Ridgway 1956: The Forgotten Prequel

The Goodhart family has a forgotten eldest son. The maxim everyone quotes was actually worded by Strathern, not Goodhart himself (B01). Push the timeline further back and the first person to write down "measures fail" systematically was neither Goodhart (1975) nor Campbell (1969) but a scholar named Ridgway, in 1956. He used eight pages to lay out everything the family would keep rediscovering afterwards, and then the entire field forgot him for sixty years.

Ridgway published a short article in Administrative Science Quarterly in 1956 with a very plain title: dysfunctional consequences of performance measurements. Those eight pages come thirteen to twenty years before Campbell's core statement and nineteen years before Goodhart, and they already have the family skeleton complete. He divided performance measures into three classes and analysed each separately: single measures, multiple measures and composite measures, each with its own way of distorting. The article is itself a review. Ridgway says at the outset that what he is doing is gathering up the knowledge about dysfunctional consequences of performance measurement that was scattered around at the time. Which is to say that as early as 1956, "a measure deforms once you grade people by it" was already field common sense in sociology and management, before it carried anyone's name. By 2024 a cross-disciplinary paper on proxy failure had put Ridgway back in as an early anchor, and the literature on gaming in healthcare performance had started citing him in the opening lines, in the tone of: we were warned about this back in 1956.

The body of fieldwork Ridgway reviewed is the most solid empirical root system on this line. The most vivid is the sociologist Blau's The Dynamics of Bureaucracy (1955). Blau studied a public employment agency where clerks were graded on the number of placements, so they started picking their customers: keeping the easy-to-place applicants for themselves and pushing the hard ones away. The reason is simple. Getting a hard-to-employ person into a job and getting an easy-to-employ person into a job score the same on the measure, and the first takes far more work. Once the measure was hung, clerk behaviour immediately slid from "help people find suitable work" to "make my own numbers look good." Another is Argyris's 1952 study of how budgets bear down on people: a budget is a planning tool, and once it becomes the nail on which individuals are assessed, it starts to deform behaviour. Each study saw only one point, and Ridgway's contribution was to string them into a general proposition. Trace further back and the root system goes one layer deeper: Merton's 1936 paper on the unanticipated consequences of purposive social action gives the general framework of "once people intervene with a purpose, chains of unforeseen consequences follow," which is the ancestral form of "a measure deforms once you grade people by it." So strictly speaking Ridgway is the first to turn scattered evidence into a proposition, but the observations he stood on are older than 1956.

Ridgway's concluding sentence is as cool as a verdict: quantitative performance measurements, whether single, multiple or composite, have undesirable consequences for overall organizational performance. But the best line he left is a metaphor. He compared the misuse of measures to treating penicillin as a cure-all: the same tool, used carelessly, brings side effects and backlash that outweigh the benefits, just as when penicillin was first treated as a wonder drug, and the cure is sometimes worse than the disease. That line, the cure is sometimes worse than the disease, is the earliest and most vivid aphorism in the whole family, and it has almost never been quoted in the popular articles that discuss Goodhart. Line the family up by date and it is clear: the earliest empirical review is Ridgway (1956), the social science formulation as a law is Campbell (thought present in 1969, written up 1975 to 1976, journal version 1979), the economic methodology theorem is Lucas (1976), and the monetary policy institutional generalization is Goodhart (1975). Ridgway stands at the front, and Goodhart, having taken the advantage of the naming, took the largest posthumous fame. That is not a coincidence, it is the recurring pattern: whoever names the phenomenon is the one history remembers. His sixty years of obscurity have a structural cause too: in 1956 the insight had no name, it was scattered through field reports in organizational sociology, and it did not travel, stick or cite like a law with a person's name on it. Only when economists in the 1970s stated similar things as Goodhart's law and Campbell's law did the family acquire a bright signboard, and Ridgway's name was not on it. It was not, in fact, an invention of 1970s economists but a discovery of 1950s organizational sociology, and the discoverer simply missed the naming dividend.

One open question: Ridgway, Campbell and Goodhart look like three parallel and mutually unaware lines of discovery, but that is only a preliminary judgment, since none of the three cites the others. If they really are independent, then the same insight being independently rediscovered by at least three groups over twenty years is itself a measure of how hard it is to avoid; if there turns out to be a hidden citation link we missed, the family tree has to be redrawn. Nobody has yet gone through the full reference list in Ridgway's review item by item to check.

The one-line takeaway: the earliest and most complete article in this family was written in 1956 by a man named Ridgway, not Ridgeway, and the family does not start with Goodhart.