Türkiye's social assistance system scores GL = 0.067 — Structural Failure.
For every 100 TL of social protection intent, only 6.7 TL of value reaches the citizen as timely, usable access to essentials. The remainder is consumed by administrative friction: waiting, redundancy, collapse, information asymmetry, psychological cost, and opportunity cost.
The system is not designed to fail. It is designed to filter. Filtering produces friction. Friction shifts cost to those least able to bear it.
| Variable | Score | Definition |
|---|---|---|
| Fs — Flow Success Rate | 0.12 | Proportion of social protection budget reaching citizens as usable access within reasonable time |
| Vn — Strategic Value | 1.5 | Social protection as critical economic stability mechanism (maximum necessity) |
| Pd — Pain Duration | 1.8× | Years of waiting, repeated applications, uncertainty borne by citizens |
| Cf — Cognitive Friction | 1.5× | Complexity of navigating 43 programs, different eligibility rules, officer discretion |
| System | GL Score | Key Characteristic |
|---|---|---|
| Estonia Digital Government | 4.20 | Pre‑loaded eligibility; automatic disbursement |
| Denmark e-Health | 3.45 | Integrated health and social data |
| Türkiye SGK (current) | 0.067 | 43 programs; manual application; 1‑12 month waits |
| UK Universal Credit (2010s) | 0.074 | Same pattern: infrastructure modernized, denominator ignored |
Estonia's GL is 63× Türkiye's current performance. Denominator‑first design works.
| Friction Source | Leverage | Reform Pathway |
|---|---|---|
| 43 fragmented programs — citizen must guess | HIGHEST | Single eligibility engine: AI determines qualification automatically. |
| Document burden (lease, bank statement) | HIGHEST | Data‑first eligibility: tax, employment, health data auto‑pull. No documents. |
| Officer discretion — outcome depends on who you talk to | HIGH | Rule‑based auto‑approval. Eligibility coded into system. |
| 1‑12 month wait — citizen waits while system processes | HIGH | Zero‑wait design: Access granted immediately, post‑hoc AI audit. |
| Recertification every 3‑12 months | MEDIUM | Continuous eligibility: system auto‑updates from live data. |
| No appeal — 85% of rejected eligible do not appeal | MEDIUM | No rejection. No appeal. Auto‑approval based on data. |
| Current System Deliverables (Infrastructure Layer) | GL Diagnostic (Outcome Layer) |
|---|---|
| Kızılaykart cards issued to 1.7M beneficiaries | Citizen waits 1‑12 months before receiving card |
| e-Devlet portal with 9,200 services | Citizen does not know which of 43 programs apply |
| Social Risk Map with 648 indicators | Data exists but does not trigger automatic disbursement |
| 43 different aid programs | Citizen must navigate fragmented rules, no single entry |
| ESSN cash transfers | Cash can be withdrawn and spent on non‑essentials |
| Scenario | Intervention | GL Score | Gain vs Current |
|---|---|---|---|
| Current | Manual application, officer review, 1‑12 month wait | 0.067 | — |
| A | Data‑first eligibility (auto‑ID, no documents) | 0.12 | +79% |
| B (Recommended) | Scenario A + zero‑wait + auto‑approval | 0.48 | +616% |
| C | Scenario B + full cross‑agency data integration, single engine | 1.28 | +1,810% |
Scenario B requires no new technology. It requires denominator redesign: measuring citizen access before payment, not after.
| # | Recommendation | Target Variable | Expected Impact |
|---|---|---|---|
| 1 | Data‑first eligibility – auto‑identify from tax/employment/health data. No application forms. | Drop‑out rate → 0 | Prevents collapse at document stage |
| 2 | Zero‑wait design – grant Access immediately (food, healthcare, transit). Post‑hoc AI audit. | Wait time → 0 | Eliminates waiting tax |
| 3 | Single eligibility engine – AI determines qualification, citizen only receives notification. | Info asymmetry → 0 | Eliminates broker culture |
| 4 | Purpose‑locked Access – QR codes locked to supermarkets, pharmacies, transit, childcare. No cash conversion. | Resource waste → 0 | Eliminates leakage and secondary markets |
| 5 | Annual reset – Access resets to zero each year. No accumulation. No permanent liability. | Psychological cost → 0 | Government controls risk; citizen cannot hoard |
| 6 | Real‑time GL dashboard – published quarterly for all social assistance programs above 1B TL. Accessible to Parliament, Court of Accounts. | Fs = continuous governance | Converts GL from diagnostic to accountability instrument |
We do not ask Türkiye to redesign its entire system overnight. We ask for one pilot: one province or city. 10 administrative processes. 14 days.
No system access needed. No upfront cost. No commitment.
After 14 days, Türkiye will see where friction is highest, who bears the cost, and the first step to remove it. Then decide: pilot, scale, or walk away.
The current system asks citizens to pay the cost of government coordination. GL + AI flips this: the government pays the coordination cost in the background. The citizen only receives and uses.
Türkiye's traffic data shows your institutions are already downloading. They are already studying. They are already preparing.
This diagnostic is not a theory. It is a working blueprint, ready for a 14‑day pilot.