The gendered effects of tariff liberalisation on local labour markets in post-apartheid South Africa

Type Thesis or Dissertation - PhD thesis
Title The gendered effects of tariff liberalisation on local labour markets in post-apartheid South Africa
Publication (Day/Month/Year) 2020
This thesis uses South Africa as a case study to examine the gendered effects of tariff liberalisation on labour market outcomes at the local level. Specifically, the thesis focuses on the effect of tariff liberalisation on regional employment growth, labour adjustment in the manufacturing sector and services sector, and internal migration over a period (1996 to 2011) in which South Africa substantially reduced tariff protection. This was also a period corresponding with low employment growth and declines in the manufacturing share of employment that vary by gender and race as well as across regions. The experience of South Africa, therefore, presents a useful context to empirically identify the association between tariff liberalisation and gendered outcomes in local labour markets. To analyse the relationship between tariff liberalisation and local labour market outcomes, the structure of the thesis chapters is configured around four research questions: 1. How is regional employment growth distributed in South Africa? 2. What is the effect of tariff liberalisation on gendered employment and wages in manufacturing? 3. How does tariff liberalisation in manufacturing affect wages and employment by gender in the services sector? 4. Does tariff liberalisation drive internal migration? The thesis draws on a regional level database of employment, wages and migration constructed using South African Population Census data for 1996, 2001 and 2011, that is combined with product-level import tariff data over the period. Chapter 2 sets the context for the thesis by assessing key sources of regional employment growth across the 234 local municipalities in South Africa. The chapter adopts a dynamic industry shift-share analysis method that decomposes the growth in national employment into three sources: (i) national growth effects, (ii) industry-mix effects and (iii) regional competitive effects. The results reveal substantial variation in employment growth across regions that is closely associated with the regions' industry composition and competitiveness. Our findings also reveal that industry-mix effects are negatively correlated with the initial employment ratio of tradable to nontradable sectors. Further, the industry-mix effect is positively correlated with the regions' initial income level, initial share of female workers, initial share of Black female workers, initial share of Black male workers, and initial share of White female workers. This descriptive review and data analysis demonstrates the importance of industry composition in driving regional patterns of employment growth and reveals how these changes are associated with the gender and racial composition of workers in the region. The remainder of the thesis analyses how tariff liberalisation may have contributed towards these outcomes. Chapter 3 investigates the gendered effects of tariff liberalisation on manufacturing employment in South Africa. The chapter adopts the Bartik (1991) approach and exploits variations in the regional composition of industries to identify how tariffs affect manufacturing employment at the local level. To examine the effect of tariff reductions on manufacturing employment, a first-difference instrumental variable estimation strategy is employed. First, the results indicate that tariff reductions had no effect on manufacturing wages, with its impact falling entirely on employment. Municipalities that were more exposed to tariff reductions experienced slower growth in manufacturing employment of both men and women. The effect was significantly stronger for women, particularly Black women, thus widening the gender employment gap. The results of testing for the various channels that may have given rise to these gendered effects reveal the dominant sources to be industry segregation combined with the comparatively large reductions in tariffs in female-intensive industries such as textiles, clothing, and footwear. Chapter 4 investigates whether tariff liberalisation is associated with structural shifts in employment from manufacturing to services by employing a similar empirical approach to that used in Chapter 3. Consistent with theoretical expectations, the chapter illustrates that tariff liberalisation was associated with strong increases in the services to manufacturing employment ratio, but this shift was not driven by the absorption of employment in the services sector. In fact, employment in the services sector also fell in regions experiencing relatively large tariff reductions. Further analysis demonstrates that the decline in services employment was driven by lower derived demand, income, and manufacturing infrastructure investment that are linked to the decline in manufacturing from tariff reductions. The results also show that tariff effects differed by gender and race. The implication of the results is that spillover effects from the decline in manufacturing diminished the absorption of labour by the services sector, thus exacerbating the regional employment impact of tariff liberalisation. The final chapter, Chapter 5, provides a descriptive analysis on how tariff liberalisation affects internal migration in South Africa. Tariff liberalisation alters relative wages and relative employment opportunities across regions, thus giving rise to internal migration. The extent to which labour responds to these changes by migrating affects the local market outcomes of liberalisation. The analysis utilises a gravity-style model and instrumental variable estimation strategy to estimate the effect of tariff liberalisation on internal bilateral migration flows. The main finding is that there was higher out-migration and lower in-migration in regions that experienced relatively large tariff reductions, as is predicted by theory. The chapter reveals that tariff-induced internal migration differed across gender in both locations with women appearing to be more spatially mobile than men. The chapter also teases out gender implication by race and family structure. The core finding is that tariff-induced gendered internal migration varied according to individual characteristics. Overall, the thesis provides new evidence regarding the impact of tariff liberalisation on local labour market outcomes in South Africa. Liberalisation had no effect on regional wages, but lowered employment and induced internal migration. The research also reveals substantial changes in the gender and industry composition of regional employment associated with tariff liberalisation. The thesis contributes to international trade theory by highlighting that the gendered effects of tariff liberalisation are country-specific, largely dependent on the intensity of tariff reductions and the gender intensities across industries. The study also demonstrates that internal migration is a mechanism for mitigating the undesirable effects of liberalisation. Accordingly, the thesis emphasises that tariff liberalisation effects are not homogenous across gender, and thus gender-specific policies may be required to ameliorate the unequal adjustment costs across different genders.

Related studies