We have been considering a KML download feature, which would contain both data and boundary information. Any thoughts you have on this would be much appreciated.

What I would like to to be able to do is request and download these composite files [spatial and aspatial] on a national basis in a defacto data standard (eg Shape files or Mid/Mif files) as opposed to having to merge the spatial/aspatial data myself. Is this a direction that ABS is going down?

A friend of mine recently had to go through 6month+ of approvals and state sanctions to carry out research with school students. I know this is a different situation, but the principle is the same, I think. EDA of government data has a lot to offer modern demography and social science, but I would think that the time lines for access place a lot of barriers on young researchers and PhD students. Even if you are researching stuff with small groups, you miss a big opportunity to study context and other broader questions without access to government data.

It looks like IATI are very closely linked with DFID in the UK and also with UNDP. Given that context, if you were interested in assistance with reviewing

(a) technical aspects of harnessing SDMX-ML to represent data and metadata structures which meet your needs (as opposed to developing an approach which is entirely IATA specific)
- SDMX, in a standard machine actionable manner, handles needs to cater for multilingual content etc

(b) applying appropriate SDMX content standards (eg countries, currencies etc) within data and metadata structures that meet your needs

then World Bank, as an SDMX sponsor, would probably be closest in terms of content/application.

IATA’s use case might (or might not) also be relevant to IMF and other sponsors reviewing possible requirements and approaches to offer an “SDMX Lite” to address concerns that SDMX in its current form may be “over-engineered” to easily support very simple use cases (although that engineering is instrumental to it meeting other use cases).

UNICEF’s DevInfo team have done a lot on delivering capabilities underpinned by SDMX to developing countries, and have worked with DFID, but my understanding is that they mainly focus on statistical data rather than financial “program management” data.

If you’d be interested in following up some of these aspects but you’re not sure of contacts in regard to SDMX then let me know.

Coming back to the theme of Don’s original post, one advantage of such an approach is that IATI should then be able to more readily draw on, and integrate with, relevant data from other sources.

Some more specific responses… If I understand it correctly, there are three key areas (from bottom of p. 2):

“Donors will publicly disclose regular, detailed and timely information on volume, allocation and, when available, results of development expenditure to enable more accurate budget, accounting and audit by developing countries.” - I’d say this is a very strong area for SDMX. One of the main user groups and indeed, five out of seven of the standard’s sponsors are financial institutes, so it’s aptly suited for this kind of work.

“Beginning now, donors and developing countries will regularly make public all conditions linked to disbursement” - there is nothing specific built into the SDMX model to help with this, but you could use annotations to document the conditions. You can do this as part of an SDMX metadata reporting structure, which allows the content and timing of future data releases to be released. You may want more than this, but it depends how “actionable” you need this information to be. If it is just to inform people, then it’s probably enough.

“Beginning now, donors will provide developing countries with regular and timely information on their rolling three- to five-year forward expenditure….” - I don’t know how much SDMX is used to exchange predicted results, as opposed to numbers that already exist. However, given the model has very clear mechanisms for handling versions of data and time series, it should work OK.