Global Warming & Supply/Demand Shock

Philip M. Boffey recently wrote an article for The New York Times entitled 'The Evidence for Global Warming'. Here is the last section of this article followed by some parallel thoughts on economic theory by Mark Thoma.

With all of the most prestigious scientific organizations convinced that global warming is an increasing menace — and with the vast majority of research articles in leading scientific journals tending to support that consensus — it would seem wildly irresponsible not to believe it is important to curb emissions. These are the institutions with the most expertise, and they have been studying the issue in unparalleled depth and breadth. Their judgment deserves the utmost respect and attention.
...

The world keeps pumping greenhouse gases into the atmosphere in what amounts to a huge uncontrolled experiment, and a gamble that all will turn out fine. But ... if the worst-case scenarios turn out to be accurate, we could be dooming much of the planet to a very unpleasant future.

In response, Mark Thoma wrote:

To answer the global warming question, scientists have to separate the cyclical part of temperature variation from the trend, and then understand the sensitivity of the trend (and cycle) to changes in greenhouse gases.

Economists face a similar problem. An important debate in economics is how much of the variation in GDP is caused by supply shocks, and how much is caused by demand shocks. To answer this and other important questions, the cyclical part of GDP must be separated from the trend component. (Demand shocks have short-run, or cyclical effects, but do not affect the long-run trend; supply shocks can have both short-run and long-run effects. Thus, the trend is dependent upon supply side factors while the cycles can be affected by both demand and supply shocks. The cyclical variation is generally thought to be dominated by demand shocks, though that is not universally accepted).

In order to differentiate a change in the trend for GDP or other macroeconomic variables from a change in GDP around the trend, long time-series are needed, and the longer the better. For example, are recent increases in GDP growth driven by high levels of productivity part of a cycle where growth and productivity will return to lower historical levels with time, or is this a change in trend so that we can expect permanently higher productivity and growth? The answer is important for all sorts of questions such as how high tax collections - and hence the deficit - will be in the future.

Unfortunately, we do not have the equivalent of samples from ice cores from the distant past to guide us -- reliable economic data doesn't exist prior to around sixty years ago, and we are often limited to forty or so years of data (since 1959 since money data didn't exist before then). Because of this, our ability to differentiate between the trend and cyclical components of economic variables is not as precise as we would like. New theory could help, but a longer span of data would be even better.