Utilitarianism lacks a measuring mechanism, as highlighted by Jesse's blog post. There is no way to quantify happiness, or pain for that matter. Pain can be physical, emotional, long term, short term. Happiness can be fleeting or permanent. Some derive pleasure from inflicting pain on others. If a sociopath kills a baby and gets as much happiness from that killing as the baby and family have net pain, then is the action truly immoral under that paradigm? While many claim this is an extreme example and we ought to do the best we can to get the best consequences, I cannot shake the qualm about selfishness. People do what is best for them. Getting a wife a birthday gift she likes yields a happier home, happier memories, and great times. But, getting her a gift she hates could yield less time together, an angry memory, and disappointment. But if one's goal is to get your wife to stop bothering you, then consequentially, you should get her a gift she absolutely hates.
We cannot know others intentions, nor does it matter for consequentialism. If I am maximizing my own utility and happiness by making others sad, and I deem these feelings worthwhile for myself, I can go about giving the worst gifts, starting forest fires, and punting puppies. I could even make a club for the sociopaths of the world to unite. If this club brings us happiness, who is to say its negative repercussions on the rest of the world have a net 'bad'?
Without a way to calculate happiness vs. sadness, there is not a surefire way to counter this claim. If I derive pleasure from other's pain, in a direct ratio, isn't hurting someone to better myself still moral because it is not increasing total pain (because of my increased pleasure)?