Mike Konczal wrote an excellent article for Democracy about the problems with a voluntary safety net and the superiority of government social insurance. The article draws on serious historical research (by other people) to prove two main points: first, there never was a Golden Age of purely voluntary charity; second, and more important, what charitable support mechanisms existed were not up to the challenges of the Second Industrial Revolution of the late nineteenth century and completely collapsed with the onset of the Great Depression.
This shouldn't come as a surprise. There are basic economic reasons why public social insurance is superior to voluntary charity. The goal here is to protect people against risk: of unemployment, of health emergency, of outliving one's savings, and so on. For a risk-mitigation scheme to work, there are a few things that are necessary. One is that people actually be covered. This is something you can never have with a private system (unless it's regulated to the point of being essentially public), since charities get to pick and choose whom they want to help. As Konczal says of private agencies before the Depression,
"They were also concerned they'd lose their ability to stigmatize—or to protect—various populations; by playing a role in determining who wasn't deserving of assistance, they could shield those they felt worthy of their support."
Another thing you want is the assurance that the system has the financial capacity to actually protect you in the event of a crisis. That's why you don't depend on your neighbors to rebuild your house if it burns down. Besides the fact that they may not like you, they probably don't have enough money—especially if you lose your house in a fire that burns down the entire neighborhood. As I've said many times before, there is no other entity in the country—and not really one in the world—with the financial capacity of the federal government. Even state governments scramble to cut benefits when push comes to shove, which is one reason why some states provide Medicaid coverage to almost no one.
We like to think that we are a nation of generous people who will help each other out, but that isn't really true. We do have a much larger charitable sector than other advanced economies, where the state shoulders more of the burden. But more than half of our total donations go to religious organizations, private schools, and medical organizations, with only 12 percent going to human services organizations. Some money does filter from other organizations to the poor, but at most you can get to one-third of the total. (The vast majority of my donations have gone to services for the poor, primarily legal services.) I've argued elsewhere that we should place limits on the tax deductibility of charitable contributions, which are effectively a way that rich people can force other taxpayers to contribute to their pet charities. But as long as we have this idealized picture of our charitable sector, it isn't going to happen.