The supposed imperative to deploy Keynesian policies to deal with current unemployment rests on the proposition that the market economy is not ‘self-correcting’; market actors do not (or ‘cannot’) change prices in the face of market disequilibrium so as to eliminate the disequilibrium.
What case can be given for this contention?
Keynes struggled over 403 pages in his General Theory to provide a cogent one.
The consensus of commentators is that best case that can be made from Keynes’ ideas for the lack of ‘self- correction’ lies in a phenomenon known as the ‘Liquidity Trap’. Keynes himself only refers passingly to the phenomenon. Indeed, the phrase ‘liquidity trap’ never appears in the General Theory. (It first appears only in 1940, in the Essays in Monetary Theory of his onetime collaborator-turned-critic, D.H. Robertson). For all that, the alleged phenomenon of the ‘Liquidity Trap’ came to assume in the post- War world a huge theoretical burden in justifying Keynesian Revolution. Remarkably, since in 2007 its galvanised corpse has been hastily twitched in the hope of it playing similar role in today’s second Keynesian revolution.
But what is the Liquidity Trap all about? It is not something that can be unpacked in one line, or a single paragraph. For all that, the quarry can be brought to ground, without mystery or evasion. And it is right to do so; for it is right the take the measure of this portentous conception.
The starting point of the Liquidity Trap scenario is that any self-correcting potential of the market economy will rest on the competition for jobs by workers in a situation of unemployment. This competition will reduce money wage, and so money prices; and all money holders will have larger money balances in real terms than they did before. People, however, need only a certain amount of money in real terms to execute the transactions they wish. The upshot is that competition for jobs will not increase employment in the first instance, but will certainly leave everyone with real money balances in excess of their needs. Therefore, it must be in the subsequent equilibration of the money market that any ‘self-correction’ is engendered.
So we are led to ask: how is equilibrium in the ‘money market is restored’?
Keynes provided one answer. It rests on the notion that money is held not only for the sake of executing transactions, but also for what he called the ‘speculative motive’; that is, the wish to optimally manage one’s wealth. But why would anyone hold money for the sake of managing one’s wealth, given that money’s nominal return is always zero? The answer lies in the fact that to be always zero is never to be negative. And there rests the idiosyncratic appeal of money; its zero rate may be low, but it will exceed the negative rate that bonds will pay whenever their prices fall sufficiently to outweigh their coupon. Thus, while those who expect bond prices to rise (call them ‘bulls’) will certainly not hold money out of a speculative motive, those who expect bond prices to fall (‘bears’) will (to ignore the coupon) willingly put all their wealth in the form of money, and hold no bonds at all.
It was Keynes' view that excess supply of money created by the competition for jobs will set in motion a trade between bulls and bears which will restore money market equilibrium. For while bears welcome their increased real money holdings resulting from falling wages, bulls, by contrast, decry their increased real money holdings as a sub-performing investment. But the bulls ‘excess supply of money’ can be corrected by a certain rise in bond prices. For any rise in bond prices will makes bears of some former bulls – since to some former (and not so bullish) bulls the rise in bond prices will entail –as a matter of arithmetic - that the level of bond prices this period will now stand higher than the level they expected for next period. And this engenders a demand for money in such new-made bears. And not only do they now willingly hold on to their increased supply of money; they also now wish to sell what bonds they have in exchange for money. And the bonds for sale of these new-made bears will be gladly bought by the remaining ‘hold out’ bulls with their own unwanted real balances.
It is easy to see that there will be some rise in bond prices such that a sufficient number of bulls are ‘converted’ into bears so that the initial excess supply of real balances is now absorbed by a expanded ‘speculative demand’ of bears, both old and new-made.
Thus the money market is restored. But there is another consequence. The higher bond price amounts to a lower yield on bonds. Consequently investment in plant and equipment will increase, so as to reduce the return on capital to the newly reduced yield on bonds. And that increase in investment will increase GDP and employment.
Thus in the end the fall in money wages has reduced unemployment. Self-correction!
If there still remains unemployment after the assumed wage cut then the process can be repeated. Clearly as this process continues, speculative money holdings will swell ever more across an ever increasing population of bears, and bonds will be increasingly concentrated in dwindling population of hold-out, ‘never say die’, bulls.
But what if there is, say, just one bull left? Consider this solitary bull, now holding all bonds. If at the existing price of bonds investment (in plant and equipment) is large enough to secure full-employment; then well and good. But what if there is still unemployment? What will a further round of money wage reductions do the trick? No. Everyone will have increased real balances, as before, and the bulls (that is, the bull) will not want to hold that increase, as before. But the difference is that the last remaining bull cannot now be converted into a bear; because for him to become a bear someone would have to buy his bonds he would now not want. And there no one left to do that! So what happens? The price of bonds rises - but not until the bull has become a bear; rather , it rises only so far as to make the bull indifferent between bonds and money. And that restores equilibrium. The bears hold more money than ever, but that is OK by them; and the solitary bull by becoming ‘a deer’ – and believing the bond price will neither rise nor fall - is indifferent between bonds and money, which is to say they have no objection to the extra money they hold.
But the key point is still to be made: that bond price at which the last bull becomes a deer constitutes a ceiling to the bond price. The bond price cannot rise any further. If wages fall again, then real balances will rise gain; but that creates no excess supply of money. For the bears are glad to hold the increased real balances, and the deer at the previously secured bond price is content to hold the increased real balances. The bond price has hit a ceiling; and the bond ‘yield’ has struck a floor, and the long rate of interest can go no lower. And if that long rate is such that investment is insufficient to employ the whole work force ... then unemployment will remain no matter how much money wages are reduced.
Such is the liquidity trap story.
What to make of it?
The story evidently is of some intricacy, but is intelligible for all that.
Its weakness is that is that it turns critically on two assumptions have no purchase on our assent. First, that bond holders are certain of their forecasts; and, two, that expected inflation remains zero throughout any deflationary adjustment.
That bond holders are supposed certain of their forecasts is essential to the feat by which a zero rate on money entails a positive floor on the long rate of interest. For it is existing bond holders all agreeing with perfect confidence what the price of bonds will be next period that makes money and bonds perfect substitutes at some critical current bond price; it is perfect confidence that makes wealth holders content to hold any amount of money at that critical bond price. But as soon as we suppose wealth holders allow for range of possibilities for the bond price next period, then money and bonds will never perfect substitutes, and the only way to persuade wealth holders to hold more money is to increase the bond price. (This point was made clearly by James Tobin in ‘Liquidity Preference Towards Risk’ in 1958 in the wake of portfolio choice theory being transformed by the theory of choice under uncertainty).
That expected inflation is supposed to remain zero throughout any deflationary episode is also essential to any liquidity trap story. The nominal rate of return of investment consists of a real component (determined by technology) and a nominal component, amounting to expected inflation. Wage and price reductions cannot affect the real component, and by Keynesian assumption cannot affect expected inflation. But is this assumption valid? Is it not possible that a reduction in current prices will not drag down expected prices by the exactly same proportion? If this is possible then a reduction in current prices will leave expected prices somewhat above current prices, and thereby create an expectation of positive inflation. And that amounts to an increase the nominal yield on capital. In other words, a cut in money wages in the current period can directly shift up the prospective nominal rate of return on investment, and so increase investment for an unchanged nominal interest rate, and so restore the equilibrating effect of a competition for jobs.
The liquidity trap, then, is not only an intricate structure, it is a fragile one.
Or is flimsy the better word?