UK Election of June 2017 – Polling review


Here are all the published opinion polls for the June 2017 UK general election, plotted as a Shewhart chart.

The Conservative lead over Labour had been pretty constant at 16% from February 2017, after May’s Lancaster House speech. The initial Natural Process Limits (“NPLs”) on the chart extend back to that date. Then something odd happened in the polls around Easter. There were several polls above the upper NPL. That does not seem to fit with any surrounding event. Article 50 had been declared two weeks before and had had no real immediate impact.

I suspect that the “fugue state” around Easter was reflected in the respective parties’ private polling. It is possible that public reaction to the election announcement somehow locked in the phenomenon for a short while.

Things then seem to settle down to the 16% lead level again. However, the local election results at the bottom of the range of polls ought to have sounded some alarm bells. Local election results are not a reliable predictor of general elections but this data should not have felt very comforting.

Then the slide in lead begins. But when exactly? A lot of commentators have assumed that it was the badly received Conservative Party manifesto that started the decline. It is not possible to be definitive from the chart but it is certainly arguable that it was the leak of the Labour Party manifesto that started to shift voting intention.

Then the swing from Conservative to Labour continued unabated to polling day.

Polling performance

How did the individual pollsters fair? I have, somewhat arbitrarily, summarised all polls conducted in the 10 days before the election (29 May to 7 June). Here is the plot along with the actual popular poll result which gave a 2.5% margin of Conservative over Labour. That is the number that everybody was trying to predict.


The red points are the surveys from the 5 days before the election (3 to 7 June). Visually, they seem to be no closer, in general, than the other points (6 to 10 days before). The vertical lines are just an aid for the eye in grouping the points. The absence of “closing in” is confirmed by looking at the mean squared error (MSE) (in %2) for the points over 10 days (31.1) and 5 days (34.8). There is no evidence of polls closing in on the final result. The overall Shewhart chart certainly doesn’t suggest that.

Taking the polls over the 10 day period, then, here is the performance of the pollsters in terms of MSE. Lower MSE is better.

Pollster MSE
Norstat 2.25
Survation 2.31
Kantar Public 6.25
Survey Monkey 8.25
YouGov 9.03
Opinium 16.50
Qriously 20.25
Ipsos MORI 20.50
Panelbase 30.25
ORB 42.25
ComRes 74.25
ICM 78.36
BMG 110.25

Norstat and Survation pollsters will have been enjoying bonuses on the morning after the election. There are a few other commendable performances.

YouGov model

I should also mention the YouGov model (the green line on the Shewhart chart) that has an MSE of 2.25. YouGov conduct web-based surveys against at huge data base or around 50,000 registered participants. They also collect, with permission, deep demographic data on those individuals concerning income, profession, education and other factors. There is enough published demographic data from the national census to judge whether that is a representative frame from which to sample.

YouGov did not poll and publish the raw, or even adjusted, voting intention. They used their poll to  construct a model, perhaps a logistic regression or an artificial neural network, they don’t say, to predict voting intention from demographic factors. They then input into that model, not their own demographic data but data from the national census. That then gave their published forecast. I have to say that this looks about the best possible method for eliminating sampling frame effects.

It remains to be seen how widely this approach is adopted next time.


Building targets, constructing behaviour

Recently, the press reported that UK construction company Bovis Homes Group PLC have run into trouble for encouraging new homeowners to move into unfinished homes and have therefore faced a barrage of complaints about construction defects. It turns out that these practices were motivated by a desire to hit ambitious growth targets. Yet it has all had a substantial impact on trading position and mark downs for Bovis shares.1

I have blogged about targets before. It is worth repeating what I said there about the thoughts of John Pullinger, head of the UK Statistics Authority. He gave a trenchant warning about the “unsophisticated” use of targets. He cautioned:2

Anywhere we have had targets, there is a danger that they become an end in themselves and people lose sight of what they’re trying to achieve. We have numbers everywhere but haven’t been well enough schooled on how to use them and that’s where problems occur.

He went on.

The whole point of all these things is to change behaviour. The trick is to have a sophisticated understanding of what will happen when you put these things out.

That message was clearly one that Bovis didn’t get. They legitimately adopted an ambitious growth target but they forgot a couple of things. They forgot that targets, if not properly risk assessed, can create perverse incentives to distort the system. They forgot to think about how manager behaviour might be influenced. Leaders need to be able to harness insights from behavioural economics. Further, a mature system of goal deployment imposes a range of metrics across a business, each of which has to contribute to the global organisational plan. It is no use only measuring sales if measures of customer satisfaction and input measures about quality are neglected or even deliberately subverted. An organisation needs a rich dashboard and needs to know how to use it.

Critically, it is a matter of discipline. Employees must be left in no doubt that lack of care in maintaining the integrity of the organisational system and pursuing customer excellence will not be excused by mere adherence to a target, no matter how heroic. Bovis was clearly a culture where attention to customer requirements was not thought important by the staff. That is inevitably a failure of leadership.

Compare and contrast

Bovis are an interesting contrast with supermarket chain Sainsbury’s who featured in a law report in the same issue of The Times.3 Bovis and Sainsbury’s clearly have very different approaches as to how they communicate to their managers what is important.

Sainsbury’s operated a rigorous system of surveying staff engagement which aimed to embrace all employees. It was “deeply engrained in Sainsbury’s culture and was a critical part of Sainsbury’s strategy”. An HR manager sent an email to five store managers suggesting that the rigour could be relaxed. Not all employees needed to be engaged, he said, and participation could be restricted to the most enthusiastic. That would have been a clear distortion of the process.

Mr Colin Adesokan was a senior manager who subsequently learned of the email. He asked the HR manager to explain what he had meant but received no response and the email was recirculated. Adesokan did nothing. When his inaction came to the attention of the chief executive, Adesokan was dismissed summarily for gross misconduct.

He sued his employer and the matter ended up in the Court of Appeal, Adesokan arguing that such mere inaction over a colleague’s behaviour was incapable of constituting gross misconduct. The Court of Appeal did not agree. They found that, given the significance placed by Sainsbury’s on the engagement process, the trial judge had been entitled to find that Adesokan had been seriously in dereliction of his duty. That failing constituted gross misconduct because it had the effect of undermining the trust and confidence in the employment relationship. Adesokan seemed to have been indifferent to what, in Sainsbury’s eyes, was a very serious breach of an important procedure. Sainsbury’s had been entitled to dismiss him summarily for gross misconduct.

That is process discipline. That is how to manage it.

Display constancy of purpose in communicating what is important. Do not turn a blind eye to breaches. Do not tolerate those who would turn the blind eye. When you combine that with mature goal deployment and sophistication as to how to interpret variation in metrics then you are beginning to master, at least some parts of, how to run a business.


  1. “Share price plunges as Bovis tries to rebuild customers’ trust” (paywall), The Times (London), 20 February 2017
  2. “Targets could be skewing the truth, statistics chief warns” (paywall), The Times (London), 26 May 2014
  3. Adesokan v Sainsbury’s Supermarkets Ltd [2017] EWCA Civ 22, The Times, 21 February 2017 (paywall)

On leadership and the Chinese contract

Hanyu trad simp.svgBetween 1958 and 1960, 67 of the 120 inhabitants of the Chinese village of Xiaogang starved to death. But Mao Zedong’s cruel and incompetent collectivist policies continued to be imposed into the 1970s. In December 1978, 18 of Xiaogang’s leading villagers met secretly and illegally to find a way out of borderline starvation and grinding poverty. The first person to speak up at the meeting was Yan Jingchang. He suggested that the village’s principal families clandestinely divide the collective farm’s land among themselves. Then each family should own what it grew. Jingchang drew up an agreement on a piece of paper for the others to endorse. Then he hid it in a bamboo tube in the rafters of his house. Had it been discovered Jingchang and the village would have suffered brutal punishment and reprisal as “counter-revolutionaries”.

The village prospered under Jingchang’s structure. During 1979 the village produced more than it had in the previous five years. That attracted the attention of the local Communist Party chief who summoned Jingchang for interrogation. Jingchang must have given a good account of what had been happening. The regional party chief became intrigued at what was going on and prepared a report on how the system could be extended across the whole region.

Mao had died in 1976 and, amid the emerging competitors for power, it was still uncertain as to how China would develop economically and politically. By 1979, Deng Xiaoping was working his way towards the effective leadership of China. The report into the region’s proposals for agricultural reform fell on his desk. His contribution to the reforms was that he did nothing to stop them.

I have often found the idea of leadership a rather dubious one and wondered whether it actually described anything. It was, I think, Goethe who remarked that “When an idea is wanting, a word can always be found to take its place.” I have always been tempted to suspect that that was the case with “leadership”. However, the Jingchang story did make me think.1 If there is such a thing as leadership then this story exemplifies it and it is worth looking at what was involved.

Personal risk

This leader took personal risks. Perhaps to do otherwise is to be a mere manager. A leader has, to use the graphic modern idiom, “skin in the game”. The risk could be financial or reputational, or to liberty and life.


Luck is the converse of risk. Real risks carry the danger of failure and the consequences thereof. Jingchang must have been aware of that. Napoleon is said to have complained, “I have plenty of clever generals but just give me a lucky one.2 Had things turned out differently with the development of Chinese history, the personalities of the party officials or Deng’s reaction, we would probably never have heard of Jingchang. I suspect though that the history of China since the 1970s would not have been very different.

The more I practice, the luckier I get.

Gary Player
South African golfer

Catalysing alignment

It was Jingchang who drew up the contract, who crystallised the various ideas, doubts, ambitions and thoughts into a written agreement. In law we say that a valid contract requires a consensus ad idem, a meeting of minds. Jingchang listened to the emerging appetite of the the other villagers and captured it in a form in which all could invest. I think that is a critical part of leadership. A leader catalyses alignment and models constancy of purpose.

However, this sort of leadership may not be essential in every system. Management scientists are enduringly fascinated by The Morning Star Company, a California tomato grower that functions without any conventional management. The particular needs and capabilities of the individuals interact to create an emergent order that evolves and responds to external drivers. Austrian economist Friedrich Hayek coined the term catallaxy for a self-organising system of voluntary co-operation and explained how such a thing could arise and sustain and what its benefits to society.3

But sometimes the system needs the spark of a leader like Jingchang who puts himself at risk and creates a vivid vision of the future state against which followers can align.

Deng kept out of the way. Jingchang put himself on the line. The most important characteristic of leadership is the sagacity to know when the system can manage itself and when to intervene.


  1. I have this story from Matt Ridley (2015) The Evolution of Everything, Fourth Estate
  2. Apocryphal I think.
  3. Hayek, F A (1982) Law, Legislation, and Liberty, vol.2, Routledge, pp108–9

Imagination, data and leadership

I had an intriguing insight into the nature of imagination the other evening when I was watching David Eagleman’s BBC documentary The Brain which you can catch on iPlayer until 27 February 2016 if you have a UK IP address.

Eagleman told the strange story of Henry Molaison. Molaison suffered from debilitating epilepsy following a bicycle accident when he was nine years old. At age 27, Molaison underwent radical brain surgery that removed, all but completely, his hippocampi. The intervention stabilised the epilepsy but left Molaison’s memory severely impaired. Though he could recall his childhood, Molaison had no recall of events in the years leading up to his surgery and was unable to create new long-term memories. The case was important evidence for the theory that the hippocampus is critical to memory function. Molaison, having lost his, was profoundly compromised as to recall.

But Eagleman’s analysis went further and drew attention to a passage in a interview with Molaison later in his life.1 Though his presenting symptoms post-intervention were those of memory loss, Molaison also encountered difficulty in talking about what he would do the following day. Eagleman advances the theory that the hippocampus is critical, not only to memory, but to imagining the future. The systems that create memories are common to those that generate a model by which we can forecast, predict and envision novel outcomes.

I blogged about imagination back in November and how it was pivotal to core business activities from invention and creativity to risk management and root cause analysis. If Eagleman’s theory about the entanglement of memory and imagination is true then it might have profound implications for management. Perhaps our imagination will only function as well as our memory. That was, apparently, the case with Molaison. It could just be that an organisation’s ability to manage the future depends upon the same systems as those by which it critically captures the past.

That chimes with a theory of innovation put forward by W Brian Arthur of the Santa Fe Institute.2 Arthur argues that purportedly novel inventions are no more than combinations of known facts. There are no great leaps of creativity, just the incremental variation of a menagerie of artifacts and established technologies. Ideas similar to Arthur’s have been advanced by Matt Ridley,3,4 and Steven Berlin Johnson.5 Only mastery of the present exposes the opportunities to innovate. They say.


This all should be no surprise to anybody experienced in business improvement. Diligent and rigorous criticism of historical data is the catalyst of change and the foundation of realising a vivid future. This is a good moment to remind ourselves of the power of the process behaviour chart in capturing learning and creating an organisational memory.


The process behaviour chart provides a cogent record of the history of operation of a business process, its surprises and disappointments, existential risks and epochs of systematic productivity. It records attempted business solutions, successful, failed, temporary and partial work-rounds. It segregates signal from noise. It suggests realistic bounds on prediction. It is the focus of inclusive discussion about what the data means. It is the live report of experimentation and investigation, root cause analysis and problem solving. It matches data with its historical context. It is the organisation’s memory of development of a business process, and the people who developed it. It is the basis for creating the future.

If you are not familiar with how process behaviour charts work in this context then have a look at Don Wheeler’s example of A Japanese Control Chart.6


Tim Harford tries to take the matter further.7 On Harford’s account of invention, “trial and error” consistently outperform “expert leadership” through a Darwinian struggle of competing ideas. The successful innovations, Harford says, propagate by adoption and form an ecology of further random variation, out of which the best ideas emergently repeat the cycle or birth and death. Of course, Leo Tolstoy wrote War and Peace, his “airport novel” avant la lettre, (also currently being dramatised by the BBC) to support exactly this theory of history. In Tolsoy’s intimate descriptions of the Battles of Austerlitz and Borodino, combatants lose contact with their superiors, battlefields are obscured by smoke from the commanding generals, individuals act on impulse and in despite of field discipline. How, Tolstoy asked in terms, could anyone claim to be the organising intelligence of victory or the culpable author of defeat?

However, I think that a view of war at odds with Tolstoy’s is found in the career of General George Marshall.8 Marshall rose to the rank of General of the Army of the USA as an expert in military logistics rather than as a commander in the field. Reading a biography of Marshall presents an account of war as a contest of supply chains. The events of the theatre of operations may well be arbitrary and capricious. It was the delivery of superior personnel and materiel to the battlefield that would prove decisive. That does not occur without organisation and systematic leadership. I think.

Harford and the others argue that, even were the individual missing from history, the innovation would still have occurred. But even though it could have been anyone, it still had to be someone. And what that someone had to provide was leadership to bring the idea to market or into operation. We would still have motor cars without Henry Ford and tablet devices without Steve Jobs but there would have been two other names who had put themselves on the line to create something out of nothing.

In my view, the evolutionary model of innovation is interesting but stretches a metaphor too far. Innovation demands leadership. The history of barbed wire is instructive.9 In May 1873, at a county fair in Illinois, Henry B Rose displayed a comical device to prevent cattle beating down primitive fencing, a “wooden strip with metallic points”. The device hung round the cattle’s horns and any attempts to butt the fence drove the spikes into the beast’s head. It didn’t catch on but at the fair that day were Joseph Glidden, Isaac L Ellwood and Jacob Haish. The three went on, within a few months, each to invent barbed wire. The winning memes often come from failed innovation.

Leadership is critical, not only in scrutinising innovation but in organising the logistics that will bring it to market.10 More fundamentally, leadership is pivotal in creating the organisation in which diligent criticism of historical data is routine and where it acts as a catalyst for innovation.11


  1., retrieved 29 January 2016, at p5
  2. Arthur, W B (2009) The Nature of Technology: What it is and How it Evolves, The Free Press/ Penguin Books.
  3. Ridley, M (2010) The Rational Optimist, Fourth Estate
  4. — (2015) The Evolution of Everything, Fourth Estate
  5. Johnson, S B (2010) Where Good Ideas Come From: The Seven Patterns of Innovation, Penguin
  6. Wheeler, D J (1992) Understanding Statistical Process Control, SPC Press
  7. Harford, T (2011) Adapt: Why Success Always Starts with Failure, Abacus
  8. Cray, E (2000) General of the Army: George C. Marshall, Soldier and Statesman, Cooper Square Press
  9. Krell, A (2002) The Devil’s Rope: A Cultural History of Barbed Wire, Reaktion Books
  10. Armytage, W H G (1976) A Social History of Engineering, 4th ed., Faber
  11. Nonaka, I & Takeuchi, H (1995) The Knowledge-Creating Company: How Japanese Companies Create the Dynamics of Innovation, Oxford University Press

Trouble at the EU

I enjoy Metro the UK national free morning newspaper. It has a very straightforward non-partisan style. This morning there was an article dealing with the European Union’s (EU’s) accounting difficulties. There were a couple of very telling admissions from an EU bureaucrat. We lawyers love an admission.

Aidas Palubinskas, from the European Court of Auditors, … described the error rate as ‘relatively stable from year to year’.

He admits that the EU’s accounting is a stable system of trouble. That is a system where there is only common cause variation, variation common to the whole of the output, but where the system is still incapable of reliably delivering what the customer wants. Recognising that one is embedded in such a problem is the first step towards operational improvement. W Edwards Deming addressed the implications of the stable system and the strategy for its improvement at length in his seminal book Out of the Crisis (1982). The problems are not intractable but the solution demands leadership and adoption of the correct improvement approach.

Unfortunately, the second half of the quote is less encouraging.

He said the errors highlighted in its report were ‘examples of inefficiency, but not necessarily of waste’.

This makes me fear that the correct approach is far off for the EU. Everything that is not efficient, timely and effective delivery of what the customer wants is waste, as Toyota call it muda. Waste represents the scope of opportunity for improvement, for improving service and simultaneously reducing its cost. The first step in improvement is taken by accepting that waste is not inevitable and that it can be incrementally eliminated through use of appropriate tools under competent leadership.

The next step to improvement is to commit to the discipline of eliminating waste progressively. That requires leadership. That sort of leadership is often found in successful organisations. The EU, however, faces particular difficulties as an international bureaucracy with a multi-partisan political master and a democratically disengaged public. It is not easy to see where leadership will come from. This is a common problem of state bureaucracies.

Palubinskas is right to seek to analyse the problems as a stable system of trouble. However, beyond that, the path to radical improvement lies in rejecting the casual acceptance of waste and in committing to continual improvement of every process for delivery of service.