This study, “America’s Shifting Statehouse Press” employed several methods to obtain a census that is as complete as possible of reporters covering the 50 statehouses in the United States.  The main methods consisted of an intake questionnaire, outreach to press secretaries, legislative staff or state government employees and direct contact via email or phone calls to remaining news outlets with missing or conflicting data.

The Intake Questionnaire

A 13-question intake questionnaire was designed by Pew Research Center as the first step in identifying statehouse reporters and then having those reporters help identify others. The questionnaire asked individuals to identify the number of full-time and part-time statehouse reporters their organization employed and to name other individuals or organizations covering the same statehouse.

From August 28 and September 1 and November 12, 2013, the questionnaire was sent via email to 1,928 reporters and editors at 839 news organizations nationwide.

The initial contact list was pulled from several different industry sources: Local television news directors (source: BIA/Kelsey; 673 members), radio program directors (source: BIA/Kelsey; 42 members from news stations), database of digital nonprofit news outlets (source: Pew Research Center; 252 outlets), AP bureau chiefs (source: AP website; 16 different individuals), AP reporters (source: Cision; 38 members), statehouse legislative correspondents associations (source: California, 89 members; Illinois, 20 members; North Carolina, 30 members; Ohio, 39 members; Virginia, 24 members; Wisconsin, 45 members; Texas, 96 members), Twitter list of statehouse reporters (source: Public Twitter list by Dan Vock, former Stateline reporter at the Pew Charitable Trust; 115 members), database of newspaper news reporters and editors (source: Cision; 456 members) and database of state and local reporters (source: Cision; 256 members). As the lists were not mutually exclusive, duplicates were removed.

The questionnaire was administered by Rational Survey. All respondents were granted anonymity in their responses to survey questions, unless they agreed to an interview. Three rounds of email reminders were sent to ensure maximum participation. In total, 360 reporters and editors completed the questionnaire. The numbers for their organizations were inputted into a main database. The additional organizations and names were put into a second database for follow-up confirmation.

In addition, respondents provided the names of 171 colleagues covering the statehouse.  Researchers collected these names and contact information and sent them the questionnaire for confirmation (in cases where contact information was not provided, researchers used Cision, Leadership Directories and/or the Web to acquire the contact information of these individuals).

Additionally, Pew Research contacted individuals at 404 organizations derived from the American Journalism Review study conducted in 2009.

Additional Steps

To complement and confirm the questionnaire data, researchers employed two additional steps.

Step 1: Between November 12 and December 2, 2013, researchers contacted, by phone or email, officials in all 50 statehouses to identify any news organization with a statehouse reporting presence that did not respond to the Web-based survey. These officials were legislative and gubernatorial press secretaries and other employees who credential and interact with journalists. In these conversations, sources also were asked to confirm that our lists of current statehouse reporters were accurate. At the end of this process, Pew Research compiled an exhaustive list of all the news organizations covering each statehouse.

Step 2: After this process was completed, researchers analyzed the data and determined where staffing information was missing for specific outlets. In this last step, between December 2013 and March 2014, researchers contacted hundreds of reporters and editors at individual outlets to confirm their staffing information. Because the Associated Press, which has bureaus in every state, declined to provide staffing information, we called individual AP bureaus. We also contacted newspaper chains, which often have joint bureaus, to make sure we had correct staffing information.

During this step of the process, researchers conducted interviews to gather further insights on statehouse coverage. Thus, throughout the report, the reader will come across a number of quotes and information provided by reporters and editors about their experiences and their take on statehouse coverage.

Definitions

For the purposes of the study, a statehouse reporter was defined as a journalist who is assigned to cover state government on a full- or part-time basis and who does so from the state capitol. More specifically, the following definitions were employed:

Full-time and year-round statehouse reporter: The journalist covers the statehouse when the legislature is in session AND when it is not in session. In other words, the statehouse is the journalist’s full-time beat.

Session-only reporter: The journalist covers the statehouse ONLY when the legislature is in session.

Part-time statehouse reporter: The journalist covers the statehouse fairly often, but has other assigned areas of coverage.

Discrepancies

Throughout this multilayered process, Pew Research maintained a detailed database for every organization. In the multiple follow-up stages, discrepancies sometimes occurred. In all, researchers identified 142 such cases. These discrepancies were classified as: a) respondents working for the same organization providing different staffing information; b) respondents that misunderstood some of the questions; and c) news organizations providing information only for one state, while employing reporters in other states, too. Researchers contacted these individuals directly and clarified all the conflicting cases.

With staffing sometimes in flux, a number of respondents provided a range for the number of reporters covering the statehouse. This occurred most in the category of “students.” Respondents told us that the number of students covering the statehouse changes from session to session. In these cases, researchers entered in the final dataset the lower number of the range provided. This was consistent in all responses where ranges were provided.

Comparisons to American Journalism Review

This study contains several references to earlier studies of statehouse staffing conducted by American Journalism Review. That report contained two tallies of statehouse coverage – full-time newspaper reporters and “session help” (journalists who contribute only when legislatures are in session). The AJR statistics referenced in the Pew Research Center’s study refer only to full-time statehouse reporters.

To keep our comparisons consistent, Pew Research Pew Research went back to the 2003 AJR list and examined statehouse staffing levels at newspapers that were accounted for in the last two AJR tallies—2003 and 2009—and in our 2014 accounting.⁷  We removed any newspapers from the original AJR cohort that had ceased publication or joined a chain/content partnership with shared employees since 2003. As a result, 220 newspapers were identified by Pew Research and are included in the comparison count.

In addition, there were two cases in which outlets included in AJR’s tally existed in our 2013/2014 count, but not within the newspaper sector. Morris Communication and the Dayton Daily News told Pew Research that their statehouse reporters work for a company that has outlets across more than one platform, including but not limited to newspapers. Thus in the current Pew Research dataset they are counted within the “multiple platform” category.

Correlations

Pew Research Center looked at the association between the number of statehouse reporters and state demographics and legislative activity. For the test of these associations, researchers computed the Pearson’s r coefficient for each pair of variables. Correlations were considered to be statistically significant if they had a p-value of less than .05. The Pearson’s r correlationcoefficient is statistically significant for the associations between the number of statehouse reporters in each state (full-time and total) and both state population (full-time: r=.88; total: r=.80) and length of session (full-time: r=.55; total: r=.43).

Note: This census began in August 2013 and continued until March 2014. In this timeframe, staffing may have shifted due to the commencement of legislative sessions, industry layoffs and restructurings and additional hiring. While this count was extensively exhaustive and used multiple methods to derive our final counts, these numbers may have shifted between data collection and the release of this report. As a result, we may have missed organizations that have journalists dedicated to the statehouse beat. To this end, we encourage submissions of any updates to this data by contacting us at journalism@pewresearch.org.