What process should organisations follow when gathering this data?

Section 3

What process should organisations follow when gathering this data?

The Global Perspective

Jason Buwanabala, Senior Manager, Inclusion and Diversity Consulting, PwC

The diagram below sets out the key stages in a typical diversity data collection exercise. The scope and complexity of each step will likely be influenced by the type of data collected.

Typical steps involved in diversity data collection (source: PwC)

Identify scope, key priorities and desired outcomes: It is important to understand key objectives and agree on an appropriate scope before undertaking your diversity data collection process. A key input into this is an understanding of long-term aims and the purpose of data collection. Once this is understood, decisions will need to be made around which diversity categories to collect (see section 4), which territories you would like to explore collecting them in, which stakeholders to involve (see section 2), roles, responsibilities and other relevant factors.

Review of permissibility of proposed approach: A review of local laws and cultural expectations or requirements should be undertaken for every territory in which data will be collected to ensure requirements are met and cultural sensitivities are considered. Conducting this initial review will help identify territories where data collection may be more complex or require additional steps. It will also help to identify any territories where local regulations or commercial considerations mean that work should be prioritised. Completing this review early in the process will help ensure a tailored approach and project plan can be developed for each of your territories and that any

required variations in tools and processes are identified. At this stage, information on diversity data categorisation sources or approaches may be collected to support later systems design and configuration (see section 4).

Detailed legal and privacy analysis: After the initial permissibility is understood, more detailed analysis should be conducted to identify key legal and privacy steps in priority territories. In relation to data privacy, for territories where data can be collected only if there is a lawful basis for processing, it is crucial to identify the appropriate lawful basis, put in place appropriate documentation (if required) and cover any additional requirements which may arise. For some territories, this may be a case of drafting an appropriate data privacy notice and consent documentation, while others will also require more detailed documentation to be put in place, or data assessments to be carried out. It will also be important to consider any discrimination laws in each territory so that relevant safeguards can be put in place to ensure data is not used in a way that may be deemed discriminatory.

Systems selection, design and enablement: A single technology tool can help to collect data in a consistent and secure way. Tool selection will likely depend on the existing tools available (and their capability) and the required use of the data (i.e. if it would need to be attributable to the employer). Most commonly, any current HR system is leveraged as this typically allows ongoing vs point in time collection and makes reporting simpler. However, where a HR system is not available, some form of online survey tool may be appropriate, providing that required steps have been taken in relation to data privacy. The tools used for data collection will also require configuration on a local level to ensure local laws and cultural adjustments are followed, with a particular focus on data privacy and handling.

Communications & engagement strategy and campaign: One of the key challenges in collecting diversity data is encouraging employees to provide this personal and sensitive data. Careful communication planning is therefore critical, with a focus on building trust with staff, assuring them that their data will be protected and reinforcing the positive reasons for data collection. In doing this planning, it is important to recognise that

employees' perspectives will differ. For example, some employees may be reluctant to provide this data as they are uncomfortable sharing this data, or may even fear it will be used for negative reasons, whilst others may assume that communications related to diversity do not apply to them. To address this, communication should start before any data is collected, using multiple channels and approaches to reach different employees, and should be seen as an ongoing process rather than a point in time communications campaign.

Support and guidance should be made available to employees and it should be made clear where further information can be found and queries and concerns can be directed. Where appropriate, it may also be helpful to engage with unions and works councils so they can support with employee engagement.

Diversity analytics: Once data has been collected, it can then be analysed and assessed to identify key trends and areas where action may be required. This analysis can also be used to look at response rates and identify where a needed. Data dashboards could also be developed to support reporting (including internally) and visualisation of data outcomes. These should be carefully designed to comply with data privacy requirements and ensure individuals can never be identified.

HSBC Asset Management

Like most, we initiated our use of diversity data with a focus on gender. Figures were accessible across markets globally, using existing data stored and readily available in the HR system. Through analysis of the data we were able to better understand disparities across levels of seniority, teams and geographies. This allowed us to set gender balance benchmarks for which we now regularly monitor progress against, inclusive of tracking leavers and joiners.

Amidst conversations and awareness sparked by 2020 social unrest and the Black Lives Matter movement, we began steps to tackle collection and measurement of ethnicity statistics as the next step in our use of diversity data. There was no existing ethnicity data held via the HR system, thus requiring us to reach out to staff directly in commencing a voluntary data collection process in markets where it was legally permissible.

As a group subsidiary, challenges in accessing the information submitted by staff became apparent. Without direct access to the HR system where the data was collected, there was an inability to view and evaluate figures specific to the asset management business. Awaiting clarifications regarding privacy rights and appropriate aggregating of the data led to significant delays in gaining the needed visibility. Furthermore, inability to see staff response rates prevented follow-ups or an active campaign to encourage self-submission, as there was no clarity on existing submission levels.

Once visibility was gained it became clear that submissions were uneven across markets. Use of employee resource groups and a more active campaign appear to have been effective in driving higher response rates in a particular market. Additional challenges of a group structure also became apparent. Given various central functions provide support to multiple group businesses, it is necessary to address how individual support staff are counted and allocated to the asset management business in order to generate the most accurate diversity statistics. We are now integrating learnings to build a more robust approach to analysis into both ethnicity and gender data moving forward.

There have been several additional learnings more broadly. First, we have realised that to enable progress in the use of diversity data, we must break the agenda up into digestible components to prevent the overall journey from stagnating. Second, and specific to ethnicity data, clear differences in representation and seniority within different ethnic minority groups means a blanket approach to address inequalities will be ineffective. Finally, and related to avoiding a blanket approach, regional differences in identification of ethnic minorities require more thought and a nuanced approach. Widely accepted government census categories make identification relatively straightforward for western markets. It is a more significant challenge in Asian markets where definitions of ethnic minorities are different and less consistent.