Looks like they just made up an explanation, "We of course have not directly tested whether social organization is indeed the intervening variable" > Instead,

Looks like they just made up an explanation, "We of course have not directly tested whether social organization is indeed the intervening variable" > Instead, the results point to the possibility that the large-scale conversion of housing units into short-term rentals undermines a neighborhood’s social organization, and in turn its natural ability of a neighborhood to counteract and discourage crime, specifically violent crime. Further, the lagged effects suggest a long-term erosion of the social organization, which would stand in contrast to the more immediate impacts that the presence of tourists would be expected to have. We of course have not directly tested whether social organization is indeed the intervening variable, but it seems clear that the issue is not the tourists themselves but something about how the extreme transience of a short-term rental unit fails to contribute to critical neighborhood social dynamics.