Journal Design Engineering Masthead
African Structural Engineering | 10 February 2003

Methodological Evaluation of Process-Control Systems in South Africa

A Difference-in-Differences Model for Adoption Rate Analysis
S, i, p, h, o, K, h, u, m, a, l, o, ,, N, a, l, e, d, i, M, o, l, o, i, ,, P, i, e, t, e, r, v, a, n, d, e, r, M, e, r, w, e
process-control systemsdifference-in-differencesindustrial policytechnology adoption
A difference-in-differences model isolates the causal effect of national policy on technology adoption.
Policy intervention increased PCS adoption probability by 18.2 percentage points (95% CI: 12.4 to 24.0).
Adoption effects were substantially stronger in capital-intensive versus labour-intensive industries.
The study provides a novel methodological framework for evaluating industrial technology programmes.

Abstract

{ "background": "The adoption of advanced process-control systems (PCS) in industrial sectors is a critical driver of productivity and safety. However, rigorous quantitative analysis of the factors influencing their uptake in emerging economies remains limited, with a paucity of robust causal methodologies applied to this domain.", "purpose and objectives": "This study aims to develop and apply a quasi-experimental econometric model to isolate the causal effect of a major national industrial modernisation policy on the adoption rates of PCS within the country's manufacturing and processing sectors.", "methodology": "A difference-in-differences (DiD) model was employed, leveraging a novel, built panel dataset of firm-level technology investments. The core specification is $Y{it} = \\alpha + \\beta (Treati \\times Postt) + \\gammai + \\deltat + \\epsilon{it}$, where $Y_{it}$ is the adoption status. Inference is based on cluster-robust standard errors at the firm level.", "findings": "The policy intervention had a statistically significant positive effect, increasing the probability of PCS adoption by 18.2 percentage points (95% CI: 12.4 to 24.0). This effect was heterogeneous, being substantially stronger in capital-intensive industries compared to labour-intensive ones.", "conclusion": "The national policy was effective in accelerating technological modernisation. The DiD framework proved highly suitable for evaluating such industrial technology programmes, controlling for time-invariant firm heterogeneity and common temporal trends.", "recommendations": "Future industrial policy design should incorporate targeted support mechanisms for labour-intensive sectors where adoption barriers are higher. Evaluations of similar engineering technology initiatives should adopt quasi-experimental methods to establish causal evidence.", "key words": "process-control systems, technology adoption, difference-in-differences, industrial policy, causal inference, manufacturing", "contribution statement": "This paper provides the first application of a difference-in-differences model to analyse the causal impact of a national policy on the