ERPZ Stop Mugging. Start Learning.

Nudging with Statistics

Beyond the standard formulas...

I was looking through some of the rally speeches for the Singapore General Elections 2011 and listening to the criticisms of the current ruling party in Singapore. The opposition have clearly grown stronger in terms of their position to criticize the ruling party. As an economist, I tend to think of this strength as not derived from politics (the opposition parties in Singapore honestly have no political power to speak of) but from economics and development. I have little interest in engaging in speculation about the General Elections results but I like to talk about some policy inspirations I got from listening to these speeches. Other inspirations for this piece comes from the concept of 'nudging' popularized by Richard Thaler, and Hans Rosling (see 'Washing Machine' & 'Joy of Stats'), who have been thinking about how to change the world with statistics.

Meta-problems
I like to think of economists as 'meta-problem-solvers'; economist don't think of solutions to specific problems - they engage in thinking out of specific problems and into what kind of system of incentives should be in place so that the problem would right itself or give rise to solutions. As I was telling a friend studying Physics at King's College, after attending a talk on 'Fuels of the Future', "economists would tell you that fossil fuel would never run out". The fact is that when you have too little supply of fossil fuels, their price simply skyrocket to a level that would make alternative energy sources much more attractive. And investment would be poured into these sectors rather than into more drilling or mining for fuels, and some day, fossil fuels would be history. Of course, the scientists are the ones who come up with ways to exploit the alternative energy sources. But the 'meta-problem', the question of 'how will the problem be solved' is answered by the market.

Nudging
And because meta-problems are usually about trying to alter behaviours (eg. switching from fossil fuels to renewables, focusing research efforts on diseases that plague developing instead of developed countries, encouraging poor families to send their kids to school, etc), there is the concept of 'nudge'. Traditional economics often consider only monetary incentives or implicit costs as carrot and sticks for behaviours. Nudging involves even the most subtle things that would alter behaviour such as manipulation of arrangements, structuring of an application form, design of a process, or posing of questions. It almost always suggests some sort of paternalism where there is an intellectual authority trying to guide actions.

Is development always about riches?

Economic Development
When I was in the military, the captain in charge of our company told us, "When doing anything, ask yourself: Are you doing the right thing? If yes, ask, Are you doing it the right way? If the answer is yes to both questions, proceed. If not, go and think about it again." In the beginning of Singapore's economic development, there was little dispute about what is the right thing to do; we just had to follow the path taken by the developed countries and stick to the fundamentals they have adhered to. Governance of the economy thus boils down to whether we are doing things the right way. In this matter, Dr Goh Keng Swee has steered economic policy pretty much in the right direction, thinking hard about the key characteristics of our economy, our idiosyncrasies and how ideas of the west can be adapted successfully. It has always been a combination of trial and error, but the design of incentives been good for bringing the economy in the direction of the shortest path to growth and development. The society was a diverse group of people but everyone starts from more or less the same point, at similar levels of income. Free market economy and market incentives operate brilliantly under such circumstances; and often, the right thing to do for the entire economy serves everyone well (the rising tide lifts all boats analogy).

Today, the tasks of governing Singapore's economy no longer rest solely on the second question, we need to think hard about whether the policies we implement are truly the 'right things'. As I mentioned quite a while back from watching Michael Sander's lectures, inequality poses a problem for free market allocation of resources. That means that policies no longer serve the needs of the entire society evenly. In the past, targeting GDP growth may be a simple useful guideline to improve the standards of living for the people. Now, we don't need the rich to grow as fast as (or even faster than) the rest of the society - we might want to target economic growth of just the bottom half, or even bottom 20% of the economy. This is in line of the tradition of our nation's governance not to be a welfare state. Essentially, targeting growth of this tier of the society is different from giving handouts or tax breaks. It is about providing the necessary community support, the right design of education infrastructure that provides them with the same opportunity for social mobility as the rest of the society.

Incentives for Governance
It has been mentioned that GDP of Singapore is used as a key performance indicator for the government. And so the salary of the cabinet are partly related to that. Given how their salaries have become a pretty controversial issue, the government cannot pretend that the big paychecks is not a problem anymore. Some people think it is a problem because it simply attracts 'talents' who are just keen about making money out of their work and perhaps not so concerned with the welfare of the typical man on the street. As an economist, I must say that I usually doubt anyone can be more interested in the welfare of others than that of his own. I'm not saying that utility cannot be intertwined, but it is rare to find such people. And while generally interested in the welfare and future of the typical man, leaders are usually a tad bit more concerned about their kids and their finances than your kids and your finances.

It would be wise, I believe, to structure incentives to socio-economic goals of the day and if it is to lift the people in the lower strata of the society, let the leader's incentives be connected to that, perhaps to the income growth of the bottom 10% of society. And it should be gross income, lest reducing tax burden of the bottom 10% could easily help to boost their key performance indicator. Sure enough, the goals will always change as with that of any entity in a dynamic environment; in order for the decisions made to be relevant for achieving the goals, the incentives should be changed accordingly.

Of course, that's just one way of thinking about ministerial compensation; there are many other ways to arrange it; you could benchmark it to the proportion of population living below certain income level, or tie it to some measure of social mobility. It is not about siding with specific segments of the economy or the electorate; it's about incentivizing policy makers to provide equal opportunities for all the the economy. At the most basic level, it is important that we start focusing on GNP, Gross National Product rather than GDP because our GDP figures includes the income accrued by foreign capital that is going to end up in foreign hands anyways.

Do some heavy lifting...

Using Statistics
In a complex, multi-tiered economy like the one Singapore have today, statisticians need to work harder to slice and dice their data into more categories and segments so that the performance of the different sectors of the economy can be evaluated carefully and the government can tackle problems at the roots when they start budding and not when they have grown into a tree. Presenting the data in different ways allows us to look at the various social problems we face in different ways. The government will have to make use of that a lot in the future of governance; there is going to be a need to scrutinize ourselves rather than look abroad for ideas.

There can be surveys done on sentiments about migration and personal prospects for the future, putting up a time series that contrasts cost of living and real income of the bottom 20% of society. Singapore Statistics needs to start being creative about the data they collect and put them together in new, imaginative way that gives us an accurate picture of the different segments of the economy. It is currently way too boring (see this). There is endless ways of doing that so it will be good to open up the data to the public for people to start toying with them. And consequently, find out what are the 'right things' to do from the statistics.

Pluralism
Even from statistics, there might not be a 'right thing' but 'right things'; when a single party crowds the parliament, there is no meaningful debate on what 'right thing' is selected as priority. This is a time when pluralism is necessary; where more approaches considered, more voices heard. Statistics is 'never wrong' (note that it is 'never right' either) because it it tries to account for pluralism. Nobody knows what's best for everyone and no one should pretend he or she can be that way. Pluralism nudges us in the way evolution selects the fit traits to propagate. The robustness of democracy comes from pluralism and it is perhaps time to try and exploit a bit of that in our country.

A Conclusion
What I have outlined in this entry is just a really random cluster of ideas I have about the meta-problem of governance in Singapore and the future direction the management of the economy can take. It is a collection of seeds ready to be sown; and it has to be done quickly because these are abstract ideas that suggests nothing particularly concrete that can be done immediately to the existing structure. They are concepts that takes time to seep in and eventually make a difference. And that, to me, is what nudging is usually about.