An epithet I've heard often in the past is that economists are simply "good at predicting the past," and in all honesty it doesn't seem like an unfair criticism. Economics is a predictive science, but it appears that economists aren't all that good at prediction. Why is this?
I think there are a number of answers to this. The first is idiology-most economists have a particular viewpoint that they are wedded to, so you don't get a lot of consensus on matters. Think about the current financial crisis...there are leftist economists arguing that the problem lies in the capitalist system and is the result of greedy rich white guys trying to profiteer from the nation's poor. On the right you might hear that it is a problem of misregulation-the necessary checks are in place but for inept bureaucrats, and that the greed isn't the problem, people committed crimes of fraud. And of course I'll tell you that the problem is entirely government created through very loose monetary policy combined with excessive subsidies in the housing market. It's not like any of us are lying-we all think we hold the truth. But it turns out that these explanations are mutually exclusive-only one can be right.
To understand this, and sort of differentiate this from a "harder" science like medicine, think about supply and demand. There is a supply of economics and a demand for economics, just as there is a supply and demand for medicine. On the supply side, you do see that occasionally doctors differ-one patient may see three different doctors about a set of symptoms and get three different diagnoses. And when it is demonstrated that one diagnosis is incorrect, that doctor can (as economists do) easily justify why their diagnosis was wrong but it was still a good diagnosis. For example, a doctor can say that their diagnosis was the best given the evidence presented, however if given another piece of evidence, they would have gotten the right answer. Economists are similar-they can make untrue predictions and blame their not coming to fruition on changing factors in the economy. This is why we always qualify statements with the phrase ceteris parebus; we are confident that if nothing changes our prediction will be right, which gives us the easy out if our prediction does not come to fruition, since something will always change. And of course, if a doctor is always wrong, he will soon lose his income stream, meaning bad doctors will tend to get weeded out by the market. I don't think that matters (at least as much) for economists because of what is happening on the demand side.
On the demand side, we see another phenomenon keeping bad predictions/explanations of the economy going, which is related to the ideology comments above. If you are sick, do what do you want? A doctor to tell you that you have disease "x" and treat disease "x" or a doctor to tell you what disease you actually have and treat that? Most people who go to a doctor simply want a correct diagnosis. It is true that there are exceptions to this, but I think that this is far less prevalent in medicine than it is in economics. If a labor union asks an economist to write a report on John Howard's Industrial Relations law reforms, and that economist writes either that they are great or that the government didn't go far enough in deregulating labour markets, what is the union going to do? Post it on their website and trumpet the benefits of free markets? Or find another economist who will argue that the IR laws will put worker rights back 50 years? I think that they will do the latter 100% of the time. Economics is unique in that the demanders of our services want us to tell them what they already "know."
Another key difference between economics and the "hard" sciences is the combination of the incredible complexity of the economy and its rapid rate of change. Yes, the physical universe is extremely complex-and yet the laws that govern the physical are immutable. The human body is incredibly complex, but human evolution occurs at a snail's pace-a drug that had a 70% success rate at curing disease "x" when it was discovered/invented in 1900 probably has the same likelihood of curing that disease today. It will probably be equally efficaceous in 3000, as would it have been in 1000 had they known about it. Sure, from time to time a new deadly disease will pop up, but for every avian influenza there are hundreds of thousands of new technologies, laws, regulations, financial/business innovations that have a huge change on the economic landscape. Every quarter in the US, there are about 7-9 million jobs that are destroyed as a result of these changes, and another 7-9 million new jobs that are created as well. I believe (and I could be wrong about this) that the rate of change in the forces that affect the economy dwarfs those of the physical world.
What drives this I think is the difference betweens the subjects of analysis in economics (and the other social sciences) and the subjects of analysis in the hard sciences. What does a white blood cell do? As a non-doctor, I may be way off here, but my understanding is that they are in your arteries and veins attacking everything that shouldn't be there. But they are not making decisions on the margin. To personify them a bit, it is as if they are hard coded to swim around in a certain way and have a list of what should be there. If they encounter an object, they check
their list to see if it should be there. If it should, they leave it alone. If not, they go Chuck Norris on it. They follow rules, never making decisions, because they are not sentient. Humans, on the other hand, are. They have the ability to choose the rules they follow, modify them if they aren't working, or be entrepreneurial and make a whole new set. If those new rules are good, they get rich. And it turns out that these rules are very private-sometimes even the people choosing them or inventing them don't know what they are exactly. It is very hard to make accurate predictions when there are no rules that we know every person will follow-however we can look at what has happened, attempt to figure out what rules were being followed, and in so doing, predict the past. Does this mean that economists have nothing to offer? If I thought that I'd have to quit my job, so I must think that we have more to offer than just being historians (not that there is anything wrong with being historian, mind you). Some things are unpredictable-new technologies, for instance. But it turns out that entrepreneurial action is quite rare, and most people roughly follow the same sets of rules most of the time. So while we can't predict WHAT will change all the time, we can predict what will happen-that is what ELSE will change-as a result. We can predict changes on the margin pretty well, we predict inframarginal changes rather poorly. And that is something.