Adam Ozimek has a good post showing that economists are not ideologues, and that they change their mind when new information comes in. So I tried to think of some examples that would apply to me. I'll start with early in my career, and then more recent:
I favored a guaranteed annual income in college, and switched to favoring a low wage subsidy in grad school. I was convinced by theoretical arguments, not empirical data.
I favored Friedman's k% money growth rule in grad school, and switched to favoring NGDP targeting in the 1980s, as velocity became unstable. My views moved away from Friedman and Schwartz's account of the Great Depression, as I began researching that area.
I became less confident that modest increases in the minimum wage would lead to substantial job loss, due to studies such as Card and Krueger. I still believe minimum wage laws have costs apart from job loss, such as reduced working conditions, and would still prefer a low wage subsidy targeted at adults. But I now think the job loss effects are much less clear than before.
I've recently become less confident that tight money leads to lower interest rates (but not this week!!) I still think that a very tight money policy leads to low interest rates, but less extreme tight money policies seem to raise long term rates more often than I had expected. I seemed to underestimate the liquidity effects of QE. I now see it as a completely open question, not well understood.
I've become less confident of the supply-side effects of state income taxes. I still think that a lack of state income taxes helps explain why places like Texas, Washington, Tennessee, South Dakota, Nevada, Florida, etc., tend to grow faster than their immediate neighbors. But I now think that in many cases the advantages of urban living in places like New York and California overwhelm modest tax cuts in places like Kansas. I've been especially influenced by the recent reversal of fortune between (urban) Massachusetts and (rural) New Hampshire, something I never would have predicted. At the national level I'm still a moderate supply-sider, more influenced by the stylized fact that Europe is poorer than America, than by the stylized fact that the economy did well in the 1990s. But I don't discount the latter experiment, which explains why I am only a "moderate" supply-sider.
The Hive Mind has made me view IQ as being more important in economic development, as well as in savings rates, good governance, etc., than I had assumed.
I'm less in favor of intellectual property rights than before, and was growing increasingly worried about their impact on economic equality even before this recent study came out.
Most often my views just get nudged slightly this way or that, as there is already so much evidence that new evidence is rarely earthshaking. Thus Tyler Cowen's recent interview with a hedge fund guy made me slightly less confident in the EMH, but didn't really shake my view that it's a useful approximation of reality.
And one final point. The complexity of the world makes it hard to do empirical tests. Tyler mentioned that the relative performance of China and Eastern Europe made him less confident that mass privatization was a good idea. I don't see the evidence that way, for a number of reasons:
1. The parts of the former Soviet bloc that privatized faster tended to do better, or at least less bad.
2. China's a very different place, which started from a position of low-income peasant agriculture. The SOEs have been the weakest part of the Chinese growth story, and it is the private sector that has done best.
When problems are very complex, such as economic development, I'm skeptical of "models". To me, the most sensible approach is to admit how little we know about linkages, and look at each issue on a case by case basis. In that case, theory would play a major role. The presumption should always be in favor of privatization, deregulation and free markets, except where there are pretty well-established exceptions, such as pollution regulations or income redistribution. There is no well established evidence that SOEs make sense, so countries should get rid of them. China probably did well despite its SOEs.