Psychologist Martin Seligman believes that the human species should have been called homo prospectus (prospectus meaning ‘to look forward’) rather than homo sapiens (sapiens meaning ‘wise’). He argues that where humankind differs from other primates is our ability to plan for the future.

Yet, Seligman’s beliefs seem baseless when taking into account mankind’s ability to plan their finances for the future. About 80 per cent of Singaporeans underestimate the amount they need for retirement by a whopping 31 per cent, according to the OCBC Financial Wellness Index.

How can governments better help citizens prepare financially for retirement? Gregory Chia, Group Director of the Policy, Research, and Statistics Group at the Central Provident Fund Board (CPFB), shares how they use data to support citizens’ retirement planning and serve them more efficiently.

Planning for the golden years with data

The CPFB helps citizens with retirement planning by granting them easy access to their pensions data. “We make it easy for members to access their CPF account information anytime through our digital platforms,” says Chia.

CPFB also provides online calculators which projects if citizens’ savings are sufficient for their retirement goals. “[These calculators] allow them to simulate the effects of a range of actions, such as housing payments, CPF top-ups and withdrawals on their retirement income,” he explains.

CPFB also analysed data of its members’ past behaviour to customise messages to encouraged them to top up their retirement savings accounts.

“Members are motivated to make top-ups for different reasons. For example, some may be motivated to help their loved ones, while tax relief may be very attractive to others,” he shares.

This approach led to 117,000 citizens receiving top ups with dollar-for-dollar matching from the government last year under a special programme, Chia reveals.

Another way that CPFB helps citizens prepare for retirement is through personalised appointments with a CPF officer when they turn 55 and 65 years old respectively.

During these sessions, CPF officers guide citizens through relevant retirement plans. Each session is customised based on their existing pensions data, such as their rate of savings and outstanding loans.

How machine learning enables better services

Machine learning has helped CPFB officers respond to queries more efficiently.

When citizens write in with their queries, they often categorise their queries wrongly, resulting in the query being routed to the wrong officers initially. To tackle this, CPFB has started to use machine learning in the form of natural language processing.

“We developed a model which automatically categorises the actual query so that it gets routed to the right party immediately. It’s still early days but we’ve found that the model is about 30 per cent more accurate,” shares Chia. “We are seeing how we can improve it further,” he adds.

CPFB also developed models to forecast citizen call volumes, as well as simulation models. “These help us to optimise resource allocation while delivering quality service at our service and call centres, says Chia.

Infusing data into an organisation’s DNA 

Data assists organisations in providing better services, but staff members first need to be comfortable with working with data.

“We are very intentional about making data fun and exciting” to embed data into CPFB’s cultural DNA, shares Chia.

They organise regular data hackathons, which injects an element of fun and friendly competition while training employees’ data skills, he continues. These hackathons were so popular that CPFB started running different leagues to make them more accessible and interesting to staff with different levels of data background.

CPFB’s senior leadership too are picking up data skills. In 2021, the senior leadership team explored machine learning through a 3D racing simulator by AWS. The programme allows participants to build a race car with advanced machine learning techniques and participate in a global racing league.

“I found it to be a great way to get some hands-on experience to learn more about this field of machine learning,” says Chia of the programme.

Levelling up data capabilities through upskilling and tech

The CPFB created a Digital Literacy Training Roadmap to improve its officers’ data skills, says Chia. This programme teaches CPFB officers about how data may be helpful. Staff members also attend data training courses as part of this Roadmap.

Additionally, the Board created an informal data science community where staff members can come together to share information about data.This allows staff members to learn data skills from their peers, explains Chia.

Besides upskilling their workers, CPFB is building a data platform to perform more advanced analytics. This allows CPFB staff to process large amounts of data quickly, so they can take on more complex projects, he says. On our own, mankind may not be the best at planning for the future. However, we are wise enough to use tech tools like data and machine learning. With data helping us to better prepare for the future, mankind may live up to both the names homo sapiens and homo prospectus after all.