« Is Ontario to Canada as Spain is to Europe? | Main | CEO Pay and the Grandfather Fallacy »


Feed You can follow this conversation by subscribing to the comment feed for this post.



I think I understand already that no real world outcome could violate an accounting identity, and agree that NFA = NFL at the aggregate level. But my instinct (often wrong), is that what I'm saying still holds.

Taxing to pay down bonds might be equivalent to default with no tax for the aggregate sector, but it is not equivalent at the level of bondholder vs non-bondholder household. The bondholder expects a certain level of real consumption from the taxpayer, equal to the real value of the bond. In the default scenario, he receives nothing.

Considered as an aggregate, household savings equals their tax liability; but the asset is not held by all households. In order to settle their obligation to the bondholders, households require infinite output, which they do not have. They cannot pay the tax; the government must default on the bond.

To get further perspective on where I’m coming from, one could consider an equivalent case in which the bond is owned by households in another country. Since the liabilities of the govt can no longer be seen as ultimate liabilities of the household sector, it’s easy to see that although we can have NFA = NFL at all times, the absence of default is not guaranteed.

And taxation is equivalent in flow terms to negative saving, so as NFA is taxed down, it unwinds the stock of NFA saving.

Right—but a real stock does not arise out of a sequence of nominal flows. As the nominal stock of debt is unwound, there’s no way to tell ipso facto what happens to the real stock. Perhaps the nominal stock is unwound to near zero, but the real stock does not change. Anything is possible when we consider the problem without conditioning on other real economic factors.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad