… is from page 20 of Deirdre McCloskey’s 2021 volume, Bettering Humanomics: A New, and Old, Approach to Economic Science:
Some definitions and their corresponding theorems are wise and helpful, some stupid and misleading. The humanities, and the humanistic step in any science, study such questions, offering more or less sensible arguments for a proposed category being wise or stupid, in advance of the counting or comparison or other factual inquiry into the world. The humanities are the study of the human mind and its curious products, such as John Milton’s Paradise Lost or Mozart’s Concerto for Flute and Harp (K. 299) or the set of all prime pairs or the definition of GDP. The studies depend on categories, such as enjambed/run-on lines or single/double concerti or prime/not-prime numbers or marketed/unmarketed products, such as we humans use. God doesn’t tell us. We do.
DBx: In no science do “the data” ever speak for themselves, and in the social sciences “the data” standing alone are especially prone to utter only unharmonious gibberish. As Deirdre says, we humans must make judgments about how to classify phenomena. If such classifications are unwise, all the data and number crunching will be, at best, pointless, and often positively menacing.
What, for example, is the meaning of “manufacturing jobs”? Jobs classified as “manufacturing” are counted very precisely. Is a truck driver employed by the steel-maker Nucor Corp. a manufacturing worker? Is this truck driver still a manufacturing worker if Nucor decides to no longer produce in-house its own delivery services but instead to purchase such services from another firm who then hires this driver to help it satisfy its contractual obligations to Nucor? Same flesh-and-blood human being performing the same physical tasks.
What about measures of economic inequality? What is income? Does it include fringe benefits? If so, what are fringe benefits? Are well-stocked employee kitchens a fringe benefit? What about company-sponsored happy hours on Friday evenings? And whatever we choose to classify (and not classify) as income, should we measure income difference across flesh-and-blood individuals? Families? (What’s a family?) Households? (What’s a household?) Why not only measure income differences of workers of the same age? (And who, exactly, should be classified as workers?) How relevant are income differences within countries? Or across countries?
And why is income – or wealth – measured in money so important? Why not measure instead actual consumption by individuals or households of goods and services? (“Oh, monetary incomes – or differences in monetary wealth – you see, are easy to measure and quantify. Doing so is much easier than measuring actual consumption of goods and services.” Well, yes. And while ease of measurement is an important consideration, surely classes of phenomena don’t become relevant or helpful simply because they are easily measured.)
Measure, count, calculate – by all means. These processes are vital tools for improving our knowledge and understanding. But don’t do so mindlessly. Don’t suppose that just because you’ve laid your hands on a large – a “Big” – data set, and then processed and regressed those data to generate calculations out to six numerals to the right of the decimal point, that you are necessarily doing science and that your results are meaningful and important. Your results might be meaningful and important, but, if so, they are not made so of their own accord. To do science correctly, serious theorizing is required. Required also is the exercise of good judgment. And to develop good judgment one must do far more than ponder data and methods of data processing.
