Kenya: “Of Perception Surveys and the Statistics Act”

The image “” cannot be displayed, because it contains errors.
Considering the source: The KHRC is so transparent that its organizational structure is completely invisible.

Although concern over corruption has moved to center stage in recent years, anti-corruption efforts have suffered from the impression that they are based on anecdotal or even biased information. For example, the widely discussed Corruption Perception Index, published once a year by Transparency International, a nonprofit advocacy group based in Germany, has been dismissed by some critics as an imprecise indicator because it is based on the opinions of international business people and research firms. —Diagnosis: Corruption (Interamerican Development Bank).

Kenya Human Rights Commission: “Of Perception Surveys and the Statistics Act” (November 3, 2006).

Apparently a “civil society” group — I say “apparently” because it does not provide an awful lot of information about itself (above) — supplementary to the governmental KNHCR, the KHCR comments on a “Statistics Act” regulating the conduct of opinion polling.

Quite a few countries regulate opinion polling now, requiring registration and certification of polling organizations and imposing polling “blackout” periods prior to the voting.

The concern apparently being that bad-faith or incompetent polling can be disruptive of the democratic information environment, rather than supporting it. See, for example,

Apparently, the Act curtailed activities by foreign groups, including the International Republican Institute.

What was that all about? Let me see what I can find out, while annotating.

For some previous (half-baked) thoughts of mine on the usefulness of “perception based” measures of public opinion as a substitute for “incidence” measures of social ills, see also

The KHRC’s complaint, which seems to be swimming against the tide of general recognition that “perception” indices are often less than meaningful as a contribution to deliberations on actions to be taken:

While civil society and media were commenting on the KHRC survey, the government through the Statistics Act draconically pulled the rag under the feet of those who have been involved in opinion polling and research. Not just the KHRC, but Kenyans are aware of Steadman Group, the Republican Institute, Afrobarometre [sic] and Gallup – including academic institutions. This Act, has been criticized variously by civil society, media and academic institutions, which view it as a violation of not only access to information (right to know) but also a serious affront on academic freedom. The freedoms to undertake independent research and thereafter disseminate that information were thus violated with the passage of this Act.

What is the Statistics Act, and how “draconian” is it in comparison with similar measures applied in other democratic countries? See, for example

Our UNAM think-tankers seem to take a very thorough comparative approach to the question, addressing the same issue addressed here:

There are those who affirm that the role of the surveys contributes to the development of democracy. Alejandro Moreno has indicated that surveys represent a “sign of liberalization in a political system that has had authoritarian features.” But there are also those who criticize the possibility of manipulation of survey results obtained by this technique.

Including me.

Interesting observation, from the same study:

The research firm Consulta Mitofsky has pointed out that electoral studies, such as exit polls, represent 8% of the total value of the research market in Mexico, but in terms of media impact, account for more than 92% of all the “noise” that is generated in relation to surveys. … At the present time there has been a proliferation in the abuse of such public opinion polls whose results are not based on the use of serious methodologies. As a result, surveys have been publicized that were undertaken without the use of representative samples or properly formulated questionnaires or that were conducted by bogus companies that impeccably report on the methodology employed, but whose results clearly reflect political bias.

The study cites a 1977 French law, still in effect, on the subject:

The image “” cannot be displayed, because it contains errors.

Back to the KHRC complaint:

In the recent past, opinion polls have found the government, including specific members of it, wanting in some respects. Similarly in the KHRC survey, the government was given good marks in some areas such as rights to life, education and worship but people also expressed dissatisfaction in other avenues such as security, food and water. For example, in total, 43 percent of the respondents gave the government a thumbs-up in protecting human rights, but 57 percent disagreed that government had a good record.

The polling was a good predictor of election results:

Whilst this was the total, some commentators were quick to point out two things: one, that result mirrored the results of the referendum, where ‘oranges’ trounced ‘bananas’ by the same margin. While these were perceptions of the state of human rights, they nonetheless mirrored the reality of the referendum results. Secondly, while this was the total, the individual ‘categories results’ were not much different. Illustratively, the exact 57-43 ratio was ‘perceived’ by those aged 18-24 years; those aged between 45-55 years; and all the women that were interviewed. The rest of other quintiles oscillated between 60-40 either way.

The provisions objected to:

To share such results, according to the Statistics Act, KHRC and other pollsters will need to submit not just their plans to conduct the survey according to Section 18(1) but will also be required to submit such results to the Kenya National Bureau of Statistics (KENBS) as provided by Section 19. What that means is that KHRC and other pollsters will have to satisfy a ‘government board’ before going public with the information. Whether plans and subsequent results, especially of those independent-minded institutions will be approved, your guess is as good as mine.

What procedures are contemplated for certifying such projects or not? Are they well-designed and transparent, and entrusted to impartial technocrats, or not? Does the Act speak to such questions?

Moreover, while the Statistics Act was being passed, civil society, media and some sympathetic MPs were unequivocal, of the need for government to either fundamentally amend the Official Secrets Act or repeal it altogether. The government’s lukewarm nature about the Freedom of Information Bill indicates its reluctance with opening the ‘highways of information’.

The information superhighway, I imagine they mean.

Is any of this an accurate description of the provisions of the act in question? You tell me:

I will say that a quick, shallow, personal survey of the transparency of Kenyan e-government — visiting Web sites in search of specific information — did at least suggest to me there is no such thing.

But the same goes for this “civil society” organization, if its level of disclosure about its organization, funding, and so on, are any indication. This “about us” statement, for example, is uninformative:

The Kenya Human Rights Commission (KHRC) was established in 1992 in response to serious human rights abuses by the government of Kenya against its people.

Established by whom? Who currently runs it? I am not finding answers to those basic questions here.

First, the KENBS and government by extension needs to be explained that perceptions are perceptions – not necessarily based on reality – but could nevertheless reflect reality. Perceptions are influenced by the media or debates about an issue under question. For example, if the Kenya Broadcasting Corporation (KBC) were to do a perception survey about the government performance, the government would score very well. First, because the KBC has overtime been a tool used by the political elite in government. On the other hand, the information reaching Kenyans, which they are supposed to ‘perceive’ and give opinion is ‘tainted and manipulated’ – propaganda by description.

The dreaded “headline effect.”

I need to look at the debate over this measure.

This seems to be one of those cases in which the argument against fails to give you a sufficient idea of the contents of the argument for.

Speaking of perception measurements, that recent BBC survey seemed to indicate that Kenyans place a great deal of trust in the fairness and accuracy of their media. (The survey press release does not choose to highlight how free they believe their press actually is.)

The image “” cannot be displayed, because it contains errors.
Speaking of perception measurements: Kenyans rate the performance of their press very highly, in terms of “honest and accuracy.” Then again, so do Nigerians. Have you ever actually read the Nigerian press? They also
value freedom of the press, according to a BBC-Synovate survey. See Brazilian News Media Garbles Survey Data on the Question, “Does Your News Media Report Honestly and Accurately?”

Secondly, the right to information albeit restricted by Acts such the Official Secrets Act, is guaranteed in the Constitution. Section 79(1) provides that unless with his or her own consent, a person is free to hold opinions and to impart them. The subsequent sub-section provides a claw-back, where such information is restricted insofar it interferes with public safety, morality, order, health or protecting reputations of others, their rights and freedoms. The formulators of the draconian, have not shown any cause that any of the past perception surveys, including the recent KHRC survey, have in way contravened section 79(2), so that any plans or results be subjected to a ‘government board’. Therefore, the constitutionality of the Statistics Act is therefore under serious scrutiny.

The Statistics Bureau, at least, does list its Board of Directors.

I am not quite clear on how the Statistics Act relates to the Official Secrets issue.

Lastly, it is evident that through the Statistics Act, government is unwilling to listen to its people – at its own peril of course. Perceptions on government performance oscillate from thumbs-up to thumbs-down depending on what government itself is doing. Illustratively, if a perception survey on media freedom was done after the Minister for Internal Security raided the Standard Media Group the results would be common knowledge – people would record their criticism of government.

Such a survey was done.

The IRI asked the question as part of a nationwide poll earlier this year:

The image “” cannot be displayed, because it contains errors.

And this, about government “keeping promises”:

The image “” cannot be displayed, because it contains errors.

On which IRI survey, more later.

Likewise, in areas northern of Kenya and North Eastern Province , the survey by KHRC indicates that Kenyans ‘perceive’ that government provision of security, food, water and documents of citizenship among others unsatisfactory. That perception is not ‘manufactured or doctored’ by KHRC, as the government is fully aware of how it has neglected these parts of the country for 43 years. Need we add more?

There seems to be a dispute here over some “perception survey” done by the KHRC — in partnership with IRI, Steadman, and Gallup? — that was accused of being “manufactured or doctored.”

As far as I can see, from a quick glance, Article 18 requires government certification to polling organizations:

Section 19 of the Statistics Act requires people doing polling to provide the underlying data, and information about collection, upon requestion.

I think I saw recently that Costa Rica also requires polling organizations to register.

The question being: How atypical is this requirement in terms of similar requirements elsewhere in the world?

When CNN-Gallup does a poll, for example, is it required to qualify as a qualified poll-taking organization — the way, for example, ratings agencies have to have SEC certifications as NSROs (“national statistical and research organizations”)

KHRC — which claims an affiliation with FIDH — could well have a perfectly legitimate case here, but I am not finding myself immediately outraged and up in arms, or seeing the draconian nature of the regulation imposed.

The devil, of course, is in the details.

And it would be helpful if KHRc took more care to practice what they preach in terms of transparency.

It reminds me a bit of the Brazilian project Contas Abertas, which recently won an Esso Prize in journalism here. Who funds and runs it? It is very, very difficult to find out.


The image “” cannot be displayed, because it contains errors.
USAID’s gift from the American people to the Elections Commission of Kenya included a modern communications network designed to make the vote-count faster and more efficient in 2007. Unexplained delays occurred, and the legitimacy of the result is not being accepted by anyone at this point. Thanks, American people, for exporting us your democracy, Texas Republic-style.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s