Finally, I would like to comment directly on the complexity of CPF policy and its implications. CPF has evolved to become far more than a simple provident fund framework that provides for retirement. In CPF Board's own words, it "also addresses healthcare, home ownership, family protection and asset enhancement".

In the years since its formulation in 1955, layers upon layers of additional schemes and initiatives have been added upon CPF, creating the complex policy artifice today.

In the course of researching and writing this paper, I am struck by this immense complexity and have tried to simplify as far as possible without losing sight of the big picture.

We have seen from earlier sections of this paper has shown how CPF policy intersects with housing policy, healthcare policy, financial and economic policy and so on. Indeed, there are probably far more interdependencies that were not explored in this paper, either because they are of marginal relevance (e.g. CPF and immigration policy: the foreign workers levy is imposed on low-skill foreign laborers, who are exempt from CPF contributions, in part to equalize wage costs for low-skilled Singaporeans) or because they are unknown unknowns.

The implications of this staggering complexity are probably underappreciated.

First, complex systems are hard to understand, to predict and to manipulate. Policymakers who are merely monitoring CPF policy must keep track of an ever-increasing number of variables and interdependencies. For example, under Singapore's current managed housing market, policymakers are constantly keeping a close eye on property prices and enacting measures to ensure the market neither overheats nor cools excessively. Any tweak to one area likely could well have unpredictable second-order effect on another area. Any large-scale reform would cut across so many different policy areas (and hence different government agencies in practice) presents significant organizational challenges.

Finally, even if the government internally decides on the trade-off to be made, this must be presented to a bewildered general public. Even as someone who is (hopefully) relatively intelligent and trying to study CPF policy in good faith, I found it really hard to understand the policy in its entirety. It is simply too complex! The average person simply cannot be expected to understand how CPF rates were an instrument of financial repression or could be used as levers for macroeconomic policy or the relative merits of different methods of financial engineering and investment approaches to managing CPF monies. Instead, collective deliberation devolves into a question of faith: do you think the government truly has your best interests at heart?

There are many areas of policy where lack of understanding by the electorate is acceptable and perhaps even welcome, especially in technocratic areas of policymaking. (Contenders include monetary policy, defense policy, foreign policy and so on.) After all, the point of representative democracy is precisely to offload decision-making responsibilities to one's elected officials. Singapore's style of governance has never been particularly consultative. But bread-and-butter social policies constitute a core part of the government's compact with the people. At the very least, the average person should be able to appreciate the fundamental trade-offs. I submit that such complexity mixed with the top-down, paternalistic-yet-self-reliant approach to CPF policy explains why it is a continued source of misguided but understandable sentiments that the government might be doing something nefarious with CPF monies.

More than responding to any specific concern with the current CPF policy, I believe this long-term trend of increasing complexity is worrisome and should be reversed.((For a similar perspective, see Devadas Krishnadas, 'Why it’s so hard for S’poreans to understand CPF', Today, 28 May 2014. https://www.todayonline.com/singapore/why-its-so-hard-sporeans-understand-cpf)) For a start, CPF should return to its original purpose of ensuring retirement adequacy. The linkage between CPF and housing policy should be reduced (e.g. using CPF for residential properties should be limited) and measures to ensure CPF monies are protected from inflation should be considered (e.g. linking interest rates to inflation or supplementing CPF with state-provided pension). Contribution rates should not be tinkered with for macroeconomic purposes and greater transparency and greater disclosure and transparency concerning the management of CPF monies and government reserves in general are welcome.