VoteIt plans

Today’s MPs are no longer scared of the whips. Instead, they are scared of their constituents. That’s a good thing. — Peter Lilley

I’ve spent the last five years or so working with Parliamentary Monitoring Groups around the world, and getting repeatedly frustrated by how very very badly most of them handle presenting information on how MPs in their country actually vote. This information is hugely important — in many ways it’s the most basic concept in a representative democracy that you are delegating your vote to them, so it’s crucial that you can see what they do with it. However there are only about 50 countries who actually publish this information on a per-person basis (according to some amazing comparative research by KohoVolit) and very few groups have succeeded in making that accessible to anyone who isn’t already a political junkie or researcher. TheyWorkForYou does it better than anyone else I’ve seen, by grouping related votes (see, for example, this summary of David Cameron’s voting history), and though it doesn’t go anywhere near as far as I would like (you can’t compare two MPs votes; or see how a Party as a whole has voted; or see where one MP has differed from the rest of their Party; or look at one issue in isolation and see how MPs or Party have grouped on it; etc), it’s several orders of magnitude better than similar sites that simply provide a list of Bills (which largely mean nothing to an average visitor), with how that MP voted on them, or the really quite bizarre graph porn that too often passes for useful information these days.

Time and time again I’ve seen groups start out with good intentions to build something wonderful, but end up spending most of their budget (money/time) just getting the basics in place, before actually getting to the meaty stuff.

So I decided to see if there was something I could do about that. Enter VoteIt.

This is nominally within the Poplus Components approach, but it’s a multi-faceted beast, the majority of which should also be useful to people doing stuff outside of Poplus too.

Writing up the entire thing is a much bigger job than I can do justice to here, but I had to jot down the basics yesterday anyway, so I should really follow my own recommendations and share that publicly here too…

• Over the past few months I’ve been working with groups from a variety of countries to try to define a standard way of expressing legislative voting data (there are enough variations in the process to make this considerably trickier than I had originally hoped!) This will become part of the Popolo standard, and after a couple of excellent sessions about this at PoplusCon last week, the first draft should hopefully appear this week.

• Once that’s in place, hopefully lots of groups will start to model as much of their legislative data as possible in this format. As mentioned earlier, there are only about 50 countries who even publish it, so in the worst case scenario I’ll start scraping some of those myself, but ideally this should be done by groups in each country. My target is that within a year we have at least a few years of data for most of those countries, stored in a common format, and available over a consistent API.

• Once even a few countries have data in a common format, groups will be able to collaborate on shared tools to analyse this data, and make it better available. Some of that can be aimed at researchers, and political “professionals”, but my own interest is in tools for the general public. OpenAustralia are currently working on reimplementing Public Whip (the site that produces the TheyWorkForYou voting record data) in Rails — so hopefully it’ll become much simpler to at least reach that level everywhere (and ideally go much further)

• Within the Poplus framework it should then also become much simpler to build other Components on top of this, all of which will be able to work for any country who have gathered the data into this common format — e.g. recording statements that politicians say about why they voted certain ways; comparing votes to political manifestos, promises, etc; cross-referencing with income/interest declarations to look for potential conflicts of interest, etc etc etc.

• If we can also get groups to model their Bill processes in a consistent manner (e.g through BillIt), and, in particular, if we could get Parliamentary Monitoring Groups to add something equivalent to EuroVoc tags to key Bills1 (and in the EU case, the originating Directive, where appropriate), then we could also start doing interesting things with comparing cross-jurisdiction voting (e.g. whether Parties who are allied in the EU Parliament vote differently in national contexts on the same laws, etc).

 Thoughts welcome on what else could become part of this; traps along the way for the unwary; etc

  1. like most of this, this doesn’t need to be completist, and doesn’t need to be approved in some top-down manner; one group in each country simply tagging key bills in a consistent manner suddenly makes all sorts of interesting things suddenly become much more possible []

Leave a Reply

Your email address will not be published. Required fields are marked *