Why I am always unlucky but you are always careless

From lost keys to failed interviews, we blame other people for their mishaps but never ourselves when we do the same. Why? Because assuming causes helps us to make sense of the world. 
When my wife can't find her keys, I assume it is because she is careless. When I can't find my keys I naturally put it down to bad luck. The curious thing is that she always assumes the opposite – that she's the one with the bad luck, and I'm the careless one.
When we observe other people we attribute their behaviour to their character rather than to their situation – my wife's carelessness means she loses her keys, your clumsiness means you trip over, his political opinions mean that he got into an argument.
When we think about things that happen to us the opposite holds. We downplay our own dispositions and emphasise the role of the situation. Bad luck leads to lost keys, a hidden bump causes trips, or a late train results in an unsuccessful job interview – it’s never anything to do with us.
This pattern is so common that psychologists have called it the fundamental attribution error. And there's a whole branch of psychology that investigates how we reason about causes for things called attribution theory. The fundamental attribution error is a good example of a quirk in the way we reason about causes, but it isn't the only one. Despite the name, it may not even be the most fundamental.
Seeking causes
Psychologists are interested in attribution of causation because it tells us important things about how the mind works. To illustrate this, imagine you see a man asleep under a tree, and a leaf fluttering down to land on his head. As the leaf touches his head he wakes up and shouts "Yikes". Anyone watching this scene would assume the man woke up because of the falling leaf.
But this simple statement is remarkably difficult to prove – you have no direct access to the cause, just the before (a leaf) and after ("Yikes"). We automatically assume the cause. We talk about it like it is a thing – somehow in the middle between the leaf and the man, but really it is just an assumption, not a thing. And indeed, some new information could come along and force us to reconsider our assumptions. We might find out later that a philosophically-minded ant had come along and, just at that minute, decided to bite the sleeping man's hand.
So our causes are assumptions, based on what we perceive but with an extra bit of imagination. They are necessary assumptions. Without looking for causes we would be stuck with a confusing picture of the world. Rather than say "the falling leaf caused the man to wake up", we have to take everything into account and say the following. "The leaf fell. The grass did the same as before. A bird flew between two trees one hundred and thirty yards away. I lost my keys. My Romanian aunt's clock in my Romanian aunt's house continued ticking (on and on and on). The man woke up."
Assuming causes in this way lets us make sense of the world. Not only is it easier to describe, the descriptions tell you how to make things happen (or avoid them – for instance, if you want the man to stay asleep next time, catch the leaf). In this way, attributions are psychological magic that help us control the future. No wonder psychologists find them interesting.
Built on sand
The fundamental attribution error is just a continuation of a wider pattern: we blame individuals for what happens to them because of the general psychological drive to find causes for things. We have an inherent tendency to pick out each other as causes; even from infancy, we pay more attention to things that move under their own steam, that act as if they have a purpose. The mystery is not that people become the focus of our reasoning about causes, but how we manage to identify any single cause in a world of infinite possible causes.
Even the way I described cause-seeking as an "inherent tendency" is part of this pattern. I have no direct access to what causes the results of experiments that have made me think this, just as I would have no direct access to what caused the man to wake as the leaf fell. I assume a thing, hidden, somehow, underneath the experiments – an inherent tendency for humans to identify each other as causes – which I then rely on to tell you what I'm thinking.
That thing might not exist, or might have a reality very different from how I describe it, but we are forced to rely on assumptions to make sense of the world, and these assumptions create a reality of causes and essences that seems solid, despite its uncertain foundation.
This all might sound overly philosophical, but once you are switched on to this tendency to invent essences you'll hear them everywhere. Generalisations or stereotypes such as "women can't do maths" or "Americans don't have a sense of humour" also rely on an invented essence of a sex, or of a nationality, a term that some psychologists have called ultimate attribution error. These views don't have a concrete existence. They are based in imagination, and are subject to all the psychological forces that are at play there. 
In more prosaic domestic moments, when it feels like such bad luck that I can't find my keys, yet my wife seems so careless when she can't find hers, I know I’m performing psychological magic. I’m observing the myriad events in the world and imagining things – my bad luck, her carelessness – which I use to explain the world with.  
With the knowledge that these explanations can only ever be built on sand, I know to be a bit more careful about how I use them.