- Cafe Hayek - https://cafehayek.com -

Economic Science and Predictability

Tweet [1]

Russ’s recent posts – here [2] and here [3] – inspire me to contribute my two-cents worth of thoughts to his debate with Paul Krugman and Simon Wren-Lewis.  (Warning: Everything that I say here is likely to be something that I’ve said in at least one earlier post.)

While an improved ability to make accurate and falsifiable predictions about empirical reality is fine, this ability is not the only, or even necessarily the finest, contribution that science makes to humankind.  Science – especially any social science, such as economics – offers contributions also, and mostly instead, on other fronts, namely, an improved understanding of empirical reality.  In short, a scientific discipline can be fully worthy of the descriptor “scientific” even when it does not generate specific predictions that in practice are falsifiable.

All theory is story-telling.

By claiming that all theory is story-telling I of course do not mean the telling of tales meant solely to entertain.  Instead, I mean that every theory – in the “hard” as well as in the social sciences – is an account that we tell about reality in our attempts to make better sense of reality.  Each such theory – each story – is ultimately judged by how well in helps us to understand reality.  A story that helps you to understand some aspect of reality better than does any other available story about that same aspect of reality is, for you, a sound theory.

Because different people inevitably have different perceptions, preconceptions, perspectives, analytical capacities, and experiences (intellectual and practical) that they bring to the table when they sit down to consider this or that story about some aspect of reality, different people will react differently to different stories.  While Smith might, say, find the story of price formation featuring supply-and-demand analysis immediately and utterly convincing, Jones might not find it so at first – or ever.

Some stories are about relatively simple aspects of reality – where “relatively simple” means that the number of ‘moving parts’ in the reality about which the story is told is relatively small.  The fewer are the number of moving parts in the reality that we are trying to better understand, the more likely it is that any story that we tell about that reality will generate specific, falsifiable predictions.

Consider the theory of gravity (which is such a powerful and convincing theory that we call it the “law of gravity”).  This theory allows us, for example, to calculate with much precision the speed at which a billiard ball will be traveling when it hits the ground on earth after being dropped from any certain height.  If a highly accurate measuring device informs us that the ball, after being dropped from that height, was traveling at some different speed, we know the relatively few factors (‘moving parts’) that we can adjust for to see if the theory is correct.  We factor in, for example, air resistance as well as the possibility that the platform from which the ball was dropped moved upwards or downwards at the moment the ball was dropped.  The relative ease of controlling for – adjusting for – all of the factors other than gravity that do, or plausibly might, affect the fall of the billiard ball to earth makes possible specific, falsifiable predictions using the theory of gravity.  We can, in other words, be so confident in our ability to isolate in practice the effect of one force (gravity) by distinguishing it in practice from the effects of other forces (for example, air resistance) we can make specific, falsifiable predictions using the theory of gravity.

In contrast, when we tell stories about the economy in attempts to improve our understanding of it, we are telling stories about a phenomenon featuring in practice countless moving parts.  And to make matters even more difficult in practice, many of these moving parts – such as, for example, changes in consumers’ preferences for apples relative to pears – are inherently unobservable to the scientist.  Therefore, the acceptability or scientific integrity of any story we tell about the economy cannot reasonably ride exclusively upon that story’s ability to generate falsifiable predictions about specific events in reality.  “Other things” are real, ever-present, and indeed often changing.

Many people, thinking that science is only that which allows for specific, falsifiable predictions, conclude from this reality about economics that economics is not a true science.  I disagree.  While economics seldom, if ever, allows for specific, falsifiable predictions about the real-world economy (as opposed to in controlled laboratory settings [4]), economics does offer stories that, if carefully followed and understood, improve our understanding of economic reality.

Before giving an example of how economics improves our understanding of reality, I pause to note that biology – which, as far as I am aware, no scientifically informed person denies is a true science – does not supply specific, falsifiable predictions about reality.  Like economics, biology can make “if-then” statements – such as, if a meteor strikes the earth and causes average surface temperatures to rise significantly above what those temperatures would have been had the meteor not struck the earth, and if no significant offsetting changes occur, then many mammals with thick fur will become extinct.  But biology cannot predict in detail just what sorts of animals (or possibly plants!) will replace the species that die out: will they mostly be winged (or not) – green-eyed (or not) – over 30 pounds in weight as adults (or not); no one can possibly say because the biological world has too many moving parts in practice to enable any such specific, falsifiable prediction to be made.  Yet nevertheless the theory (law?) of natural selection is a story that gives us the overall best understanding of the formation of species.  That this theory doesn’t permit specific, falsifiable predictions in practice does nothing to diminish its power as a tool for use by our minds to make better sense of observed biological reality.

Now back to economics.  Consider the commonly observed real-world pattern of prices after natural disasters, such as earthquakes, strike.  Prices of staple goods (such as bottled water, fuel, handyman services) rise significantly and then eventually fall back to their pre-disaster levels.  Moreover, the levels to which such prices rise as well as the time it takes for the prices to return to their pre-disaster levels is positively correlated with the severity of the disaster.

What explains this reality?  Non-economists typically tell a story centered on “greed.”  The price increases are explained as resulting from sellers’ greed.  (By the way, almost never is the greed of buyers blamed, even though the willingness of some buyers to pay the higher prices in order to ‘greedily’ enhance their chances of getting the staple goods is necessary for ‘greedy’ sellers to charge higher prices.  But ignore this little fun fact.)  The economist, in contrast – and consistent with his or her role as myth buster [5] – points out that “greed” is a poor theory, if for no reason other than the fact that there’s no reason to believe that greed is enhanced by natural disasters.  Something else almost surely must be in play.

That ‘something else’ (says the economist) is the combined forces of supply and demand.  Natural disasters simultaneously increase people’s willingness to acquire staple goods (increase the demand for staple goods) and decrease people’s willingness (or ability) to offer such goods for sale to others (decrease the supply of staple goods).  With people, because of the devastation unleashed by the natural disaster, needing more potable water and fuel and plywood and the services of home-repair experts – and with, at the same time, inventories of these things destroyed along with the destruction of supply lines (e.g., impassable roads) – each available unit of these things becomes more valuable to consumers in the devastated region.  The resulting higher demands and lower supplies push prices upwards.  These higher prices, therefore, reflect the underlying reality that a natural disaster struck.  (These higher prices, of course, also [1] cause consumers to use the now-scarcer staple goods more sparingly, and [2] give incentives to suppliers to bring  more of the goods and services into the affected region as soon as possible.)

As time passes and people repair their homes and business and as supplies and supply lines are restored, prices fall back down.

Economists’ story of such price changes is theoretically simple, yet it is about a very detailed and complex reality.  And although this story allows no detailed predictions, it remains marvelously useful for our understanding of the complex reality that we observe.  Not even the most ingenious economist could possibly accurately predict just how high the price of plywood will rise in New Orleans after a category five hurricane strikes that city.  But economics – the story of supply and demand – does permit the conditional prediction of the pattern [6] that price movements will take after a natural disaster strikes: rise and then fall.

I say “conditional prediction of the pattern” because, even when speaking only of a pattern, it’s possible that reality in an particular instance will offer evidence against the pattern prediction.  Hurricane Zeke might strike Miami without there being any consequent price increases.  Yet because the supply and demand story of prices is so compelling generally, and so useful in most cases in enhancing our understanding of observed reality, this one instance will be insufficient to cause economists to reject the supply and demand story.  Perhaps Zeke happened to destroy only very few inventories of any staple goods and few supply lines while, at the same time, an unusually large number of people out of a spirit of philanthropy contributed their time and resources to bring such a large amount of supplies of staple goods into the Miami area that all additional demands for staple goods were satisfied by these charitable contributions.  The “if-then” story remains solid: if a natural disaster destroys supplies and increases demands, and if there are no off-setting changes, then the prices of staple goods will rise in the immediate aftermath of a natural disaster.

The good and wise economist understands that understanding is not only important in its own right, it is about the best that economics can achieve in the real world.  Such understanding allows “if-then” statements – which, among their many uses, can serve to improve public policy (e.g., “If government prohibits so-called ‘price-gouging,’ then people in a natural-disaster area are likely to get fewer of the staple goods that they need.”)  But such understanding does not allow economists to make about economic reality specific, falsifiable predictions of the sort that physicists often make about physical reality.  This inability is no cause to discount or to disregard the importance of the science of economics.

Comments