Executive Summary

This research examines the health of the local journalism ecosystems in three New Jersey communities: Newark, New Brunswick, and Morristown. The goal of this research is to develop and apply a set of reliable, scalable performance metrics intended to inform funders, policymakers, researchers, and industry professionals about the state of journalism in local communities and, ultimately, its connection to healthy democracy, and to help guide decision-making about possible areas of intervention.

This report begins by defining the contours of a local journalism ecosystem and distinguishing the notion of a journalism ecosystem from other commonly employed ecosystem concepts such as media ecosystem and communication ecosystem. This report then presents a three-level conceptual and methodological framework for assessing the health of local journalism ecosystems. This analytical framework focuses on Infrastructure (the availability of journalistic sources), Output (the quantity of journalistic output from these sources) and Performance (the extent to which this output is original, is about the local communication, and addresses critical information needs).

This report then applies this analytical framework to the three selected New Jersey communities through a content analysis of a one-week sample of news stories posted on the web and social media posts provided by the journalistic sources identified in each community. The results indicate substantial differences in the journalism infrastructure output and performance across these three communities, particularly when controlling for differences in population size. Across the majority of the measures of journalistic output and performance utilized, Newark ranked the lowest and Morristown ranked the highest, with New Brunswick consistently falling in the middle. Thus, for instance, Morristown possesses more than ten times as many local journalism sources per 10,000 capita than Newark. And, during the measurement period, Morristown journalism sources produced 23 times more news stories and 20 times more social media posts per 10,000 capita than Newark journalism sources, and 2.5 times more news stories and 3.4 times more social media posts per 10,000 capita than New Brunswick journalism sources. New Brunswick journalism sources produced 9.3 times more news stories and six times more social media posts per 10,000 capita than Newark journalism sources.

The production of stories addressing “critical information needs” (Friedland et al., 2012) per 10,000 capita was almost 35 times greater in Morristown than in Newark; 14 times greater in New Brunswick than in Newark and approximately 2.5 times greater in Morristown than New Brunswick. In terms of social media output from local journalism sources addressing critical information needs, the same disparities per 10,000 capita persisted, with Morristown journalism sources producing almost 33 times more social media posts addressing critical information needs than Newark sources and almost four times more than New Brunswick sources. New Brunswick journalism sources produced almost nine times more social media posts addressing critical information needs per 10,000 capita than Newark sources. Similar disparities emerged when focusing on content that is produced by the outlet and/or that is about the local community.

These disparities appear to be a function not only of differences in the relative quantity of journalistic sources and output across these three communities, but also of qualitative differences in this output. That is, a greater proportion of the journalistic output in Morristown was original, was about the local community, and addressed critical information needs than was the case in either New Brunswick or Newark, with Newark again lagging behind in terms of the extent to which its journalism output met these criteria.

This analysis also examined the concentration of journalism output across the three communities, in an effort to determine the extent to which the journalism output in these communities is being provided by relatively few sources. In order to accomplish this, the well-known Herfindahl-Hirschman Index (HHI) was applied to the shares of journalism output within each community provided by each of the journalism sources identified. Using this measure, New Brunswick tended to exhibit substantially higher concentration of journalism output across the various content categories, suggesting that within this community, journalistic output was distributed across a smaller range of sources than was the case in either Morristown or Newark. Newark tended to exhibit the lowest output concentration levels of the three communities, suggesting that one positive aspect of the journalistic output in Newark is that it is relatively more evenly distributed across available journalistic sources.

These findings potentially point to a specific type of problem in local journalism, one in which lower-income communities are dramatically underserved relative to wealthier communities, and in which lower-income communities receive the bulk of their news from a smaller range of sources. It would be very interesting to see, if this research design were to be scaled up and applied to a larger sample of communities, the extent to which these patterns persist. Further, at such a larger scale it would be possible to conduct multivariate analyses that could identify with greater specificity the characteristics of individual communities that are predictive of overall levels of journalistic infrastructure, output, and performance. It would also be possible to explore the ramifications of these disparities, in terms of their relationship to constructs that are fundamental to well-functioning local democracies such as voting behaviors and community engagement.