Essays on the drivers and inclusive growth effects of structural change patterns in Africa

Date
2023-12-14
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH SUMMARY: As economies develop, they undergo large scale structural change – the reallocation of economic activities across three broad sectors of agriculture, manufacturing and services that accompany economic development. A fundamental attribute of structural change is the structural heterogeneity within and across countries, giving rise to cross-country variation in structural change patterns. Beyond the systematic variation in structural composition, structural transformation also induces a wide range of complementary processes and changes in technological progress, consumer behaviour, urbanisation, demographic transition, female labour participation, living standards, welfare redistribution, and even socio-political institutions. Although structural transformation has been extensively investigated in several industrialised economies and remains a pertinent policy issue amongst developing economies, empirical insights into these processes within developing economies remain unclear. The collection of four stand-alone essays in this dissertation examines the drivers and inclusive growth effects of African structural change patterns towards providing a deeper conceptual understanding of policy implications. The first essay reconsiders the stylised facts of structural change within the context of Africa. Specifically, the study summarizes broad trends and patterns in the data to understand the empirical regularities in the composition of value-added and employment shares, relative labour productivity levels and growth rates, and capital intensity levels using descriptive statistics of 20 (38) African economies from 1960 to 2018 (1970-2017). The study found considerable sectoral differences, emphasising structural disequilibrium in Africa's economic growth pattern. The following empirical insights were derived from the analyses. First, any talk of premature de-industrialisation in Africa is premature, as the manufacturing sector is still expanding, albeit more slowly than expected. Second, the service sector in Africa is equally or even more productive than the manufacturing sector, which is even more evident in the bottom-income quantile economies in Africa than in the top-income quantile. Third, the agriculture sector has overtaken the manufacturing sector as the fastest-growing sector, an empirical regularity common to emerging and industrialised economies. Fourth, while capital utilisation in agriculture has remained a paltry sum, capital utilisation in the service sector has surpassed that of the industrial sector over the last decade. The second essay empirically examines the relative price hypothesis – which suggests that structural change stems from changes in relative prices reflecting differences in sectoral technology within and across countries. Using a panel of 34 African economies from 1970 to 2017, the study used the instrumental variable and fixed effects models to examine the sources of cross-country variation in the pattern of structural change from a capital intensity and absorptive capacity perspective. Results from the study revealed the following. First, in cross-country comparative analyses, a broad (macro) level of analysis or aggregation approach, such as the two-sector models, masks heterogeneous trends in the data, thus emphasising universal similarities across countries. However, the disaggregated (micro) level analysis allows for variations specific to the peculiarities of countries making up the sample. Second, the study observed robust empirical evidence of capital-skill complementarity among the technological forces driving structural change in Africa. Third, the results showed that beneath the capital-skill complementarity, intersectoral differences abound, with cross-country variations in sectoral capital intensity exhibiting larger effects in capital-intensive agriculture and tradable goods (comprising agriculture and manufacturing), than in low capital-intensive sectors (non-agriculture, non-tradable and services). Conversely, the study observed larger cross-country variations in absorptive capacity in high-skill-intensive sectors (non-agriculture, non-tradable and services) relative to low-skilled sectors of agriculture, tradable and goods production (agriculture, manufacturing and non-manufacturing). Fourth, the study found evidence showing that at the disaggregated level, there is lower cross-country variance in sectors with a high codified knowledge base, technology transfer and absorption, such as mining and utilities, transport and communications and manufacturing, than in sectors with a tacit knowledge base and low technology transfer and absorption such as agriculture, retail and hospitality and other services. The third essay investigates the long-term consequences of commodity specialisation as a driver of structural change by decomposing commodity specialisation into commodity windfall and commodity price effects to estimate their direct and indirect transmission channels using panel quantile regression on 34 African economies from 1970 to 2017. Our results show the following. First, commodity windfall at the aggregate two-sector level diminishes the tradable and non-tradable sectors across the distributional quantile. Conversely, real commodity prices only directly depress growth in the tradable sector at the 75th quantile while damaging the non-tradable sector at both the lower and median quantile distribution. Second, commodity windfall and prices do not directly impact the agricultural sector, irrespective of the distributional quantile. Third, while commodity windfall has no direct shrinking effect on the manufacturing sector except at the 25th quantile, real commodity prices directly diminish manufacturing growth with comparable marginal effects across the distributional quantile. Fourth, contrary to the theoretical prediction, the direct Dutch Disease effect – the crowding out of other sectors of the economy, especially the tradable sector, induced by commodity specialisation in resource-rich economies – is more visible in the non-tradable sector, with the construction sector being worse off, followed by the transport and telecommunication, and retail and hospitality sectors, respectively. Fifth, the loss in non-resource commodity export competitiveness and the declining commodity terms of trade are the most debilitating indirect transmission channels. Also, contrary to theory, the study found that the real effective currency appreciation transmission channel is but a statistical mirage. The fourth and final essay explores the inclusive growth–a generic term for welfare – effects of structural change transition patterns – using the Augmented Mean Group estimator on a panel of 17 African countries from 1971 to 2017. The results revealed the following: structural upgrading from low-productive agriculture to any other part of the modern sector enhances inclusive growth, especially for low-income economies. Second, the study show heterogeneity in the welfare-enhancing effects of the different transition patterns of structural upgrading. Specifically, although structural upgrading enhances living standards, it can also induce welfare loss in many African economies, depending on the transition path followed. For instance, while the transition from agriculture to market services enhances inclusive growth in many African economies, manufacturing upgrades exhibit welfare loss or have no effects on inclusive human development. Third, the structural upgrade from agriculture to non-manufacturing is more welfare-enhancing in many African economies than the upgrade from agriculture to manufacturing, especially in middle-income economies. Third, the direct upgrade from agriculture to market services engenders inclusive growth in more African countries than the upgrade from agriculture to manufacturing. Further, the study also found that while disaggregated non-tradable service upgrades engender inclusive human development in more African economies, the reverse is true for tradable service upgrades, which foster improvements in per capita consumption in more countries. Fourth, the study provides evidence of a persistent trade-off between the welfare-enhancing effects of long-term inclusive human development and short-term household consumption – often favouring household consumption, as observed across several transition channels in several economies, including Rwanda and Lesotho, Senegal and Morocco.
AFRIKAANSE OPSOMMING: Soos ekonomiee ontwikkel, ondergaan hulle grootskaalse strukturele verandering – die hertoewysing van ekonomiese aktiwiteite oor drie bree sektore van landbou, vervaardiging en dienste wat ekonomiese ontwikkeling vergesel. Daarom is 'n fundamentele eienskap van strukturele verandering die strukturele heterogeniteit binne en oor lande heen, wat aanleiding gee tot variasie oor die land in strukturele veranderingspatrone. Buiten die sistematiese variasie in strukturele samestelling, veroorsaak strukturele transformasie ook 'n wye reeks komplementere prosesse en veranderinge in tegnologiese vooruitgang, verbruikersgedrag, verstedeliking, demografiese oorgang, vroulike arbeidsdeelname, lewenstandaarde, welsynsherverdeling en selfs sosio-politieke instellings. Alhoewel strukturele transformasie omvattend ondersoek is in verskeie geindustrialiseerde ekonomiee en 'n pertinente beleidskwessie onder ontwikkelende ekonomiee bly, bly empiriese insigte oor hierdie prosesse binne ontwikkelende ekonomiee onduidelik. Die versameling van vier selfstandige opstelle in hierdie proefskrif ondersoek die dryfvere en inklusiewe groei-effekte van Afrika-strukturele veranderingspatrone om 'n dieper konseptuele begrip van beleidsimplikasies te verskaf. Die eerste opstel heroorweeg die gestileerde feite van strukturele verandering binne die konteks van Afrika. Spesifiek ondersoek ons die neigings en strukturele samestelling van veranderinge in waardetoegevoegde en indiensnemingsaandele, relatiewe produktiwiteit en produktiwiteitsgroeikoerse en kapitaalintensiteitsvlakke. Ons het aansienlike sektorale verskille gevind, wat strukturele ongelykheid in Afrika se ekonomiese groeipatroon beklemtoon. Die volgende empiriese insigte is uit die ontledings verkry. Eerstens is enige sprake van voortydige de-industrialisering in Afrika voortydig, aangesien die vervaardigingsektor steeds uitbrei, hoewel stadiger as wat verwag is. Tweedens is die dienstesektor in Afrika ewe of selfs meer produktief as die vervaardigingsektor, 'n verskynsel wat selfs meer sigbaar is in die onderste-inkomstekwantiel-ekonomiee in Afrika as in die top-inkomstekwantiel. Derdens het die landbousektor die vervaardigingsektor verbygesteek as die sektor wat die vinnigste groei, 'n empiriese reelmaat wat algemeen is vir ontluikende en geindustrialiseerde ekonomiee. Vierdens, terwyl kapitaalbenutting in die landbou 'n skamele bedrag gebly het, de kapitaalbenutting in dienste die bedryfsektor oor die afgelope dekade verbygesteek. Laastens poog die hoofstuk om die strukturele meganismes agter die gestileerde feite van strukturele verandering te verduidelik as 'n wisselwerking tussen die ekonomiese kragte van vraag en aanbod, wat wissel van Engel se vraag-effekte tot sektor-bevooroordeelde tegnologie en handelseffekte. Die tweede opstel ondersoek die relatiewe pryshipotese empiries – wat daarop dui dat strukturele verandering spruit uit veranderinge in relatiewe pryse wat verskille in sektorale tegnologie binne en oor lande weerspieel. Die studie het die instrumentele veranderlike en die vaste-effekte-modelle toegepas in die ontleding van die bronne van variasie oor die land in die patroon van strukturele verandering vanuit 'n kapitaalintensiteit- en absorptiewe kapasiteitsperspektief deur 'n paneel van 34 Afrika-ekonomiee van 1970 tot 2017 te gebruik. Resultate van die studie toon die volgende; eerstens, robuuste empiriese bewyse van kapitaal-vaardigheid-komplementariteit tussen die tegnologiese kragte wat strukturele transformasie in Afrika dryf. Tweedens, die resultate toon dat onder die kapitaal-vaardigheid-komplementariteit, intersektorale verskille oorvloedig is met variasies in sektorale elastisiteit oor die land tot totale kapitaalintensiteit, wat groter effekte in kapitaalintensiewe landbou en verhandelbare goedere (wat uit landbou en vervaardiging bestaan) toon, as in lae kapitaal-intensiewe sektore (nie-landbou, nie-verhandelbaar en dienste). Omgekeerd het ons groter variasies oor die land waargeneem in absorpsievermoe in hoe-vaardigheid-intensiewe sektore (nie-landbou, nie-verhandelbaar en dienste) relatief tot laaggeskoolde sektore van landbou, verhandelbare en goedereproduksie (landbou, vervaardiging en nie-vervaardiging). Derdens vind ons bewyse wat toon dat daar op die gedisaggregeerde vlak laer variansie oor die land is in sektore met 'n hoe gekodifiseerde kennisbasis, tegnologie-oordrag en absorpsie, soos mynbou en nutsdienste, vervoer en kommunikasie en vervaardiging, as in sektore met 'n stilswyende kennisbasis en lae tegnologie-oordrag en -absorpsie soos landbou, kleinhandel en gasvryheid en ander dienste. Die derde opstel ondersoek die langtermyngevolge van kommoditeitspesialisasie as 'n drywer van strukturele verandering deur kommoditeitspesialisasie te ontbind in kommoditeit meevaller en kommoditeitspryseffekte om hul direkte en indirekte transmissiekanale te skat deur paneelkwantielregressie op 34 Afrika-ekonomiee van 1970 tot 2017 te skat. Ons resultate toon die volgende. Eerstens, kommoditeit meevaller op die totale twee-sektor vlak verminder die verhandelbare en nie-verhandelbare sektore oor die verspreidingskwantiel. Omgekeerd druk reele kommoditeitspryse slegs direk groei in die verhandelbare sektor by die 75ste kwantiel terwyl dit die nie-verhandelbare sektor by beide die laer en mediaan kwantielverspreiding beskadig. Tweedens, kommoditeit meevaller en pryse het nie 'n direkte impak op die landbousektor nie, ongeag die verspreidingskwantiel. Derdens, terwyl kommoditeit meevaller geen direkte krimpende effek op die vervaardigingsektor het nie, behalwe by die 25ste kwantiel, verminder reele kommoditeitspryse vervaardigingsgroei direk met vergelykbare marginale effekte oor die verspreidingskwantiel. Vierdens, in teenstelling met die teoretiese voorspelling, is die direkte Nederlandse Siekte-effek – die verdringing van ander sektore van die ekonomie, veral die verhandelbare sektor, veroorsaak deur kommoditeitspesialisasie in hulpbronryke ekonomiee – meer sigbaar in die nie-verhandelbare sektor, met die konstruksiesektor wat swakker af is, gevolg deur die vervoer- en telekommunikasie-, en kleinhandel- en gasvryheidsektore onderskeidelik. Vyfdens, die verlies in nie-hulpbron-kommoditeit-uitvoermededingendheid en die dalende handelsvoorwaardes as die mees aftakelende indirekte transmissiekanale. Ook, in teenstelling met teorie, vind die studie dat die werklike effektiewe geldeenheidappresiasie-oordragkanaal slegs 'n statistiese lugspieeling is. Die vierde en laaste opstel ondersoek die inklusiewe groei-effekte van strukturele verandering-oorgangspatrone – met behulp van die Augmented Mean Group-beramer op 'n paneel van 17 Afrika-lande van 1971 tot 2017. Die resultate toon die volgende: strukturele opgradering van lae-produktiewe landbou na enige ander deel van die moderne sektor bevorder inklusiewe groei, veral vir lae-inkomste ekonomiee. Tweedens toon die studie heterogeniteit in die welsyn-verbeterende effekte van die verskillende oorgangspatrone van strukturele opgradering. Spesifiek, hoewel strukturele opgradering lewenstandaarde verhoog, kan dit ook welsynsverlies in baie Afrika-ekonomiee veroorsaak, afhangend van die oorgangspad wat gevolg word. Byvoorbeeld, terwyl die oorgang van landbou na markdienste inklusiewe groei in baie Afrika-ekonomiee bevorder, toon vervaardigingsopgraderings welsynsverlies of het geen uitwerking op inklusiewe menslike ontwikkeling nie. Derdens is die strukturele opgradering van landbou na nie-vervaardiging meer welsynsversterkend in baie Afrika-ekonomiee as die opgradering van landbou na vervaardiging, veral in middel-inkomste ekonomiee. Derdens veroorsaak die direkte opgradering van landbou na markdienste inklusiewe groei in meer Afrika-lande as die opgradering van landbou na vervaardiging. Verder vind ons ook dat terwyl uiteenlopende nie-verhandelbare dienste-opgraderings inklusiewe menslike ontwikkeling in meer Afrika-lande teweegbring, die omgekeerde in verhandelbare dienste-opgraderings verbeterings in per capita-verbruik in meer lande bevorder. Vierdens verskaf die studie bewyse van 'n aanhoudende afweging tussen die welsyn-verbeterende uitwerking van langtermyn-inklusiewe menslike ontwikkeling en korttermyn huishoudelike verbruik - wat dikwels huishoudelike verbruik bevoordeel, soos waargeneem oor verskeie oorgangskanale in verskeie ekonomiee, insluitend Rwanda en Lesotho, Senegal en Marokko.
Description
Thesis (PhD)--Stellenbosch University, 2023.
Keywords
Structural change; Capital intensity; Absorptive capacity; Commodity specialisation; Dutch disease; Inclusive growth; Africa.
Citation