A Pew Research Center study of three U.S. media markets has lots of interesting fodder and lessons for journalists and newsrooms.
In Local News in a Digital Age, Pew studied local news coverage and consumption more thoroughly than any local news study I’ve seen. I encourage reading the full 160-page report, which provided detailed studies of the news environments in Denver, Macon, Ga., and Sioux City, Iowa.
The study includes a survey of people in each community, asking extensive questions about their community involvement and news consumption, as well as a detailed study local news providers, including all the content during one week (last July) and a computer analysis of Facebook and Twitter content and engagement with local news providers.
I’ll present my thoughts on the Pew study in three sections:
- What the study says about media and lessons we can draw from it.
- My evaluation of this study (or opportunities for future studies). I was sharply critical of Pew’s 2010 study of Baltimore’s local news market, so I think I should address what I see as strengths and weaknesses of this study. This project leaves plenty of opportunities for further study of local media, but I find it far more thorough and credible than the 2010 study, which was so biased I said it was useless.
- My Denver and Sioux City experiences (neither of them a big conflict, but both worth disclosing).
Findings & lessons from the Pew study
Pew’s story up high presents the obligatory disclaimer:
These cities are not meant to be representative of the United States as a whole, but rather serve as detailed case studies of local news in three specific, unique areas in the U.S.
Yeah, but …
Pew did the study because the data from these three specific, unique areas would have value to others in the media. And I see several areas where the study reveals or confirms facts that will be helpful beyond the communities studied:
Diversity presents a great opportunity
I’ve argued before here how important diversity is to newsrooms. I think some journalists and news media executives see diversity as a nice thing to achieve if you can, but don’t see it as an economic opportunity. The Pew study, though, shows the opportunity I’ve long argued.
The Denver Hispanic community and the Macon African American community were both large enough to provide valid data broken down by race and ethnicity. And it was interesting to see that in both communities, the minority group studied showed a notably higher interest in local news than white people did.
If your community has a large minority community, I suspect that interest in local news in that community is higher than you think it is. I think news organizations have an opportunity to increase their audience with these segments of their communities if they:
- Successfully reflect the diversity of the community in their coverage.
- Provide quality coverage of issues of interest to the community.
- Succeed in recruiting and retaining diverse news staffs.
In Denver, “about twice as many Hispanics as whites very closely follow local crime, jobs and education,” the study said. In addition, education, religion and immigration were topics covered often by media targeting the Latino community in Denver. In Macon, blacks expressed higher interest in most local news topics than whites. The gap was particularly wide in African Americans’ interest in community events and news about jobs and unemployment.
Update: The diversity angle also stood out to Poynter’s Rick Edmonds in his post on the Pew study and to Josh Stearns of the Geraldine R. Dodge Foundation.
Engagement is an opportunity
The study did not document a wave of citizen journalism or even what I would call heavy engagement with the communities.
“Citizens are a part of the news process, but only in narrow ways,” the report said.
In none of the communities were more than 1 percent of the stories studied written by citizens, rather than staffers or officials. If people in the community showed up in the media, it was usually as sources, but no community had more than 20 percent of the stories citing citizen sources. That sounds like way too much reliance on the same old official sources.
And no more than a quarter of stories provided ways for readers or viewers to get involved in a story. For all the talk in our industry of engagement (and I’ve done a lot of it), I think we still have plenty of room to grow.
Local TV dominates
I remember working at more than one metro newspaper where we boasted (and, as I recall, had the numbers to prove it) that we had a larger audience than all the local TV newscasts combined. But that isn’t true now in any of the communities studied by Pew.
I think one of the things this says for newspapers is that they need to decide what they do best (or what the market most needs from them), and concentrate their dwindling resources in those areas, rather than trying to do everything for everyone at diminishing quality levels.
Newspapers did better among people who were more engaged in community life, which isn’t surprising. But I’m not sure what newspapers can do to get people who don’t read them more engaged in the community:
The city’s main daily local newspaper, meanwhile, is more prominent among the engaged than the unengaged. In Denver, a third (33%) of the very engaged often turn to The Denver Post, compared with 20% of the somewhat engaged and 19% of the unengaged. About half (52%) of the very engaged in Macon often use The Telegraph, substantially higher than the somewhat (33%) and the unengaged (26%). And in Sioux City, about equal numbers of the very (47%) and somewhat engaged (43%) often get news from the Sioux City Journal, compared with only three-in-ten of the unengaged.
Newspapers also did better among people who love their communities:
In Macon, for example, nearly two-thirds (65%) of those who rate the city as excellent often use The Telegraph for local news, compared with 41% of those who rate it as good and 26% of those who rate it as fair or poor. And in Denver, while fewer overall get news from the daily newspaper than in the other two cities, those who rate the city as excellent do turn to the Denver Post at a higher rate.
TV and newspaper contrasts
While more people are watching TV news, the study shows that TV news tends to be shallow and that newspapers are doing more enterprise:
In the five-day periods studied, stations tended to react to breaking events rather than initiate coverage through enterprise or investigative reporting. In Denver, 31% of broadcast and web stories produced by the TV outlets were triggered by a crime activity (excluding routine weather, traffic and sports), 16% by an accident or some other mishap, and 21% by something that a government agency or official did or said. By contrast, just 4% of stories were initiated by the outlets themselves.
Macon and Sioux City had similar numbers. And TV stories are short, mostly read by anchors:
Anchor voice-overs—the brief news items that an anchor reads through—made up the majority of their broadcast reports. These reads accounted for 55% of non-sports, traffic or weather broadcast stories in Denver, 62% in Macon and fully 71% in Sioux City. Edited packages—a technique that can allow for more depth but require a fair amount of newsroom resources to produce—were rare. Just 16% of TV stories in Denver, 18% in Macon and 14% in Sioux City were packages. …
The reliance on anchor reads plays out in the average length of broadcast stories. Denver stations averaged the shortest stories. Nearly half, 45%, of non-sports, traffic and weather stories studied on local Denver stations were 30 seconds or shorter. That was true of 29% of stories studied on Sioux City stations. Macon stations were far less likely to have the very short stories, at just 17%, though 41% of stories ran between 31 and 60 seconds.
At the other end of the spectrum, only 14% of TV stories in Denver, 13% in Sioux City and 18% in Macon were over two minutes.”
Newspapers were notably more enterprising:
In The Denver Post, 12% of the stories studied (26 in all) were driven by the newspaper staff compared with 4% among local TV stations — again, not including routine weather, traffic and sports items. Seven of the 26 Post stories focused on civic issues such as government and education. In Macon, the divide was even wider, with 17% of Telegraph stories (31 in all) being press-initiated versus 6% among TV stations. Nine of the 31 focused on civic issues. And twice as many of the Sioux City Journal’s stories were press-initiated compared with local TV (14% vs. 7%), with 20 of the 28 stories focused on civic issues.
The report found that niche media and community newspapers also initiated more stories than TV stations.
TV websites are weak
While TV leads newspapers as a source of news in all three communities, TV websites in all three communities largely suck (my term, not the Pew report’s) and are hardly used at all by people other than those watching the station. The graphic above is kind of complex, but basically the gold portions of the overlapping circles above represent people getting news from a website but not using the legacy product or social media sites.
Almost everyone using a TV station’s digital products also watches the station, and those web circles tend to be notably smaller than the newspapers’ web circles.
This supports what I’ve observed about differences between TV and newspaper websites: Newspapers are doing a better job building a distinct audience and TV station websites are more likely to largely promotional vehicles for the station, posting TV content online rather than developing as digital news operations in their own right.
Denver’s ‘grumpy engaged’
Under a heading “the grumpy engaged in Denver,” the study reports some findings that should concern my former Digital First Media colleagues at the Denver Post (whom I’ll invite to comment):
Fully 44% of very engaged Denver residents say that the local news media do a fair or poor job, compared with about a quarter of the somewhat engaged (26%) and the unengaged (25%).
Further, about a quarter (24%) of very engaged Denver residents who have lived there at least five years say that the quality of local news coverage has declined in the past year, compared with 13% of the somewhat engaged and just 6% of the unengaged.
In Macon and Sioux City, there is little difference in their evaluation of the job that the media are doing and whether the quality of local news coverage has gotten better or worse.
I don’t know what to make of that. It could reflect some cuts in newsrooms (and corresponding cuts in local content), but the study doesn’t say whether Denver newsrooms have cut more severely than those in Macon or Sioux City.
Opinion content is lacking
If you think the media and blogosphere are loaded with opinion, you won’t find support for that in the Pew report, which found less than 10 percent of the content was opinion content in any of the communities:
The low amount of opinion and commentary in the local press is in contrast to trends in the national media. With cable news, talk radio, blogs and social media, the national trend is to include more opinion and interpretation in information providers. However, that trend does not seem to have taken hold on the local level.
Media variety reflected
Especially in Denver, the study reflected the increasing variety of news outlets today, and the volume of work they are doing:
The Denver Catholic Register, a newspaper produced by the Denver Catholic Church since 1900, produced 35 stories, while the nonprofit website covering Colorado’s school systems, Chalkbeat Colorado, contributed 18 and Out Front Colorado, one of the oldest LGBT-focused newspapers in the country, published 20.
The report also noted the volume of news (and comments from the public) coming from local government websites and the websites of members of Congress.
I am pleased that my former colleagues in Denver are doing a better job of linking than other media, but thoroughly discouraged at how little linking most media do. I’ve argued here many times that linking is good journalism and good business. So this paragraph from the Pew study was discouraging:
In five days of content studied in each city, 31% of online news stories in Denver (excluding traffic, weather and sports) contained at least one embedded link compared with 3% of online stories in Macon and 1% in Sioux City.
My evaluation of this study
As noted earlier, I was initially disappointed in the 2010 study of the Baltimore news “ecosystem,” and on closer examination, I said the “old-media biases by the researchers were so profound that they truly didn’t understand the ‘news ecosystem’ they were studying.”
I have no such concerns about the validity of this study. I won’t bore you with a lot of methodology, but I was impressed with the lengths the researchers went to in searching for local bloggers and in finding not only the Twitter and Facebook pages of local news providers (including non-traditional providers as well as media accounts). They even included in their data analysis the Twitter and Facebook accounts of people who interacted with the news providers’ accounts.
The volume of Twitter and Facebook data might have been so huge that it made study of breaking stories difficult. The Twitter and Facebook study included stories that produced huge numbers of tweets using a hashtag (such as a campaign to get a Macon band on a TV show). But some breaking stories might generate just a few tweets, but they are important. Or they might generate multiple hashtags or no hashtag. Still, the inclusion of this detailed Twitter and Facebook analysis in the report was a welcome improvement.
The study included helpful and revealing data about the news being provided by neighborhood groups, city government websites and websites of members of Congress, and important recognition that many organizations that aren’t primarily devoted to news are becoming news sources.
The two questions I have about this study are not really concerns about the validity of the research as perhaps an opportunity for future study and a flaw:
- Most of the analysis of local media excluded “routine weather, traffic and sports.” Those are three major areas of local news and I’d like to see a study that included them.
- I question the decision to exclude Boulder from the study of the Denver metro area.
I have two suggestions for further research that aren’t really questions or criticisms of this report, just a wish that we had this kind of research in some other areas:
- Pew researchers deliberately chose metro markets of three different sizes. I’d like to see a media study of smaller markets as well.
- I’m pleased that the study included digital news sites and non-traditional news providers, such as government agencies and elected officials’ websites, and that it studied sharing on Facebook and Twitter. The study excluded “outlets that were primarily listings of entertainment events or sports scores, public relations vehicles, advocacy organizations, local businesses or news aggregators.” Those are part of the news environment of a community, and I’d like to see a study that includes them someday.
Why exclude weather, traffic and sports?
The analysis of media content excluded three of the most voluminous and most popular areas of local news coverage (as shown by the study in the table at the right).
Again and again as you read through the report, you see the disclaimer: “excluding routine weather, traffic and sports.” As you can see at the right, that excludes the leading topic of interest in all three communities and three of the top four in Denver.
Amy Mitchell, Pew’s director of journalism research, explained the decision by email: “We have dealt with this challenge for years and have found that because these segments take up a significant portion of the newscast but are also very different in structure, it is more understandable to the reader if we put forward those numbers at the outset as a part of the whole and then focus on the remainder. If we broke everything down as a percent of total, measures like sources and format can get quite confusing for the reader. We did though try to make that break very clear in the content section so that percentage of time was accounted for high up.”
I understand the desire to collect valid data on news about crime, local government, education, jobs, arts, business and other topics that Pew studied. But weather is the leading topic of interest in each of the communities studied. And sports is followed very closely by at least 18 percent of the people in each community (and 29 percent in Denver, which has local teams in each of the four major professional sports leagues, including the beloved Broncos). As the Pew graphic here shows, the three account for about one-third of all local TV newscast content in all three markets (and substantial radio content; in fact, the Denver study mostly omitted two stations focusing exclusively on sports). I’d like to see a study that provided the same sort of data about weather, traffic and sports as this study provided about topics of lesser interest.
Interestingly, the “routine” qualifier allowed the study to sneak in big weather stories, such as a flood and tornado in the Sioux City area in the period studied.
Should the Denver study have included Boulder?
I was surprised to see that the Denver study didn’t include Boulder County. The Census Bureau regards Boulder as a separate Metropolitan Statistical Area, and some friends and family who live or have lived in Boulder say it is culturally separate from Denver in many ways. But the friends and family also confirmed what I knew: When it comes to media, Boulder is unquestionably part of the Denver market. Boulder people watch Denver TV stations (Boulder doesn’t have its own stations) and listen to Denver radio stations (except for NPR affiliate KGNU, the only Boulder outlet included in the study). The Denver Post also circulates in Boulder County, though the Daily Camera is more popular in Boulder and the Times-Call is more popular in Longmont. Neither of the Boulder County newspapers was included in the study, though weeklies and a small daily in the Denver suburbs were included.
The list of Denver news outlets identified and studied by Pew is impressive, and the data collected is valid and valuable. But a study of the Denver market would have been more valid with inclusion of Boulder County, which has nearly 300,000 people and the University of Colorado.
The importance of Boulder was acknowledged in the introduction of the Denver section, which said, “Including Boulder, the Denver region has also emerged as a tech startup hub.” But, except for KGNU, Boulder was excluded from the media study.
The study included a community paper in Highlands Ranch, which Google Maps says is 23 miles or 33 minutes from Denver, just a little closer than Boulder at 29 miles and 38 minutes.
Commuting between Boulder and Denver (both ways) is so heavy that a website, 36 Commuting Solutions (named for U.S. Highway 36, which runs between the cities), is devoted to helping commuters. I’ve driven that highway several times in rush hour. Boulder is part of the Denver metro area and should have been included in the study.
Update: Maya Gurarie, Marketing Communications Manager for 36 Commuting Solutions, answered some questions by email, citing figures from the group’s consultant Urban Trans that US 36 carries between 80,000-100,000 daily vehicle trips and operates at close to 90 percent volume to capacity both directions. Buses serving the US 36 corridor carry about 14,000 riders per day.
Those numbers support my argument about how closely linked the two areas are, but Gurarie added, “I consider Boulder and Denver to be different metro areas.” (I didn’t ask her specifically about being separate media markets.)
Mitchell’s response on the Boulder issue: “The very straight forward answer is that we followed the MSA lines. You are right in that there is some Boulder population that follows Denver news but there is also likely population from other MSA’s that follow it or in other cities. This was the cleanest way to set up the city design.”
It’s true that Boulder has its own MSA (I presume some Boulder lobbyist sometime gets credit for that), but the study actually chose the metro areas based on their ranking in Nielsen Designated Market Areas, and Boulder is included in Denver’s DMA (which is substantially bigger than the MSA). I think a media study should choose markets according to media consumption, and Boulder is part of the Denver media market.
I’d like a study of rural media markets
Pew researchers deliberately chose metro markets of three different sizes, Denver from the largest 70 of 210 Nielsen DMAs, Macon from the middle third and Sioux City from the smallest 70. All three communities, though, are still metro areas. Having worked in two markets much smaller than Sioux City (Minot, N.D., and Shenandoah, Iowa), I’d be interested in a study of media coverage and consumption in smaller markets.
I think an effective study of a smaller market might examine a granular level of news exchange that probably exists in the metro areas but didn’t surface in the study because of the volume of higher-profile news providers.
For instance, I receive a frequent email newsletter (got an installment yesterday) from a Shenandoah resident (not a journalist), who regularly shares Shen news with a rather large email list of current and former residents. Some of the news is submitted by recipients and some curated from local professional news media. I also belong to a Facebook group of my high school class, where people occasionally share news (though we shared a lot more three years ago, when our 40th class reunion was approaching). I think an effective study of a smaller community would examine the impact of websites, bulletins and newsletters published by churches, PTAs, booster clubs and other news providers one or two levels closer to people than the media studied in the metro areas.
How about studying other news sources?
As noted above, the Pew study did not audit the news from or engagement with “outlets that were primarily listings of entertainment events or sports scores, public relations vehicles, advocacy organizations, local businesses or news aggregators.”
But some of those can be vibrant news providers. For instance, I checked one type of news provider that might be considered a public relations vehicle or advocacy organization: animal shelters. The Denver Animal Shelter, Macon-Bibb County Animal Welfare and Noah’s Hope Animal Rescue and Siouxland Humane Society, both in Sioux City, all have Facebook pages with frequent posts and thousands of likes. Most of the posts are just photos of animals available for adoption, but that’s important news to some people. A dog photo shared Wednesday evening at Noah’s Hope, one of two shelters in the smallest community in the Pew study, already has 72 likes, 44 shares and 14 comments. Organizations like that (and arts groups and churches and sports booster clubs and so on) are niche news providers in communities and they merit more study.
The study hints at the potential for these types of groups, but only with a brief mention in discussing social media:
Nearly half (46%) of social news users in Macon follow local groups such as the PTA or churches, compared with 28% of Denver users and 27% of Sioux City users.
Those are some significant numbers, and I’d like to see more studies of the content and audience of these types of news providers.
My connections to Denver & Sioux City
I’ve never been to Macon, but I have extensive experience in Denver and Sioux City.
The Denver Post, which was a big focus of the Denver portion of the Pew study, is the largest newsroom in Digital First Media, where I worked from 2011 to 2014. I visited the Post a half-dozen times and also visited newsrooms of the Daily Camera in Boulder and the Times-Call in Longmont, as well as other Colorado newsrooms outside the Denver metro area. (The Colorado friends and family I consulted on the Boulder question above did not include any of my former DFM colleagues, though I will invite some of them to comment on this post if they want.)
I have been to Sioux City several times as a reporter for the Des Moines Register and Omaha World-Herald, so I’m fairly familiar with that community as well.
I don’t think that my familiarity with either community created a conflict in commenting on the Pew study, but it’s worth mentioning here. The study of both communities rang true with what I know of the media there.
Two more disclosures:
- Pew gave me an advance copy of the study, embargoed for 12:01 Eastern time today, so I could prepare this blog post.
- Robyn Tomlin, Pew’s Vice President of Digital and Communications, is a friend and former Digital First Media colleague.
I hope to see more research of this type
I would encourage media and universities to undertake studies of this type in their communities. While I have generalized from the Pew study of Denver, Macon and Sioux City, and I think many trends will be common to many or all communities, the details of each community will be helpful to media and others in that community.
The Pew study includes a detailed section on methodology that would be helpful to researchers in any community. I hope to see more studies like this, not only from Pew, but from local researchers in many communities.