One-night stands make uncooperative

To cooperate or not to cooperate is a question that every breeding species has to answer. From an evolutionary standpoint, breeding passes on half your genetic material, while your siblings will pass on a quarter of your genetic material. If life isn't too difficult, then it makes sense to do your own breeding and ignore your siblings—or, perhaps, try to outcompete them. But, if life is pretty hard—and lets face it, life is usually tough for most species—perhaps it makes sense in the long term to forgo having your own children and make sure that your relatives successfully raise as many children as possible.
This behavioral adaptation is a key concept in evolutionary theory, but it only works if parents don't cheat. One consequence of this idea is that there should be a correlation between species that breed cooperatively and the amount of cheating that goes on. But does the data support it?
To see if common wisdom was correct, a group of zoologists from Oxford examined the literature on birds, collecting papers that described the social habits of as many species as possible. They then coded the species as cooperative breeders or not, and used numerical data from the studies to codify how likely parents were to cheat on each other. As usual with population studies, the data was messy, but the vast majority of cooperative species were less likely to cheat, while most of the cheaters did not show signs of cooperative behavior.
So far so good. But there were a number of species that sat in the cooperative corner, yet still were very likely to cheat. These species indicate that things are more subtle than you'd expect based on the broadest application of the hypothesis. To untangle these relationships, the researchers looked at two things: kin recognition and evolutionary history.
Kin recognition describes the ability of a sibling to distinguish siblings from other nest mates. They found that, among the cheating species, kin recognition allowed helpers to devote their time to full brothers and sisters, although if there was enough cheating going on, this became impossible. The essential finding here is that kin recognition allows the evolutionary benefits of cooperation to be sustained in the face of a moderate degree of promiscuity.
Evolutionary history also provides some lessons. By looking at evolutionary relationships, the researchers showed that cooperative species that were also promiscuous were likely to stop cooperating over time. Promiscuous species could evolve from cooperative, non-promiscuous species as well. In short, cooperation has been gained and lost in multiple lineages, sometimes multiple times.
The conclusion is that cooperative behavior is stable in the face of some promiscuity, provided that siblings can use kin recognition to devote their resources to the right offspring. Otherwise, the benefits of cooperation are not sufficient for it to remain a stable trait.
How this applies closer to home is a bit fraught. For instance, humans are notably more flexible in their societal arrangements than birds. Furthermore, we are capable of rationalizing our way into and out of incredibly complicated sets of social rules. Add to this that nearly every species seems to exhibit some degree of promiscuity, and we can see that simplistic, sweeping conclusions are rather likely to be wrong.