Who’s Producing Local Journalism?

Assessing Journalistic Output Across Different

Outlet Types
JESSICA MAHONE, QUN WANG, PHILIP NAPOLI, MATTHEW WEBER, & KATIE
MCCOLLOUGH

AUGUST 2019

0
TABLE OF CONTENTS
Executive Summary...........................................................................................2

Introduction ........................................................................................................4

Method ...............................................................................................................5

Results .............................................................................................................10

Discussion........................................................................................................16

Endnotes ..........................................................................................................19

Appendix A: Media Databases/Directories Consulted ...................................23

Appendix B: Manual Search Protocol .............................................................24

Appendix C: Critical Information Needs Categories .......................................25

Appendix D: Descriptive Statistics ..................................................................27

Appendix E: Analyses of Combination Content Categories ...........................28

About the Authors ............................................................................................29

1
EXECUTIVE SUMMARY

The economic challenges facing local journalism and the associated declines in revenues
and newsroom staffs have generated great interest in understanding the composition and
dynamics of local news ecosystems. Much of this research has focused on case studies
of individual communities while other research has focused either on the content
produced by local news outlets in the face of these challenges or on the consumption of
local news by the American public.

However, despite what we know about the challenges faced by local journalism, the
content of local news outlets, and Americans’ preferences for local news sources, we
don’t know a great deal about how different types of outlets are serving the information
needs of their communities. This paper addresses this question through an analysis of
100 randomly selected communities across the U.S. Across these 100 communities, this
study analyzes over 16,000 stories provided by 663 local media outlets. For this analysis,
local media outlets fall into one of four categories (radio stations, TV stations,
newspapers, and online-only outlets). Each story in the sample was content analyzed to
determine whether the story was original, local, and addressed a critical information need.

To understand the journalistic performance of different outlet types, this study analyzes
each the story output of each outlet type relative to the outlet type’s numeric frequency.
Doing this allows us to assess each outlet type’s news production relative to that outlet
type’s prominence in the news ecosystem. To examine production in this way, ratios were
calculated comparing the share of total stories, original stories, local stories, and stories
addressing a critical information from each outlet type to each outlet type’s share of
outlets.

Key findings of this study include:

▪ Local newspapers significantly outperform local TV, radio, and online-only

outlets in news production, both in overall story output and in terms of stories
that are original, local, or address a critical information need. For instance:

o Local newspapers account for roughly 25 percent of the outlets in our

sample, but nearly 50 percent of the original news stories.

o Local newspapers account for nearly 60 percent of the Local news stories
in our sample – more than all of the other outlet types combined – despite
accounting for only 25 percent of the outlets in our sample.

o Local newspapers account for nearly 60 percent of the stories that meet all
three criteria (original, local, addresses a critical information need), with the
other outlet categories each accounting for only 10 to 15 percent of the
stories that meet all three criteria.

2
 ▪ Online-only media outlets remain a relatively small component of local media
ecosystems, accounting for about 10 percent of the local outlets in the sample
and generally producing only about 10 percent of the news stories in the sample,
across the various content categories (original, local, addresses a critical
information need).

o Online-only outlets do perform well in terms of the proportion of their story

output that addresses critical information needs (over 80 percent).

▪ Radio stations represent the most common type of local media outlet in our
sample, but generally are the weakest in terms of the extent to which their story
output is original, local, and addresses critical information needs.

Overall, these findings suggest that newspapers are the most important producers of local
news in terms of the volume of journalistic output being produced for local communities.
The relative paucity of online-only local media outlets, and the relatively limited (compared
with newspapers) journalistic output of these outlets suggest that online-only outlets have
yet to come close to matching local newspapers as significant sources of reporting that
is original, local, and addresses critical information needs.

These findings support the continued importance of public policy and philanthropic efforts
to support the viability of local newspapers. These findings also suggest that commercial
and philanthropic efforts to establish online-only outlets as comparable alternatives to
local newspapers remain far from this goal.

3
INTRODUCTION

In May 2019, GateHouse, one of the largest publishers in the United States, merged 50
of its weekly newspapers into 18 regional publications following its early nationwide
layoffs that were reported to affect about 200 journalistic positions.1 This is just one of the
most recent examples highlighting consolidation of local newspapers – and trend that has
contributed to the acceleration of the decline of local journalism in terms of readership,
personnel, and ad revenue. According to the University of North Carolina’s Center for
Innovation and Sustainability in Media, nearly 1,800 newspapers closed in the United
States between 2004 and 2018, leaving almost 200 counties with no newspaper at all
and half of the 3,143 counties with only one newspaper.2 Even in communities where
local newspapers are available, resources were cut to a point where there was minimal
local coverage — a situation that researchers have described as “news deserts.”3

The decline of local news outlets, even small newspapers, has a profound impact on
democratic development and public life at both the individual and community levels.4 For
example, acquisitions of and ownership changes at local TV stations and newspapers
led to the nationalization and polarization of political news, having real consequences in
citizens’ political knowledge, participation, and voting decisions.5 And the loss of
government monitoring due to local newspaper closures lead to increases in municipal
borrowing costs.6

The declines in local journalism and the associated impacts have led to a handful of policy
proposals at the state and federal levels. The New Jersey Civic Information Consortium
was established by state legislature in 2018 and funded in 2019.7 In Massachusetts,
House Bill 181 proposes a 17-member panel to research the status of local journalism in
underserved communities and review possible policy solutions. 8 In the U.S. House of
Representatives, legislators have proposed the Saving Local News Act to make it easier
for local news organizations to obtain nonprofit status and the Journalism Competition
and Preservation Act to allow news organizations to negotiate collectively with large tech
companies by creating a four-year exemption period from antitrust regulations. 9 None of
these proposals have been enacted; nevertheless, each of these legislative acts
highlights a different set of policy solutions to address change in the industry.

Both the economic challenges facing local journalism as well as the policy-related efforts
addressing these challenges require systematic empirical data and thorough accounting
of local journalism to measure and assess the state of local journalism in the United
States. Much of the existing research into local news ecosystems has tended to examine
local journalism by focusing on a single community or a limited number of local
communities, such as Pew Research Center’s 2010 study of local news in Baltimore 10 or
their later study of the news ecosystem in Macon, Georgia, Boulder, Colorado, and Sioux
City, Iowa.11 This study analyzed local media outlets across 100 U.S. communities.

In an effort to inform these discussions about the future of local journalism, this study
focuses on the question of which types of media outlets are making the most significant

4
contributions to local news ecosystems. Specifically, this research focused on the
following questions:

◼ Are local newspapers, despite their economic hardships, still the most prominent
originators of local journalism? Previous research has suggested that this is the
case;12 however, a more recent evaluation is needed.

◼ Have digital native local news outlets emerged as significant contributors of local
reporting? This has been the hope of many observers of the evolution of local news
ecosystems, but systematic data have, to this point, been sparse.

This study builds on previous research that provided a comprehensive analysis of local
journalistic output on a large scale, analyzing 100 randomly selected U.S. communities.13
While that study focused on overall levels of journalistic output, and the question of how
community characteristics factor into the health of local news ecosystems, this study
focuses on the types of outlets available in these communities; and how these different
types of outlets (TV, radio, newspaper, online-only) differ in terms of their production of
news.

METHOD

Research on local journalism has tended to focus on large metropolitan areas.14 Our goal
was to look outside of this context, to the smaller, more vulnerable communities across
the U.S. First, our focus was on communities with populations ranging from 20,000 to
300,000 residents. Using U.S. Census data, these parameters resulted in a list of 493
communities. From this list, we selected a random sample of 100 communities. Then, we
verified that sample and approximated the population based on population, income,
demographic composition.

Identifying Local Media Outlets

Our next step was to generate a complete inventory of all of the local media outlets in
each of these communities. Thus, our search criteria included local newspapers, as well
as local radio stations, local television stations and local online-only news sources. Our
inventory of local news sources was limited to those sources geographically located within
each sampled community. Therefore, we excluded news sources that might produce
news of relevance to the community but that were geographically located outside of the
community. Clearly then, we employed a very strict geographic notion of local in our
process for identifying local media outlets.

This inventory was created through a systematic process of consulting multiple media
databases and directories – 11 in total (see Appendix A) and supplementing these
database and directory scans with a multi-stage online keyword search protocol (see
Appendix B). This multi-pronged approach reflects the fact that a comprehensive portrait
of the media outlets serving local communities today can only be achieved via combining
information from a broad array of sources. Even large-scale commercial media directories

5
(e.g., Cision) had coverage gaps when compared to the manual multi-database/directory
search process that we employed.

Although our search criteria included print, radio, television, and online media, the content
gathering and archiving was conducted exclusively online. Thus, the corresponding URL
for each media outlet’s home page was located and recorded for use in the archiving
process described below. In this way, the journalistic outputs of daily and weekly
newspapers, magazines, radio stations, television stations, and local cable channels all
were assessed via their online content offerings in the same way as outputs of online
news sources such as community journalism sites were assessed.

The research approach used in this work (which is a reflection of the effort to create a
realistically scalable methodology) runs counter to the common assertion that certain
types of legacy media (e.g., local weekly print publications, ethnic media outlets) remain
slow to utilize the internet as a means of disseminating their content. Our searches
indicated that we are now at a point in the evolution of legacy media at which this
generalization no longer holds true. The economic and strategic pressures and incentives
to have an online presence, combined with the inherent economic imperative to distribute
content production costs across as broad an audience base as possible,15 suggest that
the content available online can serve as a reliable indicator of the relative journalistic
output across individual outlets, regardless of their “native” platform.

The key term here is indicator: we are not seeking to produce a comprehensive inventory
of journalistic output, only a set of indicators that are conceptually and methodologically
robust and that can be employed in comparative analyses across communities or over
time. In support of this position, the data gathered on the 100 selected communities show
that only 37 out of 791 (5%) media outlets that did not have a corresponding online
presence, with the bulk of these being low-power and translator radio stations. 16

Categorizing Outlets

For this analysis, each outlet identified was placed into one of four broad outlet type
categories: 1) newspapers; 2) television stations; 3) radio stations; and 4) online only
outlets. Categorization was accomplished through a combination of online research (e.g.,
consulting outlet home pages) and consultation with the relevant media databases
described in Appendix A.

Content Archiving

Our content data gathering relied on a partnership with the Internet Archive. 17 We used
the Archive-IT community archiving platform to create our own web archive, as the locally-
oriented web sites that are central to this study generally are not part of the Internet
Archive’s routine archiving work. Moreover, custom archiving allowed the team to specify
the frequency of crawling, further enhancing the reliability of the data. The Archive-IT
platform allows users to specify the websites for collection and to set parameters including
the frequency and depth of crawling. In this case, we had 733 distinct URLs that we

6
identified for content archiving.18 In cases where a media outlet’s content was behind a
paywall, subscriptions to those sites were obtained.

In terms of the depth of the content crawl, we focused on the home-page coverage of
each media outlet and crawled to a depth of one. This means that in addition to archiving
the home page, we archived each web page that was one hyperlink (or one click) from
the home page. This approach draws from the premise in the journalism studies literature
that news source front pages/home pages represent a meaningful indicator of the most
important news events and issues affecting a community and thus represent a useful
means of assessing media outlet performance.19

The last methodological question to address was the sampling used to create the web
archive. For this project, we created a constructed week sample, selecting a Monday,
Tuesday, Wednesday, Thursday, Friday, Saturday and Sunday at random over a two-
month period during the summer of 2016. Websites were crawled on July 27, August 2,
August 6, August 28, September 9, September 21 and September 26 of 2016.

Using this methodology, we created a robust record of the home-page presentation of

local news in 100 communities. The web crawl across the sample week generated a
collection containing 1.6 million documents (html files, pdfs, images, audio files, etc.), 2.2
terabytes of total data, and, an archive of over 16,000 news stories. 20

Content Analysis

The primary goal of our content analysis was to determine the extent to which journalistic
output was serving the information needs of local communities. To do this in a way that
could be reliably and efficiently accomplished at the scale at which we were operating,
we focused on three criteria:

1. Whether the story was original;

2. Whether the story was about the local community;

3. Whether the story addressed a critical information need.

These criteria can be seen as fairly superficial – but fundamental – indicators of the
complex notion of the “quality” of the journalism being produced within communities. 21
These criteria provide a relatively simple, economical, straightforward, and replicable set
of indicators of journalistic performance that address the fundamental concern about
whether journalistic sources are addressing communities’ information needs.

The notion of critical information needs has been central to the ongoing discourse about
the performance of local journalism. 22 This approach draws upon this discourse, and the
research it has inspired.23 Specifically, each story was content analyzed to determine
whether it fit into one or more of the critical information needs categories identified in the

7
literature review by Friedland et al. that was prepared for the Federal Communications
Commission.24 Friedland et al. provide eight categories of critical information needs.
These categories are as follows:25

1. Emergencies and risks

2. Health
3. Education
4. Transportation systems
5. Environment and planning
6. Economic development
7. Civic information
8. Political life.

Descriptions for each of these categories can be found in Appendix C. Stories that did not
address one of these categories were coded as a 9. This category is intended to capture
less substantive story types (celebrity news, sports reports, etc.). For the purpose of the
analyses presented here, categories 1-8 are collapsed in order to create a dichotomous
variable (i.e., story does/does not address a critical information need) that reflects the
frequently-employed distinction between “hard” and “soft” news.

Each story also was coded for whether it was original (i.e., produced by the media outlet
rather than reprinted, linked, retweeted, or shared from elsewhere) and whether it was
about the local community.

The emphasis here on original stories is intended to separate aggregation, linking,

sharing, retweeting, and re-publication activities from original reporting. 26 This originality
measure seeks to discern the robustness of local journalism sources by determining how
active they are in producing news stories. A story was considered “original” if it included
a byline by an outlet’s reporter as well as if it had no indicators that it was a reposting of,
or hyperlink to, content produced elsewhere.

The emphasis on locality is employed to capture the extent to which the output of local
journalism sources focuses on the local community. This measure is intended to address
the extent to which local journalism is truly local, providing community members with news
and information about, and directly relevant to, their communities. This measure reflects
the long-standing localism principle, which has featured prominently in democratic theory
perspectives on media, and in media policymaking. From this perspective, the extent to
which citizens are able to engage in informed democratic participation in their
communities is a function of the availability of local news and information about their
communities.27 In order to assess locality, we opted for a strict geographic definition of
community, where we identified an item as about the community only if the subject was
an issue/event oriented around the specific municipality.

8
Together, these three variables reflect some of the primary concerns raised in ongoing
critiques and analyses of the state of local journalism today:

1. That the economic pressures on local journalism create overwhelming incentives

to aggregate and repurpose existing content rather than engage in original
reporting28

2. That the changing technological and economic dynamics for news distribution and
consumption are exacerbating the extent to which large-market or out-of-market
news can infiltrate local communities, thereby undermining local journalism 29

3. That the increasing challenges associated with attracting and retaining an

audience for news are compelling local news outlets to neglect substantive topics
in favor of an emphasis on “soft” news, celebrity, and sensationalism. 30

For these reasons, we think these particular variables of focus represent a useful set of
top-level indicators of how well local journalism is fulfilling its central purpose of facilitating
informed participation and engagement in local community affairs.

Content analysis of the online news stories was conducted by a team of trained coders.31
From these data, four measures were constructed:

1. Stories: (the total number of news stories produced for the community)

2. Original: (the total number of original news stories produced for the community)

3. Local: (the total number of local news stories produced for the community)

4. CIN: (the total number of news stories addressing a critical information need
produced for the community)

In addition, measures were created for each combination of measures 2-4 (i.e., counts of
stories meeting two or more of the robustness criteria), leading to:

5. Original + Local

6. Original + CIN

7. Local + CIN

9
 8. Original + Local + CIN

Community Characteristic Variables

In order to get a sense of the characteristics of the communities in our sample, a number
of municipality characteristics were gathered from 2010 census data. From the census
reports, we gathered data on the size of the population (POPULATION) in each
community. We also gathered data on the population density (DENSITY) of each
community, median household income (INCOME), the percentage of the population that
is African-American (AA%), and the percentage of the population that is Hispanic/Latino
(HL%). Descriptive statistics for these community characteristics are presented in
Appendix D.

RESULTS

As a starting point for this analysis, Table 1 presents an overview of the journalistic output
for each outlet type, in terms of how story output fared across the various content
categories that we created (described above). So, for instance, in Table 1 we see that
almost 53 percent of the stories produced by local newspapers were original, compared
with 32 percent for TV and radio, and 48 percent for online-only sources.

When it comes to local stories, roughly 26 percent of the newspaper stories we analyzed
were local, compared with less than 9 percent for TV, less than eight percent for radio,
and roughly 25 percent for online-only outlets.

When it comes to stories that address a critical information need, 63 percent of newspaper
stories met this criterion, compared with 72 percent for TV, 50 percent for radio, and
almost 82 percent for online-only.

Table 1: Comparing Outlet Type Story Output

Original Local CIN O&L O&C L&C O&L&C

Newspaper Stories 52.92% 25.75% 63.38% 24.80% 33.32% 17.43% 16.71%
(n = 6,129)
TV Stories (n = 32.41% 8.70% 72.08% 8.06% 26.21% 7.47% 6.83%
2,521)
Radio Stories (n = 32.36% 7.48% 49.62% 6.29% 17.16% 4.46% 3.90%
8,002)
Online-Only Stories 47.65% 24.82% 81.57% 23.86% 36.20% 19.97% 19.24%
(n = 1,697)

The remainder of Table 1 presents the proportion of stories from each outlet type that met
multiple criteria (i.e., Original and Local [O&L]; Original and addresses a Critical
Information Need [O&C]; Local and addressed a Critical Information Need [L&C]; and
stories that met all three criteria [O&L&C]). So, for instance, we can see that newspapers
(16.71 percent) and online-only outlets (19.24 percent) performed best in terms of

10
producing stories that met all three criteria; with television stations (6.83 percent) and
radio stations (3.90 percent) performing substantially worse.

As the numbers indicate, generally, newspapers and online-only outlets perform best
when looked at based on serving critical information needs. This is not surprising given
that, unlike TV and radio stations, newspapers and online-only outlets are generally more
narrowly focused on providing news. Local television stations and radio stations (and their
associated home pages), on the other hand, tend to have broader – or (in the case of
radio stations focused on specific music genres) different areas of focus.

In addition, relatively few communities of the size analyzed in this study have local
television stations. And for those communities that do have local television stations, the
low percentage of stories that are local in focus indicates that these stations are not
particularly focused on providing news coverage of their communities of license.
Rather, local television stations are more broadly focused in their news coverage from a
geographical standpoint.

Nonetheless, this analytical approach gives a sense of how the different outlet types
compare in terms of the composition of their story output. What this analysis does not tell
us, however, is the relative magnitude of the journalistic contribution that each outlet type
is making to their local communities.

To address this question, we next we examined each outlet type’s contribution in terms
of the share of the stories that they account for. The first component our findings is
presented in Table 2. So, for instance, in column 3 of Table 2, we see that newspapers
account for just over 36 percent of the news stories produced across the 100
communities; TV stations account for 15 percent; radio 40 percent; and online-only outlets
eight percent.

These percentages give us a sense of the relative prominence of each outlet type in terms
of the journalistic output the outlet provides. We see, for instance, from these numbers
that newspapers and radio stations are providing the bulk of the journalistic output in the
communities we studied. Online-only outlets provide a relatively small proportion of the
totality of news reporting available.

Table 2: Share of Journalistic Output by Outlet Type: Total Stories

% of outlets (N = 663) % of stories (N = 16,763)

Newspaper 25.27% 36.52%
TV 13.62% 15.08%
Radio 51.14% 40.25%
Online only 9.99% 8.16%

However, these percentages don’t take into consideration variation in the number of
outlets of different types available across the communities analyzed. As column two of
Table 2 indicates, there are significant differences across outlet types in terms of their

11
prominence in our sample. As the table indicates, radio stations account for over 50
percent of the local media outlets in our sample, followed by newspapers (27 percent);
TV stations (almost 15 percent); and then finally online-only outlets (10 percent).

We can get a better sense of the journalistic performance of different outlet types by taking
into account their story output relative to each outlet type’s numeric frequency. Doing this
allows us to assess each outlet type’s news production relative to that outlet type’s
prominence in the news ecosystem.

To examine production in this way, ratios were calculated comparing the share of content
to the share of outlets by category. And so, returning to the total number of stories, Figure
1 illustrates that radio has an output ratio of 0.79. This ratio of less than 1 means that
radio produces a share of the total news output that is substantially less than radio’s share
of the total number of outlets in our sample. In fact, radio produces fewer news stories
relative to its proportion of outlets than does any other outlet type in our sample, followed
by online news sources with a ratio of 0.82. Newspapers, with a ratio of 1.45, produce
more content relative to their share of outlets than any other outlet type; while TV stations
produce stories nearly even to their share of outlets, with a ratio of 1.11.

Figure 1: Ratio of Stories to Outlets: Total Stories.

We continue this analytical approach of looking first at total output, and then at relative
output, for each of the three primary content criteria that we applied.

12
First, we look at original stories (i.e., stories actually produced by the media outlet, rather
than repurposed from elsewhere). As Table 3 indicates, a similar pattern emerges with
original stories that we saw with total stories. As we can see in Table 3, newspapers
produced nearly half of original stories in our sample, even though newspapers make up
roughly a quarter of the outlets in our sample. As Figure 2 indicates, this results in a ratio
of 1.86 for newspapers, meaning that newspapers are significantly overproducing original
stories relative to their share of the outlets in our sample.

In what will prove to be a recurring theme, online-only news outlets produced original
stories roughly in proportion with the prominence of online-only outlets in our sample
(about 10 percent) (see Table 3); for a ratio of .95 (see Figure 2). Both TV and radio
stations underperformed in the production of original stories relative to their share of
outlets. TV stations produced a slightly lower share of original stories (about 12 percent)
than their share of outlets (almost 14 percent). Radio, with a ratio of 0.62, produced just
over 30% of the original stories in our sample, despite accounting for half of the outlets in
the sample.

Table 3: Share of Journalistic Output by Outlet Type: Original Stories.

% of outlets (N = 663) % of original stories (N = 6,949)

Newspaper 25.27% 46.99%
TV 13.62% 11.88%
Radio 51.14% 31.67%
Online only 9.99% 9.45%

13
Figure 2: Ratio of Stories to Outlets: Original Stories.

Focusing next on local stories (i.e., stories that are about the local community),
newspapers and online news sources perform even stronger in terms of local content
while TV and radio stations perform worse. In fact, the share of local stories produced by
newspapers (almost 60 percent) was more than double their share of outlets, resulting in
a ratio of 2.36. As Table 4 indicates, newspapers produced more local stories in our
sample than all of the other outlet types combined.

Online news outlets produced a slightly larger share of local stories than their share of
outlets (10 percent v. 13 percent), for a ratio of 1.29, continuing the trend of story output
being roughly proportional to outlet frequency. Finally, although radio stations make up
roughly half of the outlets in our sample, radio stations produced only 19% of local stories,
resulting in a ratio of 0.37 (see Figure 3).

Table 4: Share of Journalistic Output by Outlet Type: Local Stories.

% of outlets (N = 663) % of local stories (N =

2,646)
Newspaper 25.27% 59.7%
TV 13.62% 8.33%
Radio 51.14% 19.13%
Online only 9.99% 12.85%

14
Figure 3: Ratio of Stories to Outlets: Local Stories.

Turning next to stories that addressed a critical information need (CIN), newspapers’ story
output far exceeded their outlet frequency, producing 38 percent of the CIN stories while
accounting for 25 percent of the outlets (see Table 5), for a ratio of 1.51 (see Figure 4).
TV stations performed comparably, accounting for almost 18 percent of CIN stories
against roughly 15 percent of outlets, for a ratio of 1.32. Online-only news sources, with
a ratio of 1.1, provide a share of CIN stories that is roughly equal to their share of local
media outlets. Radio stations are the only outlet type that makes up a smaller share of
CIN stories than their share of outlets, with a ratio of 0.64 (see Figure 4).

Table 5: Share of Journalistic Output by Outlet Type: CIN Stories.

% of outlets (N = 663) % of CIN stories (N =

10,927)
Newspaper 25.27% 38.17%
TV 13.62% 17.92%
Radio 51.14% 32.94%
Online only 9.99% 10.97%

15
Figure 4: Ratio of Stories to Outlets: CIN Stories.

With this approach, we are also able to focus more narrowly on stories that meet multiple
criteria. We have included most of these in Appendix E. However, below we present the
distribution of the production of stories that meet all three criteria (original, local, and
addresses a critical information need). These stories, it could be argued, represent the
highest quality journalism in our sample.

When we focus on stories that meet all three of these criteria, we see that newspaper
performance far exceeds that of other outlet categories (see Table 6). Newspapers
produced nearly 60% of the stories meeting these criteria, while accounting for roughly
25 percent of outlets, for a ratio of 2.35 (see Figure 4). TV and online-only outlets had
ratios at or near 1. Radio, once again, performed the worst, with a .3 ratio.

Table 6: Share of Journalistic Output by Outlet Type: Original/Local/CIN Stories.

% of outlets (N = 663) % of O, L, & CIN stories (N =

1,726)
Newspaper 25.27% 59.41%
TV 13.62% 15.28%
Radio 51.14% 15.28%
Online only 9.99% 10.03%

16
Figure 5: Ratio of Stories to Outlets: Original/Local/CIN Stories.

DISCUSSION

Newspapers remain most important source of local journalism in communities

As this analysis makes clear, as much as newspapers may represent a dwindling
presence in the local news ecosystem of many communities, they still provide nearly half
of the original reporting to be found in our sample; more than half of the locally oriented
reporting in our sample, and almost 40% of the stories that addressed a critical
information need. Newspapers provide the largest proportion of the stories meeting these
criteria amongst the four categories of outlet types that we analyzed.

These numbers take on added significance when we take into account the fact that
newspapers accounted for only 25% of the media outlets located within the 100
communities that we analyzed. Essentially, newspapers provide a share of the reporting
meeting our analytical criteria that was 1.5x to 2.5x larger than newspapers’ share of the
total outlets in our sample.

In combination, these numbers suggest that newspapers remain by far the most important
source of local journalism serving local communities.

This finding is magnified when we look at the performance of other types of local media
outlets. Radio, for instance, consistently underperformed according to our metrics. That
is, while radio stations are quite numerous in the communities we analyzed, the reporting
found on their home pages seldom met the original, local, and critical information need
criteria that we applied.

17
Online-only outlets are still a small presence in local news ecosystems
More important, however, may be the findings related to online-only media outlets. One
of the key points of contention within the journalism field is whether, or to what extent,
newer online news sources have been able to meaningfully compensate for the declines
that have been affecting local newspapers. The data presented here suggest that online-
only news outlets remain a relatively small presence in local news ecosystems,
accounting for about ten percent of the outlets in our sample.

In addition, these online-only outlets generally provide news output that is in proportion
with their share of the total outlets in the community. That is, online only news outlets
represent about 10 percent of the media outlets in the 100 communities that we analyzed
– and produced about 10 percent of the stories in our data set. When we focused on
specific story criteria (original, local, critical information needs), online only news outlets
generally accounted for about 10 to 15 percent of the stories that met these criteria.

There are a number of key takeaways from these findings regarding online-only outlets.
The first, obviously, is that online-only outlets, from a numerical standpoint, still represent
a fairly small chunk of the local media ecosystem. The second is that, unlike newspapers,
online-only media outlets do not over-perform relative to their frequency. They have not
taken on the characteristics of local newspapers in this regard, in terms of their overall
journalistic productivity. This was consistently the case across all of the evaluate criteria
that we applied to each story. The journalistic output for online-only news sources remains
quite modest in relation to the journalistic output being provided by local newspapers.

It would be particularly useful to be able to place these findings into a larger historical
context. That is, how different might these numbers have looked five years ago or ten
years ago? How might they look five years down the line? While archival data are not
available to replicate this analysis, moving forward it is possible to apply our
methodological approach in order to track future trends.

In terms of the broader implications of these findings, the results of this research provide
compelling support for public policy and philanthropic initiatives directed at preserving the
viability of local newspapers in the face of the economic hardships that they are enduring.
These findings would also seem to suggest that all of the commercial and philanthropic
efforts and resources that have, to this point, been directed at developing online-only
news sources that can compensate for the declines of local newspapers have a long way
to go before these outlets are performing in a way that is comparable to local newspapers.

18
ENDNOTES

1 Meyer, P., & Strattan, L. (2019, May 31). Merging publications – internal communication
– confidential, Memo to GateHouse New England Staff, Retrieved August 15, 2019, from
https://drive.google.com/file/d/0B-
rEQi3pqA3jVDZpUmZNTDBDMnkxNXpkNDJVV0FPZHkzLXJZ/view.
2 Abernathy, P.M. (2018). The Expanding News Desert. Center for Innovation and
Sustainability in Local Media, University of North Carolina at Chapel Hill School of Media
and Journalism, Retrieved August 15, 2019, from https://www.usnewsdeserts.com/.
3 Abernathy, P.M. (2016). The Rise of a New Media Baron and the Emerging Threat of
News Deserts, Center for innovation and Sustainability in Local Media, University of
North Carolina at Chapel Hill School of Media and Journalism, Retrieved August 15,
2019, from https://www.usnewsdeserts.com/reports/rise-new-media-baron/executive-
summary/; Ferrier, M., Sinha, G., & Outrich, M. (2016). Media deserts: Monitoring the
changing media ecosystem. In M. Lloyd & L.A. Friedland (Eds.), The Communication
Crisis in America, and How to Fix it (pp. 215-232). New York: Springer.
4 Schulhofer-Wohl, S., & Garrido, M. (2009). Do newspapers matter? Evidence from the
closure of the Cincinnati Post. Discussion papers in economics, Princeton University,
Woodrow Wilson School of Public and International Affairs.
5 Shaker, L. (2014). Dead newspapers and citizens’ civic engagement, Political
Communication, 31(1), 131-148; Darr, J.P., Hitt, M.P., & Dunaway, J.L. (2018).
Newspaper closures polarize voting behavior, Journal of Communication, 68(6), 1007-
1028; Hayes, D., & Lawless, J.L. (2018). The decline of local news and its effects: New
evidence from longitudinal data, Journal of Politics, 80(1), DOI: 10.1086/694105; Martin,
G.J., & McCrain, J. (2019). Local news and national politics, American Political Science
Review, 113(2), DOI: 10.1017/S0003055418000965.
6 Gao, P., Lee, C., & Murphy, D. (2019). Financing dies in darkness? The impact of
newspaper closures on public finance, Journal of Financial Economics, DOI:
10.16/j.fineco.2019.06.003.
7 Karr, T. (2019, June 20). Free Press Action and New Jersey Policy Perspective urge
Gov. Murphy to support $2 million in funding local journalism for local news.
FreePress.net, Retrieved August 15, 2019, from
https://www.freepress.net/news/press-releases/free-press-action-and-new-jersey-
policy-perspective-urge-gov-murphy-support-2.
8 House No. 181. (2019, January 17). The Commonwealth of Massachusetts, House
Docket No. 2360.
9 Office of Congressman Mark DeSaulnier. (2019, June 6). Congressman DeSaulnier
introduces legislation to eliminate hurdles for newspapers to become non-profits,
Retrieved August 15, 2019, from https://desaulnier.house.gov/media-center/press-
releases/congressman-desaulnier-introduces-legislation-eliminate-hurdles;
McCormick, A. (2019, June 13). Two new bills in Congress propose a helping hand for

19
 a hurting industry, Columbia Journalism Review, Retrieved August 15, 2019, from
https://www.cjr.org/politics/congress-bills-support-local-news.php.
10 Pew Research Center (2010). How news happens: A study of the news ecosystem of
one American city. Retrieved August 15, 2019, from
https://www.journalism.org/2010/01/11/how-news-happens/.
11
Pew Research Center (2015). Local news in a digital age. Retrieved August 15, 2019,
from https://www.journalism.org/2015/03/05/local-news-in-a-digital-age/.
12 See Pew Research Center (2010).
13 Napoli, P.M., Weber, M., McCollough, K., & Wang, Q. (2018). Assessing local
journalism: News deserts, journalism divides, and the determinants of the robustness of
local news. DeWitt Wallace Center for Media & Democracy. Retrieved August 15, 2019,
from https://dewitt.sanford.duke.edu/wp-content/uploads/2018/08/Assessing-Local-
Journalism_100-Communities.pdf.
14 See Ali, C., Radcliffe, D., Schmidt, T.R., Donald, R. (in press). Searching for
Sheboygans: On the future of small market newspapers. Journalism (describing a
“knowledge gap” in the study of small market news organizations).
15 Hamilton, J. (2004). All the News that’s Fit to Sell: How the Market Transforms
Information into News. Princeton, NJ: Princeton University Press.
16 Translator stations generally just relay the signal of a larger nearby station. Thus any
news provided by such stations is already accounted for in our capturing of the news
output of the “parent” station. Low power stations may be more likely to be producing
journalism that is not captured by our online archiving protocol.
17 https://archive.org/.
18 It is worth noting that, in some instances, more than one outlet shared the same URL.
This typically occurred in relation to radio stations. Because radio stations often are part
of large ownership groups, it was sometimes the case that an individual radio station’s
home page would be (or would redirect to) the home page for the entire station group.
In such instances, the home page was analyzed each time it arose in connection with
an individual community. It is also important to note that some of the sites that were
archived produced no stories during the sampled days, which is why our content
analysis encompasses 663 total outlets.
19 For a more detailed discussion, see Napoli, et al. (2018).
20 The complete archive is available online at https://archive-it.org/collections/7520. Note
that the actual archive today is larger than described in the text. The archive was created
with specific boundaries for the purposes of this research, but the Internet Archive
decided to continue to data collection effort on a monthly basis to maintain the record
of local news websites.
20
21 Lacy, S., & Rosenstiel, T. (2015). Defining and measuring quality journalism. Report
from the News Measures Research Project, http://mpii.rutgers.edu/wp-
content/uploads/sites/129/2015/04/Defining-and-Measuring-Quality-Journalism.pdf.
22 See, e.g., Knight Commission (2009). Informing Communities: Sustaining Democracy
in the Digital Age, http://www.knightcomm.org/read-the-report-and-comment/;
Waldman, S. (2011). The Information Needs of Communities: The Changing Media
Landscape in a Broadband Age. Federal Communications Commission, Washington,
D.C., http://www.fcc.gov/infoneedsreport.
23 E.g., Napoli, et al. (2018).
24
Friedland, L., Napoli, P.M., Ognayanova, K., Weil, C., & Wilson, E.J. (2012). Review of
the Literature Regarding Critical Information Needs of the American Public.
Washington, DC: Federal Communications Commission.
25 See Friedland et al. (2012) for extensive and detailed descriptions and discussions of
each of these critical information needs categories.
26 See, e.g., Pew Research Center (2010).
27See, e.g., George, L., & Waldfogel, J. (2006). The New York Times and the market for
local newspapers. American Economic Review, 96(1), 435-447.
28
See, e.g., Anderson, C.W., Bell, E., & Shirky, C. (2012). Post-industrial journalism:
Adapting to the present, http://towcenter.org/research/post-industrial-journalism/.
29 George & Waldfogel (2006).
30See, e.g., Patterson, T. (2013). Informing the News: The Need for Knowledge-Based
Journalism. New York: Vintage.
31 Prior to training, two pilot tests on a sample of data were conducted by the researchers
in order to identify data-gathering challenges and difficulties interpreting or applying the
coding categories. These pilots refined the training and content analysis protocols.
Coders underwent two trial sessions using data samples (followed by assessment and
debriefings) in order to identify any challenges or uncertainties associated with the
content analysis protocol and code book. In identifying and coding individual news
stories, coders were instructed and trained to exclude social media feeds, widgets,
photo galleries, event calendars, outlet promotional announcements, and
advertisements from their coding activity. Coders also coded only those stories with
publication dates that matched the dates in the constructed week sample. Google
Translate was used to facilitate coding of foreign-language content. A dedicated
communication channel facilitated with the app Slack was established and maintained
during the entirety of the coding process, so that coders could share questions or
21
challenges that emerged during the coding process and obtain immediate feedback and
guidance.

Using Riffe, et al.’s (2013) formula for determining an intercoder reliability sample, a
random sample of cases was selected for an intercoder reliability check by an additional
trained coder. Using Krippendorff’s alpha, intercoder reliability was .81 for the Original
variable, .84 for the Local variable, and .73 for the Critical Information Needs variable.
See Riffe, D., Lacy, S., & Fico, F. (2014). Analyzing Media Messages: Using
Quantitative Content Analysis in Research (3rd ed.). New York: Routledge.

22
APPENDIX A:
MEDIA DATABASES/DIRECTORIES CONSULTED

The following were the key sources consulted in searching for news outlets within a
given community.

Television
1. Association of Public Television Stations’ Station Directory
2. FCC Broadcast Television License Database
3. NPR Labs Mapping and Population System

Radio
4. FCC AM and FM Broadcast License Database
5. Radio Locator

Newspapers
6. Library of Congress Directory of Newspapers

Online News Sites

7. Knight Foundation’s Directory of Community News Sites
8. Columbia Journalism Review’s Guide to Online News Startups
9. Online Newspaper Directory for the World
10. Michelle’s List

Alternative Sources
11. Mondotimes

23
APPENDIX B:
MANUAL SEARCH PROTOCOL

In order to identify additional sources missed in the online search, a manual search
protocol was developed in order to search for additional sources that were not readily
accessible via traditional search.

1. Start with Wikipedia and search for each community

a. May list community’s media
b. May provide community nickname(s) to utilize in search engine queries
c. May describe large minority populations that could be useful in search for
minority/foreign language media outlets

2. Go to Patch.com
d. Enter community name into Find Your Patch pulldown menu

3. Google Search
a. Key terms:
i. “[Community Name][Community Nickname] News”
ii. “[Community Name][Community Nickname] Journalism”
iii. “[Community Name][Community Nickname] Hyperlocal”
iv. “[Community Name][Community Nickname] Blog”
b. Repeat in Spanish
i. News = Noticias
ii. Journalism = Periodismo

24
APPENDIX C:
CRITICAL INFORMATION NEEDS CATEGORIES

1. Emergencies and risks

Individuals, neighborhoods, and communities need access to emergency information on
platforms that are universally accessible and in languages understood by the large
majority of the local population, including information on dangerous weather;
environmental and other biohazardous outbreaks; and public safety threats, including
terrorism, amber alerts, and other threats to public order and safety. Further, all citizens
need access to local (including neighborhood) information on policing and public safety.

2. Health
All members of local communities need access to information on local health and
healthcare, including information on family and public health in accessible languages and
platforms; information on the availability, quality, and cost of local health care for
accessibility, lowering costs, and ensuring that markets function properly, including
variations by neighborhood and city region; the availability of local public health
information, programs, and services, including wellness care and local clinics and
hospitals; timely information in accessible language on the spread of disease and
vaccination; timely access to information about local health campaigns and interventions.

3. Education
Local communities need access to information on all aspects of the local educational
system, particularly during a period when local education is a central matter for public
debate, decision-making, and resource allocation, including: the quality and
administration of local school systems at a community-wide level; the quality of schools
within specific neighborhoods and geographic regions; information about educational
opportunities, including school performance assessments, enrichment, tutoring,
afterschool care and programs; information about school alternatives, including charters;
information about adult education, including language courses, job training, and GED
programs, as well as local opportunities for higher education.

4. Transportation Systems
All members need timely information about local transportation across multiple accessible
platforms, including: information about essential transportation services including mass
transit at the neighborhood, city, and regional levels; traffic and road conditions, including
those related to weather and closings; timely access to public debate on transportation at
all layers of the local community, including roads and mass transit.

5. Environment and Planning:

Local communities need access to both short and long-term information on the local
environment, as well as planning issues that may affect the quality of lives in
neighborhoods, cities, and metropolitan regions, including; the quality of local and
regional water and air, timely alerts of hazards, and longer term issues of sustainability;
the distribution of actual and potential environmental hazards by neighborhood, city
region, and metropolitan area, including toxic hazards and brownfields; natural resource

25
development issues that affect the health and quality of life and economic development
of local communities; information on access to environmental regions, including activity
for restoration of watersheds and habitat, and opportunities for recreation.

6. Economic Development
Individuals, neighborhoods, and communities need access to a broad range of economic
information, including: employment information and opportunities within the local region;
job training and retraining, apprenticeship, and other sources of reskilling and
advancement; information on small business opportunities, including startup assistance
and capital resources; information on major economic development initiatives affecting
all local levels.

7. Civic Information
Communities need information about major civic institutions, nonprofit organizations, and
associations, including their services, accessibility, and opportunities for participation in:
libraries and community-based information services; cultural and arts information;
recreational opportunities; nonprofit groups and associations; community-based social
services and programs; and religious institutions and programs.

8. Political Life
In a federal democracy, citizens need information on local, regional, and county
candidates at all units of governance, including: information on elected and voluntary
neighborhood councils; school boards; city council and alder elections; city regions; and
county elections; timely information on public meetings and issues, including outcomes;
information on where and how to register to vote, including requirements for identification
and absentee ballots; information on state-level issues where they impact local policy
formation and decisions.

26
APPENDIX D:
DESCRIPTIVE STATISTICS

DESCRIPTIVE STATISTICS FOR COMMUNITY VAR IAB LES

(N=100)
Variable Minimum Maximum Mean SD
POPULATION 20256 291707 59408.05 50797.90
DENSITY 628.5 17346.30 3258.92 2830.43
INCOME 21301 156439 68775.30 25586.94
AA% 0 86 10.78 15.28
HL% 1.0 96.0 15.73 19.05

27
APPENDIX E:
ANALYSES OF COMBINATION CONTENT CATEGORIES

% of outlets % of original & local stories Ratio

(N = 663) (N = 2,476)
Newspaper 25.27% 61.39% 2.43
TV 13.62% 8.24% .6
Radio 51.14% 17.17% .34
Online only 9.99% 13.2% 1.32

% of outlets % of original & CIN stories Ratio

(N = 663) (N = 4,394)
Newspaper 25.27% 46.84% 1.85
TV 13.62% 15.21% 1.17
Radio 51.14% 26.58% .52
Online only 9.99% 11.36% 1.14

Local and CIN Stories

% of outlets % of local & CIN stories Ratio
(N = 663) (N = 1,837)
Newspaper 25.27% 58.31% 2.31
TV 13.62% 10.31% .76
Radio 51.14% 16.46% .32
Online only 9.99% 14.92% 1.49

28
ABOUT THE AUTHORS

Jessica Mahone is an Associate in Research at the DeWitt Wallace Center for Media
and Democracy in the Sanford School of Public Policy at Duke University. She
previously worked with the News Measures Research Project in 2016 and 2017.
Jessica received her Ph.D. in political communication from the University of Florida. Her
research interests include local news and civic engagement, research methods in
journalism, and news media and activism.

Qun Wang is a Ph.D. at Rutgers University’s School of Communication and

Information. She was a research assistant on the News Measures Research Project
assessing the robustness of local journalism in 100 U.S. local communities. Her
research focuses on journalism, technology, and society. Her work has been published
in Digital Journalism, Journalism Practice, Journalism, and more.

Philip M. Napoli is the James R. Shepley Professor of Public Policy in the Sanford
School of Public Policy and Professor of International Comparative Studies at Duke
University. Professor Napoli is also a Faculty Affiliate with the Sanford School’s DeWitt
Wallace Center for Media and Democracy, and is an Associate with the Duke Initiative
on Science and Society. Professor Napoli is the author of the new book, Social Media
and the Public Interest: Media Regulation in the Disinformation Age (Columbia
University Press, 2019).

Matthew Weber is an Associate Professor and the Cowles Fellow of Media

Management in the Hubbard School of Journalism and Mass Communication at the
University of Minnesota. Matthew is an expert on media industries, organizational
change and large-scale Web data. His recent work includes a large-scale longitudinal
study examining how media organizations evolve with technology. Additional research
includes an examination of technology in local news organizations, and the use of social
media within organizations.

Katie McCollough is an Assistant Professor in the Communication Studies department

at Augustana University in Sioux Falls, SD. McCollough received her Ph.D. from
Rutgers University where she served as the Graduate Assistant on the News Measures
Research Project as well as a Tow-Knight research fellow studying the local news
ecosystem in New Jersey. Her research interests include local news audiences and
other sites of participatory media.

The authors also wish to thank the undergraduate and graduate student research
assistants who worked on this project: Esko Brumel, Ian Dunham, Chioma
Ihekweazu, Christy Kuesel, Kreste Patrow, Lu Wu, and Shannon Zenner.

29