Most charities worry about wasting donor money. The money is being given to them to help people, and any they spend on themselves is a waste. So they operate in small offices, don't pay well, and have their employees use slow and flaky donated computers. This is shortsighted thinking. They are making the employees they do have less effective, and there are good people who they will not be able to hire at all. If increasing administrative expenses from 15% of the budget to 30% doubles how effectively the remaining program funds are spent, they should do this.
Assuming charities are trying to maximize the amount of good they can do, why aren't they already doing this? A big part of the answer is the state of external charity evaluation. It's difficult to evaluate how much good a charity is doing, but the amount of money they spend on various things is public knowledge . So people look at the ratio of administrative expenses to program expenses in order to see how well a charity is spending the money that comes in. This is the logic that a charity that spends 15% of its money on overhead and 85% directly helping people is going to do more good than one that spends 30% and 70%. Unfortunately, this logic is flawed, because charities make tradeoffs between money spent on overhead and the effectiveness of their programs. This strong external pressure to maximize the fraction of money spent on programs is limiting the ability of charities to do what they're supposed to.
There are a few things I can think of to do about this. One is to use overall efficacy ratings instead of just financial efficacy ones. Another is to find a charity that you trust to do as much good with your money as they can, even if that entails paying to hire good people. And especially be wary of charities that trumpet extremely high financial efficacies; wonder what they've given up to get there and don't reward them for competing in a game that is hurting everyone. 
 In the US, charities file a form 990, which gives information on spending broken down in various ways. This is the data charity navigator works off of.
 Trumpeting high overall effectiveness ratings is good, though.