That's not always true at all and varies greatly from one sub-field to another. I have read countless top-conference papers where the authorship are obviously attributed in alphabetical order.