One of the frequent grumbles I hear around Westminster is about the website TheyWorkForYou, which aims to provide citizens with information about how MPs vote on legislation. A debate about the use and future of the website has erupted after John Ashmore, the editor of the centre-right news site CapX, suggested it should be shut down.
TheyWorkForYou is a very good illustration of GK Chesterton’s quip that “if a thing’s worth doing it’s worth doing badly”. It’s vital that people are able to scrutinise the voting record of their MP; TheyWorkForYou facilitates this, although often pretty poorly. But a bad TheyWorkForYou is better than no TheyWorkForYou.
TheyWorkForYou and the semi-affiliated website The Public Whip, from which it scrapes data, were both established in the aftermath of the Iraq War and votes on university tuition fees. Although “division lists” – lists of who has voted on what and when – have always been published, before these websites, there was no reliable, publicly available archive of them.
The Public Whip provides that archive, but it isn’t, and doesn’t aspire to be, particularly accessible to the lay reader (I use it frequently for work, but I have to lean heavily on other sources of information to understand it properly). TheyWorkForYou aims to take the information on The Public Whip and make it accessible – no easy task, when the budget is limited and much of the process is automated.
TheyWorkForYou is good at differentiating between MPs’ voting records on foreign and security policy – the main dividing issues within the Labour Party when the website was established in 2004. It is worse at differentiating between socially conservative and socially liberal MPs, and its record is fairly uneven when it comes to differentiating between MPs on most of the more recent political divides, such as Brexit, the climate crisis and equalities issues.
So if you want to tell the difference between the dovish Conservative MP John Baron and the hawkish George Osborne, then TheyWorkForYou is fairly reliable. But it provides less nuance when it comes to public policy issues: for example, TheyWorkForYou describes not only Baron and Osborne, but the Labour MP John McDonnell and the Green MP Caroline Lucas as having “Voted a mixture of for and against greater regulation of hydraulic fracturing (fracking) to extract shale gas”. Why? Because McDonnell and Lucas voted against government measures to regulate fracking, believing them to be inadequate, and voted for opposition amendments to make them tougher. It is thus strictly accurate, but entirely misleading to describe Lucas and McDonnell as having voted “a mixture of for and against” measures on fracking. Anyone looking at TheyWorkForYou would conclude that Caroline Lucas was about as anti-fracking as the median Tory MP, which patently isn’t the case.
This is, in part, because of TheyWorkForYou’s age. Under New Labour, opposition and back-bench amendments from the Conservatives tended to shift legislation away from its intention – to water down or blunt government initiatives. This made it easier to work out how to automate what votes “meant” because they were ostensibly consistent. It is quite easy to interpret the meaning of a Conservative MP voting for an amendment to gut the functions of the devolved parliaments and then against the creation of them in 1999. It is slightly more difficult, but still relatively easy, to work out the overall inclination of a Labour backbencher who votes for an amendment to add further scrutiny to a draconian bit of anti-terror legislation and then votes against the bill.
But under the Conservatives, opposition and back-bench amendments have tended to make government legislation more radical, leading to what can appear to be contradictory voting patterns: an MP might vote for a full-fat version of a government proposal in an amendment, only to vote against the diluted version offered by the government. It is a lot harder to work out how to automate an MP’s voting to toughen regulations on fracking, and then voting against a bill which contains some regulations on fracking that they believe to be deficient. A top-of-the-line voting record website should not, as TheyWorkForYou does, describe MPs who have consistently voted against or abstained on any measure to increase the rights of LGBT people (such as the right to have a civil partnership) but have voted for “cleaning up” exercises (such as the right to get divorced in a civil partnership) as having voted “for and against” LGBT rights.
Improvements have been made to TheyWorkForYou since the site’s creation. Under Labour, the preferred tactic of socially conservative MPs was simply to give votes on equality issues a miss, in part because during the Blair and Brown years, many of these votes were on a so-called “two-line whip” (i.e. if you don’t want to vote for this, don’t turn up; but if you do turn up, you’d better vote for it). By contrast since 2010, under the Conservatives, socially conservative MPs across the House are much more likely to turn up and vote their consciences, because votes on equality issues have tended to be treated as “free votes”. As a result, TheyWorkForYou has got better at differentiating between socially conservative and socially liberal MPs over the last decade, simply because the former group has been voting rather than abstaining.
The small team at MySociety, the civic participation charity which runs TheyWorkForYou, is incredibly responsive when these problems are flagged and has implemented a number of spot fixes (for example, there is now a category for “policies contained within the coalition agreement”) to ameliorate this issue. But they’ve yet to find a way to properly resolve it. But when you start looking at how to improve TheyWorkForYou, you quickly run into resource constraints: you need to be able to make difficult judgement calls about what an abstention means; whether a vote can be considered a “cleaning up” exercise; how to assess a government bill and an opposition amendment that might be said to do the same thing. All of these cases essentially require a human to be on call to make these judgements and to explain to users of the site what these judgements mean, why they’ve made them and what the implications of making different choices would be.
For example, I think the government’s Police, Crime, Sentencing and Courts Bill should primarily be seen as a piece of legislation that severely limits the right to protest and effectively criminalises traveller communities simply for existing. But you could also argue that the Bill primarily lengthens jail sentences and clamps down on a variety of serious crimes. How you define the Bill, and thus assess an MP’s vote for it, is a big call, but one you have to make if you wish to communicate what votes mean “in plain English”, as TheyWorkForYou aims to do.
Although TheyWorkForYou often falls short of its objectives, and sometimes gives users a false level of certainty about what their MPs’ voting records mean, a world without it would be much worse and much less transparent. The existence of TheyWorkForYou has helped drive parliament to become much more transparent online; it has helped make Hansard, the record of everything that is said and done in parliament, easily searchable on the parliamentary website; it has provided easy-to-use archives of MPs’ written and spoken contributions, and a more accessible record of their voting history.
I think it is unlikely that, in a world without TheyWorkForYou, we would have instead a new and accessible UK parliament website, a searchable Hansard, a free online version of Erskine May, the guide to the rules and procedures of the House of Commons, and much else besides. TheyWorkForYou could be more transparent about its limitations, and could be better at directing people to the raw figures on The Public Whip.
After the launch of TheyWorkForYou, it took 17 years for parliament to set up its new, more transparent website, and 15 years to create its handy CommonsVotes app. That it took so long does, I think, show there is some value in the imperfect service provided by TheyWorkForYou.