Forecasting: The Paradox of Predicting Our Future
Since time immemorial, mankind has been looking for a way to predict the future. This is especially true for the United States government, which seeks to forecast the effects of policy and legislation.
But what is forecasting and why is it useful?
Just as many historians and political scientists study the past in order to understand the present, others study the past in order to better grasp the future. In many ways it is eerily similar to what you see the local weatherman do every evening. Forecasters study trends and developments and try to make guesses as to what the likely outcomes are, and what could happen if other events occur. This is a more common event than you may think, with government organizations ranging from the Congressional Budget Office (CBO) to the CIA involved in this type of work.
Forecasts serve as important tools for policymakers to understand possible future scenarios and prepare for them. This work is important in trying to understand the impact of legislation on the economy and on government spending, and can hold sway over arguments.
The CIA and other groups do this type of work on topics involving U.S. foreign policy and national security. For example, since China has experienced annual GDP growth hovering at 10 percent over the last two decades, the most likely future scenario is that rapid economic growth in China will continue in the near future, assuming current policies are maintained. This is a sound line of reasoning: it takes into account historical trends and analysis of the current economic situation within China and provides the most likely outcome if these trends continue.
This work is crucial to defining American foreign policy in the long-term, as in this case towards an economically empowered China. If you think China will continue to grow economically, it certainly changes the way in which American-Chinese relations are conducted. Similarly, if one believes that growing instability is the problem, then one must be prepared for this.
The problems that arise from forecasting can be divided into two basic areas: problems with the forecast itself, and then what we do with the forecasts.
One big issue in forecasting is time. If one is creating a forecast for the near future, it is more likely to be accurate, but is less helpful in creating long-term solutions and policies. Long-term forecasts can be useful to policymakers in trying to create long-term solutions, but these forecasts are more likely to miss critical events, tarnishing their worth.
While one can predict something when speaking vaguely, it is nearly impossible to forecast specific events. Many foresaw some form of civil disorder in the Middle East occurring sometime in the near future because of unemployment and other economic/political troubles. No one predicted a 26-year-old Tunisian setting himself on fire would ignite the collapse of the Tunisian government and send waves throughout the Middle East. While this event may be too specific, I have not heard of anybody that thought the Egyptian and Tunisian governments would be overthrown in 2011, that a civil war would break out in Libya, and that protests would roil Syria, Yemen, and Bahrain.
These can be called ‘outlier’ events in terms of forecasting. These events have a small chance of occurring at any given moment, but have a very large impact when they do. Two other events which stand out as outliers from the past 20 years are the collapse of the Soviet Union and the events of Sept. 11. Neither event could have been reasonably predicted at any moment to have been a likely occurrence, yet these are two of the most defining events of the last two decades.
While outliers may be the events that have the most impact, they are outliers for a reason. They are simply unlikely to occur. The more time we spend researching and preparing for these events, the less time and money is spent preparing for the more likely outcomes. The obvious answer would then be to balance our attention between the scenarios, but this is a trap that only can be conceived as effective in the present. When an outlier occurs, people will be outraged that more time and money was not spent preparing for it. When an outlier is prepared for but does not occur, people will view it as a waste of taxpayers’ money.
So if the problems of forecasting are so vast, then what good is it to even try to do so?
First, one must realize the importance of these forecasts is in their utility as tools to help prepare for the future, not predict the future. So long as policymakers and the public understand that there is a certain risk of something happening, then it is no longer entirely in the hands of those who did the forecasting.
Finally, we just need to accept that some future events are unknowable, and expect that our intelligence agencies and others involved in forecasting will not always foresee important events. As the Director of National Intelligence James R. Clapper Jr. stated, intelligence agencies can “reduce the uncertainty for decision makers, but not necessarily eliminate it.”
That being said, I still don’t trust the weatherman.