SACRAMENTO — California’s governor has spent the better part of two years telling the state that housing production is up 59 percent under his administration, a figure that, as it turns out, was computed from data that did not include San Diego, and a handful of other cities with the bad habit of actually building houses, a methodology that one expert reviewed for POLITICO and called, in the most generous phrasing available to a professional, “a pretty basic error.”

The number itself is not in dispute. The 59 percent is real, in the sense that it was calculated, printed, and delivered to the public with the full ceremonial weight of a state press conference. What is in dispute is the denominator’s neighborhood, which, per POLITICO’s analysis of the underlying data, was assembled from a dataset that, at the moment of assembly, was missing the construction figures for some of the state’s largest jurisdictions, including San Diego, a city that, by most available counts, has been building at a rate that, had it been included in the tally, would have made the 59 percent look less like a boom and more like a rounding decision.

The discrepancy is, by the account of the experts who have looked at it, not a small one. The Dreier Roundtable, in its breakdown of the flawed methodology, notes that the incomplete dataset led the governor to inflate the rate of construction by more than three times, which is the kind of arithmetic that, in other contexts, would be described as a clerical error, and in this context is being described, by the people whose job it is to describe it, as a “methodology question,” which is the term a government uses when it has looked at a number, decided it is not a number it wants, and is now deciding what it is.

"It's a pretty basic error." — a housing economist, on a figure that was, until this week, the centerpiece of a state housing strategy

The administration’s response, such as it has been, has been to note that the 59 percent “reflects the data as it was available at the time,” a position that is, in the technical sense, correct, and in the practical sense, the equivalent of a restaurant explaining a missing course by pointing to the menu. The underlying data does exist; the cities in question report their construction figures on a regular schedule; the figures, had they been included, would have been available for the same stretch of years the 59 percent covers. They were not included. This is, as one state housing official put it in a statement that did not name the error and did not, in any sense, apologize, “a matter of data timing.”

THE 59, BY THE NUMBERS

  • Claimed increase in housing production: 59%
  • Increase, per the corrected data: "less than a third of that, by some estimates"
  • Cities missing from the original tally: San Diego and "other large jurisdictions"
  • Expert's description of the error: "a pretty basic error"
  • Administration's description of the error: "a matter of data timing"

The larger story is the one that does not depend on the arithmetic at all, and it is the one the housing industry has been telling, in a lower and more persistent register, for several years: that the state’s production numbers, good or bad, have not kept up with the state’s population, and that the gap between the two is the reason a studio apartment in the city the missing data comes from costs, by most available comparisons, roughly the same as a mortgage payment in a state where the governor has, until this week, been describing the housing situation as a 59 percent success story.

At press time, the governor’s office had not commented on the corrected figures, had not retracted the 59 percent, and had, in a statement on other matters, described the state’s housing agenda as “the most ambitious in a generation,” which, in the language of the data, is a claim that has, as of this morning, a new denominator attached to it, and a new set of people looking at it.