A lot of them aren't really irrational, or maybe it depends on your definition. For example the endowment effect - it would be stupid to part with your only ticket unless the buyer pays a large premium.
You didn't understand. The point was that the same person would pay $X for good A, but will only sell good A for $B, where $B >> $A..
Let's pretend that if I want a widget I am willing to pay $1 (but not $2) for it. That means I would prefer to have the widget over $1, but prefer to have $2 over the widget. However, if I have a widget I'm willing to sell it for $3 (but not for $2). That implies that I am would rather have the widget than $2, which is clearly a contradiction with the previous conclusion. This is irrational. (And it violates several of the key assumptions underlying classical economics).
The endowment effect is causing very serious problems right now.
This is probably the reason why people are unwilling to sell houses at below-bubble prices, even though they would be unwilling to buy at current prices.
If someone figures out how to arbitrage this, they will be rich.
Yeah I got that. There's a number of possible resolutions, eg selling the widget might be an unnecessary hassle. The problem is we're starting with different sets of premises. As I see it, a rational person considers many factors in the decision to sell good A for $X. The amount he'd pay for it is just one factor & probably the least important one.
In the real world there are transaction costs - but for most people in the example given (college students selling event tickets) they don't come anywhere near $1000.
