Association, Causation, Expectations and Caveat Civis

Economics focuses on how changes in incentives faced by self-serving (not necessarily selfish, but willing to command more resources to promote what matters to them) lead them to change their behavior in an almost unlimited range of ways, reflecting the fact that the range of choices in which scarcity is a factor is also huge. To understand such a choice, it is necessary to “think soberly”, so many logical errors that can lead our reasoning astray should be avoided. Principles of Economics text I currently use such errors as part of the very first chapter.

One such trap is to equate association with causation. As my text says: “In economics, it is very important to identify causal relationships. But statistical association alone cannot establish this causal relationship.”

In my class, I sometimes use a silly example to illustrate the problems that interpreting an association or correlation as implying causation can cause. Let’s say ice cream sales are positively related (or correlated) with property crime rates. This does not mean that one of them caused the other. Higher ice cream sales may have led to an increase in property crime. An increase in property crime may have led to an increase in ice cream sales. Some other variable or set of variables may have caused both the increase in ice cream sales and the increase in property crime. Or it could have been a random outcome (the main problem was the likelihood that we could reject randomness).

Students quickly see that none of the lines of causation between ice cream sales and property crime seem plausible (although there are a surprising number of cases in which important underlying incentives are not obvious on the surface, so this standard is not definitive) . Then it won’t take long for anyone to assume that summer is the main cause of both. Warmer weather is expected to boost ice cream sales, and summer means no school hours and more time “outdoors”, providing more opportunities for property crime.

Recognizing this allows me to draw a key policy-related conclusion that I want students to understand from this section. I’m just asking, “How effective would our policy decisions be if we believed that there was a causal relationship between ice cream sales and property crime in either direction?” In both cases, this can lead to very inefficient policies. If I had restricted the sale of ice cream because I thought it would reduce property crime, I would have wasted a lot of resources and not achieved anything I wanted to. If I were to limit the use of property crime laws as a means of increasing ice cream sales, that would be just as ineffective. And any other misattribution of causality can have the same adverse consequences.

Moreover, in our complex world, where we often move from “ceteris paribus” assumptions that make it easier to study the mechanisms of economic relations separately, to the need to weigh the numerous and often conflicting incentive stories that are relevant to a particular situation, the number such opportunities are very large.

This is one of the great advantages of market systems in a complex world. Anyone who thinks there is a causal relationship between two variables and that they could profit from that relationship can test this belief in the market and the process will show what works best. But public institutions tend to be monopolies that are not subject to market testing for profitability and do not face potential bankruptcy (except in a moral sense). This opens up much greater opportunities for the implementation of public policy in the absence of a reliable understanding of the underlying causal mechanisms. Thomas Sowell characterized difference as “replacing what worked with what sounded good”, as evidenced by the fact that in area after area – crime, education, housing, race relations – the situation worsened after brilliant new theories were put into action . The amazing thing is that this story of failure and disaster did not discourage or discredit the social engineers.”

This potential confusion, often at very high stakes, is one of the many reasons why good intentions often lead to inefficient results, creating many problems. Pathways to policy failure. It also requires us to analyze more carefully the cause-and-effect relationships if we are to implement more effective policies.

One obvious approach is to ask if one variable changed before another. But must we conclude that if one variable (A) changed first, and then another variable (B) changed, then the first caused the second? No. This is such a well-known misconception that it has a name – Post Hoc Ergo Propter Hoc delusion (often shortened to after the fact delusion).

While we cannot logically conclude that what happened first caused what happened next, that does not rule out that A could have caused B. It is still possible. In fact, such a correlation is often the subject of research aimed at finding a causal mechanism (or mechanisms) that could explain it.

But since time cannot be reversed, this result rules out that B is the cause of A, which can significantly narrow the “list of suspects” of causation. One description Isaac Newton’s approach to physics illustrates this. “Newton thought about cause and effect as sequential… Since movement occurs in time, cause and effect must be ordered in time. The effect can happen before the cause only in fantastic stories about time machines, that is, in reality (as far as we know) cannot happen. In other words, what happens after cannot be the cause of what happened before.

Unfortunately, what may be true in physics is not necessarily true in humans. There is an important way in which time can effectively move backwards by changing people’s behavior.

This is because people often change their behavior when they first begin to expect something to happen, which may be substantially before rather than after it happens, as shown in the downward movement of stock prices in the 1930s. , which occurred as the likelihood that protectionist smoothing Hawley’s tariff bill will be increased.

Other examples include expectations of future tax changes. If I begin to expect taxes on a certain class of assets to increase in the not so distant future, reducing their after-tax value to me, I can sell those assets before this happens, and the effect will precede the end result. cause. Such reasons may even be in the relatively distant future. Let’s say I majored in a field that I thought was relatively uninteresting but very lucrative. The expected increase in marginal tax rates for higher income individuals before I reach my peak earnings will reduce the income side of the comparison compared to an interesting but less taxed alternative and may even change my specialization today. decades ahead.

How such expectations affect policy choices is shown in introductory courses in macroeconomics. Initially, when building a model of aggregate demand and supply, texts often simply postulate that the change in aggregate demand is unanticipated, without specifying how we know it was unanticipated or when it would actually occur in the real world. This postulate leads to an easily understandable story about how people will predictably be deceived in such a situation and what they will do in response until there is enough time to answer more fully. This story becomes the main macroeconomic storyline illustrated by the model.

However, at some point later, the question arises of how expectations are formed and how this can change the main storyline, usually under the rubric of rational expectations theory. In a nutshell, the result is what has been one main story, in which any decision the macroeconomic change planners deceive people in the direction of policy change, resulting in the same reaction each time, becomes any one of four different stories (with options), where we don’t know which one will actually take place.

Suppose the government is trying to stimulate the economy through fiscal and/or monetary policy. The story with unexpected changes suggests that people are completely fooled by what the government is doing. Thus, such policy attempts, at least initially, move output in the desired direction and by the intended amount (if government planners also get a lot of other things right). But it can also happen that people “see” some political changes in advance, but underestimate their magnitude. If so, you will get the same qualitative history (in which direction the variables change), but the size of the effects will change depending on how much people anticipated the policy change. Unfortunately, policy makers need to get the scope, timing, and direction right in order to effectively stimulate or stabilize the economy. What could be the right policy may turn out to be wrong, at least to some extent. In addition, people can correctly anticipate policy changes and their magnitude. In such cases, these people cannot be fooled, and the expected impact on people’s behavior will disappear as people’s reactions offset policy changes. Policy in this case would not achieve the desired goals. It would be useless. And it is also possible that people will not only see the stimulus coming, but also overestimate its magnitude. In this case, output and its associated variables will move in the opposite direction to what is expected. Such a stimulus could lead to recession rather than growth, as evidenced by the stagflation of the 1970s that we hear so many warnings against repeating today.

This multifaceted set of possibilities, none of which can be relied upon with any degree of certainty, is very different from the theoretical world in which policy change is simply defined as unexpected. However, this is where we are now in macroeconomic policy. Many voices claim they have the ANSWER about what fiscal and monetary policy should be, both now and in the future, delivered with a big show of confidence. But the truth is that no one knows exactly what will happen when macroeconomic policies change, unless they know how people’s expectations will react. And if one’s confident self-assessment on this score is wrong, the results can be very different from what one expects, including the significant possibility that things will get even worse. Thus, we must recognize that, in our present circumstances, honesty requires that any serious response (at least in part) include “it depends” rather than the abundance of “trust me; follow my plan because I know what to do” from the snake oil vendors on the ring road. It would seem that we need a new Latin phrase to protect ourselves –warning (let the citizen beware).

Gary M. Galles

Gary M. Galles

Dr. Gary Galles is professor of economics at Pepperdine.

His research focuses on public finance, public choice, the theory of the firm, the organization of industry, and the role of freedom, including the views of many classical liberals and America’s founders.

His books include Pathways to policy failure, Faulty premises, Wrong Policies, Apostle of the worldas well as Freedom lines.

Get notifications of new articles from Gary M. Galles and AIER.