That’s definitely be much better than what they’re doing but it’s too low-level to make big savings. The things which cost large sums usually go back to high-level policies. In the case of the post office, that’s the congressional mandate to pre-fund employee pensions - they have to pay in your full estimated decades of pensions costs before you retire, whereas normal pension plans assume the company won’t disappear the day you retire – and the expectation that they provide service in rural locations. That’s a public benefit, vital for many people, but the post office can’t book that public benefit on their financial ledger.
A more widespread one is also more familiar to most HN readers: governments need a ton of IT, like everyone else, but are prohibited from offering competitive salaries or hiring staff instead of engaging a few large contracting companies. That has the direct impact of doubling costs and increasing turnover, but even worse is the higher likelihood of failure caused by not having institutional knowledge and expertise: contracting only works when you can provide clear instructions and hold the vendor accountable, but that requires you to have staff capable of doing so. Some agencies have been able to recruit skilled people based on their mission, but those are many of the teams being gutted now and it’ll take many years of better governance to convince people to come back after seeing how their public-spirited predecessors were treated.