Surprisingly small share of headline Chinese project spending typically stays circulating in African economies. Quantitative estimates vary by project and sector, but most credible studies and case-work put the local “value retained” in many Chinese-financed infrastructure deals in the range of roughly 20–40% — meaning 60–80% of the money effectively flows back to Chinese suppliers, contractors, banks or leaves as repatriated profits.

That’s a blunt summary; below I explain why the share is low, why estimates vary, and where the money goes — with concrete evidence and examples. 

Why so little stays local (the mechanics)

1. Tied financing and tied procurement. A large fraction of Chinese loans are “tied”: the financing is conditioned implicitly or explicitly on using Chinese contractors, equipment and inputs. That means the loan disbursements often return to Chinese firms that supply materials, machinery, and labor. AidData’s mapping of Chinese contracts finds many such contractual clauses (special accounts, procurement conditions) that effectively route cash back to China. 

2. State banks and Chinese supply chains. Most project finance comes from Chinese policy banks (China Exim, China Development Bank). Those banks lend to governments or to special purpose vehicles that then hire Chinese SOEs. Because the supply chain (steel, heavy machinery, engineering services) is largely Chinese, much of the contract value is spent on goods and services produced in China — not locally. World Bank analyses of China-financed projects in Africa document this pattern repeatedly. 

3. Use of Chinese labor and expatriate management. For complex infrastructure, Chinese firms frequently bring in Chinese technical teams and sometimes large numbers of Chinese workers for skilled roles. That reduces wage and skill-transfer benefits to local labor markets and increases repatriation of salaries and per diems. Case studies show local labor often fills lower-paid roles while higher-value consulting/technical work stays foreign. 

4. Resource-backed and collateralized deals. In many resource-backed loans, repayment is structured in oil, minerals or future commodity flows rather than cash retained in country. That structurally funnels value away from local economies and ties repayment to exports. The World Bank’s work on resource-backed loans documents how repayments and offtake arrangements reduce local capture of value. 

Evidence and ballpark figures

• AidData/academic synthesis: AidData’s project-level work and briefs on “how China lends” show that a large share of loan dollars remain linked to Chinese banks and contractors through contractual mechanisms (special accounts, tied purchasing). While they do not publish a single continent-wide “percent retained” number, their evidence implies substantial repatriation from large projects. 

• World Bank project assessments: The Bank’s reviews of China-backed infrastructure in Africa have repeatedly found that local procurement content is often low, with indirect costs and foreign inputs dominating large projects. One earlier World Bank synthesis noted that in strong export-performing economies indirect costs absorb only 10–12% of production costs; in many African projects indirect costs (infrastructure-related) can be 20–30% — showing how infrastructure investment patterns and procurement choices affect local value capture. Extrapolating from dozens of case studies, many analysts use a conservative 20–40% local retention assumption for China-built megaprojects. 

• Sectoral variation: Mining and energy projects financed by China tend to keep less value locally (because of equipment imports and commodity export structures). Urban infrastructure (local roads, small bridges) can keep more value locally if local firms and labor are used. AidData and SAIS-CARI case work show this heterogeneity. 

Concrete examples that illustrate the pattern

• Large rail/highway projects (e.g., Kenya’s Standard Gauge Railway): construction contracts were awarded to Chinese SOEs, large segments of procurement and equipment came from China, and operation/maintenance arrangements kept revenues and technical control tied to Chinese firms — reducing net local retention. This is a representative template for many megaprojects. (See Reuters and project case reporting for specifics.) 

• Resource-backed deals (various Angolan, DRC, and others): loans repaid in oil, minerals, or via offtake agreements make the “staying” share especially small because repayment itself is extraction, not domestic investment. The World Bank’s resource-backed loans report discusses how repayment structures limit local capture. 

What “percentage stays local” depends on (key drivers)

1. Procurement rules: If host governments insist on open procurement and local content requirements, more of the cash can stay. If contracts are negotiated bilaterally and procurement is effectively closed, less stays. 

2. Project scale and complexity: Small, labor-intensive civil works retain more local wages and small-firm contracts. Mega technical works (dams, railways, power plants) use expensive imported kit and expertise. 

3. Use of Chinese vs local suppliers: The proportion of materials and machinery sourced locally vs imported is decisive. Where steel, cement, or equipment must be imported, retention drops. 

4. Contract structure (loans vs equity vs joint ventures): Equity or true FDI that builds local firms can create higher retention than off-balance-sheet loans that require China-supplied inputs. AidData documents that many projects are effectively loans tied to Chinese supply chains. 

Caveats and nuance

• No single continent-wide number: There is no single authoritative percent that applies to every Chinese project in Africa — estimates depend on methodology, project type, and data availability. But the consistent theme across SAIS-CARI, AidData, World Bank, IMF and peer research is that a majority of dollar value in large China-funded infrastructure projects tends to flow back to China, not remain as local retained value. 

• Some projects do deliver strong local impact. Not all China-backed projects are identical; health clinics, smaller roads, and some PPPs have higher local content and genuine spillovers. The pattern is heterogeneous. 

Policy implications — what African governments can do to keep more value

1. Demand local-content clauses and enforce them (training, procurement, materials sourcing).

2. Publish contracts and run open tenders so local firms can compete.

3. Structure deals with equity or joint-venture elements that give local partners a real ownership stake, not just debt obligations.

4. Negotiate technology transfer and operation/maintenance clauses that build domestic capacity rather than lock in foreign operators. 

                        +++++++++++++++++++++++++++++++

There’s no mystery: much of the headline value of Chinese-financed projects tends to flow back to Chinese banks, contractors and suppliers, especially for large, technical, tied contracts. Reasonable, evidence-based ballpark estimates—acknowledging variation—place local retention often in the 20–40% range for major projects, with the remainder effectively returning to China through procurements, repatriated profits, and resource-backed repayment mechanisms. If African policymakers want infrastructure that genuinely seeds domestic industry and jobs, they must insist on contract transparency, enforceable local-content rules, and deal structures that prioritize local ownership and value-addition.