Archive

Tag Archives: model versus reality

The dictum is to have one subject per letter. This paradise is no longer possible when time passes and letters and subjects accumulate. Let me take stock of some findings on democracy.

Economic theory needs a stronger defence against unwise application of mathematics. Mathematicians are trained for abstract thought and not for empirical science. Their contribution can wreak havoc, for example in education with real life pupils and students, in finance by neglecting real world risks that contribute to a world crisis, or in voting theory where they don’t understand democracy.

Nowadays, though, I am also wary of students from the Humanities who rely upon legal views (their version of mathematics) instead of empirical understanding.

For the following, distinguish single seat elections (president, prime minister) and multiple seats elections (parliament). There is also a key distinction between Equal Proportional Representation (EPR) with proper elections and District Representation (DR) that has contests rather than proper elections.

Key findings

(1) Montesquieu’s Trias Politica of the separation of powers is failing, and we need the separation of a fourth power, an Economic Supreme Court, based upon science, with a position in the constitution at the same level as the Executive, Legislative and Judiciary. The current setup allows too much room for politicians to manipulate the information for policy making. This need for separation can also be proven logically in a model using stylised facts, see the book DRGTPE. A short discussion on the 2007+ European crisis is here.

(2) Kenneth Arrow in his Impossibility Theorem has a correct deduction (there is an impossibility) but a wrong interpretation. He confuses voting and deciding. For this debunking of Arrow’s Theorem, see Chapter 9.2 of Voting Theory for Democracy (p239-251). Sheets of a presentation in June 2018 are here.

(3) A voting method that many might find interesting is the Borda Fixed Point method. See the counterfactual example of selecting a Prime Minister for Holland.

(4) Political science on electoral systems is no science yet but still locked in the Humanities, and comparable to astrology, alchemy and homeopathy. People in the USA, UK and France still have taxation without representation.

(4a) The key paper is One woman, one vote. Though not in the USA, UK and France.

(4b) A supportive paper develops the SDID distance measure for votes and seats.

(4c) This paper reviews the role of statistics for the latter measure. Sheets of a presentation in June 2018 are here.

(4d) An earlier comparison of Holland and the UK in 2010 (update 2015) contains a major stepping stone, but is not as critical as (4a). This analysis resulted in a short paper for Mathematics Teaching 222 (May 2011) at the time of the UK referendum on Alternative Vote.

Minor results because these lead to dead ends

(5) There are some supplementary findings, that I do not regard as major, but as roads that you might need to walk in order to discover that they do not lead far.

(5a) There are Two conditions for the application of Lorenz curve and Gini coefficient to voting and allocated seats. The Lorenz curve is a neat way to graphically show the disproportionality and inequality of votes and seats. The Gini is its associated measure. However, above measure SDID is to be preferred, since it is symmetric and doesn’t require sorting, has a relation to the R-squared and the Weber-Fechner law.

(5b) We can compare votes and seats but also use a policy distance. A crucial question is who determines the distance between policies ? When we have a distance, how do we process it ? I am not convinced by the method, but a discussion is here.

(5c) The Aitchison geometry might present a challenge to SDID. This paper provides an evaluation and finds this geometry less relevant for votes and seats. Votes and seats satisfy only two of seven criteria for application of the Aitchison distance.

(5d) This paper tries to understand the approach by Nicolaus Tideman and compares it with the distinction between voting and deciding.

(5e) Mathematician Markus Schulze was asked to review VTFD but did not check his draft review with me, which caused needless confusion, see here and here. PM. Schulze now has this 2017 paper, but doesn’t refer to Borda Fixed Point, perhaps thinking that he understands it, but he apparently is not open to the diagnosis that his “review” is no proper review.

Conclusion

For the above, it is pleasant that a distinction can be made between key results and findings about dead ends. I listed my debunking of Arrow’s Theorem as a key result, but it also identifies this theorem as a dead end. Thus, it is also a matter of perspective. When you are at the dead end, and turn around, the whole road is open again.

PM. Earlier weblog entries on democracy are here.

Advertisements

Karl Pearson (1857-1936) is one of the founders of modern statistics, see this discussion by Stephen Stigler 2008 (and see Stigler’s The Seven Pillars of Statistical Wisdom 2016).

I now want to focus on Pearson’s 1897 paper Mathematical Contributions to the Theory of Evolution.–On a Form of Spurious Correlation Which May Arise When Indices Are Used in the Measurement of Organs.

The main theme is that if you use the wrong model then the correlations for that model will be spurious compared to the true model. Thus, Pearson goes at length to create a wrong model, and compares this to what he claims is the true model. It might be though that he still didn’t develop the true model. Apart from this complexity, it is only admirable that he points to the notion of such spurious correlation in itself.

One example in Pearson’s paper is the measurement of skulls in Bavaria (p495). The issue concerns compositional data, i.e. data vectors that add up to a given total, say 100%. The former entry on this weblog presented the inequality / disproportionality measure SDID for votes and seats. These become compositional data when we divide them by their sum totals, so that we compare 100% of the votes with 100% of the seats.

Pearson’s analysis got a sequel in the Aitchison geometry, see this historical exposition by Vera Pawlowsky-Glahn and Juan José Egozcue, The closure problem: one hundred years of debate. Early on, I was and still am a fan of the Aitchison & Brown book on the lognormal distribution but I have my doubts about the need of this particular geometry for compositional data. In itself the Aitchison geometry is a contribution, with a vector space, norm and inner product. When we transform the data to logarithms, then multiplication becomes addition, and powers become scalars, so that we can imagine such a vector space, yet, the amazing finding is that rebasing to 1 or 100% can be maintained. It is called “closure” when a vector is rebased to a constant sum. What, however, is the added value of using this geometry ?

It may well be that different fields of application still remain different on content, so that when they generate compositional data, then these data are only similar in form, while we should be careful in using the same techniques only because of that similar form. We must also distinguish:

  • Problems for compositional data that can be handled by both Sine / Cosine and the Aitchison geometry, but for which Sine and Cosine are simpler.
  • Problems for compositional data that can only be handled by the Aitchison geometry.

An example of the latter might be the paper by Javier Palarea-Albaladejo, Josep Antoni Martın-Fernandez and Jesus A. Soto (2012) in which they compare the compositions of milk of different mammals. I find this difficult to judge on content since I am no biologist. See the addendum below on the distance function.

In a fine overview by sheets, Pawlowsky-Glahn, Egozcue & Meziat 2007 present the following example, adapted from Aitchison. They compare two sets of soil samples, of which one sample is contaminated by water. If you want to spot the problem with this analysis yourself, take a try, and otherwise read on.

When the water content in the sample of A is dropped, then the test scores are rebased to the total of 100% for B again. E.g. for the 60% water in sample 1, this becomes:

{0.1, 0.2, 0.1} / (0.1 + 0.2 + 0.1) = {0.25, 0.5, 0.25}

PM. A more complex example with simulation data is by David Lovell.

Reproduction of this example

It is useful to first reproduce the example so that we can later adapt it.

In Wolfram Alpha, we can reproduce the outcome as follows.

For A, the input code is: mat1 = {{0.1, 0.2, 0.1, 0.6}, {0.2, 0.1, 0.2, 0.5}, {0.3 , .3, 0.1, 0.3}}; Correlation[mat1] // Chop // MatrixForm.

For B, the input code is: mat1 = {{0.1, 0.2, 0.1, 0.6}, {0.2, 0.1, 0.2, 0.5}, {0.3 , .3, 0.1, 0.3}}; droplast[x_List?VectorQ] := Module[{a}, a = Drop[x, -1]; a / (Plus @@ a)]; mat2 = droplast /@ mat1; Correlation[mat2] // Chop // MatrixForm.

The confusion about the correlation

In the former weblog entry, we had SDID[v, s] for the votes v and seats s. In this way of thinking, we would reason differently. We would compare (correlate) rows and not columns.

There is also a difference that correlation uses centered data while Sine and Cosine use original or non-centered data. Perhaps this contributed to Pearson’s view.

One possibility is that we compare sample 1 according to A with sample 1 according to B, as SDID[1A*, 1B]. Since the measures of A also contain water, we must drop the water content and create A*. The assumption is that and are independent measurements, and that we want to see whether they generate the same result. When the measurements are not affected by the content of water, then we would find zero inequality / disproportionality. However, Pawlowsky et al. do not state the problem as such.

The other possibility is that we would compare SDID[sample i, sample j].

Instead of using SDID for inequality / disproportionality, let us now use the cosine as a measure for similarity.

For A, the input code is: mat1 = {{0.1, 0.2, 0.1, 0.6}, {0.2, 0.1, 0.2, 0.5}, {0.3 , .3, 0.1, 0.3}}; cos[x__] := 1 – CosineDistance[x]; Outer[cos, mat1, mat1, 1] // Chop // MatrixForm.

Since the water content is not the same in all samples, above scores will be off. To see whether these similarities are sensitive to the contamination by the water content, we look at the samples according to B.

The input code for Wolfram Alpha is: mat1 = {{0.1, 0.2, 0.1, 0.6}, {0.2, 0.1, 0.2, 0.5}, {0.3 , .3, 0.1, 0.3}}; cos[x__] := 1 – CosineDistance[x]; droplast[x_List?VectorQ] := Module[{a}, a = Drop[x, -1]; a / (Plus @@ a)]; mat2 = droplast /@ mat1; Outer[cos, mat2, mat2, 1] // Chop // MatrixForm.

Since the water content differed so much per sample, and apparently is not considered to be relevant for the shares of the other components, the latter matrix of similarities is most relevant.

If we know that the samples are from the same soil, then this would give an indication of sample variability. Conversely, we might have information about the dispersion of samples, and perhaps we might determine whether the samples are from the same soil.

Obviously, one must have studied soil samples to say something on content. The above is only a mathematical exercise. This only highlights the non-transposed case (rows) versus the transposed case (columns).

Evaluation

Reading the Pearson 1897 paper shows that he indeed looks at the issue from the angle of the columns, and that he considers calibration of measurements by switching to relative data. He gives various examples, but let me show the case of skull measurement, that may still be a challenge:

Pearson presents two correlation coefficients for B / L with H / L. One based upon the standard definition (that allows for correlations between the levels), and one baptised “spurious”, based upon the assumption of independent distributions (and thus zero correlations for the levels). Subsequently he throws doubt on the standard correlation because of the high value of the spurious correlation.

One must be a biologist or even a skull-specialist to determine whether this is a useful approach. If the true model would use relative data with zero correlations, what is the value of the assumptions of zero or nonzero correlations for the absolute values ? What is useful depends upon the research question too. We can calculate all kinds of statistics, but what decision is intended ?

It is undoubtedly a contribution by Pearson that looking at phenomena in this manner can generate what he calls “spurious correlation”. Whatever the model, it is an insight that using the wrong model can create spurious correlation and a false sense of achievement. I would feel more comfortable though when Pearson had also mentioned the non-transposed case, which I would tend to regard as the proper model, i.e. comparing skulls rather than correlating categories on skulls. Yet he doesn’t mention it.

Apparently the Aitchison geometry provides a solution to Pearson’s approach, thus still looking at transposed (column) data. This causes the same discomfort.

Pro memori

The above uses soil and skulls, which are not my expertise. I am more comfortable with votes and seats, or budget shares in economics (e.g. in the Somermeyer model or the indirect addilog demand system, Barten, De Boer).

Conclusion

Pearson was not confused on what he defined as spurious correlation. He might have been confused about the proper way to deal with compositional data, namely looking at columns rather than rows. This however also depends upon the field of interest and the research question. Perhaps a historian can determine whether Pearson also looked at compositional data from rows rather than columns.

Addendum November 23 2017

For geological data, Watson & Philip (1989) already discussed the angular distance. Martin-Fernandez, Barcelo-Vidal, Pawlowsky-Glahn (2000), “Measures of differences for compositional data and hierarchical clustering methods“, discuss distance measures. They also mention the angle between two vectors, found via arccos[cos[v, s]], for votes v and seats s. It is mentioned in the 2nd row of their Table 1. The vectors can also normalised to the unit simplex as w = v / Sum[v] and z = s / Sum[s], though cos is insensitive with cos[w, z] = cos[v, s].

In sum, the angular distance, or the use of the sine as a distance measure and the cosine as a similarity measure, satisfy the Aitchison criteria of invariance to scale and permutation, but do not satisfy subcompositional dominance and invariance to translation (perturbation).

This discussion makes me wonder whether there are still key differences between compositional data in terms of concepts. The compositional form should not distract us from the content. For a Euclidean norm, a translation leaves a distance unaffected, as Norm[x – y] = Norm[(x +t) – (y + t)}. This property can by copied for logratio data. However, for votes and seats, it is not clear why a (per party different) percentage change vector should leave the distance unaffected (as happens in logratio distance).

An election only gives votes and seats. Thus there is no larger matrix of data. Comparison with other times and nations has limited meaning. Thus there may be no need for the full Aitchison geometry.

At this moment, I can only conclude that Sine (distance) and Cosine (similarity) are better for votes and seats than what political scientists have been using till now. It remains to be seen for votes and seats whether the logratio approach would be better than the angular distance and the use of Sine and Cosine.

This weblog entry copies the earlier entry that used an estimate.
Now we use the actual YouGov data, below.
Again we can thank YouGov and Anthony Wells for making these data available.
The conclusions do not change, since the estimate apparently was fairly good.
It concerns a very relevant poll, and it is useful to have the uncertainty of the estimate removed.

The earlier discussion on Proportional Representation versus District Representation has resulted in these two papers:

Brexit stands out as a disaster of the UK First Past The Post (FPTP) system and the illusion that one can use referenda to repair disproportionalities caused by FPTP. This information about the real cause of Brexit is missing in the otherwise high quality overview at the BBC.

The former weblog text gave an overview of the YouGov polling data of June 12-13 2017 on the Great Britain (UK minus Northern Ireland) preference orderings on Brexit. The uncertainty of the estimate is removed now, and we are left with the uncertainty because of having polling data. The next step is to use these orderings for the various voting philosophies. I will be using the website of Rob LeGrand since this makes for easy communication. See his description of the voting philosophies. Robert Loring has a website that referred to LeGrand, and Loring is critical about FPTP too. However, I will use the general framework of my book “Voting theory for democracy” (VTFD), because there are some general principles that many people tend to overlook.

Input format

See the former entry for the problem and the excel sheet with the polling data of the preferences and their weights. LeGrand’s website requires us to present the data in a particular format. It seems best to transform the percentages into per-millions, since that website seems to require integers and we want some accuracy even though polling data come with uncertainty. There are no preferences with zero weights. Thus we get 24 nonzero weighted options. We enter those and then click on the various schemes. See the YouGov factsheet for the definition of the Brexit options, but for short we have R = Remain, S = Soft / Single Market, T = Tariffs / Hard, N = No Deal / WTO. Observe that the Remain options are missing, though these are important too.

248485:R>S>T>N
38182:R>S>N>T
24242:R>T>S>N
19394:R>T>N>S
12727:R>N>S>T
10909:R>N>T>S
50303:S>R>T>N
9091:S>R>N>T
22424:S>T>R>N
66667:S>T>N>R
9091:S>N>R>T
36364:S>N>T>R
6667:T>R>S>N
3636:T>R>N>S
12121:T>S>R>N
46667:T>S>N>R
15758:T>N>R>S
135152:T>N>S>R
9697:N>R>S>T
9091:N>R>T>S
8485:N>S>R>T
37576:N>S>T>R
16970:N>T>R>S
150303:N>T>S>R

Philosophy 1. Pareto optimality

The basic situation in voting has a Status Quo. The issue on the table is that we consider alternatives to the Status Quo. Only those options are relevant that are Pareto Improving, i.e. that some advance while none lose. Commonly there are more Pareto options, whence there is a deadlock that Pareto itself cannot resolve, and then majority voting might be used to break the deadlock. Many people tend to forget that majority voting is mainly a deadlock breaking rule. For it would not be acceptable when a majority would plunder a minority. The Pareto condition thus gives the minority veto rights against being plundered.

(When voting for a new Parliament then it is generally considered no option to leave the seats empty, whence there would be no status quo. A situation without a status quo tends to be rather exceptional.)

In this case the status quo is that the UK is a member of the EU. The voters for R block a change. The options S, T and N do not compensate the R. Thus the outcome remains R.

This is the fundamental result. The philosophies in the following neglect the status quo and thus should not really be considered.

PM 1. Potentially though, the S, T and N options must be read such that the R will be compensated for their loss.

PM 2. Potentially though, Leavers might reason that the status quo concerns national sovereignty, that the EU breaches upon. The BBC documentary “Europe: ‘Them’ or ‘Us’” remarkably explains that it was Margaret Thatcher who helped abolish the UK veto rights and who accepted EU majority rule, and who ran this through UK Parliament without proper discussion. There seems to be good reason to return to unanimity rule in the EU, yet it is not necessarily a proper method to neglect the rights of R. (And it was Thatcher who encouraged the neoliberal economic policies that many UK voters complain about as if these would come from the EU.)

Philosophy 2. Plurality

On LeGrand’s site we get Plurality as the first step in the Hare method. gets 35% while the other options are divided with each less than 35%. Thus the outcome is R.

(The Brexit referendum question in 2016 was flawed in design e.g. since it hid the underlying disagreements, and collected all dissent into a single Leave, also sandwiching R between various options for Leave.)

Philosophy 3. Hare, or Instant Run-off, a form of Single Transferable Vote (STV)

When we continue with Hare, then R remains strong and it collects votes when S and N drop off (as it is curiously sandwiched between options for Leave). Eventually R gets 45.0% and T gets 55.0%. Observe that this poll was on June 12-13 2017, and that some 25% of the voters “respect” the 2016 referendum outcome that however was flawed in design. I haven’t found information about preference orderings at the time of the referendum.

Philosophy 4. Borda

Borda generates the collective ranking S > T > R > N. This is Case 9 in the original list, and fortunately this is single-peaked.

Philosophy 5. Condorcet (Copeland)

Using Copeland, we find that S is also the Condorcet winner, i.e. wins from each other option in pairwise contests. This means that S is also the Borda Fixed Point winner.

Conclusions

The major point of this discussion is that the status quo consists of the UK membership of the EU. Part of the status quo is that the UK may leave by invoking article 50. However, the internal process that caused the invoking of article 50 leaves much to be desired. Potentially many voters got the suggestion as if they might vote about membership afresh without the need to compensate those who benefit from Remain.

Jonathan Portes suggested in 2016 that the Brexit referendum question was flawed in design because there might be a hidden Condorcet cycle. The YouGov poll didn’t contain questions that allows to check this, also because much has happened in 2016-2017, including the misplaced “respect” by 25% of the voters for the outcome of a flawed referendum. A key point is that options for Remain are not included, even though they would be relevant. My impression is that the break-up of the UK would be a serious issue, even though, curiously, many Scots apparently rather prefer the certainty of the closeness to a larger economy of the UK rather than the uncertainties of continued membership of the EU when the UK is Leaving.

It would make sense for the EU to encourage a reconsideration within the UK about what people really want. The Large Hadron Collider is expensive, but comparatively it might be less expensive when the UK switches to PR, splits up its confused parties (see this discussion by Anthony Wells), and has a new vote for the House of Commons. The UK already has experience with PR namely for the EU Parliament, and it should not be too complex to use this approach also for the nation.

Such a change might make it also more acceptable for other EU member states if the UK would Breget. Nigel Farage much benefited from Proportional Representation (PR) in the EU Parliament, and it would be welcome if he would lobby for PR in the UK too.

Nevertheless, given the observable tendency in the UK to prefer a soft Brexit, the EU would likely be advised to agree with such an outcome, or face a future with a UK that rightly or wrongly feels quite maltreated. As confused as the British have been on Brexit, they might also be sensitive to a “stab-in-the-back myth”.

The earlier discussion on Proportional Representation versus District Representation has resulted in this paper: Two conditions for the application of Lorenz curve and Gini coefficient to voting and allocated seats, MPRA 80297.

Brexit stands out as a disaster of the UK First Past The Post (FPTP) system and the illusion that one can use referenda to repair disproportionalities caused by FPTP. This information is missing in the otherwise high quality overview at the BBC.

In the earlier Puzzle on the YouGov poll I estimated Brexit preference orderings from a summary statistic published by YouGov. The next step is to use these orderings for the various voting philosophies. I will be using the website of Rob LeGrand since this makes for easy communication. See his description of the voting philosophies. Robert Loring has a website that referred to LeGrand, and Loring is critical about FPTP too. However, I will use the general framework of my book “Voting theory for democracy” (VTFD), because there are some general principles that many people tend to overlook.

Input format

See the Puzzle weblog text for the problem and the excel sheet with the estimate of the preferences and their weights. LeGrand’s website now requires us to present the data in a particular format. It seems best to transform the percentages into per-millions, since that website seems to require integers and we want some accuracy even though the estimate is tentative. We can also drop the preference rankings with zero weights. Thus we get 14 nonzero weighted options. We enter those and then click on the various schemes. See the YouGov factsheet for the definition of the Brexit options, but for short we have R = Remain, S = Soft / Single Market, T = Tariffs, H = Hard / WTO. Observe that the Remain options are missing, though these are important too.

261841:R>S>T>H
53499:R>S>H>T
38386:R>T>H>S
60161:S>R>T>H
30087:S>R>H>T
44443:S>T>R>H
34960:S>T>H>R
22354:S>H>T>R
24777:T>S>H>R
15640:T>H>R>S
181873:T>H>S>R
49951:H>S>T>R
20475:H>T>R>S
161553:H>T>S>R

Philosophy 1. Pareto optimality

The basic situation in voting has a Status Quo. The issue on the table is that we consider alternatives to the Status Quo. Only those options are relevant that are Pareto Improving, i.e. that some advance while none lose. Commonly there are more Pareto options, whence there is a deadlock that Pareto itself cannot resolve, and then majority voting might be used to break the deadlock. Many people tend to forget that majority voting is mainly a deadlock breaking rule. For it would not be acceptable when a majority would plunder a minority. The Pareto condition thus gives the minority veto rights against being plundered. (When voting for a new Parliament then it is generally considered no option to leave the seats empty, whence there would be no status quo. A situation without a status quo tends to be rather exceptional.)

In this case the status quo is that the UK is a member of the EU. The voters for R block a change. The options S, T and H do not compensate the R. Thus the outcome remains R.

This is the fundamental result. The philosophies in the following neglect the status quo and thus should not really be considered.

PM 1. Potentially though, the S, T and H options must be read such that the R will be compensated for their loss.

PM 2. Potentially though, Leavers might reason that the status quo concerns national sovereignty, that the EU breaches upon. The BBC documentary “Europe: ‘Them’ or ‘Us’” remarkably explains that it was Margaret Thatcher who helped abolish the UK veto rights and who accepted EU majority rule, and who ran this through UK Parliament without proper discussion. There seems to be good reason to return to unanimity rule in the EU, yet it is not necessarily a proper method to neglect the rights of R. (And it was Thatcher who encouraged the neoliberal economic policies that many UK voters complain about as if these would come from the EU.)

Philosophy 2. Plurality

On LeGrand’s site we get Plurality as the first step in the Hare method. gets 35% while the other options are divided with each less than 35%. Thus the outcome is R.

(The Brexit referendum question in 2016 was flawed in design e.g. since it hid the underlying disagreements, and collected all dissent into a single Leave, also sandwiching R between various options for Leave.)

Philosophy 3. Hare, or Instant Run-off, a form of Single Transferable Vote (STV)

When we continue with Hare, then R remains strong and it collects votes when S and H drop off (as it is curiously sandwiched between options for Leave). Eventually R gets 44.4% and T gets 55.6%. Observe that this poll was on June 12-13 2017, and that some 25% of the voters “respect” the 2016 referendum outcome that however was flawed in design. I haven’t found information about preference orderings at the time of the referendum.

Philosophy 4. Borda

Borda generates the collective ranking S > T > R > H. This is Case 9 in the original list (including zero weights), and fortunately this is single-peaked.

Philosophy 5. Condorcet (Copeland)

Using Copeland, we find that S is also the Condorcet winner, i.e. wins from each other option in pairwise contests. This means that S is also the Borda Fixed Point winner.

Conclusions

The major point of this discussion is that the status quo consists of the UK membership of the EU. Part of the status quo is that the UK may leave by invoking article 50. However, the internal process that caused the invoking of article 50 leaves much to be desired. Potentially many voters got the suggestion as if they might vote about membership afresh without the need to compensate those who benefit from Remain.

Jonathan Portes suggested in 2016 that the Brexit referendum question was flawed in design because there might be a hidden Condorcet cycle. The YouGov poll didn’t contain questions that allowed to check this, also because much has happened in 2016-2017, including the misplaced “respect” for the outcome of a flawed referendum. A key point is that options for Remain are not included, even though they would be relevant. My impression is that the break-up of the UK would be a serious issue, even though, curiously, many Scots apparently rather prefer the certainty of the closeness to a larger economy of the UK rather than the uncertainties of continued membership of the EU when the UK is Leaving.

It would make sense for the EU to encourage a reconsideration within the UK about what people really want. The Large Hadron Collider is expensive, but comparatively it might be less expensive when the UK switches to PR, splits up its confused parties, and has a new vote for the House of Commons. The UK already has experience with PR namely for the EU Parliament, and it should not be too complex to use this approach also for the nation. Such a change might make it also more acceptable for other EU member states if the UK would Breget. Nigel Farage much benefited from Proportional Representation (PR) in the EU Parliament, and it would be welcome if he would lobby for PR in the UK too.

Nevertheless, given the observable tendency in the UK to prefer a soft Brexit, the EU would likely be advised to agree with such an outcome, or face a future with a UK that rightly or wrongly feels quite maltreated. As confused as the British have been on Brexit, they might also be sensitive to a “stab-in-the-back myth”.

If you don’t have proportional representation (PR) then some voters get representatives they did not vote for. Thus it isn’t very democratic not to have PR.

The last weblog criticised the UK Electoral Reform Society (ERS) for erroneously claiming that Single Transferable Vote (STV) was PR.

ERS namely adopts districts, which causes STV to lose the limited PR properties that it has.

A persons affiliated with ERS answered to this criticism:

“We are well aware of the tension between the desire for (overall) proportionality and the desire for guaranteed local representation.  This tension is apparent among British electors when opinion polls have asked relevant questions about the outcomes of voting systems.  British electors want both overall proportionality (of parties) AND the local representation provided by exclusively single-member districts.  That is just not possible, so we aim for a compromise between local representation and overall proportionality through appropriately-sized multi-member districts. (…).” (Personal communication)

This is an unsatisfactory answer since there simply is no such “compromise”. When one must choose between a square and a circle then the answer is not some other graphic with some measure of deviation. If there is no PR then there is no PR, and then ERS should not claim that they have PR. To express their “compromise”, ERS speaks about “STV-PR” but this is like speaking about square-circles, and comes with the grating sound from nails across a blackboard.

If n is the number of voters, s the number of seats, then q = n / s is the threshold or quota, of voters per member. A candidate can be elected when he or she meets the quota. When the district size is 2q, then the district representative must get 50%+1 of the vote to attain the quota. At best s / 2 seats can be filled in this manner, since s / 2 * 2q = n again. All unfilled seats can be allocated using overall PR. This shows that districts are not a key design feature, while PR is. (These formulas can be adjusted for turnout, when district size is defined in terms of the electorate and not actual voters. See here.)

By focusing on districts, ERS loses track of the key design feature, and it lets its logic be occluded by a less relevant issue.

Wikipedia follows ERS

Apparently the editors at wikipedia follow ERS rather uncritically. The wikipedia statement in red is what ERS claims falsely and what is adopted by wikipedia too. The statement in green is true. Since the statement in green is true, the statement in red can only be true by chance.

Proportional representation (PR) characterizes electoral systems by which divisions in an electorate are reflected proportionately in the elected body. If n% of the electorate support a particular political party, then roughly n% of seats will be won by that party. The essence of such systems is that all votes contribute to the result: not just a plurality, or a bare majority, of them. Proportional representation requires the use of multiple-member voting districts (also called super-districts); it is not possible using single-member districts alone.[1][2][3] In fact, the most proportional representation is achieved when just one super-district is used.

The two most widely used families of PR electoral systems are party list PR and single transferable vote (STV).[4][5] Mixed member proportional representation (MMP), also known as the Additional Member System, is a hybrid Mixed Electoral System that uses party list PR as its proportional component. MMP has the potential to be proportional or semi-proportional depending on a number of factors such as the ratio of first past the post (FPTP) seats to PR seats, the existence or nonexistence of compensatory seats to make up for overhang seats, and election thresholds.[6][7][8][9]   (Source: Wikipedia on PR)

ERS thus is confusing the world including wikipedia. My advice for the editors of wikipedia (and the ERS) is:

  • Maintain conceptual integrity.
  • Restrict PR to the notion that p% of the votes translates into p% of the seats.
  • For PR the first preferences are relevant and not what is done with the subsequent preferences. Thus do not label STV as a PR-system but as “potentially PR”, or as STV-PPR.
  • For PR it suffices when the electorate selects parties. A single candidate is a party with a single candidate.
  • The professionals in parliament can use more complex systems like STV. The use of STV (there) must be compared to other systems, like Borda Fixed Point.
  • Get rid of the hangup on district representation.

Unfortunately, the person affiliated with ERS writes to me, with an unrelenting hangup about districts, and neglecting that PR should hold nation-wide:

“Neither the ERS nor I would be prepared to label STV-PT as “potentially PR” or anything similar.  I have seen some academics describe STV-PR as “a semi-proportional” system.  That is just nonsense.  For the same district magnitude, STV-PR and party-list PR both deliver the same degree of proportionality.  The fact that some electorates are prepared to accept electoral districts that cover the whole country for party-PR but don’t like the idea of “large” electoral districts for STV-PR is completely irrelevant.  It is the district magnitude that is the determining factor, not the voting system.”  (Personal communication)

Scotland is an example

Scotland has four electoral systems, and I copy from Wikipedia:

Does this mean that Scotland comprehends democracy or that they don’t ?

The Party List System as used for the EU Parliament generates proportional representation (PR), and this would be the criterion for representative democracy.

(Obviously, for the election of a local council, the norm for PR are the local votes, and not nationwide PR. Once the issue here is reduced to apportionment, then STV is one of the options and a choice depends upon one’s criteria.)

Let us look at the Scottish implementation Additional Member System (AMS) a.k.a. Supplementary Member System a.k.a. Mixed Member System (MMS). I would prefer the latter term, since there is nothing “additional” about an elected MP. Sometimes the term “Mixed Member Proportional” (MMP) is used but this is only warranted when there really is overall PR.

The current Scottish system

Scotland has 73 constituencies, in which the candidate is selected by FPTP. There are 8 regions with 7 seats per region, to a total of 56 regional seats. These “additional seats” are used to make the outcome more proportional. Brief explanations of the current Scottish system are by the Parliament itself and The Scotsman. The Scottish Parliament elections of May 5 2016 have these full data. The turnout was 55.6%.

Scotland like the UK has a hangup on the distinction between the local candidate and the party. It is claimed: “In the second vote the voter votes for a party rather than a candidate.” Indeed, when the first vote has a FPTP selection, then voters may be forced to vote strategically for a candidate of reduced preference, in trying to prevent that a worst candidate is selected. Thus the explanation about local representation may be a misrepresentation about what might really motivate voters.

When we compare the votes for the constituencies (districts) and the regions, then we don’t see much of a difference, except for the Greens and Others. (This are totals though, and there might be differences over districts.)

District

Region

Party

Votes

Votes

Con

501,844

524,222

Green

13,172

150,426

Lab

514,261

435,919

LD

178,238

119,284

SNP

1,059,898

953,587

Others

11,741

102,314

Total

2,279,154

2,285,752

We take the summed region vote as determining what the proportions for the parties should be. The additional 56 seats and their restriction to regions are not enough for correction of the error in the local vote. The SNP got 7% more seats than warranted under PR.

District

Region

 All
Party

Seats

Seats

Seats

%Seats

%Votes

%S-%V

Con

7

24

31

24.0

22.9

1.1

Green

0

6

6

4.7

6.6

-1.9

Lab

3

21

24

18.6

19.1

-0.5

LD

4

1

5

3.9

5.2

-1.3

SNP

59

4

63

48.8

41.7

7.1

Others

0

0

0

0.0

4.5

-4.5

Total

73

56

129

100.0

100.0

An alternative for Scotland

Let us consider a rough alternative for Scotland:

  • A local winner must get at least 50% of the vote of a district (constituency).
  • All 129 seats are allocated in proportion to the summed region vote.

The data file allows us to determine which candidates are elected now. This generates a quite different result. In the local vote, only 29 candidates manage to get at least 50% of their district (constituency). 95 candidates are selected via the Party List, which puts the ERS argument for locality into perspective. In this rough alternative, there are 5 seats that cannot be allocated due to rounding errors. But having 4% empty seats is not unfair given that 4.5% of the votes are wasted on the small parties.

District

Region

 All
Party

Seats

Seats

Seats

%Seats

%Votes

%S-%V

Con

1

29

30

23.3

22.9

0.3

Green

0

8

8

6.2

6.6

-0.4

Lab

0

25

25

19.4

19.1

0.3

LD

2

5

7

5.4

5.2

0.2

SNP

26

28

54

41.9

41.7

0.1

Others  0  0  0

0.0

4.5

-4.5

Total

29

95

124

96.1

100.0

A general observation

The quota is q = 2,285,752 / 129 = 17720. Above criterion of 50% of the local vote might be too lax. With 73 districts, the number or electors per district might be too small. If the number of districts is 129 / 2 ≈ 65, then the average district has size 2q, and the criterion of at least 50% of the votes would fit the overall condition of winning a seat via satisfying q.

A google showed this page by Andrew Ducker who also wondered about PR in Scotland. He mentions: (1) The region votes must be summed for nationwide PR indeed. (2) A 50%:50% distinction between local and national seats would be helpful indeed. In reply to this: why still allow FPTP when it may cause that a minority winner would become the “representative” ? It is better to require at least q and/or at least 50% of the district.

The UK Electoral Reform Society (ERS)

The UK ERS falsely claims that STV applied to districts would be PR while it is not. The ERS also criticises the Scottish system, but perhaps for the wrong reasons.

In 2011, the current Scottish system was already in place, and the ERS advised a change. See the Guardian or the BCC:

One of the authors of the report, Prof John Curtice of Strathclyde University, said: “The widespread expectation that the Scottish Parliament would be a multi-party parliament, in which no party would ever have an overall majority, has been dashed. “In truth, although the electoral system bequeathed to the Scottish Parliament by Labour was far more proportional than first-past-the-post, it was never one that was best fitted to the realisation of that original expectation. “It still favours larger parties over smaller ones, who, indeed, are actually being discouraged from standing in the constituency contests.” (BBC 2011-11-15)

A similar critique is given w.r.t. the 2016 outcome. Some changes like an “open party” list (i.e. the ability to vote for individual party candidates on the list) and the allocation of Sainte-Laguë may indeed be mentioned.

But this is small beer compared to the major critique on the Scottish system, that it still allows for the hangup on district representation.

While the ERS should warn voters and legislators about this hangup, the ERS suffers from this hangup itself too, and propounds STV for districts, which destroys PR.

The person affiliated with ERS writes to me:

I am not in favour of electing MPs (or other representatives) in two different ways. In Scotland we have experience of the Scottish Parliament where MSPs are elected by the Additional Member System (AMS = a regionalised version of MMP). Some of the worst problems of electing MSPs in two different ways (Constituency and Region) have abated over the years, but the tension remains and surfaces from time to time. It would have been much better if all the MSPs had been elected by STV-PR, but AMS was a political compromise as one the major parties (Labour) would just not accept STV-PR at any price. (Personal communication)

Again, this person at ERS suggests that STV would be PR, even calling it STV-PR, while the very application of STV to districts destroys the PR.

Missing Scottish voters

ERS Scotland director Willie Sullivan wrote a book about the structurally low turnout for Scottish elections: “The Missing Scotland: Why over a million Scots choose not to vote and what it means for our democracy” (publisher).

In an article, Sullivan summarises:

“If the working people wanted democracy, why do so many now not vote? Surely these are the people that should be most eager to flex their democratic muscle? In research for my book, Missing Scotland, I tried to find out why more than a million Scots choose not to vote. What I found is worrying. Most important of all, people don’t think voting will make anything better. They have tried voting, and they have tried not voting, and there is no difference. They think politicians are all the same, don’t understand their lives and they make promises they never keep. This is not a question of not caring. The people I spoke to care a lot about their families and communities. They are worried about losing their homes or their jobs. They even like the idea of democracy, they just don’t think we have it. Not voting is often a deliberate act.”

I haven’t read this book, but only find it relevant enough to mention its existence. My guess is that Sullivan hasn’t mentioned two elements:

  1. When the Scottish electoral system was changed, they didn’t adopt the PR system like in Holland, but kept their hangup on districts, and ERS itself was an agent in this. Potentially Sullivan might feel ashamed that he doesn’t quite comprehend what democracy is.
  2. Scotland isn’t aware of the failure of the Trias Politica model and the need for an Economic Supreme Court.

Conclusion

My finding is that major political distortions in the UK, France, USA and India arise because of lack of PR.  A lobby for STV for districts doesn’t resolve this, and it is falsely claimed to be PR. Thus I would tend to advice electoral reform in this order of priority:

  • first PR, like the system in Holland or the EU Parliament (Open Party List)
  • if this is up and running, secondly allow for an element of locality for half of the seats (s / 2, district size 2q, and the district representative is elected with at least 50% of the district vote, potentially corrected for turnout)
  • if this is up and running, improve the system by allowing voters freedom on how they vote
  • compare STV and Borda Fixed Point and other methods for the selection of the local representative.

The UK Electoral Reform Society hinders clarity on electoral reform since they show a hangup on districts. They better focus on establishing Proportional Representation (PR), while regarding the issue of districts as of secondary importance.

Given overall PR, one might even let voters determine on the ballot how to deal with the district representation, for the s / 2 seats available for district representation.

  1. Some voters might vote for a party, and be done with that. Seats are allocated to the party in proportion to the total number of votes. (Closed Party List) Some voters might wish to select a party but also a particular person in that party, so that the party order takes over if the person would not be elected. (Open Party List) These approaches can be combined (as in Holland) when the Closed List voters vote for the party leader.
  2. Some voters might indicate where their vote would go, if their party of choice isn’t elected. (Remember that a single candidate is a party with a single candidate.)
  3. Other voters might wish to vote for particular candidates across parties, and then might want to indicate how votes would have to be transferred if the candidate doesn’t get elected. (Otherwise it is apportioned automatically.) There is still the comparison between STV and e.g. (repeated) application of the Borda Fixed Point method. STV runs the risk of eliminating a compromis candidate, who receives few votes in the initial stage, but who can collect support because of secondary preferences. This might not be relevant for the party proportion but be quite relevant for voters and the candidates themselves. This would not be an issue of PR but of Quality Representation (QR).

PS. Dan Hodges (Telegraph June 1 2015) has a very entertaining article “No, Britain does not want proportional representation“. The weak spot in his argument is that the 2011 referendum on AV was misrepresented as a referendum on PR while it actually was a referendum on AV. The strong point is that ERS cannot be convincing if its arguments are confused. There still is a case for sound arguments and good education.

PPS. The subtle relation between proportional representation (PR) and district representation (DR) shows also in the existence of a Senate or House of Lords, in which districts / States might be represented by 2 senators per State like in the USA. For a Senate the DR might be acceptable since the Senate has the role of guardian for the nation itself. The House would be sensitive to the preferences of the electorate, and in that case PR would be logical.

There is a bizarre incomprehension of democracy.

I was inclined to say in the English speaking world – UK, USA, India – but there is also France, with the upcoming elections for the French National Assembly 2017. Also France has a system of District Representation (DR) rather than Proportional Representation (PR). Examples of PR are Holland, Germany and to some extent the European Parliament.

Democratic theory favours PR above DR. It is a historical mistake that countries have DR.

The UK Electoral Reform Society (wikipedia) protests about the UK general elections 2017.

Electoral Reform Society, website June 2017

Indeed, there is this difference (wikipedia) between the UK system of DR with the better system of PR. While the popular vote gives a majority to Lab + SNP + LibDem, the majority in seats goes to Con + DUP.

Seats Votes
Con

317

48.8%

13,632,914

42.3%

DUP

10

1.5%

292,316

0.9%

Lab

262

40.3%

12,874,985

40.0%

SNP

35

5.4%

977,569

3.0%

LibDem

12

1.8%

2,371,772

7.4%

Other

14

2.2%

2,047,362

6.4%

Total

650

100.0%

32,196,918

100.0%

There is also the political dynamics of proportionality. Once a proportional system is in place then new parties will have a larger chance to get elected, and then they also have more scope to grow and to challenge the existing parties. We may observe that the UK 2017 outcome may be more proportional than in the past, but this still neglects the dynamics and the build-up of frustration amongst minorities that aren’t represented well.

The UK already has experience with the system of PR, namely for the European elections. While UKIP didn’t do so well within the UK system, Nigel Farage gained the limelight via the elections for the European Parliament.

The UK Electoral Reform Society fails too

Part of the bizarre situation is that the UK Electoral Reform Society (ERS) fails too. They favour the system of “Single Transferable Vote” (STV) and they call this a system of PR while it isn’t PR. It is rather bizarre that they do not comprehend this. The UK had a so-called “referendum on PR” in 2011, but this was actually on the system of “Instant Run-Off” (a.k.a. “Alternative Vote”).

Apparently, the UK has a hangup on DR. They think that districts allow voters to connect directly with the local politicians, and that this reduces the influence of the party bosses. This is a dubious argument. If a representative and party member is out of line with the party then the party might still sack him or her nevertheless. In PR it is easier to start a new political party and be elected (when the issue likely isn’t just local).

Thus the statements by the ERS about district representation derive from historical bias and not from clear theory and practical experience with PR.

Let me give an example how STV favoured by ERS is not PR at all, even though they claim that it would be PR.

Example that STV is not PR

Consider two districts with 30000 voters each. Each district has 2 seats, so that the Droop Quota is 30000 / (2 + 1) + 1 = 10001. Parties contending for these seats are Con, Lab, en LDP. In District 1, the Con are popular, and they present there a list with two candidates. In District 2 the Lab are popular, and they present there a list with two candidates. We consider a rather symmetrical situation as in the following table, also with the STV results.

In STV, voters vote only once, but they can assign a rank order of the candidates.

In District 1, 7503 Con1 voters give Con1 as their first choice and Con2 as their second choice. Also 7501 Con2 voters give Con2 as their first choice and Con1 as their second choice.

  1. In the first round, no candidate meets the quota. LDP1 has the minimal number of votes, 7494, and is eliminated.
  2. In the second round, Con2 has the minimal number of votes, 7501, and is eliminated.
  3. In the third round, the Con2 votes are allocated to Con1, and Con1 meets the quota and is elected.
  4. In the fourth round, Lab3 is the remaining candidate and thus is elected, though the 7502 votes are below the quota.

The situation in District 2 is analogous.

District 1 Votes District 2  Votes
Con1

7503

Elected Con3

7502

Elected

Con2

7501

Lab1

7503

Elected

Lab3

7502

Elected Lab2

7501

LDP1

7494

LDP2

7494

30000

30000

Let us now join the two districts, and look how STV works for the national vote.

There are 60000 voters and 4 seats, so the Droop Quota is 60000 / (4 + 1) + 1 = 12001.

The Con voters put the Con candidates of their district in the first places, and then the Con candidates of the other district. For example, the 7501 Con2 voters have the rank order {Con2, Con1, Con3}.

Because of the symmetrical structure of this example, there are some ties. Rather than using a coin, we use the alphabetical order.

  1. In the first round, no candidate meets the quota, and LDP1 is eliminated.
  2. In the second round, the LDP1 votes go to LDP2, and it is elected.
  3. In the third round, alphabetically Con2 has the minimal number of votes, and is eliminated.
  4. In the fourth round, Con2’s votes go to Con1, and it is elected.
  5. In the fifth round, alphabetically Lab2 has the minimal number of votes, and is eliminated.
  6. In the sixth round, Lab2 votes go to Lab1, and it is elected.
  7. In the seventh round, alphabetically Con3 has the minimal number of votes, 10506, and is eliminated.
  8. In the eighth round, Lab3 remains, and is elected, though with only 10506 votes.

Thus now LDP2, Con1, Lab1 and Lab3 are elected.

Upshot:

  • As the UK Electoral Reform Society (ERS) states that STV would generate proportional results, both {Con1, Con3, Lab1, Lab3} and {Con1, Lab1, Lab3, LDP2} would be proportional results, which however are quite different results, which destroys the meaning of proportionality.
  • With a hangup on DR, there will be little scope for fair representation of the minority LDP.
  • PR would require party representation with {37.5%, 37.5%, 25%} of the seats. Admittedly, this is difficult to achieve with 4 seats, but if the situation persists then one might change the number of seats. This uses the PR criterion rather than the STV criterion.

In these considerations, the notion of PR dominates DR.

Confusing information from Holland

Holland is a small country with some 10 million voters. One might think that its proportional system might not easily scale up to a large unit like the EU. If the EU Parliament would be fully proportional, then Germany might have too large a weight. However, there are also divisions along party lines, and proportionality still would be a fair choice.

Holland had general elections on March 15 2017, and political parties are still negotiating about a coalition government. This is bad advertisement for the system.

There is a confusion in Holland about desiring the minimal majority coalition, that bargains for an agreement that would apply for the next 4 years. It would be more rational to look for a larger majority, and rather set for a flexible agenda, so that issues can be dealt with in varying manner. See this paper of mine. It may also be better to have elections every year, so that Parliament becomes more sensitive to the popular vote. (If you would replace only 25% of the seats, then the electoral quota becomes 4 times larger, and this might be too high again.)

Most bizarre is that the Dutch party D66 wants to change the Dutch PR system into more use of districts … see here. Thus the good news about democracy in Holland is killed again by the campaigners on the D66 hobby horse.

Conclusion

We observe that the UK ERS protests against the failure of the UK electoral system, but we can also conclude that ERS doesn’t comprehend democracy.

We find similar confusions all over the world. The best advice is to change your national electoral system to the PR system like in Holland (or to some extent the EU parliament), yet many campaigners ride their hobby horses of wildly confusing varieties, and calling it “democracy” while it actually isn’t.

PM. A paper of mine on comparing PR and DR is here.

This discussion continues with the exchange of Flassbeck & Lapavitsas versus Storm.

My own position is as follows. Holland creates its own unemployment since 1970 by a wrong policy on taxes and premiums. Holland has the entrenched government policy of “solving” this by wage moderation and exporting its own unemployment to other nations. I present my alternative analysis since 1990. Economists at TU Delft have been arguing against this policy of wage moderation for decades too. They overlook the cause in taxes and premiums, and focus on technology. Schumpeterian innovation requires higher wages to get rid of obsolete technology. Now that Germany has had wage moderation too (because of the fall of the Berlin Wall and the Mark = Mark policy) the discussion on wage moderation moves to center stage for the survival of the Euro. Servaas Storm at TU Delft enters the European discussion again with arguments about technology, and again neglecting taxes and premiums, and neglecting the censorship of science by the Dutch government (what this weblog is about). Storm has an innovation in his analysis by including banking, and how international credits drive international trade, yet, he seems to neglect the phenomenon that trade surpluses generate funds that look for opportunities, often by providing credits that generate more surpluses. Thus Dutch and German wage moderation would be causally more important than bank credits.

PM. See also former IMF director Johannes Witteveen’s lecture on the Dutch export surplus and a need for an investment policy. There is also my discussion in 2009 with a chart of the Dutch export surplus in 1971-2010 (forecast). This is already 8 years ago.

Let me restate some basic economics for some readers who lack this.

Basic macro-economics

  • Let real national output (GDP) be y and the price level be p.
  • Let labour input be x and the wage be w.
  • Then labour productivity is λ = y / x and the Labour Income Quote is LIQ = w x / (p y).
  • Let there be a Cobb-Douglas production function: y = β x^α, with β containing capital and technology.
  • When producers maximize their profits π = p y w x subject to labour input x, then we can derive:
  • The first order condition: dπ / dx = 0 gives p β α x^(α – 1) – w = 0.
  • Or the wage can be set at w = p β α x^(α – 1) = p α y / x = p α λ, since the national labour supply is given as x.
  • This w = p α λ is the rule mentioned by Flassbeck & Lapavitsas: let wages grow with labour productivity and the agreed target of inflation of 2%.
  • From w = p α λ we can derive α = w x / (p y) or α = LIQ.
  • Unit labour costs are ULC = w x / y = w / λ = p LIQ. Thus alternatively w = p LIQ λ or w = ULC λ.

The assumption of the Cobb-Douglas function seems somewhat specific, but given the relatively small changes that we are considering the approximation is often so good that we almost seem to have a definition. The LIQ has the character of a structural parameter α, at least for annual changes.

If prices p and wages w and labour input x remain the same from one year to the other, and productivity rises by rate g, so that  =  (1 + g) y[-1], then α = LIQ = w x / (p y) = w x / (p (1 + g) y[-1]) = LIQ[-1] / (1 + g) = α[-1] / (1 + g). For example, if α[-1] = 80% and g = 2% then α ≈ 78%. In this case α would be stable if wages would rise by 2% too.

The w = p α λ condition is not in the EMU rules. The Eurozone countries apparently are less aware of the notion of “national bargaining” (as in the Dutch Polder model) and have been hesitant to include national wage agreements in the EMU and Stability & Growth Pact (SGP) and its updates. (Check for the word “wage” on this wiki page.)

Another possible rule might be a tax of 5% on the three year cumulative trade surplus (which may be seen as 15% for a single year), to be invested in productive capacity in the deficit countries via national investment banks. Such a tax would not be on export items (like a tariff) but levied on the Eurozone member governments of surplus countries. (At this applet, set the color bar to a score of 0, and slide over the years.)

It is unavoidable to think about such rules. Holland has been moderating its wages long before Germany did. The policy put pressure on the exchange rate of the guilder, but this was resolved by joining the Euro. Holland still is a small country and the impact wasn’t much felt. Now, Europe must explain to Germany that a raise of German wages is required, whatever they fear about inflation. It should help Germany to grow aware that my analysis (see DRGTPE) allows full employment at stable prices, not only by exports but also by stimulating the domestic market.

Shifting the blame

Both North and South Europe deviated from w = p α λ. Some Northerners blame the South, and some accept some blame themselves.

  • Sinn and Schäuble argue that Southern Europe should moderate their wages like Germany.
  • Bofinger and Flassbeck & Lapavitsas argue that Germany (and Holland) should raise their wages.

As Storm states:

“Their main point is that there would not have been large unsustainable current account imbalances within the Eurozone, and consequently no sovereign debt crisis in the deficit countries, if all member states had kept their nominal wage growth equal to labor productivity growth plus 2% (the inflation target). Professor Wren-Lewis (2016) has been making the same point. In this account, this delicate equilibrium has been deliberately upset by nominal wage moderation in mercantilist Germany, with a growing German trade surplus just being the flipside of the growing trade deficit in Southern Europe. It is rather ironic, in my opinion, that a similar logic is used by mainstream observers such as Sinn (2014) or even Mr. Schäuble himself, with this difference: Sinn and Schäuble argue that the current account imbalances were caused by a failure of the crisis countries to follow Germany’s successful example in cutting down their unit labor costs.”

Towards a collapse of the Euro

Sinn and Schäuble want to control inflation and they lack instruments to make sure that Southern Europe adheres to the EMU rules. Thus Sinn and Schäuble take the hard line that it is up to Southern Europe to choose themselves:

  • either unemployment because of high wages
  • or internal devaluation, and subsequent unemployment because of deficient internal demand.

Hence we can understand Flassbeck & Lapavitsas:

“Germans ought to know better than all others about the difficulties caused by wage divergences in a currency union. The deviation of East German wages as measured in international currency, following the German Monetary Union of 1990, destroyed East German industry and forced a transfer union. Unfortunately, for the EU and the EMU the option of a transfer union is simply not available. As long as Germany persists with its policy of wage moderation, the only future for the EMU is collapse.”

Check how I criticised Angela Merkel for her deceit at the German elections in 2013. Given German policies on wage moderation, standard economic theory allows her the choice between a transfer union or a breakup, but she kept silent about this. Of course there is my amendment to the theory of the optimal currency area, see MPRA or RWER, but as long as German policy makers do not indicate that they understand his amendment, we must conclude that they disinform their electorate.

How does Storm handle this ?

How does Storm handle this reference to basic economics ? He misstates the argument, and then rejects it.

“(…) that Eurozone imbalances were driven by (exogenous) losses or gains in unit labor cost competitiveness (…) is a myth (…)”

Storm’s problem is on causality: “what drives what”. Yet this is not quite what this discussion is about. What Storm calls a myth are basically accounting rules.

  • Use GDP = Y = p y = C + I + G + X – M, with consumption C, investments I, government G, exports X and imports M.
  • The current account CA = X – M is also the increase in foreign assets FA = X – M (NY FED).
  • National income, employment and wage translate into LIQ and λ. This is mere accounting.
  • Compare two situations for the same country with only a difference in M. In the first situation there is Y1 with a surplus on the current account, or M < X. In the second situation there is Y2 with a deficit or M > X. Thus Y1 > Y2. Assume the same output price p and working force x so that y1 / x > y2 / x, or λ1 > λ2. The productivity with a surplus is higher than with a deficit. For example, in the second case the country worked as hard as usual, but also imported a car by borrowing from abroad. Mere accounting causes that observed productivity drops. Similarly we have w x / y1 < w x / y2 or ULC1 < ULC2, or that the deficit situation has higher unit labour costs.

Economics is about causality and not about accounting, but it is important to be aware of accounting effects. Regressions with statistical data that contain these accounting effects must be judged carefully.

In above example of importing a car, causality seems to run from first importing to secondly a statistical observation on productivity. This is Storm’s view. But this is not the only causal possibility. Sinn and Schäuble might argue that higher productivity might have been feasible if the car hadn’t been imported but e.g. produced in the country itself with a creditor in the country itself. Thus there seems to be more complexity than Storm allows for (though he already makes a complex case). And Sinn and Schäuble might state more clearly that they also plea for the demise of the German car industry.

Storm’s five arguments

Storm has five arguments that we may indicate shortly. Apparently he repeats himself at points, but this is okay since we look at the arguments and not their number.

  1. Banks in Northern Europe lent to customers in Southern Europe, assuming that loans in Euro were safe anywhere. (Comment: True. However, if there hadn’t been surpluses on the Northern current accounts, then these banks would have had less funds. We are not speaking about a single year, but a prolonged period of surplus funds looking for “investment” opportunities.)
  2. German firms, producing high-tech, high value-added, high-priced and mostly very complex manufacturing goods, do not directly compete with Spanish, Portuguese, Greek or even most Italian firms, which are specializing in lower-tech, lower value-added, low-price and less complex goods (Simonazzi et al. 2013).” (Comment: This is not relevant, since differences in quality are corrected by differences in wages, whence we compare w1 / λ1 and w2 / λ2.)
  3. Four empirical “facts”. (a) Elasticities. (b) In Spain imports grew while exports were unaffected. (c) World income explains exports, and national income explains imports. (Costs might have a one-time effect but then are stable.) (d) There were first the imbalances and only later the worse ULCs. (Comment: Basically agreed on (a)-(c). However, this (d) is the same as (1). We are not speaking about a single year, but about a prolonged period of imbalance and funds looking for profit.)
  4. A more theoretical discussion of (3c), with the example of (2). “These asymmetric growth patterns are the direct consequence of structural differences in productive specialization (Simonazzi et al. 2013).” (However, see (2). Obviously, the EMU doesn’t have an exchange rate regime to correct sustained imbalances. Apparently governments must impose what otherwise would have been done by exchange rate markets.)
  5. “Higher Wages and Higher Inflation in Germany Will Not Help.”

Storm on point 5:

“German exports and imports, as I argued above, are not very sensitive to changes in relative unit labor costs, however, and hence there will be only a limited amount of expenditure switching (away from German products and toward foreign goods), as has also been convincingly shown by Schröder (2015). Let me repeat for clarity’s sake that I am strongly in favor of higher nominal wage growth (in excess of labor productivity growth plus 2%) in Germany. It will definitely help Germany. But it will not help the crisis-countries of the Eurozone.”

“The assumption is that German GDP increases by € 100 billion (which means German GDP is growing at 3.7%). Through global production chains, [my emphasis] German growth creates € 29.5 billion of income in the rest of the world and about € 7 billion in the selected European countries listed in Table 1.”

This looks at production chains (Germany, USA, Korea) ! This may well be. But higher German wages would also mean higher German imports for consumption.

Storm’s view on the real issues (again)

Storm repeats what he regards as the real issues:

“(…) the common currency and monetary unification have led to a centrifugal process of structural divergence in terms of structures of production, employment and trade (as explained in my earlier notes).”

“German wage moderation mattered a lot, not through its supposed impact on cost competitiveness, but via its negative impacts on (wage-led) German growth and inflation, which in turn prompted the ECB to lower the interest rate in the first place.” (Comment: This “negative impact” is TU Delft slang for the idea that low wages reduce the need for Schumpeterian innovation.)

“The consequent crisis of the Eurozone is a deep crisis of inadequate aggregate demand in the short run and unmanageable structural divergence between major member states in the long run.”

I wonder. If Germany provided the European industrial zone and Southern Europe provided the European vineyards, olive trees and universities, then this might still work and everyone might be happy, provided that the prices of cars, wine, olive oil and Ph. D. doctorates would be right. Wage levels in Southern Europe might still be lower, but with a purchasing power parity (PPP) living standards might still be quite comparable. Sinn and Schäuble might like an argument that EU support for investments in Southern Europe should not be competitive to the German car industry (see here on the restauration of the Colosseum).

But this is not the full story. The Po valley has fine cars and machinery too. Italy itself has a North-South problem. Spain has the difference between Catalunya and Andalusia. And Germany has Laender who don’t do as well as Bayern.

Closing this review

This exchange started with Bofinger’s argument that German wages should be raised. This argument is fine. It will stimulate Germany’s domestic economy and imports. The obvious ceiling is provided by risks of unemployment and inflation, but the rule of a wage rise with productivity and the target of 2% inflation is fine too. Germany also has some catching up to do.

It is correct that German exports might not be much affected, and thus neither employment in the exporting sector, because the productivity growth in the exporting sector likely is larger than this growth in the domestic sector. But the rise of imports would still help in reducing the surplus on the external account.

Storm’s arguments on competitiveness & wage moderation are a different subject. This is basically the subject of investments and regional development, and the role of banking. Germany is advised to focus on domestic investments.

Economic analysis would be served by having another indicator alongside GDP, namely a correction of GDP for borrowed funds. The X – M correction works fine for foreign assets, but a correction for domestic borrowing would be helpful too. If one buys a domestic car with credit, then this domestic car really has been produced, but it would be indicative to know whether 10% or 25% of GDP would be from credit.

Overall I can repeat that my analysis of 1990 is still very relevant for understanding and solving the Great Stagflation since 1970. There are DRGTPE dating before the 2007+ crisis and CSBH after it. DRGTPE already has a chapter on the distinction between the exposed and sheltered sectors, and CSBH has a refinement of that argumentation.

It is unfortunate that our fellow economists at TU Delft have been neglecting that analysis since 1990, whence they still lack the full picture. But every day starts with a new sunrise.