If the economy improves and unemployment drops, Obama can take credit. If it fails to improve and unemployment rises, though, he can say he averted an even worse showing. Republicans will take the opposite tack — attributing any improvement to the natural resilience of the economy and blaming the administration if things get worse. And neither side will really know who’s right.
A scientifically trained politician (or journalist) often has good reasons to simply say, “I do not know.” But in politics, or especially in politics, such statements are considered a sign of weakness, and therefore, political suicide. To be a successful politician you need to signal strength, not epistemological sophistication. To a lesser extent this applies to (partisan) journalists who write about macroeconomic matters as well. If Paul Krugman would just confine himself to sketching a number of different scenarios without taking sides, many people would find his columns boring and would look for ammunition to engage in political debate elsewhere. The addiction to politics is so strong that we are prepared to throw everything we have been taught about valid reasoning and how science operates out of the window.
Related reading: Looking at the world through politically-colored glasses
On economic forecasting: A positive-sum game against nature