Policy, production, and productivity : spatial dynamics in the South African maize industry during the 20th century

Thesis (PhD)--Stellenbosch University, 2019.ENGLISH SUMMARY : This dissertation investigates the way distortionary agricultural policies, responding to political tensions during South Africa’s structural transformation, also distorted the location of production. Taking the example of maize, it explores the interplay between changes in policy, production and productivity, and changes in the spatial footprint of production. The study covers more than a century (1904–2015), so as to include all the agricultural policy phases typical of an economy undergoing structural transformation. It finds that changes in agricultural policy after the mid-1930s, enacted in response to the political tensions that emanated from the converging and diverging interests of groups within the mining and agricultural sectors, are reflected in changes in maize production and prices. With the implementation of supportive policies, production expanded into areas previously supporting little maize, thereby undermining environment-based comparative advantages of production. Using spatial indexes, the study estimates that at its peak this policy-induced shift in the location of production reduced productivity by between 7.9 and 15.3 per cent. The dismantling of supportive policy during the 1980s and 90s coincided with the removal of land from maize production by farmers. By 2015 the area planted to maize had reverted back to the level it had reached almost 80 years earlier in 1935, before supportive measures were implemented. But spatial inefficiency partly persisted because some production continued in drier, lower-yielding regions. After the distortionary policies were removed, some of the spatial distortion remained, despite the lower productivity, because of region-specific investments in improving plant material, farming practices and infrastructure. So, while some of South Africa’s maize production ended up in the wrong places, technological improvements eventually made the wrong places more right. Despite the misguided policies, drought-focused research-and-development investment in technologies such as hybrid maize generated a sequence of innovations which more than quadrupled the maize yield per unit of rainfall between 1950 and 1993. The South African case serves to show that distortionary policies carry both short- and long-term costs. This is particularly relevant to sub-Saharan Africa, several of whose countries have pursued or are still pursuing various forms of maize and other agricultural intervention. The case also offers an example of successful adaptation to adverse weather conditions and suggests that a change in the location of production can serve as a proxy for climate change. Three new historical datasets for the period were created specifically for this study: maize price, trade and production data; hybrid adoption, replacement and yield trial data; and district-level maize output and area planted data compiled from 17 digitised agricultural censuses, standardised to current spatial boundaries. Although the datasets are limited to maize, the procedures devised to construct them can be used by future researchers to extend the analysis to other crops and regions.AFRIKAANSE OPSOMMING : Hierdie proefskrif ondersoek die manier waarop verwronge landboubeleid, in reaksie op politieke spanning gedurende Suid Afrika se strukturele transformasie, ook die ligging van produksie verwring het. Vanuit die perspektief van mielies, verken dit die wisselwerking tussen beleidsverandering, produksie en produktiwiteit, en veranderinge in die ruimtelike voetspoor van produksie. Die studie strek oor meer as ‘n eeu (1904–2015), sodat dit al die onderskeie landboubeleidsfases insluit wat tipies is aan ‘n ekonomie wat ‘n strukturele transformasie ondergaan. Dit vind dat veranderinge in landboubeleid na die middel-1930s, in reaksie op die politieke spanning wat voortgespruit het uit die gedeelde- and botsende belange van groepe in die myn- en landbou sektore, gereflekteer word in veranderinge in mielie-produksie en pryse. Met die implementering van ondersteunende beleid, het produksie uitgebrei tot gebiede wat voorheen min mielies verbou het, sodoende is die omgewingsgebaseerde vergelykbare voordeel van produksie ondermyn. Ruimtelike indekse word gebruik om te beraam dat hierdie beleidsgeïnduseerde verskuiwing, produktiwiteit met tussen 7.9 en 15.3 persent verlaag het. Die aftakeling van die ondersteunende beleid gedurende die 1980s en 90s strook met die onttrekking van grond van mielieproduksie. Teen 2015 het die aangeplante area teruggekrimp tot ‘n vlak wat dit byna 80 jaar vantevore in 1935 gehandhaaf het, voor die ondersteunende maatreëls. Maar die ruimtelike ondoeltreffendheid het gedeeltelik voortgeduur omdat produksie steeds plaasgevind het in sommige droër gebiede met ‘n laer opbrengspotensiaal. Nadat hierdie beleide verwyder is, het die ruimtelike verwringing gedeeltelik voortbestaan, ten spyte van die laer produktiwiteit, as gevolg van streekspesifieke beleggings in die verbetering van plantmateriaal, verbouingspraktykte en infrastruktuur. So, alhoewel van Suid Afrika se mielieproduksie in die verkeerde plekke beland het, het tegnologiese vooruitgang hierdie verkeerde plekke uiteindelik meer reg gemaak. Ten spyte van onbedagte beleid, het droogte gefokusde navorsing-en-ontwikkeling beleggings in tegnologieë soos batermielies, opeenvolgende innovasies tot gevolg gehad wat mielie opbrengs per eenheid reënval meer as viervoudig verhoog het tussen 1950 en 1993. Die Suid Afrikaanse geval wys dat verwronge beleid beide kort- en langtermyn kostes dra. Dit is besonder relevant tot sub-Sahara Afrika waar verskeie lande histories of tans verskillende vorme van mielie en ander bemarkingsintervensies voortsit. Die Suid Afrikaanse geval bied ook ‘n voorbeeld van ‘n suksesvolle aanpassing tot ongunstige klimaatsomstandighede and stel voor dat ‘n verskuiwing in die ligging van produksie kan dien ‘n proksie vir klimaatsverandering. Drie historiese datastelle vir die periode is spesifiek vir hierdie studie ontwikkel: mielieprys, handel en produksie-data, bastermielie aanneming-, vervanging- en proef opbrengs-data, and distriksvlak mielieproduksie en geplante area data saamgestel uit 17 gedigitaliseerde landbousensusse, gestandaardiseer tot huidige distriksgrense. Alhoewel die datastelle beperk is tot mielies, sal die prosedures wat ontwikkel is om hulle daar te stel ook deur ander navorsers toegepas kan word om die analise uit te brei na ander gewasse en streke