Gavin Chait hates walking past empty shops.
We’re talking about shops where the only person inside is a bored cashier, waiting for customers. Gavin sees it as a sign that the business should never have been set up in that location, and, more importantly, as something that’s completely avoidable.
With his company Whythawk, he’s on a mission to get that changed — and he’s using Freedom of Information to do so. It’s a very interesting case study that shows just how WhatDoTheyKnow, our Freedom of Information platform, can be used for the social good.
So, if you have a few minutes, sit back and watch Gavin explain what led him to make 350 FOI requests, one to each local authority in England and Wales — and what he did when many of them were turned down.
You can read more about the whole project at Pikhaya.com.
Thanks very much to Gavin for taking the time to talk to us.
Do you have a story to tell about how you’ve used one of mySociety’s sites? We’d love to hear from you: just drop us a line on email@example.com.
Researching in an unstable environment
It’s been nearly two years since the InfoLib Liberia project with iLab Liberia started. In that time the project has faced many hurdles, some predicted, and some completely unforeseen.
The iLab team have seen their country devastated by Ebola, only 11 years after the end of their second civil war, bringing tragedy and instability along with it. As you can probably imagine, the impact of curfews, fear and death in communities has made it difficult for people to continue with their daily lives. The social impact of such a disease is wide-reaching. Distrust, marginalisation and exclusion can be directed at those who show symptoms, or even who suffered and survived.
These are challenges that our local partners have had to contend with every day, both when holding training sessions and more crucially when researching the impact of the project on people’s lives.
However, by far the largest hurdle for this particular project has been a mixture of low internet penetration and lack of government will to release information. The team on the ground have been working tirelessly to create an ecosystem of requesting and training Public Information Officers (PIOs) to reply – even providing them with tablets to scan documents without needing electricity, let alone a computer. But if those officers have no access to the information that has been requested, their jobs become virtually impossible.
The project is now drawing to a close and we’re undertaking our final research survey. It seemed like a good time to take a look at what we’ve learnt about the impact of our joint Freedom of Information project in Liberia.
When designing the project we decided that impact could best be measured in terms of whether or not the project increased confidence in government transparency.
We carried out surveys in January 2016 and April 2016, to provide a baseline picture and then an assessment of impact at midline. The final survey is being conducted in August 2016 just as the project ends.
The first survey – the baseline – was carried out mainly in the rural areas. iLab Liberia teamed up with LFIC to survey 152 participants who had been involved in the FOI workshops that LFIC had held in the counties.
We had to attempt the second survey twice, as it turned out to be more challenging than we’d expected. We needed the participants from the first questionnaire to answer the same questions we’d asked them initially, in order to measure change — but it proved hard to locate all of them.
There were many factors which caused this, but the main one was economic drivers, forcing people to move to where the opportunities are. It’s a problem many researchers must run into working in the field.
Carter, the project lead at iLab Liberia told us:
“There are several reasons why this happens […]. People migrate a lot between markets, farms. Several persons who participated in the baseline could not be reached as they [had] travelled to other cities/counties. [Or] the job that allowed them to reside in that city/county is no longer available so they might have left seeking after another job.”
Our second attempt was more successful. We managed to contact a large percentage of the original participants in the survey: 112 of the 152.
We’ve found out some interesting things from doing this research. We saw that 74% of people who use the internet daily say it’s their main source of information, though it is still only a small percentage of the population who have access to the internet.
So the next biggest source of information? Radio! 85% of people with with no access to the internet give radio as their main source of information. Thinking of the migration of workers between cities and counties – you suddenly appreciate why Radio is such an important medium for getting hold of information. Thankfully, as you’ll remember from our original blog post, we’re covering both of these media in the InfoLib project.
In the months since we began studying the impact of this project we also learned that fear of making a request has dropped by 5% in the individuals surveyed . The amount of people who reported that they didn’t know how to ask for information dropped from 24% to 21%. This is pretty great news to us as it shows that our training and our encouragement is working – albeit slowly.
Finally we saw the percentage of people who believe government would be more transparent if citizens could see the information they hold rise by 3% to 93% of the surveyed respondents. Even if this figure hadn’t risen, this demonstrates a clear existing demand from the citizens of Liberia for the Government to release more information about its activities which is great news overall!
No project is without its challenges, and as you’ve seen above one of the big ones is ensuring that the same people respond from survey to survey. Not being able to pin down precisely the same set of people means that we can’t say with 100% certainty that we have a true measure in the difference in attitude.
As a result of the economic and social drivers mentioned above, the workforce in Liberia is very transient. This makes disseminating information through radio and internet mediums even more important. This research has shown that these are the primary sources of news and official information for the majority of Liberians, and continuing to improve knowledge about, and access to, information via these sources will empower the population further.
Finally, it can be challenging to demonstrate impact in projects like these, simply because research is not the main focus for our local partners. We partner with local groups because they are passionate, capable, and able to engage and mobilise citizens around a certain issue. We cannot expect small grassroots groups to have the resources or experience to conduct academic surveying, sampling or interviewing that could detect and definitively isolate the short term impact of a small project. This piece of research has provided some encouraging interim results, but most of all, it has provided valuable lessons to us at mySociety in trying to conduct this kind of impact research remotely and in partnership.
While we wait for the outcome of the final survey we can feel cautiously hopeful that this project has caused a small change in the way access to information operates in Liberia. infoLib will continue to run after the project officially ends, and mySociety will continue to support the work that iLab does in this area . However it may take longer than we had expected or hoped, to see the governmental shift towards releasing information.
We’ve released Alaveteli 0.25! Here are some of the highlights.
Visible delivery status
Gareth and Zarino have added a delivery status feature that shows whether a message has been received by the authority’s mailserver. This should provide reassurance for site users that messages are getting through and makes it difficult for an authority to successfully claim that they didn’t receive the request.
Clicking on the delivery status indicator reveals a bit more detail about the status itself. Admins are shown more detail here including relevant mail logs to diagnose problems or provide proof to the authority if required.
We’ve upgraded from the so-called “Legacy” (ga.js) version of Google Analytics to Universal Analytics. For most Google Analytics users there’s nothing to do here except sit back and enjoy continued technical support and new feature rollouts from Google but if your Alaveteli theme has custom analytics scripting, you should check Google’s upgrade guide as well as our upgrade notes to see if you need to make changes. If you’re not ready to move to this release yet, don’t panic – you may not get any shiny new features from Google but they haven’t published an end date for support yet.
In addition to spam email, there’s been an increase in the number of accounts that create profiles containing spam links – presumably to boost their search engine ranking score rather than to trick people into clicking through from the site. Having spent some time going through accounts on WhatDoTheyKnow to look for patterns, we’ve added some tools to this release to try to discourage this use of Alaveteli and to make it easier for admins to discover and ban offending user accounts.
(We also looked at extending our reCAPTCHA use for new account signups but this didn’t seem to help so we are not offering it to reusers.)
Martin has been working away on improving page load times and accessibility compliance to make the pages faster to load and easier to navigate. (A process we’re continuing into the next release.) We’ve also updated the help template code so that the examples are in the example theme rather than the core code and added a rake task to help check whether your theme implements the help pages correctly.
The full list of highlights and upgrade notes for this release is in the changelog.
Thanks again to everyone who’s contributed!
This is Part three in a blog post series highlighting information that has been disclosed thanks to Freedom of Information (FOI) requests on Alaveteli sites across the world. Here are part one and part two.
Australian police use banned restraint technique on asylum seekers
Detention Logs is a project that publishes data, documents and investigations that reveal information on conditions and events inside Australia’s immigration detention network.
As part of the project, Detention Logs used Alaveteli site RightToKnow to ask the Australian Department of Immigration to disclose incident reports from detention facilities.
One such incident report revealed that the Department of Immigration approved the use of a controversial body lock technique on an asylum seeker.
The incident report describes the restraint as follows:
“The seat belt was fastened. The client and the escort staff were the first passengers to board the plane. As the client continued screaming and resisting, DIAC staff issued instruction to the escort staff to ‘lock the client down’ by pushing her chest towards her knees. However, the client still continued screaming loudly and attracting attention of the cabin crew.”
According to New Matilda and Detention Logs, the account of the restraint strongly resembles the “seated double embrace” technique banned by police in Victoria and New South Wales in Australia, and some government agencies in the United Kingdom.
The United Kingdom Ministry of Justice banned this technique in juvenile detention facilities, following the death of 15-year-old Gareth Myatt in 2004.
Why the EU’s taking its time on restricting harmful chemicals
Endocrine disrupting chemicals (EDCs) are chemicals that are present in everyday products – from plastics and cosmetics to pesticides. Because of their ability to interact with the hormonal (endocrine) systems of living organisms, they are suspected of having serious health and environmental impacts.
The European Union is supposed to regulate EDCs, in order to protect citizens from harmful effects. Both the EU’s 2009 pesticide regulation and the European chemicals package (REACH) demand that the EU take action on these chemicals.
However, several years later, the European Commission is still no closer to taking concrete action.
The documents have helped journalist Stéphane Horel and research and campaigning group Corporate Europe Observatory (CEO) to piece together why the process has stalled, and what has been said behind closed Commission doors.
To communicate their findings, they produced a book and documentary, which explains how corporations, and even actors within the Commission, are stalling the process of this key public health and environment legislation.
Now Rwandans can find out when their parliament is in session
The Rwandan parliament now publish the Chamber of Deputies’ schedule of debates/activities for each day on the homepage of their official website (under ‘Today in Parliament’).
This follows an FOI request on Rwandan Alaveteli site Sobanukirwa, which urges the parliament to do just that. The Sobanukirwa team don’t know for sure if their request influenced parliament’s decision to publish this information, but we are pretty sure it had an impact.
The Australian and EU examples above show the power of making a lot of FOI requests, across authorities, and then grouping them together to create/support a project or investigation about a wider issue of public concern.
The Rwandan example shows that perhaps, just perhaps, a single request can cause a big impact.
So, what do you want to investigate? What have you always wanted to know? Start your own investigation to unearth truths that may just surprise you, and use an Alaveteli site to help you.
If you know of any interesting requests made on Alaveteli sites (or other online FOI portals) that you’d like featured in this blog post series, then please do get in touch.
We recently made some changes to our Freedom of Information software, Alaveteli, that makes it twice as easy for partners to change the look and feel of their site. This post goes quite in-depth, so it’s probably best suited for designers, developers and other technically-minded readers
This year we embarked on a ground-up redesign of Alaveteli’s theming capabilities to make customising Alaveteli easier and faster. We used a mixture of techniques to try to achieve this and since releasing the update we’re happy to have seen a decrease in the complexity of our themes, and the time we spend making them.
Alaveteli, our Freedom of Information platform is one of our most deployed software packages and each deployment requires some customisation, from simple branding changes to complex functionality to support a country’s FOI process. This is either done internally by our own developers, or our partners take this on themselves.
An Alaveteli deployment consists of two things, the core: the guts of Alaveteli with no customisations, and the theme: the aesthetic and functional customisations made to each site. We ship Alaveteli with a fully-featured base theme that gives partners a good starting point to work from, but they can also start their own theme from scratch.
After talking to our partners and the Alaveteli team we felt that the process of theming was too complicated and involved. It required a lot of knowledge of Alaveteli’s inner workings, was coded in a way that made it hard to understand, and almost always required the intervention of a front-end expert, even to make trivial changes.
Our vision for how Alaveteli’s theming should work was:
- Intuitive Front-end developers and designers should be able to follow and understand the templates without needing expert knowledge about Alaveteli’s inner-workings.
- Flexible Overriding Alaveteli’s core code should be easy, so a theme should use low-specificity, componentised code to enable this.
- Supportive Should a partner decide to only change a few colours, or embark on a large-scale redesign Alaveteli, the theming process should never get in the way.
How we did it
Using the tools we have
Alaveteli is built on Ruby on Rails, so our first goal was to make better use of Ruby on Rail’s template inheritance.
Firstly, we split all of the Sass files into logical components.
Each component has two associated Sass files, a layout.scss and a style.scss. layout.scss describes only layout of a component, e.g. the physical dimensions and position on a page. style.scss describes the styles, for example typefaces, colours, borders and other presentational properties.
Should a partner wish to totally redesign or replace a component, they can replace both files with their own. Alternatively they can just choose to replace one part of the component, for example keeping the layout, but replacing the styles.
Ensuring the lowest possible specificity
When customising Alaveteli we want to be sure that a partner never has to fight with the base theme to achieve their goals. The most important part of this is to ensure CSS styles are easy to override.
Firstly we removed all ID selector’s from Alaveteli’s stylesheets and replaced them with class selectors, giving us a good jumping off point as class selectors have much lower specificity than IDs. We then rewrote and refactored Alaveteli’s CSS using the Block Element Modifier methodology (BEM).
BEM helped us in two ways: it’s written in a way that makes CSS self-describing, for example the code snippet
.header__navigation__listtells us that this list is a child of the navigation element and a grandchild of the header element.
As well as this, using a single class to define the styles for this element means that using BEM gives our styles the lowest possible specificity, so partners can override them with little resistance.
Using a settings partial
All of our theme’s styles use the variables defined in _settings.scss when referencing colours, typefaces, spacing, and even logo files so the bulk of Alaveteli’s customisation can be done in our base theme’s settings partial _settings.scss.
This allows partners to make simple changes which have a huge impact on the look and feel of the site. We’ve found that the large majority of our theme customisation can be done by just changing the variables in this file and it has sped up the theming process markedly.
We’ve quietly rolled out these updates to Alaveteli over the past 12 months and the impact of the changes has been encouraging.
Internally, we’ve found our theming projects are faster and more efficient. We conservatively estimate that it takes half the time to make a new theme than it used to.
Thanks to the self-documenting nature of BEM and the styles encapsulated in our settings partial, Alaveteli theming requires a lot less understanding of the platform’s inner workings. Our partners are able to be more self-sufficient and rarely require our intervention to assist with simple customisations.
Maintenance and support of our themes is greatly simplified, as the structure is consistent across projects, we’re able to roll out further changes and fixes in a more predictable way.
Overall, we’re really happy with how the changes worked out and the things we’ve learnt from this project we’ll be carrying into our others too, so look out for changes like across our work in the future.
This is Part Two in a blog post series to highlight interesting and impactful requests that have been made through the 25 websites running our Alaveteli FOI software across the world. You can see Part One here.
The European Commission really doesn’t want us to know about their correspondence with tobacco lobbyists
Last year, the research and campaigning group Corporate Europe Observatory (CEO) sent a request for documents relating to all correspondence/meetings between DG Trade officials and tobacco lobbyists from January 2014 to March 2015 on AsktheEU.org.
Following this, the European Commission ‘released’ very heavily redacted documents concerning their contacts with the tobacco industry on EU trade negotiations, including the ongoing EU-Japan and EU-US trade talks (TTIP).
In all four documents (correspondence with and minutes of meetings with tobacco lobbyists) virtually all the content is blacked out, including the names of all tobacco lobbyists and Commission officials involved. Only around 5% of the text is visible.
As CEO points out, the Commission’s secrecy around its relations with tobacco industry lobbyists and its international trade negotiations causes great concern and highlights the lack of transparency within the Commission.
To read the full story, see this article on CEO’s website.
There are now Freedom of Information websites running on our Alaveteli software in 25 jurisdictions worldwide, which between them have processed more than 330,000 FOI requests.
But what sort of information is being revealed through these sites? And what impact has this information had? In our new series of posts, we’ll be giving you a roundup of some of the most interesting and impactful requests made on Alaveteli sites from across the world.
mySociety’s flagship Freedom of Information (FOI) request portal WhatDoTheyKnow.com, operating in the UK since 2008, has amassed a whopping 330,000 FOI requests (and counting) from citizens over its eight year life-span.
That equates to approximately 15-20% of all FOI requests made in the UK. It also represents the largest database of FOI requests in the country, having provided a platform for requests and responses to over 17,000 UK public authorities to be published publicly online.
Those are some impressive numbers: however, until now we haven’t known much more about what requests are being made, whether there are trends, or indeed, whether the responses that people are receiving are satisfactory.
We thought it was about time that we took a look under the bonnet of WhatDoTheyKnow to find out the answers to some of these bigger questions.
We decided first to look at what themes and policy areas people were most interested in when making an FOI request. We chose this area because we suspected that many people would be asking for similar things from similar authorities. If this is the case, then this would be a clear evidence-based argument for authorities to increase pro-active publication of certain information.
The task itself was not an easy prospect. WhatDoTheyKnow does not have a tagging or categorising system, so there are over 330,000 requests that we had no quick or easy way of comparing. The volume of data was also so high that we couldn’t reasonably extract every request and analyse which policy area(s) it was relevant to.
To solve this issue, we decided to focus on the 20 authorities receiving the highest volume of FOI requests between 2008-2016. This way we could rely on a large sample of requests for both both local authorities and government departments. The list of authorities is below.
Department for Work and Pensions
Department for Education
Wirral Borough Council
UK Borders Agency
Birmingham City Council
Brighton & Hove City Council
Liverpool City Council
Transport for London
Westminster City Council
Ministry of Defence
Metropolitan Police Service
Bristol City Council
Ministry of Justice
Lambeth Borough Council
Camden Borough Council
Department of Health
Kent County Council
Taking all the requests made to these public bodies gave us a total of 49,500.
With the generous support of Thomson-Reuters, we were able to use OpenCalais, their automated tagging system, to assign one or more thematic tags to each FOI request made. Over 100,000 hyper-granular tags were automatically applied in this way.
Once that was complete, we went through each tag and the requests it was associated with. We grouped tags into policy areas and checked for any that had been incorrectly assigned. We then split the authorities into two groups: Local Authorities and Departmental Bodies, to compare the most requested information.
Among Local Authorities, the top requested information concerned:
- Housing Specifically, information on social housing stock/occupancy/waiting lists, facilities for homeless and at-risk individuals, and planning permission
- Social Care Information concerning care providers and their funding/operations, care in the community arrangements, social worker qualifications and staffing levels, and information concerning the operation and monitoring of social work departments
- Accounts and Budgets Citizens commonly request accounting and budgetary information at a far more granular level than authorities are currently publishing.
- Authority management Citizens also wish to know with greater detail how authorities are operating internally, requesting management and meeting information, emails about decision-making, and structural information concerning development, contracting and relationships with private companies
- Business rates Concerning long-term empty properties, the impact of rates on town centres, charitable or other discounts, and regeneration.
These are the top five of thousands of tags, but common themes were clear when comparing these authorities.
Generally, requesters have shown they want information in a more detailed form than authorities are currently providing, in particular in the expenditure of public funds and those services catering for complex cross-cutting social issues. Given the ongoing housing crisis in the UK, coupled with the ageing population, it is likely that information concerning these policy areas will be in increasing demand.
Conversely, among Departmental Bodies, the top requested information showed few common themes. This is primarily due to the differences in policy areas between the departments. There were, however, significant spikes in certain policy areas within departments, particularly around immigration, and this will be the focus of future investigation.
In conclusion, we understand that very few FOI requests are completely identical in subject matter, but broad trends are clearly evident.
If Local Authorities proactively publish more granular information about the policy areas we now know citizens are actively interested in, they may see a dip in formal FOI requests.
Publishing information and data in a machine-readable format may even enable other civic technologists like ourselves to build tools to assist councils in their delivery of vital services. Whilst this will not eradicate FOI requests completely, it would hopefully begin a shift in behaviour.
In short: wouldn’t it be great if, instead of authorities seeing FOI as an administrative burden, they began to see pro-active publication as an opportunity to harness the skills, initiative and flexibility of citizens.
If you’re a UK citizen, it probably won’t have escaped your notice that we have a rather important vote coming up.
On June 23, a referendum will decide whether or not we remain in the European Union. It’s a divisive subject, with strong advocates and emotional arguments on both sides. But here at mySociety, we know what we believe.
We believe in an informed vote.
That’s why we advise you to analyse the facts before making up your mind where to place your cross. And to help you do that, here’s a list of impartial resources, from us, from our partners, and from other organisations.
Check the facts
Just as they did for the UK general election, our friends at Full Fact will be setting out the truth behind the emotive speeches, claims and counterclaims around the referendum. Here’s where you can find all their EU analysis.
They started off with a good check of the government’s EU leaflet.
Ask some questions
Wondering about something specific? Or perhaps you’ve seen claims flying about on social media which you’d like to check for accuracy. In some cases, a Freedom of Information request will help you source the facts and figures you need to understand the truth.
But hurry: by law, requests to the EU can take up to 30 working days to process (20 in the UK) and in actuality they often take longer.
Know where to vote
Of course, for the referendum, there are no candidates — but you do need to know where to vote. Democracy Club’s Open Polling Stations project is attempting to make that information easier for everyone to locate: you can start by inputting your postcode on WhoCanIVoteFor. Where they don’t have the polling station data, you’ll see a phone number for your local council.
Many care worker positions require regular overnight shifts. Depending on the job, you might be there ‘just in case’, with an expectation that you’ll generally get a night of uninterrupted sleep; or you might be there specifically so that you can respond to regular calls for help from clients.
Either way, you’ve been contracted to be away from your home, ready for work if needed. That’s why the law states that the national minimum wage should apply to sleep-in shifts — but as this user discovered through a systematic series of Freedom of Information requests, many councils fail to meet this standard.
It’s always good to see WhatDoTheyKnow being used to uncover this kind of important data. You can read more, check how your own council fares, and see the conversation unfold via some interesting users’ comments on Reddit, or see the original requests on WhatDoTheyKnow here.