If managed well, crowdsourcing can help strengthen government decision-making and enhance the democratic experience of citizens. Done poorly — by a self-selecting online poll, for example — it can be a hollow tool to avoid real engagement, and may even damage trust in government.
So how can it be used effectively? Two academics have made an attempt to answer that question, using the example of crowdsourced reform of off-road traffic laws in Finland.
Tanja Aitamurto and Hélène Landemore, who designed and led the crowdsourcing operation, argue in an open-access paper that building in inclusiveness, apart from being morally desirable in a democracy, increases the range of perspectives to be available to decision-makers, allowing for better informed policy. Opening up participation on specific issues also gives citizens who may not normally engage with politics a stake in the system.
The participants enjoy a “really strong sense of empowerment when they participate”, Aitamurto says.“… people actually learned to understand other people’s points of view.”
“One of the participants we interviewed said, ‘this is the first time in my life I feel that I am really participating in the making of democracy and influencing the decision-making in this society.’ It feels much more real than just voting for some person,” she explained at a recent conference.
Another told researchers the crowdsourcing platform was the only way they felt they could properly engage from the remote area where they lived.
Accountability is a key ingredient. In a crowdsourcing context, it increases the need for policymakers to provide justifications for decisions, creating positive incentives for the use of evidence.
Transparency not only means citizens are kept in the loop about issues that matter to them, but allows them to observe the input of others, many of whom have a different perspective. This “horizontal transparency” wasn’t intended — initially the project was designed to elicit information from participants, rather than be a place for deliberation between them.
“This was a little bit unexpected, but people actually learned to understand other people’s points of view,” she said. Deliberation also helps with expectation management. “They don’t think that their idea will be taken to the law as is, they are more realistic.”
Another positive outcome was that analysis showed engagement went beyond the “usual suspects”. Participants were evenly distributed between urban and rural areas, and geographically distributed across Finland. They were fairly well educated, and most were full-time workers. But only around one-third were the type to write to an MP or participate in town hall meetings, meaning the other 70% were those who didn’t normally take part in civic activities.
It’s also important to discuss what crowdsourcing does not do, Aitmamurto argues. It’s not a decision-making tool, or a method of direct democracy.
“Crowdsourcing is not a public opinion poll,” she contended. “Crowdsourcing is always inherently based on self-selection, meaning that it’s the people who are interested, who want to participate, who will participate. So it’s not a random sample, it doesn’t have any statistical representativeness.”
The Finnish process
The law, which regulates the use of vehicles such as snowmobiles and all-terrain vehicles off-road, was seen as having fallen behind, in part due to increases in traffic. It has two basic goals: to protect nature from the harm that off-road traffic causes and to ensure the safety of the drivers and other people.
There are a few complications to what might sound like a simple law: indigenous Sami people use snowmobiles for herding reindeer, and many Finns use them for hunting and fishing in the vast wilderness. People who live or holiday in cabins may also not be too happy if someone set up a new track next to their house.
The Ministry of Environment decided it would experiment with crowdsourcing as a public engagement tool in the reform process, setting up a website on which citizens could propose ideas, vote ideas up or down and comment on each others’ proposals. The platform was moderated, though of 4000 comments, only around 10 were removed in each phase for being inappropriate.
Another website was set up to provide more information about off-road traffic law, which included a video with the minister of environment explaining the aims and reasons behind the crowdsourced process and rolling updates on the process.“… researchers were able to use network analysis to rank the popularity of ideas and figure out their appeal to different groups of people.”
In the first part of the process, the ministry invited policy experts to identify 10 areas for discussion by citizens, such as problems related to off-road traffic, age limits for off-road traffic, the environmental impact of off-road traffic, and so on.
Participants were then asked to propose ideas and share perspectives relating to the problem areas of their choice. These were posed as questions, such as “is the current age limit of 15 years sufficient, or should it be lowered to 12?”. Public demand also led to the creation of a free-form section where anyone could suggest problems outside the 10 posed at the beginning.
The first crowdsourcing phase engaged about 700 users, who generated around 340 ideas and conversation starters, 2600 comments and 19,000 votes in reaction to the ideas.
In the second phase, the participants were invited to propose solutions to the problems identified in the first round. To do this, the broad challenge areas were divided into more narrow topics. The “safety” section, for example, was subdivided into “improving safety in off-road traffic” and “safe transition traffic off- and on-road”. The second phase generated less engagement — around 170 ideas, 1300 comments and 6000 votes.
The third phase saw each participant given a random sample of ideas to be evaluated. Two evaluation methods were used: rating and comparison. In a rating evaluation, participants scored an idea from 1 to 5, while for comparison, participants were asked to either choose a favourite idea out of two or rank several ideas in the order of their preference. Experts then rated ideas according to different criteria.
From this evaluation, the researchers were able to use network analysis to rank the popularity of ideas and figure out their appeal to different groups of people.
Identifying challenges for future
Unfortunately, while this was occurring the minister changed. The new boss is not as keen on crowdsourcing, so the evaluation reports are sitting on the new minister’s desk.
A similar process for reforming housing legislation “is going really well”, however. That one is being driven by the civil servants themselves, meaning it’s more likely to reach an outcome. Still, Aitamurto believes the road reform journey has helped identify some issues for the future.
The biggest challenges are due to the sheer volume of information crowdsourcing produces. Whereas when governments consult lobby groups they are usually presented with amendments for consideration that are coherent and synthesised, and can be copied and pasted into the law, “the input that comes from the crowd is really atomic, it can be scattered”, she says.
Understanding the range of preferences can take time. In traditional stakeholder engagement preferences tend to be fairly well-known in advance, as they often result from the ideological differences held by political parties or interest groups. Sifting through thousands of inputs from all different participants means it takes longer.
Aitamurto is working on a tweaked approach to crowdsourcing with the innovative city of Palo Alto in California, where the city creates a synthesised policy and asks people to comment on it.
Despite the challenges, the fact that so many people are willing to spend their own time to share their ideas and engage with government shows the potential for crowdsourcing in policy reform.
“We can see that the participants feel like they become closer to the decision-making process, the policy making process when they actually participate online,” Aitamurto argued.
“This type of simple process can actually make a difference in these participants’ lives, and also their perception about what their role in democracy is.”
Tanja Aitamurto was a presenter at the 2015 Australia and New Zealand School and Government conference in Melbourne.