Measure Up Initiative: Fortune and Refinitiv Partnership (Methodology)

Executive summary

Following heightened international focus on social injustice and racial inequality, Fortune and Refinitiv have joined forces to help businesses across all industries address one of the most pressing challenges we face – building truly inclusive and fair workplaces.

The Measure Up initiative aims to make diversity, equity and inclusion (DEI) disclosure and performance a critical metric for successful businesses. For this analysis, we leverage Refinitiv Diversity and Inclusion (D&I) data to create a relative benchmark and identify the most progressive companies in diversity and minorities inclusion.

Refinitiv Measure Up scores rank companies based on their disclosure and performance relative to the Fortune 500 companies. This assessment is initiated alongside the 2021 Fortune 500 companies and across Fortune 1000 company pages to track progress over time

Data collection

We’ve encouraged Fortune 500 companies plus others to contribute their minority group inclusion data using Refinitiv’s ESG Data Contributor tool by February 2021. Refinitiv also sourced data from public disclosures to ensure a critical mass of information was available to effectively rank performance of companies in terms of minority group inclusion. We list the 14 key metrics below that formed part of this analysis.

Ranking

Refinitiv adopted a percentile scoring approach to ranking companies as part of Measure Up. The key takeaway is that disclosure of data plays a critical role in how well a company has been ranked for this analysis. Companies that disclosed most and have the best percentage breakdowns for the numeric data measures come out better in the final ranking list.

Key points around the Percentile Ranking methodology are:

  • Rankings have been calculated based on relative performance versus other companies in the Fortune 500
  • All data measures are equally weighted
  • Percentile scores at a data-measure level are calculated based on data reported by companies for all 14 data measures
  • Non-disclosure of a Boolean data measure results in a score of zero for that item
  • Data measure scores are summed up at company level
  • Final percentile scores between 0-1 are derived for each company based on summed data measure scores
  • Percentile scores can be converted to ranks

Ranking process steps

Step 1: Disclosure of data

The 14 data measures listed in the table below formed part of the Measure Up survey and have been considered when ranking the Fortune 500 companies. Fortune 500 companies were encouraged to contribute three years of data for these measures to support trend analysis.

MetricData type
Policy Board DiversityBoolean
Policy Diversity and OpportunityBoolean
Targets Diversity and OpportunitiesBoolean
Day Care ServicesBoolean
Employee Resource GroupsBoolean
Minorities board PercentagePercentage
Women EmployeesPercentage
New Women EmployeesPercentage
Women ManagersPercentage
Employees with DisabilitiesPercentage
Gender Pay Gap PercentagePercentage
Minorities Employees PercentagePercentage
Minorities Managers PercentagePercentage
Minorities Salary GapPercentage

Step 2: Percentile scoring formula applied at data measure level

For each of the 14 data measures listed above, a score was generated using the percentile formula below.

  • How many companies are worse than the current one?
  • How many companies have the same value?
  • How many companies have no value at all?

A percentile rank score is based on the rank and therefore is not very sensitive to outliers.

Boolean data measureYes = 1, No = 0Boolean data measures are typically answered with Yes/No. Yes, denotes that the company has a Policy for a particular data measure & No denotes that the company does not have a policy. Companies that answered with a ‘Yes’ get a percentile score and answered with a ‘No’ get a score of zero.
Numeric data measureHigher the percentage, better the percentile scoreIf a numeric data is reported by a company, then a relative percentile ranking is applied considering all the companies reporting the respective data measure.

Illustration—Boolean data measure percentile ranking

The illustration below focuses on the Day Care Services data measure for 10 sample companies = “Does the company claim to provide day care services for its employees?” 

In the illustration, 7 companies have policy for Day Care Services and 3 companies do not have a policy. Those who responded “yes” they have a policy are awarded a score of 1 and those with “no” policy is assigned a 0. 

Company-actual nameTitleData measure valueData measure value converted to numeric valueCategory for percentile score calculationData measure percentile score calculationsPercentile formula calculation
Amazon.com, inc.Day Care ServicesYes1Same Value0.65(3+(7/2))/10
UnitedHealth Group IncorporatedDay Care ServicesYes1Same Value0.65(3+(7/2))/10
AT&T Inc.Day Care ServicesYes1Same Value0.65(3+(7/2))/10
AmerisourceBergen CorporationDay Care ServicesYes1Same Value0.65(3+(7/2))/10
Ford Motor CompanyDay Care ServicesYes1Same Value0.65(3+(7/2))/10
CIGNA CorporationDay Care ServicesYes1Same Value0.65(3+(7/2))/10
Chevron CorporationDay Care ServicesYes1Same Value0.65(3+(7/2))/10
Exelon CorporationDay Care ServicesNo0Worse00
Northrop Grumman CorporationDay Care ServicesNo0Worse00
Plains GP Holdings, L.P.Day Care ServicesNo0Worse00

Once the scores above are in place, the percentile scoring methodology is applied, as per the table below.  Other Boolean Data Measures in the survey are scored in an identical fashion:

  • No: of Companies with worse value = 3 companies
  • No: of Companies with same value = 7 companies
  • Total No: of Companies with a value = 10 companies
  • Data Measure Percentile score will be (3+(7/2))/10 = 0.65
  • Companies who do not have the policy will get a score of zero

Illustration—Numeric data measure percentile ranking

The illustration below focuses on the Employees with Disabilities data measure, which is the percentage of employees with disabilities or special needs.

Companies that employ a high percentage of disabled employees receive a better percentile score than those with lower percentages. To illustrate, the steps below show how the percentile score is calculated for Northrop Grumman Corporation versus the Employees with Disabilities data measure. The same percentile scoring methodology is applied as set out in Step 2.

No: of companies with worse value = 13In total there are 14 companies that have reported Employees with Disabilities of which 13 companies have a lesser data measure value when compared to Northrop Grumman Corporation
No: of companies with same value = 1No other company has data measure value of 8.1, so number of companies with same value will be 1
Total No: of companies with a value = 14Denotes total number of companies which have reported Employees with Disabilities

Data Measure Percentile score for Northrop Grumman Corp is calculated as (13+(1/2))/14 = 0.964285714

Company-actual nameUnitData measure valueData measure percentile value calculationsPercentile formula calculation
Northrop Grumman CorporationPercentage8.10.964285714(13+(1/2))/14
Microsoft CorporationPercentage6.10.892857143(12+(1/2))/14
L3Harris Technologies, Inc.Percentage60.821428571(11+(1/2))/14
CMs Energy CorporationPercentage50.75(10+(1/2))/14
Kinder Morgan, Inc.Percentage40.642857143(8+(1/2))/14
Hilton Worldwide Holdings Inc.Percentage40.642857143(8+(1/2))/14
OshKosh CorporationPercentage30.535714286(7+(1/2))/14
Qualcomm IncorporatedPercentage2.70.464285714(6+(1/2))/14
BlackRock, Inc.Percentage2.40.392857143(5+(1/2))/14
AT&T Inc.Percentage2.1610.321428571(4+(1/2))/14
Waste Management, Inc.Percentage20.25(3+(1/2))/14
C. H. Robinson Worldwide, Inc.Percentage1.980.178571429(2+(1/2))/14
3M CompanyPercentage1.40.107142857(1+(1/2))/14
AbbVie Inc.Percentage1.20.035714286(0+(1/2))/14

All other numeric data measures in the survey are scored in the same way.

Step 3: Summing up percentile scores for each company

Once data measure percentile scores have been calculated, the next step in the process is for these values to be summed up. The table below uses Microsoft as an example, where the data measure values are converted to data measure percentile scores and summed up to a score of 9.498901527.

Company-actual nameTitleData typeData measure valueData measure percentile value calculations
Microsoft CorporationDay Care ServicesBoolean10.862
Microsoft CorporationEmployee Resource GroupsBoolean10.75
Microsoft CorporationEmployees with DisabilitiesFloat6.10.892857143
Microsoft CorporationGender Pay Gap PercentageFloat1000.729741176
Microsoft CorporationMinorities Board PercentageFloat39.70.889937107
Microsoft CorporationMinorities Employees PercentageFloat49.80.769417476
Microsoft CorporationMinorities Managers PercentageFloat41.30.881294964
Microsoft CorporationMinorities Salary GapFloat100.60.944444444
Microsoft CorporationPolicy Board DiversityBoolean10.574
Microsoft CorporationPolicy Diversity and OpportunityBoolean10.566
Microsoft CorporationTargets Diversity and OpportunityBoolean10.902
Microsoft CorporationWomen EmployeesFloat28.60.345
Microsoft CorporationWomen ManagersFloat26.30.394009217
  Sum of percentiles of all the reported data measures9.498901527

Step 4: Ranking companies based on the sum of the data measure percentile score

The final step is to take the sum of percentiles of data measure scores for all the companies and apply the same percentile score methodology outlined in Step 2.  This will give a final company level percentile score which enables the list to be ranked.  The table below illustrates this with a subset of examples companies. 

Company-actual nameSum of data measure percentile value calculationsPercentile scoresRanks
Microsoft Corporation9.4989015270.9988913531
Target Corporation8.6111928750.9966740582
The Gap, Inc.8.46832250.9944567633
Intel Corporation8.2200768270.9922394684
Centene Corporation7.9137767240.9900221735
The Allstate Corporation7.7566525420.9878048786
PVH Corp.7.707454880.9855875837
Amazon.com, Inc.7.6777279470.9833702888
Bank of America Corporation7.637180040.9811529939
Levi Straus & Co.7.6019213120.97893569810
Wells Fargo & Company7.5956158030.97671840411
Verizon Communications Inc.7.551596660.97450110912

Appendix 1: Single company scoring examples

Step 1: Scoring at data measure level

Fourteen data measures form part of the Measure up survey with companies awarded scores for each measure disclosed. For Boolean data measures, where a company has a disclosed the fact it has policies, targets etc, they are converted to a score of 1 and scored accordingly for that data measure. Non-disclosure, or no policy is converted to a score of 0, resulting in a 0 score for that data measure.

Example: Sample Corporation has not set targets or objectives to be achieved on diversity and equal opportunity, so was awarded a 0 score for this measure. Sample Corporation does have a policy on Board Diversity, so the data measure value is represented as 1, which equates to a percentile score of 0.574 for this data measure.

Numeric/Percentage Data is scored based on the value of the Percentages reported by the company. The higher the percentage, higher the percentile score. Sample Corporation percentile score was high for both Women Employees and Women Managers as data measure value percentages were high in comparison to other companies in the Fortune 500.

Table 1 is an illustration of data for Sample Corporation, showing the data measure values and the percentile scores which are benchmarked against other Fortune 500 companies in the survey.

Company-actual nameCategoryTitleData typeUnitData measure valueData measure percentile value calculations
Sample CorporationWorkforceTargets Diversity and OpportunityBooleanTRUE/FALSE00
Sample CorporationManagementMinorities Board PercentageFloatPercentage290.688679245
Sample CorporationWorkforceMinorities Employees PercentageFloatPercentage500.774271845
Sample CorporationWorkforceMinorities Managers PercentageFloatPercentage360.762589928
Sample CorporationWorkforceNew Women EmployeesFloatPercentage730.981481481
Sample CorporationManagementPolicy Board DiversityBooleanTRUE/FALSE10.574
Sample CorporationWorkforcePolicy Diversity and OpportunityBooleanTRUE/FALSE10.566
Sample CorporationWorkforceDay Care ServicesBooleanTRUE/FALSE10.862
Sample CorporationWorkforceWomen EmployeesFloatPercentage750.961666667
Sample CorporationWorkforceWomen ManagersFloatPercentage700.993087558
Sample CorporationWorkforceEmployee Resource GroupsBooleanTRUE/FALSE10.75

Step 2: Generation of a company level percentile values

Once data measure percentile scores have been calculated as part of step 1, step 2 is to sum up the data measure percentile values at company level. Using Sample Corporation as an example, 11 data measure values were summed up to give a company percentile score value of 7.913776724.

Table 2 highlights the data measures disclosed by Sample Corporation represented as data measure value plus data measure percentile scores and summed company level percentile score for the company.

Company-actual nameCategoryTitleData typeUnitData measure valueData measure percentile value calculations
Sample CorporationWorkforceDay Care ServicesBooleanTRUE/FALSE10.862
Sample CorporationWorkforceEmployee Resource GroupsBooleanTRUE/FALSE10.75
Sample CorporationManagementMinorities Board PercentageFloatPercentage290.688679245
Sample
Corporation
WorkforceMinorities Employees PercentageFloatPercentage500.774271845
Smaple CorporationWorkforceMinorities Managers PercentageFloatPercentage360.762589928
Smaple CorporationWorkforceNew Women EmployeesFloatPercentage730.981481481
Sample CorporationManagementPolicy Board DiversityBooleanTRUE/FALSE10.574
Sample CorporationWorkforcePolicy Diversity and OpportunityBooleanTRUE/FALSE10.566
Sample CorporationWorkforceTargets Diversity and OpportunityBooleanTRUE/FALSE00
Sample CorporationWorkforceWomen EmployeesFloatPercentage750.961666667
Sample CorporationWorkforceWomen ManagersFloatPercentage700.993087558
Sum of percentiles at company level: 7.913776724

Step 3: Generation of final percentile ranking score for all companies

The percentile score formula is then applied to the sum percentiles at company level for each company in the Fortune 500 to identify the best forming companies and ranked accordingly. Using Sample Corporation as an example, their summed data measure percentile value of 7.913776724 was taken along with those of all other Fortune 500 companies and the percentile score formula was applied to generate a final list of rankings. Sample Corporation came out at number 5 on the list vs. peers in the Fortune 500 list.

Table 3 shows an illustration for 5 sample companies, their summed data measure percentile scores and final company level percentile scores and ranks.

Company-actual nameSum of data measure percentile value calculationsPercentile scoresRanks
Microsoft Corporation9.4989015270.9988913531
Target Corporation8.6111928750.9966740582
The Gap, Inc.8.46832250.9944567633
Intel Corporation8.2200768270.9922394684
Sample Corporation7.9137767240.9900221735

Refinitiv, an LSEG (London Stock Exchange Group) business, is one of the world’s largest providers of financial markets data and infrastructure. With $6.25 billion in revenue, over 40,000 customers and 400,000 end users across 190 countries, Refinitiv is powering participants across the global financial marketplace. We provide information, insights, and technology that enable customers to execute critical investing, trading and risk decisions with confidence. By combining a unique open platform with best-in-class data and expertise, we connect people to choice and opportunity—driving performance, innovation and growth for our customers and partners.