, in which the average salary for either protected group (i.e., females or minorities) is below the average salary for the relevant reference group (i.e., males or whites), [the contractor] will evaluate the Job Grouping or Consolidated Job Grouping using a t-statistic. This t-statistic analysis will be performed using a t-test for any Job Grouping evaluated that has 30 or more employees. For any Consolidated Groupings evaluated, this t-statistic analysis will be performed using ordinary least squares regression relating salary to a protected group indicator and indicators for each of the component Job Groupings."

"The t-statistic analysis indicates that, for any Job Grouping or Consolidated Grouping evaluated, the difference between the average salary of the protected group (females or minorities, as applicable) and the average salary of the reference group (males or whites, as applicable) equals or exceeds 2.0 standard deviations in absolute value, [the contractor] shall address the disparities in the Job Grouping or Consolidated Grouping by either:

1. Making salary adjustments within the Job Grouping or Consolidated Grouping sufficient to bring the average salary of the relevant protected group to within 1.0 standard deviation of the average salary of the reference group; or
2. Performing a least squares regression analysis to analyze the relevant disparity (females v. males or minorities v. non- minorities, as applicable)."

"If [the contractor] chooses to perform a least squares regression analysis to address the disparities noted in the t-statistic analysis, the regression will relate salary (in dollars) to a protected group indicator (a variable that is 1 for members of the protected group and 0 otherwise), indicators for Job Groupings, and additional factors that may affect an employee's salary and that are appropriate to the data. If the regression analysis indicates that there is a disparity disfavoring the protected group (i.e., female or minority, as applicable) that equals or exceeds 2.0 standard deviations, [the contractor] shall:

1. Make salary adjustments sufficient to bring the average salary of the relevant protected group to within 1.0 standard deviation (using the least squares regression model) of the average salary for the reference group; or
2. If [the contractor] determines that there are legitimate non-discriminatory reasons that are causing or have caused the disparity, provide OFCCP with particularized information in support of those legitimate non-discriminatory reasons."

"If [the contractor] determines that adjustments will be made as a result of the t-statistic or least squares regression analysis, it shall calculate the total amount of the adjustment using the following formula: ADJ = Np(DIF - SE), where ADJ is the total amount of the adjustment, Np is the number of members in the relevant protected group (women or minorities), DIF is the average pay gap, and SE is the standard error of the DIF. The total amount of the adjustment will be distributed proportionally to those individuals in the protected group (women or minorities) whose salaries are below the average of the relevant reference group (males or whites) based on the following formula: AMT = ADJ x PCT, where AMT is the amount each woman or minority whose salary is below the average will receive, ADJ is the total amount of the adjustment as calculated above, and PCT is the percentage of the adjustment each such woman or minority will receive, calculated by dividing the amount of that person's pay gap (average salary of the relevant reference group - salary of that person) by the sum of all the pay gaps of the women or minorities whose salaries fall below the average for the relevant reference group."

If you have been in the EEO/AA business for a while, you will notice that OFCCP is requiring the contractor in this Conciliation Agreement to bring compensation rates within one standard deviation of the comparative group. We have typically used a two standard deviation threshold. This change will make such settlements much more costly for contractors who accept them.

There is legal precident for the two standard deviation comparison. And, it turns out, there may also be a legal precident for the one standard deviation comparison. In 1992, General Motors Corporation entered into a Consent Decree that used one standard deviation as its threshold.

When it comes to Conciliation Agreements, federal contractors had better plan to have a statistician join in on the legal review process. Whatever you do, don't allow your CEO to sign such an agreement without having it reviewed by these specialists and getting from them a simple explanation of the work involved in the implementation of such an agreement. It could be substantial.

A Gentle Word on Behalf of Our Business

When you need help developing your affirmative action program, give us a call. We specialize in AAP development, implementation training and compliance review support for clients all over the country. Find out more about our AAP development service by looking at our sample agreement and other information. You will find it all at http://www.management-advantage.com.

You wouldn't go to an IRS audit alone. Why think about going into a Department of Labor compliance review without professional support? The stakes are just as high either way.

We are ready to give you the support you need.

And while you're at it, think about ordering a copy of our reference and training book on preparing affirmative action plans and managing compliance reviews. You will find it an invaluable resource at a price that just can't be beat.

Secrets of Affirmative Action Compliance, new 6th edition, contains over 500 pages of the latest and current regulation requirements and practical suggestions for your organization. Includes new Federal Regulations and rules for federal construction contractors. $99.95 plus $9. shipping/handling and CA sales tax for CA destinations. Credit Card Orders ... Call Toll Free:

We can help with your other human resource management needs as well. Think of us the next time you need:

• Employee Handbooks
• Management Training in Compliance Issues
• Affirmative Action Plan Development
• Affirmative Action Statistical Analysis
• Disparate Impact Testing for New Hires, Promotions, Transfers, Terminations
• Expert Witness
• Books, Software or Other Support Materials for HR Professionals

YOUR FEEDBACK IS INVITED

Thanks for taking the time to read our newsletter. We would enjoy receiving your thoughts about its value to you. You can e-mail your message to tmainc@management-advantage.com

or simply give us a call and tell us in person. Our office number is 925-671-0404. We appreciate your feedback.

If you ever wish to unsubscribe, please send an email message to: majordomo@management-advantage.com and in the body of the message write: "unsubscribe theadvantage (your email address)." Do not use the parentheses or quotation marks. No other message is required. Your address must match the one you used to subscribe. If it does, you will be automatically removed from our subscriber list.