> For a profit making company, this means it loses its competitiveness and then falls. For government, this means it gets a bigger budget.

I would argue that with companies, the expected outcome is that they accumulate and then stabilize at a "typical" level of bloat. For software, Google is a good example: Much more bloated than they used to be, but still competitive enough, and looking stable.

Governments also have some ways to de-bloat, mostly from internal and external pressure (like Greece).

I think that consistently getting services at no-bloat prices is just not realistic in general over longer timeframes; Netflix, Google, Uber, Amazon, ... give countless examples of excellent early price/quality ratios that got inevitably worse over time.

My hypothesis is that for our current understanding of what a government is supposed to provided (which did expand significantly since WW2, but mostly for good reasons) we have now reached a "bloat-stable" budget, and there is just no "easy" way to slash this budget without hurting the services (even though in a hypothetical bloat-free environment these could be provided for less).

Here is interesting data which seems to kinda support my point, with government spending as GDP percentage for the US, Germany (less GDP/capita) and Switzerland (more GDP/capita): You can see budgets rising post WW2 across the board, and then stabilizing in the 30-40% range (slightly higher for the poorest country, which makes sense). I would have honestly expected to see those trending up even for the last decades (which they don't really) and Covid was also less prominent than I would have assumed:

https://ourworldindata.org/grapher/government-spending-vs-gd...