Better Roads for Less: Brazil’s Performance Contracts Offer a New Playbook for Highway Investment

Brazil’s road experience shows performance-based contracts can cut rehabilitation costs by 65–70% and maintenance costs by 50–55%, helping governments stretch limited infrastructure budgets. However, better pavement alone does not improve road safety, highlighting the need for dedicated safety measures, smarter contracting and stronger transport data systems.

Brazil's experience in managing its vast federal highway network offers an important lesson for governments struggling with tight infrastructure budgets: better contract design can significantly reduce road rehabilitation and maintenance costs, but smoother roads alone will not necessarily make travel safer. A 2026 World Bank study, based on data from Brazil's National Land Transport Agency (ANTT), Federal Road Administration (DNIT) and Federal Highway Police (PRF), analysed much of the country's approximately 64,000-kilometre federal paved road network between 2013 and 2025. It found that performance-based contracts can be 65–70 percent cheaper for rehabilitation works and 50–55 percent cheaper for routine maintenance than comparable traditional contracts, potentially giving governments more infrastructure for limited public spending.

Performance Contracts Could Stretch Every Infrastructure Dollar Further

Brazil uses three main approaches to manage federal roads: traditional input-based contracts, performance-based contracts (PBCs), known locally as CREMA, and private concessions. About 25 percent of the federal network has been transferred to private concessionaires, largely supported by toll revenues.

However, concessions cannot work everywhere. Lower and medium-traffic highways may not generate enough toll revenue to cover investment and operating costs, while tolls may also face public acceptance challenges.

This is where PBCs could provide an alternative. Traditional contracts consist mainly of routine maintenance, representing about 90 percent of interventions by number, while rehabilitation accounts for roughly 10 percent. PBCs have a very different composition: around 80 percent rehabilitation and 20 percent routine maintenance.

Although this makes PBCs appear more expensive when total spending is compared, the World Bank finds that similar activities cost substantially less under performance contracting. Rehabilitation is approximately 65–70 percent cheaper, while routine maintenance costs 50–55 percent less than under traditional contracts.

For governments, particularly those facing fiscal constraints, these savings could allow limited transport budgets to cover larger road networks or redirect resources towards other infrastructure priorities.

Concessions Bring Better Roads, But at a Higher Price

Private concessions offer a different value proposition. The study finds that concession roads receive substantially higher investment and consistently achieve better pavement conditions than publicly managed roads, even after accounting for spending and road characteristics.

Road quality was assessed through the International Roughness Index (IRI). Lower IRI scores indicate smoother roads. Values below 2.5 can indicate relatively good road conditions, while levels above 3.5 may indicate inadequate conditions.

Higher road asset management investment was significantly associated with better pavement quality. However, concession maintenance is more expensive partly because concessionaires provide a wider package of services, including road-user assistance and emergency response.

The findings suggest governments should avoid applying one contract model across every highway. High-traffic corridors capable of supporting toll revenues may be suitable for concessions, while medium-traffic roads could benefit from performance-based contracts that deliver rehabilitation and maintenance without requiring tolls.

For international development partners, this creates an opportunity to support governments in classifying road corridors according to traffic levels, financial viability, maintenance requirements and expected economic benefits before choosing financing and contracting models.

Smoother Highways Alone Will Not Stop Road Deaths

One of the study's most important findings challenges a common assumption in transport investment: improving pavement quality does not automatically reduce crashes, injuries or fatalities.

After controlling for differences between roads, researchers found no clear statistical relationship between improved pavement conditions or concession status and better safety outcomes. Traffic volumes and multilane road configurations showed stronger relationships with crash outcomes, although the study stresses that this largely reflects greater traffic exposure rather than evidence that multilane highways are inherently dangerous.

The implication for policymakers is significant. Governments should not treat road rehabilitation programmes as substitutes for dedicated road-safety investments.

Maintenance contracts and transport projects need specific safety measures, including better road geometry, signage, speed management, enforcement and emergency response. Development banks and other financing institutions could similarly integrate measurable safety requirements into road projects instead of focusing primarily on kilometres rehabilitated or pavement quality.

Better Contracts, Better Data and Smarter Investment Choices

For private-sector stakeholders, wider adoption of PBCs could create opportunities beyond conventional toll concessions. Construction and infrastructure companies capable of maintaining roads over several years and meeting measurable performance standards could participate in corridors that are not commercially suitable for toll-based concessions.

The model also carries risks. Contractors become responsible for maintaining agreed performance levels rather than simply completing individual works, while governments must have sufficient fiscal and institutional capacity to make payments throughout multi-year contracts.

Brazil's experience also exposes the importance of better infrastructure data. Researchers had to combine financial information, pavement measurements and geolocated accident records from several institutions. More standardized reporting and systematic network-wide road-condition surveys could strengthen investment decisions and help governments identify which contracts deliver the best value.

The report cautions that its statistical findings represent strong correlations rather than definitive causal effects, meaning they should not be interpreted as proof that contract type alone produces particular outcomes.

Still, its policy direction is clear. Governments can potentially make road budgets work harder by matching contracts to individual corridors, international development partners can support stronger procurement, safety and data systems, and private firms can develop performance-based infrastructure services. Brazil's experience ultimately shows that cheaper maintenance, better roads and safer travel are separate objectives and achieving all three requires smarter contracts, sustained investment and dedicated safety measures.