Assessing Culture (Research)
- Who is doing/funding this research?
- What is the purpose of the research?
- Is it possible to 'measure' risk culture?
- Why do we need a 'statistically valid' tool for assessing risk culture?
- How have we validated the survey instrument?
- Who can participate in the research?
- What's the time-frame?
- What's involved for participating banks?
- What do participating banks get out of it?
- How is this different to the services provided by many consulting firms?
- What information is available to local prudential supervisors?
- What about data security and confidentiality?
- What has to happen before the survey is administered?
- What happens after the survey is administered?
Who is doing/funding this research?
A cross-disciplinary group (finance and organisational psychology) based at Macquarie University in Sydney, Australia. The research is funded by Macquarie University and the Centre for International Finance and Regulation (CIFR). CIFR is a centre of excellence for research and education in the financial sector CIFR. The Australian Prudential Regulatory Authority (APRA) has assisted with the research design.
What is the purpose of the research?
The ultimate goals are to: build greater understanding of risk culture in order to inform the creation of more robust culture in financial institutions and support more effective prudential supervision.
Our research will help achieve this goal by the following means:
- We have created the first statistically validated survey instrument for risk culture.
- We will implement the survey instrument in up to 400 business units in up to 15 large banks spread across five countries.
- We will analyse the data to explore the following questions:
Does business unit culture vary significantly according to location, business line or bank? Does culture vary significantly with potentially causal factors such as governance structures, remuneration policies, education and training, resourcing of the risk management function, status and authority of risk managers, business strategy and risk appetite? *Is bank culture a significant driver of individual behaviour (e.g. speaking up, lack of gaming behaviour) and organisational outcomes such as staying within risk appetite?
Is it possible to 'measure' risk culture?
Yes. Organisational psychologists have developed statistically reliable survey instruments to assess other aspects of culture such as safety in construction and other industries. We know these instruments help predict important outcomes such as 'safe behaviour' and lower injury rates. We have now produced something similar in relation to risk culture.
Why do we need a 'statistically valid' tool for assessing risk culture?
Without a rigorous validation process, it is not clear what a questionnaire is measuring, if anything. A reliable survey instrument will allow financial institutions to test the uniformity of culture throughout the firm, identify pockets of 'problematic' culture, demonstrate the effectiveness of the culture (if it is strong) to outsiders and track how culture changes over time.
How have we validated the survey instrument?
We have followed a three step process. 1.Generated an inventory of possible questions based on: analysis of the disaster cases of the past, interviews with many industry experts (consultants, supervisors, risk managers), previous research in safety culture. 2.Pilot tested questions to a) identify the underlying dimensions of risk culture, and b) test whether the risk culture dimensions are correlated with desirable (self-reported) risk management behaviours and attitudes in bank employees. The initial pilot study used alumni of the Macquarie Master of Applied Finance program with a second pilot performed in an Australian bank. 3.A confirmatory study in a large bank was the final step.
Who can participate in the research?
We are seeking large banks represented in multiple countries. At this stage we hope to include banks headquartered in Australia, Canada, Singapore/Malaysia, UK and the USA i.e. banks based in English speaking countries. Other possibilities may be considered on a case by case basis. However, the identity
of participating banks will remain confidential.
If your organisation does not meet the criteria for the research (too small, not international, not a bank) we may be able to assist with risk culture assessment on a consulting basis. We can only offer free risk culture assessment if your organisation meets our research criteria.
What's the time-frame?
All surveying should be completed before June 2015.
What's involved for participating banks?
- Banks complete an 'institutional profile' capturing information on company demographics, remuneration policies, training/education programs, governance structures etc.
- An anonymous online survey, which takes around 15 minutes to complete, is administered to staff. The researchers will work with participating banks regarding sampling, but in summary we require approximately 100 employees in at least 30 separate teams or business units spread across different geographical locations and business lines. All staff in the risk management function will be surveyed.
- An anonymous online survey is administered to the Directors and Senior Executive team which takes about 10 minutes to complete.
What do participating banks get out of it?
They receive a free independent report on their own risk culture, showing uniformity across business units, whether culture is "shared" in surveyed teams, and the relationship of culture to a set of drivers and risk behaviours. This includes the opinions of risk managers and assurance. A second report (delivered direct to the board) examines how the 'tone at the top' compares with employee perceptions of risk culture. They also receive a free benchmarking analysis to assess how their own culture compares with other participating banks on a no-names basis. The benchmarking study will also report statistics related to a set of drivers of culture. *Banks also receive a Corporate Social Responsibility benefit by contributing to this crucial research.
How is this different to the services provided by many consulting firms?
Our service is of the highest quality and similar to services you might receive from other firms, but differs in several key ways: There is no financial cost during the research development phase. *The survey instrument is statistically validated. *The results will ultimately be published in scholarly journals and industry papers on an aggregated, no-names basis. We use a state-of-the-art 'within group agreement index' to assess whether there is a shared perception of culture across the organisation (most consultants provide individual analysis only).
What information is available to local prudential supervisors?
In several countries local prudential supervisors are taking a keen interest in the research findings. They will see research outcomes only on an aggregated basis. Participation in the research is confidential unless banks choose to discuss their participation with their supervisors.
What about data security and confidentiality?
We take these issues very seriously. We have developed protocols to protect the identity of participating banks and to protect the personal data of participating staff. More details are available on request.
What has to happen before the survey is administered?
A research contract has to be signed by both the university and the participating banks. This spells out the rights and responsibilities of both parties. The bank has to determine its sampling strategy i.e. which particular business units will be included. Macquarie research team to advise. The CEO sends out an email to staff in the selected business units announcing the survey and encouraging candid participation. Approximately one week later emails are sent to all staff in the selected business units containing a link to the secure online survey platform. Emails are usually sent by the bank's HR function. The survey is usually open for two weeks, with a reminder email after one week.
What happens after the survey is administered?
You receive a confidential report within three months of the close of the survey (provided you have completed the institutional profile). The final benchmarking analysis is likely to be available in mid-2015 (when all banks have completed data collection).