Crowdsourcing International Development Transparency



"Citizens of Sarpallo VDC in Mahottari district waiting for the VDC office to open."

The money was supposed to buy bicycles for ten girls attending public school in Mahottari, Nepal. The head of the village government allocated 150,000 rupees—about $1,400—for the project, which he said would encourage girls from the marginalized Dalit (so-called “untouchable”) community to attend school. An auditor signed the paperwork and the money left the government’s coffers.

“Not a single student in the village got a bicycle,” says Pranav Budhathoki, the founder and CEO of the Local Interventions Group. Budhathoki received an anonymous report of the missing funds during the pilot of an anti-corruption project he is looking to launch throughout the country of Nepal. He immediately sent his regional representative, a respected journalist in Mahottari, to find out what happened to the money.

Budhathoki is trying to curb corruption in Nepal by publishing information about local government budgets—about where money should have gone—and then eliciting feedback from citizens about what projects actually happened. In urban, wealthy contexts, people can use smartphones and apps to crowdsource information about corruption, but Budhathoki knew that wouldn’t work in rural, undeveloped parts of Nepal. So he came up with an alternative: People could submit reports of corruption using very simple tools—text messages and phone calls—and he would use technology on the back end to aggregate, analyze, and publish the data. This is the key to development, he says: not to increase funding or develop new projects, but rather to give citizens the information and the tools they need to demand accountability.

“The problem is not resources,” he explains. “The problem is citizens not knowing how much the government has allocated in their name.” To solve the problem of corruption, he says, it’s necessary for people to be “engaged and involved and reporting what they see around them, the malpractices and corruption issues.”

According to Budhathoki, the Nepali government sends money to Village Development Committees to address every imaginable social injustice, but “up to 60 percent of that allocated budget gets sent back to national coffers because the local governments don’t spend it,” he says. Local news reports corroborate that claim. As a result, Budhathoki says, “health centers are running without medicines and doctors, and kids are attending schools with no books or teachers.”

Budhathoki founded the Local Interventions Group in September 2011. Since then, he’s launched an app aimed at reducing police abuses and increasing police effectiveness and has used the open-source software Ushahidi to address cheating and violence in the 2013 Nepali elections and to ensure that earthquake relief services and supplies reached the right people after this year’s earthquake. His newest project tackles corruption head-on by publishing information about budgets and then crowdsourcing reports of corruption. He ran a pilot project last year in two of Nepal’s 75 districts, and, over the next two years, he plans to scale up the program so it will reach the entire country.

Here’s how it works: First, the Local Interventions Group’s regional representatives, all well-connected reporters, ask government officials how much money was earmarked for, say, schools or clinics in a certain area. Then, the Local Interventions Group crowdsources reports about what is needed, what is being done, and the quality of the work.

During a three-month pilot, Budhathoki solicited reports in each of three categories: absenteeism of local officials, misspent or disappeared development funds, and missing pension payments. At “Mobile Help Desks,” volunteers gathered information from text messages and phone calls. Budhathoki received 1,300 citizen reports. Then, the regional representatives verify the reports with local government officials and threaten to make the information public unless the grievance is resolved. Over the course of the three months, the Local Interventions Group verified about 600 to 700 genuine, specific, and actionable reports.

Take the bikes for female students, for example. Budhathoki’s contact in Mahottari took the report of the stolen bicycle funds to the district government. “Within three days, he got a call from the chief district official,” Budhathoki says. The Village Development Committee secretary admitted responsibility and promised to return the money. It took three more months, but “the 150,000 rupees was finally returned back to the national coffers,” Budhathoki says. No girls have gotten bicycles, since there’s a new government with new priorities in office, but the money is back where it belongs.

As Budhathoki scales up his anti-corruption project to the entire country of Nepal, he hopes to streamline his interactions with the government. His NGO will collect, vet, and verify grievances; map them onto a GIS platform; and then share the reports both with government officials and civil society organizations. That way, “each district official, head of the village government, [will] see these reports at his desk, on his computer, early morning every day so that he can talk to his colleagues on the ground and do something about that,” Budhathoki says.

Several researchers who have investigated anti-corruption programs say such collaboration is the key to success. Carla Miller is the founder and president of City Ethics, a nonprofit that aims to help local governments develop programs to prevent corruption, and has prosecuted federal corruption cases involving elected officials. She’s found that anti-corruption programs entirely internal to the government with no line to citizens can themselves be corrupted, while citizen groups can lose momentum and don’t necessarily understand the inner workings of government. “Ultimately, you have to have communication, collaboration,” she says.

The spirit of collaboration with government officials was behind Budhathoki’s decision to take down a map of crowdsourced corruption reports he’d initially published on the Local Interventions Group’s website. “We put it up for a month and then we shut it down because the government was getting antsy because they didn’t want citizens to know there’s a lot of corruption in the government agencies,” Budhathoki says. “Of course everyone knows that there’s a lot of corruption in government agencies at the local level but they didn’t want that to be up on the website for everyone to see.”

So, in the end, Budhathoki came to the conclusion that taking the map down was the right decision. “We’re okay with that because my approach has been collaborating with government officials rather than confronting them,” he says—and, now, he adds, he can walk into government offices and say, “Alright, let’s try to change something on the ground.”

This is a controversial approach. Some analysts, such as Miller, agree that taking down the map was the right call. “Maps of corruption and press and then a reaction from the government” is just a short-term solution, she says, that “doesn’t integrate citizens into long-term planning and identifying those issues faster.”

Yuen Yuen Ang, an assistant professor of political science at the University of Michigan, has looked at the platform I Paid A Bribe, which crowdsources reports of corruption in India, and similar platforms in order to identify what works and what doesn’t. She agrees that working with the government should be Budhathoki’s priority.

“Budhathoki made a pragmatic choice in agreeing to the government’s request—which, importantly, was a request, not a demand,” she wrote in an email. “By doing so, he avoids antagonizing the government, making it more likely for state authorities to work with his organization to take concrete steps in fighting petty corruption.”

“The purpose [of crowd-sourcing initiatives] is not to shame the government or individual officials, because, as Budhathoki himself points out, everybody already knows there is a lot of corruption,” she adds.

The Local Interventions Group isn’t combating corruption within elite circles, which Ang says would call for a political response, but rather what Ang refers to as “petty corruption”—that is, “low-stakes, diffused corruption that directly affects the lives of regular citizens.” Changing out government officials has no effect on small-scale corruption, Ang says, since it’s a systemic problem. Instead, the best way to approach the problem is through procedural changes: for example, “reducing bureaucratic discretion, centralizing budgetary management, reforming public compensation practices, etc.”

Budhathoki was able to make such a collaborative procedural change when it came to paying pensions. During his pilot, he found many people didn’t get their pensions on time, or at all. The government officials decided that, when the next month of pension payments came due, they’d have the Local Interventions Group’s regional representatives oversee the process, verifying who was owed a pension and who received one. That way, “we know how much money is being given to whom,” Budhathoki says. “The whole process [is] transparent.”

That’s the kind of change he hopes to continue to make in the future. “That happened because we’re seen as partners,” he concludes.

Citizen Science Crowdsourcing Data Science



About six years ago, Shane Davis quit his job as a biologist to spend all of his time aggregating and analyzing data about oil and gas.

Davis says he gives citizens the tools “to fight off” the oil and gas industry. With information he provides, Davis says communities can go to companies and tell them:

‘Wait a minute. We have information here showing that over the last X amount of years your operations have already caused groundwater contamination at this rate or your patterns of spills are at this rate.’ Or this many people got hurt. Or there are complaints, laundry lists of complaints about your operation. Or maybe these operators have had huge problems with their well casings. Maybe they’ve contaminated private water wells or an aquifer.

In other words, Davis is stepping in where he says state regulatory agencies have failed to protect the people, earth, and water of Colorado from pollution. Indeed, all over the U.S. and world, environmentalists have stopped relying on government agencies to monitor everything from oil spills to the spread of invasive species. Instead, calling themselves “citizen scientists,” they’re taking matters into their own computers and smartphones: gathering, analyzing, and publishing data. Some, like Davis, have adopted a combative stance toward the government. Many others would rather work with government agencies and say they’re contributing information the state simply doesn’t have the capacity to gather.

Davis gets his information from a state agency, the Colorado Oil and Gas Conservation Commission, or COGCC. When companies send mandatory reports about their operations to the state, the COGCC uploads them to their website as PDFs. Since PDFs are nearly impossible to aggregate and analyze, Davis developed tools to scrape information from them and dump it into an Excel spreadsheet—“a really, really topnotch spreadsheet, a spreadsheet that has incredible functionality,” he says. Then, he looks for patterns: spills, groundwater contamination, well casing failures.

At first, Davis spent 70 to 80 hours a week doing this work. “I was basically in my own solitary confinement for two years,” he recalls. “Now, it’s gotten a lot easier. I’m pretty quick at it.”

When Davis finds useful information, he brings it to communities and shares it.

“My presentations are not data-heavy. You’re going to lose everyone if you just jam it packed full of data,” he explains. “I’ll put in images, satellite images of the shale formation, and I’ll pick a bunch of people—politicians or environmental whatever. And I’ll show where they live, right on top of that shale formation. But then I’ll show them all the other well bores that are around their house and some failures that have happened. Maybe there’s a benzene spill.”

Davis is not the only citizen scientist to adopt this tactic: to take advantage of the troves of data available on government websites to tell stories. Adrian Cotter, who’s been with the Sierra Club for 13 years, says environmentalists have always considered themselves “data-based” and the only difference is the accessibility of data: “There’s just a lot of resources online for finding everything from all the oil rigs in Alaska to all the [oil] leases in the Gulf.” Overlaying that data with maps, as Cotter did with oil rigs and the migratory paths of the caribou herds of the Alaska National Wildlife Refuge, can tell powerful stories.

In 2012 and 2013, six Colorado communities voted on bans or moratoriums on fracking within their borders. Davis visited all of them, bearing information about the oil and gas industry. He recalls telling them, “Hey, look what’s happening in your backyard! Near your schools, your playgrounds, your universities, public parks. This is the information that I can give you.” In its turn, the Colorado Oil and Gas Association, the industry trade association, spent hundreds of thousands of dollars fighting the measures. It’s unclear to what extent Davis’ data managed to sway voters, but he’s been credited with inciting anti-fracking sentiment in Colorado and with coining the term “fractivist.” Five of the measures passed, one of them by only 13 votes.

Since then, the Colorado Oil and Gas Association has sued all five communities, saying only the state has the power to regulate the industry. The state has joined the industry in the suits, which Davis views as yet more evidence that “regulatory agencies are designed by those they benefit the most” and are not capable of protecting the citizenry. Courts struck down three of the fracking bans, and the Colorado Supreme Court is considering the other two.

Some citizen scientists find it more efficacious to work with government agencies rather than fighting them. David Newell, a professor at Southern Cross University, in New South Wales, Australia, and a researcher of frogs and toads, says, “We need to work collaboratively to be able to bring about change.”

For example, he wanted to find out how far the invasive cane toad had spread across his state of New South Wales. It would have been expensive to launch a study—and, besides, people already had the information; there just wasn’t an easy way for them to share it.

“Well, they can ring up their local park service office and somebody would write it down on a piece of paper and if we’re lucky it might end up in a database,” he says he thought at the time. “But let’s actually come up with a mechanism by which people can use technology, log their record, and also at the same time gain additional information around what it is that they can actually do—so give them a portal to be able to contribute to the database and then get additional information.”

That’s why Newell decided to build Toad Tracker, which has since become Toad Scan, a means for average Australians to report cane toad sightings. Toad Scan is the opposite of a PDF: Instead of uploading information in such a way that anyone interested in the data has to wade through reams of documents in order to find anything useful, all the data is right there, instantaneously available to the public.

That democratization of information at first made government regulators uncomfortable, Newell recalls, because “they see themselves as the knowledge-holders of databases.” That said, he adds, once “government agencies see that this is a really valuable tool to be able to capture spatial information,” they generally come around. His goal is for citizens, academics, and the public to work together towards conservation goals.

As for Shane Davis in Colorado, he’ll continue to fight. “I’m not stopping until we change law so it favors communities and the environment and does not favor corporations, corporate capitalism, oil and gas,” he says.

Crowdsourcing Mapping open data



MafiaMaps is an app to crowdmap the mafia phenomenon all over Italy.

  • “We want to make mafia visible to everyone, city by city, region by region.”

    I am not at a press conference held by the Italian Ministry of Internal Affairs or the Head of the police; I’m with a group of young political science students. And this is not just wishful thinking: MafiaMaps is an app they are building to map the mafia phenomenon all over Italy.

    Pierpaolo Farina, Hermes Mariani, Claudio Ripamonti, and Samuele Motta are part of the core group of MafiaMaps volunteers, about 15 people in their early- to mid-20s. We meet in the courtyard of the Political Science faculty of the University of Milan, where they’re studying or recently graduated.

    The group met and bonded during a political science class on the sociology of organized crime.

    While in Italy there is—predictably—a lot of research on the mafia, the young students felt that there wasn’t a structured organization of all that knowledge.

    So, two years and a half ago, they started WikiMafia, an online encyclopedia (Creative Commons-licensed) that now counts more than 200 full articles, and another 1,000 partially completed or draft articles.

    On WikiMafia, you can find anything from mafia organizations and their historical development to power structures and infiltration in the public administration and private sector.

    But as the WikiMafia volunteers’ work and their academic careers progressed, they felt that something was missing: How do you give people immediate access to all this knowledge? How do you make people understand that mafia is everywhere in Italy and closer than people think?  

    They eventually figured out that an app would make all of their research immediately available to anyone.

    That idea became MafiaMaps, a constantly updated map pinpointing the last-known location of convicted criminals and of mafia killings, as well as where anti-mafia organizations are at work, creating projects and organizing events.


“We start from the judiciary inquiries,” explains Pierpaolo Farina, the project manager. “It’s the only way to have data that is compelling and precise: you have names, dates. Then we broaden the scope and keep researching, interviewing people and fact-checking everything.”

Analysis includes books and other research materials, sometimes even “oral history”: in many cases, I’m told, the “everyday” mafia victims do not make it to the news, so nobody writes about them.

Farina mentions the case of Pasquale Campanello, a prison guard in Poggioreale, Naples. A father of two, Campanello was only 32 when he was shot by four killers in 1993 for refusing to help convicted criminals receive messages and gifts from their outside accomplices. There are no books celebrating him, killed just for doing his job.

The team went to talk to his widow and created a lasting trace of his sacrifice: now WikiMafia has an article on Campanello and he will soon be on the map.

WikiMafia cost about 150 euros and many hours of volunteer work. Since its inception, MafiaMaps has aimed to have a life of its own: “We decided to launch a crowdfunding campaign to create a community that uses the app: it would be useless to develop it, otherwise,” says Farina. At 26, he is one of the few graduates of the group, as well as a published writer and prolific blogger.


Colors correspond to various mafia groups: The Sicilian Cosa Nostra is purple; the Calabrian 'Ndrangheta is blue.

Colors correspond to various mafia groups: The Sicilian Cosa Nostra is purple; the Calabrian ‘Ndrangheta is blue.

For months, the team studied other successful crowdfunding campaigns and then set up their own, with a number of intermediate goals: the first 10,000 euros will provide mapping for the Lombardy region, reaching the 20,000 threshold will allow them to map other two regions and so on. The final goal is 100,000 euros for all 20 regions.

“We wish we had Kickstarter in Italy: it would make things so much easier!” Farina jokes. But I interview them on a productive Monday morning: Farina has just returned from a popular morning radio show. “We raised 400 euros in 10 minutes!” he says.

Next, MafiaMaps will go “the start-up way,” they say, looking for foundation grants, sponsors, and other forms of financial support.

“We do not want public money, in order to avoid the controversy that often arises for those who work on the matter: that you do it because you expect to live with government funding,” Farina declared in an interview to prominent newspaper La Repubblica, earlier this month.

The Faculty of Political Science will soon give them a room to use as their headquarters and they are already in touch with a number of possible sponsors.

The campaign was launched on March 21 and has raised about 14,000 euros so far. As the final day is May 23, MafiaMaps will likely have only enough money to start mapping Lombardy.

The guys do not seem worried: they have already started developing the app: “We’re gonna start with that and do it in the best way possible,” says 23-year-old Hermes Mariani, a MafiaMaps co-founder from Lecco. “We will release the app as scheduled, next March. When people will see the results, they will want to help and support MafiaMaps in their region.”


MafiaMaps will allow users to contribute, pointing out news, events, but it will be a selective crowdsourcing.

“We’ll verify everything before putting it on a map, as we check carefully the contributions to WikiMafia,” 26-year-old Samuele Motta clarifies.

Many of them have studied the mafia for years and are critical of the sloppy work they often see on the issue. “If you say everything is mafia-related, then it’s easier to argue that nothing is really mafia-related,” says Claudio Ripamonti, a 20-year-old volunteer. They later explain that, in white-collar crimes, mafia affiliates are often a loose connection, but their role is often amplified. Also, despite many connections between mafia, politics, and enterprise, “we’ve never been sued in two years of work,” says Farina, laughing.

While big and small anti-mafia organizations are already supporting and contributing to the project, MafiaMaps has an ally in the Italian open data community, and, fittingly enough, from Sicily. Confiscati Bene (“well confiscated”) is a participatory project stimulate an effective re-use of buildings and other assets seized from the mafia.

The project investigates their current condition and potential through the analysis of relevant data coming both from official sources and from bottom-up, citizen-monitoring initiatives, as previously reported on techPresident.

It is not an easy task, Confiscati Bene project manager Andrea Borruso tells Civicist in a Skype interview: “Public datasets date back to 2013, many information are still missing and the quality hasn’t improved at all in the past two years.”

Confiscati Bene was born from a hackathon a little more than a year ago and has been nurtured by a small and active volunteer community. The founders recently created an association and are looking for a business model, Borruso says: “This year has been great but this project, this topic, deserves more, it deserves actual everyday work…instead of nights!”

He’s only half-joking: Borruso says that their work got them prizes and acknowledgments (they were mentioned as a best practice by former World Bank open data specialist Samuel Lee during the last PDF Italy) but it hasn’t got any easier in the 14 months since Confiscati Bene’s inception.

And the Italian institutions are not helping: “We’ve recently been told that the agency [ANBSC, the Italian National Agency for the Management and Disposal of Assets Seized and Confiscated from Organised Crime] is updating the data but at the moment nothing is available, not even the old datasets,” he explains.

As I write, an independent and volunteer initiative is the only place showing the national datasets of the confiscated assets, while the government page displays a “coming soon” sign.


For many years, the mafia was perceived as only affecting the south of Italy and as a rough, unrefined form of organized crime. Inquiries and trials in the early 90s showed a rather sophisticated systems with ramification all over the country and abroad, including the United States.

One of the most prominent judges at the time, Giovanni Falcone, worked closely with the FBI, the NYPD, and federal prosecutors in a case known as the Pizza Connection, busting an international heroin smuggling ring that laundered drug money through pizza parlors. (His statue can be found at the FBI Academy in Quantico, Virginia.)

The fact that this kind of project started in northern Italy is quite significant: “We’re proving that civic antibodies work,” says Farina, a reference to the common metaphor of the mafia as Italy’s disease.

Still, it’s not easy to talk about it: “We recently presented MafiaMaps in my hometown,” adds Mariani. “People came to us saying that they know the mafia is there, but it’s better not to know all these things. We wanna show them they’re wrong.”

The MafiaMaps team follow the words of Judge Falcone, displayed on their website: “The mafia is not invincible, it is a human fact, therefore, it has a beginning and an end.”

One of the heroes of the anti-mafia movement, Falcone was killed with his wife and three police officers in 1992: his car exploded as he was reaching his native Palermo.

“Italy has been called ‘the mafia country’ for many years. We’re very proud when we’re interviewed by foreign media and hear that we’re now the country with a strong anti-mafia movement,” says Pierpaolo Farina at the end of our interview.

Last Monday, Judge Falcone would have turned 76. Saturday will mark the 23rd anniversary of his death.

It will also be the last day of the MafiaMaps fundraising campaign.