Capitalism in America; Review

What It's About

Alan Greenspan ran the Federal Reserve, the government bank that steers the US economy, for almost twenty years. He wrote this book with Adrian Wooldridge, a longtime editor at the Economist magazine. Their argument covers three centuries and stays simple. America grew rich by letting old ways of working die when a better way shows up. Economists call that creative destruction. The refrigerator killed the ice-harvesting business, and the people and money that once cut lake ice went off and built better things. Most countries protect the dying industry. America mostly lets it go, and the authors say that habit is the whole secret. They measure the gain in productivity numbers, and after three centuries the numbers are hard to argue with.

They do not hide the cost, which surprised me. The churn wrecks real livelihoods and real towns, and the book says so. Their answer is that the trade is worth it. Their proof is the 1930s, when the government tried to hold the churn still. The country got poorer and the recovery ran slower.

What I Think

Somewhere in the middle of the book, the American worker leaves the farm and lands at a desk. The authors write it as a handover. The manual worker gives the economy to the knowledge worker and steps out of the story.

I read that passage twice because it did not sit right. A week earlier I stood in front of a fire alarm panel and had to rebuild its maintenance history myself. Most of the paper trail never existed. Programming that panel was knowledge work under any definition Greenspan uses. An hour later I was up a ladder wiring devices.

No handoff happened in that room. The technician kept both kinds of work, and the book never sees him. That gap is what this review is about.

What It Gets Right

I opened this book expecting a fight. I ended up agreeing with its main idea, because I have watched it run inside my own trade and cannot pretend otherwise.

Fire alarm panels used to be zone systems. A zone panel groups devices by area, so an alarm only tells you which hallway to go search. The newer addressable systems give every device its own address, and the panel names the one device that is in trouble. The old hands fought the change, and they were wrong. A callback is a return trip to fix something that did not stay fixed. Addressable panels earned fewer of them, and the callback settles that kind of argument. The field figured it out faster than the debates did. The same thing happened when cellular communicators replaced the dialers that ran on copper phone lines. I burned years chasing trouble signals, the panel's way of reporting a fault. The cause was corroding copper somewhere on the run to a phone office nobody maintained anymore. The trade is better off with that gear gone.

The book gets a quieter thing right too. Letting old methods die takes discipline, and the pressure always pushes the other way, toward protecting whatever is dying. A tradesman feels the same pull on a smaller scale every time the code cycle retires an old habit.

What It Gets Wrong

The book counts destruction in the units it can see. Those units are industries and jobs. Lost knowledge never gets counted, so the book cannot price it. When a trade loses its old hands, it loses things nobody ever wrote down. One example: what a smoke detector actually does after twenty years of salt air, no matter what the spec sheet promised at install. No productivity chart shows that loss. You find out about it during a fire.

The deeper miss is the workforce itself. In this book the worker is an output. The economy turns over, and the workers it needs are supposed to appear on the far side, retrained by markets and schools. Schooling gets pages. Apprenticeship gets nothing, even though five-year apprenticeships built the workforce the book calls manual. The book never wonders what happens when the next era needs a worker that nobody is building on purpose. That worker has to do the hand work and the thinking work at the same time, on one paycheck. The trades have been that worker for a while now.

That worker exists. The federal government has already built a safety model that depends on him.

What I'd Add

Here is the book's argument at work in my part of the world, in a document almost nobody reads.

NFPA 72 is the national fire alarm code. It requires the devices that start an alarm, like smoke detectors and pull stations, tested 100 percent every year. That schedule did not come from a desk. Like most fire code, it came out of real fires where the devices did not work.

The military runs on a different set of books. The Unified Facilities Criteria (UFC) are the standards the Department of Defense writes for its own buildings. The fire protection volume is UFC 3-601-02, dated 7 Oct 21 with Change 1 on 18 Apr 25. It tells its users to apply its own task frequencies "in lieu of the tasks and frequencies in the National Fire Codes." Its Table 2-1 puts those same alarm devices on a two-year test cycle. The cycle is a straight two years, with no plan that rotates half the devices through each year. That cuts the testing in half, on purpose.

The two-year cycle came from a reliability study an Air Force engineering agency published in 1999, Technical Report 01-10. The study assumed a building sees a fire demand about once in fifty years. It also assumed the tech does the testing right 99 times out of 100. Feed those numbers in and the math holds. The annual cycle dies and a cheaper one replaces it, with system reliability still rated at 99 percent. Greenspan could read that document and find his whole book in it. So do I. An old method died and a better-priced one took its place, applied to the systems that keep a building's people alive.

Now walk the field. On some bases I walked, 60 percent of the fire alarm systems sat disabled, some fully and some in part. Missed maintenance stacks on missed reporting, year after year. The deficiency list, the running list of known problems, quietly becomes the only real record of the system. The study is still defensible on paper. The assumption underneath it has not held.

Look at what 99 percent actually demands. A tech stands on a ladder with a can of test smoke the manufacturer approves. He trips the detector, then checks its sensitivity, meaning how much smoke it takes to trip. The criteria track those readings against limits across five years, so this year's number only makes sense next to the old records. That is knowledge work, and he does it by hand. The write-up is its own second task on top of the testing. It has to come out clean enough that the next tech who opens that panel can pick it up and work from it.

The criteria assume a tech who is fluent in both kinds of work. They assume his paperwork holds up the same way his testing does. The study that halved the testing also raised the bar on the person doing it, and no one paid to build that person. No new training money came with the new frequencies. The trade kept turning out manual workers for a job the model had already redefined.

The UFC's writers get no blame from me. They built a model, and every model stands on its assumptions. This one stands on us. Sixty percent disabled is what that unfunded assumption looks like twenty years on.

The techs inside those buildings were right the whole time, for whatever that is worth. They said for years that the cycles cannot survive staffing gaps and turnover. Being right earned them nothing. A two-year cycle assumes the same crew is still there in two years. Turnover means they often are not. The study never had field data. Only the people who kept touching the gear know how long these systems really last. The deferred work is coming due now, and there is exactly one labor pool that can price the recovery.

So the work runs toward us, and it is the unglamorous kind, basics done at the level the study counted on all along. Build techs into knowledge workers, because the federal criteria already treat them as such. Start with an apprenticeship that treats the write-up as seriously as the wiring. The paperwork counts as the work now, and the report stays with the building. Some day another tech will stand where you stood, and what you left behind is what he gets.

Verdict

I expected to fight a central banker for three hundred pages. Instead the book handed me the clearest description I have read of why my own trade's testing schedules got rewritten. Its one blind spot is the person its newest models depend on. Read it to understand why old methods die and keep dying. Then train for the part the book left out, because the government's own criteria already assume you have.