This article first ran on April 11, 2017 on the Columbia Journalism Review's website.
By Anya Schiffrin
At a time when U.S. journalism is being hit by the collapse of advertising revenue, ongoing uncertainty about business models, and a continual assault from the alt-right and the White House, a new book explains why investigative reporting is essential for policy making and social well-being.
Democracy’s Detectives: The Economics of Investigative Journalism (Harvard University Press, 2016) by economist James T. Hamilton is probably the most detailed, comprehensive study ever published of how US investigative reporting has evolved since Watergate. Hamilton, director of the journalism program at Stanford University, provides a taxonomy of subjects covered by different types of outlets and estimates the cost-benefit to society of the reporting.
Hamilton’s economic perspective is particularly relevant at a time of resource constraint. Many of the investigative stories Hamilton examines cost their newspaper $200,000 or $300,000 to report, but the savings to society were far larger. The following Q&A has been edited for length and clarity.
One of the remarkable things about Democracy’s Detectives is you looked at data mined from 12,000 entries in the journalism prize contests held by the Investigative Reporters and Editors organization from 1979 to 2010. What did you find?
Investigative reporting involves original work, about substantive issues, that someone wants to keep secret. The economics of this type of work are often discouraging. Original work means there is a fixed cost of finding and telling the story. When your reporting changes laws or policies, the benefits spill over onto many people who will never be your readers or viewers, so it is hard for the media outlet to monetize the value generated by these policy changes. By denying access and resisting FOIA requests, the government can raise the transaction costs to finding out what is going on in an agency. These are some of the reasons why, on average, a story submitted to the IRE prize competition takes six months to produce.
Despite these calculations, investigative work does get done. In newspapers, investigative work appears to happen at outlets with higher circulations, which allows a paper to spread the costs of investigation across many readers. If you are offering important stories that others are not, this provides a reason for consumers to seek out your work.
In Democracy’s Detectives you remind us of a key point made in Journalism of Outrage which is that the impact of journalism coverage can unfold over time and in three phases: individualistic, deliberative and then substantive. In other words, the impact on an individual, of a piece of investigative reporting, can be transformed into a broader discussion about how to address the problems revealed by journalists and then finally into a policy change. After analyzing so many pieces of investigative reporting, what kind of impact did your study find from that reporting, and how did that change over time?
Nearly 15 percent of IRE prize contest entries noted their work led to investigations by others. The most likely individual effects were resignations (6%), indictments (4%), and firings (3%). Slightly more than one percent of the IRE stories resulted in the enactment of new laws.
Through case studies, I found that each dollar spent on stories can generate hundreds of dollars in benefits to society, though gains are distributed in ways hard for news organization to translate into additional reporting resources. In one example, I studied the December 2008 News and Observer series “Losing Track: North Carolina’s Crippled Probation System.” The paper found that between 2000 and 2008, 580 people on probation in North Carolina were convicted of killing someone. The three-day, 10-article series took six months of reporting, and cost $216,500 to produce. The reporting generated benefits across the state, but only 6 percent of state households paid for print version of the newspaper. The series changed personnel, law, policies, and expenditures. I estimate that for each dollar in investigative costs, $287 emerged in net policy benefits in the first year of full implementation of probation reforms. There is no process, though, that links brand reward for the paper to the magnitude of these net policy benefits. If the News and Observer had captured just 10 percent of net policy benefits, the paper could have nearly doubled the size of its newsroom.
I also studied the impact of investigative reporting by looking at the work of Pat Stith, who did investigative work in North Carolina for nearly 36 years. During that time, he produced more than 300 investigations. Of those, 149 generated substantive changes, 110 produced deliberative outcomes, and 43 generated individual impacts.
In 10 percent of Stith’s investigations, the result was a new law passed in North Carolina. Across nearly 36 years on investigative beat, he generated 31 new laws. The legislation he generated affected multiple policy areas: public safety, environmental protection, criminal justice, civil rights, and health care. In his last four years of investigative work, he prompted significant legislative change each year.
As you note, there doesn’t seem to me much of a business model for investigative reporting. What did you learn from the past that can help us in the future? How can investigative reporting be supported, especially in the small towns where it’s under threat or no longer exists, according to you.
There are five incentives that lead to the creation of information about public affairs: Pay me (the subscription model); I want to sell your attention (advertising); I want your vote (partisan); I want to change how you think about the world (nonprofit); and I like to talk (expression). Investigative reporting can be generated by different combinations of these incentives, and I think we’ll continue to see a reweighting of these incentives as a source for investigative work.
I’m optimistic about the future of investigative reporting in part because of the evolution of computational journalism. New combinations of data and algorithms can lower the costs of discovering stories. Telling stories in more personalized and engaging ways holds the prospect of the type of product differentiation which can raise the probability that you could charge for news.
It is true that outlets in smaller towns are at a disadvantage, since there are fewer people whose resources or attention could support the creation of costly stories about their local community. Smaller outlets, though, can seek help. IRE has provided training subsidies for smaller newsrooms. ProPublica will often partner with small outlets who can tell the local version of a national story. Right now, state and local governments are releasing data online that can be the source of new accountability stories in small communities.
What lessons do you have for the media under Trump? What should journalists do?
It is always cheaper to repeat a story than to find one. Through his tweets and his free-wheeling management style, President Trump will generate many stories that have a relatively low cost to cover and a high appeal to entertainment demands. The challenge for journalists will be to tell the stories of policy implementation on the ground, to describe the lives of voters as policies are changing (rather than waiting four years to check back in during a campaign).
A key tool may be FOIA. I found in an examination of FOIAs at 14 federal agencies, media FOIAs dropped overall by about a quarter and by nearly 50 percent for local newspapers. In an era when career executive branch workers see the very basis for their agencies challenged, responses to FOIAs may offer a way to provide journalists with the type of data that gives rise to more accountability reporting.
Looks like you haven't made a choice yet.