Are the tech hipsters disproportionately in the areas already covered by AT&T? It seems plausible, if Apple and AT&T were targeting similar demographics that Apple would choose AT&T as its first partner. But it's on the edge of weird correlations where the burden of proof starts to fall on the person disputing that the study is unbiased, rather than on the researchers to rule out every possible correlation.

Alternatively, maybe you are arguing that even if iPhones caused a decline in birth rate among tech hipsters, that doesn't transfer to the population at large. This is both less believable and less valuable as a criticism of the study: even if the result only holds in one demographic group it's still an interesting finding.