←back to thread

Scale Ruins Everything

(coldwaters.substack.com)
175 points drc500free | 1 comments | | HN request time: 0.323s | source
Show context
efitz ◴[] No.41848475[source]
It’s not that scale ruins everything, it’s the pursuit of scale that ruins everything.

I have thought a lot about antitrust recently and realized that it’s an overlapping problem with the VC unicorn problem. People get so hung up with being bigger/biggest, faster/fastest, that they minimize or ignore pathological side effects.

What if we have a progressive tax structure on business based on scale? Partly on size, partly on market power?

For example, what if corporate income tax were tied to the logarithm of the number of employees+contractors? What if the corporate income tax increased with the market share in markets served?

These kind of ideas have very low impact at small scales, but after a certain scale they start to have such a huge effect on bottom line that they disincentive growth for its own sake.

And they’re self-regulating and largely objective, unlike our current antitrust laws in the US.

replies(5): >>41848614 #>>41849918 #>>41850221 #>>41851217 #>>41851734 #
userabchn ◴[] No.41849918[source]
> what if corporate income tax were tied to the logarithm of the number of employees+contractors? What if the corporate income tax increased with the market share in markets served?

Is there a reason for choosing those rather than the simpler alternative of the corporate income tax rate increasing progressively with revenue (as it does for individuals)?

replies(2): >>41850114 #>>41850705 #
1. efitz ◴[] No.41850114[source]
I dunno. I have a natural affinity for continuous vs discrete functions. But I also think thresholds cause bad behavior when you are near a threshold; you try to game the system by exceeding the threshold while trying to appear to be below it.

Corporate accountants are sophisticated enough to deal with formulae, and small businesses would be unaffected as the function changes very slowly at small numbers so even an error would likely not materially affect your return.

Calculation of market share would be more problematic.

And of course if you don’t like logarithms then choose your favorite exponential function.