DeLong on interest rates

There’s an interesting piece by Brad DeLong in Saturday’s Australian Financial Review, which I can’t find anywhere online.He lists four reasons why interest rates can be expected to rise in the future. The first two are just two facets of the same thing – the embrace of deficit finance by the US Republicans in both Congress and the executive branch. The third is closely related: the failure of Western European governments to address the coming fiscal crisis associated with retirement income policies. All of these are currently being offset by the willingness of the Chinese and Indian central banks to buy US and Euro bonds in an effort to maintain competitive exchange rates. But this can’t last forever.

The fourth factor is “the inability of Western European governments top enact sufficiently bold liberalising reforms to create the possibility of full employment, together with the failure of Western European monetary policy to be sufficiently stimulative to create the reality of full employment”. I have some doubts about this. Leaving aside any questions about the efficacy of “liberalising reforms”, it’s far from clear that the adoption of stimulatory monetary policy in the short run is conducive to low nominal interest rates in the long run. I would have thought that just as stimulatory deficits in a recession need to be offset by larger surpluses during booms, an active monetary policy implies a larger variance in interest rates over the cycle.

More significantly, it seems to me that, on a crucial point Brad has the argument backwards. He says “Believers in low interest rates … point to rapid technological progress, which has boosted output.” In general, technological progress ought to create new investment opportunities at high rates of return, while the associated increase in asset values should raise current consumption. This should raise the real rate of interest, not lower it. In fact, this is precisely the argument Brad makes in relation to India and China.

On checking, I was surprised to find out that the ratio of nonresidential gross investment to US GDP is near an all-time low, at 10.0 per cent (you can get the data from the Bureau of Economic Analysis . Taking account of the increasing share of computers and software, which have high depreciation rates, it seems likely that the net investment share is lower than ever. Maybe this can be explained if all the technological progress takes the form of capital-saving reductions in the cost of computing and telecommunications. This would reduce demand for capital (but ought to increase demand for labor, something that has evidently not taken place).

3 thoughts on “DeLong on interest rates

  1. This has not much to do with this post, but I just wanted to inform you that I stumbled across your curriculum vitae online and am dumbfounded by how productive you appear to have been, and the sheer volume of newspaper and journal articles you’ve been churning out. I’m impressed, but that’s probably just because I’m a schoolkid whose pet obsession is economics, but who wants to study Arts and Law so he can work at the Financial Review someday.

  2. This is a puzzle to non-economists . Surely such a fundamental thing as interest rates over the short term would be a source of agreement amongst them. Since it isn’t there seems to be something rotten in the state of Denmark. Either the models are not taking into account enough variables ( and hence are too variable ) or the models are wrong .

  3. It would seem to me that ‘the’ interest rate movement over time, could well be similar to the CPI, in the sense that while they are useful for comparison in short time frames, they are less useful over longer ones. This is largely the problem of qualitative changes in the measures. A simple consumer example would be that VCRs or DVDs, or mobile phones and internet, have not even existed for price comparisons over our lifetimes. In terms of qualitative differences, both in the consumer and investmenty fields, think of the qualitative improvements in the motor car in the last 30yrs. These have never been cheaper or more reliable than they are today. In general this is reflected in the calibre of machinery and equipment for investment purposes also. Computer controlled machinery and equipment has also come ahead in leaps and bounds. For example, laser cutting equipment has the capacity to produce complex metal shapes, without costly tooling up, in a fraction of the time previously. A CAD drawing fed into the machiine can be replicated on any base material loaded on the cutting bed. In general though, the quantum leap in tecnology was achieved some years ago with the initial integration of computer interface and machine. The latest piece of equipment, will have little advantage in productivity and features, over its 5yr old predecessor. In fact it may simply have a new software interface which involves lost productivity to master. You might liken the improvements in office productivity from DOS to Windows/Office3.1 compared to Windows/Office98 and Windows/Office XP today. Computer/CAD controlled equipment as well as logistics and accounting have a similar history. Diminishing technological returns is quite apparent, with the reluctance of industry to wear the relatively high training upgrade costs of newer equipment.

    This sort of qualitative analysis may explain the declining relative proportion of investmennt expenditure. As well this may be coupled with a demographic exogenous shock that causes interest rates to rise unexpectedly. The rapid new household formation by boomers in the early 70s, may have produced a sudden investment demand for capital expenditure, to facilitate demand. We may be experiencing the reverse of this now, with mature demand ushering in a long period of historically low interest rates.

Comments are closed.