The Royal Statistical Society’s been in the data busi­ness now for a hun­dred and eighty years. We love work­ing with ODI, who’s been around for three years, because we find we’re say­ing the same thing with dif­fer­ent voic­es, with the same agenda. 

For the RSS, the data agen­da’s absolute­ly embed­ded in our his­to­ry. Charles Booth was one of our ear­ly pres­i­dents, map­ping where pover­ty lay side by side with wealth in London. Florence Nightingale was our first female fel­low, map­ping how dis­ease was spread­ing in hos­pi­tals in the Crimea. And we’ve been think­ing at the RSS about how can data help sup­port pros­per­i­ty. How can it help sup­port democ­ra­cy? And how can it help sup­port bet­ter policy?

And our answer was this, our lit­tle Data Manifesto. It’s got ten rec­om­men­da­tions, and it I don’t have time to take you through all of them in detail in the that time that I’ve got. But I want­ed to give you the head­lines. If you drop me an email or tweet me, I’m very hap­py to send you a copy of it. And here we’re talk­ing about the full range of the data spec­trum that has been described ear­li­er. Let me pick out five key themes that real­ly come out of the Data Manifesto.

The first is to invest to cre­ate and get the data that we need. And this par­tic­u­lar­ly means invest­ing in sci­ence, and research, and devel­op­ment. As a coun­try, we’re falling behind in this area at the moment, com­pared to oth­er coun­tries. And this is real­ly the source of long-term pros­per­i­ty. Now, good sci­ence does need to be open sci­ence. We’re see­ing an increas­ing prob­lem about the repro­ducibil­i­ty of sci­ence. If you actu­al­ly re-run exper­i­ments, you’ll find that you can’t get them to give the same results that you got the first time round. So the data must be open, and we must encour­age peo­ple to repro­duce the results of sci­ence. That’s got­ta be the quid pro quo, but we’ve got to invest in the first place.

Similarly, gov­ern­ment data. If we’re going to change the way the gov­ern­ment takes its approach to learn­ing about the world that we live in and the soci­ety we are, we need to invest in good-quality sta­tis­tics and data gath­er­ing. One exam­ple of that is the cen­sus. The gov­ern­ment has com­mit­ted to chang­ing the way we do the cen­sus by 2031, going away from the old mod­el which is gath­er­ing data once every ten years, to mov­ing to much more real-time data from your dri­ving license appli­ca­tions, from your GP records, and so on. But to make that switch, we’re going to have to have invest­ment and a new way of thinking.

The sec­ond is to open the data. I’m not going to say too much about that, because you already bought into this idea. But let me just pick out a few things which per­haps haven’t been talked about so much. The first is pri­vate sec­tor data. The pri­vate sec­tor’s got a crit­i­cal role to play. When we talk about open data, we’re often talk­ing about open­ing up gov­ern­ment data. But actu­al­ly the pri­vate sec­tor is hold­ing more and more impor­tant dates which needs to be opened up.

One of the areas the RSS has been cam­paign­ing is around phar­ma­ceu­ti­cal com­pa­nies. We’ve been work­ing with the AllTrials group which Ben Goldacre set up, look­ing at can we make sure that phar­ma­ceu­ti­cal com­pa­nies reg­is­ter their tri­als so that actu­al­ly you know the results that you’re get­ting from phar­ma­ceu­ti­cal tri­als are not skewed towards the pos­i­tive. Every tri­al is reg­is­tered and open.

Similarly, as we’re find­ing more and more pri­vate sec­tor com­pa­nies are deliv­er­ing pub­lic ser­vices, we’re not hold­ing them nec­es­sar­i­ly to the same stan­dards in terms of their account­abil­i­ty around data. So schools, hospitals…if it’s a pri­vate sec­tor com­pa­ny, our view is that they’ve got to be held to the same data stan­dards as the pub­lic sector.

And one oth­er thing I’d pick out, which I haven’t heard the open data com­mu­ni­ty say enough about is the recent threats to the Freedom of Information Act. That the gov­ern­ment has set up a new FOI com­mis­sion to review the Freedom of Information leg­is­la­tion. Hilariously, the FOI com­mis­sion is not FOI-able itself. The first meet­ing that it gave, jour­nal­ists had to redact var­i­ous things because they could­n’t even say who was giv­ing the infor­ma­tion to them. So, this slight­ly Kafkaesque pic­ture I think needs the open data com­mu­ni­ty to inter­vene and show that open data sits side by side with free­dom of information.

So, I’ve talked about invest­ing to get the data, open­ing up the data. The third thing is that you’ve got to have the skills to ana­lyze the data. And so, some­times we, the data kind of geek com­mu­ni­ty, for­get that what peo­ple want is not data. They want answers. And actu­al­ly, to go from data to answers, you need real­ly good ana­lyt­i­cal, sta­tis­ti­cal skills. 

We asked MPs, ​“If you toss a coin twice, what are the chances of get­ting two heads in a row?” And the answer of course is 25%. Only 40% of them could announce that ques­tion cor­rect­ly. Say you know, these are the peo­ple that are hav­ing to make deci­sions at coun­try lev­el on our behalves all the time. And their data skills and not nec­es­sar­i­ly where they should be.

So, off the back of that, we ran a cam­paign in the run up to the elec­tion. We asked our six thou­sand mem­bers to write to their can­di­dates and say, ​“If you get elect­ed, do you promise to take sta­tis­ti­cal train­ing from the Royal Statistical Society?” Three hun­dred can­di­dates said yes. Of them, fifty-five got elect­ed. And only a cou­ple of weeks ago on World Statistics Day (yes, such a thing exists) we held our first train­ing ses­sion for a batch of MPs. And they were delight­ed. Because it was­n’t about telling them ​“this is what you’re doing wrong.” It was say­ing, ​“We need to inoc­u­late you against the lob­by­ists and the kind of bewil­der­ing num­bers that you’re going to come across. Most of you dropped maths at 16. Data is scary for you. Let’s open that up and let’s build you confidence.”

But it can’t just be about MPs. We’re try­ing to work with the pol­i­cy com­mu­ni­ty. And we’re work­ing right the way through the school of edu­ca­tion sys­tem, and uni­ver­si­ty lev­el as well. Data ana­lyt­ics is the future of our econ­o­my and of soci­ety, and we’ve got to make sure we’ve got the skill base to inform that.

The fourth area we’ve we stressed is to actu­al­ly use the bloody data to inform decision-making. And it’s a scan­dal how lit­tle this hap­pens. Government has made a cou­ple of steps for­ward in this area. I think the What Works Centres, which are actu­al­ly look­ing at bring­ing togeth­er the data that we have in par­tic­u­lar areas and bring­ing them togeth­er with the pol­i­cy­mak­ing com­mu­ni­ty has been a real­ly pos­i­tive step for­ward. But this is in only small areas. So, open does­n’t end with open­ing up the data. The next step is to then shove it in the peo­ple’s faces and say, ​“What’re you going to do with it? This is what the evi­dence tells us.”

And final­ly, we need insti­tu­tions to cre­ate trust. Every week in the news now, we have TalkTalk, Ashley Madison, what­ev­er it might be. Now, these are not nec­es­sar­i­ly about open data, but as we already know, our per­son­al data, very impor­tant to us, is actu­al­ly… The scan­dal sur­round­ing some of this, and the con­cerns that peo­ple have around their pri­va­cy, has the abil­i­ty to actu­al­ly bring down the open data agen­da because in most peo­ple’s minds, these agen­das are not separate.

And so we have to have a robust insti­tu­tion­al frame­work to give peo­ple the con­fi­dence. We did some research about a year and a half ago show­ing what we said was a data trust deficit. If you ask any­body how much they trust any insti­tu­tion, and then how much they trust that insti­tu­tion with their per­son­al data, it’s always low­er. So there is this gap around per­son­al data, which is crit­i­cal. And if we don’t address that lev­el of trust, all of the wider ini­tia­tives that we have will skupper.

So, that was a very quick ground tour. But if you want to know more, our Data Manifesto is on the RSS web sites. We are cam­paign­ing for a bet­ter soci­ety based on good use of data, and I hope that— If you’re inter­est­ed, we’re open, an open body. You don’t have to be a sta­tis­ti­cian. You just tough to care about data. So, you’d be very wel­come to join us in our cam­paign. And we work very close­ly with the ODI in tak­ing all of this for­ward. Thank you very much.