Logic, facts and context

A good example paying attention to the framing of an argument, a matter that is critical to effective communication and yet rarely performed well.

In this case, two excellent economists (dismal science, the study of scarcity), advance an argument in the WSJ that the federal government ought to subsidize economics research in the same fashion as it does scientific research. They make their case broadly in terms of logic and empirical evidence that research in economics has a positive return.

Tyler Cowen in Does the dismal science deserve federal funding? backs up the argument to the real question. Given limited federal spending, which of the many productive ways of spending those scarce resources will have the greatest return? It is not enough that economics research have a positive return, it has to have a positive return greater than all the other competing ways of spending those limited resources. That is the first order question that rarely gets asked.

But there is a second level of sophistication where the decision-making gets even more tangled, sophisticated and interesting. This has to do with what I call the shopping mall paradox. If you own a shopping mall with 100 retail stores and you want to maximize your profit, you would look at the collection of 100 stores and calculate which stores have the highest profit per square foot. I do not have real world experience of this but from what I have read, the answer is that shoe stores have the highest profit square foot.

So if your goal is to achieve maximum profit, then the logical thing to do is to convert all 100 stores into shoe stores.

The paradox is that, absent some particularly unique circumstances or marketing capability, a mall entirely devoted solely to shoe stores will actually have less profit than a mixed-use mall. Even though shoe stores have a very high profit per square foot of space, there is rarely community demand for a mall entirely dedicated to shoes.

The real challenge for a mall operator is to optimize profit for the overall organic whole by striking the right balance of parts. There are an inordinate number of factors and considerations to take into account in guessing how to optimize profit. If people go to a mall primarily because it is more efficient to do most their shopping in a single location, then you have to have a portfolio of stores that meet most their needs, even though those different types of stores will have widely varying profits. A general department store or a grocery store may have a lower profit per square foot than a shoe store but you need those anchor stores to draw the shopper volumes necessary to support the shoe store.

Now multiply these trade-off decisions across a mix of 100 stores and trying to find the optimum mix between meeting basic consumer needs and maximizing profit and you can see how challenging that task is.

Clarity of goal (maximize profit), logic (identify the highest profit stores and maximize) are both necessary to a good argument but are not sufficient. You have to have context which is what Cowen is pointing out in his post. Yes, economic research likely pays dividends. That is insufficient knowledge though. We have to know that it pays greater dividends than alternative uses. But that also is insufficient knowledge. It comes down to the portfolio - what is the optimum mix of investments in order to achieve some mix of overarching goals. Even if economics research had a higher return than, say defense spending, you wouldn't want to invest everything in economic research and nothing in defense. Optimisation of a variety of outcomes in a system with multiple independent but contingent subsystems is problematic and depends greatly on contextual knowledge. Contextual circumstances is what so often leads to unintended consequences.