When is enough good enough?

By Rick Ashburn

Given the complex relationships among the vast number of variables involved in any economic system, students are taught to examine economic questions in a very limited way. There is a common – and dangerous – phrase used in economic analysis. The Latin term “ceteris paribus” means “with other things equal.” An analyst will say, “holding all else constant…” The method of examining the relationship between two variables, while ignoring and ruling out any other variables, allows us to understand economic concepts.

Ideally, we would examine the entire economic environment and make comprehensive, all-encompassing conclusions about every little piece. This would be known as a general equilibrium solution; general in the sense that we have covered everything there is to cover. While it is the purist’s objective, a general equilibrium analysis is fiendishly difficult to achieve. And, so, we tackle just little bits of the puzzle at a time. We usually use the shorthand of only considering the interaction of two variables at a time. This is known as the partial equilibrium approach.

Yet this shorthand places us on a slippery slope. All else is never equal.

To illustrate the ceteris paribus concept, we can examine the effect of price changes. Let’s say we want to figure out the effect of gasoline price increases on a household’s consumption of gasoline. It should be obvious that, as prices rise, the household will have to spend more for a given amount of gas. This is called the price effect: Higher prices push up our spending on gas. On the other hand, another effect is pushing us to figure out ways to avoid using so much gas. This would be known as the income effect or the substitution effect since we will shift our spending to other things. On the one hand, we have upward pressure on our gasoline spending. On the other hand, we have downward pressure on our gasoline spending.

The partial equilibrium approach to the problem would involve only considering, say, the price effect. We can say with some certainty that, ceteris paribus, a household will spend more on gasoline as its price increases. Yet, all else is not the same – the household might decide to stop driving so much. Only a comprehensive general equilibrium analysis can get at the answers we seek. As you might surmise, the general answer is not so clear-cut. Will the family spend more, or less, on gasoline? We really can’t tell without doing a whole lot more work and analysis.

The shorthand of using partial analysis carries over, and dangerously so, to the stock and bond markets. A lot of stock market analysis these days turns on the effect that interest rate changes will have on stock prices. If we ignore the larger world, we can reliably conclude that higher interest rates are bad for stocks. This is the ceteris paribus point of view. If everything else in the world were frozen right where it is, higher interest rates would make stocks less attractive.

However, we cannot ignore the ever-changing larger world. We have to look more generally at the situation. In the case of rising interest rates, we need to consider what might be causing rates to rise. If rates are rising because the lending environment is tightening, and foreign capital is leaving American markets, then rising rates are probably bad for stocks. If rates are rising because the economy is booming, and investors have high forecasts for corporate earnings, then rising rates are clearly good for stocks. I have been reading a lot of punditry these days that concludes that, should the Fed begin to raise rates, stocks will fall. This point of view fails to consider just what sort of circumstances would convince the Fed to raise rates.

The daily short-hand analysis you read about stock market activity is of the partial equilibrium variety. Pundits and commentators have about 15 seconds to tell you what happened, so they pick a single variable that supports the day’s price movement. They might as well be ascribing closing prices to the phases of the moon. When I read statements along the lines of, “Prices rose today as investors shrugged off fears of higher inflation,” I know they’re just guessing. Absent the time and analytic skill to make sense of a complex situation, analysts just pick one thing. They rule out any other potential variables so as to make their own simplistic analysis appear more robust.

As one economist put it, “If we don’t make these assumptions, we can’t do anything at all.” That is actually good advice. Either do it right, or admit right up front the severe limitations of the analysis.