by Mario Rizzo
In a recent talk on Keynes, Roy Weintraub refers to an idea he finds in Keynes that can be characterized as the confidence multiplier. Spending in particular areas builds confidence in other areas because businesses in the latter can rely on complementary spending in the former. Thereby a generalized resumption in spending can take place, perhaps lifting the economy out of a slump.
This idea many Keynesians have promoted is seriously flawed. Under appropriate conditions, spending can beget spending. But so what?
Some Keynesians will say that the spending will lift resources out of idleness and almost any use of these resources will be of greater value than idleness.
Today’s idleness is waiting for the sky to clear, that is, it is the consequence of trying to figure out, amid the process of liquidation of resources in over-expanded sectors, the appropriate areas for new investment. The true value of this idleness is imputed from the value of the improved resource allocation that it permits.
It would be great, of course, if (1) it took no time to re-allocate resources, or (2) it did not matter where resources were employed as long as they are employed. But each of these statements is clearly untrue.
No doubt there is a kind of “you first” phenomenon at work. Economic activities are interdependent. Therefore, when one agent tries to determine where to allocate resources he must determine where others will want to allocate their resources. The greater the uncertainty with respect to this the more disadvantages will be associated with being the first mover. On the other hand, successful first-movers will gain greater profits. This is a coordination problem.
Is stimulating general “confidence” likely to alleviate this coordination problem?
Suppose government through fiscal spending policy tries to stimulate aggregate spending. It cannot do this in a neutral way. It will have to spend in particular directions. And then those who receive the initial money will spend in particular ways. Additionally, the government will have to stay-the-course with clearly announced specific spending over a period of time that is likely to correspond to the planning horizons of economic agents. At low interest rates, this is likely to be a fairly long horizon.
Keynes himself recognized the problem. Accordingly, he advocated consistent and persistent public direction of the major part of investment. The idea is that confidence would grow out of the certainty created when agents know the level and direction of investment spending and can count on it for extended period of time.
We do not live in the world that Keynes wanted both because government cannot be counted upon to pursue consistent long-term policies and because Keynes did not accept that resources were systematically misallocated during the boom.
In our world economic agent must deal with several uncertainties at the same time.
1. How will the political system allocate stimulus resources in the concrete and for what length of time?
2. In which directions will those who receive income, subsequent to the government-round of spending, choose to spend?
3. What is the sustainable pattern of spending, saving and investment that will emerge as the government stimulus declines?
Investors do not invest in the abstract or in the aggregate; they invest in specific areas. Stimulus, as it is actually practiced, compounds the coordination difficulties that agents face. They must now predict the behavior of political agents who do not respond to conditions of supply and demand in the market. What will Tim Geithner, for example, come up with next? What new conditions will Obama or Congress set for the carrying out of business. All of this gets mixed in with attempt to discover new market equilibria relative to sustainable preferences. The informational quality of prices tends to deteriorate. (Roger Koppl has important things to say on the subject of such “big players” here.)
Simply generating confidence that government will be spending on something for some indefinite period of time does not correct the underlying problem. There is a double coordination problem: spending must be coordinated with other spending, and it must be coordinated with the sustainable preferences of consumers and savers.
There are three types of confidence relevant here. The first (“regime confidence”) is created by stable rules or policies. Agents know that government will not introduce additional uncertainties or noise (relative to the underlying economic data). Thus prices will not needlessly mislead. Lack of confidence of this kind can be generated by unpredictable shifts in economic policy and by stimulation of unsustainable lines of spending and production. The second type of confidence (“monetary confidence”) is an indirect effect of satisfying an increase in the demand to hold money (by the monetary authorities or the banking system increasing the supply of money). This would prevent prices from falling by an uncertain amount. When entrepreneurs fear deflation they are likely to withhold resources, that is, they are likely not to commit them to specific market uses since they are unsure of the future course of prices.
The third type of confidence (“resource allocation confidence”) is the effect of entrepreneurs discovering, in the context of the two forms of confidence mentioned above, the improved – more profitable – lines of production in the wake of the prior allocational distortions. It is hard to see how the state can directly help entrepreneurs to solve this problem. The market is a discovery process. Specific markets require entrepreneurs with specific, detailed and transient factual knowledge who can act promptly. This is not any state I know about.
So now we can answer our question: Can stimulus spending, by building general confidence, alleviate the coordination problems agents face?
Stimulus spending will not increase, but reduce, regime confidence. Policy uncertainties will increase. The monetary confidence will not be affected by fiscal policy. It also seems that outright deflation of a persistent sort is not a real threat at this time. And resource allocation confidence is not government’s forte.
We can see what is wrong with the Keynesian confidence multiplier. It multiplies error; it multiplies the obstacles in the way of market corrections, and it multiplies the time it will take for recovery to ensue.
Once we go behind the veil of macro-aggregation, we can see the complexities of the real-world problem and the utterly simplistic nature of policies based on stimulus and confidence multiplication that have been proposed as the cure for recession.
(Note: John Maynard Keynes himself had a good deal to say about “confidence” especially as this idea was used by U.K. Treasury Officials. This is a topic for another day.)