[jerf]

"but this would create a perverse incentive to ALWAYS default on loans."

Statically, yes. However I'm having a harder time mentally analyzing what this would do dynamically, which is a much more interesting problem, since the real world is not static.

My first approximation is that because "everyone" would know this is something they can do with their debt, that lending would consequently become much more rare, and lenders would be much more careful about securing their loans. The initial impact on the economy would probably be sharply negative; what happens after that is probably a "your guess is as good as mine" situation, even amongst economists. Attitudes about debt have varied widely throughout time and space. Some will say tightened lending will wreck the economy longterm, others would suggest that loose lending has already wrecked the economy and the initial shock would simply be paying down damage already done, after which the economy would be much healthier. Which side you fall down on probably has more to do with ideology than education; given how much trouble economists have explaining even our current economy I'm not willing to give even experts much credence for explaining how such a different one would work.

[dchmiel]

You're right on that dynamically it is a very interesting problem and something that from my exposure to economics, economists seem to spend very little time analyzing.

Changes beget other changes and so on. Finding the steady state level of change would be challenging yet interesting to model.

[yetanotherphd]

I have a PhD in economics and that is the first thing that an economist would think of (in my case it wasn't, but I did eventually realize it...).

One reason why many of the the "crazy" ideas posted on HN don't make it to policy, is that there are people trained in economics who are able to see the flaws in such policies.

EDIT: and in case you were curious, this wouldn't generally be considered a dynamic problem in economics. Once the policy was announced, interest rates would immediately change to a new equilibrium, which accounted for people's higher tendency to default. So the problem is simply to calculate the new static equilibrium.

[jerf]

Yes, in this case I merely mean to bring in the idea that changes are reacted to. I was using static vs. dynamic more in the physics sense, where "static" can be looked at as a snapshot of the system you're looking at, whereas dynamic brings in the concept of time. I agree that all else being equal, there's every reason to believe that this sort of change would produce a new static equilibrium, at least in terms of debt itself. (My doubt is more about the net effect on the economy as a whole, as debt seems to be a matter of some debate even today. The dominant economic thought seems to be pretty comfortable with it, I have to admit I'm coming around to more of a "Black Swan" sort view where I think the dominant consensus is overvaluing it, which interestingly also accords with a lot of historical thinking on the topic albeit not with that amount of mathematical backing.)

[yetanotherphd]

We usually call the immediate effect (in this case on interest rates) partial equilibrium, and the "full" effect including all flow on effects, "general equilibrium". And there are models which include dynamic and random effects (dynamic stochastic general equilibrium - DSGE - models). Can't tell you much about them apart from the name :-)

Anyway it seems like messing with debt contracts is a bad way to change the interest rate, if that's all you want to do, since that is precisely the lever that the Federal Reserve controls anyway.