Site Meter

Monday, April 16, 2012

Simon Wren-Lewis again.  I really don't mean to criticize everything he writes, but I find his blog extremely stimulating.

He wrote

Monetary policy works by changing the real interest rate. At the zero lower bound monetary policy loses its power, unless it can influence inflation expectations. However inflation expectations for consumers will also depend on the evolution of sales taxes (indirect taxes like VAT). A pre-announced increase in sales taxes will raise expected inflation, and so reduce the real interest rate faced by consumers at the zero lower bound. In this sense changes in sales taxes can mimic monetary policy.


update: After thinking more (roughly that means after thinking at all) it is really hard for me to see how Wren-Lewis's proposal could fail to work (not impossible - it is never impossible -- but requring one extreme implausible assumption after the other). I will critique my critiqu below said critique.


I comment:


There are three relevant real interest rates.  One is nominal interest minus consumer price inflation -- this is the rate which should affect consumption.  The other is the nominal interest rate minus producer price inflation.  This should affect investment along a balanced growth path.  OK also the nominal interest rate minus the rate of increase of prices of capital goods should shift the timing of investment.

My concern is that the second real interest rate is not reduced by rising VAT.  If I am deciding whether to invest, I compare my nominal revenues which depend on the price I get for my product to depreciation plus nominal interest.  The fact that consumers will have to pay more does not make it profitable for me to invest.

I admit again that if I am deciding when to invest, I will invest now when I can get equipment and building materials without paying higher VAT.

But it still seems to me that you are thinking of a model in which real interest rates have an economically important effect on consumption.  I think there is essential no evidence of any such effect (last I heard GMM estimates of the aggregate intertemporal elasticity of substitution were negative.  In any case, empirical estimates of it are tiny.

I guess that it is still true that DSGE models require reasonably high intertemporal elasticities of substitution to come close to fitting the data.  But they don't fit the regression coefficient of consumption.   growth on the real interest rate.  Here I fear that focusing on a micro founded model which most nearly fits the data has possibly lead you to believe something about the relationship between real interest rates and consumption which is simply inconsistent with the data.

update: OK two things. First the pulling forward of investment almost has to be important. The idea is that firms invest now to avoid paying VAT on equipment and such which they buy (I just paid VAT to have my research project buy a new computer -- in theory that is productive capital although I will have to work on producing something with it). The only case would be irreversible investment in capital with low resale value by a firm which is only investing at all because of extremely low real interest rates right now. The plausible case is that investment is currently low, since demand is low so there is spare capacity (or in more neoclassical theory terms the capital labor ratio is high and the marginal product of capital is low). In this plausible case, firms will invest when things are back to normal and will invest right now not later to avoid paying taxes later.

Also, the stylised fact that the intertemporal elasticity of consumption seems to be low is based on regressing the growth rate of consumption on an econometricians estimate of the expectable real interest rate. This might mainly show that econometricians and normal people have extremely different estimates of inflation. A very clear extremely forecastable increase in prices due to a scheduled increase in VAT might have a large effect on consumption even if the econometricians' estimates generally don't have much to do with real world consumers.

No comments: