Keith Rankin's Thursday Column
Business Confidence and Goodhart's Law
6 July 2000
Last week's National Bank Survey of Business Confidence revealed the biggest loss of confidence since 1988. Yet, unlike 1988, there is no obvious reason for the apparent pessimism. It's no good blaming government policies such as slightly higher top tax rates and slightly less laissez-faire in the labour market. Australia, with its higher taxes and more unionised workforce, has had a more ebullient business sector for most of the last decade.
Is the problem really wimpish business sector? Maybe we are not reading the survey correctly. I can think of four reasons why this may be so.
First, the number of firms questioned who expected their own activity to rise was nearly as great as the number expecting their market conditions to deteriorate. The main problem is that at most 22% of respondents expect other firms' outlooks ("general confidence") to improve. We read to much into businesspersons' second guesses of other businesspersons' expectations.
Second, surveys of this type do not measure business pessimism or optimism in absolute terms. Rather, if taken at face value, they indicate changes in business sentiment. Thus a fall in confidence may simply mean that activity had been so high that few firms could conceive of any increases in their rates of investment; hence planned falls in investment would outweigh planned rises. One is much more likely to fall from the crest of a wave than from the trough.
Third, business confidence survey's are usually taken to represent the business view of the national government's performance. In fact, business confidence could be low despite rather than because of government. A problem of confidence could lie with the policies and practices of another agency (eg the Reserve Bank) or in some aspect of the global economy that no national policymaking agency has any control over.
Fourth, these surveys have become very political, and the survey participants know that. The surveys have become means by which businesspersons communicate their policy preferences, to the Reserve Bank and to the government. More than ever before, the survey question on "general confidence" is, I would argue, being answered tactically rather than truthfully.
The real story in mid-2000 is that businesses are conditionally confident. They will raise their investment and employment plans if they can be reassured that the Official Cash Rate of interest (OCR) will not be raised further this year. While the economists at the National Bank have not claimed, as I do, that their respondents are tactically overstating their pessimism, the National Bank itself is unequivocal. "There is no excess domestic demand in the economy ... we will repeat last month's message to the RB: DO NOTHING."
The tactic worked. Yesterday the Reserve Bank did not raise the OCR, suggesting a strategic rethink by the Reserve Bank, which had previously given the impression that the OCR would be subjected to incremental rises throughout 2000.
If I am right, attempts to manipulate business confidence represent an interesting extension of Goodhart's Law.
In a revealing 1992 BBC documentary ("A Fable from the Age of Science"), Professor Charles Goodhart states: "Goodhart's Law said that if ever the Government decides to rely on any particular statistical relationship as a basis for policy, then, as soon as it did that, that relationship would fall apart. And that's just what happened." Goodhart served as Chief Economic Adviser on Monetary Policy for the Bank of England from 1980 to 1985.
Goodhart had in mind, in particular, the relationship between monetary aggregates (eg M1 or M3) and prices (eg the CPI or CPIX). A Friedmanite monetarist, in the 1970s he believed that there existed a straightforward causal relationship from money supply to prices. Prices, he believed, were a simple function of money. The monetarist interpretation of the so-called Quantity Theory was supported in his and many others' minds by historical data.
But when he and others advised Margaret Thatcher's British government to constrict the growth of the money supply as the correct means to bring down inflation, the policy was a conspicuous failure. "The economy did not behave in the way that monetarists predicted. The squeeze on money led to a wave of factory closures while inflation continued to rise. ... Even more mystifying was the behaviour of the money supply. Despite the squeeze it was still growing, something the monetarists thought impossible." (New Zealanders can relate to this. When the "squeeze" was on in the mid-1990s - and before that in the mid-1980s - money flowed conspicuously, at least in Queen Street and Lambton Quay.)
Goodhart generalised his experience to create an economists' cross between Murphy's Law and Schroedinger's Uncertainty Principle. The moment a causal variable is controlled by a policymaker, then the causal relationship disappears. The understanding had been that rising monetary aggregates caused increased inflation and that rising interest rates caused disinflation. But, Goodhart said, once someone like Don Brash tried to control the quantity or the price of money, then those cause-effect relationships failed. Catch 23. The political implications were profound. No aspect of a nation's economy could be managed by an authority figure pulling policy levers this way or that.
Business confidence came to be seen as a critical causal variable in the 1990s. The National Government thought that the expectations of businesspersons were self-fulfilling. So policymakers tried to manage the economy in a way that would induce business confidence; eg by reducing uncertainty, taxes, real wages and benefits. In doing so, they changed the (always tenuous) cause-effect relationship between "net confidence in general business conditions" and subsequent "actual business conditions". Reducing uncertainty, taxes, real wages and benefits did not create an investment nirvana in Aotearoa.
We now have a perverse yet more democratic relationship between confidence and subsequent reality. Since business confidence came to be politically manipulated, business confidence surveys have become self-refuting rather than self-fulfilling. Businesspersons, many of whom are confident with respect to their own businesses, deny that they are confident about the general state of the economy. As a result of these tactical responses, the demon of gratuitous monetary policy is tamed. Businesses invest and employ; the economy grows.
© 2000 Keith Rankin
Thursday Column Archive (2000): http://pl.net/~keithr/thursday2000.html