Are We in a Liquidity Trap ?

In which Robert Waldmann finds economic theory useful.

A question: when does deficit financed public spending cause higher interest rates which cause private investment to be lower than it would be other things (including the spending) equal.

Proposed answer, not when we are in a liquidity trap.

Depending on the definition of liquidity trap this answer is wrong or not relevant to the current situation.

Now I confess that some economists (including me) sometimes rely on an equivocation to argue that we are in a liquidity trap and therefore there will be no crowding out via interest rates. The evidence that we are in a liquidity trap is that safe short term interest rates are very low. However, for there to be no crowding out via interest rates it is necessary either that deficit financed public spending doesn’t affect long term interest rates, or that long term interest rates don’t affect investment.

Also note that investment depends on factors other than just interest rates. In the data, high investment is associated with high GNP growth as well as with low interest rates, so even if there is some crowding out via interest rates, the net effect of the spending on private investment can be positive.

One justification for looking only at safe short term interest rates and the whole yield curve is that, if safe short term interest rates are zero expansionary monetary policy is pushing on a rope (as is absolutely demonstrated by recent experience). This means fiscal policy is the only available demand policy so “is there any crowding out via interest rates” is not the key issue. So we should define a liquidity trap as a period with safe short term interest rates are zero and define a new term for the whole term structure is zero — say a super liquidity trap and get to a second proposed answer

Not when we are in a super liquidity trap. After the jump I criticize my second proposed answer.

It is clear from the data that increased deficits and expected future deficits cause higher long term nominal interest rates. Direct measurement of real interest rates via markets for TIPS (indexed to the CPI) is a recent phenomenon. The anticipated effect of Bush deficits on long term interest rates didn’t IIRC show up. The counter-argument basically is that the period whith both TIPS and insane US fiscal policy is identical to the period of insane People’s Bank of China policy.

However, it is also possible that deficits affect long term interest rates only via inflation. Economic theory suggests that investment should depend on real not nominal interest rates. So it is possible that there is no crowding out via interest rates due to deficit spending.

Oddly, there is a strange argument which is the same up to the last sentence and then draws the opposite conclusion. The argument is high deficits cause high expected inflation which causes high nominal long term interest rates which cause reduced investment. One often meets this argument in non-academic discussion of the economy. My reaction has always been “huh?!?” (also before I took my first economics course). I don’t think it is possible to write down an economic model in which people are rational and price indices are available and indexed contracts can be written, in which long term nominal interest rates matter.

I now discover that I find economic theory useful in my efforts to understand this argument. I don’t generally find economic theory useful. Certainly my opinions on the economy and economic policy haven’t been influenced at all by my own work in economic theory which is modest but not zero.

However, I didn’t find economic theory useful in a way which is flattering to economic theory. My reasoning is that an argument which is so theoretically unsound probably wouldn’t survive if it weren’t supported by actual experience. Now this argument might rely on an underestimate of human idiocy, but I find it convincing.

If all firms considered only their expected long term real interest rate when deciding on investment, then people would notice.

It is very possible that nominal interest rates matter because debt contracts aren’t indexed. For example take floating interest rate loans with fixed nominal repayment schedules. Higher inflation implies a more rapid rate of required real repayment. Ooops. There are many many such contracts. It is hard to reconcile the existence of such contracts with rationality but they exist. If people look at debt service ratios and count nominal not real interest as the cost of debt then they will be confused by inflation. And so forth and so on.

To me the fact that the argument makes no sense in theory leads me to suspect that it is important in practice.