Can Transparency International be trusted
for Integrity?
Dr.
Atam Vetta
Special to South Asia Tribune
Transparency
International (TI) has published its 2002 report. It claims that
Bangladesh is the most corrupt country. India ranked 71st and Pakistan
is just below it (respective scores 2.7 and 2.6 out of 10). South
Asian newspapers have dutifully, reported the findings of TI. http://www.transparency.org/cpi/2002/cpi2002.en.html
This
is understandable because journalists have limited time and some
of them are over-impressed by European Organisations. TI lists about
118 'national chapters' and a few Regional Groups throughout the
world and ‘employs’ some people including retired bankers
and military men. A TI Report is in 2 parts; (1) A long book of
reports on different regions and (2) a short Statistical appendix.
The appendix gives the statistical justification for the Index.
The question is, does TI have enough integrity to be supported?
I regret my answer is, NO. My experiences relate to the 2001 report
which
was published a few weeks before the Bangladesh General Election
and was well publicised in South Asia (SA). The then Prime Minister
Sheikh Hasina was forced to defend her country. I was appalled by
the lack of sensitivity of the Bangla situation among TI and decided
to read the Report. My reading of the Report indicated that TI knew
nothing about corruption and wrote a Rejoinder (available: vettaatam@aol.com).
I explained
that there are 5 types of corruption, namely, (1) Selfish corruption
(by a person in authority and which most South Asians face every
day), (2) Egalitarian corruption (practiced by politicians in SA
where a small proportion of the corrupt money filters down to glue
the ‘vote bank’. This explains why corrupt politicians
are elected time and again). (3) Leech like corruption (practiced
by armed forces when they award big contracts). (4) Jehadi type
corruption and (5) Politico-economic corruption where rich individuals
or big companies give money to politicians to influence decisions.
TI look only at the fifth type of corruption i.e. the big business
corruption.
Why?
It believes that politico-economic corruption discourages big business
from investing in third world countries. False. Big companies and
rich individuals are only interested in the ‘return’
on the capital invested and the safe return of the capital itself.
In Suharto’s days, many companies and individuals queued up
to invest in Indonesia. Some countries that allowed enormous incoming
investment are in ruins. Which country would now want investment
from Enron and the World.com?
Serious
research in corruption can be dangerous. An example of serious research
is the South Asian Times’ investigation of land sales to high
military officers in Pakistan. The 2001 Report appears to be written
by people who relied on old newspapers. It has no value year after
year. Occasionally, an author knows very little about a country.
For Example, the 2001 Report said (pages 47-48) that “In the
late 1980s, the Swedish firm Bofors paid vast kickbacks to President
Rajiv Gandhi and middlemen in what became the worst corruption scandal
in Indian History” (italics added). It is the first time,
to my knowledge, that any one has accused the late Mr Gandhi of
having received kickbacks himself. Nor was he “President”
of India. About Bhutto, it said that “In March 2001, the Supreme
Court set aside the conviction of former prime minister Benazir
Bhutto and her husband on corruption charges and ordered a retrial
in the wake of an expose~ of intelligence tapes revealing the manipulation
of her trail by the Sharif government”(p.41). On page 44,
it said“ the alleged manipulation of her trail” (italics
added). There were also unsubstantiated statements about the Prime
Minister and Leader of the Opposition of Bangladesh. I was appalled.
I do not claim that they are not corrupt, only that TI had lost
touch with civilised behaviour. Moreover, the statistical knowledge
of the authors of CPI is abysmal and we discuss it later.
I sent
my Rejoinder to TI and put a note on their Visitors book with my
email address. TI ‘s reaction is interesting. First they removed
my email address. Secondly, Dr Robin Hoddis of TI in his letter
dated 20 November 2001, admonished me for circulating my paper,
something a scientist does instinctively. He said “Given that
we are the report’s editorial team, it might have been more
appropriate if you sent this rejoinder to us in the first instance,
rather than sending it to what appears to be dozens of TI and non-TI
recipients”. Frankly, I thought that, given my limited resources,
my efforts to publicise the deficiencies of the Report 2001 would
earn me praise from TI. I was disappointed by their reaction. Third,
he promised to correct the error pertaining to Rajiv Gandhi. Incidentally,
I followed his advice and sent a copy of my Comments on the Index
to TI Chairman. Despite repeated requests, I have received no reply!
In
my Rejionder, I made some specific points about Pakistan; (1) a
coalition government after elections, (2) The setting up of a Reconciliation
Board with the sole purpose of getting the ‘stolen’
money back. Later PPP stated that they will set up a ‘Truth’
Board. I welcome it, if the aim is to get the stolen money back
from American and European banks. (3) The civilian Pakistani government
should convince UN that the overthrow or attempted overthrow of
an elected government is a crime against humanity.
Corruption
Perception Index:
Does
the Corruption Perception Index (CPI) have any value? I believe
not and give below a summary of my criticisms;
1.
Social scientists construct an Index to compare yearly changes.
With CPI yearly comparisons are not possible. A country’s
position depends on the number of countries considered in a given
years. This number varies.
2.
The authors of CPI do not appear to know some basic statistical
concepts. For example, they said “Dividing the standard deviation
by the square root of the number of sources minus one yields an
unbiased measure which represents the standard error of the mean
score” (italics added). This statement is wrong. Actually,
standard error = standard deviation/Ön. Division is by Ön
and not by Ö(n-1).
3.
Their understanding of the concept of the 95% confidence interval
is deficient. They said “The true mean score can be said to
be within this interval with a 95% probability”. False, actually,
it means that in a large number of experiments, 95% of sample means
will lie in this interval. For one sample, the mean either lies
in the interval (probability =1) or it does not (probability=0).
4.
They said “The benefit of combining data in this manner is
that erratic findings from one source can be balanced by the inclusion
of other sources, which lowers the probability of misrepresenting
a country’s level of corruption” (Global Corruption
report 2001, p.233). False. We teach children that errors accumulate
and do not balance each other. A mathematically minded person would
know the formula Var (X1 ± X2) = Var. X1 +Var. X2.for two
independent variable. (Please ask for ‘Comments’, vettaatam
@aol.com)
5.
The same or similar ‘experts’ are used to rank a set
of countries each year. Therefore, big changes in the table are
unlikely. It was reported that General Musharaaf has asked TI to
advise the Pakistani government. I, therefore, suspect that Pakistan
will go above India in the 2003 Report.
6.
They said “However, determining this standard error requires
the assumption that there is no imprecision associated with the
values given by the individual sources”. I do not understand
this statement. Actually, we calculate standard errors when we take
samples.
7. CPI is an orgy of simple linear transformations. A group of countries
are standarised and standardised again.
Clearly,
I neither have confidence in the integrity of TI nor in their CPI.
What then needs to be done? South Asians are intelligent and resourceful.
21st century may become theirs. Their social scientists and statisticians
are second to none. They need to take their destiny in their own
hands. SAARC needs to look into the activities of TI and ask, should
the international community continue to spend millions on producing
junk Reports. Indeed, SAARC could take the initiative and after
consultation with the appropriate Ministers of Member countries,
ask International bodies that keep TI afloat if they would finance
a SAARC Committee to study corruption in SA to alleviate it. Given
the £millions involved, it will not be easy. Here is a challenge
for you, Mr. Rahim, Chief of SARRC and it can be achieved despite
Indo-Pak rivalries.
The
writer is a London-based Statistician who served as Chair of the
Association of Statistical Lecturers in Polytechnics (later universities)
and as Governor of the Abingdon College. He also served on the National
Council of NATFHE and as Chair of its Southern region. He also served
on the Oxfordshire Education Committee as HE representative for
about 10 years and on the Governing Body of Oxford Brookes for about
the same time.
Email Story | 
Discuss Story
