Money shortages create all manner of frictions.

Say you have a large group that believes, with varying levels of truth, that it's underpaid. Say that there's nowhere near enough money floating around to bring the entire group up to the level it wants. (Not that this ever happens, but bear with me.) Barring a visit from the Money Fairy, or a really drastic, from-the-ground-up restructuring in which absolutely everything is on the table, you basically have four choices:

1. Just say no, and leave it at that. Let the discontent and/or injustice continue.
2. Take what little money you have, and divide it across the board. Hope that a truly paltry increase will paper things over.
3. Take what little money you have, and ask the group to propose a fair internal division.
4. Do triage, and target relatively meaningful increases to the worst cases. Leave the rest alone.

The astute reader will notice quickly that all four options suck.

The first option is the easiest, and in some ways, the best. It's evenhanded, it saves money, and it's easy to implement. Its only drawback is that it completely ignores the underlying problem.

The second option is also evenhanded, and also easy to implement. But it manages to both burn your budget and leave the problem unsolved, making it the worst of both worlds.

The third option looks good on paper, but flops in practice. In the absence of a union, how do you even know who to talk to? Who gets to represent whom? And with a union, the first move will always be "the total number has to be much bigger." Well, yes, and I'd love to have washboard abs. Not. Gonna. Happen. (The second move is usually "that's your problem," thereby defeating the purpose of the conversation.) It's theoretically possible that this could work. In the settings I've seen, nope.

The fourth option sometimes wins by default, but that doesn't make it right. The impulse to craft a political compromise will quickly crash headfirst into legalistic arguments from parity and precedent. "How come they got increases and we didn't? How was this decision made?" Suddenly, you're defending your right to even make a decision. And heaven help you if the protected classes aren't evenly distributed between groups. Allegations of favoritism – and worse – will fly from the first day.

And anything like this will be the hardest, by far, to implement, since you'll need documentation for criteria, verification, application, etc. You'll annoy nearly everybody, make a bunch of new enemies, and put the beneficiaries in an awkward position. Although in some ways it's the 'right' answer, it's also the most dangerous and internally costly.

The shame of it is that the fourth option actually comes closest to fairness, assuming that the decision-maker is accountable. Options one and two don't even try to address fairness. Option three sounds fair, but in practice it usually rewards the squeakiest wheels, rather than the most deserving.

The tension between legal arguments and political ones is chronic and painful. Legal arguments are based on uniformity, process, and elephantine memory. Political arguments are based on what can actually be done at the moment. With infinite resources, it's possible to satisfy both; simply raise the floor as high as it takes to stop the complaining. But with strapped resources, you have to make a choice.

So a quick survey of my wise and worldly readers: which option would you choose?