Adverse Impact
Adverse impact occurs in a personnel action when the selection rate of a one class of applicants or employees falls below 80% of the selection rate of a complement class. Traditional adverse impact considers the selection rate of a protected class (i.e., race or gender) against the complement non-protected class.
Affirmative Action Plan
From the Code of Federal Regulations 41 Section 60-2.10, an Affirmative Action plan is “a set of specific and results-oriented procedures to which a contractor commits himself to apply every good faith effort.”
An Affirmative Action planplan, prepared annually, consists of statistics and narratives, in the format specified by the Office of Federal Contract Compliance Programs (OFCCP). The statistics look at both a “snapshot” or moment in time around the plan date and a period of time, usually 12 months, leading up to the plan date. The Agency performs periodic audits of contractors to review these documents.
Update plans for interim periods are developed to measure progress, based on information provided in the Annual plan.
Any Difference Rule
For a protected class in a job group, a placement goal is set when availability exceeds
employment by any difference.
Applicant
A person considered for a job within an organization. An applicant for the purpose of Affirmative Action reporting should be defined to ensure conformance to the regulations and to comply with organizational policy and
procedures.
Availability
The percentage of a protected class, calculated to be available for employment openings within a job group at the facility. Data collected by the U.S. Census Bureau are used to compile these demographics by occupation and geographic area.
Census Occupation Codes
Standardized occupation categories, contained in the U.S. census data set. For the accuracy of an AAP, each job within an organization must be matched to a census occupation code.
Company Name
The Company Name will appear at the top of all reports printed in balanceAAP. Location names or information can e provided and included, separately.
Comparator
A comparator is a member of the non-protected class who is successful in being hired, promoted, transferred, not terminated, or receiving a salary above the median). Protected class members who were not successful are then compared to the comparators in a Cohort Analysis.
EEO Category
EEO Categories are job classifications published by the Equal Employment Opportunity
Commission. Most federal contractors use the EEO-1 schedule. However, governmental organizations use the EEO-4 schedule and educational institutions use the EEO-6 schedule. Use EEO report instruction booklets as a guide in categorizing jobs but do not be afraid to be flexible when assigning categories to meet the needs of the organization. In general, positions categorized in the EEO-1 Officials and Managers category should include organization officers, whether they supervise people, work or accounts, and all supervisors and managers properly exempted from overtime payments. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 254 Flexibility is a must when assigning EEO Categories due to their importance in the development of job groups. For example, an employee who works on a personal computer may be listed as a Computer Operator in the Technical Category if the employee is working with specialized software evaluating engineering or other technical information, or in the Clerical category if he or she is inputting information or using word processing and spreadsheet software. To avoid confusion, make it a point to use different titles and job codes for similar positions in different EEO categories. Similarly, the position Maintenance Mechanic may be listed as an operative or craft position, depending upon the skill level required to perform the position. Certain Administrative Assistant positions may be classified as clerical or professional positions, again depending on pay status and required education and experience. Eighty Percent Rule A placement goal is set when 80% of availability is greater than employment. Formula: (.8 × availability percentage) is more than employment percentage Example: Availability = 26.8% Employment = 23.5% (.8 × 26.8 availability percentage) = 21.4 21.4 is NOT more than 23.5 employment percentage In this case, no placement goal would be set. While the availability did exceed employment, 80% of availability did not exceed employment. Employment The number or percentage of minorities or females among those employed at a facility. Protected class employment percentages are compared to availability percentages in the Utilization Analysis. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 255 Errors Errors are records containing blank or invalid data. Examples of errors include roster records with blank or missing race, sex, job, or department codes. Duplicate records are also errors. Records with errors are excluded from reports. Executive Order 11246 Signed by President Johnson, Executive Order 11246 is the basis for federal contractors’ affirmative action obligations. It also applies to most banks and contractors and subcontractors on construction projects financed in whole or in part by federal funds. Originally it covered only minorities. A later executive order, 11375 which covered women, has since been combined into it. One evidence of this is the different wording for Factors 1a and 1b. Revised Order 4, based on E.O. 11246, requires contractors and subcontractors with 50 or more employees and a contract of $50,000 or more to develop and carry out a written Affirmative Action Program. Facility Within an organization or a set of related organizations, a location or a set of locations sharing an immediate hiring area. In general, separate affirmative action plans should be prepared for each facility of 50 or more employees. If you are unsure on defining your facility and the employees included in the plan, call Berkshire Associates Inc. at (800) 882-8904 or consult a labor attorney experienced in affirmative action plans under Executive Order 11246. Factor Weights In a weighted analysis, a weight is a statement of how important the factor being weighted is to the whole. A weight of .300 indicates that this factor makes up 30% of the whole. Note that the weights must add up to 1.000. BALANCEaap automatically sums your weights and warns you if you try to leave a job group with an invalid weight sum. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 256 Feeder Jobs Staffed positions that can provide a normal promotion path into a particular job group. Feeder jobs to a particular job group must, by definition, be from a different job group. File Consistency Issues File Consistency Issues are logical inconsistencies in your personnel actions files. One example is an employee whose job in the New Hires file is different from his/her job in the Roster, with no Promotion or Transfer to explain the switch. Another common example is an employee found in both the Roster and the Terminations file. File Consistency Issues may well be explainable. In the second example above, the employee may have been rehired. Thus it is your choice whether or not to correct File Consistency Issues. Fisher’s Exact Test Fisher’s Exact Rule is a test of statistical significance, similar to standard deviation, that is more appropriate for small groups. Fisher’s Exact scores of .025 or less are generally regarded as statistically significant. Placement Goals If utilization in a job group for a protected class falls below availability to the level defined by the Utilization Rule selected, a placement goal will be set for that group equal to its availability percentage. History Data History data includes employment and availability numbers from the previous year’s affirmative action plan. Previous year’s employment numbers are needed to determine the number of incumbents for promotion, transfer, and termination impact ratio analyses. Previous year’s availability numbers are needed for the Goal Attainment and Share of Opportunities reports. Job Group One job or a group of jobs having similar content, compensation or opportunity. Usually an EEO Category or a subset of an EEO Category. Job groups are the fundamental unit BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 257 of analysis in an affirmative action plan. Reports such as the Job Group Analysis, Availability Computation, Utilization Analysis, Goals, Impact Ratio Analysis, and Goal Attainment are organized by job group. Lines of Progression Formal or standard unbroken career paths within the confines of one department. Narratives Two narrative reports are required in your affirmative action plan: one for minorities and females and one for the disabled and veterans. The narrative reports describe your organization’s policies and practices with regards to affirmative action. OFCCP The Office of Federal Contract Compliance Programs is a part of the U.S. Department of Labor and has overall responsibility for ensuring that government contractors, subcontractors and certain financial institutions comply with federal regulations including those concerning affirmative action. Personnel Actions Personnel actions monitored in BALANCEaap include applicants, hires, promotions, transfers, and terminations. In addition, two user-defined personnel actions can be tracked. Personnel action data for the calendar year preceding your plan date should be included. Plan Name The plan name is typically something like “January 1, 2003- Annual Affirmative Action Plan.” For an update plan, a typical name is “July 1, 2003- Update to Annual Affirmative Action Plan.” BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 258 Promotion Calculations An alternative to using identified feeder jobs; promotion calculations reflect the employment statistics of those jobs from which promotions actually occurred. Promotion data for the calculations can be drawn from the current year promotion file or the promotion history file. Protected Class A category of employees covered by provisions of regulations or statutes. In preparing affirmative action plan statistics, the protected classes of Females and Minorities are required to be analyzed. BALANCEaap allows you to analyze any or all of: Blacks, Asians/Pacific Islanders, Hispanics, American Indians/Alaskan Natives minority subgroups in addition to Females and Minorities. These too can be referred to as “protected classes.” Veterans and the disabled are usually covered in a separate plan. No statistical analyses are required for these protected classes. Recruitment Area When using 1990 or 2000 Census data, recruitment areas can be places of 50,000+ people (1990 only), counties/townships, MSA’s, PMSA’s (1990 only), CMSA’s (1990 only), states, the entire U.S. summary, or weighted or unweighted combinations. Each job group can use one census area as its Reasonable Recruitment Area in determining availability. Research Lists Research Lists are audit support tools used to help analyze and explain any adverse impact in the personnel actions or the salary analysis. Research lists include the members of a protected class who were adversely affected (not hired, not promoted, not transferred, terminated, or whose salary is below median), protected class members who were successful, and successful and unsuccessful non-protected class members. BALANCEaap allows you to fill in a comment for each employee or applicant in the research list. You can use these comments to provide explanations or extenuating circumstances, or other relevant information. The comments will print out on the Research List roster. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 259 Rule of Nines A test for the existence of “small groups” used in the Significance of Incumbency vs. Estimated Availability report. Where small groups exist, the validity of a large sample estimator like standard deviation is in question. The formula is the total number in the group multiplied by the protected class availability multiplied by its complement. The result should be greater than nine to rely on standard deviation as a test of significance. Example: If a job group has 76 employees and the availability for Hispanics is 7.8%, the formula would be 76 x .078 x (1 - .078) = 5.47 In this instance, the validity of standard deviation to measure the significance of any differences between Hispanic availability and employment would be questionable. The use of this rule is infrequent because it is not well-understood and it may establish too high a “small group” threshold. For example, under the Rule of Nines any job group with less than 36 employees will score less than 9. Some statisticians would set a lower absolute threshold. The rule of thumb for non-statisticians is to rely more on common sense than statistics when assessing situations where a placement goal has been set or adverse impact has been calculated. Salary Analysis The Salary Analysis is actually a series of analyses that can identify salary inequities within jobs, job groups, or salary grades. BALANCEaap’s salary analysis reports represent the latest in OFCCP audit techniques – the reports generated by BALANCEaap are the same as those prepared by the OFCCP when auditing your plan. The Salary Analysis is a useful audit support tool, and can be especially effective when used in conjunction with BALANCEaap’s Research List tool. Salary Grade Salary grades are groups of jobs with similar compensation structures. Salary grades are used to sort jobs in the Work Force Analysis and Job Group Analysis. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 260 Salary Growth Rate Salary growth rate is the annual percentage growth rate since the hire date. To compute salary growth rate, the current salary, starting salary, and hire date are required. Salary growth rate is computed using the “compound” method, according to the following formula: growth rate = (current salary/starting salary)^(1/y) - 1 (y = number of years since hire date) Sector Reports Sector or Communication Reports are management tools that can be used to show individual department, division, or cost center heads how they are doing. A plan can be broken into sectors by departments. Sectors can also be defined in terms of other sectors, allowing for “roll-ups” into the total plan level. Sector reports all use the availability calculated for the whole plan, but each sector report calculates its own Job Group Analysis and Work Force Analysis. Thus each sector report has its own Utilization Analysis. Also, if valid department information is present in the personnel actions files, personnel action summaries and the goal attainment report may be printed by sector. Significant Difference Rule Also known as the Two Standard Deviation Rule. A placement goal is set when availability exceeds employment by a statistically significant amount. When using this rule, you have the option of using an alternate comparison rule for small job groups that fail the Rule of Nines Test. Formula: [(Availability x total job group employees) - number of protected class employees] divided by the square root of [(1 - Availability ) x Availability x total job group employees] Example: Availability = 26.8% Employment = 16 Total Job Group = 68 employees BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 261 [(.268 x 68) -16] / SQRT[(1 - .268) x .268 x 68] = .61 Standard Deviations In this case, no placement goal would be set. Standard Deviation A large sample estimator of statistical significance based on the Central Limit Theorem. For utilization – the difference between the expected number (availability) and actual number of protected class employees is divided by the square root of the total employees in the job group multiplied by the availability ratio multiplied by the complement of the availability ratio (1 - availability). Formula: [(Availability x total job group employees) -number of protected class employees] divided by the square root of [(1 - Availability ) x Availability x total job group employees] Example: Availability = 26.8% Employment = 16 Total Job Group = 68 employees [(.268 x 68) -16] / SQRT[(1 - .268) x .268 x 68] = .61 Standard Deviations Start Point Analysis The Start Point Analysis is an estimation of the prior year’s affirmative action plan based on the current roster and the personnel actions that occurred during the plan date range. The Start Point Analysis is generated by taking the roster and “rolling back” the personnel actions. Hires are removed from the roster, terminations are added to the roster, and promotions and transfers are undone. The end result, if the personnel action files are complete and accurate, should be equal to last year’s roster. Utilization Rule A utilization rule is used to compare employment and availability percentages and determine if a placement goal should be set. The simplest rule is the Any Difference Rule. A placement goal is set when availability exceeds employment by any difference. Other rules include the Whole Person Rule, Eighty Percent Rule, and Significant Difference Rule. BALANCEaap Desktop User’s Guide Copyright © 2003 by Berkshire Associates Inc. Page 262 Update Plan An update is usually done for quarterly or semi-annual monitoring of personnel actions and utilization. It is frequently required for compliance reviews that occur more than several months after the annual plan. In an “update” or interim plan, availability is not recalculated as in an annual plan. Instead, availability figures are imported from the annual plan. Whole Person Rule A placement goal is set when availability exceeds employment by at least one whole person. Formula: (availability - employment ) × total in job group is more than 1 whole person. Example: Availability = 26.8% Employment = 23.5% Total Job Group = 68 employees (26.8% availability - 23.5% employment) x 68 persons employed in job group = 2.24 persons In this case, a placement goal would be set in this job group for this protected class because the difference of availability over employment represents 2.24 persons, more than one whole person.