For this week’s datajournalism assignment, Godwin Nnanna and I looked into the admitted theft of $250 million of Delta State money by James Ibori, who pled guilty in London in late February. (get the data)
We wanted to find out just what he stole, explain how it fit into the context of corruption more generally in Nigerian states, and clearly illustrate the magnitude of his actions. We’re still working on the article.
To tell this story, we needed information on what Ibori stole as well as more general information about Nigerian government budgets. We had to compile our own datasets from a variety of incomplete sources.
- The Metropolitan Police Press Bulletin contained detailed information about the guilty plea, including addresses, value amounts, and photographs of Ibori’s highest value assets. This information was the basis of all of the newspaper reports we saw.
- Values were not provided for some of Ibori’s UK properties, so we used Zoopla’s property database to arrive at reasonable price estimates based on comparable properties in the vicinity
- We obtained the guilty plea from the writer who covered the issue for the BBC. The prosecutor did not respond to our emails.
- Nigerian government budget information is very hard to get, and actuals are almost impossible to obtain. We were able to cobble together enough information for the story however:
- Godwin called contacts at the Nigerian Economic and Financial Crimes Commission
- I scraped federal and state budget data from YourBudgIt.com, a scrappy government transparency initiative out of Nigeria’s CoCreation Hub. YourBudgIt (who launched a redesigned website today) then sent me updated data after I requested it on Twitter.
- A number of citizens’ advocacy groups track budgets and expenses of Nigerian states. Many will look through the budget for building projects and then take photos of those projects to monitor if the money is being used.
- Our datasource is here, which includes inline links to sources where possible
We want to use this data to add context to the story– to tell people that focusing on Ibori’s luxury assets actually minimizes our impression of how much he stole. The assets reported by Scotland Yard account for less than half of what he stole.
James Ibori and his associates pled guilty to stealing a lot of money– half as much as the Nigerian federal government spends on agriculture in a year and several times the annual budget for education and health capital projects in Delta State, where he was governor.
Then we created a series of dashboards for the four major delta states, showing a variety of figures about budget allocations, health, education, and poverty. Here for example is Delta State, where inequality is rising rapidly, poverty is widespread, health and education are a small part of the budget, and most of the money goes into capital projects, which sometimes go into people’s pockets rather than the infrastructure they supposedly support.
Partnerships between Techs and Journalists
This was the first time I partnered with a journalist who was unfamiliar with what it takes to write software or do data wrangling. As a result, I think we ended up doing our own thing toward the end, and I expect that we’ll have to put significant work into making our stories converge. Between the growing popularity of data pieces and Initiatives like the Data Journalism Handbook, I hope it will become easier for these collaborations to go smoothly.
Overall, I think it’s important to communicate the stylistic affordances of data journalism as well as the constraints on scope created by committing to collaborate on a data piece. Overall, I would love to learn more about successful working practices of data researchers who collaborate with journalists.
Tech Design & Recommendations
This was a *hard* project to do. Here are some recommendations:
- We need to encourage and support more NGOs to release data alongside their reports.
- We need to design databases which are capable of containing information about the source of figures in different rows/columns
- We need to do more to make journalists aware of data resources in their own area, as well as support NGOs towards sharing more of their data sources with journalists
- We could crowdsource the creation of datasets from disparate sources if we had the tools for crowd researchers to document the source of a particular number, and tools for users of that dataset to evaluate the sources of those numbers
- Lots of organisations are independently pressuring government for the same data. A “What Do They Know” app for Nigeria which helped them pool efforts would be awesome.