9.3 Adverse Impact Analysis (The Four-Fifths / 80% Rule & Standard Deviation Tests)

9.3 Adverse Impact Analysis (The Four-Fifths / 80% Rule & Standard Deviation Tests)

In public sector human resources, ensuring that selection instruments are free from unlawful discrimination is both a legal mandate and a core ethical responsibility. Under Title VII of the Civil Rights Act of 1964 (as amended by the Civil Rights Act of 1991, 42 U.S.C. § 2000e-2(k)), employment discrimination is evaluated under two distinct legal theories: Disparate Treatment and Disparate Impact (Adverse Impact).

While disparate treatment requires proving intentional, discriminatory animus, disparate impact focuses strictly on the consequences and statistical outcomes of neutral employment policies. Understanding how to calculate, detect, and legally analyze adverse impact using the Four-Fifths (80%) Rule and Standard Deviation Tests is a foundational competency tested on the PSHRA-SCP examination.

---

1. Legal Foundations: Disparate Treatment vs. Disparate Impact

+-----------------------------------------------------------------------------+
|                   TITLE VII EMPLOYMENT DISCRIMINATION THEORIES              |
|                                                                             |
|   DISPARATE TREATMENT (Intentional)       DISPARATE IMPACT (Adverse Effect) |
|   - Unlawful intentional discrimination   - Facially neutral policy/test    |
|   - Evaluated under McDonnell Douglas     - Discriminatory outcome/effect   |
|     burden-shifting framework:              on protected classes            |
|     1. Prima facie case (qualified,       - Intent is completely IRRELEVANT |
|        rejected, position open)           - Established in Griggs v. Duke   |
|     2. Legitimate Non-Discriminatory        Power Co. (1971) & codified in  |
|        Reason (LNDR) articulated            Civil Rights Act of 1991        |
|     3. Pretext proven by plaintiff        - Defense requires proof of       |
|   - Affirmative defense: BFOQ (Bona         Job-Relatedness & Business      |
|     Fide Occupational Qualification)        Necessity under UGESP           |
+-----------------------------------------------------------------------------+

The Landmark Precedent: Griggs v. Duke Power Co. (1971)

In Griggs, the Duke Power Company required a high school diploma and passing scores on two standardized aptitude tests (Wonderlic Personnel Test and Bennett Mechanical Comprehension Test) for transfer into higher-paying operational departments. Neither requirement was shown to measure ability to perform the jobs. The Supreme Court unanimously held:

1. Title VII prohibits employment practices that are "fair in form, but discriminatory in operation."
2. Lack of discriminatory intent does not redeem employment mechanisms that create disproportionate exclusionary barriers.
3. The employer bears the burden of proving that any test or educational requirement with an adverse impact is job-related for the position in question and consistent with business necessity.
4. "The touchstone is business necessity. If an employment practice which operates to exclude Negroes cannot be shown to be related to job performance, the practice is prohibited."

---

2. The Four-Fifths (80%) Rule: Calculation & Methodology

Under UGESP Section 4D (29 CFR § 1607.4(D)), federal enforcement agencies utilize the Four-Fifths (80%) Rule as an administrative rule of thumb to determine whether a selection procedure produces adverse impact.

The Four-Step Calculation Algorithm

1. Step 1: Calculate the Selection Rate ($SR$) for Each Demographic Group: $SR = \frac{\text{Number of Applicants Selected / Passed}}{\text{Total Number of Applicants in Group}}$
2. Step 2: Identify the Benchmark (Highest) Selection Rate: Identify the demographic group with the highest selection rate ($SR_{\text{max}}$).
3. Step 3: Calculate the Adverse Impact Ratio ($AIR$) for Each Group: $AIR = \frac{SR_{\text{Focus Group}}}{SR_{\text{Benchmark Group}}}$
4. Step 4: Compare Against the 80% (0.80) Threshold:
   • If $AIR \ge 0.80$ (80%), adverse impact is not indicated under the administrative guideline.
   • If $AIR < 0.80$ (80%), prima facie adverse impact is demonstrated, requiring validation or process modification.

---

Step-by-Step Worked Mathematical Example

A municipal police department administers an entry-level physical agility examination to 300 applicants (200 Male applicants and 100 Female applicants). The results are as follows:

+-----------------------------------------------------------------------------+
|                      PHYSICAL AGILITY ADVERSE IMPACT TABLE                  |
|                                                                             |
|   Group     Applicants  Passed / Selected  Selection Rate (SR)   AIR vs. Benchmark |
|   -------   ----------  -----------------  -------------------   ----------------- |
|   Male         200             120         120 / 200 = 60.0%     60.0% / 60.0% = 1.00 (Benchmark)|
|   Female       100              30          30 / 100 = 30.0%     30.0% / 60.0% = 0.50 (50.0%)    |
+-----------------------------------------------------------------------------+

• Calculation: $AIR_{\text{Female}} = \frac{30.0\%}{60.0\%} = 0.50 \quad (50.0\%)$
• Conclusion: Because $50.0% < 80.0%$, the physical agility examination produces severe adverse impact against female applicants under UGESP. The agency must either validate the physical test under UGESP standards or modify the test format/cutoffs.

---

3. Statistical Significance: The Two Standard Deviation Test

While the Four-Fifths Rule is an administrative guideline, federal courts require inferential statistical significance to determine whether disparities are due to chance or systemic bias. In landmark cases—Castaneda v. Partida (1977) and Hazelwood School District v. United States (1977)—the Supreme Court established the Two Standard Deviation Rule.

The Standard Deviation ($Z$-Score) Formula

When analyzing binary hiring outcomes (selected vs. not selected) from an applicant pool, the selection process follows a binomial distribution:

1. Expected Hires ($E$): $E = N \cdot p$ (where $N$ = Total applicants selected/hired from the total pool, and $p$ = Protected group's proportion of the total applicant pool)
2. Standard Deviation ($SD$ / $\sigma$): $\sigma = \sqrt{N \cdot p \cdot (1 - p)}$
3. Difference ($Z$-Score): $Z = \frac{|\text{Observed Hires} - \text{Expected Hires}|}{\sigma}$

+-----------------------------------------------------------------------------+
|                   COURT INTERPRETATION OF STANDARD DEVIATIONS               |
|                                                                             |
|   Z-Score Disparity           Statistical Probability     Legal Conclusion  |
|   -------------------------   -----------------------     ----------------  |
|   Less than 2.0 SD (Z < 1.96)  p > 0.05 (Greater than 5%)  Attributable to   |
|                                                            random chance;    |
|                                                            No adverse impact |
|                                                                             |
|   2.0 to 3.0 SD (Z >= 1.96)    p < 0.05 (Less than 5%)     Statistically     |
|                                                            significant;      |
|                                                            Prima facie impact|
|                                                                             |
|   Greater than 3.0 SD (Z > 3)  p < 0.001 (Less than 0.1%)  Overwhelming proof|
|                                                            of disparate      |
|                                                            impact            |
+-----------------------------------------------------------------------------+

Standard Deviation Worked Example

An agency hires 100 firefighters ($N = 100$) from an applicant pool where Black applicants comprise 30% of all applicants ($p = 0.30$).

• Expected Black Hires ($E$): $100 \times 0.30 = 30$ hires
• Observed Black Hires: 16 hires
• Standard Deviation ($\sigma$): $\sigma = \sqrt{100 \times 0.30 \times (1 - 0.30)} = \sqrt{100 \times 0.30 \times 0.70} = \sqrt{21} \approx 4.58$
• $Z$-Score Disparity: $Z = \frac{|16 - 30|}{4.58} = \frac{14}{4.58} = 3.06$
• Legal Conclusion: Because $Z = 3.06 > 1.96$ ($p < 0.002$), the disparity exceeds 3.0 standard deviations. This provides statistically conclusive evidence of adverse impact under Castaneda and Hazelwood.

---

4. Limitations of the Four-Fifths Rule & Small Sample Corrections

UGESP Section 4D notes that the 80% rule is unreliable when dealing with small sample sizes due to the "flip-flop" anomaly:

• Example: If an agency hires 2 out of 4 male applicants ($SR = 50%$) and 0 out of 1 female applicant ($SR = 0%$), the $AIR = 0%$. However, had the single female applicant been hired, the female selection rate would jump to $100%$, completely reversing the impact ratio.
• In small sample sizes ($N < 30$), courts and psychometricians reject the Four-Fifths rule and instead mandate Fisher's Exact Test (hypergeometric distribution) or multi-year data pooling over successive hiring cycles.

---

5. The Three-Stage Burden-Shifting Defense in Title VII Litigation

Under Section 703(k) of Title VII (codified in the Civil Rights Act of 1991), disparate impact lawsuits follow a structured, tripartite burden-shifting framework:

+-----------------------------------------------------------------------------+
|                  DISPARATE IMPACT BURDEN-SHIFTING FRAMEWORK                 |
|                                                                             |
|   STAGE 1: PLAINTIFF'S PRIMA FACIE BURDEN                                   |
|   - Plaintiff must identify the specific selection procedure and prove      |
|     statistically significant adverse impact (via 4/5ths Rule & SD Tests).  |
|                                                                             |
|   STAGE 2: EMPLOYER'S BURDEN OF PROOF (Job-Relatedness & Business Necessity)|
|   - Burden of PROOF (persuasion) shifts to the employer.                    |
|   - Employer must demonstrate that the challenged selection device is       |
|     job-related for the position in question and consistent with            |
|     business necessity through formal UGESP validation.                     |
|                                                                             |
|   STAGE 3: PLAINTIFF'S REBUTTAL (Less Discriminatory Alternative)           |
|   - If employer proves validity, burden shifts back to plaintiff.           |
|   - Plaintiff must prove that an equally valid alternative selection        |
|     procedure existed with less adverse impact, and the employer refused    |
|     to adopt it.                                                            |
+-----------------------------------------------------------------------------+