I always say, never underestimate the power of Spite.
This method reveals that the majority of individuals exhibit consistent (non-)spitefulness over time and that the distribution of spitefulness is bipolar: when choosing whether to be spiteful, most individuals either avoid spite altogether or impose the maximum possible harm on their unwitting victims.
To the extent that this holds up, it would be quite useful in a predictive sense.
The problem with this sentence is that "evil" can mean so many different things -- some of them fit well to some people, some of them are strawman versions.
We need to be more specific. Otherwise it's easy to get into a "reversed stupidity" pattern, where we notice that the strawman version of the argument is obviously wrong, and therefore... no people are morally worse (e.g. more likely to cause others harm, even without personal benefit) than others.
Imposing "maximum possible harm on their unwitting victims" sounds pretty unambiguously evil. If you can come up with a reasonable definition that disagrees, feel try to type it out.
Here is a rather curious paper describing psychology researchers' attempts to investigate "spitefulness" - I think they define spitefulness roughly as "hurting others without any benefit to oneself".
Does this include the subset "vindictiveness"? That is, "hurting others without any benefit to oneself when done against those who have previously hurt that which you care about". I endorse that subset as respectable, sometimes practical and even sometimes virtuous. Sometimes it is even heroic.
As an example, if I am playing a strategy game and an ally breaks an alliance in violation of whatever explicitly specified termination conditions we had arranged I will be 'spiteful'. That is, I will assign some degree of value to doing harm to that enemy above and beyond whatever is useful for my future success. The spite (vindictiveness) may be overridden by other priorities if there is still a chance for me to win the game. Being the type of person who is vindictive in this manner (and easy to predict) can be useful even though at the time of the vindictiveness it can be too late to help.
Seems like game theory + signalling can provide an explanation.
In game theoretical sense, if you precommit to harm, even at cost to yourself, those who defected on you, this can be a good strategy because it make others less likely to defect on you.
But to make it work, you have to signal in advance that your precommitment is credible. Harming someone, even at cost to yourself, even if they didn't harm you before, could be such signal.
There is probably some optimal amount of this senseless harm in a given situation, because it costs you, because it makes you enemies, and because if you'd do it all the time to everyone, then the other players would have no reason to not defect on you.
But if you harm people regardless of whether they've harmed you first, it doesn't disincentive people from harming you. It only does if you're more likely to harm people who have harmed you than people who haven't.
But if you harm people regardless of whether they've harmed you first, it doesn't disincentive people from harming you.
Indeed, all else being equal it gives an incentive. Hurting you reduces your ability to do harm.
http://www.plosone.org/article/info:doi/10.1371/journal.pone.0041812
Here is a rather curious paper describing psychology researchers' attempts to investigate "spitefulness" - I think they define spitefulness roughly as "hurting others without any benefit to oneself". References the Stanford Prison Experiment. Concludes, more or less, that some people are spiteful, sometimes.
I have many reservations about the methodology used in this experiment (main one: not sure if the entire process really reflects any real-world motivations, and hence results might not mean much), but I thought it might be of interest to people on this site. Also, of the 30-odd references cited at the end of the paper some sound rather interesting and many are available online.