Annette Bongardt and Francisco Torres (ABFT) argued about the UK Parliament: “Parliament has a strong and clear mandate for Brexit, Remainers and EU politicians shouldn’t question it“. Their reasoning can be rejected on key steps.
(1) Parliament already has a mandate
Parliament has a mandate, potentially provided by the Magna Carta of 800 years ago, though historians and legal experts will discuss aspects.
- There is no reason to argue that this mandate would vary with a particular issue.
- When someone questions Brexit then this doesn’t put Parliament’s mandate into doubt.
(2) Don’t confuse a general vote and a single issue
It might be possible for political scientists to look at a particular issue and argue that the mood in Parliament concurs with the public mood. This would require a poll targeted on that issue, both for Parliament and the public, so that we can do a statistical test on the difference of the averages. ABFT provide no such a poll. They refer to the general election, but the election concerned more issues. It is too simple to argue that the positions of the parties and candidates were clear on Brexit and that the popular votes concurred with such views.
(3) The Brexit referendum result doesn’t provide evidence
ABFT refer to the Brexit referendum of 2016 as one way how the electorate has expressed its view. However, scientists should clarify that the Brexit referendum question was flawed in its design. I explained this in an earlier LSE Brexit blog text of May 17, see here. My text relies upon a sound theoretical foundation. Representative democracy should not be confused with populist referenda, since voters in the ballot box cannot bargain.
There is no need to repeat my text here. The discussion on the weblog however was helpful in highlighting the distinction between the legal situation and the moral issue. Whether one legally invokes article 50 is a binary Yes / No issue. The Brexit referendum question was not about this legality but about policy preferences.
The referendum question wasn’t:
- “Britain has to invoke article 50 in 2017 whatever happens”
- versus “It is not so that Britain must invoke article 50 in 2017 whatever happens”.
For policy, there are various options, and electors have (conditional) expectations. With this ambiguity, the referendum question generated a situation of “Garbage in, garbage out“. We really don’t know what the electorate wants.
For example, some Scots who voted for both independence and Remain might now decide to stay in the UK with Brexit, which means both dependence and Leave and thus is quite opposite to their original views. Apparently they fear many years in the wilderness when out of both EU and UK. Clearly these voters have had conditional expectations and they have not been invoking legal rules. The Brexit referendum question clearly doesn’t provide the clarity that many attach to it.
(4) The role of scientists
Article 50 has been invoked and Brexit will happen, potentially abruptly on March 29 2019 when there is no agreement. One might argue that it is silly to continue to mention the problem of the Brexit referendum question.
However, my impression is that the 27 EU members may well be sensitive to the scientific observation that the Brexit referendum question was flawed in design. An international agreement on this observation, potentially also at the ECJ, might create an option for the UK to propose to annul the invoking of article 50.
If the UK should wish to annul Brexit before March 29 2019 then this can only be done with agreement of the 27 EU members. These members will not be swayed easily when the UK would use economic and financial arguments to explain its regret. But the issue of the referendum question is clear and impartial. The diagnosis may well be that the UK has fallen victim to a populist mood, and up to now lacks insufficient checks and balances to correct this. It fits the role of scientists to clarify this, which in itself is a check and balance.
ABFT propose that scientists defend the Brexit referendum question as sound, or at least want them to be silent, given that article 50 has been invoked and Parliament apparently has no intention of trying to annul it. I would hope that they are open to the idea that scientific criticism still would be possible.
They present events as a rational story from referendum to Parliamentarian mandate, while the events might well be diagnosed as an accidental course of history, in which the UK happened to take a one-way street because a street sign was flawed.
A YouGov poll also showed that people show loyalty to a seeming majority:
“There is a third group who change the dynamics of EU-related arguments – the “Re-Leavers.” These are people who voted to Remain in the EU and many still think that leaving was the wrong decision, but crucially now believe the government has a duty to carry out the will of the British people.”
“When taking this into account, we can split the country into three groups instead of two: The Hard Leavers who want out of the EU (45%); the Hard Remainers who still want to try to stop Brexit (22%); and the Re-Leavers (23%). The other 9% don’t know.” (YouGov May 12 2017)
Also the UK audience would be served with a better understanding that the Brexit referendum question was flawed.
(5) Democracy versus populism
My comments w.r.t. Brexit are scientifically warranted, yet ABFT in their final paragraph suggest that this criticism is “subverting the popular mandate”, “undermining the position of the EU chief negotiator” and “a subversion of the democratic process that led to the country’s decision to exit the EU”. Referring to the Brexit referendum result they ask: “Were there no lessons learnt from the rise of populism?”
ABFT are confused on the distinction between representative democracy (Parliament) and populism (referenda, neverendum). David Cameron’s decision to have a referendum was populism itself, while a democrat would have adopted proportional representation (PR), see below. See this interview by Protesilaos Stavrou on the distinction. Potentially one might use a referendum to ratify a constitutional change, but then one would tend to require that at least 1 / 2 of the population would decide, say a 3 / 4 majority of a turnout of 2 / 3. In Holland, a constitutional change, adopted (in proposal) by one parliament, requires ratification by another parliament after new elections. The Brexit referendum did not meet such standards (a turnout of 72% and a majority of 52%), apart from the design flaw on the question. It is a bit too simple to hold that there are now two UK Parliaments who are on a course towards Brexit, since article 50 was invoked by only one Parliament (referring to that flawed referendum).
ABFT are also confused on subversion. When there are serious points on content then those have to be dealt with, and it doesn’t help to neglect or denounce valid criticism.
(6) Proportional Representation (PR) versus District Representation (DR)
At first it seems as if proportional representation (PR), differing from current district representation (DR), doesn’t matter, and would be another topic than Brexit. It is a dead topic too, as John Cleese already showed in 1987, here. The ABFT article however clarifies that they entertain some assumptions that better be discussed too.
- The use of DR rather than PR is part of the lack of checks and balances that explain the populist drift in the UK. While DR strengthens existing powers, PR allows easy entry of challengers, which is the competition that economists like to see. Brexit would quite likely never have happened if the UK had had PR. In 2015 UKIP got 12.5% of the votes and only 1 seat. With a similar share of seats, UKIP would have become just one of the parties instead of the nightmare of the mainstream parties. In Holland Geert Wilders with a similar percentage has a marginal existence. If needed, mainstream politicians can agree on a Grand Coalition, like in Germany.
- ABFT suggest that support for Conservatives or Labour might mean support for either Hard or Soft Brexit but then they neglect the logic of DR, that voters must vote strategically.
- It is better to discuss annulment of Brexit within an environment with PR than one with DR. Britain might have a year to redesign its electoral system and have proper elections.
PR is the key design criterion for democracy, and districts are mainly a historical hangup. Consider n voters and s = 650 seats, so that the quota is q = n / s. This allows s / 2 = 325 districs of size 2q, since s / 2 * 2q = n again. With 45 million voters there might be an average distict size of about 140,000, correcting for turnout. Let a district candidate be elected when he or she has more than the quota q. Normally this amounts to q / (2q) = 50% of the district. One might allow a lower turnout in districts, and let a candidate be elected with at least 50% of the local vote. The unfilled seats can be filled by the nationwide PR criterion. Thus, a fundamental choice for PR still allows a feature of districts if so desired.
In France, the UK, USA and India DR is the key design feature whatever happens to PR. They haven’t overcome the historical hangup yet.
The distance of DR to PR can be expressed in the PR Gini coefficient, see the excel sheet and graphs. David Cameron in 2015 had a coefficient of 29.7%. In 2017 Theresa May has a PR Gini coefficient of 15.6% while Holland has 3.6%. The UK thus is strikingly disproportional.
The major non-PR impact in Holland is from the 2% wasted votes for small parties that get zero seats. In 2017 the UK had 3.5% wasted votes for parties that got no seats. One might argue that the wasted votes should be omitted from the Gini, yet, they rather stand out as a sore spot in current representation. A proportional representation of the wasted vote w in total n is possible by leaving seats empty or by filling the seats and taking a qualified majority f = 1/2 / (1 – w / n). In 2017 f = 50% / 96.5% = 51.8%. A representative majority in a full House of 650 seats then requires 337 seats, and not 326. See here.
There is a curious miscomprehension in the UK about what democracy actually is.
(a) This miscomprehension also exists in the Electoral Reform Society (ERS). There is a ranked system called “Single Transferable Vote” (STV) that has some PR properties when applied to the whole country. ERS however applies it to districts, and DR destroys nationwide PR again. Incongruously, the ERS keeps saying that their STV would be PR while it isn’t. See my criticism and counterexample here. The continued support of ERS for districts is part of the confusion that blocks the change to PR. (And with 4 candidates per district, their districts might need to have some 500,000 voters, curiously undermining the reason of “closeness” to have a district.)
(b) The 2011 referendum on PR was actually a referendum on the method of “Alternative Vote” (AV), which is a suboptimal suggestion too.
(c) It is apparently a misconception and dogma amongst UK voting theorists too that the (more complex) methods of ranked voting like STV or AV must be used at the level of voters. There is no need for this. It would suffice if the UK would adopt the Dutch system of Open Party Lists. The (more complex) methods of ranked voting can be applied by the professionals in Parliament, if needed, and those methods would still be proportional if the party weights were proportional.
(d) There is now an initiative “Make Votes Matter” (MVM) that strictly targets PR but that suffers from such conceptual problems too.
- ERS signed their declaration though ERS, as said, also proposes a system of STV with districts that is not PR.
- MVM might generate a discussion on ranked systems while such complexity might block PR again. It is better to clearly state a preference for “PR in Open Party Lists”, for otherwise we might end up with another situation that it is unclear what people signed up or voted for.
Parliament’s mandate comes rather from Magna Carta, not Brexit, and scientists have every reason to question both Brexit and the lack of proportional representation (PR).
I now notice the contribution by Jonathan Portes on the LSE Brexit blog about the influence of the Condorcet paradox (June 15 2016, that appeared on the NIESR blog on June 6 2016). I referred to voting theory too, though after the referendum of June 23, in this comment of June 29 2016, that was later summarized on the LSE Brexit blog in May 2017.
- Portes suggests that there is a Condorcet paradox, while my argument only mentions that there might be one, and that we basically don’t know, even after the Brexit referendum “outcome”. The argument by Portes is less strong since he provides a speculation, and others might argue that it is only a speculation.
- My main reason to refer to the Condorcet paradox was to explain that many voting theorists themselves do not properly understand voting theory. See the argument about the distinction between voting and deciding. Indeed, Portes repeats the confusion: “We may not be inconsistent individually but we can be so collectively.” Potentially a collective might be inconsistent, of course, as an individual might be. Yet, the point concerns the distinction between intransitive vote counts and the transitive decisions (e.g. indifference) based upon such vote counts.
For the Brexit referendum question, the problem might likely be a bit different. With R for Remain, S for Soft (EEA) and H for Hard (WTO), there are 6 possible strict preferences, from R > S > H (Theresa May before the referendum) to H > S > R (Theresa May after the referendum). If S and H are collected in L then there arises the claimed binary choice between R and L. Voters who are in the categories S > R > H or H > R > S would face a hard question, and might decide not to turn out, or develop assumptions about what the L actually might mean given what what they think about future developments. Similarly for the versions of R.
Overall, the argument suffices that Proportional Representation is the proper method for modern democracy, and that referenda are an instrument of populism, notably within systems with District Representation. See the next weblog entry.