Tyler Cowen and Brad DeLong are fessing up. It's a Friday in August, with nothing to report but the dismal GDP figures we were all expecting.  So I'll start with the Iraq War:

1)  I erroneously believed that I could interpret the actions of Saddam Hussein.  He seemed to be acting like I'd act if I had WMD.  Whoops!  I wasn't an Iraqi dictator, which left huge gaps in my mental model of Hussein.

2)  I erroneously extrapolated the experience of World War II to Iraq.  This took several forms:

a)  I overlooked the fact that Japan and Germany were both stable bourgeois nations with solid industrial bases long before we got into the act.

b)  I overlooked the fact that we completely destroyed this nations before occupying and reconstructing them.

3)  I was insufficiently empathetic in imagining how Iraqis would feel about our invasion.  We liked the French for giving us military help during the Revolution.  Now imagine that France had invaded in order to liberate us from the British.  Even if they really did eventually leave, this would have had much worse results.  Looking back, my confidence in our liberatory powers seems terribly callous, and it doesn't really do the dead Iraqis much good that I'm sorry for it.

These things led me to underestimate the time and expense of the war (both fiscal and in human lives), and underestimate the benefits.  Maybe history will vindicate the invasion, but I can't say this seems likely.

Onto the financial crisis, where my erroneous beliefs are probably pretty typical:

1)  I recognized the housing bubble pretty early (the first mention I can find on my old blog is in 2002)--but I had no idea it would have these kinds of broad, devastating effects.  If you had asked me in 2006 what would happen as a result, I would have pictured

a)  a wealth-effect lead recession, as consumers realized they weren't as rich as they thought
b) a decline in the construction industry
c) some bank failures. 

I would not have pictured wholesale runs on the money markets, the collapse of the shadow banking system, and 10+% unemployment.

2)  I believed in the "Great Moderation".  That is, I believed that the Fed and prudent fiscal policy had, to a large extent, tamed the business cycle.  I did not believe that there was even a small risk of another Great Depression; I believed that the Fed could and would prevent the contagion from spreading.  Arguably they (and the Treasury) did, but I did not imagine anything close to that level of intervention being necessary.

3)  I believed regulators were smarter than they were.  In 2004, when the SEC decided to let the investment banks lever up to 30-to-1 instead of 12-to-1, because after all, the SEC had the tools to quickly identify and stop any contagion, I would have said they were probably right.  (I'm not sure I was aware of it).

4)  I believed bankers were smarter than they were.  Or rather, I believed the system was smarter than it was.  Individual bankers making idiotic mistakes?  Absolutely.  The occasional bank being brought low?  Sure--it happens pretty regularly, in fact.  But the whole banking system taking its entire balance sheet to the roulette table and laying it all down on a single bet? Ridiculous. 

5)  I expected any crisis to come from America's gaping current account balances and its long-term entitlement problems.  Again, arguably this was true, if you believe the "Global Savings Glut" theory (I'm inclined to).  But I expected the problems to come via a currency crisis, not a global meltdown touched off by crappy US mortgage bonds.

6)  I believed that over reasonably long time-frames, modest investments in equities would allow you to retire in comfort.

7)  I believed that securitization mitigated risk by spreading it around, rather than enhancing risk by reducing transparency.

8)  As a corollary, I believed that on the whole, Fannie/Freddie were harmless--not a libertarian ideal, of course, but hardly the worst thing the US government was doing.  I still don't think they were the worst thing the US government was doing, but I think that their distortions of the market were toxic, and have been especially so in the wake of the crisis*.

9)  I believed we knew a lot more about the Great Depression, and how to fix such a thing, then we turned out to.  I don't think we know much at all about the various roles of fiscal and monetary policy; I think the only lessons we know for certain are "don't peg your currency", and "don't let the banking system collapse".

10) I underestimated the danger that new financial instruments pose to a system in which neither the bankers nor the regulators understand their unexpected effects.

Random policy things I was wrong about:

1)  The bankruptcy reform reduced bankruptcies more, and for longer, than I expected.  Still don't think it was on net good policy, but about this empirical effect Todd Zywicki was right, and I was wrong.

2)  I think it's possible the Medicare Prescription Drug benefit may actually be saving money in other parts of the system.  (Sorry, George!)

3)  I was too optimistic about Doha; I no longer expect any serious trade liberalization for at least the next decade, maybe more. 

On the political side:

1)  I would never have imagined the 21st century United States Government effectively nationalizing an automaker and an insurance company.

2)  I was astonished that Democrats managed to hold their coalition together to pass an incredibly unpopular policy in the odd belief that it would somehow get more popular later. (Oops)

3)  I would never have predicted the emergence of the Tea Parties as an important phenomenon.

Obviously, this is just a partial list.  Long-running commentary is very humbling; you cannot retroactively revise your beliefs the better fit your predictions.  Which is why I try not to make too many predictions.  At the very least, it hopefully keeps me from firing the outrage cannon at everyone who disagrees with me on ultimately empirical questions; it's a lot easier to eat humble pie if you've sweetened it.

* Let me be clear--I'm not saying that Fannie and Freddie caused the crisis all by their little lonesomes, or even in tandem with the CRA.  I'm just saying that I think that Fannie and Freddie are a bad idea, whether they're quasi-government institutions with wink/wink guarantees, or actual government institutions with wink-wink mandates to support housing prices.