How to use data and engagement journalism to hold government accountable

State and local government agencies give away millions of dollars in incentive funding to companies and individuals each year, often with little transparency or public scrutiny. That’s a lot of unchecked power. 

Financial incentive programs often have lofty goals: to reduce air pollution, expedite medical research or boost technological development, for example. Taxpayer dollars are given to private organizations in the name of public good, often for public health, yet the public knows little of the success or failures of the programs or the particulars of who benefits. 

The programs are ripe for investigative journalism. While it’s unlikely that reporters have resources to fully audit a program, we can ask questions and publish records that shine light on the process and players, and potentially spark public policy conversations about the programs’ efficacy and fairness. 

That’s what I did in partnership with Kerry Klein at KVPR in a project supported by the USC Center for Health Journalism’s Impact Fund.

Here in the San Joaquin Valley, our air basin is among the most polluted in the country. Financial incentives are a key strategy used by the local air district to reduce emissions. Each year, the San Joaquin Valley Air Pollution Control District distributes millions of dollars to farmers to help them buy cleaner tractors. We wanted to know whether the program was working, who benefits and who gets left behind. 

We coupled data and engagement journalism to answer those questions. We fought for public records, interviewed program administrators and participants, analyzed the data sets and shared our results with some of the agency’s biggest critics for feedback. 

We found some of the world’s largest agricultural corporations have received millions of dollars, while some small farmers can’t afford to upgrade their dirtiest equipment. Meanwhile, ag equipment is still responsible for a substantial portion of air pollution in the valley and regularly puts farmworkers’ health at risk. That all begs the question: Are incentives the best strategy to tackle pollution and protect public health? We published our findings, as well as the raw data of award recipients, in hopes that others might raise additional questions and critiques to continue the conversation. 

Here’s how to do it: 

Learn the basics

Once you’ve narrowed in on the program you’d like to focus on, set up interviews with administrators and review any documentation you can find online. Find out how the program is supposed to work. Who qualifies? Have there been any recent changes to the program? Who decides where the money goes? What kind of oversight is there? Has there ever been an audit? 

Initial conversations might help you narrow your focus early on. Also find data or context about the larger problem the program is trying to solve. For example, we found government data that estimated the amount of pollution emitted from the sector of vehicles the program sought to improve. 

It can be helpful to also speak with advocates or agency critics at this point in your reporting. They may have insight, suggestions or ideas based on their own research. 

Get the data 

It’s possible that no one has ever asked the agency for this dataset before. For that reason alone, plan to submit a public records request for data early in your reporting process. Here are some tips to help you get your hands on what you need. 

• Request a key to the data set you are inquiring about. That will provide you with a kind of glossary of the categories of data available. 
• Ask for award recipients, date of award, expected outcome (such as jobs created, or emissions reduced) and other program-specific data. Here’s an example. 
• Request data be provided in a machine-readable format, such as CSV or Excel spreadsheet. We initially received data in a lengthy PDF, which was unusable. With the help of public record law experts, we learned California law states that information retained in an electronic format must be made available in any electronic form in which the agency holds the information. The “agency shall provide a copy of an electronic record in the format requested if the requested format is one that has been used by the agency to create copies for its own use or for provision to other agencies.” 
• If the agency is not responding or sends incomplete data, send a strongly worded letter, on letterhead, and quote relevant sections of public record law that remind them of their obligation. Here’s an example. If your news outlet doesn’t have an attorney on retainer for this kind of assistance, First Amendment organizations are available to help. Consider MuckRock, the California News Publishers Association legal help team, or the First Amendment Coalition.   
• Learn from our error! Be sure to review your conclusions, totals or findings with the agency. Do this especially if the agency gives you multiple sets of data. We failed to compare what we thought was the same data sent two different times. In reality, the agency had sent us updated data. We didn’t realize it was different from the original data, and we published an error that had to be corrected. 

Engage potential program participants 

People who participate in incentive programs and those who are the target of such programs have valuable insight into what is or isn’t working, or why they do or don’t participate. We wanted to hear from farmers about their thoughts on the program, so we did strategic outreach to build sources and hear multiple perspectives. Here are quick steps to broaden your pool of sources and gain valuable insight. 

1. Create a survey using Google Forms. Introduce yourself and your project at the top, including your contact information. Keep the survey short. Use plain language. Use both open-ended questions and specific questions. Here’s an example. 
2. Go to where your target audience already talks. We promoted our survey with an audio call out that aired on KVPR several times. We got a few responses that way. What was more effective, was introducing our survey and project with a link in a Facebook group heavily used by farmers in our region. We did receive more responses there, but even more helpful was the conversation that took place in the comments. 
3. Engage more deeply. We followed up with many survey participants with phone calls and in-person interviews to learn more about their experiences.