Post mortem: dissecting letters from experts

(Get it? POST mortem? Because it came by… po- never mind.)

This is the unexpected third part of what has turned into a disinformation trinity. I hadn’t expected the first part in late November (the Conservatives’ factcheckUK saga), I was marginally less surprised at the second instalment in early December (the three big parties’ “local newspapers”), and by this third volume, I have hit a wall of weary acceptance. A single instance can be discounted as anomalous, but three is a pattern. A tiresome, dangerous, corrosive pattern that suggests a much wider comfort with behaviours that the electorate find problematic than I, at least, had previously thought.

So what’s the story for this latest and (hopefully!) final chapter before the General Election strikes? Somewhere throughout November 2019, the Liberal Democrats started sending out letters. Here are two exemplars – one that is fairly representative of those distributed in England, and one that seems to be fairly representative of those sent in Scotland.

The letters sparked a flurry of tweets that you can find by searching for phrases like “Mike Smithson” and “polling expert”, with plenty of examples of individuals considering these letters to be very problematic, e.g.

And so on.

The cast of characters

Our story today has a cast of two characters.

Let’s deal with the obvious one first. Prominently positioned top-right, bespectacled and smiling, the bio-picture is captioned with…

Mike Smithson

Polling and Elections Expert

According to Wikipedia, Smithson was born 1946 and is therefore around 73 as I write this. He spent his twenties working for BBC News as a journalist. In his early forties, he joined the newly formed Liberal Democrats, and four years later, in the 1992 General Election, he contested the seat for North Bedfordshire as a Liberal Democrat candidate. Ultimately he came in third behind the winner, Trevor Skeet (Conservative) and the runner-up, Patrick Hall (Labour). In his late forties Smithson shifted to university fundraising directorships, and then in 2004, in his late fifties, he founded the gambling website, Political Betting. (Amusingly, this site describes itself as “Britain’s most read political blog”. I’d be very intrigued to know how one measures the quantity of reading done by any given site visitor, since the duration that a page is open tells you precisely nothing, and most computers are not fitted with eye-trackers, let alone ones that send reading-time statistics back to random blogs, but… I digress.)

In 2007, now in his early sixties, Smithson released a book on how to make money betting on politics, and two years after that, he joined Twitter. Since then the account has posted around ten (re)tweets every day ever since, and in 2014 it was described as one of the most influential over-50s accounts on Twitter.

What I haven’t been able to uncover is Smithson’s academic credentials. His Wikipedia page notes that he is “a former fellow of Queens’ College, Cambridge and Magdalen College, Oxford”, but that’s difficult to nail down in terms of substantive academic achievement. Does this indicate a bachelor’s degree? A Master’s (as they’re understood outside of the weird Oxbridge system)? Something else? Any degrees were likely awarded somewhere around the late 1960s, and I’ve searched as best as one can for degrees obtained half a century ago, but to little avail. He could have a PhD or a post-16 diploma. I honestly couldn’t say. I should stress here that I in no way suggest that one cannot become an expert without going through some sort of formal academic system. There are all sorts of fields for which academia is singularly unsuited, and the number of successful entrepreneurs with little to no formal educational certifications is substantial proof that there are many roads to excellence. However, later in the analysis I’ll refer back to the stereotypical notion of expertise and this will come up again.

Overall, then, Smithson was a journalist, an attempted-politician, and a devotee of the Liberal Democrats who has since evolved into a betting site owner and editor, and whose credentials appear to be predicated mainly on a lifelong combined experience of politics and gambling. He may also have directly relevant academic experience, but if that’s the case, I can’t find any evidence of it.

So, this is the individual who supposedly sent these “expert letters”. But what was it about the content of the letters that proved so contentious, and triggered numerous individuals to describe them as misleading, manipulative, or even outright deceptive?

Let’s take an examplar apart line by line…

Putting the “er” in “expert”

Alright, so we already know that the letter opens with this claim that Mike Smithson is a “Polling and Elections Expert”. The use of the term “expert” is a first catchpoint. Whatever dictionaries say, rightly or wrongly, there is a chance that some proportion of the readership of this letter will equate “expert” with academic credentials. Whatever the case, people also associate experts with neutrality. In a search for the term in a large dataset, it frequently hangs out close to words like…

  • arbitrator: an independent person aiming to impartially and fairly resolve a problem
  • determination: an expert determination is a process of resolving disputes, e.g. through arbitration
  • witness: an expert witness usually provides impartial evidence in a criminal or civil case
  • tuition: the teaching of others in a subject, art, or skill
  • opinion: an expert witness’ formal and impartial findings in a case

…and so on. In short, the word “expert” carries a strong sense (semantic prosody) of neutrality, objectiveness, and impartiality. It’s also presented in bigger font, and it’s at the top of the page, so it’s the first indication of the authorship of this piece, and that tells the reader how to interpret what follows.

Why is this important? Well, It’s one thing, for instance, to be told that the small fire in your kitchen is extremely dangerous by your nervous neighbour who thinks that caffeine after lunch is a borderline safeguarding issue, and another thing to hear the exact same words from a fire fighter who is an expert in gas explosions. The degree of social credit we extend to an author is heavily predicated on how impartial, informed, and reliable we deem them to be, and this letter starts out with a strong intimation that the apparent author is well endowed with both objectivity and expertise.

Between night and shade

What then? Depending on the letter, the next line reads…

The election in [____________] is between the Labour [sic] and the Liberal Democrats.


The election in [____________] is between the Conservatives and the Liberal Democrats.

(You can see how they used some sort of mail-merge system here and forgot to account for the determiner before Labour. But again, I digress.)

So, those two fickle words, “is between”…

I did a corpus search to see what sort of constructions that phrase brings up, and you’ll be shocked to know that it tends to represent two “contestants”. Sometimes they’re points on a numerical scale (between £300 and £500). Sometimes they’re physical locations (between Epping and Surrey). Sometimes they’re conceptual phenomena (science and religion, hope and disaster). And often, they’re groups (men and women, firms and markets, Marxists and non-Marxists). Essentially, however, they all function on the basis of being two distinct poles. Why does this matter? Well, when it comes to competitive formulations in particular, the construction “between [X] and the Liberal Democrats” sets up a presumption that there is no third opponent, or at least, not one worth mentioning. It suggests that the two named are in direct contention with each other, and they encapsulate the most immediately relevant participants.

If I suggest, for instance, that I’m taking part in a 100m sprint race that is between me and Usain Bolt, it would perhaps surprise you to then learn that Shelly-Ann Fraser-Pryce (the fastest woman in the world) is also taking part. You’d probably feel compelled to gently point out that this race is really between Usain and Shelly-Ann, and I’ll be lucky to be dead-level with them at the starting blocks. Essentially, I’m not a relevant competitor in any “between [x] and [y]” construction about that race.

So, by phrasing the apparent competition as between [x] and the Liberal Democrats, the Liberal Democrats are now positioned as one of the two most competitive options for this constituency in the upcoming election. And you’re being told this by an expert in polling and elections. Who are you to argue? Are you an expert in polling and elections? Do you have time to fact-check this random letter? Or will you just take it as presented that this information is accurate?

Worry not, I did the heavy lifting for you. Testing whether an election outcome is “between [x] and [y]” is difficult, because it’s essentially trying to guess at how people will vote before they’ve actually voted, and as we know, there have been a lot of developments since the 2017 General Election (hereon GE17). However, what we can do is look back on the GE17 results to see how the targeted constituencies voted historically. I could’ve even gone back to GE15 but I value my time and sanity.

What of the constituencies in this case? Well, I currently have a dataset of twenty-two [update: twenty-three] letters from 2019 (you can access that here along with other examples from 2018 and 2017 but note that the link will expire on the 06th of January 2020). Notably, two of the Scottish letters are more generic than the rest, having no reference to their specific voting locale, so that left me with twenty-one constituencies that I could objectively check against past results. Based on the GE17 results, I graded how competitive the Liberal Democrats were on a +/-3 scale. You can find the key for that at the bottom of the table.

Here’s the outcome:

Altrincham Sale West Conservatives CON: 26,933 (51%)

LAB: 20,507 (38.8%)

LDEM: 4,051 (7.7%)

Battersea Labour LAB: 25,292 (45.9)

CON: 22,876 (41.5%)

LDEM: 4,401 (8%)

Chelmsford Conservatives CON: 30,525 (53.7%)

LAB: 16,953 (29.8%)

LDEM: 6,916 (12.2%)

Cheltenham Conservatives CON: 26,615 (46.7%)

LDEM: 24,046 (42.2%)

LAB: 5,408 (9.5%)

Chipping Barnett Conservatives CON: 25,679 (46.3%)

LAB: 25,326 (45.7%)

LDEM: 3,012 (5.4%)

Cities London Westminster Conservatives CON: 18,005 (46.6%)

LAB: 14,857 (38.4%)

LDEMS: 4,270 (11.0%)

Edinburgh South Conservatives LAB: 26,269 (54.9%)

SNP: 10,755 (22.5%)

CON: 9,428 (19.7%)

LDEM: 1,388 (2.9%)

Esher Walton Conservatives CON: 35,071 (58.6%)

LAB: 11,773 (19.7%)

LDEM: 10,374 (17.3%)

Finchley Golders Green Conservatives CON: 24,599 (47%)

LAB: 22,942 (43.8%)

LDEM: 3,463 (6.6%)

Guildford Conservatives CON: 30,295 (54.6%)

LDEM: 13,255 (23.9%)

LAB: 10,545 (19%)

Hitchin Harpenden Conservatives CON: 31,189 (53.1%)

LAB: 19,158 (32.6%)

LDEM: 6,236 (10.6%)

North Devon Conservatives CON: 25,517 (45.8%)

LDEM: 21,185 (38%)

LAB: 7,063 (12.7%)

Oxford West Abingdon Conservatives LDEM: 26,256 (43.7%)

CON: 25,440 (42.4%)

LAB: 7,573 (12.6%)

Penistone Stocksbridge Labour LAB: 22,807 (45.8%)

CON: 21,485 (43.2%)

UKIP: 3,453 (6.9%)

LDEM: 2,042 (4.1%)

Portsmouth South Conservatives LAB: 18,290 (41.0%)

CON: 16,736 (37.6%)

LDEM: 7,699 (17.3%)

Putney Conservatives CON: 20,679 (44.1%)

LAB: 19,125 (40.8%)

LDEM: 5,448 (11.6%)

Twickenham Conservatives LDEM: 34,969 (52.8%)

CON: 25,207 (38%)

LAB: 6,114 (9.2%)

Warrington South Labour LAB: 29,994 (48.4%)

CON: 27,445 (44.3%)

LDEM: 3,339 (5.4%)

Warwick Leamington Conservatives LAB: 25,227 (46.7%)

CON: 24,021 (44.4%)

LDEM: 2,810 (5.2%)

Watford Conservatives CON: 26,731 (45.6%)

LAB: 24,639 (42%)

LDEM: 5,335 (9.1%)

Wimbledon Conservatives CON: 23,946 (46.5%)

LAB: 18,324 (35.6%)

LDEM: 7,472 (14.5%)


✓✓✓ = in first place, well ahead of the second place contender
✓✓ = in first place, but not far ahead of the second place contender
 = in second place
= in third place, but could contend for second place
✕✕ = in third place, but not a realistic contender for second place
✕✕✕ = in fourth place

In case tables full of numbers don’t send you into frissons of shivering delight, I will summarise. Of the twenty-one letters in my dataset, based on the GE17 results, the Liberal Democrats could be characterised as directly competitive, or indeed, likely GE19 winners in… <drumroll> … five cases. So for three-quarters of the letters in my dataset, the Liberal Democrats came in third place or lower in GE17, and are therefore very unlikely to be even remotely competitive. In fact, in nine constituencies they scored in the single digits for their percentage of the vote share. And worst of all, in two constituencies (Edinburgh South, Penistone & Stocksbridge) they actually came a distant fourth.

To conclude this part of the analysis, no matter how you slice it, for the majority of the letters in my dataset, the opening sentence, “between the [x] and the Liberal Democrats” essentially sets up the reader to make a false assumption.

Puff pastry – delicious, empty calories

Let’s move onto the next few sentences. If it helps to follow along, here’s an exemplar letter I made by amalgamating the majority of the letters. You can do whatever you want with/to it. The letter again excludes the two generic Scottish examples because they’re substantially different, but plenty of the points I make below also apply.

Let’s tuck in:

I take polls seriously, having earned my living by cutting through the spin to see what the data really says.

He takes polls seriously. Unlike other, less informed people, this might suggest, who don’t. Perhaps unlike the mainstream media who aren’t “experts”. Who knows. And then he provides some credentials. He’s earned his living doing… what now? I’m intrigued to know what the job title to that role is. “Spin-cutter and data spokesperson”? It’s very intangible. When he sits at his desk each day, what does he actually do?? Anyway, essentially here he throws out the notion that others (of course not him) are creating spin, and he is doing us a priceless service by revealing The Truth. Selflessly. Magnanimously. Because he’s a neutral expert and that’s what neutral experts do. He presents the notion of “seeing what the data says” as though data reveals one coherent, logical, self-evident truth to anyone who could just be bothered to look at or understand it, when in reality, data has to be analysed (already a subjective process), and the results of that analysis must then be interpreted. By humans. Whose interpretations are necessarily and always entrenched in multiple forms of bias.

Anyway, let’s have a look at more of his credentials:

I set up Britain’s first political betting website in 2004

I find this one slightly amusing. Here, in plain words, there is an indication that he runs a gambling operation. A political gambling operation. More on that later. Presumably he tells us he that he got there first because that somehow demonstrates to us… his genius?

and have spent the past 15 years analysing what voters are saying.

Now THIS I like. I would absolutely delight in seeing his research. Does he use a corpus linguistic methodology? Discourse analysis? Conversation analysis? Something else? Does he collect comments on his blog? Tweets? Questionnaire responses? How does he analyse the free-text language? What statistical analyses does he apply to the quantitative data? There is a world of implied methodological and analytical competence and effort in here, and I would be more than thrilled if Mr Smithson would be willing to share his datasets, or his methods, or his analytical frameworks, or indeed any part of his process of “analysing what voters are saying” with me. I certainly wouldn’t imagine that he simply means that he impressionistically eyeballs the odd comment online here and the occasional statement there and glances at polls and makes subjective assessments because I’m sure he would know as well as I do that such an approach would be desperately unrepresentative and prone to wild extrapolation fallacies.

Anyway, onto some crystal-ball truisms:

When looking at the recent election polls one thing is clear – this General Election is going to be a memorable one.

Well. Yes. General Elections generally are memorable. Not often for the right reasons. They tend to be excruciatingly dull for one significant part of the electorate and fraught with stress and disappointment for another significant part. You literally don’t need to “look at recent election polls” to come to that conclusion. In fact, how polls would measure how memorable an event will be that has not yet even taken place, I cannot begin to fathom.

But then we look for a moment like we might start to get a little more concrete:

From the analysis I’ve done,

Ooooh he’s done an analysis! Singular! Just the one but maybe he’s the silver bullet of political insight! How exciting! I wish I knew literally any single detail about this analysis whatsoever! But maybe the next bit will say more…

the data

OH GOD THERE’S DATA wouldn’t it be great to know what kind of data it is? Where it’s from? How much? If it’s representative? How stable it is? What it consists of? Is it betting data? Voting data? Patterns of dogs scratching their ears data? WHAT IS IT EVEN??? I need to know maybe he’ll tell us next THIS IS GETTING EXCITING

gives us a good idea


of the choice


people face on Thursday 12th December –

I- what?

The choice people face in December? How does the analysis, or the data, inform the choice people face? The choice is essentially whoever is on the ballot. No amount of analysis or data will change that. So… not quite as revealing as it seemed.

Also I don’t know who actually typed this but their computer is neither auto-superscripting dates, nor is it auto-correcting hyphens. There’s clearly cut-and-paste going on too because in other letters where dates are used lower down the superscripting has happened, presumably automatically.

Anyway, then we get into a complicated mess of bullet-points, and for the sake of brevity, I’ll only focus on little bits, e.g.

nearly 200 seats across the UK could be decided by [Labour supporters/Remainers] voting tactically for the Liberal Democrats…

Notice the words, “could be”? Jeremy Corbyn could be the unacknowledged child of Queen Elizabeth I and Spencer Tracy. I could be a millionaire disguised as an academic (see what I did there?). Anything could be anything. So this expert isn’t really making any statements. Just observing possibilities that anyone could arrive at without need for more than the most elementary political knowledge. Of course, hedging a statement isn’t de facto bad, but when you have so far offered no data, no method, no analysis, and only vague truisms, it starts to feel a bit suspicious that this “expert” really has any specific, insightful, useful information to share at all.

Then suddenly, this bombshell drops:

Most people who support the Conservatives want to use their vote in whatever way is most likely to stop Jeremy Corbyn becoming Prime Minister.

I’m sorry what now? How can he possibly know voter intention? It’s one of the biggest questions and areas of study in political science today – understanding what influences people to vote certain ways, and whether individuals are (consistently) one-policy voters, or best-compromise-voters, or fickle monsters of fate and chance. How can Mike Smithson possibly know what “most people who support the Conservatives” think?

Let’s make this really objective. Let’s imagine that “most” means just 51%, and let’s take the last best figure we have to work from. Thirteen million people (13,636,684) voted for the Conservatives in GE17. 51% of that number is seven million people (6,954,709). So if we take this at face value, Mike Smithson apparently knows what roughly seven million people think.

Uhuh. Sure. Right.

At the risk of seeming contrary I think I preferred the vague factless assertions better.

Anyway, how does our independent, neutral, objective expert who analyses data and makes incredible non-findings and knows around seven million voters’ intentions close out his missive?

I’m not [x], but…

For his parting persuasions, Smithson decides to call upon the wisdom and experience of his reader, by saying:

You know from previous elections that a vote for any one other [sic] the Liberal Democrats can end up with [a Labour MP/a Conservative MP/an MP who backs Brexit/the bogeyman].

It’s an interesting one because it presumes that the reader somehow has an encyclopaedic knowledge either of their own local past election outcomes, or of countrywide seat losses and gains, and more than that, of how other people voted and the consequences that this had. It bestows upon the recipient a flattering sense of being equally as sage and wise as this lofty and knowledgeable writer. And it does so carefully, via the word “can”, rather than “will”. Here’s a possible outcome you won’t like, it whispers, and here’s how you can stop it from happening. Vote for us or get what you deserve. But no guarantees or refunds.

The final sentence is perhaps the most spectacular:

I’m not here to tell you who to vote for, but I hope this information is useful to you when making a decision in this election.

Let’s be fair to Smithson. He really does not give any direct imperative that tells you to “VOTE LIB DEMS!” or “GO DUCKS!” or “EAT MORE PIE!”. And yet, that asserted neutrality is beautifully and comprehensively counteracted by the tiny three-letter word that follows it….


“I’m not racist, but… [racist thing]”

“I don’t want to hurt your feelings, but… [upsetting revelation]”

“I’m not here to cause trouble, but… [starts riot]”

“I’m not here to tell you who to vote for, but…”

But does a world of work here, in that it says “Yeah, I see everything that came beforehand, and now I’m going to present a contrasting scenario”. In this case, the sentence doesn’t go on to bare its yellow soul and tell the reader which party is “the best choice”. Instead it goes back to its opening stance of apparent impartiality, characterising all that has been said as “information” – another word from the playbook of neutrality and objectivity. And the word “useful” also suggests something disinterested, like a tool. Or a fact. In reality, of course, the “information” is election material, and “useful” could be substituted with a word like “persuasive”.

So how do we know that this is election material? Because at the bottom, it has a tidy little imprint, per the Election Commission regulations, and that imprint, in something like size 6 font, reads:

Published and promoted by M. Dixon on behalf of the Liberal Democrats. 8-10 Great George Street, London, SW1P 3AE

Printed by ProCo, Parkway Close, Parkway Industrial Estate, Sheffield, S9 4WJ

Dixon who?

This leads us to the second and final cast-member in our dramatis personae.

M. Dixon is likely to be Mike Dixon. And who is that? Well, Mike Dixon was appointed the Liberal Democrat CEO on the 21st of October 2019, a mere seven weeks ago as I write this. You can find him on Twitter, but not on Wikipedia, and the fact that his name is on this imprint suggests one of two things. Either this train was in motion as he arrived and he was encouraged to sign off on it, or this was a quick initiative from him to start stamping his mark on the party. It’s a difficult one to be sure about because as it turns out, this is not the first round of letters ever sent out by Mike Smithson.

Let’s wind the clock back to the heady days of GE17…

(I wonder why the signature is so different on this letter and the 2018 ones below versus all the 2019 ones above?)

So here we have a similar letter to the 2019 offerings, less prettily formatted, and more explicitly associated with Smithson’s gambling and betting persona. There is no mention of his being a “polling and election expert” even though it’s only two years ago. In fact, the very prominent web address and the self-congratulatory description of Smithson’s blog being “one of the biggest political websites in the UK” makes me wonder if this wasn’t a mutually beneficial endeavour for the Liberal Democrats, where they used his guise as a politico who is so savvy, he can make a living off betting on election outcomes, and in turn he used their election material to advertise his site.

Er… thereby… allowing him to… make a living off of other people betting… on… You know what? Who am I to question whether his business succeeds because he genuinely possesses rare political acumen and succeeds in his own bets or because his election material brings gambling punters to his site and puts their money in his pockets instead.

Back to the case. We can’t know for a fact, but it is interesting to note the similar wording – that he makes his living by cutting through the spin (again what does that even mean??) then this much more chatty digression onto what the data is “telling” him (how does one get data to confess its secrets so readily? Does he torture his? I’d really like to know) and then his magnanimous declaration that he’s not here to tell the recipient who to vote for, followed yet again by that fatal “but”.

And again in 2018, in the Lewisham East by-election, his letters were popping through doors:

Again, the references to his business are literally writ large and boasted about in the heading, the copy, and the end signature, and again he’s “had a look at the factors at play” (much looser language than the analysis of data he suggests in his 2019 version), and this time he opens with his noblesse oblige ideal of letting the voters choose for themselves, rather than ending with it.

But there’s a problem.

Actually, there are two problems.

They’re quite different but they intertwine, so let’s just get stuck in with them both at once…

Executive authors and insider betting

In June of 2018, just after the Lewisham East by-election, a little site by the name of the SKWAWKBOX (that is an exceptionally difficult word to type; who thought that was a good name??) reported on an “analyst/gambler” who was sending letters for the Liberal Democrats in Lewisham East… and betting heavily on the outcome. You should go read the article at their site since they deserve the traffic and revenue, but to quote the key little bits:

Mr Smithson, speaking from a Spanish train, told the SKWAWKBOX that the letter had not been his idea but was sent at the direct request of the Lib Dem party, who knew what he was going to say because they had seen similar comments on his website.

I’m sorry wait what.

He… did not write the Lewisham East letter? But… it says, in bold, right in the middle of the letter:

I’ve had a look at the factors in play here in Lewisham East and it’s an interesting one.

Yes, the conclusion (“it’s an interesting one”) to the premise (“looking at the factors”) is literally fluff, and therefore commits him to absolutely nothing, but any reasonable reader would assume that the next sentence about “The key factor appears to be…” is directly informed by what he has looked at. And the ordinary reader is also likely to infer that the statistics that follow have also been drawn from the “factors” that he has “looked at”. I mean, if he didn’t write it, then who did? And how could they possibly make these sorts of apparently evidence-based claims?

And then there is the second aspect, which I’ve hinted at throughout, and that is the frankly bizarre juxtaposition of a political gambler – a person who places bets on political outcomes – presenting supposedly neutral information (that is in fact election material) to the electorate to persuade them to vote in certain ways and then, I dunno, betting on the outcome that he has himself partially shaped??

The SKWAWKBOX captures it nicely here:

He also admitted that he had ‘taken a very large position’ on the LibDems coming above the Tories at tonight’s count – “as much as I’ve been able to get on at Ladbrokes and I stand to do reasonably well if it happens”. He did not, however, consider there was a conflict of interest or ethical issue in placing a bet on an outcome and then presenting his expertise in a letter to voters on behalf of the LibDems as a means of trying to make his desired outcome happen.

So that one’s pretty breath-taking. I have absolutely no clues how electoral law stands on it but this feels like something that should be very not allowed.

Experience is making mistakes, wisdom is learning from them

At this stage, you may well be thinking, well, Smithson seems smart. Or if not smart, at least savvy. Probably. And with hindsight, he isn’t likely to have let a situation like this 2018 by-election hiccup occur again, right? He seems to have had his fingers slightly singed by the SKWAWKBOX here, so you’d imagine he would be wiser the next time round. Surely for the 2019 letters, there would be no such issue around his authorship.

Well, on the 02nd of December (2019) just five days ago as I write this, he put out a little statement on his blog, It’s strangely buried at the end of a totally different blog post about “cross-breaks for GE2019 pollsters”, and it’s indented as a quote that appears to have been authored by him personally (he is not the only author on his blog) and it reads thus:

My Lib Dem Mailings

There’s has been much discussion on the site and elsewhere about a tactical voting letter from me that was sent out last week by the Lib Dems to selected voters. Quite a few PBers have received one. As is widely known I have been a member of the Lib Dems since its foundation and make no apologies for seeking to help the party during elections.

I should explain that while I approved the text of the letters I did not have a prior view of the list of constituencies they were going to. This was unlike GE2017 when a similar exercise was carried out with me approving every single seat on the constituency list. If I had had an input this time most of the seats would have been the same but I would have had a different list.

For this election letters were put into the Royal Mail system near the start of the election campaign where the post European election result dominated the most recent polling landscape. Since that time of course the YouGov MRP poll and other events have moved the needle.

The party has given me assurances about the future.

Okay so three things.

Firstly, “I approved the text of the letters”. Uhuh. So he didn’t write them himself. I think no one is shocked by that idea by this stage, but when you look at one of those letters, fresh out of the envelope, having never seen one before, I would imagine nearly everyone who read it would presume that the person whose picture, name, and signature is on that letter and who uses “I” and “me” throughout actually wrote that letter personally. Especially when it’s set up to look like independent, impartial, expert information, unsolicited by any political party, just sent out of the goodness of his heart.

Secondly, the third paragraph (“For this election…”) essentially implies that any analyses carried out (and again, I would love to know what they consisted of) were effectively out-of-date by the time the letters were sent, so the logical conclusion is that the “expert information” the letters contain is, at best, only accidentally accurate, or it is inaccurate, or it is outright wrong. But of course no such caveats or considerations are highlighted in the letters. They are presented as if the one analysis he carried out stands as the monolithic “truth” that the reader can implicitly trust from the moment of receipt to the day in the election. If the landscape really is changing so much and so fast that the difference of a few days or weeks is responsible for all the pushback then surely an expert would either make that clear in their letter or they wouldn’t attempt to advise voters at all for fear of muddying the water with outdated information.

Finally, that last line is a joy to behold. What a world of meaning is captured in “assurances”. One can only guess at what those assurances may have been, and how they came about. I have enjoyed dreaming up the various phone-calls or emails or Twitter DMs between Mike Smithson, and, I’m guessing, Mike Dixon, Liberal Democrat CEO, only a few weeks in post, perhaps encouraged to sign off on this mailshot, or eager to make his mark, and then ending up as the imprint on a letter with which the executive author seems not entirely pleased.

I look forward to seeing what other imprints M Dixon finds himself named in.

To infamy, and beyond

In conclusion, factcheckUK was mindboggling in its stupidity, because it corroded the trust that voters have in both politicians and our new final bastion of hope, the factchecker, and it did that all through the medium of Twitter, the stronghold of the younger voter.

Then the local newspapers debacle was frustrating because it corroded the trust that voters have in both politicians and our previous final bastion of hope, the newspaper, and it did that all through the medium of the local press, a favourite with the 65+ working class voter.

And then this. The  Mike Smithson letter. This one  is personally irritating, because it corrodes the trust that voters have in both politicians and in experts. Real experts carefully collect data and ensure that it is representative and/or fit for purpose. Then we apply methods carefully, and quality check our processes as we go along. Then we derive rigorous and careful results. Then we explain them as impartially as possible, complete with caveats and provisos. We don’t, generally speaking, try to push people to our gambling website to make money on the outcome of an event that we’ve actively tried to shape through our supposedly impartial advice that we didn’t write ourselves anyway.

If politicians could have come together, as a team, to work out how best to sabotage their own credibility and the electorate’s faith in the democratic process as comprehensively as possible, along with a range of other traditionally respected and valued avenues of impartial information, I actually can’t envision a more complete campaign than this one. From factcheckers to local papers to actual experts, from Twitter to the press to personal letters, it has struck across all avenues of credible information and all platforms, thereby reaching all voting groups. Routinely I encourage people to vote and engage with the democratic process and when people point out that politicians are often not playing on the same field as voters, I think of cases like this and I have very few answers.

In simple words, our democratic system appears to be sinking under a toxic swamp of disinformation – not from those terrible outside actors in different countries, but from our own parties. As each tribe goes information-nuclear, the rest seem to be scrambling to follow suit. Reversing the trend and fixing the damage will require a seismic culture shift away from trickery and manipulation, towards transparency and responsibility. But each person who can make a difference in however small a way should. Disinformation should be called out not just by recipients of it, but by those who see it passing through the system, and by those who are asked to participate in it. Underhanded tactics should be resisted, quietly if it cannot be done loudly, and passively if it cannot be done actively. We all know how resistance can build in the machine if each person puts the brakes on their part in a process, and where disinformation is concerned, we should be applying the brakes as hard as we can. I’m asking all our politicians, civil servants, volunteers, supporters, campaigners, and adherents across every party to take care of your own back yard. Tidy it. Weed it. Don’t let the poisonous tendrils of “just this one time” or “for their own good” start to creep through the grass. The problem may seem insurmountable at this point, but to continue the nature metaphor, in the evergreen words of Spock,

One can begin to reshape the landscape with a single flower, Captain.

Okay I’m done.

That’s the end of this three part flustercluck of pre-election analyses. I’m going to spend the remaining five days hiding under a pile of chocolate.