In international pension discussions, it is common practice to use Sweden, Chile and Australia as examples of Defined Contribution (DC). Even though they are located on three different continents, each of these countries has a distinctive culture and unique historical context when it comes to social security and pension policy.

About a year ago, I led a pilot study for the Swedish Government covering the Swedish Premium Pension which is unique as a mandatory individual DC scheme integrated in the Swedish State Pension. It is a sibling to the more well-known mandatory pay-as-you-go system with a capital buffer (Notional DC). The first pillar system covers everyone who is currently working, or has been working, in Sweden. Data on individuals’ choices is available since its inception in 2000. It is a rich source of individuals’ information that can be cross-referenced with background data such as age, income, marital status, postcode and so on, so there is no lack of information regarding scheme members’ actual behaviour.

During this study we looked into both Chile and Australia to learn from their long experience with DC pensions. This provided us with the opportunity to observe full policy circles typically starting with an expert commission identifying problems, followed by the introduction of new pension policies aimed to deal with the identified problems. Moreover, it gave us the opportunity to learn to what extent the policies actually resolved these problems. This historical analysis provided invaluable information as it helped us to assess the potential effectiveness of different policies.

In my most recent trip outside Europe, I attended a conference where Latin American regulators assembled to discuss ways to improve their different pension systems. This was an eye-opener for me; it was amazing to see how similar their problems were even if specific details were different depending on the local context. The core challenges centre on:

  • how to get individuals to save for their pension in the first place, and
  • how to mitigate bad market practices.

A common denominator

Most countries face the same basic problem: consumers are not acting as policy makers expect them to. To explain this uncomfortable deviation, consumers are typically labelled as ‘irrational’ by policy experts. Significant effort has been spent on improving financial education and market transparency in the hope that consumers will fundamentally change their behaviour to become more like the expectations of the policy makers. There is some irrationality present in this logic, which should not be attributed to the consumer. A more sensible approach is to apply an evidence-based approach to pension policy. That means trying to understand how consumers actually behave and then building a system around that.

Sweden and Chile

A good lesson from Sweden is the national administrative infrastructure around the Premium Pension. With a national administration of pension rights, providers are ‘only’ competing by offering different investment services, which eliminates expensive transfer and administration costs.

However, the main problems in Sweden were the poorly designed choice architecture resulting in over 850 mutual funds with no gate-keeper. This poor design was based on a misplaced belief that the competition between mutual funds on the retail level would automatically clear the market with respect to price and quality. In fact, it did the opposite – this ideologically driven design allowed for fraud and bad marketing practices.

Chile, on the other hand, has a better designed choice architecture where each provider can only offer five different types of funds. Each type is prescribed by regulation and, to keep prices low, there is an auction for the inflow of new participants. One would expect this to be a recipe for a user-friendly low-cost system but, despite the products being almost identical and relatively easy to compare, most switches are the result of aggressive retail sales practices fuelled by high commissions paid via administrative transfer costs.

There are five factors that require extra consideration in pension design.

1. Political truism

Political ideology is tightly linked with economics; socialists base their world view on Marxist theory; market liberals have built their world view around Milton Freidman’s theories and so forth.

The challenge is that academic economic theories have been upgraded to political beliefs that, over time, have solidified into political truisms. Politicians from both the left and right are trying to reshape the world into a ‘better’ place, where ‘better’ reflects their particular ideal for how the world should work.

The biggest challenge for evidence-based pension policy is to convince politicians to look at how people are behaving, not what their particular political ideology predicts people will do.

2. Local context

Despite the similarities, each country across the world has a unique context in terms of state pension, old age poverty protection and other forms of social security for its citizens. The design of the occupational pension is conditional on the local context and how that is expected to develop over time. The latter is especially important in those cases when the current welfare solutions are not deemed to be sustainable in the long-term, for example due to an aging population.

Australia and Chile have mandatory participation of workers in their second pillar pension system. In Sweden and The Netherlands the workplace pension is quasi-mandatory and supervised by the so-called social partners (employer organisations and trade unions). But in all these cases, the growing group of self-employed workers is not covered by the second pillar pensions.

In countries with a large informal sector, pension coverage has been a problem as an average worker is only officially employed for a portion of their career. In Chile, the average worker contributes to workplace pensions in 40 per cent of their career and this is referred to, in the academic literature, as density of contributions. In a way, the Latin American countries have been dealing with the self-employed challenge for many decades, so it would be natural for us to learn from their first-hand experiences.

3. Governance

Most individuals are not engaged with their pension savings and many would not even save unless it was mandatory. It is well-known that pension products are sold, not bought and there is a large imbalance of information between buyer and seller.

If most of us do not act as the principal for our own pensions, to whom can we trust the stewardship? Stewardship includes setting the goal or target of the system, tilting it in favour of the end-user and making sure it delivers value for money. The natural list of candidates includes: government, private sector and social partners.

Although the government seems to be the obvious candidate to assume the role of steward, in countries with a history of corruption and/or a fast-increasing populist political class, consumers’ trust in the government, as well as the private sector, is low. Perhaps social innovation based on new technology could become a game-changer to fill the role that social partners had in the past?

4. Choice architecture

Each pension system has a choice architecture; offering no choice is also a choice architecture. Practical experience and insights from behavioural economics suggest that a choice architecture should contain a default option and offer a limited number of meaningful choices that make clear what impact the choices will have on an individual.

A broad and complex choice architecture will inevitably lead to choice-overload, which will result in the individual either abstaining from making a choice or asking for advice. This then will lead to the emergence of various sales practices dressed up as ‘free advice’.

Nudges implemented in the choice architecture will dictate individuals’ decisions. Experience shows that no matter what the default choice is, the majority of members will end up in it. This requires us to design the default choice carefully to meet the goal/target of the pension system in a robust way. If there are no clear goals articulated for the pension system, it is a sign of governance failure.

5. Implementation

Badly implemented, a good and well-considered pension design can end up failing miserably. It is necessary to monitor and supervise the implementation carefully and if it does not work as intended, public interventions are warranted.

For the market mechanisms to work, it is crucial that governments intervene to restore a healthy balance between informed buyers and sellers of products. The DC market is unique because most buyers are largely uninterested but mandated to save for their pension. Sellers provide complex products that are very difficult to evaluate even for a pension specialist. Practical experience shows that relying on retail competition as the main clearing mechanism between quality and prices has not worked well. Mis-selling scandals and market practices that result in higher than necessary costs for members seem to occur all over the world.

A better approach, which requires strong governance, is a system anchored in a good choice architecture for the members. Different components in the choice architecture could be based on collective procurement, benefitting from institutional competition and levelling of the playing field between the consumer and the financial industry.

The art of what is politically possible

Last year, Sweden was able to make a breakthrough in policy regarding the Premium Pension. This was done on the back of fraudulent behaviour in a few of the mutual funds. This created a crisis, which forced politicians to climb out of their ideological trenches and look for a solution. I am fully aware that pension reform is the art of the possible (rather than the optimal) and that the historical path determines what is politically possible at certain points in time.

Different approaches to DC policy have been tried with various degrees of success across the globe. Sweden, Chile and Australia all have a long history of DC policy design and providing consumer choices. The data from these countries is a veritable gold mine for those who are interested in producing evidence-based pension policy.

Although it might seem futile to think outside the current political constraints we occasionally experience a crisis that opens a window of change. We should mentally prepare to move towards more of an evidence-based policy for DC; as Louis Pasteur said “Chance favours the prepared mind”.