A data revolution in

development?
Rupert Simons
13 May, 2016

Topics for discussion

• Describe some development data gaps

• Understand initiatives being taken to close those gaps

• Assess the transparency of aid spending and results

Data gaps in an emergency: Sierra Leone
Managing Ebola

Mining industry collapses

Emergency budgeting

Iron ore price, USD per mt
180
135
90
45
0



Where are the patients?
Where are the
ambulances?
Have the Ebola relief
workers been paid?



Source: WHO, FT, IMF

2007 2009 2011 2013 2015

Mining companies all
privately held or quoted
on AIM, outside SL
Taxes, royalties fell from ~
$50m to ~$10m
Contracts, payments not
published




GDP fell by ~20% in 2015
Weekly tax take fell from
~$10 to ~$5 million in
worst weeks of Ebola
70% of budget is wages
>$100million emergency
loan needed from IMF

Data gaps in development: a global problem
Economic growth

Women who die giving birth

Nigeria GDP, USD per capita

Agricultural production

Maternal mortality, thousands

3000
2250

532

1500
750
0

1990
2007

2009

2011

2013


2015

2015


303

1950s methods measure
agriculture, modern
methods services
‘Rebasing’ doubled GDP in
Nigeria and Ghana
Lower income countries
get cheaper loans

Only 1 in 5 African
countries derive death
rates from national civic
registries
Maternal mortality
calculated from algorithm
based on GDP, fertility
rate and “skilled
attendance at birth”

Source: Morten Jerven, Development Initiatives, FAO



Ethiopian maize, tef
production used to grow
by 2-3% per year
Since 2008, production has
nearly doubled
10-15 million Ethiopians
still require food aid

Topics for discussion

• Describe some development data gaps

Understand initiatives being taken to close those gaps

• Assess the transparency of aid spending and results

Top down: Donors and UN agencies have helped fund
a growing number of household surveys

Unfortunately global priorities often trump national ones:

Source: UN report ‘A World that Counts’

Bottom up: there is growing use of citizen generated data . . .
Thousands of Indians report when they were asked for, paid or
refused to pay a bribe

Nepalese citizens use a local
helpdesk after the earthquake

Source: Open Nepal, ipaidabribe.com

. . . and there may be some big data
applications, though there are risks too
Example: Using cellphone calls to track malaria in Cote d’Ivoire

Source: UN report ‘A World that Counts’

Meanwhile, civic registration and
administrative data lags far behind
Quality and availability of data in African countries,
percentage of countries
Unusable or
unavailable

9
9

15
38

Out of date

58

40

Recent (last 10
years for census,
last 3-5 years for
survey)

60
78

29
82

13
45

Census

33

Poverty census

Source: Development Initiatives “Data Revolution in Africa”

29

18
9
22
Birth statistics

13

Estimates for closing data gaps range widely,
with $100-200 million per year as the median
Cost of building and maintaining statistical systems to
monitor the Sustainable Development Goals
USD million per year
1,100
900- 1,000
100-200

17-66

Source: Brookings; CGD; SDSN

New money
Existing aid
>10,000

Own resources
Not specified

Topics for discussion

• Describe some development data gaps

Understand initiatives being taken to close those gaps

• Assess the transparency of aid spending and results

Publish What You Fund reviewed the quality of
data produced by major aid donors
Over a data collection
period of 14 weeks

We worked with
people from

46

To assess 39 indicators of
data on aid and development
Maximum possible score

100

agencies in
65

22
and

countries

36

independent reviewers
and CSOs

Activity level data
(28 indicators)

Organisation level
data (8 indicators)
Commitment to
transparency (3
indicators)

25
10

What we found: Aid is getting more
transparent
Performance of agencies in the Index
Total

67

68

Very Poor
Poor
Fair
Good
Very Good

46
11%

39%

32%

15%
Ten donors did very well:

22%

35%

24%
24%
24%
12%
7%
6%

10%

2013 Index

2014 Index

Source: Aid Transparency Index 2013-16

17%

22%
2016 Index










UNDP
U.S. – MCC
UNICEF
UK – DFID
The Global Fund
World Bank – IDA
IADB
AsDB
Sweden
AfDB

The norm is gradually shifting from annual
reporting to quarterly, monthly or real time
Real-time data sharing
• Uploaded automatically
• Daily decision-making
• Limited editing

Regular updating
• Monthly or quarterly publication
• Machine-readable and web formats
• Edit for confidentiality

Annual reporting
• Companies House/ Charity Commission returns
• OECD-DAC Credit Reporting Statistics (for official donors)
• Annual report on website (usually in PDF)

Over 400 organisations are now publishing data
in the IATI standard
Number of IATI publishers by type
2011-15

300+

Source: IATI annual report 2015

However, only a minority of donors are
publishing forward-looking spending data
Forecast aid flows recorded in IATI
November 2015 snapshot

Source: IATI annual report 2015

Some governments are using the data
IATI import system in
use
IATI import system
under development
IATI partner country

NGOs are using it too,
for example:

Source: IATI, Development Gateway

Very few other people are using it

Flood victims have demanded that
some of Britain’s £12billion foreign
aid budget is used to help them
after a catastrophic few weeks.

Source: Press searches

Questions