Governance and Global Affairs Blog

Making policy with big data

Making policy with big data

Governments have increasing amounts of data at their disposal. How can big data be used in policymaking? And are governments ready to deal with all this data? That is what Sarah Giest is interested in. Read the interview here.

Is it a recent development that governments use big data for policymaking?

‘Not really. Governments have always used large amounts of data: personal data, tax data, administrative data. The main difference is that now, new types of data and new techniques to understand it have become available.’

Could you give some examples of these new types of data?

‘Think of mobile phone data, like gps locations of citizens, and transport information. These are more dynamic, realtime types of data, that for instance a city can use to look at how citizens move around. What I want to know, is: does it actually help governments to have bigger amounts of data? How do they use it for the benefit of citizens?’

What can governments do with all that data?

‘They could, for instance, make an app with traffic, public transport and weather information, that could help citizens in their daily lives. Including all that information in one app would make it easier for people to make travel decisions. Or governments can use big data at a policy level, to make policy decisions about things like carbon emission reduction, health, or public services. But actually, successful examples are still rare. A lot of processes go into working with the data, before it can actually be useful to policymakers.’

You collaborate with computer scientists and Gemeente Den Haag in the research project READ- URBAN. Does that project help the local government to make more use of data?

‘Yes, in that project we look at a specific group of people, the ‘working poor’: people that have jobs but live below the poverty line. The Gemeente Den Haag wants to take new policy initiatives for this specific group of citizens. So we help them figure out where these people are, what they do and what kind of services they need in order to move ahead. We help the Gemeente by bringing together data from different sources: Google Maps, social media, CBS, etcetera. This way, we can learn more about these people’s social networks, for instance, and the kind of neighbourhoods they live in.’

What’s the main challenge of that project?

‘The hardest part is linking all the data. Gathering a lot of data about someone is actually a very tricky process. In this case, we need to get permission to get the official data, understand how it is structured, and then combine it while taking into account its complexity in terms of time, space and type. We actually hired a theoretical physicist to make sense of some of the spatial processes that are going on in the data.’

Do you think governments are ready to deal with such complex data sets?

‘Many governments have difficulties finding people who have the skills to analyse this kind of data. A solution to this is to outsource some of these things to companies. But then the problem is that governments don’t get to see the raw data, they just get a report that summarizes it. Data is not neutral – choices are made on what data to include or exclude, and on what algorithm to use. So I think outsourcing these things is a dangerous route to go, but building the skill within governments will take time.’ 

Are there big differences between countries in how governments make use of data?

‘Yes, but it’s tricky to compare countries. Quite often, Estonia and Singapore are mentioned as examples of countries that are more ahead in this regard. But I believe these are dangerous examples to use: Estonia used to be a communist country and Singapore is a one-party state, where it’s easier to decide things. In the Netherlands, for instance, things are different: there’s a set institutional structure that has been in place for years, which determines how data is collected, shared and linked.’

Do these differences between countries also have to do with cultural differences?

‘Yes, definitely. In Sweden, for instance, a lot of data is just publicly available and nobody feels weird about it. Anyone can look up your school record and all your grades online there. Whereas in Germany, where I’m from, things are a lot different. When Google Maps took pictures of streets, people were protesting to get their houses off Google Streetview. How people feel about privacy and data sharing can really differ between countries.’

How about privacy issues in the Netherlands: is big brother watching us?

‘I think it’s a misconception that government looks at individuals all the time. Yes, a lot of data is collected, but that usually happens at a very aggregated level. So a city will have numbers of how many people are in the city at a certain time, for instance, but they’re not tracking me personally when I go to work or when I leave the city. Data at the individual level only comes into play if they identify a person of interest – a criminal suspect, for instance. But to actually look at individuals, a government needs to meet legal requirements.’

Big data seems to be everywhere these days. Do you think it may be a trend that will pass?

‘No, I think we are really at a point where we need to take big data seriously, especially when it comes to governments using it. So I think diminishing it as a trend, like many people do, is dangerous. If we do that, we won’t do proper research into the effects and use of big data. It doesn’t really matter what we call it - ‘big data’, or just ‘data’ – but it’s here to stay.’

Add a Comment

Name (required)

E-mail (required)

Your own avatar? Go to www.gravatar.com

Remember me
Notify me by e-mail about comments