Mapping Corruption Risks in Public Procurement Through Data

Up to 50% of public spending in developing countries and, on average, 29% of public spending in Organisation of Economic Co-operation and Development (OECD) countries is spent through public procurement. The process is complex, with many opportunities for manipulation, and the OECD estimates that 20%-30% of the value of most projects is lost to corruption (OECD 2016, ‘Preventing Corruption in Public Procurement’). This means that procured goods and services are often poor quality or dangerous, or fail to reach the intended beneficiaries. Corruption in public procurement also undermines confidence in government and excludes good companies from markets.

This research uses big data analytics to identify corruption risks in public procurement and provide an evidence base about which reforms are most effective in reducing corruption risks in this important area of public spending. It consists of detailed work to: collate and clean very large datasets; develop and test red flags or proxy ‘corruption risk indicators’; analyse changes in public procurement rules and practices; understand the dynamics of corruption in public procurement in different contexts through case studies; and undertake statistical analysis to test the impact of reforms on the distribution and scale of corruption risks across 100 countries.

Key Findings

• Corruption risks in public procurement are greater in countries where political regimes have shorter time horizons and weaker state capacity.
• Partisan favouritism, whereby a political party in office abuses its power to allocate contracts to its favoured allies, flourishes in conditions where politicians are able to capture and dominate the institutions responsible for (a) implementing and (b) monitoring procurement processes.
• Donor efforts to curb corruption by increasing oversight and transparency are effective in reducing some corruption risks in procurement [R5], and are particularly valuable in countries with low state capacity. This suggests that donors can control aid in weak state-capacity countries by substituting their own controls for the lack of local controls.

Impact and Implications

Public procurement accounted for around half of all public spending in developing countries and was a major channel for international development aid. Yet it remained highly vulnerable to corruption. This project helped to deepen understanding of how procurement processes were manipulated – and which interventions were most effective in reducing corruption risks. The findings contributed to efforts to safeguard public funds, improve service delivery, and strengthen trust in public institutions and markets.

Impact:

International Donors

The research team engaged heavily with donors throughout the research, e.g., presenting the methods and findings at internal workshops with the Foreign, Commonwealth, and Development Office (FCDO) (formerly the Department of International Development (DFID)), the World Bank, and the Inter-American Development Bank (IDB), to inform their practices and rules about disbursing aid and monitoring procurement and to provide evidence to support their advocacy work with national governments, so they could collect and publish better quality procurement data and introduce better monitoring systems.

The work fundamentally changed the approach of the World Bank, particularly its Solutions and Innovations in Procurement (SIP) team, which works to identify risks in Bank-financed contracts and to assist governments in improving their own risk management. The team also worked with some World Bank country offices to build awareness of the potential of big data analytics, eg, in August 2017, Sussex research team co-organised an event with the Bank’s country office in Dar es Salaam, the Tanzanian Public Procurement Regulatory Authority (PPRA), and the Tanzanian Prevention and Combatting of Corruption Board.

National governments

The research team also worked directly with public procurement regulators in two countries, Jamaica and Uganda, to develop online tools to assist their work. These interactive portals allow the regulators (Integrity Commission of Jamaica (ICJ) and Ugandan Public Procurement and Disposal of Assets Authority (PPDA)) to analyse their own procurement data, helping them to spot systemic corruption risks as well as high-risk individual transactions, hence informing policy change and investigation of cases.

Civil society

The team also worked with the African Maths Initiative (AMI), a Kenya-based non-governmental organisation that works on improving maths education in Africa, to incorporate the red flags methodology into their open-source, user-friendly software package, R-Instat (a front-end to R). The team also worked with AMI to organise workshops for maths students, civil society activists, and researchers in Tanzania (March 2017), Ghana (May 2018), and Uganda (October 2018).