File - Muskoka

The North-South Institute/Canadian International Development
Platform/People & Code
The challenge(s)
• Muskoka – MNCH is Canada’s signature foreign aid initiative, at
$2.85bn over 5yrs (announced 2010) it makes up the single largest
share of Canadian aid
• The initiative has placed substantial emphasis on
transparency/accountability from the start
• Lot of very useful “open data” related to the initiative is available
• But this is fragmented, across sources, format types
• The complicated nature of the commitment ($1.1bn in “new”
funding + $1.75bn in “baseline”) makes analysis challenging,
despite open data
• IATI doesn’t contain ‘marker’ or ‘coding’ for Musk/MNCH
• Partner level traceability is not possible as no Canadian
Musk/MNCH executing partner (currently) publishes to IATI
• Impact/results info is at a general level (compared to partner info)
The solution(s)
• Work with a master list of Musk/MNCH unique IDs (across sources) to first
conduct a meta analysis of data sources
• Convert and link various data sources and types into one hub (e.g.
MongoDB) to ready API for JSON access which would allow visualization
(e.g. d3.js)
– Sources: DFATD-IATI; DFATD-Open Data; IATI registry; IATI datastore
– Types: XML, CSV; including manual upload to DB
• Demonstrate visualization that enables:
Tracking commitments vs. actual expenditures (burn rate)
Baseline vs. new funding breakdowns
Network/cluster graph of partnerships (type of organization, CSO, multilateral etc.)
Executing partners and organizations involved; subsectors/priorities
Country level mapped view; with disaggregation to project level
Allow extensibility to new data feeds (e.g. interchange with Can-MNCH alliance data)
Summary analysis
• There are 687 unique projects overall, however:
– If you only query current projects browser (DFATD open data) you will
find 574
– Musk-MNCH browser has fewer
– If you combine with historical (DFATD open data) you will find 496
– If you looked in current IATI you will only see 158
– There are differences in what info is available via which source
• There are several good reasons for these differences (even though
all are “open data” they have different ends); not just a problem of
data joins, quality (there are some issues but not insurmountable)
• Not only do sources have differences in the extent of coverage, but
also in attributes (which fields are covered)
• Overall, the initiative seems on-track, but it would be difficult to
reach this assessment without triangulating various sources
Summary analysis
• Data currently allows coverage from 2010-11 to 201213 (prelim).
• By year 3 (of 5) $1.77bn of $2.85bn has been spent
– $1.15bn in baseline and $0.62bn in new
• Vast majority of projects executed by foreign non-profit
partners, i.e. multilateral agencies, directly through
foreign governments; $1.13bn or 63.8%
• Canadian non-profits are second largest exec partner
$347mn –dominated by few large partners like
Micronutrients I; Aga Khan; Care; Plan; Save
– Several “nulls” and breaks in data yet to be worked out
Given the small number of large partners (Canadian) involved, and an already
established network (Can-MNCH alliance), Muskoka-MNCH implementers could be
excellent candidate for IATI publication pilot
Canadian Musk-MNCH partners could play valuable role in enhancing IATI results
tracking fields by leveraging their own results/impact ME systems to enrich IATI
data (and possibly the wider schema)
IATI publication would be better positioned as lessons/Knowledge-Sharing exercise
(than accounting/reporting)
Pilot could be worked into org work plans, could be limited (to start) to just Musk-MNCH projects
In the process, IATI often reduces or eases other reporting burdens
Official figures likely understate Canada’s contribution to global MNCH; not having
complete picture of Canadian orgs resource allocation, investment, partnerships,
we estimate, leaves out about 50% of the Canadian contribution
Bringing this in view could enhance results & impact analysis, narrative and build
further support for efforts
Combination of open source tools, IATI data standard, community (tech, data,
policy) support can make this happen

similar documents