Working with IATI - Mark Brough - Publish What you Fund

Working with IATI

Presentation to Open Data for Development Camp

Amsterdam, 12th May 2011

The aid revolution starts with <?xml?>

Presentation to Open Data for Development Camp

Amsterdam, 12th May 2011

What’s the problem?

Publish many times, use rarely

Publish Once, Use Often

The problem

• Lots of information in lots of different places• Of varying quality...• Different formats / classifications; not

compatible / comparable• Often not current or forward looking

Afghanistan• Of the $32 billion pledged by the US for

2001-2008, less than 20 percent ($6 billion) is recorded in the government’s aid database.

• That means Afghans have no way of knowing what’s happening with the other $26 billion the US has been spending in their country.

Towards a common standard

• It’s not a silver bullet – but it’s hard to see how aid effectiveness can be delivered without aid transparency

• Commitments under AAA and to deliver on PD• Key vehicle: International Aid Transparency

Initiative– Donor-led initiative to publish information in a

standard, comparable format• 8 EU Member States are signatories to IATI

The solution: open data and standards

• Use standard format for publishing raw aid information (IATI-XML)

• Publish it to your website• Register this data with the IATI Registry– www.iatiregistry.org

• Multiple infomediaries can access and use this information to meet specific stakeholder needs

• Set up in Accra, Ghana in September 2008• 19 Signatories– African Development Bank, World Bank, Asian

Development Bank, European Commission, United Nations Development Programme, Global Alliance for Vaccines and Immunisation, Hewlett Foundation

– Australia, Denmark, Finland, Germany, Ireland, Netherlands, New Zealand, Norway, Spain, Sweden, Switzerland, UK

• 2 Observers– France, US

19 developing countries have endorsed IATI

• Sierra Leone• Liberia• Bangladesh• Honduras• Republic of Congo• Democratic

Republic of Congo• Ghana• Rwanda• Indonesia• Nepal

• Viet Nam• Papua New Guinea• Moldova• Montenegro• Colombia• Burkina Faso• Malawi• The Dominican

Republic• Syria

Where we’re at now

• 9th February – Standard agreed• 28th January – DFID published all its projects to

IATI• 1st April – Hewlett Foundation published all its

projects to IATI

A project in DFID’s project-level database

The same DFID project in the IATI XML format

Hewlett Foundation’s IATI data

How are the donors doing?

• AidWatch Report: aid transparency• Sneak peak at the results so far, for 25

European donors• Full report launched 19th May 11am, Brussels

UK**

Denmark

**

Swed

en

Estonia

European

Commission*

Czech Rep

ublic

Luxem

bourg

Belgium**

Finlan

d

Austria

Slovak

iaSp

ain

Lithuan

ia*

Sloven

ia

German

y

Latvia

**

Netherl

ands* Ita

ly

Portugal

**

France*

*

Hungary**

Greece

Poland*

Cyprus

Malta

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

Information collected but not published

Information only sometimes published

Information systematically published

Commitment to Aid Transparency (FOIA and IATI)

Aid Transparency of 25 European Donors

* Donor did not have the opportunity to review the initial results, as results were collected too late; ** Donor was given the opportunity to review the initial results, but did not reply within 4 weeks; *** No information was collected on this donor

Ove

rall

scor

e ou

t of 3

8Embargoed until 19th May

UK**

Netherl

ands IA

TI

Denmark

**

Swed

en

Estonia

European

Commission*

Czech Rep

ublic

Luxem

bourg

Belgium**

Finlan

d

Austria

Slovak

iaSp

ain

Lithuan

ia*

Sloven

ia

German

y

Latvia

**

Netherl

ands* Ita

ly

Portugal

**

France*

*

Hungary**

Greece

Poland*

Cyprus

Malta

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

32

34

36

38

Information collected but not publishedInformation only sometimes publishedPyramide + AKVOInformation systematically publishedCommitment to Aid Transparency (FOIA and IATI)

Aid Transparency of 25 European Donors

* Donor did not have the opportunity to review the initial results, as results were collected too late; ** Donor was given the opportunity to review the initial results, but did not reply within 4 weeks; *** No information was collec -

ted on this donor

Ove

rall

scor

e ou

t of 3

8Embargoed until 19th May

Netherlands scores much better in IATI (it publishes more information)

• 12 Fields in internal database which are similar to IATI

• 5 Fields mapped across to IATI by AKVO• 2 Fields added by AKVO by looking in other

documents to enhance IATI data• Netherlands moves from 17th to 2nd if it

publishes this data for all its projects

Challenges for IATI

• Still some questions unanswered• Recipient budget identifier field still TBD: so

not yet linked to recipient country budgets.• Voluntary... Members and observers represent

over 2/3 of all ODA, but a lot of aid left out! • How many signatories will implement?• Optional components – not all fields are

compulsory, so how many will be used?

Alternatives to IATI

• OECD’s CRS / CRS ++– High quality statistics; data verified by OECD– Not detailed enough; not timely (latest is 2009)

• EC’s TR-AID– Should be IATI compatible, but not clear yet

• Bilateral initiatives

What’s next?

• More donors publishing their data to IATI – probably at least another 6 before November/December (at HLF4)

• IATI standard starts to get data fed through it; let’s see how it works

• People start to use IATI data!

Aid and domestic spending in Uganda

Questions• Sampled: for 1 country and 1 project• Three parts:– Organisation-level information– Country-level information– Activity (or project)-level information

• For each piece of info:– Is it published? (Y/N)

• systematically for all recipients/projects all of the time• just for some recipients/projects some of the time?

– If you don’t publish it, do you collect it?– Evidence: show where this information is (the URL)

Organisation-level questions• Does this donor publish aid allocation policies and

procedures?• Does this donor publish its procurement

procedures?• Does this donor publish the total development

budget for the next three years, as submitted to parliament?

• Does this donor publish their annual forward planning budget for assistance for the next three years?

Country-level questions• Choose your donor’s biggest recipient country

(e.g., India). Then answer these questions:– Does this donor publish the country strategy

paper for India?– Does this donor publish forward planning budget

or documents for the institutions they fund in India for the next three years?

– Does this donor publish its annual audit of its aid programmes in India?

Activity-level questions• Activity-level aid information– This is the specific detail about aid flows, needed for

informed decisions about where aid is / should be going

– This is the level needed for coordination, alignment, results/evaluation, ownership, and esp. accountability

– E.g. Is it good use of Polish aid to build a school in Gikongoro or Butare? • USAID is already building a school in Gikongoro. • The Rwandan government can support the ongoing cost of

only one school – in Butare or Gikongoro.

Proposed Process• Choose (or find) a current activity (or a

“project”) in the chosen recipient country• Answer as many questions as you can find info

on• NB not finding info is data!• Then send it to your donor agency

NB: could just send it all straight to the donor, but in our experience asking them to “check/correct” gets best results

Activity-level questions• Does this donor publish which organisation

implements the activity?– Needed for accountability; performance

evaluation• Does this donor publish the current status of

the aid activity?– Is it still ongoing? Is it supposed to be completed?

The Tracker• Online, frequently updated tool – in real time.• A page per country showing the relative

performance of donors across Europe, and then across the world.

• Will allow partners to monitor and encourage progress up to Higher Level Forum 4 in Busan.

• We’re consulting on the methodology to aggregate the data up and would welcome feedback.