Minority Report software can predict the future

Business as unusual.

Who'd have thought that Tom Cruise could be a harbinger of our global future? But silly sci-fi in the vein of Minority Report, where future events are predicted and dealt with before they even occur, may well have the last laugh if the findings coming out of the Microsoft Research lab are a sign of things to come.

Teamed with the Technion-Israel Institute, researchers set out to design software that can predict events before they happen. This includes epidemics and outbreaks of violence. It turns out that with 22 years worth of New York Times articles and a data-based Wikipedia the results are surprisingly accurate.

In fact, the software was able to correctly predict outbreaks of cholera in Angola based mainly on the occurrence of droughts in the area. The system signalled that incidents of cholera were likely because previous news reports had shown that cholera is common following droughts - perhaps due to poor sanitation through lack of water. Sure enough, days later in drought-stricken Angola the first cholera cases were being reported.

The researchers also found that the system they had developed could foresee violent and political unrest by combining factors such as location, citizen's earnings, and GDP. After testing the software on everything from political instability to epidemic outbreaks it has proved to be correct between 70-90 per cent of the time.

When it comes to violence, the findings are slightly controversial. The researcher's report concludes: "The system identified, in an automated manner, that for locations with large immigrant populations (e.g. Ohio and New York), the shooting of an unarmed person by the police can cause protests". This result suggests that the automation process may oversimplify complex events.

But the implications when it comes to disease are promising. If an epidemic can be predicted then aid agencies and governments can prepare for the worst, ensuring that help is on hand when the first incident hits.

Tom Cruise: harbinger of our global future? Photograph: Getty Images
Getty
Show Hide image

A small dose of facts could transform Britain's immigration debate

While "myth-busting" doesn't always work, there is an appetite for a better informed conversation than the one we're having now. 

For some time opinion polls have shown that the public sees immigration as one of the most important issues facing Britain. At the same time, public understanding of the economic and social impacts of immigration is poor and strongly influenced by the media: people consistently over-estimate the proportion of the population born outside the UK and know little about policy measures such as the cap on skilled non-EU migration. The public gets it wrong on other issues too - on teenage pregnancy, the Muslim population of the UK and benefit fraud to name just three. However, in the case of immigration, the strength of public opinion has led governments and political parties to reformulate policies and rules. Theresa May said she was cracking down on “health tourists” not because of any evidence they exist but because of public “feeling”. Immigration was of course a key factor in David Cameron’s decision to call a referendum on the UK’s membership with the EU and has been central to his current renegotiations.  

Do immigration facts always make us more stubborn and confused?

The question of how to both improve public understanding and raise the low quality of the immigration debate has been exercising the minds of those with a policy and research interest in the issue. Could the use of facts address misconceptions, improve the abysmally low quality of the debate and bring evidence to policy making? The respected think tank British Future rightly warns of the dangers associated with excessive reliance on statistical and economic evidence. Their own research finds that it leaves people hardened and confused. Where does that leave those of us who believe in informed debate and evidence based policy? Can a more limited use of facts help improve understandings and raise the quality of the debate?

My colleagues Jonathan Portes and Nathan Hudson-Sharp and I set out to look at whether attitudes towards immigration can be influenced by evidence, presented in a simple and straightforward way. We scripted a short video animation in a cartoon format conveying some statistics and simple messages taken from research findings on the economic and social impacts of immigration.

Targeted at a wide audience, we framed the video within a ‘cost-benefit’ narrative, showing the economic benefits through migrants’ skills and taxes and the (limited) impact on services. A pilot was shown to focus groups attended separately by the general public, school pupils studying ‘A’ level economics and employers.

Some statistics are useful

To some extent our findings confirm that the public is not very interested in big statistics, such as the number of migrants in the UK. But our respondents did find some statistics useful. These included rates of benefit claims among migrants, effects on wages, effects on jobs and the economic contribution of migrants through taxes. They also wanted more information from which to answer their own questions about immigration. These related to a number of current narratives around selective migration versus free movement, ‘welfare tourism’ and the idea that our services are under strain.

Our research suggests that statistics can play a useful role in the immigration debate when linked closely to specific issues that are of direct concern to the public. There is a role for careful and accurate explanation of the evidence, and indeed there is considerable demand for this among people who are interested in immigration but do not have strong preconceptions. At the same time, there was a clear message from the focus groups that statistics should be kept simple. Participants also wanted to be sure that the statistics they were given were from credible and unbiased sources.

The public is ready for a more sophisticated public debate on immigration

The appetite for facts and interest in having an informed debate was clear, but can views be changed through fact-based evidence? We found that when situated within a facts-based discussion, our participants questioned some common misconceptions about the impact of immigration on jobs, pay and services. Participants saw the ‘costs and benefits’ narrative of the video as meaningful, responding particularly to the message that immigrants contribute to their costs through paying taxes. They also talked of a range of other economic, social and cultural contributions. But they also felt that those impacts were not the full story. They were also concerned about the perceived impact of immigration on communities, where issues become more complex, subjective and intangible for statistics to be used in a meaningful way.

Opinion poll findings are often taken as proof that the public cannot have a sensible discussion on immigration and the debate is frequently described as ‘toxic’. But our research suggests that behind headline figures showing concern for its scale there may be both a more nuanced set of views and a real appetite for informed discussion. A small dose of statistics might just help to detoxify the debate. With immigration a deciding factor in how people cast their vote in the forthcoming referendum there can be no better time to try.