On military genius

There is a way of learning about warfare that many men (and a few women, but this is an overwhelmingly male approach) have from reading popular histories of warfare, from certain TV documentaries (especially from the era of the History Channel when it was nicknamed the “Hitler Channel”) and from lots of wargaming, both table-top in the seventies and eighties and computer since the mid-eighties. This is about almost ignoring the political/diplomatic context of wars and perceiving the economics as purely about the ability of national economies to supply the military machine. If your first thought about the US economy in WWII is about the mass manufacture of Sherman tanks and comparing that to the higher-quality tanks produced in lower numbers by the German economy, then this is the way of learning that you have.

This approach thinks mostly about battles in terms of the military skill displayed by the commanders on both sides.  As a grognard wargamer, I absolutely acknowledge this approach, but it isn’t history. It’s a form of art appreciation. The beauty of Hannibal’s double envelopment at Cannae, or of Rommel’s use of the desert to create encirclements in battle after battle in the North African campaign, or of Napoleon at Austerlitz, or of Grant at Vicksburg: does it really matter that these decisive victories were won with style, where Gettysburg, or El Alamein, or Friedland, or Zama, or were not?

But there is a real lesson, though: decisive victories – excepting those achieved by overwhelming numbers (e.g. the German defeats of Luxembourg in the two World Wars) nearly always come from the enemy commander making a bad mistake.

If you look at good generals, they create situations where there is an opportunity for the enemy to make a mistake, but where they themselves are not committed until the enemy does. If you want to see flashy maneuvring on both sides, then look at battles and campaigns where both generals were bad. Campaigns where both generals are good see them offering battle at an advantage and it being declined, until eventually the battle that is actually fought becomes a grinding battle of attritition and of the skill of lower-level officers and NCOs. Battles like Gettysburg or the Wilderness in the American Civil War, like El Alamein or Tunis in WWII, like Zama in the Second Punic War, like the battles of Alte Veste, Fürth and Lützen in the Thirty Years’ War – all of these involved skilful generals on both sides, and in none did an army get outmaneuvered. Victory came down to the quality and numbers on each side. These battles tend to be bloody even when there is a clear winner on one side or the other.

When I first started looking at history from this perspective, I was full of admiration for the so-called “military geniuses”. I can still name a few dozen. But, the more I looked at battles, the more it became clear that the big gaps in skill weren’t among the good generals, but the bad ones. Chancellorsville is more about how bad Hooker was than how good Lee was. The decisive battles of Sedan are about Napoleon III and Gamelin’s lacks of military skill, not about von Moltke or Manstein/Rundstedt/Guderian’s genius. Cannae is much more about Paullus and Varro than about Hannibal.

When people talk about an admiration for Robert E. Lee because of his military genius, I certainly understand that of which they speak – Fredericksburg and Chancellorsville are both fine works of the military art – but I no longer regard this as a high art where a handful of greats can exceed the merely good, in the way that a Mozart or a Beethoven can in music. I take the view that the few generals who look greater than other competent and skilful generals do so more because of the (lack of) qualities of their opponents. The result of this is that I’m much less inclined to honour great generalship for itself.

Also, of course, I’m interested in history and not just in the appreciation of the art of generalship. Saying that Erich von Manstein was a great general is like saying that Leni Riefenstahl was a great film director – it doesn’t in any way excuse the appalling nature of the cause for which he fought and she directed films. The same can and should be said for Lee and Jackson and Longstreet commanding the armies of the Confederacy.


A Brexit Plan

As I’m sure most people will know, I’m a Remainer and a pretty hardcore one – I’d like to join the Euro and Schengen, and I think that referendums are of pretty dubious legitimacy (and especially so for ill-defined binary choices used to claim democratic legitimacy for things that are being made up years after the votes were cast).

However, if I’m being forced to write a Brexit plan at gunpoint and I’m not allowed to write either “don’t do it” or “just stay in the EEA you fools”, then here goes:

  1. Slow down. Don’t trigger Article 50. If you already have (you fools), then negotiate for a long extension. If you can’t do that either, then EEA (on Norway’s terms and only very minor fiddling around the edges, because the 2 year time limit is very narrow and there isn’t time to negotiate much) as a transition with no defined end date, but rather a defined end state.
  2. By “defined end state”, I mean that the transition ends when we have an FTA with the rest of the EU and FTAs or other bilaterals/multilaterals with enough non-EU countries to feel comfortable (ie at least USA and China), and we have also agreed terms for an individual membership of the WTO. I’d expect that to take at least a decade and possibly two.
  3. Now we get to the difficult question.

Northern Ireland

Honestly, if I was advising the Tories, I’d say “Have a border poll and then get Boris Johnson to say a bunch of offensive things a week before polling day, and then it won’t be a problem.” but supposing we’re not allowed to just get rid of the place, the basic problem is that you can’t set up border patrols (because that’s a hard border and a breach of the Good Friday Agreement).

The current solution has been the CTA (a sort of mini-Schengen for the UK and Ireland), which will pretty much work fine for allowing people to physically cross the border at will (anyone with a visa from either the UK or Ireland is already allowed in the other country, so as long as EU citizens get visa-free travel to the UK as tourists then we don’t need border patrols) but there is then the problem of goods. If goods in a certain category can be imported into the UK  but not the EU (or vice versa) or would pay much more in tariffs in one or the other, then having no border controls at all would let people import through whichever is cheaper (or legal) and then load it on a lorry and drive over the border. If you can then go from NI to GB or from Ireland to the mainland EU by ferry, then Ireland could be used for a massive exercise in tariff evasion and even getting around legal restrictions if regulations vary between the UK and the EU after Brexit.

This is solvable for people, because short-term stays aren’t a big deal (and long-term stays have to be deal with in-country rather than at the border). But it’s harder for goods. Now, as long as the UK continues to have high regulatory standards (even if different from the EU) so it doesn’t become a massive loophole for dangerous goods (etc), then there might be an agreement to be made. For instance, suppose the UK agreed to require anything in the UK that isn’t allowed in the EU to be labelled as such (ie “Not for Sale in the EU”) and the tariffs are kept close enough that the extra costs of importing through the other country make tariff evasion rarely cost-effective in country (and you run the risk of getting caught – see tobacco smuggling within the EU as presently constituted) then it might be workable. That would be a tough and long negotiation. Remember, you could land a container at the new Liverpool2 port, load it on a lorry, drive to Birkenhead, cross to Belfast on a ferry, drive to Dublin and ferry to Cherbourg. So as long as that costs too much compared to just driving to Dover and paying the tariff, plus the risk of being caught for tariff evasion, then there might be a deal to be done between the UK and EU to allow there to be no customs at the Irish border.

You’d still have to do a deal on Ireland-Northern Ireland goods, but you could almost certainly make a deal where goods produced in either can be sold in the other without restriction. Both Ireland and NI would politically favour that in a big way.

But getting everyone to agree? You can see why I’m setting aside a decade!

Our European Future

This is a policy paper from a group I chaired, which became the policy of LDYS (Liberal Democrat Youth and Students, as they then were) at their Autumn Conference in 1997.  It’s not policy of anyone, and I no longer agree with all of it, but I’m putting it here so there is a copy available on-line; it was only available on paper hitherto.

Introduction and Remit

The LDYS EU working group has a remit limited to the consideration of the structures of a federal EU government and to the external aspects of foreign and security policy. We intend to present other papers dealing with other specific policy areas to future conferences. The general question of the division of powers between the European, national and regional levels is expressly excluded from the remit of this paper,
since it would in effect require a policy paper that covered everything – considering who should make a decision is difficult without thinking about the sort of decision they should make. Most decisions can already be taken at EU level, but the member-state veto tends to result in few actually being taken. In other words, it is the structure by which the EU takes decisions, rather than a lack of actual power to take them, that keeps so much decided by the member-states. We hope that future policy papers and motions will consider not only what LDYS policy should be, but where the power to take such decisions should reside – local, regional, national or European level.

Section One: Structures

The European Union’s structures are in desperate need of reform. The EU was established as an international agency, in the same manner as UN agencies such as UNESCO. This means that it operated by the consent of every member state and its decisions were implemented by the member states. Over the years it has evolved in to a body that is much more governmental in its nature. It has not changed its structures adequately to allow for the changing role of the organisation and for the expansion
from the original ECSC of Six to the current EU of Fifteen and probable future expansion to twenty or more. The most powerful body in the modern EU is the Council of Ministers; still the lineal descendant of the diplomatic conferences that established the EEC in the first place. Even in the days of Jacques Delors, the most powerful person within the EC was the head of government of a member state – Helmut Kohl. This structure is illiberal, unrepresentative and undemocratic.

The European Parliament over-represents the smaller member-states and underrepresents the larger ones. This – and the similar imbalance in the Council of Ministers when operating by Qualified Majority Voting (QMV) derive from a compromise between equality between states and equality between citizens. This compromise is false.

LDYS propose the following reforms to the structures of the EU:

1. In the immediate future, LDYS will pursue the following:

  • Incorporation of the European Declaration of Human Rights into EU law.
  • Opening up the Council of Ministers to public scrutiny.
  • Equalisation of the membership in the European Parliament and a common electoral system of STV in multi-member constituencies.
  • Increasing the powers of the European Parliament and in particular increasing the usage of the co-decision procedure
  • Increasing the use of QMV in the council and lowering the threshold for a qualified majority from five-sixths to two-thirds.

In the longer term, we aim for a complete restructuring along the following lines:

2. To replace the Council of Ministers with a Senate: Each member state would have five representatives in the Senate, elected by the national parliament immediately after each general election on a basis of STV.

3. To adjust the membership of the European Parliament so that each MEP represents a million citizens. The smallest member state should still have one MEP. A common electoral system of STV in multi-member constituencies should be introduced. It would be elected for fixed four year terms. The European Parliament should decide for itself where its sittings should be.

4. The European Parliament would have the final decision on all legislation, although the Committee of the Regions and the Economic and Social Council would be consulted.

Once passed, it would move to the Senate, which can accept, amend or reject. If the Senate amends or rejects a bill, it would return to the Parliament which can then pass it in whatever form it chooses without further reference to the Senate.

5. A new European executive should be established to replace the executive responsibilities of the Commission and the Council of Ministers. The administrative aspects of the Commission would continue as a civil service. It will be called the “European Cabinet” and will be directly elected by the European Parliament by a separate AV vote using a secret ballot for each position. If there is a vote of no
confidence in the Parliament, then there are new elections to the Cabinet. Members who resign would be replaced in the same manner. The European Parliament would decide number of cabinet ministers, the titles of the various ministers and their exact responsibilities and whether or not to have a “Prime Minister”.

6. All of these structures should be part of a federal constitution, which should also incorporate a “Declaration of the Rights of the European Citizen”. The European Court of Justice would be a court of final appeal in regard to both the constitution and European legislation. It will consist of one judge from each member-state, appointed by their national judiciaries. A President will be elected by the judges from amongst their own number. Cases will be heard by groups of five judges, sitting as the Court of First Instance. Several such Courts can hear cases at the same time. In cases regarding the
constitution where there has been a split verdict or where two judges who did not hear the case at first instance request it, there would be a right of appeal to a hearing of the full court.

7. The constitution can be amended by a two-thirds majority in a referendum called by the European Parliament.

Section Two: Foreign and Security Policy

The European Union does in theory have a Common Foreign and Security Policy (CFSP). At the moment, this means that the Foreign ministers have regular meetings to co-ordinate their policy; it has also involved the mission to Bosnia-Herzegovina. There is some degree of co-operation between the diplomatic and particularly consular services of the member-states.

In the Security (that is, defence) arena, the major developments of recent years have been the work of Panavia on the Tornado and now the Eurofighter and the setting up of Eurocorps. This is a joint force consisting of contingents from several member states.

LDYS propose:

1. That the diplomatic and consular services of the EU member-states increase their co-operation. In particular, in some parts of the world it is proving very expensive for even the larger countries to have an embassy in every state. If embassies and consulates in some of the more distant parts of the world were merged into EU embassies and consulates, it would be possible to have full-time ambassadors in every country in the world (only the USA can do this at present). Also an EU consulate in some of the smaller or more remote cities would be far better than three member-state consulates in one and then none in the next city. Of course, member-states would retain independent embassies and consulates in the more important countries and cities (Washington, Moscow, Beijing, etc.)

2. That the EU be given full, rather than advisory membership of the UN and its agencies (UNCTAD, UNESCO, WHO, etc.), with an eventual aim of a permanent seat on the Security Council, replacing the UK and French seats.

3. An EU foreign policy should be a matter for the EU minister with responsibility for foreign relations, and the Foreign Ministers of the member-states will be responsible for their own policies.

4. There should be a single European agency for defence procurement, to ensure effective co-ordination and to reduce the political pressure for national preference. European militaries are bottom-heavy; they rely on the USA for top-level command, control and communications (C3). Also, the diversity in weapons systems is extremely wasteful; a European agency would resolve many of these problems.

5. The process of European integration was started in order to and will continue to seek to prevent and avoid war. European trading of arms with oppressive regimes should stop at once.

6. Nuclear weapons should remain exclusively a matter for the member states. LDYS restate their desire for an end to all nuclear weapons.

Worst-case scenario, February 2017

Content note: This is a worst-case scenario. People die.

If this would stress you out, don’t read it.

So, the story so far (ie, this is what has happened in actual January 2017).

President Trump issued an executive order that was so blatantly discriminatory that it couldn’t pass Rational Basis review (that’s the lowest level of review under the fourteenth amendment’s guarantee of the equal protection of the laws).

Customs and Border Patrol started holding various people in detention on landing at the airports.

The American Civil Liberties Union went to a whole bunch of federal district courts, for the various courts covering key airports like JFK in New York, Dulles in Washington DC, etc.

The judges gave out emergency injunctions requiring CBP to release people from detention, give them access to lawyers, etc.

In some cases (e.g. Dulles), CBP ignored the court orders and continued as they were, including deporting at least one person in defiance of a court order.

So that’s where we are as of Monday morning, 30 January 2017.

This is where the fiction starts.

The ACLU goes to the Federal District Court for the Eastern District of Virginia (EDVa) and asks for contempt proceedings. Given that there cannot be a civil remedy for the deported person, this means criminal contempt of court. Given the significance, a three-judge panel, headed by Rebecca Beach Smith, Chief Judge of the court, is convened, and issues an arrest warrant for “person or persons unknown, being or claiming to be officers of the United States Customs and Border Patrol, preventing the execution of an order of this Court” and directs Federal Marshalls to go to Dulles to execute the warrant.

There are three phone calls made simultaneously.
The ACLU contacts Terry McAuliffe, the Governor of the Commonwealth of Virginia, asking him to consider putting the Army National Guard of Virginia at the disposal of EDVa to enforce the law.
The US Attorney, who has just failed to prevent the warrant being issued, calls the Federal Circuit Court of Appeals for the Fourth Circuit to apply for a writ of certioriari to stay execution of the warrant awaiting appeal. This takes some time to be issued.
The CBP officer who was representing his department there calls Dulles to warn them that the Marshalls are coming.

Four Marshalls arrive at Dulles with their warrant, and CBP officers draw their weapons and refuse them access. At this point the Supervisory Deputy US Marshall (commanding) is handed a telephone by an ACLU lawyer. On the other end of the line is a Lieutenant-Colonel in the Virginia Army National Guard, informing him that two hundred guardsmen of the 2nd Squadron, 183rd Cavalry Regiment were on the road from Staunton and would be at Dulles airport within an hour.

The Marshalls decide to wait rather than start shooting at CBP, but do not inform CBP what they are waiting for. CBP assumes that they have stopped the Marshalls and are in a stand-off and, rather nervously, continue to process passengers.

The US Attorney finally gets his writ of certiorari, and heads to the airport to break up the stand-off. As she arrives, she hears the first gunshots…

When the VANG arrive, they place themselves at the disposal of the Supervisory Deputy US Marshall to enforce the law. He goes to the CBP and informs them (through a bullhorn) that he will be compelled to use force if they do not comply. He gives them a few minutes to clear civilians out of the way. The cavalry are also preventing civilians from entering Dulles by land at this point, which is what finally alerts the media.

The CBP duty manager, finally, gives some urgency to their higher authority over the phone, warning them that there are going to be four dead US Marshalls pretty soon and they’d better be prepared to deal with that. He hasn’t noticed the National Guard. He is given orders to resist with force. Who gave those orders will ultimately be a matter of some contention.

The Supervisory Deputy US Marshall decides that the clock has run out, and the first Stryker reconnaissance vehicle rolls into the arrivals area of the Dulles terminal.

Stryker Armored Reconnaisance Vehicle
This is not a tank, honest. See, it has wheels, not tracks.

At this point, CBP panic. Unfortunately, armed federal officers panic by opening fire. The Stryker returns fire with its machine gun, killing 12 CBP officers, including the duty manager, two civilians and injuring what are ultimately determined to be no less than 91 other people, many of them actually taking injuries not from bullets but from diving to the ground to avoid the bullets that are flying.

The surviving CBP officers surrender, and the First Battle of Dulles, the first action of what would later become known as the Second American Civil War is over.

At this point, Donald Trump is finally informed that there has been a shoot-out at Dulles. The smart response – entirely within his powers – would be to federalise the Virginia National Guard, order them to stand down, send some new CBP officers, and reopen the airport.  Instead, advised by Steve Bannon, he heads for AF1 ordering them to take him to a secure location (ie to STRATCOM in Nebraska), declares Virginia to be in rebellion and orders the US Army to retake Dulles from the rebel forces.

Given the distance from the Pentagon to Dulles (not much), there are real tank regiments wrecking freeways on their way to the airport within a few minutes. ATC diverts all civilian aircraft away, while F-16s are being launched from Andrews Air Force base to provide top cover.  The Second Battle of Dulles lasts about as long as the first, as Abrams and F-16s go through a reconnaissance troop like a hot knife through butter.

M1A2 Abrams Tank in desert colours
This is a tank. An Abrams M1A2 to be precise.

Since this is a worst-case scenario, the US military now has conflicting orders, and divides. STRATCOM orders all nuclear elements to stand down and to refuse all launch orders until the political situation is clear (even in a worst case scenario, the US isn’t nuking itself). A whole series of states with Democratic governors end up in rebellion, while the midwest tears itself apart. Large parts of the official military disobey all orders and refuse to fire on fellow Americans and a chaotic multi-sided civil war lasts nearly a year until Trump finally ends up swinging from a lamppost by his heels.

On European co-operation

I did this on twitter earlier, so it’s a series of tweets.

In case you’re wondering why 1815 is a relevant year, that’s the year of the Battle of Waterloo, when Napoleon was finally, definitively, defeated, and the almost-continuous wars of the previous 23 years ended. Exhaustion and the mass slaughter (the total was in the region of five million military deaths, from a far smaller population than the fifteen million of WWI) meant that European countries were looking for a more reliable way of keeping the peace.

There had been European wars in the mid-century period, and one battle, Solferino, inspired the creation of the Red Cross. But the total deaths in European wars between 1815 and 1914 add up to less than a million.

To be fair to this period, most of the so-called “laws of war” (the Geneva and Hague conventions, in particular) come from this time.

World War I.

And, to some extent, Italian Fascism. But really, it was Manchuria in 1931 that broke the League.

There’s more, a long, long list of more international institutions. Including several specifically European ones, like the ECHR.


One more tweet (to get the formatting to work):

Voting systems and the French presidency

I’m seeing some people criticise the two-round system in France for the way it enhances the Front National’s prospects.  This is reasonable, but they go on to suggest AV, which wouldn’t make it better. France has a host of problems, and changing the voting system for president won’t help any of them.

French politics has the same problem as most countries – the median voter is centrist, but very few others are.

If you look at the left vote in the opinion polls – and that’s 43% or so of the country – then 20% are intending to vote for a candidate to the left of Hollande. Hollande was originally a candidate from the left of the PS and has largely governed as such. Think more Ed Miliband than Corbyn, but still on the left of the main French left party. Yet nearly half of the left electorate will vote for candidates to his left – mostly (15% or so) for Mélenchon (who is ex-PS and left in protest at how rightwing the party was before Hollande), with a few (2%) voting for the Greens or one of the two Trotskyist candidates (about 3% between them).

This leaves Hollande on 8%, with Macron (essentially a Blairite in UK terms, though I’m sure he’d reject that – still, he left Hollande’s government in protest at how left-wing it was) on 15%.

So if you have AV, how are those transfers on the left going to work out? Well, the Trots would go to Mélenchon, the Greens will likely do the same, which means that Macron would need to either pick up transfers from the right, or win over virtually all of Hollande’s vote to stay ahead.  I’d guess that neither of those happens, and Mélenchon ends up ahead of Macron.

And on the right?

Obviously, Le Pen is on 25% and getting lots of publicity, but the mainstream right has more votes – Juppé was on 28-30% in polls because the centrist candidate (Bayrou) had a pact to step down for Juppé. However, Bayrou will run against Fillon (and would have done so against Sarkozy), so that splits out at 20% for Fillon and 8% for Bayrou.  If you add the numbers up, you get 96% – this is because there is Dupont-Aignan, who is a euroskeptic rightwinger with about 4%. His vote is one of the very few for whom Le Pen is transfer-friendly.

So the next question is how the Bayrou transfers move – there aren’t enough of them to put both Fillon and Macron ahead of the two extreme candidates. I suspect they mostly go to Fillon. At which point, Macron is eliminated in fourth place (numbers will be something like Mélenchon 23, Macron 22, Fillon 26, Le Pen 29). Now, there might be enough transfers there to put Le Pen from first to third – at which point Fillon wins in a landslide, or it could end up with Mélenchon v Le Pen, which is probably the only way Le Pen can win a final round.  If somehow Macron sneaks past Mélenchon, then he will grab transfers from first Mélenchon, then from Fillon and win a landslide over Le Pen in the final round.

Regardless of the system, the odds are heavily that one of the two moderates (Fillon or Macron) wins, and the ideal system would have them face each other in the final round – after all, if they’re the only ones who can win, shouldn’t the final decision be between them?  The problem is that when you get down to the four candidates in double digits, there is probably certainly a majority who prefer either Mélenchon or Le Pen (my back-of-an-envelope has this as about 52-48), which usually means that either Macron or Fillon gets squeezed out in fourth (probably Macron).

There is no step in the process – either in AV or in the French two-round system – that is likely to result in a Macron v Fillon contest; they are both trying to consolidate their left and right coalitions, and usually one of them fails.

Condorcet, incidentally, probably elects Bayrou as President, because that’s where the median voter is. I don’t think there are a lot of people who prefer both Fillon and Macron to Bayrou, which is what you need to stop Bayrou under Condorcet.

Most systems do elect Fillon – Le Pen will really struggle to break past 32-33% of the first-round vote in the second round – so even if a lot of the left refuses to vote for Fillon, he only needs about 20% of the left to switch rather than stay at home.

Just to be clear, this election is very close to being a pathological worst-case scenario for voting systems. It’s the sort of election that gets used in proofs of Arrow’s Theorem.  Criticising the two-round system is fine; it’s not a good voting system. But don’t go thinking you have a better system, because I’ve looked at a whole bunch (including approval) and they’re all awful – there are so many ways to elect one of the extreme candidates very narrowly.

My solution, by the way: don’t have a Presidency in the first place. Have a proportional election to your parliament and then you can work the mess out in coalition negotiations. You probably end up with a left-right coalition of everyone bar the extremes (a coalition from Hollande to Fillon has a bare majority), which is, well, the same as Germany’s fairly effective current government.

Quick set of rough analogies for British readers:
Trots : Trots
Mélenchon: Corbyn
Greens: Greens
Hollande: Ed Miliband (or maybe Gordon Brown in 2010, worn out by governing)Macron: Blair (young Blair of 1997, but less charismatic and in a much tougher political environment)
Bayrou: Nick Clegg
Fillon: Thatcher.
Dupont-Aignan: the sort of Tory/UKIP backwoodsman who can’t shut up about Europe
Le Pen: Farage’s younger, better-looking, be-suited successor

And for American readers:
Trots: Trots
Mélenchon: Sanders but leftier.
Greens: Greens who care more about the environment and less about leftiness
Hollande: Sanders if he’d been President four years and was worn out by failure
Macron: Bill Clinton in 1992. Cory Booker if he was white.
Bayrou: Huntsman? Do you have any liberal Republicans left?
Fillon: Ryan is probably the closest, but much more boring. I can’t name someone, but he’s a non-evangelical religious conservative who mostly cares about economics?
Dupont-Aignan: Imagine if Trump lost and a boring incompetent tried adopting his coalition.
Le Pen: Imagine if Trump lost and competent people tried adopting his electorate.

On referendums on constitutional change

In the UK, we have had two sorts of referendum on constitutional changes. We’ve had ones like the 2011 AV referendum, where a full Act of Parliament has been passed, and the referendum is on whether or not to implement it. At this point, we, the voters, know (insofar as it is possible to know anything in politics) what the consequences of the decision will be.

The other sort are like the 2016 EU referendum or the 2014 Scottish Independence referendum, where the specifics haven’t been spelled out. Now, to be fair to both of these, they would involve international negotiation to implement, and the international negotiators would not be much inclined to do that negotiation without a reasonable prospect of it being implemented.

Somewhere in-between are the 1997 referendums on Scottish and Welsh devolution, where the terms of the deal had been worked out, but there wasn’t a complete Act of Parliament. Still, I don’t think many people who voted in 1997 would have said that the Scottish Parliament and Welsh Assembly weren’t what they voted for.

Whereas, I suspect that quite a few people who voted on the EU referendum are saying “this isn’t what I voted for”. And, equally, quite a few people voted against Scottish independence on the grounds that they didn’t know what it would look like. I think this kind of referendum is a problem, and, after the EU referendum, I think I have worked out a solution to the having these sorts of questions resolved through referendum without creating the problem of voting for one thing and getting something else altogether.

My proposal is that, as a constitutional convention, we should hold two sorts of referendums on constitutional changes.  One sort are what I call a “ratifying referendum”. A ratifying referendum is one where there is a specific proposal, either a complete Act of Parliament just awaiting the referendum result to implement, or a detailed proposal like the Scottish Parliament and Welsh Assembly.

As a general principle, where possible, we should be presented with ratifying referendums, where a specific, detailed proposal is placed before the people to say “yes” or “no” on.  In Parliamentary languate, a Third Reading not a Second.

The other sort are what I call “pre-negotiation referendums”. These are a referendum on whether, in principle, to do something. In general, these should be avoided, as they risk the whole “not really knowing what you’re voting for” problem, but there is a case where they are legitimate, and why I’m calling them “pre-negotiation”. If, to implement the change, representatives of the people have to negotiate with outside powers – ones outside of the electorate of the referendum, that is, then those outsiders (other countries, or other parts of the UK) could, quite legitimately, refuse to negotiate seriously if the government was saying “we’ll put this to a referendum”, especially if the polls say it will be rejected.  However, in this case, there should also be a requirement for a ratifying referendum once the details have been worked out.

So, for Scottish independence, there would be a vote on whether in principle, Scotland should become independent and start negotiations to that effect. Then, before making any irreversible decisions, the Scottish Government would return to the people (in, just to be absolutely clear, a second, ratifying, referendum) with a detailed, agreed, negotiated plan – agreed with the British government and with all the relevant international institutions (EU, UN, etc). Then the Scottish people could ratify that, or not.

The same would apply for leaving the EU – they wouldn’t trigger Article 50 until the second referendum (and, equally, Cameron’s renegotiation wouldn’t happen until after the first; so you could have two parallel negotiating tracks with the British people able to choose between them).

Obviously, it could be that the EU refuses meaningful negotiations without triggering A50, but in that case, you’d put to the people explicitly that we’d trigger A50 immediately on a Leave vote and be out on our ear two years later with no prospect of retaining any relationship with the EU.