Note: I'm trying to write a blog post a day for 100 days. This is day 1, and let's just say... I'm kind of rusty on writing. Let's hope it gets better from here...

The recent Trump administration's focus on tariffs have made me think about the US economy and its structure. There's a clear reverence for manufacturing and manual labor in the US, and not limited to the new administration. I've seen an nostalgia for manufacturing jobs in the US for most of my life (probably not surprising as this was the period during which most of these jobs were offshored or otherwise disappeared). More recently we've seen the US care about industrial policy from a security angle, e.g. with the CHIPS act.

So I get why manufacturing is important to a lot of people from an employment perspective. But recently I was thinking about how an important aspect of manufacturing or manual labor is that, in general, it is universally productive. When we make a physical item, we're producing something of societal value. (Of course you might have externalities like pollution, but let's assume we've priced those in). Value needn't be physical: a massage therapist produces value by leaving you feeling better than when you left. But when you're producing or transforming a physical thing, by and large someone will purchase and use that thing. This can fail to be the case in the short-term: you can build a toy no one wants, but in the long term you won't stay in business. The big exception here is military spending, where the good outcome is that whatever your produce isn't used and instead sits idle.

But not all work is productive, and I don't mean just "bullshit jobs": the US has a large fraction of its economy tied up in occupations that counterpose each other and produce little to no societal value.

The Service Economy

Once we leave manufacturing behind, it becomes a bit more difficult to measure what productivity means. We often use something like GDP to measure such things, but it's easy to produce GDP without being useful. The oft mentioned "bullshit" administrative jobs "produce" GDP, but don't provide a valuable service as a result. If there's a city regulation that you have to pay me a 1% fee to get a special stamp for your car, I've got a good business, but I'm not exactly contributing much to society: no one is going to miss me if I'm gone.

There are lots of examples of regulatory capture out there: given our example above, car dealerships come to mind. But I'm more interested in the moment in how local-minima in private markets can result in similarly pointless (but importantly only pointless in the large, not in the small!) work, which we see exhibited clearly in health insurance and law.

Counterposing Forces, Zero-Sum Games and why your Visit to the Doctor is so Terrible

My main thesis here is that the US has developed advanced industries, with millions of workers whose primary job is competing in zero-sum games. The archetypal zero-sum games is of course gambling: no money is generated by the gambling industry, just moved around. You might argue that a casino is like a massage therapist, but I have to assume there are a lot fewer massage addicts out there than gamblers...

But while gambling is clearly a complicated way to transfer money around, we also have more obscure zero-sum games we play, for example our health insurance system. Most anyone who has gone to the doctor in the US will be familiar with the bizarrely complex billing process, with astronomically high prices for items such as e.g. $76 rolls of gauze. What's going on? It's a game the hospital administrators and coding specialists are playing with the insurance claims adjusters. And not a cheap game! If you go by the BLS Estimates the US employs ~1 million people in these roles (if you assume some amount of hospital administration and management is also involved in the coding). Doctors in the US are also compelled to produce the proper codes and justifications for procedures, reducing the amount of time they can spend on the productive work of helping people, but even without that loss in productivity you can see this is a large time and monetary cost.

Especially frustrating in healthcare is that for a variety of reasons this "game" is that it's not a fixed fraction of the cost, but escalating over time, e.g. from the Commonwealth Fund report

OECD Health Statistics data show the U.S. spent $1,055 per person on “governance and health system financing administration” in 2020, compared with the OECD12 average of $193 per person

What's fascinating is the local minima we've constructed here: for any particular actor, they have to participate in the game, but as a whole system, the societal value is effectively nil. The value of a dentist for having an administrator that is good at billing is huge, but wouldn't be necessary if there was a higher trust environment, if cash payments were the norm, or if billing was standardized. If you start looking around, you see these contrasting forces all over the place. Early in my career I worked on identifying "web spam" at Google - filtering out low-quality web pages that served just ads. But both my work and the spammers only existed because the system for web search created a zero-sum game for this type of activity. Lawyers are another good example. Winning a particular case might mean a billion dollars either way, but systemically lawsuits don't produce value for society: a society that was identical but just had fewer lawsuits would presumably have more smart people available to do something more productive.

In all these cases, a better system would do exactly that: leave us the same effect but with less effort. For insurance, government provided healthcare is an obvious alternative, but even regulation which e.g. standardized the billing rate for hospital procedures would radically reduce the space for conflict between hospitals and insurers. For web-spam, the increase in the use of LLMs for search dramatically reduces the value of positioning in web results and may mean the end of this game as well.

Solutions are Hard

In all these areas, it seems like we'd be better off as a society if we could eliminate or reduce how much time and effort and money we spend playing these games. But because they're systemic, no single actor can break the cycle. You effectively need an external force, most likely government, to intercede and change the rules of the system. As these sectors become larger, however, they gain more polictical clout, and so it becomes harder to intercede without risking political ruin. For example, imagine a world where the US moved to single payer health insurance - even in the case where the care was identical to today, you would be eliminating millions of middle-class jobs!

In lieu of government, I assume these systems will continue to grow until they become untenable. Perhaps an economic collapse or AI will obviate the need for humans in these roles, or more hopefully they are replaced by an alternative model.