Politicians and Paradoxes

By Kevin D. Williamson
Thursday, May 05, 2016

Some of you computer scientists and economists out there will be familiar with Bonini’s Paradox, which holds that making scientific models of real-world phenomena is difficult because as the models grow more accurate they become as difficult to understand as the underlying realities they represent; conversely, the more simple a model is, the less accurately it represents reality. As the French philosopher-poet Paul Valéry put it, “Everything simple is untrue, everything complex is useless.”

Or: A complete and accurate map of the world would be world-sized.

This is why some economists (and some critics outside the field) have doubts about the real long-term usefulness of economic models. There is a great deal of mathematics in economics, and a great deal of mathematics in engineering, but economics isn’t very much like engineering: Engineers can generally be quite confident about whether suspension bridges of a certain design will stand or fall, whereas economists cannot honestly say with the same confidence what the results of any given economic policy will be. (Unhappily, this does not stop some of them from making pronouncements with unwarranted certainty.)

Understanding the workings of a modern market economy is more like observing the development of an ant colony over a period of years than it is like setting a broken humerus or designing a microprocessor. General patterns of behavior emerge and can be anticipated to some degree, but what ant business any given ant is going to be getting up to on any given day is very difficult to predict. Other complex systems present similar challenges: We know about general weather patterns, and we might even have some ideas about long-term changes in future atmospheric conditions, but nobody can tell you when the first day of snowfall is going to be in Topeka, Kan., this coming winter.

Economics is better at telling us what is happening than what is going to happen. There’s an art to asking the right question and a science to finding the answer.

A very interesting informal discussion has been going on among economists, including Dietrich Vollrath, of the University of Houston, and Ernie Tedeschi, formerly of the Treasury Department, about a seeming paradox: Employment is growing, and improvements in key technologies continue at a steady pace, so why is growth in productivity relatively weak? There are more people working, and the tools are getting better at a steady (and in some cases remarkably fast) clip, so why isn’t the output per man-hour more impressive than it is? Why did productivity fall in the first quarter of 2016?

The answer seems to be, in part, that workers aren’t actually using those awesome new tools.

Put another way, what Tedeschi found is that the growth in employment is skewed toward lower-productivity industries and away from the high-productivity ones in which those technological advances presumably would have the most impact. Firms in industries such as finance, information, and mining (in which oil-and-gas production has a very large economic footprint) have hired fewer people, and those in sectors such as education, leisure and hospitality, and health-care services have hired more. The technological innovation is in one part of the economy, and the job growth is in another.

(This scenario obviously is complicated by the fact that one of the purposes of innovation in fields such as automation is to reduce the need for human labor. I trust the authors will forgive this and other simplifications.)

Vollrath gets into the math and details in an interesting post headlined “Technology versus the Distribution of Workers in Aggregate Productivity,” in which he finds: “Workers are not only moving into low-productivity sectors to begin with, but into sectors with low productivity growth.”

Much of our political debate, about economic matters as well as others, happens at a level of abstraction that makes it in Valéry’s term a worst-case scenario: so simplified that it is almost useless. Politicians talk about creating jobs, and commentators (myself included) often are fixated on measurements such as the employment rate or the work-force participation rate. This is in no small part because our hippic understanding of politics demands a scorekeeping device. A shopworn but nonetheless true piece of advice for business managers warns that focusing only on what is measureable (and conveniently measureable at that) can lead us astray.

Because our attention is dominated by that which is more easily and more directly measureable, we may miss other, crucial, information. Yes, it matters how many people are working, but it also matters — and matters a great deal — what kind of work they are doing.

In much the same way, our discussion about government spending probably focuses too much on spending levels (not that those are unimportant) while giving insufficient consideration to the fact that it matters — and matters a great deal — what we spend the money on. Federal spending levels were quite high during World War II, and the great libertarian and economist Milton Friedman to his credit never regretted his main contribution as a young economist working in the Roosevelt administration: designing the income-tax withholding system. Beating Hitler and saving the world was worth the expense. Getting monkeys high on cocaine, less so, though surely those experimental subjects were the envy of lab monkeys everywhere.

Ideologies are models, too. And, like a good map, they may help to guide us, but they are simplifications. The map isn’t the territory, and “there’s a difference between knowing the path and walking the path,” in government as in all things. Economies simply are not manageable in the way our political rhetoric assumes. What is manageable is the protection of property rights, including the right to produce and to trade without political interference, the maintenance of stable rules and independent judiciaries for the enforcement of contracts, and transparency in public affairs. That’s more likely to lead to the desired outcome than is a targeted seven-year partial tax credit for . . . whatever.

But politicians don’t talk that way, and we cannot any longer pretend not to know why.