Popular opinion, and every Liberal on the planet, would say "liberals", hands down. In fact there's one person in my life who has expressed surprise at how charitable I am for being such a conservative.
However, what if we frame the notion of the uncaring conservative as a hypothesis and propose tests? I suppose the first thing to do is define what we mean by "caring". How about defining it in terms of donating time, energy, and money to the needy? Here's a piece by Thomas Sowell:
A new book, titled "Who Really Cares" by Arthur C. Brooks examines the actual behavior of liberals and conservatives when it comes to donating their own time, money, or blood for the benefit of others. It is remarkable that beliefs on this subject should have become conventional, if not set in concrete, for decades before anyone bothered to check these beliefs against facts.
Some of the findings:
• People who identify themselves as conservatives donate money to charity more often than people who identify themselves as liberals. They donate more money and a higher percentage of their incomes.
It is not that conservatives have more money. Liberal families average 6 percent higher incomes than conservative families.
• ...most of the states that voted for John Kerry during the 2004 election donated a lower percentage of their incomes to charity than the states that voted for George W. Bush.
• Conservatives not only donate more money to charity than liberals do, conservatives volunteer more time as well. More conservatives than liberals also donate blood.
According to Professor Brooks: "If liberals and moderates gave blood at the same rate as conservatives, the blood supply of the United States would jump about 45 percent."
Professor Brooks admits that the facts he uncovered were the opposite of what he expected to find -- so much so that he went back and checked these facts again, to make sure there was no mistake.
Then there's the problem of accounting for the difference.
I see three basic possibilities.
First, maybe the act of identifying as a Liberal makes one less caring.
Second, we may be seeing the effect of an ideological "filter" which selects the less caring as members and the more caring as non-members.
The third possibility is that any measured relationship is purely due to coincidence. I think Brooks has probably looked at the problem in enough different ways that we have to assume the difference is real. If it were coincidence, we'd expect to see at least a few cases where the difference tipped the other way.
So if it's a real difference, why does it exist?
Sowell likes to examine differences between ideological groups through their basic vision of reality:
Fundamental differences in ideology go back to fundamental assumptions about human nature. Based on one set of assumptions, it makes perfect sense to be a liberal. Based on a different set of assumptions, it makes perfect sense to be a conservative.
The two visions are not completely symmetrical, however. For at least two centuries, the vision of the left has included a belief that those with that vision are morally superior, more caring and more compassionate.
In other words, identifying as Liberal flatters a person's self-esteem, and relieves him or her of the need to actually do anything estimable.