Tanzania is about to outlaw fact checking: here’s why that’s a problem

September 6, 2018 4 By Duncan Green

Guest post from Aidan Eyakuze, Twaweza’s Executive Director

Experts say it took just four minutes from beginning to end. First, some sensors failed. Then the pilots lost control of the plane, it stalled, went into freefall and smashed onto the surface of the Atlantic Ocean at a force 35 times greater than that of normal gravity. None of the 228 people on board Air France 447 from Rio de Janeiro to Paris on May 31, 2009 survived. The tragedy that unfolded that night was triggered by failed airspeed sensors. Without the air speed reading, the computer systems failed and the pilots, flying literally data blind, were unable to regain control of the aircraft.

Statistics are society’s sensors. Independently collected, processed, disseminated and debated, they are vital to the health of a country. We supress, fabricate or ignore them at our peril. At the risk of stretching the Air France 447 example to breaking point, sensory failure can be fatal.

Amending Tanzania’s Statistics Act

The parliament of Tanzania is considering a Written Laws (Miscellaneous Amendments (NO.3) Act 2018. It contains nine substantive amendments to the Statistics Act 2015. Most are positive – for example the offence of publishing statistics that are “false” or “may result to the distortion of facts” has been removed. But one proposed change is truly alarming. The amendments introduce the following new text in Article 24A(2):

‘A person shall not disseminate or otherwise communicate to the public any statistical information which is intended to invalidate, distort, or discredit official statistics.’

Sorry, that’s illegal now

Should this Article pass unchanged, if any official statistics happen to be incorrect (or even just disputable), then pointing out the problem and correcting it will be illegal. Any commentary querying or challenging official data would arguably be illegal under the amended Act, regardless of whether such commentary was correct or not. Indeed, this clause effectively outlaws fact-checking, unless any fact-checking confirms that the facts being checked are correct. Further, publication of any statistical information that contradicts, or merely cast doubt on, official statistics, could be prohibited under this amendment.

The value of independent statistics

Independent statistics can save lives. The Ramani Huria project has mapped poor, flood-prone areas in Dar es Salaam for flood modelling and therefore better upkeep of the infrastructure, improved warning systems, and improved and more accurate response in event of a flood crisis. Under the proposed amendments to the Statistics Act, these independently produced maps and the potentially life-saving information they contain could become illegal.

Independent statistics – when credible and transparent – can help boost the economy. Small business retailers are important to Tanzania’s economy, and could also be an important source of revenue. But, according to a 2017 report, “there are no official statistics on the number of small business retailers in Tanzania. However, secondary research data … indicates that the Total Addressable Market (TAM) is quite large and unpenetrated: one interviewee made a rough approximation of between 350,000 and 500,000 small retailers in the country.” These numbers would not exist were they not collected – and made public – by independent (non-state) entities.

Independent statistics, or data collected using methods pioneered by non-state actors, can improve inclusiveness. A recent water point mapping (WPM) exercise carried out jointly by the World Bank and the Tanzanian government and inspired largely by WPM methodology developed by WaterAid, helped to make a strong case that the resources mobilized during the first phase of the Water Sector Development Program (2007-2014), which increased Tanzania’s spending on water by a factor of four, had not led to the anticipated improvements in access to clean water. This motivated reflection by government and donors to ensure that future investments would have their desired impact. In addition, the government has recognized that WPM data can be a useful input to identify wards and villages with the greatest need and opportunity.

Governments benefit from understanding what citizens say they want and need. Unfortunately, those sentiments are almost never accurately captured in official data and statistics. Afrobarometer surveys inform government policy-makers about what citizens want. Of course, this does not always mean that policy-makers will decide to prioritize goals in exactly the same way, because of competing opportunities and constraints on policy-making and implementation. But without this feedback, governments may be surprised by negative reactions to their efforts. While public sentiment is often expressed in other ways than public opinion surveys — such as through protests, social media, etc. — the discipline of rigorous independent statistics allows researchers to be sure that such opinions actually represent the views of a larger population and to identify differences of opinion within key segments of the population.

Independent actors can fill crucial gaps in how public service delivery is monitored. This example from Uganda is instructive. In June 2003, the results of the first directly observed study of teacher attendance in a national sample of 100 Ugandan schools was presented to the Ministry of Education. The data revealed that more than 1 in 4 teachers, who were supposed to be in class, was away from school. The unanimous official review was that the methodology was invalid and the problem was nowhere near as large. Three years later, in May of 2006, another

anyone seen teacher?

independent national survey of schools yielded similar results to the 2003 survey. This time the review was mixed. A reluctant minority of officials acknowledged the problem. By the beginning of 2008, the ministry started discussing teacher absenteeism as an important challenge in service delivery. In May 2018, the Office of Prime Minister announced an initiative to use biometric machines and monetary penalties in 20 pilot districts to address the problem of teacher absenteeism. This example suggests that the government agencies responsible for teacher supervision are under-resourced and/or reluctant to publish embarrassing statistics. But, independent statistics can help reform-minded colleagues in government to act in the interest of development and their people.

Avoiding disasters through statistics

There is little visible drama in the lives damaged by flooding, poor access to clean water, or even teacher absenteeism, when compared to the instant tragedy of the AF 447 disaster. However, the underlying principle is the same: all sorts of disasters can be averted and avoided when decisions about what to invest in are informed by as complete a set of data and statistics as can be mustered.

Official statistics alone are a small and incomplete part of the full picture. Independent statistics make a huge difference. We need them to get a more accurate reading of our airspeed and to ensure there is sufficient lift under our collective wings. Amendments to Tanzania’s Statistics Act must promote, protect and defend independent statistics.