Skip to main content
SA Policy Space

SAFTU

LabourEngagement posture: ConcernedInfluence 6/10
Pkg 1: Infrastructure UnblockPkg 2: SMME & Employment AccelerationPkg 5: State Capacity & Governance

How these positions were derived

Every stance on this page is AI-derived from parliamentary and public records — committee transcripts, submissions, and published statements — and each carries the one-line reason the model gave for it. They are a research starting point, not a position an organisation has confirmed. Verification is tiered. 235 of 451 stance records have been verified, and 37 corrected, by a paired AI review — two independent AI reviewers per actor plus an adjudicator — that checked each stance for coherence against the actor’s recorded brief and the idea text. That check did not reach back to primary parliamentary sources, so an AI-verified badge means internally consistent, not independently confirmed. A corrected row shows the revised position, with the originally recorded stance and the adjudicator’s reason inspectable on its badge. The distinct top tier is human review: 0 of 451 stance records across the whole site have been checked by a human reviewer, and a human review always outranks the AI tiers. None have yet. Human-reviewed rows get their own badge as soon as they exist; the 179 rows with neither badge are the model’s unchecked call.

Stakeholder brief

Primary interests

Anti-austerity fiscal policy; nationalisation of strategic sectors; worker and community ownership of public enterprises; opposing structural adjustment that SAFTU argues deepens inequality. Represents ~800,000 workers primarily through NUMSA and other progressive unions.

Key concerns

Structural reform agenda reflects IMF and World Bank conditionality that prioritises fiscal consolidation over poverty reduction; privatisation and concessioning transfers public wealth to private capital without distributional benefit; water and energy reforms raise prices for poor households; weakening of labour protections through 'flexibility' reforms.

What they bring

Grassroots knowledge of how structural reform affects low-income households and informal workers; mobilisation capacity that if engaged could legitimise reform; analysis of distributional impacts that should inform reform design; community-level implementation intelligence.

Conditions for engagement

Ownership models that include worker cooperatives, community benefit sharing, and significant state stakes; robust price regulation protecting low-income households in reformed energy and water sectors; democratic accountability for reform decisions; redistribution mechanisms funded from efficiency gains.

Reform design insight

SAFTU's concerns about distributional impacts are a legitimate policy design input even where preferred policy solutions differ. Reform designs that demonstrably improve access and affordability for low-income households — through lifeline tariffs, cross-subsidies, and community benefit sharing — are more durable politically and constitutionally than those that optimise for efficiency metrics alone.

Recorded positions

10 positions across 1 reform package. Opposed more often than not.

4

Support

0

Conditional

6

Oppose

0

Neutral

Two scales are in play. The engagement posture above (Concerned) is one actor-level judgement, the axis the feasibility simulator scores. These four counts are the separate per-idea scale: support reads as champion, conditional reads as constructive critic, oppose reads as concerned, neutral has no posture counterpart at all — the posture scale has no neutral. An actor’s posture and its per-idea positions can disagree, and where they do the disagreement is the point.

Stance portfolio

Grouped by reform package, then by the binding constraint each idea addresses. The line under every title is the reason the stance was assigned.

Package 2: SMME & Employment Acceleration

10 ideas4 support6 oppose
Labour Market10

Agreement & opposition

Scored only on ideas both actors hold a recorded position on. Identical positions score +1, support against opposition −1, support against conditional +0.5, and any pair involving a neutral 0. The figure shown is the mean over the shared ideas. Pairs sharing fewer than 5 ideas are not scored — a one-idea overlap produces a perfect agreement that means nothing.

Most aligned

  • COSATU+0.05Little in common10 shared ideas of 10 · 4 agree · 6 clash
  • National Treasury-0.10Little in common10 shared ideas of 10 · 4 agree · 6 clash

Most opposed

Data as of 2026-10-10 · latest PMG meeting 2026-10-02