LIES, DAMN LIES >

The trou­ble is soon you won’t be able to call Dar’s of­fi­cial sta­tis­tics that.

The East African - - FRONT PAGE - Ai­dan Eyakuze is ex­ec­u­tive di­rec­tor of Twaweza. This ar­ti­cle was first pub­lished on Ox­fam­blogs.org

A new clause in the Sta­tis­tics Act ef­fec­tively out­laws factcheck­ing, un­less any fact-check­ing con­firms that the facts be­ing checked are cor­rect.”

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 into the sur­face of the At­lantic Ocean at a force 35 times greater than that of nor­mal grav­ity. None of the 228 peo­ple on board the May 31, 2009 Air France flight 447 from Rio de Janeiro to Paris 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, sen­sor fail­ure can be fa­tal.

Amend­ing Tan­za­nia’s Sta­tis­tics Act

Ac­cord­ing to the web­site of Tan­za­nia’s par­lia­ment, MPS last week passed the Writ­ten Laws (Mis­cel­la­neous Amend­ments (No.3) Bill 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 that is in­tended to in­val­i­date, dis­tort, or dis­credit of­fi­cial sta­tis­tics.”

This ar­ti­cle means that, 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 life­sav­ing 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 mo­bilised dur­ing the first phase of the Wa­ter Sec­tor Development Pro­gramme (20072014), 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 recog­nised 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 pol­i­cy­mak­ers about what cit­i­zens want. Of course, this does not al­ways mean that the pol­i­cy­mak­ers will de­cide to pri­ori­tise goals in ex­actly the same way, be­cause of com­pet­ing op­por­tu­ni­ties and con­straints on pol­i­cy­mak­ing 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.

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 one in four teach­ers who were sup­posed to be in class were 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 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.

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 with the in­stant tragedy of the AF 447 dis­as­ter. How­ever, the un­der­ly­ing prin­ci­ple is the same: All sorts of dis­as­ters can be averted 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 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 in­de­pen­dent sta­tis­tics.

Newspapers in English

Newspapers from Kenya

© PressReader. All rights reserved.