Private investment is a smaller share of the economy today than in 1994
In 1996 the economist Nicoli Nattrass showed that the business camp and the labour camp were betting on the same unknown: what makes firms invest. One side said confidence, the other said demand. Neither delivered the investment it promised.
Fixed investment as % of GDP, nominal. Stats SA/SARB national accounts via Econdata, pulled October 2026.
Demand and confidence both matter. Interest rates don't add much.
What predicts next year's private investment growth, 1995–2025? Each bar shows the effect of a typical (one standard deviation) move in last year's conditions. "Confidence" is business confidence net of what growth already explains.
OLS on 31 annual observations, Newey–West standard errors, R² = 0.47. Grey = not statistically significant. A predictive association, not a causal estimate.
Both 1996 camps were half right, which is why the argument never ends.
Brazil invests about as little as we do, yet far more of its adults work
Low investment explains slow growth. It doesn't explain why fewer than four in ten South African adults have a job.
Averages for 2010–2024. World Bank WDI: total fixed investment (% of GDP) and employment-to-population ratio, 15+ (ILO modelled estimates).
Thirty minutes, one argument, about a dozen charts
This week's argument
Why a row over US–SA relations reopened the oldest question about the transition.
1996: two bets
Nattrass's reading of GEAR and its labour-movement rival: both were gambling on investment.
The scorecard
What GEAR hit and missed, and thirty years of growth, investment and jobs.
Testing A and B
How much we redistributed, what confidence did, and a simple test of both channels.
The next bet
Why neither bet paid out, and what the GNU has to get right this time.