RAR
Resilience-Adjusted Redundancy
SRI crossed with fragility: 31 countries show adaptive resilience, 26 wasted redundancy.
OSE reads entrepreneurial ecosystems as cultivated orders: the Cultivated Order Theory (COTEE) frames the meta-organization measures, the Aligned Gender Index and the AEEI. Every figure on this page is recomputed live on 60,600+ qualified actors across 150+ countries.
Level 1 sets the law of combination; Level 2 measures the architecture; Level 3 will connect both to what happens to entrepreneurs.
The ecosystem is cultivated: the seed (actors present), the soil (a country's capacity to enforce its rules) and the gardener (deliberately built support). Level 1 sets the reading rule for everything below it.
Integrative · liveSubordinate to COTEE. Measures the architecture (CI, FDI, turnover, density) and the orchestration (VRIO, SRI, EES) of the organizations of organizations that make up an ecosystem.
Subordinate · liveNot yet instrumentable: it requires venture-level data (who is affiliated with which structures, and what happens next). Planned as the outcomes layer of the framework.
PlannedThe support apparatus only pays off where rules are binding — inert as a main effect, powerful as a complement of the institutional regime.
Everything that is there: finance, talent, knowledge, the organizations counted in a registry. Necessary — never sufficient on its own.
A country's capacity to make its own rules bind (rule enforcement, World Governance Indicators 2023). The soil decides whether the seed yields anything.
Incubators, accelerators, programmes, agencies — the apparatus. A multiplier where rules bind; close to inert as a main effect where they do not.
Cross the soil (rule enforcement) with the gardener (apparatus builder vs endowment-based) and the four regimes appear. The same apparatus predicts 1.3 new firms per 1,000 adults where rules are weak and 6.4 where they bind. Replicated on 111 countries (WGI 2023 + World Bank outcomes).
α = 0.001 between components: COTEE yields a profile, not a score. Adding the layers destroys the information.
Every reading is conditional on the enforcement regime. The same apparatus means different things in different soils.
The national figure is an aggregate of urban modules. Diagnose and act at the module, not the envelope.
Table 14 of the theory, recomputed on OSE data: typology 43 apparatus-builders · 68 endowment-based (paper 58 / 55; bootstrap stability ARI 0.631); cultivation interaction β = 0.23 (p 0.076; paper 0.396); effect on FDI +0.60 (p 0.051); no effect on patents / high-tech exports.
Rule enforcement is a necessary condition for venture creation: no ecosystem creates at scale where rules do not bind.
Result · Necessity consistency 0.854 (threshold 0.80), coverage 0.775.
Replicated (fsQCA necessity)
The relevant unit of an ecosystem is the urban module, not the national envelope.
Result · 68.4 % of composition variance is within countries (paper: 72 %).
Replicated (variance decomposition)
The effect of interface legibility (portals, one-stop shops) inverts with enforcement: helpful where rules are weak, harmful where they are applied.
Result · Press panel, 19 countries, 2014–2023, 154 country-years: +0.12 (n.s.) under weak rules, −0.46 (significant) under enforced rules; interaction β −0.341 (t −3.82).
Replicated (longitudinal press panel)
Why H5 matters for policy. Portals and one-stop shops make an ecosystem more legible. Where rules are weak, that legibility is harmless (+0.12, not significant). Where rules are applied it becomes a cost (-0.46, significant): the interface overlays what already works.
Hence the design rule — fund pathways, not portals, and put sunset clauses on one-stop shops.
Enforcement first: revocable accreditation, rules that bind. Contents poured on weak soil stay inert.
The urban basin, not the national envelope: 68.4 % of composition variance sits inside countries.
Fund pathways, not portals. Put sunset clauses on one-stop shops — legibility without enforcement inverts.
Lock in the ties and structures that proved themselves — turn cultivated order into rules.
Orchestrating is not conducting the orchestra — it is tuning the room.
Enforcement -0.20 (rules weakly applied), built support 31 %, base endowment 63 %, 21 % of actors created by the State or donors, 8 urban modules. Expected creation 1.5 → 4 per 1,000 adults if rules were enforced — the soil, not the seed, is the binding constraint. Morocco report →
Organizations of organizations: authority without hierarchy, membership boundaries, and the orchestration of complementarities. Ecosystem advantage is a governance achievement, not a resource endowment.
Level 2 is read inside COTEE. Its first tier measures the architecture and governance of the meta-organization (coverage, differentiation, turnover, density); the second measures resource orchestration (VRIO coverage, redundancy, effectiveness). A third tier — the entrepreneur interface (ACI, cross-affiliation) — awaits venture-level data. All values below are computed live on the OSE base.
| Code | Measure | Definition | World value |
|---|---|---|---|
| CI | Coverage Index L1 · Architecture & governance | Share of the 14 canonical boosts covered by at least one actor. | World 100 % |
| FDI | Functional Differentiation Index L1 · Architecture & governance | Normalised Shannon entropy of the boost mix — how evenly functions are distributed. | World 0.87 · Gulf 0.79 · Oceania 0.93 |
| MTR | Membership Turnover L1 · Architecture & governance | Founding cohorts of member organizations — renewal of the meta-organization. | Peak cohort 2015 · 19.9 % founded since 2020 |
| D | Network density L1 · Architecture & governance | Density of the top-30 co-occurrence network between actors. | Qatar 0.146 vs Saudi Arabia 0.063 |
| VRIO | VRIO coverage L2 · Resource orchestration | Share of the 5 OSE layers where the ecosystem holds valuable, rare, inimitable and organised resources. | Computed per country on 5 layers |
| SRI | Service Redundancy Index L2 · Resource orchestration | Average number of actors delivering the same service — duplication of the offer. | World 2.01 · Africa 2.08 · Americas 2.83 |
| EES | Ecosystem Effectiveness Score L2 · Resource orchestration | mean(CI, 1 − SRI, VRIO, FDI), each term normalised. | World 0.72 |
| RAR | Resilience-Adjusted Redundancy Reading · Architecture × Substance | SRI × fragility — is redundancy a buffer or a waste? | 31 countries adaptive resilience · 26 wasted redundancy |
| P1 | Governance vs Endowment Reading · Architecture × Substance | Ecosystem advantage as a governance achievement rather than an endowment. | R² 0.20 · 63 countries above the expected line |
| P3 | Form vs Substance Reading · Architecture × Substance | Countries whose architecture runs ahead of their substance. | 93 of 115 countries (81 %) |
| TMI | Three-Mechanism Index Reading · Architecture × Substance | Weight of the normative, resource-mediated and platform coordination mechanisms. | Normative 36 % · Resource 33 % · Platform 31 % |
Crossing what an ecosystem is built like with what it actually delivers.
SRI crossed with fragility: 31 countries show adaptive resilience, 26 wasted redundancy.
Ecosystem advantage is a governance achievement, not an endowment — R² 0.20, 63 countries above the expected line.
93 countries out of 115 (81 %) have their architecture ahead of their substance.
Normative 36 % · resource-mediated 33 % · platform-based 31 %.
Membership turnover by founding year — peak in 2015, 19.9 % of actors founded since 2020.
Strongest bundle: L2 Cognitive ↔ L3 Relational. Association rules: Test→Proto (×20.7), Sales→Market (×11.5), Team→Mentor (×9.0).
Every actor is qualified by the boosts it actually delivers, sorted into five layers from technical productization to psychological support. Layers are a profile — never summed.
An equality index aligned with the Africa Gender Index (AfDB / UNECA), computed from World Bank Gender Statistics — female labour-force participation, Women, Business & the Law, female tertiary enrolment, parliamentary seats, on three dimensions — economic, social, representation — and crossed with the women-focused structures of the OSE registry.
Crossed with the women-focused structures of the OSE registry (‰ of actors). The correlation between the AGI and OSE density is weak (ρ ≈ +0.13): gender equality and support density are two different questions, which is why Expert reports them side by side rather than merged. Morocco, for instance: AGI 50/100, rank 21 / 53 in Africa, 112 / 176 globally; 2 women-focused structures out of 604 (3.3 ‰).
The Africa Entrepreneurial Ecosystem Index (AEEI) applies the Stam & van de Ven (2021) framework to 54 African countries. OSE interfaces it live and re-reads each pillar through COTEE: contents only pay off where the soil holds.
| AEEI pillar | Stam elements | COTEE layer |
|---|---|---|
| Formal institutions | Formal institutions | Soil |
| Culture | Entrepreneurial culture | Soil |
| Finance | Finance | Seed |
| Talent & knowledge | Talent · Knowledge | Seed |
| Networks & intermediaries | Networks · Intermediaries | Gardener |
| Physical infrastructure | Physical infrastructure | Environment |
| Demand | Demand | Environment |
| Leadership | Not a quantity but a mode (gardener vs architect): not measured by the AEEI, captured by COTEE through the apparatus × enforcement interaction (H3). | |
54 African countries covered, 30 ranked with a composite score, 20 underlying indicators. Source: africa.ecosystem.build.
AEEI pillars are sorted into Soil, Seed, Gardener and Environment, then read conditionally on enforcement. The composite score is useful to rank; COTEE warns that the components do not correlate (H1) — read it as a profile.
Advanced ~13 (incl. MA, TN, ZA, KE) · intermediate ~22 · emerging ~19. Available on every Expert country profile, benchmark and insight.
Composite 3.47 / 7, rank 3 / 30, percentile 7. Strengths: talent & knowledge (0.77), infrastructure (0.68), finance (0.39, +0.28 vs mean). Weaknesses: networks & intermediaries (0.18), demand (0.26).
Reference: Stam, E. et al. (2025). Measuring National Entrepreneurial Ecosystems in Africa. SSRN, https://doi.org/10.2139/ssrn.5254687
Each hypothesis of the Cultivated Order Theory is recomputed on the OSE base. Status is updated as the base grows; the full programme is on the research page.
The components of an ecosystem do not correlate and no single configuration is sufficient — an ecosystem must be read as a conditional profile.
Result · α = 0.001 between components on the OSE base; typology replicated (43 apparatus-builders · 68 endowment-based; paper 58 / 55; bootstrap stability ARI 0.631).
Replicated on 111 countries
Rule enforcement is a necessary condition for venture creation: no ecosystem creates at scale where rules do not bind.
Result · Necessity consistency 0.854 (threshold 0.80), coverage 0.775.
Replicated (fsQCA necessity)
Deliberate apparatus multiplies venture creation where rules bind and stays inert where they do not (apparatus × enforcement).
Result · Interaction β = 0.23 (p 0.076) on the OSE base vs 0.396 in the paper; predicted creation 1.3 → 6.4 (apparatus builders), 1.5 → 4.0 (endowment-based). Effect on FDI +0.60 (p 0.051); no effect on patents or high-tech exports.
Direction replicated · significance in progress
The relevant unit of an ecosystem is the urban module, not the national envelope.
Result · 68.4 % of composition variance is within countries (paper: 72 %).
Replicated (variance decomposition)
The effect of interface legibility (portals, one-stop shops) inverts with enforcement: helpful where rules are weak, harmful where they are applied.
Result · Press panel, 19 countries, 2014–2023, 154 country-years: +0.12 (n.s.) under weak rules, −0.46 (significant) under enforced rules; interaction β −0.341 (t −3.82).
Replicated (longitudinal press panel)
Each measure on Expert traces back to one of these papers. Working drafts are shared on request.
Guéneau, G. et al. · Journal of Business Venturing Insights
Ecosystem advantage is a governance achievement, not a resource endowment. Three coordination mechanisms (normative, resource-mediated, platform-based) and a multi-level architecture; longitudinal evidence from seven MENA ecosystems.
Guéneau, G. · Working paper — H5 replicated on a longitudinal press panel (19 countries, 2014–2023)
The integrative Level 1 framework. Deliberate support pays off only where rules bind; the module is urban; interface legibility inverts with enforcement (H5).
Guéneau, G., Rouabeh, L., Chabaud, D. · Academy of Management Proceedings
The three-scale analysis framework: resource configurations across organizations, VRIO-consistent bundles, and the measures that Expert now computes live on the OSE base.
Guéneau, G., Chabaud, D., Chalus-Sauvannet, M.-C. · Technological Forecasting & Social Change
Longitudinal network analysis across four territories: the best-performing ecosystems rely on strong, persistent ties with practical support actors.
Guéneau, G., Chalus-Sauvannet, M.-C. · Management & Prospective (Gestion 2000)
Structure and process of a regional ecosystem co-evolve — neither can be assessed in isolation.
Guéneau, G. (sup. Chabaud, D.) · Université Paris 1 Panthéon-Sorbonne — doctoral dissertation
The founding measurement framework of OSE: graph theory, web-scale data collection and configurational analysis in African contexts.
14 observed indicators per actor, continuous requalification and 10+ years of history — the conditions under which a score can be defended in front of a scientific council.
14 observed quality indicators per actor; no self-reported surveys.
Every actor is re-checked by the pipeline: defunct programmes, ghost incubators and duplicates are removed, not archived as live.
10+ years of data — before/after readings of policies and programmes, not snapshots.
Louvain communities, centralities and co-occurrence networks: who works with whom, not only who exists.
60,694 qualified actors, 153 countries, read with the same layers, boosts and rules.
A public API (api.ose.africa) and CSV/PNG exports on Expert let any reviewer recompute what we publish.
Definitions of every term used here are in the glossary.
The framework, its measures and each replication are challenged by an independent scientific council.
Every measure on this page is recomputed on Expert — pick a country, read its cultivation regime, compare it with up to six others.
Curated insights from our 117 ecosystems — new country reports, ranking shifts, methodology notes. No fluff, no spam, unsubscribe in 1 click.
By subscribing you agree to our privacy policy. We never share your data.
The observatory. Data, diagnostics and decision tools for the builders of entrepreneurial ecosystems.
The academy. Capacity building deployed where it's needed — online, on-site, in ecosystem residency.
The research society. The scholars advancing the science of entrepreneurial ecosystems.