This section provides a basic overview of the study’s main findings. More detailed tabular presentations of the results can be found in Appendix F. Table 2 presents some general descriptive data about each of the three communities, such as population, number of journalism sources identified, number of sources per 10,000 capita, and the proportional participation of these sources on the two major social media platforms (Facebook and Twitter). These data speak to the relative health of the Infrastructure Layer of the local journalism ecosystems across these three communities.

Table 2: Descriptives

Town Population Per Capita Income # Journalism Sources Sources/10k Social Media Presence Score
Newark 277,00 $13,009 16 .58 80
New Brunswick 55,000 $16,395 13 2.36 81
Morristown 18,000 $37,573 11 6.11 68


One point worth noting in this table is the substantial variation in the number of sources per 10,000 capita across these three communities. As the table indicates, the smallest, wealthiest community (Morristown) has, proportionally, substantially more journalism sources than the largest, lowest-income community (Newark), with New Brunswick situated between the these two communities in terms of population, per capita income, and sources per 10,000 capita.[1] As we will see, this disproportionate availability of local journalism sources dramatically impacts the volume of journalism output across these three communities (see below).

Next, we look at the overall levels of journalism activity across the three communities. Here, we control for population size in order to have a relative sense of the quantity of journalism output being produced, both in overall terms and in terms of journalism meeting the various criteria – and combinations of criteria – discussed above (e.g., original, about community, meeting critical information needs). We look first at our analysis of stories available on the journalism sources’ home pages. We then turn to the journalistic output on social media platforms

[1] It is important to note that there are a number of radio stations licensed to the city of Newark, but many of these stations’ studios and transmission towers are based in New York City, and the stations essentially operate as New York City-focused radio stations. These stations (e.g., WQXR, WNSH, WHTZ) were not included in this analysis as local journalism sources for Newark.

Web Sites

Figure 2 depicts the differences in journalistic output across the three communities, with a focus on the news stories that were present on the home pages of the sources located within each community. This graph provides breakdowns across each individual coding category, as well as all combinations of coding categories. At the most basic level (the top category in the graph) – stories per 10,000 capita – we can see that Morristown journalistic sources presented nearly 200 stories per 10,000 capita in the sample week, compared with less than ten for Newark and approximately 80 for New Brunswick.

Figure 2: Journalistic Output Per 10,000 Capita Across Three NJ Communities (Web Sites)


As we work our way down the graph, we see that this pattern persists for each way in which the story output was filtered. Thus, for instance, Morristown journalism sources produced over 130 stories per 10,000 capita that were coded as Original, compared with just over 50 for New Brunswick and less than ten for Newark. At the very bottom of the graph, we focus on stories that met all three of the coding criteria (stories that were original, about the community, and that addressed a critical information need). When these filtering criteria are all applied, Morristown journalism sources produced 50 stories per 10,000 capita, compared with just over ten for New Brunswick and less than one for Newark.

Figure 3: “Quality” of Journalistic Output Across Three NJ Communities (Web Sites)


Another way to examine story output is in percentage terms. That is, what proportions of the stories being produced in these communities have met the various criteria? Figure 3 presents these results, showing the proportion of the stories available on the home pages of the journalism sources in each community that met each coding category (individually and in combination). As we can see in Figure 3, some of the patterns seen in Figure 2 persist, though not to the same extreme degree. Morristown journalism output tends to perform better on each of the evaluate criteria than Newark journalism output. New Brunswick journalism output approaches or exceeds that of Morristown in some instances (e.g., % original; % original and addressing critical information needs).
Starting at the top of the graph, for instance, the percentage of news stories produced by Morristown journalism outlets that was original approached 70 percent. In terms of originality, the proportion of news stories produced by New Brunswick journalism sources meeting this criterion was slightly higher (70%). For Newark the proportion was just under 60 percent. As Figure Three also indicates, while over 30 percent of the Morristown news stories analyzed were about the community and addressed critical information needs, this percentage was less than 20 percent for New Brunswick and just over ten percent for Newark.

Finally, we look at the concentration of the journalistic output found on the home pages for the local journalism sources. As Figure 4 indicates, New Brunswick exhibited consistently higher levels of output concentration than either Newark or Morristown across all of the content coding categories. So, for instance, New Brunswick’s HHI for web story output was 4559.18, compared with 2062.20 for Morristown and 1902.58 for Newark. The levels of output concentration in Newark and Morristown tend to be similar. These patterns suggest, compared to Morristown and Newark, a substantially larger proportion of the journalistic output in New Brunswick is produced by fewer sources.

Figure 4: Concentration of Website Stories Across Three NJ Communities

Social Media

We turn next to social media output. Figure 5 presents the same breakdown as Figure 2, with the focus this time on social media posts rather than stories available on the sources’ home pages. As should be clear from Figure 5, the same pattern that was found for home page output persists when we focus on the social media output of these journalism sources. The social media output of Morristown’s journalism sources far exceeds that of Newark and (to a lesser extent) New Brunswick across all of the coding categories, ranging from the broadest (posts per 10,000 capita) to the narrowest (original posts about the community addressing critical information needs per 10,000 capita). For instance, Morristown journalistic sources produced over 200 posts per 10,000 capita addressing critical information needs during the measurement period, compared with 60 for New Brunswick and less than ten for Newark.

Figure 5: Journalistic Output Per 10,000 Capita Across Three NJ Communities (Social Media)


To some extent this pattern persists (though is not as extreme) when we shift our analytical focus from social media posts per 10,000 capita to the proportion of social media posts meeting the various coding criteria. As can be seen in Figure 6, the journalistic sources in the three communities were roughly comparable in the extent to which their social media posts met the originality criteria (all in the 90 percent range). However, when additional criteria were applied to these postings (whether they were about the community, or addressed critical information needs), the Morristown – New Brunswick – Newark high-to-low pattern re-emerged.

Figure 6: “Quality” of Journalistic Output Across Three NJ Communities (Social Media)


Finally, we turn to journalistic output concentration in the social media context. As Figure 7 indicates, there is a substantial amount of variation across the communities in terms of their relative social media output concentration across the various coding categories. Thus, for instance, while Newark exhibits substantially higher output concentration than either Morristown or New Brunswick in terms of overall social media posts and in terms of original social media posts, when the focus is on posts about the community, or on any of the combinations of content categories, Newark’s output concentration is by far the lowest. New Brunswick tends to exhibit the highest levels of social media output concentration across all of these other categories.

Figure 7: Concentration of Social Media Output Across Three NJ Communities