I'm really enjoying the sub-discussion with horatio, though I suspect it's only of interest to the two of us. Sorry to everyone else for it being in the thread, but it's still vaguely relevant as it will help you keep your lottery winnings for longer... 
There are two issues to separate out here:
1) Does rational choice theory (and therefore classical economics) always explain real people's actions?
2) Does it actually matter if it doesn't, provided the model generated by classical economics is sufficiently close in describing the phenomenology anyway?
Taking these in reverse order, because the second is the deeper conceptual issue....I think it does matter.
I know Friedman's (I think it was him anyway) falling leaves argument but to extend his analogy, that means that at most the model can predict are the falling leaves, not the tree's biology. This failing doesn't matter provided that all you're interested in is the phenomenology of falling leaves... but what about all the trees around that are not deciduous?
I would suggest that this pattern of thinking regarding models and what their job is, is unhelpful in the longer term task of understanding the reality of human actions including what they do with money. It's akin - although not as extreme, obviously- to Ptolemaic models of the solar system: it works at the job of prediction (mostly) but at the cost of introducing assumptions and complexities that do no reflect reality.
Classical economic theory has the same issue: it describes some of reality wonderfully well; it also abstracts some other aspects of reality well enough to explain them sufficiently (albeit with incorrect assumptions as to motive); but it doesn't explain everything, and some of the explanations invoked in the marginal cases are so jarring to how human psychology actually functions that if there's another model available that better approximates reality, it seems correct to me to incorporate it into our range of models to use, at least for those contexts.
It is possible for both classical economics and newer schools to co-exist as heterodoxy; they simply do different things because they approach the problem from opposite directions. I don't view the conflict as being destructive, but synergistic.
Returning to point 1, I think your explanation of why equivalence breaks down is a prime example of one of those cases where classical economics gets you to the right result (we both agree on the phenomenology of equivalence breaking down) for psychologically very jarring reasons. Your explanation is perfectly fine within classical theory and since experimental testing is impossible, on one level that's enough ("falling leaves" again).
But human behaviour is actually rarely rational, even using the accepted narrow economic definition of rational choice being simply about maximising perceived gain. The reason is that more often than people like to believe, our actions are done so automatically/unconsciously as to eliminate any forethought about the potential impact of our action (this is actually not unique to our interaction with money, but it's only when it occurs with money that it has an impact on economic theory). This has absolutely nothing to do with intelligence or stupidity, by the way, it's just the way humans behave when not actively shining the "spotlight of attention" onto a topic.
Now, historically, spending money has been something humans paid lots of attention to. But increasingly, money has become abstracted from something they feel and have to choose to spend. This creates an increasing number of opportunities for money to be spend incidentally (or even accidentally!) without a person choosing to spend it. And if you're not actively choosing while you spend, ideas of rational choice maximising gain go out the window. Non-classical theories can (sometimes) cover these circumstances so have valid utility. You don't need ALL behaviour to be irrational in this sense for heterodox economics to be useful, merely enough of it to cause problems or unwarranted assumptions in classical models alone.
To return to our jumping-off point of equivalence: if you actually sat down with someone and asked them whether they viewed current national debt as equivalent to future taxation, of course they would view them as equivalent. However, for the vast majority, it's only because you're temporarily constraining their freedom of action by forcing them to answer the question that they've even bothered to consider it. If they're naturally thoughtful people, they might even remember the fact (most will not). However, even that thoughtful subset will not necessarily alter their behaviour in light of that fact. And without that explicit sit-down conversation with them about equivalance, lots of expenditure is irrational This is partly due to pressure from organisations (eg advertising, which certainly doesn't work by providing information to optimise gain, as classical theory would have it), partly due to habit persistence, partly due to time preference, and partly because a large mass of people simply do not pay attention to what the government does and so cannot incorporate that knowledge into their choices. If you don't know the government is building up dangerous amounts of debt, you cannot possibly consider it equivalent to increased future taxation.
Here's a little unproven pet theory of mine that I haven't seen discussed before:
- classical economics is a better way to model recessions
- heterodox economics does better in periods of steady growth.
My feeling is that when the economy is doing well, a larger mass of people do not pay attention to economics. Therefore there is more opportunity, on average, for irrational action to seep in at the edges. At times of crisis, people pay closer attention, actually listening to the news, and so rationality looms larger, making classical theory more valid on its own.
As you can tell from the above, while I certainly highly value the contributions of (neo)classical economics, my natural leanings are inclined to heterodox interpretations. This is perhaps an unsurprising position for me to take, given that I'm a shrink by profession so naturally I think human psychology is important in governing action, and I see lots of evidence of irrationality of action (again, using the economic sense of the word here). Economics is a just little side-interest that I indulge occasionally when I have time to read about it, as I find it an interesting case of applied psychology, so forgive me I'm less efficient with the terminology and concepts than an economist would be.
There are two issues to separate out here:
1) Does rational choice theory (and therefore classical economics) always explain real people's actions?
2) Does it actually matter if it doesn't, provided the model generated by classical economics is sufficiently close in describing the phenomenology anyway?
Taking these in reverse order, because the second is the deeper conceptual issue....I think it does matter.
I know Friedman's (I think it was him anyway) falling leaves argument but to extend his analogy, that means that at most the model can predict are the falling leaves, not the tree's biology. This failing doesn't matter provided that all you're interested in is the phenomenology of falling leaves... but what about all the trees around that are not deciduous?
I would suggest that this pattern of thinking regarding models and what their job is, is unhelpful in the longer term task of understanding the reality of human actions including what they do with money. It's akin - although not as extreme, obviously- to Ptolemaic models of the solar system: it works at the job of prediction (mostly) but at the cost of introducing assumptions and complexities that do no reflect reality.
Classical economic theory has the same issue: it describes some of reality wonderfully well; it also abstracts some other aspects of reality well enough to explain them sufficiently (albeit with incorrect assumptions as to motive); but it doesn't explain everything, and some of the explanations invoked in the marginal cases are so jarring to how human psychology actually functions that if there's another model available that better approximates reality, it seems correct to me to incorporate it into our range of models to use, at least for those contexts.
It is possible for both classical economics and newer schools to co-exist as heterodoxy; they simply do different things because they approach the problem from opposite directions. I don't view the conflict as being destructive, but synergistic.
Returning to point 1, I think your explanation of why equivalence breaks down is a prime example of one of those cases where classical economics gets you to the right result (we both agree on the phenomenology of equivalence breaking down) for psychologically very jarring reasons. Your explanation is perfectly fine within classical theory and since experimental testing is impossible, on one level that's enough ("falling leaves" again).
But human behaviour is actually rarely rational, even using the accepted narrow economic definition of rational choice being simply about maximising perceived gain. The reason is that more often than people like to believe, our actions are done so automatically/unconsciously as to eliminate any forethought about the potential impact of our action (this is actually not unique to our interaction with money, but it's only when it occurs with money that it has an impact on economic theory). This has absolutely nothing to do with intelligence or stupidity, by the way, it's just the way humans behave when not actively shining the "spotlight of attention" onto a topic.
Now, historically, spending money has been something humans paid lots of attention to. But increasingly, money has become abstracted from something they feel and have to choose to spend. This creates an increasing number of opportunities for money to be spend incidentally (or even accidentally!) without a person choosing to spend it. And if you're not actively choosing while you spend, ideas of rational choice maximising gain go out the window. Non-classical theories can (sometimes) cover these circumstances so have valid utility. You don't need ALL behaviour to be irrational in this sense for heterodox economics to be useful, merely enough of it to cause problems or unwarranted assumptions in classical models alone.
To return to our jumping-off point of equivalence: if you actually sat down with someone and asked them whether they viewed current national debt as equivalent to future taxation, of course they would view them as equivalent. However, for the vast majority, it's only because you're temporarily constraining their freedom of action by forcing them to answer the question that they've even bothered to consider it. If they're naturally thoughtful people, they might even remember the fact (most will not). However, even that thoughtful subset will not necessarily alter their behaviour in light of that fact. And without that explicit sit-down conversation with them about equivalance, lots of expenditure is irrational This is partly due to pressure from organisations (eg advertising, which certainly doesn't work by providing information to optimise gain, as classical theory would have it), partly due to habit persistence, partly due to time preference, and partly because a large mass of people simply do not pay attention to what the government does and so cannot incorporate that knowledge into their choices. If you don't know the government is building up dangerous amounts of debt, you cannot possibly consider it equivalent to increased future taxation.
Here's a little unproven pet theory of mine that I haven't seen discussed before:
- classical economics is a better way to model recessions
- heterodox economics does better in periods of steady growth.
My feeling is that when the economy is doing well, a larger mass of people do not pay attention to economics. Therefore there is more opportunity, on average, for irrational action to seep in at the edges. At times of crisis, people pay closer attention, actually listening to the news, and so rationality looms larger, making classical theory more valid on its own.
As you can tell from the above, while I certainly highly value the contributions of (neo)classical economics, my natural leanings are inclined to heterodox interpretations. This is perhaps an unsurprising position for me to take, given that I'm a shrink by profession so naturally I think human psychology is important in governing action, and I see lots of evidence of irrationality of action (again, using the economic sense of the word here). Economics is a just little side-interest that I indulge occasionally when I have time to read about it, as I find it an interesting case of applied psychology, so forgive me I'm less efficient with the terminology and concepts than an economist would be.
