samenwerkingsinstituut Sociale Keuze Theorie"
(Members taken from the
April 20, 2001
I think that it is useful
that I inform you that there has been a collective misconduct on your part,
something that goes against the integrity of science. I would advise you
to consider my enclosed clarification, and I would gladly accept the group’s
apologies, both to me and professor Donald Saari.
The issue is described in
[Adres] Scheveningen, Holland
cc. professor D. Saari, professor
R. Gill, and the chairpersons of Wiskundige Genootschap and American Mathematical
A note on the scientific
misconduct of the Dutch working group on Social Choice Theory (SCT) ("Interuniversitair
samenwerkingsinstituut Sociale Keuze Theorie") in relation to the 37th
"Nederlands Mathematisch Congres" (NMC)
Thomas Cool, April 20 2001
Which completes my experience
with the SCT working group in February, March and April this year.
As a preamble, it is useful
to restate that I protest against the abuse of power by the directorate
of the Dutch Central Planning Bureau (CPB). As an econometrician, I held
a position as a ‘scientific co-worker’ there from 1982-1991.
In 1989/90 I developed an analysis on unemployment, of which a part deals
with Arrow’s Theorem (for the social choice involved). The CPB
directorate blocked my analysis from discussion, while as a scientist I
should have the opportunity for internal discussion first before I would
submit the analysis for publication by the bureau. The directorate abused
its power by turning the issue into a labour dispute, and it used some
lies to get me fired. The issue is still in court, and I am asking the
academy of science (KNAW) to investigate te matter. Professor Richard Gill
(RUU, KNAW) has taken an interest in the case, and supports my request
Given this background, it must
be obvious that I dislike disputes in general. I dislike what CPB directorate
does, as I similarly dislike writing these lines. In particular since a
dispute may cause people to hypothise that perhaps the CPB directorate
had an issue. Yet in some cases, clarity is essential. In this case, I
think that there can be some educational value to all persons involved,
and the general public at large, that there are some costs when the integrity
of science is not respected. E.g. Holland has had a needless but huge (hidden)
unemployment for 10 years now (which can be generalised for Europe). I
am sorry to report, in this respect, that economists tend to disregard
my work and they tend to accept the abuse of power by the CPB directorate.
Therefor, it is useful to call for the cavalry, and ask the other professions,
and in particular the mathematicians, to investigate the case and to defend
the integrity of science.
Still in the preamble: Since
1990/91 I have hardly had time for research. Protesting against the abuse
costs time, just as it costs time to write (and read) these lines. I have
had various temporary jobs in different industries, and it took time to
learn the different fields. Yet it was a great joy in 1993 when I discovered
Mathematica, and since then I have been working steadily towards my major
recent publication "Definition and Reality in the General Theory of Political
Economy" (DRGTPE) (2000). I also have published "The Economics Pack, applications
of Mathematica" (TEP) (1998, 2001) and this January I put out "Voting Theory
for Democracy" (VTFD). All these can be ordered online at www.gopher.nl.
TEP has had a good review in the Economic Journal, it has sold about 100
copies and the software has brought me into contact with some fine economists.
Clearly, part of my strategy is to use this quality software to also draw
attention to my economic analysis and to my protest against the abuse of
science by the CPB directorate.
When I had written VTFD, this
January, I contacted professor Pieter Ruys (KUB) again, who had been a
contact person for me in 1990 on my analysis on Arrow’s Theorem.
In 1990 he did not understand the analysis, and my hope was and is that
VTFD will make a difference. Because I had seen a booklet on the subject
by professor Harrie de Swart (KUB), I contacted him as well. Both professors
were interested in the book, so I took the expense to send these to them.
It also appeared that there existed some working group on Social Choice
Theory, and to my surprise and delight, I was even invited to join.
For all clarity: the 1990 paper
on Arrow’s Theorem that the CPB directorate blocks from discussion
(and eventual publication by the CPB) is part of VTFD, and only reordered
since the book is about voting in general and not just Arrow’s
Leaving the preamble phase:
On February 23 2001 I first met the SCT working group. It appeared that
professor Donald Saari would come to Holland, and that there would also
be a SCT session at the 37th Nederlands Mathematisch Congres
(NMC). In that meeting of February 23 it was decided that I would present
my book VTFD at NMC. I mentioned, then, that this would be very useful
for Saari as well, since I have some useful comments to make on his work.
I also had brought along some
copies of my book. But nobody bought a copy.
In the train back home I encountered
Eliora van der Hout again, one working group member. She said that she
would study the book. It also appeared that she was a room-mate of dr.
Annemarie ter Veer - who I had met a few years ago - and who also appeared
to be a member of the SCT working group. In that period, Annemarie was
with the political party De Groenen, and this party had taken an interest
in my economic analysis. In fact, a delegation of that party took me along
to the CPB, to ask CPB to do some calculations on the analysis. Eventually,
though, at some discussion with De Groenen, there was some political manipulation
- not by Annemarie ! - and I decided that I would no longer advise this
A few days after the meeting,
Harrie informed me that my presentation was annulled. Some people, anonymous
to me, had contacted him after the meeting, and had said that they feared
that my work would not be up to the standards.
I hope that it is clear what
happens here. (a) Since the objections to a presentation were not voiced
in my presence, I have not been able to give answers. (b) See the psychological
effect on Saari: If my book has been recently published and there is no
session on it, such that Saari could attend, then he might think that the
book has no relevance for him.
One should know that I, through
my work on Mathematica, have had some contact with Alex Tabarrok, and that
I, through him, had learned about Saari’s work at an earlier
stage. I have indeed tried to contact Saari a few years ago. In fact, both
Tabarrok and Saari moved to California afterwards. Unfortunately, no contact
with Saari was established back then. Given Saari’s importance
for the field, one can understand that I think that it is important that
we are in contact.
Thanks to Harrie de Swart, two
thinks happened: (a) He gave me an introduction to professor Saari by email,
so that I could send him my book VTFD and so that he could look at it before
he came to Holland. Saari replied with an email that he would be delighted
to look at the book. (b) I was invited to give a presentation about my
book for the working group. This presentation was on March 16.
Present at my March 16 presentation
were Ruys, Storcken, De Swart, Monsuur, Bosch, Rusinowska and some students
who are unknown to me. Afterwards I got some applause, which sounded honest.
Ton Storcken promised to study my book (of which I have taken the expense
of sending him a copy as well).
My problem with the meeting
however was: (a) I was told that some people had decided not to come, since
they thought that the presentation would not be up to their mathematical
standards. The group officially has 28 members or so, my impression is
that some 14 turn up regularly, but at my March 16 presentation only 6
turned up (excluding the students). (b) There was no further discussion
about the issue of me and Saari.
Hence, I wrote a letter to Harrie
and Pieter to emphasise that the mathematical level of my analysis should
not be underestimated, and that the analysis should deserve serious attention
by the working group.
There was a strange problem
with dates. Harrie first announced that Saari would give a full day workshop
on Tuesday April 17. Even though my work is of the highest quality, I still
have no official Ph. D. title yet, and for the current universities in
Holland that means that I cannot work at an university. Currently, I teach
parttime at a professional college. When I applied for a leave, my new
head decided, suddenly, that lecturers are allowed only one day of leave
per semester. So I had to choose between a presentation on Webmathematica
of April 24 and Saari’s lecture of the 17th. I arranged
my leave for April 17. However, a week before the 17th, Harrie
wrote an email in which he proposed a change of this workshop to the 18th.
I wrote back that any change was OK, but that I needed to know this sufficiently
in advance because I have to make arrangements with my school. When I tried
to contact Harry what the final day would be, I could not reach him. Since
it was quite uncertain on which day Saari would have the workshop, and
since I could not take 2 days of leave, I cancelled the 17th.
On the 17th (after Easter) it appeared that the workshop would
be on Wednesday the 18th - but it was too late to make arrangements.
It so happens, for Wednesday’s,
that I give only morning lectures. So I could attend Saari’s
Wednesday afternoon session. Due to a delayed train, I came in late just
after lunch and just before the lecture started at 14:00 hours. Finally
I met Saari, but thus severely handicaped in way of introduction.
In the break of the lecture,
Harrie told me that he had received my letter, and that he would take the
time to really look at my book. I would have to wait till September before
he could have a reaction, and he warned me that it even might be later
than that. As such, I could only accept this, since I am happy will all
real interest that people can foster.
The workshop ended around 16:00
hours. Since I had no questions on this afternoon lecture, and since I
missed the morning lecture, I considered it proper to remain silent on
my views concerning Saari’s confusion, and thus I did not engage
the whole working group into a discussion on that.
When everybody appeared to leave,
I finally had time to ask professor Saari whether he had looked at my book,
what his schedule was, and whether he would have time to discuss issues.
He said that he had been ill, and had had no time to look at my book.
Thus: Given that Saari had had
no time to look at my book, and given that I arrived only just before his
second lecture, Saari had no way of knowing about my knowledge of his work.
Thus Wednesday April 18th, Saari and I talked for perhaps an
hour, but this was a talk with lots of misunderstandings on his part, while
I felt at a great disadvantage since I had expected that he would have
had a look at my book.
It appeared that the AMS has
put out a publication by Saari, "Chaotic Elections", 2001. The flyer of
that book of course restates Saari’s confusions. The AMS flyer
also states: "Saari’s book should be required reading for anyone
who wants to understand what happened not only in the presidential election
of 2000, but also how we can avoid similar problems from appearing anytime
any group is making a choice using a voting procedure." Well, I think that
the book would be useful reading - but it is my VTFD that is required reading.
Which is why I think that protesting against these events in the SCT working
group is so important as well.
Professor Saari and I may agree
98%, but the 2% difference is the difference between a human being and
a chimpansee. I can explain the difference in a few statements, and I have
done so for Saari, but one needs to think about it, and the statements
apparently need their context before people understand them.
Thus for example, professor
Saari says that Arrow’s Theorem is caused by the lack of distinction
between rational and irrational voters, he uses the same cause for Sen’s
Theorem, and he uses this commonality as an argument for having identified
the true cause of the paradoxes. Yet, in my analysis, the cause for Arrow’s
Theorem is the confusion between ‘voting’ (fields of
voting results) and deciding, while Sen’s Theorem is based on
a confusion between personal and social choices. For both points, see VTFD.
In both cases, we have paradoxical theorems, and it is proper indeed to
look for causes why we think that these results are paradoxical. Both Saari
and I accept the theorems as mathematical truths, but we reject their common
interpretation. But since our causes are different, the implications are
different. For all clarity, I also say, and discovered this independently
from Saari, that Arrow’s Theorem (and the ‘axiom of
pairwise decision making’ in particular) does not use all information
that is available, yet this is not the cause for the theorem.
Arrow’s wrong interpretation
of his theorem has been confusing SCT for 50 years. It would be a real
pity if Saari’s wrong interpretation would give us another 50
years of confusion. The danger of that is large, since Saari has presented
some wonderful work in geometry, and has given some wonderful new tools.
Mathematicians are in danger of embracing all Saari’s work, the
tools and the interpretation, in one sweep - just as they did with Arrow.
The effect of ingrained biases can be very strong. I have presented the
proper analysis since 1990, and people, such as also the CPB directorate,
have not treated this analysis with the proper respect. Criticising Arrow
was ‘not done’. I am very grateful to professor Saari
for creating room to be more critical of Arrow’s interpretation.
But it would be wrong when Saari’s interpretation would cause
people to stop thinking again. In fact, we could already see this happen
at the 37th NMC.
Thus, at the 37th
NM Congres, Saari presented the main theme of his work in a noon lecture.
I did not see Harrie de Swart
before that lecture started, and thus I could not inform him that Saari
had had no time to look at my book.
At one moment Saari asked the
audience how they would explain Sen’s Theorem. Since Saari and
I had not discussed Sen’s Theorem, I raised my hand to offer
my explanation. However, Saari rejected my offer, saying ‘We
had a discussion yesterday’, and thus I was not allowed to speak.
Clearly, Saari allowed his confusion about the two theorems to dominate
the discussion. Well, at that moment he had the floor, so one should allow
for this, and it is proper that people can hear what he thinks.
At the end of the lecture, there
was time for questions. The evening before I had decided that it would
be proper for me to speak up. Thus I did. And I stated for the audience
that I have a great respect for professor Saari, and that he has done some
brilliant work, but that he entertains a confusion on some crucial points,
and that it is important to study those points.
Professor Saari gave the impression
to be angered by my statement. He even said: "You say that I am wrong,
but you don’t give arguments why." This apparently got approval
with the audience, since, indeed, at that moment, it seemed as if I had
not given any argument. Yet, it should be obvious that I have given my
arguments - I had sent him my book - and that professor Saari only has
had not enough time to properly consider them.
Of course, I also can understand
professor Saari a bit here. He will have encountered, as I have, lots of
unspoken opposition to his criticism of Arrow’s interpretation.
Perhaps my remark reminded him of that.
Yet, of course, the proper reaction
would have been to invite me to present my arguments to the audience. I
had my sheets of my presentation of March 16 with me, and it would have
taken only 20 minutes, which time was available. The audience could have
voted on this, or only the interested people could have stayed - the others
voting with their feet.
After the lecture, I checked
with Saari whether he was angry indeed, and it appeared that this was not
the case. I also reminded him that I did have arguments.
Harrie de Swart appeared unhappy
with my remark, but I think that I have succeeded in explaining that I
have only decent objectives in having made it.
After these events professor
Saari and I did not speak with each other again, though there seemed to
be ample opportunity for that.
The afternoon session of the
SCT group had only two presentations, by Rob Bosch and Hans Peters. Again,
there would have been ample time to allow me a presentation as well. But
nobody invited me to do so. Rob Bosch in fact started copying Saari’s
I can conclude that the SCT
working group collectively misconducts in the following way:
Clearly, since there are lots
of signs of good intentions, there is reason for some optimism. Also, there
have been some unfortunate events, like Saari’s illness and the
change of the lecture dates. Yet still, I think that the group owes an
apology to professor Saari and me. Of course, at this moment professor
Saari has not read my book yet, and thus he does not know how important
the issue is. But eventually I can only think that he would agree.
There are biases against the
quality of my work, and the group allows those biases to run affairs instead
of that the group deliberately tries to get clarity. Fortunately, some
individuals like Harrie de Swart and Ton Storcken in particular show good
intentions, but apparently they also have busy agenda’s, and
there is a large risk that they simple do not take sufficient time to get
to the root of the analysis.
The process of the selection
of speakers for the NMC was improper - and it allowed those biases to grow
I told the group that they did
not understand social choice theory, that they needed to study my book,
and that it would be important that we tried to get the argumentation across
to Saari. Just the same as I told in 1990. I might as well have said this
to a wall. As a scientist, however, I must expect that if people disagree
on a statement, that they give counterarguments. The group did not do this.
The group created a situation
such that I was forced to a make a formal statement at Saari’s
NMC lecture - and such that Saari made the error of falsely accusing me.
Various people in the audience, who are no regular visitors of the SCT
working group, will have had a strange idea of what was happening.
My other suggestion is that
the group now starts studying the book, that we have some sessions on it
in September, and that we afterwards invite professor Saari again.
Appendix: Professor Saari's
email about my book
>From mailnull Wed Feb 28
Resent-from: "H. de Swart"
<H.C.M.deSwart @ kub.nl>
Resent-to: cool @ dataweb.nl
Resent-date: Wed, 28 Feb
2001 21:22:26 MET
X-Sender: dsaar @ e4e.uci.edu
Date: Wed, 28 Feb 2001 12:18:27
To: "H. de Swart" <H.C.M.deSwart@kub.nl>
From: Donald Saari <firstname.lastname@example.org>
Subject: Re: book by Thomas
>In our research group on
Social Choice Theory, you will meet
>Thomas Cool, who has recently
written a book "Voting Theory
>for Democracy", in which
he also discusses your work.
>He has designed a Borda
Fixed Point, which differs from your
>preference for Borda itself.
>He would be delighted to
send you his book, so that you
>could read about his approach.
If you would be interested,
>please tell me, so that
I can tell Thomas that he can send
>you the book indeed. Otherwise,
I'm sure that you will be
>able to discuss matters
when you meet him.
>With kind regards and looking
forward to see you, Harrie de Swart.
>H.C.M. de Swart e-mail:
>H.C.M.deSwart @ kub.nl
>Chair in Logic and Linguistic
Analysis tel. (0031) 13 4662415
>Tilburg University fax.
(0031) 13 4662892
>P.O. Box 90153
>5000 LE Tilburg http://cwis.kub.nl/~fsw_2/fww/home/swart/index.htm
This would be delightful!
If I have the book in advance, I
could (hopefully) have it
read before I arrive so I could be more
informed in our discussions.
Either mailing address given
below would suffice.
Donald G. Saari, UCI Distinguished
Professor of Mathematics and Economics
Economics: 3295 Social Science
Plaza, (949) 824 5894
Mathematics: 233 Multipurpose
Science & Eng. Blg, (949) 824-7121
Mail: Either Department,
University of California, Irvine,
Irvine, CA 92697, dsaari
Appendix: My April 19 email
to professor Saari on his false accusation
Dear professor Saari,
Today, at the 37th Nederlands
Mathematisch Congres, held at Vrije Universiteit Amsterdam, I again informed
you that you are on an off-track course concerning voting theory and the
theory of social welfare. What your articles 'crucially' say about voting
theory and social welfare is unwarrented, and your statements could be
detrimental to democracy and social welfare themselves.
In reply to this remark of
mine, you replied that I did not give any reasons to substantiate my statement.
Your comment caused the audience to neglect my remark, at that moment.
However, your reply is improper,
since I did send you my book, that contains this substantiation. You told
me earlier - Wednesday, when we met for the first time - that you had been
ill before, and that you did not have time to look at the book. Below email
exchange shows that there has been some effort to get my book to you. I
take your word for it that it has arrived. Thus there is a clear difference
between the existence of my analysis and your lack of time to study it.
I would like to have it on
record that I greatly appreciate your work.
Thousands of years hence,
students will use your representations to understand voting paradoxes.
You created a tool, in the same way as Descartes created analytical geometry.
Your contribution to the subject is brilliant - and a proof of a determined
intelligence trying to discover a hidden truth. Yet, there is a fine line,
separating your true contribution from the 'off track' conclusions. On
some aspects, you are too rash, and some of your conclusions are not valid.
I truely forgive you your
remark, on my supposed lack of providing argumentation, since I know how
hard it is to work on the frontier. Yet I also know that our colleagues
are easily mislead. For 50 years they followed Kenneth Arrow, and now they
might as well follow you - but also on you rash conclusions!
Currently, I am quite at
a loss as to the current situation. The best thing probably is that you
inform all people involved, that your statement, that I did not provide
any substantiation, was incorrect. I please ask you to do so, though the
damage already has been done.
My best regards,