A question often posed on Facebook groups is "Do authors get paid more for a book if it's read via Kindle Unlimited or bought as a Kindle ebook?"

The answer is surprisingly complicated!

It depends on:

• the cost of the book as an ebook
• the number of pages in it
• the country it's read/sold in
• the currency exchange rate
• the KENP rate (amount paid in each country per page read) - which is different in each country and changes month to month, depending on how many people are in KU and how many books they read!

I said it was complicated!

Since an author is paid per pages read, clearly a shorter book will generate less money via Kindle Unlimited than a longer book. But if the book is priced at 99c/99p then it will probably net the author more if the book is read in KU rather than bought.

What has often interested me is how different my royalties are across different countries, for the same book.

Let me use "Aegyir Rises" and the box-set as examples, and look at UK v US v Canada v Australia.

Amazon tells me "Aegyir Rises" is 454 KENP (Kindle pages) and the box-set is 1606 KENP. If I look at a month where I have good data for those four countries, I can calculate how many pages need to be read to earn me a pound in each country. From that, I can calculate how much I would earn for 454 pages, and 1606 pages. I can also compare the royalties generated by reading the book in comparison with buying the book.

Okay, so in the table, you can see in the pages/£ column that an Australian has to read a lot more pages (512) to earn me the same amount as money as a Brit reading the books (332 pages)! I think this must be a reflection of how many people are signed up for KU in Australia, how much they pay for it and how many books they read each month.

If you then look at how much each book sale gets me (in £) the data are all pretty similar. The killer column is the final one: how much I get if they read the book in comparison with if they bought the book. 41%... OUCH! Even 66% isn't great!

Is the box-set any better? Well... yes, and no

The final column looks better, right? But compare how much I get paid if someone in Australia reads the whole of the box-set in comparison with someone buying "Aegyir Rises"... £3.14 for 1606 pages read (three decent-sized books) compared to £2.14 for just the first book. The only reason the percentages in the final column look better is because there is a price-cap on the box-set in Canada and Australia. You can't charge more than $9.99 for an ebook in Canada or $11.99 in Australia. Since individual books are priced $4.99 and $5.99, you essentially get a book for free by buying the box-set there.

Are you surprised by how little an author gets? Minimum wage in the UK is £8.91/hour for my age. £8.91 x 40 hours/week x 50 weeks per year = £17,820. The average payment for the ebook if it's read is £1.17. If it's bought it's £2.09. To earn just the minimum wage, I need to sell ~4 books every hour. Or 7.5 people have to read the entire book every hour.

I can tell you for free, I get nowhere near that! It's a really good job I didn't hope to make any money by writing books!

Wines are often described as tasting ‘green’. But what does this mean? Jamie Goode takes a look at the science of green wine.

Can you taste a colour? In some cases, it seems that you can. Green. It’s a descriptor often used in tasting notes, and often it’s pejorative, especially in the context of red wines. Think of some of the tasting terms that are used to describe greenness: herbal, grassy, vegetal, asparagus, green pepper, and even ‘green tannin’.

There are a number of chemicals present in wine that can contribute to these green flavours, but there is one group of compounds in particular that is strongly associated with green aromas and flavours in wine. This is the alkylmethoxypyrazines, and in our exploration of greenness we will begin with them. They are present in green tissues of plants, and they were first discovered in green peppers in 1969.

Sauvignon Blanc is one grape variety where it seems that green is desirable. Good Sauvignon often has an element of greenness. Think of a New Zealand Sauvignon Blanc from the Marlborough region, and especially the Awatere Valley. Key descriptors include green pepper (bell pepper to Americans, also known by many as capsicum), tomato leaf, herbal and even cut grass. These aromas and flavours have chiefly been attributed to methoxypyrazines.

With the resounding success of New Zealand Sauvignon, winemakers around the world began working to increase the levels of methoxypyrazine in their Sauvignons. [Later, they were to attempt the same with the polyfunctional thiols, but that’s another story.] Some South African winemakers famously took the illegal short-cut of adding methoxypyrazines to their wines directly, but they got caught. Now, we think that green alone isn’t enough, and fortunately the excessively green Sauvignons with little else going on are becoming rarer.

Green is also a frequent contributor to the aroma profile of the Bordeaux grape varieties Cabernet Sauvignon, Cabernet Franc and Merlot. Carmenère, a Bordeaux variety that relocated to Chile, also has high levels of green. Good red Bordeaux often has green in its signature, although in some new world countries winemakers seem terrified of greenness. In most other well-known red varieties, methoxypyrazines are below detection levels in ripe grapes.

Isobutyl-methoxypyrazine (IBMP) is the most important methoxypyrazine in wine, and typical concentrations found in wines would be in the range of 5-30 nanograms/litre. That’s significantly above the sensory threshold in water, which is just a couple of nanograms a litre (the sensory threshold in wine, especially red wine, is higher at around 15 nanograms/litre). The other ones identified in wine are isopropyl- methoxypyrazine (IPMP) and sec-butyl- methoxypyrazine (SBMP), which are usually at much lower levels, but could still play a role contributing to these green flavours.

So how do these green flavours get into the wine? Methoxypyrazine levels gradually increase in grapes as the berries grow, reaching a maximum just before veraison, and then they start to decrease as ripeness approaches. Unripe grapes typically have high levels of methoxypyrazines, which may be acting as an anti-feedant, along with high tannin levels. The idea here is that the grapevine doesn’t want its berries eaten by birds until the seeds are ready to be dispersed. Certainly, if you taste an unripe grape, it tastes pretty disgusting: there’s no sugar, very high acidity, grippy tannins and excessive greenness.

Vines with grape leafroll virus often have delayed berry maturation, and can make red wines with high levels of green in them. Often, winegrowers leave the grapes on the vine in an attempt to ripen out the greenness, but it’s not as simple as this: extra hang time may not be effective in reducing methoxypyrazine levels beyond a certain point, so if they are present they can’t just be ripened out.

A common viewpoint among viticulturists is that one way of reducing methoxypyrazine levels is to expose the grapes to the sun. So they pluck the leaves in the fruit zone, thinking that the sunlight will degrade the methoxypyrazine. But there’s very little evidence for this photodegradation actually occurring. Still, leaf removal seems to have some effect, especially when it’s done early on. This might be because of reduced accumulation rather than increased degradation. In fact, there is evidence that if the vine is growing rapidly during this period of methoxypyrazine accumulation, then more methoxypyrazine is accumulated, independent on the level of cluster shading. So for red wines you really want vine growth to slow down just before the methoxypyrazine accumulation phase.

If methoxypyrazines are present in the grapes at harvest, because they are very stable compounds, they tend to persist through fermentation and then stay at the same level in the final wine.

One other source of greenness in wine is from the use of stems in winemaking. This is an interesting topic in its own right, and deserves its own discussion, but when whole bunches are used in fermentation they can contribute green flavours to the wines.

Interestingly, methoxypyrazines are also produced by some insects. There are notable incidences where ladybirds present in grape bunches at harvest cause green taints in wine, because they are capable of producing heroic levels of methoxypyrazine. This can lead to the wine being ruined. Vineyards near soybean farms can be at particular risk, because the soybeans are harvested and then the ladybirds find a new home in the vineyard. You don’t need many in a vat for this green taint to occur, so a vibrating sorting table is a must where this is potentially an issue.

So what can you do if you want lower methoxypyrazine levels in your wine, but they are present in the harvested grapes? It’s possible to mask their presence using oak, and it seems that more gentle pressing results in lower levels in white wines. Some methoxypyrazine is lost in settling and clarifying the must of white varieties before fermentation. And thermovinification also reduces levels. But, generally, if you want lower levels, you need to address this issue in the vineyard.

But there’s more to greenness in wine than just methoxypyrazines. The other main contributors are a bunch of chemicals called green leaf volatiles. These are produced by a chemical reaction chain called the oxylipin pathway, and they are involved in plant defence. One of these compounds is called hexanol, which is described as green, herbaceous, grassy, woody and fruity. Then there’s cis-3-hexen-1-ol, which is green and grassy (and also acts as as semiochemical, alterting predators of herbivores that dinner is ready), and trans-3-hexen-1-ol, which is green, earthy and fatty.

Finally, there are green notes in wine that could come from the environment. Vineyards in proximity to both eucalyptus and pine trees can make wines that are influenced by aromatics from the leaves of these trees, which impart distinctive characters to the wine including an element of green.

What about green tannins? Personally, I don’t think they exist: this is just picture language. Tannins are sensed as astringent (this is the sense of touch), and sometimes also as bitter (the sense of taste). But I can’t see a way for them to taste green. Perhaps people using this term are describing tannins with a certain mouthfeel as green because they co-occur with green flavours from slightly underripe grapes.

World’s First Open Source Insurance Policy

Why we’re throwing out the traditional policy and creating Policy 2.0!

Open to change

“Every act of creation is first an act of destruction” — Pablo Picasso

To create an unconflicted insurance company you have to rethink the very business model of insurance; to make it transparent you have to take a bulldozer to its foundational document: the Policy.

Lemonade is committed to both.

In late 2016 we launched Lemonade with a novel business model, but sensed that razing the ‘industry standard’ insurance policy was more than our regulators and reinsurers could stomach.

That was then. This is now.

Transparency is the best policy

“A lack of transparency results in distrust and a deep sense of insecurity” — Dalai Lama

To normal people, insurance policies are the ultimate word salad. The renters policy we launched with is ‘industry standard,’ which is to say it’s 40 pages long, and contains some 20,000 words, many of them from middle-English.

But more troubling than words like ‘pewterware,’ ‘smudging,’ and ‘bailee,’ is how the policy contorts even simple words, like ‘Employee,’ into gobbledegook:

The most lethal ‘transparency-killers,’ though, aren’t quaint words or wordy definitions — they’re the exceptions. As John Verdon put it, “an exception is a resentment waiting to be born.” And since insurance policies read like a laundry list of exceptions, with exceptions to those exceptions, and exceptions to those exceptions, they ooze resentment.

Take the simple question of how much ‘personal property’ is insured by your policy. If you bought a $20,000 personal property policy, you might expect the answer to be $20,000. Which is true, except…

• There’s a $1,000 limit for the removal of your neighbor’s fallen trees… But there’s a $500 limit for any one tree… And you get nothing if the trees didn’t damage ‘a covered structure’; Except, that is, if they block the driveway; On condition the blockage ‘prevents a “motor vehicle” from getting by; Except, that is, if the “motor vehicle” isn’t “registered for use on public roads.”
• There’s a $2,500 limit for pewterware (whatever that is).
• Your claim’s capped at $1,500 for “furnishings” of a “watercraft” (huh?)
• You’ll see only $250 for “antennas, tapes, wires, records…” (naturally…)
• And you’ll get zero if your pewterware, furnishings, antennas, tapes, wires or records were in the hands of a “bailee” at the time they were stolen.

And so it continues for 20,000 words! Any wonder people feel the deck is stacked against them?

Aims of Policy 2.0

“Everything should be made as simple as possible, but not simpler” — Albert Einstein

We’re taking a run at creating ‘Policy 2.0,’ with 4 overarching goals: Make a policy that’s simple, approachable, relevant, and digital.

Simple:

The key to simplicity is consistency. Exceptions make actuaries happy, but their costs, in complexity and transparency, are just too high.

Zero exceptions may be unrealistic, because the price of such a policy would be unattractive, but the goal must be a radical reduction in exceptions, and ensuring they’re easy to understand and remember.

Approachable:

There are good reasons why policies use lingo that requires a law degree and a broker’s license to understand, but Policy 2.0 is about using contemporary English that everyone understands.

And the policy has to be way shorter. No document is ‘readable’ if it’s so long that no one actually reads it.

Relevant:

Policy 2.0 doesn’t aim to increase or decrease coverage per se. It aims to balance relevant coverage with affordability, while allowing users to choose additional coverage that makes sense for them.

Today’s policy deals at length with volcanic eruptions, nuclear fallout and ‘civil commotion,’ but says nothing about laptops or smartphones. Say what?!

So Policy 2.0 will drop coverage for ‘volcanic action’ and suchlike, while doing away with ‘gotcha’ limits for jewelry and electronics (not to mention ‘antennas, tapes, wires and records.’)

After all, people are more likely to be hurt by incomprehensible coverage, than by a volcano.

Digital:

Each Policy 2.0 will be unique and dynamically-generated, based on the choices the user made. While people can print Policy 2.0, it’ll be at its most powerful on a screen. When the policy says that $20,000 of property is covered, for example, our Live Policy technology makes that sentence clickable, so the user can instantly change that to $30,000. If the user wants to add earthquake coverage, to take another example, they can initiate that from within the policy itself, and the policy will morph to include earthquake coverage.

Being digital means Policy 2.0 will shape-shift on command.

Our ‘open policy’ policy

“Sunlight is said to be the best of disinfectants” — Louis D. Brandeis

We’ve received tens of thousands of questions about our ‘industry standard’ policy, and these conversations have guided our work on Policy 2.0. But we’d like more input. Writing an entirely new policy hasn’t been done in generations, and there are many perspectives to bear in mind:

• Consumers (trade-offs around making it comprehensive yet comprehensible, relevant yet affordable.)
• Contract-law (making it certain and enforceable, while avoiding legalese or industry jargon.)
• Regulatory (complying with regulations, on a state by state basis, that weren’t written with Policy 2.0 in mind.)
• Actuarial (the further we travel from the traditional policy, the harder it will be to use historical loss-data to price Policy 2.0.)
• Mobile-first (ensuring the policy can be communicated on a 5 inch screen, in just a few seconds.)

So before we bring Policy 2.0 to market, we invite consumers, advocacy groups, regulators, insurance enthusiasts, data scientists, designers, competitors, and techies, to all jump in. We’re big believers in ‘wisdom of the crowd,’ and hope broad participation will make Policy 2.0 all it can be. We invite you to join the conversation and contribute on Github.

Open source

Policy 2.0 is open in another sense: not only do we want broad participation, we want broad adoption.

Policy 2.0 is available under a ‘Copyleft’ license to all our competitors, free of charge (Open Source GNU FDL license).

Making insurance simple, approachable, and relevant should be a common goal, a communal effort, and a universal benefit.

Insurance needs you!

“Act as if what you do makes a difference. It does.” — William James

The first Policy 2.0 we’re publishing is ‘Renters Insurance Policy 2.0’, though more lines, countries and languages will follow. Weighing in at 2,300 words, this Policy 2.0 is almost 90% lighter than its predecessor, and hopefully far simpler, more approachable and more relevant too.

This draft was created in consultation with state regulators, and will be submitted to all states once the community has had its say. Until the approvals are in, Lemonade will continue to offer an industry-standard policy, with a view to letting users switch to Policy 2.0 once its available in their state.

But for now, please click here to read Policy 2.0 and help make it that much better. As Che Guevara put it, “The revolution is not an apple that falls when it is ripe. You have to make it fall!”

Join the conversation

• Read Policy 2.0 in full on Lemonade.com
• Contribute to Policy 2.0 on Github

---

World's First Open Source Insurance Policy was originally published in Lemonade Stories on Medium, where people are continuing the conversation by highlighting and responding to this story.

I realise that's an odd title. It comes from my increasing realisation that there are far too many people in the world who DON'T know what they're talking about, but refuse to accept that someone who is an expert in the topic might know more than them...

BBC radio 4 recently (and deservedly) got flack from scientists (and others) who were understandably outraged that Lord Lamont was brought in 'to add balance to the climate change discussion'. Lord Lamont may know a lot of things but he patently knows nothing about climate change science. And patently told a heap of lies that went unchallenged by the presenters.

Donald Trump doesn't appear to be able to distinguish between fact and fiction (and won't accept being told that). If he doesn't like something or doesn't understand it, it's 'fake news'.

Most of the UK government are not scientists and yet make wide-sweeping (and incorrect) statements or policies about a whole heap of things - neonics and the effect on insects, whether badger culling reduces the incidence of TB in cattle to name but two.
Read more »

Soil and nutrients washing off agricultural land is the single most common reason for our rivers and streams not being as healthy as they should be. Across England it accounts for a quarter of the reasons for waterbodies not achieving good status under Water Framework Directive classifications. In Herefordshire where soils are very light and agricultural practices intensive the figure is a worrying 48%.

Tackling the problem obviously involves finding out where the polluting runoff is coming from. This sounds simple but isn’t! River and stream catchments (the area of land that drains into them) are large. Often the runoff originates quite a distance from the polluted river or stream, finding its way by gravity via roads, paths, ditches and other fields. Also agriculture is a highly mobile industry. 

What is a bare arable field one year might be pasture the next, so the locations of high risk move around.

To help us find the needles in the haystack we’ve turned to modern technology. And it’s working.

Fields of bare soil with no growing crops or grass covering them are clearly much more likely to lose their soil. We can now identify where they are by using the excellent images from the Sentinel 2 satellite which was launched in 2015. 

It crosses the UK every few days so gives us an almost real time picture of land use, although we do have to have clear skies!

We can then overlay other useful information over the top of the satellite imagery which further narrows our search.

Environment Agency LIDAR data provides incredibly accurate information on the height and levels of land. For our search we identify all land that has a slope of more than 6%. This is because soil runoff is much more likely to happen when there is a steep slope.

The next step is to overlay a highly detailed map of the river network. This includes tiny ditches and streams as well as larger rivers and gives a good idea of how soil and nutrients running off the land might find their way into the main river network.

Finally we put on details of soil types. Light, sandy soils are much more likely to be washed off than heavy clays.

By doing all this we can massively reduce the area of the catchment we target for visits by our Environment Officers. 

Our first trial over a whole catchment was carried out on the River Lugg in Herefordshire. Its catchment has a total area of 170 square kilometres. Where to start?!

By selecting areas of bare soil from satellite data we reduced 170km to 30. A lot better but still a massive area to walk over!

By overlaying data layers of steep slopes and soil types we got down to 6.5km and 50 discrete fields. That’s less than 4% of the total catchment area. Much more like it!

Following some typical winter rainfall in early February we mounted an operation to send officers to visit each of the identified 50 locations. They were equipped with iPads (other tablets are available!) with pre-loaded maps and a specially designed app to upload details of what they found. This can include text, photos and video footage. A network connection allows details to be uploaded in real time, although if there’s no reception (fairly common in Herefordshire) data is stored and uploaded as soon as a connection is found.

 
Data and images can been collated immediately and activity coordinated by the operation manager back at the office and also viewed, in real time, by other teams of officers at different locations.

Initial results have been very encouraging. Of the 50 locations visited, 13 were assessed as presenting a high risk of causing immediate pollution problems. 5 were actually causing pollution at the time of the visit despite only relatively small amounts of rainfall.

During the period of the operation no high risk fields were found that hadn’t been highlighted by our targeting system.

We will be contacting all the owners of the 13 high risk locations soon.

Perhaps the most encouraging aspect of the technology (in addition to its accuracy) has been the reaction of our staff who are using it, several remarking it’s revolutionising the way they’re working and making their job far more productive and rewarding.

This is just the start of our journey in using data and technology together in the field to more accurately target our agriculture inspection and regulatory work. It should bring benefits both to the environment and the industry.

The so-called Dublin Case seeks to establish whether we can unilaterally – ie without the consent of the other 27 member states – withdraw our Article 50 notification.

What follows is a list of frequently asked questions, which I reserve the right to add to over time.

Why is the case important?

Once Theresa May has triggered Article 50 – likely to be in March of this year – we will leave the EU unless we change our mind about the wisdom of that course.

If we do change our mind, there is no doubt that, if the other 27 Member States agree, we could withdraw our Article 50 notification. But it is also possible that we have a unilateral right, which we could exercise without needing to seek their agreement. And it is obviously preferable that we have control over such a decision.

How far has the case progressed?

Because the question whether a member state can unilaterally reverse its Article 50 notification is one of European Union law, it can only be answered by the Court of Justice. And obtaining an answer involves two stages. First a national court needs to refer the question to the Court of Justice. And, second, the Court of Justice needs to answer it.

On or before 27 January we will issue proceedings in the High Court in Dublin. We are targeting a hearing in early April on the question whether it should make a reference to the Court of Justice.

If it does, we know from past experience that the Court of Justice can give a decision within three months of a question being referred to it.

Of course, these timescales will depend, in particular, on the attitude that the Irish courts takes towards the question of a reference. And the speed with which the Court of Justice is prepared to hear the case.

You can read the most recent press release here.

Is this an attempt to block Brexit?

The question seeks an answer to a legal question ‘can a state that has triggered Article 50 later unilaterally withdraw its notification?’

If the Court of Justice answers that question ‘yes’ it leaves the question ‘does Parliament want to withdraw its notification’ for our democractically elected Parliament.

So the case does not enter the political sphere – it leaves the political questions, rightly, for the politicians.

I should add that I did vote to Remain in the Referendum and, were we hypothetically to have another referendum tomorrow, I would vote the same way.

Who is funding the case?

The £70,000 anticipated cost of the proceedings before the Irish High Court was crowd-funded from small donations. There were almost 1,300 donations of £25 or less and over 1,700 donations of £50 or less.

It is possible there will need to be further monies raised to fund the Irish litigation. And if a reference is obtained to the Court of Justice in Luxembourg there will certainly need to be a further fundraising.

You can read what I said at the time of the first fundraising here.

It should be noted that the lawyers are acting at below market rates. And that I will publish as much of the material as I can. You can read our Letter to the Irish Advocate General here: ireland-letter-before-action

What is the Government’s stance on revocability?

The Government’s stance is a political one: it says that it does not intend to revoke Article 50 and David Davis claims not to know whether it could if it wanted to:

one of the virtues of the article 50 process is that it sets you on way. It is very difficult to see it being revoked. We do not intend to revoke it. It may not be revocable—I don’t know. That is the route we are going down. I expect, at least at that point, that people’s calculation will change from, “How can we make them change their minds?” to, “How can we best deal with this?”

However, it is at the very least highly surprising that it has not sought legal advice on the question (although it is possible that David Davis is not aware of that advice). And the fact that the Government does not say that Article 50 can’t be unilaterally revoked is telling.

It is also worth saying this. No one, acting rationally, chooses to make a momentous decision earlier than they need to and before they have the fullest possible evidential picture before them. A Government driven by the interests of the country should want to preserve its optionality until the last possible minute. Moreover, it is hard to understand why the United Kingdom Government might argue for an outcome that denuded itself of a unilateral right and left it instead at the mercy of the agreement of the other 27 member states.

If a reference is made to the Court of Justice, what will the Court say?

Of course, nobody knows.

If we did there would be no need for a reference. But the clear preponderance of legal opinion is that a member state could revoke a notification if it had a real change of mind.

There are many examples but:

• here’s Donald Tusk (then President of the Council) expressing a view (at 24.55);
• here’s Lord Kerr who is credited with having drafted Article 50;
• here’s the former Director of the Council’s Legal Service;
• here’s Sir David Edward, former Judge of the Court of Justice; and
• here’s a good academic law treatment.

Of course, when the country is making what Theresa May has rightly described as “momentous” decisions we ought to know – not merely think we know – the legal framework within which those decisions are being made.

Why is the case being taken in Dublin?

Whoever makes the reference to the Court of Justice in Luxembourg the answer will be the same. The same court will hear the case and the United Kingdom will have the same opportunities to advance written and oral arguments.

That having been said, the case rests on alleged breaches of the Treaties by the other member states; those breaches can only be asserted in the courts of those other member states; and the Irish courts are the natural choice because Ireland shares an operating language with the UK, has a very similar legal system, and is profoundly affected by Brexit.

What is the position of the Irish court likely to be?

I have taken this case with the benefit of advice from specialist Senior Counsel – the Irish equivalent to our own Queen’s Counsel – and solicitors experienced in obtaining references to the Court of Justice. I am confident it is the right course for me to be taking.

However, I cannot disclose their advice. That having been said, the widely respected Fellow and Associate Professor at the School of Law, Trinity College Dublin Eoin O’Dell has written:

the balance of the argument is that a referral to the CJUE is likely if a case does indeed get off the ground.

What happens if Article 50 is found to be revocable?

In her Brexit speech of 17 January 2017 Theresa May promised:

the Government will put the final deal that is agreed between the UK and the EU to a vote in both Houses of Parliament, before it comes into force.

Michel Barnier has said that the negotiations with the UK will need to be concluded within 18 months of the trigger date to allow time to ratify the deal. And David Davis agrees this is achieveable. So there will be ample time after the deal is concluded for Parliament to consider whether it wishes to accept the deal and act accordingly.

Parliament is supreme and sovereign. Should MPs or the House of Lords reject the deal they would have the (theoretical) so-called ‘cliff-edge’ option of leaving the EU without a deal. But they will also have the (viable) option of remaining in the EU. Or (more likely) putting the question back to the electorate in the form of a Referendum on the Final Deal.

A ‘Final Deal’ referendum has previously commended itself to a number of prominent Brexiters, including Dominic Cummings.

Whether one is likely to happen will be a function of the prevailing popular mood at the time Parliament is required to consider the question. As I explained here, it would be a mistake to assume that the mood today (which is broadly static since the referendum) remains static over time. We have not, after all, at the time of writing, even triggered Article 50 and many things can and will change.

Finally

You can see an interview with me discussing the case here.