Raw Thought

by Aaron Swartz

Thinking Clearly About Piece-Work

My friend Jonathan Zittrain has been working on a book about how the return of piece-work will destroy America. As someone who kind of misses piece-work, it made me want to think through the issues involved.

Because, in the real world, competition is imperfect, when a company sells something it earns a surplus. That surplus then must be divided between capital and labor, or more concretely, the company and the employee. The terms of distribution are determined by the employment agreement (i.e. how much the employee will get paid).

If labor has all the power, it will want almost all of the surplus, so it’d demand an employee agreement where it gets paid a big chunk per good produced — this is basically piece-work. If capital has all the power, it will want to pay employees barely enough to stay alive and be able to show up for work — this is basically wages. So, to a first approximation, wages are what you get when capital is in charge and piece-work is what you get when labor is in charge.

But there are some additional considerations. The transaction costs for capital are relatively low — you can invest in a whole bunch of companies at a time, whereas you can really only do a couple jobs at a time (or usually just one) and it’s quite painful to find another one. So employees also want insurance — they want to guarantee they’ll keep getting paid even when the market for the stuff they produce declines. Again, we can imagine two extremes: if labor is in charge, it’ll want a job for life, where getting downsized is inconceivable; if capital is in charge, it’ll want to employ people for exactly as long as their marginal product is profitable — which means staffing up in booms and downsizing freely in busts (and not just business cycle booms and busts — companies want to hire lots of temporary workers for Christmas, for example).

Now obviously a job for life doesn’t make a ton of sense in a piece-work world. If AT&T has guaranteed you a job for life with steady wages, then it makes sense for it to invest in retraining you for some different part of the business when mechanical telephone switches get replaced by computer servers. It’s less clear how this would work in a piece-work deal. But the fact is there’s no reason the insurance part should be tied to the wages. In Denmark, for example, if your job disappears the government simply pays you 90% of your old salary until you find a new one (up to some cap, of course). There’s no reason piece-work can’t coexist with this kind of social insurance system, which would seem to be the choice of our all-powerful labor.

But that’s treating labor as a monolithic entity. The third thing wages do is collapse differences in pay between workers. For example, it’s well established that programmer productivity varies by an order of magnitude — the best programmers at a company can be ten times as productive as the worst programmers — but I’ve never heard of a company where programmer pay varies by anywhere near that much. Really great programmers might be paid double or perhaps even triple the worst programmers, but I’ve never heard of anything close to a 10x difference.

The one place where you do see these really huge differences is in CEO pay, but even this isn’t really a counterexample since CEO is a job with no intra-company reference class. That is, it’s not like companies will have one CEO getting paid $1.2M and a another getting paid $12M (indeed, I’d guess companies with co-CEOs find them getting paid the same amount) — instead, the order-of-magnitude differences are all found across companies.

But order-of-magnitude differences are totally possible with piece-work, especially the kind of intellectual piece-work that Zittrain is concerned with. Krugman has bragged that he “writes faster than anyone in journalism” and it’s quite possible to imagine him turning out columns in a tenth the time of Barbara Ehrenreich (who turned down the NYT op-ed columnist job because it was too time-consuming). So wages might be a way of quietly redistributing money from the speedy Krugmans to the dawdling Ehrenreichs. But just as with social insurance, you can imagine this role being taken up by the government instead: through progressive taxation.

Now according to classic economic theory, these changes wouldn’t just be details of style, but would increase the size of the overall pie. And, on the squishy side, they’d provide much greater scope for human freedom. Assuming I was guaranteed a decent wage either way, I’d far rather be able to stay up late working one night in exchange for blowing off work the next. Not to mention getting to work the hours I want, from the place I want, in the way I want, etc.

Now the practical fact is that most jobs don’t have a concrete enough product to be amenable to piece-work. They’re a mixture of all sorts of different tasks, require interaction with a specific group of other people, and have all sorts of other features that kind of force them to be your usual office job. But none of that applies to Zittrain — he’s talking about jobs that are already piece-work and arguing that they shouldn’t be. But it seems to me like, if you’re on the side of labor, your preferred solution should be more social insurance and progressive taxation instead.

You should follow me on twitter here.

July 5, 2012


Regarding “… working on a book about how the return of piece-work will destroy America.” - I hope the book is as you describe! (i.e. has that perspective - I’m looking forward to it).

On “But it seems to me like, if you’re on the side of labor, your preferred solution should be more social insurance and progressive taxation instead.”

OK, but how about an alternate more “local” solution if such a preferred “global” solution seems too difficult politically?

posted by Seth Finkelstein on July 7, 2012 #

You can also send comments by email.

Email (only used for direct replies)
Comments may be edited for length and content.

Powered by theinfo.org.