Meeting a challenge: Using government budgets to estimate annual public investment priorities in infrastructure

Globally, infrastructure investment is predominantly driven by governments. Public investment in infrastructure far surpasses private investment. For example, in developing economies, public investment is reported to represent around 83% of total infrastructure investment.

Yet, there is a remarkable lack of data available on a global scale about the infrastructure priorities of governments and the specific areas in which they are investing. Information on current government priorities can be invaluable to stakeholders like private investors, who use governments’ infrastructure project pipelines to inform their own investment strategies. It is also useful for countries to be able to see how and how much their peers are investing.

To help fill this gap, the GI Hub recently launched InfraTracker 2022, the first annual tracker of public investment in infrastructure for the G20. By diving into the details of central government budgets across 26 G20 member and guest economies, we estimate annual planned investment in infrastructure for 2022. InfraTracker also presents what transformative outcomes can be achieved through this investment to improve sustainability, inclusivity, and resilience while enhancing economic development.

This article provides a behind-the-scenes look at how we estimated public investment priorities, and why doing so isn’t as straightforward as it might seem.

Data on public investment in infrastructure is important, but existing estimations are limited

An important part of closing the vast global infrastructure investment gap is knowing the status quo – how much is currently being invested by both the public and private sectors. Unfortunately, there is a global and longstanding lack of consolidated, consistent, and current data on infrastructure investment levels and trends. In the absence of such data, policymakers and researchers often rely on what is known as Gross Fixed Capital Formation (GFCF) for General Government (a measure within the System of National Accounts), as a proxy for public investment in infrastructure.

However, using GFCF as a proxy comes with significant known limitations. For example, GFCF includes many items that are not considered to be infrastructure (such as intellectual property products and investment in defense and mining).  Sectoral breakdowns of GFCF figures are also generally not available, limiting the ability to extract more detailed insights. Moreover, there is significant time lag associated with the publication of GFCF data – typically around two years. This precludes a real-time view of government spending.

Various policy tracking tools are also available as an alternative source of information on government infrastructure priorities. However, these tend to have a narrower focus, either on a single sector such as energy, or a single priority such as climate change. They also tend to report high-level expenditure amounts across multiple years, rather than annual estimates, and do not disentangle capital expenditures from operational expenditures.

InfraTracker aims to address some of these issues by delving into the details of the latest annual government budgets to present planned capital expenditures for all infrastructure subsectors.

Estimating public investment in infrastructure from government budgets also has many challenges

Extracting public investment in infrastructure data from government budgets is a novel approach with many advantages, but is not without its challenges.

Firstly, budget documents are not uniform across countries. There is no single, standard way of reporting on investment in infrastructure, and the degree of granularity presented in publicly available budget documents varies significantly across countries. Some may only report expenditure details by ministry, while others report program- and project-level details. Some may report annual investment amounts, while others report only amounts for multi-year periods.

Secondly, ‘infrastructure’ is generally not an identified sector within budgets, and countries do not typically report budget estimates for ‘infrastructure’. This means that data must be extracted from the budgets of multiple line ministries, such as Ministries of Transport, Energy, Education, and Health.  An additional complicating factor is that ‘infrastructure’ is not uniformly defined across countries – there is no internationally agreed definition or classifications for infrastructure.  What’s more, the scope of infrastructure is constantly evolving in response to global trends such as digitalisation and climate change. What may be considered infrastructure today, such as nature-based solutions (NBS) or electric vehicle charging infrastructure, may not have been considered yesterday.

Thirdly, government budgets often do not identify capital expenditure for infrastructure, making it difficult to disentangle from operating expenses. This distinction is particularly important for social infrastructure sectors, such as health and education, where a large portion of government expenditure would be classified as operational, such as the payment of teacher salaries. While some countries do identify capital expenditures and capital transfers to subnational levels of government through budget codes (or otherwise), it is not common practice.

Lastly, while data from central governments can shed some light on public investment, the decentralised nature of function and finance in many countries means that investment in infrastructure is frequently carried out at a subnational level or by government and state-owned enterprises (SOEs). Such investment is difficult to estimate given the immense resources required to collect such granular data.

InfraTracker does include capital transfers by central governments to other levels of government, but it does not capture investment directly planned by other public entities like subnational governments or SOEs, given these significant data challenges. Such investments can be significant in many countries. For example, in the OECD economies, subnational governments are responsible for 57% of public investment.

All of these challenges mean that collecting and analysing data from government budgets is not a simple exercise. However, the approach does deliver new insights into the infrastructure priorities of G20 governments, and does so much more quickly than other methods, thereby providing high-level data for the private sector to identify government trends and areas of opportunity.

2022 results

The following insights were our top three findings from InfraTracker 2022:

1. G20 central governments budgeted almost USD1 trillion for infrastructure investment in 2022, equivalent to around 1% of total G20 GDP.

Of this trillion, 50% was central governments investing directly into infrastructure, and 50% was transfers to other levels of government for investment in infrastructure. 71% of the USD1 trillion was in advanced G20 economies and 29% in emerging G20 economies. The amount represents only a small share of the annual global infrastructure investment needed to keep pace with economic growth and achieve global climate goals and the UN Sustainable Development Goals (SDGs). For example, McKinsey estimates that USD3.7 trillion is needed annually to keep pace with projected GDP growth and up to a further USD1 trillion to meet the SDGs. These are also underestimates as they only include economic, and not social (such as health, education, or housing), infrastructure needs. Meanwhile, the IEA estimates that total annual capital investment in energy must reach almost USD5 trillion by 2030 in order to achieve net zero by 2050.

2. G20 central governments invest heavily in the transport sector, particularly roads.

The transport sector was allocated 42% of G20 governments’ total investment in infrastructure in 2022. This is more than double the second most popular sector (social infrastructure, at 17%) and around 2.5 times more than was invested in the energy, communications, water, and waste sectors combined. The dominance of the transport sector is evident in both advanced G20 economies and emerging G20 economies. It receives the highest share of central government infrastructure investment in both groups.

3. Inclusivity is the most targeted transformative outcome by G20 central governments, although climate goals also feature prominently.

Beyond economic growth and development, inclusivity was the most targeted transformative outcome by G20 central governments in 2022, with around half (51%) of investment in infrastructure targeting at least one inclusive outcome, such as inclusive mobility or affordability and access to services. Climate goals also featured prominently, with 39% of G20 central government investment in infrastructure targeting either environmental sustainability or resilience. The transport and social sectors made up around half of this investment (24% and 21% respectively). This pattern of public investment contrasts with trends in private investment in infrastructure; among private investors, there is minimal green investment outside of the renewable energy sector. Some examples of green investment within transport and social sectors include the UK’s Transport Decarbonisation Plan  and Singapore’s HDB Greenprint, which aims to transform public housing estates into more sustainable homes.

Conclusion

Data gaps are significant in the infrastructure space and have long been a challenge for infrastructure policymakers and practitioners. Current data on public investment in infrastructure, for example, are very limited, particularly on a global scale. Delving into the latest government budgets is one way to gather more insight into current government infrastructure spending trends and priorities, though not without its challenges. Understanding the current state of public investment in infrastructure is critical to understanding the global infrastructure investment gap and how to close it, ensuring that the right infrastructure decisions are made for a sustainable and inclusive future.