There are systems that encourage the opposite behavior, but they tend to be damaging in other ways. It all comes down to the metrics that you are responsible for as a manager.

I used to manage a grocery department at a well-regarded supermarket chain. In retail, one of the only inputs you can affect at the store level that impacts the bottom line is labor, so metrics are usually organized around labor optimization. Each week, you would be expected to exceed the previous year's Units Per Labor Hour (UPLH). This can only be accomplished two ways: increasing unit movement (which you really have no control over and is largely driven by consumer sentiment and population density) or cutting labor hours. In fact, even if volume is increasing, you can't even add hours even though labor needs scale with volume in a brick-and-mortar retail setting. To exacerbate things, the reductions in labor have already compounded yearly since the policy was put into place. They want you to squeeze water from a stone. "Well, So-and-So (who you've never met) was able to increase UPLH by 10% 5 years ago! Why can't you?"

I guess my point is that systems that discourage waste often lead to insane work environments rather than cleverness or innovation, whereas encouraging waste leads to bloat and inefficiency.

In my experience, it's rarely about flagrantly wasting money for the sake of running up expenses to a budget level. Rather, it's usually more about pulling in expenses, doing something you had wanted to do anyway but didn't think you had enough money for, etc.

Budgets are a pretty powerful and widely used tool for managing organizations. And there are other checks and balances. It's not like a manager can necessarily just decide to have a team off-site in Hawaii because there's room left in the budget.

Maybe I just put too much stock in that one episode of The Office.

I think the disconnect is that you are describing it as a system, when in reality it's just a bunch of misaligned incentives and plain old human nature. It's unfixable.

It might be both of those things at the same time. This book claims so https://en.wikipedia.org/wiki/Systemantics.

The money isn't wasted, it just goes somewhere else. Better that it go through you (and to the contractor), than through someone else at the company.

There is, but it’s hard—like, theoretically murky hard.

So the basic inefficiency is that you want to do top-down resource allocation, “I approve of this much budget going to that project.” We could call that bureaucracy, or The State, or whatever. There is a reason that every modern military in the world has this bureaucratic gene: you can track who is responsible for every dollar easily, which limits the scope of corruption. Corruption does not itself kill the other countries: it just places an upper bound on how much money, how many resources that military can effectively put to use. But the militaries who can inefficiently use unbelievable resources clobber the ones who efficiently use fewer, and you get a survival of the fittest thing.

So the basic problem is that bad actors exist within a sufficiently large organization, and the bureaucratic solution incurs the cost of making everybody into bad actors, but with the benefit of limiting the badness of their action by top-down accountability. It is also somewhat bounded in how much it wastes: non-bad-actors who really don’t need their big budgets do have a weak vested interest in allowing it to be cut, as it frees up resources for the organization as a whole and this can improve their job stability, year-end bonus, etc.; also business units that really are not pulling their weight can be reorganized over long time scales. That is bureaucracy in a nutshell, the natural top-down solution.

To solve the corruption problem with a bottom-up approach requires connecting individual interests to organization interests, so that in a game-theoretic sense there are no bad actors (albeit there may be irrational ones who want to hurt themselves in order to hurt society). This is a really hard problem in accounting. The basic thing that you want to do is to make sure that everyone gets paid some baseline amount, plus some proportion of “what they make for the company.” In cases where this is really easy to determine, nobody does it any other way. Salespeople get commissions, and they get them fairly universally. This solves any corruption problem bottom-up. [It is also 100% transparent: “Why did she earn more than you? Because she sold more than you”—top-down budgets are frequently confidential wherever possible due to the risk of one subunit (could be larger than an employee) discovering that another subunit which “does less” in whatever sense gets more of the pie.]

The problem is, we occupy complicated systems and it is not easy to determine how much the organization’s bottom line will be impacted by the loss of a particular individual. What is the “commission” that I should be paying to a janitor? Am I supposed to pay developers money for completing “story points?” And how do I do that without creating a toxic atmosphere where everybody wants to overestimate the number of story points in a task—how do I objectively measure those story points in terms of the hard cashy business value they create? What about managers or recruiters; how do I reward you for the business value of the people you managed/recruited?

Without solving this sort of hard problem, you can’t guarantee that when someone uses nepotism to the organization’s great loss, that they don’t feel the full brunt of that loss and therefore have a selfish incentive to be fair in the first place.