The effects of information communication technology policy alternatives on South Africa's agro-processing industries

Abstract

Since 1994, the South African Government has developed various policy plans to ensure South Africa’s economic growth and development. The agro-processing subsector has been earmarked in the policy plans as one of the sectors with the potential to achieve South Africa’s economic growth and development. Despite this, statistics show that the subsector has been ineffective in driving the required growth and development. Against this backdrop, the study aimed to examine the contribution of ICT investment to the growth of the agro-processing subsector. Four objectives were identified as follows: (1) to examine whether ICT policies contributed to the labour productivity growth of the agro-processing industries; (2) to estimate effects of ICT intensity on the growth of labour productivity, output and employment; (3) to examine the relationship between ICT intensity and the growth of labour productivity, output and employment; and (4) to forecast the potential effects of ICT intensity on growth of labour productivity, output and employment.

The ICT intensity index was applied to rank 10 agro-processing industries into two groups: “more ICT-intensive industries” and “less ICT-intensive industries”. Thereafter, the annual growth rates of labour productivity, output and employment for all the industries were calculated. Four econometric techniques were applied to achieve the objectives of the study as follows. The Difference-in-Differences (DD) technique was used to achieve the first objective. The Pooled Mean Group (PMG) estimations were performed to achieve the second objective. The Toda and Yamamoto (TY) Granger non-causality tests were applied to achieve the third objective. The Impulse Response Function (IRF) and Variance Decomposition (VDC) analyses were conducted to achieve the fourth objective.

The findings from the ICT intensity index indicated that the more ICT-intensive industries (i.e. food, beverages, textile, paper and rubber) accounted for 78% of ICT investment, while the less ICT-intensive industries (i.e. tobacco, wearing apparel, leather, wood and furniture) accounted for the remaining 22%. In the case of individual industries, the food industry accounted for the largest share of ICT investment (37.20%), while the tobacco industry accounted for the smallest share (0.84%). This implies that the food industry has invested the most in ICT, whereas the tobacco industry has invested the least.

The DD findings showed that the more ICT-intensive industries experienced a slightly higher acceleration (i.e. increase overtime) in labour productivity growth than the less ICT-intensive industries. However, the DD estimator was insignificant, which means that the difference in labour productivity between the two groups cannot be attributed to ICT use. The policy implication is that there was no evidence to support that ICT policies contributed to the labour growth of industries.

The PMG findings indicated that ICT intensity had no significant effect on the growth of the aggregated industries. These findings conform to studies that found zero and negative significant effects of ICT when industries were aggregated. This implies that the lower growth contributions of the industries that invested less in ICT outweighed the relatively higher growth contributions of the industries that invested more ICT.

The findings further showed that positive and significant effects were notable only in the short run for the more ICT-intensive group. Furthermore, whereas in the long run ICT intensity yielded positive and significant effects on the output growth of both the less and more ICT-intensive industry groups, its effect was higher for the more ICT-intensive group. The findings conform to previous studies that found higher effects for the more ICT-intensive industries. This implies that the impact of ICT on agro-processing industries varied per group of industries, such that an industry group that invested more in ICT (i.e. more ICT-intensive industries) experienced higher growth than those that invested less (i.e. less ICT-intensive industries). The findings for individual industries revealed that ICT intensity contributed more to the growth of an industry that invested more in ICT (i.e. food industry) and less to the growth of an industry that invested less (i.e. tobacco industry).

The findings from the TY Granger non-causality tests showed that there was no evidence of causality for the less ICT-intensive industries. In contrast, the results showed that there was evidence of a causal relationship for the more ICT-intensive industries. This implies that causal effects occur in line with ICT intensity given that evidence of causality was evident for the industry group that invested more in ICT. The findings for individual industries showed that there was evidence of causality for the food industry. This implies that evidence of causality was notable for an industry that invested more in ICT.

The IRF findings showed that, in the long run, ICT intensity would impact positively on the growth of labour productivity, output and employment of both the less and more ICT-intensive industries. This finding varied from the TY test in which there was no causal relationship between ICT intensity and growth for the less ICT-intensive industries. Therefore, the fact that positive effects for the less ICT-intensive industries were only detected in the long run implies that the returns on ICT investment for the less ICT-intensive industries would be notable over a long period. However, the VDC results, which captured the magnitude of the effect, showed that, while ICT intensity would contribute to the growth of all industry groups, its contribution would be higher for the more ICT-intensive industry group. The findings for individual industries indicated that, while ICT intensity would contribute to the growth of all agro-processing industries, it would contribute more to the growth of the industry that invested more in ICT (i.e. food industry). In contrast, it would contribute less to the growth of an industry that invested less in ICT (i.e. tobacco industry).

Two main findings have been derived from this study for policy decision making. Firstly, the study found that the existing ICT policies have not yet contributed to the labour productivity growth of the agro-processing industries. This is attributable to the observation that the existing policies do not currently focus on ensuring access to and usage of ICT by other non-ICT industries, including the agro-processing industries.

Secondly, the study found that, whereas ICT investment would contribute to the growth of all agro-processing industries, in the long run, it would contribute more to the growth of the food industry, which invested more in ICT, and less to the growth of the tobacco industry, which invested less in ICT. Against this backdrop, the key drivers that could have restricted ICT use by the tobacco industry were identified as (a) value-chain information intensity, (b) size of the industry and (c) policy environment. In particular, the tobacco industry invested less in ICT and consequently realised low growth because it is low-value chain information-intensive and is comprised of fewer firms. In addition, lower ICT investment and growth is attributable to policies that prohibit the advertisement and promotion of tobacco products, and the distribution and sale of tobacco products through postal services, the internet, or any electronic media.

This study recommends the integration of ICT into the growth and development policy plans for agro-processing. This can be achieved through the inclusion of skills development and ICT infrastructure development in the policy plans for the effective use of ICT. Where private firms invest in skills development and ICT infrastructure, it is recommended that the government should compensate such firms through the proposed ICT Tax Incentive Programme. The aim of the proposed ICT Tax Incentive Programme should be to stimulate investment in ICT given that the present study has established that the more an industry invests in ICT, the higher the growth in labour productivity, output and employment. The proposed programme should be implemented over a longer period, in line with the study’s finding that the returns on ICT investment take time to materialise. It is further recommended that priority should be given to the more ICT-intensive industries as the returns on investment would be higher.