How inewsource analyzed federal whistleblower reports

inewsource broke the news Tuesday that the House Committee on Veterans’ Affairs will hold a hearing on the VA Office of the Medical Inspector, a powerful agency that oversees VA medical care around the country and investigates allegations of wrongdoing in the healthcare system that serves more than 9 million veterans.

As part of that story, we wanted to know how good of a job the medical inspector’s office does when it investigates claims made by whistleblowers about the VA healthcare system.

There was no database to answer that question, so we built our own.

inewsource compiled information from hundreds of reports published online by the Office of Special Counsel, an independent federal agency reporting directly to the president, which reviews whistleblower investigations conducted across the executive branch of government.

We found that the medical inspector’s investigations of the VA healthcare system have been “unreasonable” in about 16% of cases reviewed by the Office of Special Counsel in the past decade.

Two weeks before publication, inewsource provided the findings and methodology to spokespeople at the House Committee on Veterans’ Affairs, the Office of Special Counsel and the Department of Veterans Affairs to give them a chance to review our work.

inewsource developed careful criteria for how exactly to include and describe details from the special counsel’s reports in a dataset to ensure the results could be replicated. Here’s background information on the reports we compiled, how we built our dataset and what it took to fact check our work.

How the special counsel operates

The Office of Special Counsel is an independent agency that reviews whistleblower allegations brought by current and former federal employees, along with applicants for federal employment, in the executive branch of government. The special counsel protects the identity of whistleblowers and ensures that alleged wrongdoing is investigated and corrected.

The special counsel reviews allegations, but only refers cases for investigation where a determination has been made that the information provided establishes a substantial likelihood of one of the following:

  • Violation of a law, rule, or regulation;
  • A gross waste of funds;
  • An abuse of authority;
  • A substantial and specific danger to public health or safety; and/or
  • Censorship related to research, analysis or technical information.

Once the special counsel decides to refer a case, the request for investigation is sent to the head of the agency involved. This agency is usually the agency where the whistleblower is employed. (For example, if an employee at the San Diego VA hospital files a complaint with the special counsel, it will be investigated by an entity of the Department of Veterans Affairs, such as the VA Office of Inspector General, or the VA Office of the Medical Inspector. The head of the agency determines what entity conducts the investigation.) The agency will then provide a report to the special counsel with its findings, describing whether it substantiated the whistleblower’s allegations.

Federal law requires that when the special counsel receives an agency report, it must determine whether the findings appear “reasonable.” The office has broad discretion in deciding whether the findings appear reasonable or not, and in making its determination will consider the scope of the agency’s investigation, whether the findings are inconsistent or do not align with the facts of the case, or whether any proposed corrective actions address the underlying problems. Whistleblowers have an opportunity to comment on the agency report, and the special counsel may require the agency to follow up and provide a supplemental report.

The special counsel closes a case by sending a letter to the president with the agency’s report, the whistleblower comments and the special counsel’s decision on whether the findings appear reasonable. These reports and letters are publicly available on the office’s website, dating back to 2009.

Compiling public files

inewsource hand-compiled a dataset based on reports available on the Office of Special Counsel’s website. Every whistleblower investigation reviewed by the special counsel’s office that was publicly available on the website was included in inewsource’s data.

The data compiled for each case includes case numbers, the date the case was closed by the special counsel’s office, the locations where the alleged incidents took place, the federal agency responsible for investigating the allegations, whether the agency’s investigation substantiated the whistleblower’s allegations and whether the Office of Special Counsel found the agency’s investigation reasonable.

Search inewsource's database of federal whistleblower complaints

Click here to search the inewsource database of federal whistleblower complaints within the executive branch.

In cases that were referred to the Department of Veterans Affairs for investigation, inewsource also noted which entity within the department conducted the investigation. The VA’s whistleblower investigations are conducted by many entities — including the VA Office of the Medical Inspector, the VA Office of Inspector General and others — depending on the nature of the allegations. Cases referred to the medical inspector’s office generally relate to the VA’s healthcare system.

The special counsel’s public files section of its website is broken down into shaded boxes, with each containing information on a given case with links to reports about that case. These documents include reports from agencies that do the investigations (usually labeled “Subject Agency Reports”), responses by whistleblowers to the agency investigation (usually labeled “Whistleblower Comments”) and the special counsel’s conclusions about the investigation (usually labeled “Letter to the President”).

Some pieces of information compiled by inewsource about each case were taken directly from the counsel’s website, because it was an available description of the case. That includes the location of the alleged violations and the date the case was closed. Other pieces of information had to be compiled by reviewing the documents associated with each case, including whether the special counsel found the investigation “reasonable.”

inewsource established standards for compiling each of the pieces of information included in its dataset. Here are some important details about how our data was compiled:

      1. Case number: Each employee who files an allegation with the Office of Special Counsel is assigned a case number. If multiple case numbers are referenced in a single counsel report, that means several employees reported information related to the same allegations. inewsource grouped case numbers together if they were referenced in the same special counsel reports. Each case only represents one line of data in inewsource’s dataset. Case numbers were taken directly from the special counsel’s website.
      2. Date closed: Occasionally, the special counsel’s office reopens a case, if for example, the agency conducting the investigation provides more relevant information after the office officially closes the case or if the agency fails to take corrective action. inewsource’s dataset references the most recent closure date for each case. Closure dates were taken directly from the special counsel’s website.
      3. Agency: The special counsel refers allegations to a federal agency for them to investigate. inewsource took the agency names from the special counsel’s website and standardized the spellings (for example, changing “Department of Defense” to “Defense.”) These agency names sometimes refer to whole departments, such as “Homeland Security,” and in other cases refer to specific agencies within those departments, such as “Customs and Border Protection.”
      4. Location: Each case has a description of where the allegations occurred. For example, an employee might claim wrongdoing at an Army base or VA medical facility anywhere around the country or the world. If more than one location is involved, all of them are included in inewsource’s data. This information was taken directly from the special counsel’s website.
      5. Findings: Information on whether the Office of Special Counsel found an agency report reasonable or not is available in the closure letter for each case. (The file is usually labeled on the website as a “letter to the president.”) inewsource reviewed the closure letters for each case on the counsel’s website and searched for the counsel’s standard language describing whether the findings were reasonable or not.If the Office of Special Counsel found any aspect of any agency report in a given case to be unreasonable, then inewsource considers the findings unreasonable. This includes cases in which the special counsel found a report to be “unreasonable in part.”If the counsel concluded it could not determine whether the report was reasonable or not, inewsource said the case had “no determination” in its data. In the few cases where reports available online did not clearly state whether the findings were reasonable, inewsource denoted, “information not available.” If the counsel stated that the findings were reasonable but it still had concerns about an agency’s conduct, inewsource still marked these cases as “reasonable.”
      6. VA Office: Many entities within the VA could be tasked with reviewing allegations of wrongdoing. Information on which VA office investigated the case is usually available in the closure letter (“letter to the president”) for each case. inewsource reviewed the closure letters for each case on the counsel’s website and searched for the counsel’s standard language describing which office was tasked with the investigation, noting the office in its dataset. If the closure letter did not describe this information, inewsource reviewed the case files called “Subject Agency Reports” to determine which office led the investigation.
      7. Agency conclusions: Agencies reviewing cases for the Office of special counsel will state whether they substantiated the whistleblowers’ allegations. The special counsel summarizes these conclusions as “substantiated,” “partially substantiated” or “unsubstantiated.” This information was taken directly from the special counsel’s website when available. If the website states the case was “totally unsubstantiated” or “unsubstantiated but led to agency changes,” inewsource considered those “unsubstantiated” allegations in its dataset to minimize the number of categories for analysis.If summaries were not available on the website, inewsource reviewed the closure letter (or “letter to the president”) in each case and determined whether the agency’s conclusions appeared to substantiate the whistleblowers’ allegations. inewsource denoted these cases with an asterisk in its dataset.inewsource determined the agency “substantiated” the whistleblower’s allegations if all of the allegations were substantiated and the agency “partially substantiated” the allegations if at least one allegation was partially or fully substantiated but not all allegations were substantiated. inewsource considered the allegations “unsubstantiated” if none of the allegations were substantiated at all.

    Fact checking

    inewsource investigative reporter Jill Castellano reviewed all the cases on the Office of Special Counsel’s website twice to ensure the information was entered properly into the dataset. Intern Natallie Rocha reviewed the special counsel’s cases a third time, fact checking each piece of information in the dataset. Any discrepancies or complicated cases were reviewed by Managing Editor Laura Wingard. The data analysis was reviewed by Brandon Quester, director of data and visuals.

    Two weeks before publication, inewsource provided the findings and methodology to the House veterans affairs committee, the VA and the special counsel’s office for review.

    • We'll do the work. You just read it.
      Sign up for our newsletter.

    Zachary Kurz, spokesman for the special counsel’s office, said the office does not dispute inewsource’s findings. He said the office plans to independently verify the statistics, but it could not do so before our story published because of issues with the office’s new electronic data system.

    “There are several discrepancies between our internal case management system and information presented on our website’s ‘public file,’ that we are working to resolve,” he said. “Both the case management system and website were launched last month and require some cleanup before we would be able to verify your findings.”

    Analyzing the compiled data

    After compiling and standardizing the data, inewsource used Microsoft Excel to perform pivot tables that provided summaries of the Office of Special Counsel’s cases.

    inewsource grouped together all the cases investigated by each federal agency (e.g. one line item for “Health and Human Services” and another line item for “Department of Agriculture”) and summed the total number of “reasonable” and “unreasonable” investigations in each agency. We divided the total number of unreasonable cases for each agency by the total number of cases the agency reviewed to calculate the percent of cases in each agency that the special counsel’s office considered “unreasonable.”

    The analysis showed that 15.95% of Veterans Affairs investigations were considered “unreasonable” by the Office of Special Counsel.

    Since VA investigations are conducted by many entities, we also specifically wanted to know how many investigations led by the VA’s Office of the Medical Inspector were “unreasonable.” inewsource grouped together the cases in each VA office (e.g. one line item for the medical inspector and another line item for the VA inspector general’s office). We divided the total number of unreasonable cases for each office by the total number of cases reviewed by each office to calculate the percent of “unreasonable” cases in each office.

    The analysis showed that 15.69% of the VA medical inspector’s investigations were considered “unreasonable” by the special counsel, close to the average for the whole VA.

    We then added up the total number of “unreasonable” cases in the data and divided that by the total number of cases in the data to calculate the percent of “unreasonable” investigations across the federal government. This number was 11.28%, lower than what we found when looking just at VA cases.

    Finally, inewsource wanted to know how much of the special counsel’s workload comes from each federal agency. We grouped together all the cases investigated by each federal agency (e.g. one line item for “Health and Human Services” and another line item for “Department of Agriculture”) and summed the total number of cases for each agency. We then divided the sums by the total number of cases in the dataset.

    The analysis shows that 40.85% of the special counsel’s cases were related to the Veterans Affairs.

    We divided the number of VA medical inspector cases by the total number of special counsel cases, concluding that 25.56% of the special counsel’s cases were specifically related to the VA medical inspector’s office.

shadow-ornament

We'll let you know when big things happen.

About Jill Castellano:

Jill Castellano
Jill Castellano is an investigative data reporter for inewsource. To contact her with tips, suggestions or corrections, please email jillcastellano [at] inewsource [dot] org.