That's not true. Even in the hypothetical scenario that all humans reproduced exactly enough to replace themselves then died immediately (i.e. a completely constant human population), there would almost certainly be gradual change in the inheritable traits of the human population (i.e. evolution). If sexual partners are chosen, and those choices are inheritable traits (i.e. genes influence your choice of mate), that's textbook natural selection.

> even in the least developed areas of the world, this is no longer the general case.

A very large portion of the most successful pride of lions probably also breed, since they're not worrying as much about food or predators. Does that means the entire species is not evolving?

  > Natural selection only works if unfit individuals
  > are killed off prior to breeding.

  > since nearly everyone survives to breed.

Natural selection works strongly when group A dies before reproducing and group B does not. It still works, though more weakly, when group A has an average of 2.0 offspring and group B has an average of 2.1 offspring.

In some parts of the world, natural selection is still significant (the common example: sickle cell anemia drastically increases malaria survival). In many parts of the world, though, differential reproduction appears to be a result of culture / religion / personal choice; it's not well understood how much this relates to genetics, and therefore not clear how much it deserves to be called "natural selection".