Reading your reference quickly, what stuck out to me is whether the prisoners have already existing better relationships with family prior to incarceration, leading to more regular communication while in prison, and greater integration with family and society after release resulting in lower recidivism. Those who aren't staying in touch may have gone into prison without ongoing family relationships, reduced connections to society, and later, increased risk of recidivism. Perhaps your link addressed that, I didn't notice it though.
I think the system is screwed, don't get me wrong, but the reference may be attributing a surface level effect (allowed to make calls) to a more latent thing (already have decent integration with family and society).
No direct answer but I think the argument is that no matter what restrictions that make communication difficult or intentionally unpleasant have a negative impact. Your question somewhat assumes that the communicate/no communicate is mostly a manifestation of prior familial relations rather than prison actions that prevent contact...i.e.,:
>For example, one female attorney said she was told by prison officials that she could not visit a prisoner because her underwire bra set off the metal detector. After leaving, removing her bra and then returning, she was told she could not visit because she wasn’t wearing a bra.
The possibility for latent effects obviously exists, but there are actions taken by the prisons that can only be seen as attempts to reduce contact, which we can see has a negative impact. Whether its a formative or reflective latent doesn't seem like something that should be a driving factor in decisions here to me.
I think the system is screwed, don't get me wrong, but the reference may be attributing a surface level effect (allowed to make calls) to a more latent thing (already have decent integration with family and society).