Much of our prosperity stems from our ability to identify important relationships from masses of data. This is as true in economics as in the sciences. Generally, the more information, the easier it is to determine important relationships. But data can be treacherous if underlying cycles or trends are hidden. This is as true for information arising from an economy, like interest rates or inflation, as it is for data from nature like hailstorms or droughts.
Weather prediction gives many examples of the dangers of assuming that what is true over some limited period will be true forever.
The 1980s and 1990s were relatively calm in terms of severe hurricanes. During the 1950s and 1960s, destructive storms were more common. Now, we are back in a severe storm phase of a long-term cycle.
Unfortunately, a great deal of coastal development took place since 1990. The relatively quiet 1980s induced complacency. People built homes and businesses in areas that, over the long term, were very susceptible to storm damage.
Hurricane patterns are an example of a cycle that was not recognized for a long time. The cycle itself does not necessarily point to a long-term trend. But some scientists do predict more extreme weather from human-induced climate change. In either case, limited information could result in bad decisions.
We have temperature and rainfall data for Minnesota for about 150 years. However, information on hail and damaging winds is more sparse.
All sorts of insurance, including that for wind damage to buildings and hail damage to crops, is based on historical averages. But such averages may have been calculated over a span of years that is not a good predictor for the longer run.
Human actions can change causal relationships, even those based on natural events. Maps showing flood plains are used for zoning, permitting housing developments and insurance decisions. They are based on stream-flow records that may go back decades.
But changes in land-use patterns in the watershed alter how much flooding results from a given rainfall. Paving farmland for suburban malls increases the speed and volume of runoff. Cropland also accelerates runoff compared with prairie or forests. The Conservation Reserve Program and reduced-tillage farming can reverse these changes somewhat. But it may be dangerous to base important investment decisions on flood data from the past.
Weather is not the only area where naive reliance on apparent relationships can cause problems. Over the last 60 years, as economists applied John Maynard Keynes’ ideas to industrialized countries, some saw clear links between output, employment and prices. Okun’s Law described a relationship between changes in gross domestic product and employment. The Phillips Curve asserted a clear tradeoff between inflation and unemployment. Milton Friedman saw a direct link between the money supply and price levels, even in the short run.
But as time went on, none of these relationships proved as reliable as first thought. GDP-employment links depended in part on how sticky labor markets are. What was true when a significant part of U.S. industry was unionized and international trade was small was no longer true when unions shrank and trade grew.
It turned out that if one tried to use Phillips’ inflation-unemployment tradeoff as a guide to minimizing both, you instead increased both over time.
The 1990s showed that with increased trade, unemployment could be lower – without inflation rising – than many economists had thought. Foreign competition kept domestic producers from raising prices the way they easily could have in the 1960s.
Friedman’s theory rested on the “velocity of money” – how fast money changes hands – being stable and predictable. But over time, it became clear that velocity varies.
For instance, innovation in banking and payment systems produced more variation. Negotiable Order for Withdrawal accounts, essentially mutual funds one could write large checks on, made money turn over faster. So did debit cards and the rapid growth in credit eagerly thrust at consumers by credit-card companies.
Once the dust from the current financial brouhaha ends, we probably will conclude that the Federal Reserve was lulled into complacency by easier international flows of capital. Increases in the money supply after Asian financial problems in 1997, the collapse of Long Term Capital Management in 1998 and the attacks of Sept. 11, 2001, did not raise the inflationary pressures that equal increases would have in the past. Slowing of money growth over the past two years has not pushed up interest rates as it would have in the past.
Easy imports suppressed the kind of consumer inflation that fast money growth would have caused 30 years ago. Enormous flows of capital from East Asia suppressed the interest rate increases Fed tightening usually caused. Decisions that seemed sound in the light of historical data turned out bad because underlying relationships have changed.
© 2007 Edward Lotterman
Chanarambie Consulting, Inc.