Why are interest rates so low (II)12906
I take up the second part of Ben Bernanke's analysis of why interest rates are so low. He rejects the "secular stagnation" hypothesis advanced, among other, by Larry Summers (see here for more contributions and here for Summers' counterpoint to Bernanke's post). Like Bernanke I do not find the secular stagnation hypothesis convincing at all. But I also find his analysis far from satisfactory, even incoherent sometimes. I do not see why "facts" that are not facts are being accepted as the starting point of analysis, and why the latter has to turn immediately into questions of policy framed in the most narrow format. If we are not even sure of what is going on, what is the point of debating policies?
Larry Summers has proposed "secular stagnation" as his explanation for the low rates we observe on a wide array of debt instruments. According to his definition
The essence of secular stagnation is a chronic excess of saving over investment. The natural question for an economist to ask is how can such a chronic excess exist in flexible markets? In particular, shouldn’t interest rates adjust to equate saving and investment at full employment? The most obvious answer is that short term interest rates can’t fall below zero (or some bound close to zero) and this inhibits full adjustment.
Because my intention was that of commenting on Bernanke's argument, I will comment first on his post and then take up a few issues in Summers' argument I do not find convincing and Ben did not discuss. Should I find the time I will get back next week to the more structured arguments advanced in the Vox-EU volume linked above in support of the secular stagnation argument.
Discussion of Bernanke's post
As said, Bernanke's post is disappointing. He accepts as "facts" statements that are not supported by data or are, literally, untestable (see below) and then makes a number of statements I fail to understand.
- How downward rigid must prices be for an "excess supply" to last "secularly"? Everything we know about price stickiness is measured in quarters, a year at the very most. In fact, for a number of years in the USA, the EU and Japan, the policy talk has all been about "deflation", meaning that prices are dropping. In fact, they are while unemployment, in the US at least but not only, also keeps dropping and is now at a relatively low level. Employment is also growing even if, and this seems the one worrisome issue to focus upon on the employment side, the employment to population ratio has recovered little and is at its lowest levels in thirty plus years. Notice, first, that it is well above the level of the decades 1950-1980, which were, according to this discourse, decades of high growth. Notice, second, that the picture changes substantially once we look separately at the two key age groups: 16-19 and 25-54. This is not the place to get into the demographic and socioeconomic details of this issue but, it seems to me, there is enough evidence to suspect it has little to do with some secular stagnation of demand and a lot to do with demographic and behavioral change.
- Bernanke never asks what the theoretical and empirical meaning of "excess saving", on which the secular hypothesis rests, is supposed to be. I will come back to this later while discussing Summers, but I should point out that either income produced at time t is stored in the form of currency and is not used to purchase anything (consumption or investment goods, it makes no difference) at time t+1 (and this continues for many periods) or the fact that people "save" and do not "consume" simply changes the composition of final demand but does not necessarily lowers it.
- Unless - and here Bernanke misses the opportunity of showing a logical contradiction in Summers' argument - such income is spent to purchase already existing assets, which are being passed around from one investor to another in a closed circle that only increases their prices without generating any demand for new output. This infernal cycle of "tresaurization" must, on the one hand, create a bubble (thereby contradicting Summers' claim that we need bubbles to sustain aggregate demand: in fact bubbles suck away income from final demand) and, on the other, begs the question of what the hell the initial owners of the bubbling assets are doing with the capital gains they obtained by selling the asset in the first place. Start another bubble? In other words, is secular stagnation equivalent to "bubbles all the way down"?
- By accepting the idea that the whole problem can be reduced to a "lack of demand" and that the only way to increase demand is to lower interest rates (!) Bernanke gets himself tangled up again in the discussion of the mythical Wicksellian Rate. He comes up with an incoherent argument (which he attributes to Paul Samuelson: I did not have time to check it, but Samuelson has made so many jokes that it may be well true he said this nonsense as well) according to which when the real interest rate is negative (or even zero) it is profitable to level the Rocky Mountains in order to save the fuel we consume driving through their passes. Is he kidding or what? Incoherent example aside, the whole argument is neither here nor there: on the one hand, when people are impatient enough, they will consume and not invest if the rates or return offered are negative, while, on the other, the WR are measured by the rate of return on T-bond can be as negative as you like and still there may exist profitable investment opportunities that are not exploited either because they are risky or because the credit system fails to "complete markets". He does recognize this fact saying that he "concede(s)", while there is nothing to concede here to the secular stagnation argument: if the issue is with the credit system and with the lack of instruments/markets to appropriately cover the risk of profitable investments, "secular stagnation" is just the wrong way to look at it!
Discussion of Summers' secular stagnation
First, a brief note about the "model". What on earth is secular stagnation? Summers cites Alvin Hansen, but the latter had no coherent argument (let alone a model) and, in fact, his prediction was complete nonsense, as the facts have proved. So, what does it mean to say that we have too much saving and too little demand? Like it or not, bot national income accounting and plain economics tell us the following: EITHER you take your money income and hide under the mattress or you purchase some security that finances some kind of investment somewhere, thereby generating demand for some factors of production of some kind somewhere else. The composition of demand changes, relative prices have to adjust, some firms have to shut down and other have to be created, workers need to change jobs BUT total demand needs not decrease at all!
The only other theoretical alternative is to use the income of period t (and of many subsequent periods, t+1, t+2, etcetera) to sustain a bubble in the price of some pre-existing asset. This amounts to the "tresaurization cycle" I described above in which B uses her income from period t to purchase the existing asset X - say: Renoir's painting - from A. Then C uses his income to purchase the same asset from B (at a higher price) and so on. The Renoir's painting changes hands in each period and incomes get passed from B to A, from C to B, from D to C and so on. Which begs the question: what on earth will A (and B, and C, and ...) do with the money they receive when they sell the asset (at a higher price, hence the bubble)? They can either start another bubble (and another, and another ...) thereby purchasing turtles at an increasing price all the way to the end or ... or hide the proceeding under the matress or ... spend them!
I stop here and wait for a less turtle-like model of the economy than this. In the meanwhile I should notice, though, that even this turtle-bubble "theory" contradicts Summers' claim according to which financial bubbles were the sources of growth in recent decades: the model predicts the opposite. If, in stead, he meant to say that (for the half dozen reasons we have been studying for years now) an ill-guided expansion of credit and ill-designed monetary and regulatory policies created incentives for investing in the wrong assets between 2001 and 2007 and, once those investments failed, we realized we were poorer than we thought and allocated resources to profitable activities, then we are in business and we can start talking. But, again, this leads us very very far from any secular stagnation theory!
The facts supporting Summers' view in BB's rendition:
For a number of reasons—including the contemporary decline in population growth, the reduced capital intensity of our leading industries (think Facebook versus steel-making), and the falling relative prices of capital goods—Larry sees Hansen’s prediction of limited investment in new capital goods and an economy that chronically fails to reach full employment as relevant today.
Population growth, in the US at least, does not seem to have particularly slowed down in recent years. Civilian labor force has grown less than before since the beginning of the Great Recession but is now picking up again and I find it a very long shot to decide that the trend has really changed in the long run (we are talking about "secular" stagnation, after all :)). I already discussed employment/unemployment above and showed that IF there is an issue it has nothing to do with demand but with demographic and social-behavioral changes that require adjustments in the structure of the economy's supply side and not just some generic "stimulus" to aggregate demand.
Further, the "new industries" are as capital intensive as the old ones, or even more more: only they use different kinds of capital (here and references therein for details). Finally, what the falling relative prices of investment goods (a fact that has been true since the 1950s!) would have to do with secular stagnation and reduction of demand for investment escapes me and I have no idea what kind of causation Summers has in mind. Elementary partial equilibrium logic would suggest the opposite, if anything and I see no general equilibrium argument or model suggesting that when, ceteris paribus, the price of investment relative to the price of consumption decreases one should see fewer and less profitable investments!
Another point - that Bernanke also takes up at the end but I believe deserve being stressed - has to do with international trade. Summers seems to be arguing as if the US economy - and almost all the European ones, as interest rates are even lower there and growth has been weaker in the EU than in the US for decades ... - were closed economies: they are not and some of them export more than 50% of their output. Internal demand is important, certainly, but 1/3 or 1/2 of your GDP is exported, then it is global demand that matters. When one takes into consideration the global landscape all the pseudo-facts Summers mentions become even less true than for the US: world population is growing and getting younger, employment is growing worldwide and in recent decades a few billions people have entered the market economy. Their real incomes are increasing and their demand for both consumption and investment goods is doing the same. In the whole history of humanity this century is as far from economic "stagnation" as any has ever been!
If these are all the "facts" available to support the claim of a secular stagnation, the claim is groundless. And, barring turtles all the way down, the theory does not even exist.