If a weather expert tells you what the weather will be on a specific day next year, you can safely ignore him. If he tells you a hurricane is heading your way, it's a good idea to get out of the way, even if the storm ends up turning. That's playing the odds.There's much more in his blog. Scott is also funding a survey of what some 500 economists think about some economic policies of the two candidates running for the American presidency. The results will be interesting.
Likewise, if an economist tries to tell you where the stock market will be in a year, you can safely ignore that. But if he tells you a gas tax holiday is an unambiguously bad idea, that's worth listening to, especially if economists on both sides of the aisle agree.
If you think it is okay to ignore economists because they are so often wrong, you're looking at the wrong questions. Economists are generally wrong with complicated models but right about concepts. For example, they know that additional domestic drilling won't make much of a dent in the energy problem. And they know that free trade is generally good for all economies. (You can argue with my examples, but the point is that some things are generally known by economists while not being understood by the general public.) [link]
As Scott finds in his survey, most economists are moderate Democrats. But also the kind who likes things like free trade, and in general freer markets than your average democrat. Jagdish Bhagwati, perhaps the one of the world's most staunchest free trade economists is a registered democrat and an Obama supporter from the start, although he disagreed with Obama's populist NAFTA bashing.
My general observation is that if you have a liberal/cosmopolitan upbringing and eventually learn economics, you tend to be more sympathetic to policy proposals and ideas of classical liberalism (i.e. support more individual freedom and freer markets) or as Scott says in the current American political language, be more socially liberal and economically conservative.