The Missing Middle: What Australia’s Aid Transparency Reveals

Australia’s aid transparency architecture has changed significantly in recent years.

Through AusDevPortal, renewed reporting to the International Aid Transparency Initiative (IATI), and enhanced performance reporting, substantially more information is now publicly available about Australia’s development program than was the case even a few years ago. For the first time, it is becoming possible to explore parts of the aid program as a dataset rather than simply read about it through annual reports.

For researchers, policymakers and taxpayers alike, this is a welcome development.

Yet after spending time exploring the architecture, I have come away with a simple conclusion.

The next step in Australia’s transparency journey is not more indicators, more reports or more dashboards.

It is the publication of two datasets.

A Three-Tier Framework

The Australian development program is organised around a three-tier Performance and Delivery Framework.

Tier 1 measures development outcomes and trends across the Indo-Pacific, including poverty, economic growth, governance, health, gender equality, climate resilience and fragility.

Tier 2 measures Australia’s contribution to development, reporting results “directly attributable to Australian development efforts”.

Tier 3 measures the performance of Australia’s development investments.

Introduced under the 2023 International Development Policy, the framework provides a clear architecture for understanding how aid is delivered, what it achieves and how it contributes to broader development outcomes.

The logic is intuitive. Well-managed investments should generate development results, which should contribute to improved development outcomes.

Tier 3 → Tier 2 → Tier 1

Investment Performance → Results → Development Outcomes

The framework itself is not the problem.

The challenge lies in the availability of the underlying data.

Tier 1: Transparent but Not Published

Tier 1 is built largely from publicly available international datasets. These indicators draw on sources such as the World Bank, WHO, Freedom House, the Notre Dame Global Adaptation Index and the World Economic Forum.

The underlying data is generally accessible and well documented.

However, as far as I can tell, DFAT does not publish the complete Tier 1 dataset used to generate the reported indicators. Instead, it draws on these external sources and reports selected results.

Reconstructing the Tier 1 dataset is possible. But it requires querying multiple databases, harmonising country names and years, and assembling the information manually.

The data exists.

The dataset does not.

Tier 3: Increasingly Accessible

Tier 3 appears to be the most mature part of the transparency architecture.

Australia’s performance system relies on Investment Monitoring Reports (IMRs), Humanitarian Investment Monitoring Reports (HIMRs) and Final Investment Monitoring Reports (FIMRs). AusDevPortal, IATI reporting and annual performance reports now expose substantially more information about investments, financial allocations, policy markers and performance assessments than was previously available.

Independent analysts have already begun extracting and visualising parts of these datasets. Tools such as Scrittle demonstrate the potential of combining AusDevPortal and AusTender data to analyse the portfolio in new ways.

This represents a significant improvement over earlier periods when much of the performance architecture was visible only through summary reporting.

Questions about how investments are performing are becoming increasingly answerable.

Tier 2: The Missing Dataset

Tier 2 is where things become interesting.

DFAT publishes a substantial amount of information about Tier 2 results. Annual reports contain aggregate indicators, technical notes explain how those indicators are calculated, and case studies illustrate how results have been achieved.

The technical notes reveal a sophisticated results architecture. They contain detailed definitions, attribution rules, data quality standards, disaggregation requirements and reporting methodologies. These are not the characteristics of a system built around anecdotal reporting.

Indeed, it is difficult to see how figures such as:

165 million people reached through social protection programs;
54 countries supported to strengthen health systems; or
2.9 million people supported to improve climate adaptation and resilience

could be produced without aggregating detailed investment-level results.

The implication is straightforward.

A substantial Tier 2 dataset almost certainly exists.

What does not appear to exist publicly is the dataset itself.

Instead, external users see the outputs of the system:

aggregate results;
technical methodologies; and
selected case studies.

The underlying investment-level results data is not readily available for analysis.

Why This Matters

This matters because the current architecture makes some questions much easier to answer than others.

For example, it is increasingly possible to identify:

how much Australia spends;
where it spends it;
which investments are performing well; and
how overall development outcomes are changing.

What is much harder to determine is:

which investments contributed most to reported results;
which sectors generate the strongest results relative to expenditure;
whether highly rated investments produce stronger results; and
which delivery partners appear to perform most effectively.

Answering these questions requires moving beyond published aggregates to the underlying data that generated them.

At present, that is difficult.

Two Datasets

The good news is that the solution appears relatively straightforward.

The first step would be to publish the full Tier 1 dataset used to generate the reported indicators. The underlying data is already public, but DFAT could dramatically improve transparency and usability by consolidating it into a single downloadable dataset.

The second step would be to publish the underlying Tier 2 results dataset that sits behind the aggregate indicators.

Together, these two datasets would transform the analytical value of the framework.

Researchers could examine relationships between investment performance, development results and development outcomes without having to reconstruct the evidence base themselves.

More importantly, they would allow independent analysts to test assumptions, identify patterns and generate insights that are currently difficult to obtain.

The Next Frontier

Australia deserves considerable credit for the progress that has been made.

The transparency ecosystem is richer and more sophisticated than it was only a few years ago.

The next frontier, however, is not transparency alone.

It is analytical usability.

DFAT has already built much of the underlying architecture. Tier 1 data exists. Tier 2 results are being collected. Tier 3 performance information is increasingly available.

The missing step is to make the underlying datasets available in a form that allows others to analyse them.

If Australia’s aid transparency revolution has a next chapter, it may be as simple as publishing two spreadsheets.

The impact would be far greater than it sounds.

Australia’s aid transparency agenda has advanced significantly in recent years.

The launch of the AusDevPortal, renewed reporting to the International Aid Transparency Initiative (IATI), and publication of annual Performance of Australian Development Cooperation Reports have made more information available about Australia’s development program than at any point in its history. Independent initiatives such as Scrittle are beginning to demonstrate what can be done with these newly available datasets.

For researchers, policymakers and taxpayers alike, this is a welcome development.

Yet after spending time exploring the architecture, I have come away with an unexpected observation.

The challenge is no longer simply one of transparency. It is increasingly one of connectivity.

The Australian development program is organised around a three-tier Performance and Delivery Framework. Tier 1 measures the development context in the Indo-Pacific. Tier 2 measures Australia’s contribution to development. Tier 3 measures how Australia delivers development cooperation. The framework was introduced as part of Australia’s 2023 International Development Policy and is intended to improve transparency, accountability and learning across the development program.

The logic is intuitive. Well-managed investments should generate development results, which should contribute to improved development outcomes.

Yet when viewed through the lens of publicly available data, the relationships between these tiers are less clear than the framework itself suggests.

Tier 1: Transparent but Distributed

Tier 1 is built largely from publicly available international datasets.

DFAT explicitly notes that Tier 1 indicators are measured using open-source data. These indicators draw on organisations such as the World Bank, WHO, Freedom House, the Notre Dame Global Adaptation Index and the World Economic Forum. The data is available, documented and generally accessible.

Reconstructing Tier 1 would require effort, but not because the data is hidden. The challenge is that it is distributed across multiple systems and must be harmonised across countries, indicators and years.

In short, Tier 1 is transparent, but fragmented.

Tier 3: Increasingly Accessible

The second observation is that Tier 3 appears to be becoming more accessible.

Australia’s performance system relies on investment-level monitoring and assessment processes, including Investment Monitoring Reports (IMRs), Humanitarian Investment Monitoring Reports (HIMRs) and Final Investment Monitoring Reports (FIMRs). Annual PADC reports publish portfolio-level assessments, while AusDevPortal provides detailed investment information and access to a growing collection of documents and evaluations.

Investment-level information on effectiveness, efficiency, gender equality, disability equity, financial allocations and policy markers is increasingly visible through these systems. Independent analysts have already begun extracting and visualising parts of this information.

This represents a significant improvement over earlier periods when much of the performance architecture was visible only through summary reporting.

The result is that questions about the management of the aid program are becoming easier to investigate.

Tier 2: More Complex Than It First Appears

Tier 2 initially appeared to be the weakest part of the framework.

DFAT publishes aggregate results such as the number of countries supported, people reached or organisations strengthened. It also publishes technical notes and narrative case studies illustrating how development results were achieved.

What remains less clear is how much of the underlying evidence base is publicly accessible and in what form.

The existence of aggregate indicators implies that detailed underlying information exists somewhere. However, it is not obvious whether that information is:

publicly available through existing systems;
embedded within investment documents and evaluations;
accessible through unpublished datasets or APIs; or
held only within internal reporting systems.

The answer may be some combination of all four.

This distinction matters because it shapes the kinds of questions that can be answered.

The Questions the Current Architecture Struggles to Answer

The public reporting system provides valuable information about development outcomes, investment performance and aggregate results.

What it does not readily provide is a way to trace the relationships between them.

For example:

Do highly rated investments produce stronger development results?
Does effective investment management translate into better outcomes?
Which sectors generate the greatest development impact per dollar spent?
Which delivery partners consistently perform best?
Which interventions contribute most strongly to improvements in poverty, governance, resilience or health?

These are not criticisms of the framework. They are questions about what can be learned from the data that is currently visible.

At present, answering them would require bringing together information from multiple sources and, in many cases, reconstructing the relationships manually.

A Connectivity Problem Rather Than a Transparency Problem

The more I explored the data, the more I became convinced that the central issue is not necessarily a lack of transparency.

Rather, it is a lack of connectivity between different layers of evidence.

The framework is effective at describing three things:

the development context (Tier 1);
Australia’s reported contribution (Tier 2); and
the quality of aid delivery (Tier 3).

What is harder to see is how those three layers interact.

The public reporting architecture makes it difficult to trace the relationship between investment performance, development results and development outcomes. Whether the underlying evidence exists publicly, privately, or only in fragmented forms remains unclear.

This is an important distinction.

The challenge may not be that the evidence is absent. It may be that it is dispersed across different systems, reports, evaluations and datasets that are not yet connected in a way that enables systematic analysis.

The Next Frontier

Australia deserves considerable credit for the progress that has been made. The transparency ecosystem is richer and more sophisticated than it was only a few years ago. The OECD’s recent peer review of Australian development cooperation specifically noted improvements in transparency, online reporting and performance reporting.

The next frontier, however, may not be publishing more data.

It may be connecting the data that already exists.

Imagine being able to move seamlessly from:

Investment → Performance Rating → Development Result → Development Outcome

Such a capability would allow researchers, policymakers and citizens to explore not only what was achieved, but how and why.

The greatest opportunity in Australia’s aid transparency agenda may therefore lie not in the publication of additional indicators, but in improving the traceability of evidence across the three tiers of the performance framework.

The architecture is increasingly visible.

The relationships within it are still emerging.