PDF The Importance of Accurate, Reliable and Timely Data ...

TH E I M PORTAN CE OF ACCURATE, RELI ABLE AN D TI MELY DATA

Discussion Paper prepared for a Group of `Em inent Au st r a lia n s'

working wit h t he I ndigenous com m unit y of t he Goulburn Valley, Victoria to assist in independently m easuring and analysing the success of initiatives designed to foster an equitable lifestyle for

the region's I ndigenous people

Rum balara Foot ball Net ball Club Shepparton, Victoria

19th May 2006

by Saul Eslake Chief Econom ist Aust ralia & New Zealand Banking Group Lt d

Accurat e, reliable and t im ely inform at ion is vit al t o effect ive decision- m aking in alm ost every aspect of hum an endeavour, whether it be undertaken by individuals, com m unit y organizat ions, businesses or governm ent s. I t is an essential com ponent of any effort to persuade individuals, businesses or governm ents to m ake different decisions from the ones which they m ight m ake in t he absence of part icular pieces of inform at ion. And it is an int egral part of any attem pt to hold those who m ake decisions accountable for the consequences of the decisions which they m ake.

I n the absence of accurate, reliable and tim ely inform ation, people and organizations will m ake bad decisions; they will be unable to help or persuade others to m ake better decisions; and no- one will be able to ascertain whether the decisions m ade by particular individuals or organizations were the best ones that could have been m ade at the tim e.

I n short , inform at ion is a source of influence and power. That is why aut horit arian governm ent s seek t o cont rol access t o inform at ion. Open, enlight ened societ ies do not have `Minist ries of I nform at ion'; rat her, t hey put in place procedures ( such as `freedom of inform ation' legislation) which attem pt to prevent governm ents from restricting access to inform ation.

Not all inform ation can be sum m arized in num erical form s ( what are com m only called dat a) . And dat a ( such as weight s, volum es, counts, and m onet ary values) cannot capture every aspect of individual, business or governm ent act ivit y. But it is a truism that `what gets m easured gets valued'; and that what is not, or cannot be, m easured tends to be ignored.

The scope of what can be m easured and enum erat ed in t he form of dat a is widening alm ost continuously, not least as a result of advances in inform ation t echnology. Two of t he m ore obvious exam ples of t his are in sport and healt h care. The variet y and volum e of dat a on at hlet e or player perform ance available t o coaches, spect at ors, officials and players or at hlet es t hem selves has m ushroom ed in recent years, changing the way in which athletes and players are evaluat ed and t rained. Likewise t he abilit y t o m easure and calibrat e in m uch finer detail patients' breathing, blood flow and other physiological and psychological responses has dram atically changed the way in which illnesses are treated.

I n the business sector it is becom ing increasingly com m on ? som e m ight say essential ? to extract and analyse m inutely detailed inform ation about custom er behaviour, and about the profitability of individual products or services, points of sale and business units or even individual em ployees ? in pursuit of goals such as m arket share, productivity or profit.

Rising expect at ions ? including t he growing expect at ion t hat t hose ent rust ed wit h scarce resources ( such as public funds) have t o dem onst rat e t hat t hey have used them wisely, for the purposes for which they were provided ( that is, to be accountable for their use ? have also contributed to an increasing dem and for dat a.

From a developm ent or povert y alleviat ion perspect ive, as t he World Bank put s it ,

" St at ist ics are crucial in t he fight against povert y. They are t he essent ial starting point, telling us how m any people live below the poverty line in developing countries, how social, econom ic and environm ental conditions differ throughout regions, what infrastructure, health and education services are lacking across the world.

2

St at ist ics also t ell us how successful policies designed t o alleviat e povert y are: whether m aternal m ortality has decreased through investm ent in healt h care; whet her m ore children are at t ending school t hrough increases in teacher training; whether fewer people are starving through the im plem ent at ion of focussed agricult ural proj ect s for exam ple. They reveal if our goals are achievable and if our st rat egies are on t rack. They alert us when we need to m odify developm ent program s and re- direct resources"1.

Accurate, reliable and tim ely statistics are sim ilarly essent ial t o program s aim ed at alleviating poverty, deprivation and m arginalisation of I ndigenous com m unit ies, in Aust ralia and elsewhere. The explicit exclusion of I ndigenous inhabit ant s from st at ist ical collect ions such as t he Census prior t o 1968 was, arguably, part of a broader pattern of exclusion of I ndigenous com m unities from full part icipat ion in Aust ralian life.

And alt hough t he Aust ralian Bureau of St at ist ics, in part icular, has for m any years sought assiduously to collect and publish statistical inform ation relating to Aust ralia's I ndigenous inhabit ant s, explicit ly seeking t o involve I ndigenous people and com m unities in the gathering and use of statistics2, there are num erous gaps in essential inform ation ? particularly at the regional or local level ? and the accuracy and validity of the data is widely questioned, underm ining its relevance to decision- m aking and program evaluation.

As John Taylor of t he Cent re for Aboriginal Econom ic Policy Research at t he Aust ralian Nat ional Universit y point s out ,

" there appears to be a growing m ism atch between the broad direction that I ndigenous affairs policy is taking ? focussing effort on partnerships with specific regions, com m unities and even fam ilies ? and the availability of inform ation ( except from the five-yearly census) at these detailed levels ... Today, what is fram ed by governm ent as a new part nership approach exist s in I ndigenous affairs in Aust ralia, but invariably t he st at ist ical basis for assessing it s effect iveness is lacking"3.

The inadequacy of exist ing st at ist ics regarding I ndigenous populat ions ext ends t o the m ost basic question of how m ay I ndigenous people there are in particular locat ions. I n Tasm ania, for exam ple, census st at ist ics ( which suggest t hat a higher proportion of that State's population is of I ndigenous origin t han of any ot her j urisdict ion except t he Nort hern Territ ory) are widely believed ( including by prom inent representatives of the I ndigenous com m unity) to over- state the num ber of persons of Aboriginal descent . Conversely, in t he Sheppart on dist rict census st at ist ics are widely believed t o under- st at e t he size of t he local I ndigenous com m unity4. I n such circum st ances it is alm ost im possible for discussions about the effectiveness of program s aim ed at im proving I ndigenous education or health outcom es, or increasing I ndigenous participat ion in the labour m arket, not to be underm ined by doubts about the reliability of data.

1 Shaida Badiee ( Direct or, Developm ent Group) , Building St at ist ical Capacit y t o Monit or Developm ent Progress, The World Bank, 2005. 2 M. Walt er, `Collect ion and Disaggregat ion of Dat a Concerning I ndigenous People: An Aust ralian Perspect ive', Workshop on Dat a Collect ion and Disaggregat ion for I ndigenous Peoples ( New York, 19- 21 January 2004) , PF11/ 2002/ WS.1/ 9. 3 John Taylor, `I ndigenous Peoples and I ndicat ors of Well- Being: An Aust ralian Perspect ive on UNPFI I Global Fram eworks', Meet ing on I ndigenous Peoples and I ndicat ors of WellBeing ( Unit ed Nat ions Depart m ent of Econom ic and Social Affairs, Ot t awa, 22- 23 Mar ch 2006), p. 8. 4 Personal conversat ion wit h Paul Briggs, Direct or of t he Koori Research and I nform at ion Cent re, Sheppart on, 23rd March 2006.

3

And yet t he availabilit y of accurat e, t im ely and reliable dat a ? dat a which is accept ed wit hout quest ion as such by all st akeholders - could do so m uch t o inform discussion about the effectiveness or otherwise of such program s, or about the design of alternative program s where existing ones have been shown to have been ineffective, and to build persuasive argum ents for m aking strategic investm ents in new program s.

As but a sim ple exam ple, if there were a widely accepted estim ate of the num ber of I ndigenous people of working age ( 15- 65) in the Shepparton area, of t he num ber of those who were in paid em ploym ent, and of their average rem uneration, it would be possible to calculate the benefit to the broader com m unity ( for exam ple in term s of the net increase in purchasing power) of a program designed to increase the em ploym ent rate of I ndigenous inhabitants towards that of the com m unity as a whole.

Suppose that of the 2,500 I ndigenous inhabitants estim ated by the Shepparton Koori Research and I nform at ion Cent re t o be living in t he dist rict , 1,500 are of working age and that of them 300 or 20% are thought to be in paid employment earning an average of $30,000 per annum . Raising t his num ber t o 915 or 61% , roughly in line with the national average, at the sam e average incom e ( which is well below the national average) would thus inj ect an additional $27? m n ( gross) int o t he broader Sheppart on com m unit y each year.

Sim ilarly, accurate inform ation on I ndigenous housing occupancy would enable calculations to be m ade of the net com m unity benefit (for exam ple by way of rent paym ents to local landlords) of narrowing the readily observable wide gap between the num ber of I ndigenous residents per dwelling and that of the broader com m unity.

Such inform at ion would m at erially assist in m aking decisions about t he `ret urn' on any public ( or private) funds devoted to im proving em ploym ent or housing outcom es for I ndigenous citizens, and subsequently in evaluating the effectiveness of such program s.

However it is im possible to have a conversation am ong stakeholders on the basis of m utually understood and accepted data at present because it sim ply does not exist . That part ly reflect s t he ( t o som e ext ent underst andable) wariness of I ndigenous people in providing accurate inform ation about them selves and their fam ilies to governm ent agencies; and it partly reflects the lack of im portance attached to collecting such data by governm ent agencies them selves.

To be useful in ident ifying t he ext ent of problem s or t he effect iveness of program s designed to deal with them , data has to be collected and assem bled in a transparent, rigorous fashion, in accordance with established sam pling and other statistical procedures and free of any suggestion of pre- determ ined bias. There is no point in collect ing dat a and publishing dat a if it is open t o suggest ions of being derived from sam ples which are too sm all, or using processes which have been queried or discredited, so that subsequent argum ent revolves around the integrity of the data itself rather than the problem which it is intended to enum erate.

On t he ot her hand, it is also essent ial t hat t he way in which dat a are collect ed, assem bled and published respects I ndigenous sensitivities, not least so as to ensure that I ndigenous people understand the purposes for which the data is being collected and willingly provide accurate answers to data collect ors.

4

I n t hat sense, as John Taylor not es, `t he real challenge for st at ist icians is how t o achieve m easurem ent whilst respecting ( and incorporating) the cultural integrity of the subj ects'5. I n sum m ary, accurate, tim ely and reliable data cannot by them selves solve problem s of I ndigenous disadvantage and detachm ent ( or any other problem ) . But it is im possible t o im agine how t hose problem s can be solved ? or indeed, the com m itm ent and resources required to solve them can be m arshalled - without it.

5 John Taylor, op. cit ., p. 15.

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download