Reporting on how houses get built

What this section is for

The rest of this publication is reference material. A sheet on foundations explains what governs a foundation, and it is written to be consulted rather than read through. The journal is the opposite: pieces with an argument, written to be read once, about questions the reference sheets raise and cannot answer.

Those questions are usually about why an industry behaves the way it does. Why a construction method with obvious advantages has a shrinking market share. Why two adjacent counties enforce different requirements. Why the same change costs four times as much in month five as in month two. None of those is answerable by describing a building, and all of them affect what a building costs and how long it takes.

How these pieces are sourced

Every figure traces to a primary source, and the sources block at the foot of each piece names the publisher, the document and the date it was read.

For this section that mostly means two kinds of document. Federal statistics, principally the Census Bureau series recording how each year of new houses was actually built, which now runs back more than three decades. [1] And federal audits, which are unusually valuable in this subject because they record what a programme was supposed to achieve alongside what it did, decades after everyone involved has moved on. The 1976 General Accounting Office report on Operation Breakthrough is the clearest example, and it turns out to describe the present accurately. [2]

What these pieces avoid is the industry's own commentary about itself. Trade coverage of offsite construction, in particular, has been announcing an imminent transformation continuously since the 1960s, and it is a poor guide to what has actually happened. The completion counts are a better one.

How a piece gets chosen

The test is whether a question keeps arriving from a different direction. The pieces here all started as a sentence that had to be written repeatedly on reference sheets, at which point it was clearly doing more work than a sentence should.

The code adoption piece exists because every page that mentions a requirement has to explain that requirements are local, and the explanation deserved its own place. The change order piece exists because the same cost mechanism turned up in budgeting, in contracts and in scheduling, described three times and understood once. The factory housing piece exists because the offsite sheets kept gesturing at a market share that needed accounting for.

That origin has a consequence worth stating. These are not surveys of a topic, they are arguments about a specific question, and they take positions. Where a position rests on a figure, the figure is cited and the reader can check it. Where it rests on a judgement, the piece says so rather than dressing the judgement in numbers.

What is not here

No product reviews, no recommendations and no affiliate links. Nothing in this section is chosen because a manufacturer asked for it, and no figure in it was supplied by one.

No predictions either, or as few as the subject permits. The pattern in the data is that predictions about this industry have been wrong in the same direction for fifty years, which is a good reason to describe what has happened and let a reader draw the forecast.

And nothing published under an invented name. Pieces here are unattributed, under the publication, which is the arrangement most trade and reference publications use and which is explained on the standards sheet. the contents sheet carries the full contents.

A word on corrections, since a publication that takes positions will eventually take a wrong one. Where a piece is found to be wrong on a fact, the fact is corrected and the correction is stated on the page rather than made silently, and the revision date at the head of the sheet moves. Where it is wrong on a judgement, that is harder to adjudicate and the honest response is usually to say what changed the view rather than to quietly redraft. Neither of those is unusual practice. Both are worth committing to in writing, because a publication that has never said what it does when it is wrong has not thought about it.