IF there had been any lingering doubts at the beginning of the year that 2009 would bring a recession to remember, January quickly put an end to them. Within the first three weeks of the month, economic conditions deteriorated so rapidly that the Ministry of Trade and Industry (MTI) had to lower its growth forecast for this year twice.Its most recent revision, made two weeks ago, marked an unusually sharp and swift downgrade.

[…]

But if there is one lesson we are likely to learn this year, it is that during this unusually stormy skies, economic projections have become even less reliable than weather forecasting. […] Perhaps the takeaway, then, is simply be prepared for the unexpected. (The Straits Times, p. A2, 2 Feb 2009)

When I first saw this article in the world-class newspaper The Straits Times yesterday, my instinctive reaction was this: of what use are economic forecasts if they are unable to serve us any real purpose except to create confusion and uncertainty? Is it even necessary to have forecasts if we are told to always ‘be prepared for the unexpected’?

I believe these questions have a significant impact on the reputation of ‘economics’ as a discipline and, more importantly, its quest to be recognized as a ‘science’. Indeed, economists gain eminence within academia and policy-making circles because of their perceived ability to predict how the world will develop. But how often has it come up short? Perhaps the situation cannot be worse than what we see now: As 2007 came to a close, did economists forecast Lehman imploding in 2008? Was it foreseen that a whole country could go bankrupt (Iceland) because of poor regulation of financial markets elsewhere? Apparently not! What then has gone wrong?

From the way I see it, there is a massive lacuna in the methodologies (notably the Autoregressive Integrated Moving Average, or ARIMA, method) that economists employ to obtain figures and make forecasts. I would argue that the conventional way of relying on time-series data to create regression analyses is not feasible at all, for the simple reasons that time-series data reflects the past and there is a lag period before the data is even ready for analysis. If we are to successfully predict the future when the present is turbulent and fraught with uncertainties, how pertinent would time-series data be? How would these data help us ascertain trends and make responsible economic policy decisions? Furthermore, many variables are omitted from analyses or wrongly classified as ‘disturbances’ or ‘error terms’ when they directly impact on economic decisions. Is it then possible to know whether we are analyzing the right variables?

Then again, the aforementioned questions presuppose that economic projections are necessary to aid policy-making, but is that really the case? Are policy-makers any less ‘responsible’ if projections go totally wrong (just like they did recently all over the world)? Of course, most of us would argue that policy-makers need to conduct due diligence in order to make informed decisions, and I have no issue with that. The main problem is rather the inability of economic forecasts to ‘inform’. The challenge, therefore, is to bridge this methodological chasm.

In the aftermath of the Wall Street scandals, I have argued that there is a greater need to strengthen regulation. I would like to add that for decisions to be truly ‘informed’, policy-makers need to get a better grasp of what private companies are doing in order to enhance the relevance of their policies. Put simply, policy-makers need to alter the way data is collected.

I propose two ways to go about it. First, and a highly radical one at that, is to introduce secret agents into companies’ middle and higher management. No, I am not obsessed with Infernal Affairs or James Bond! Indeed, I truly feel that this method is more rewarding than the traditional economic forecasts as it provides access to fresh data (especially those that may not even be made public eventually) and top-secret management decisions. Imagine this: Three months prior to the Lehman Brothers crash, George W. Bush gets alerted by his informants in the Lehman top management about problems with the internal accounts and cash flow. Could he – despite his very limited intellect – have more time to react and prevent Wall Street from spiraling into stygian depths?Certainly, this approach is unscrupulous and operates on a principle of distrust, because it assumes that corporations are not ‘transparent’ after all. It also demands a morally upright government to make judicious use of the data collected rather than sell them to other companies/governments or hold the companies involved to ransom. As such, policy-makers would have to decide whether this is a necessary evil to stem more evil.

The second method, on the other hand, operates on a principle of trust. Policy-makers could conduct dialogue with the top 200 or top 500 firms in the economy (depending on its size) once in two months, as well as shorten the auditing cycle to 4 months (3 months for checks, one to follow-up). Policy-makers also need to engage more widely researchers from a wider field (not just economists, but also economic sociologists, economic geographers and economic psychologists). This is a much more plausible way to collect fresh information, understand problems/challenges companies face and pruchasing/production developments/strategies at present and in the near future, but there are three conditions: one, policy-makers have to be reflexive and be able to change their policies according to developments in the economy based on the data collected; two, companies must be forthright in their dialogue with policy-makers, and must be willing to highlight problems and share data about their companies; and three, there is a need to widen significantly the pool of regulators in both the public and private sectors).

As global policy-makers concoct programs to overcome the current economic crisis, it has become manifest that depending on economic forecasts is not the way to go, because economics is never a 'science', however much it tries to appear as such (note its widespread use of terminologies also used in Physics). There is really no point stating 'Data from whatever country hints at recovery/recession' when we do not understand whether the data represents a one-off development that may not develop into a trend! What we need is grounded and in-depth information that ‘informs’ rather than figures that keeps on mutating because someone tweaks the mathematical formula. After all, humans are the ones making conscious decisions and formulating strategies, not abstract numbers and mathematical models. For sure, the ‘unexpected’ will always occur in life (and this is what makes life exciting, isn't it?), but if the ‘unexpected’ could have been discovered and preempted in the first place, then at the very least, policy-makers have the responsibility to the people to execute this task successfully.