Creating stable and well-paid jobs is central to eliviating poverty in low-and-medium income countries. International trade is sometimes seen as an instrument to achieve this goal. This column looks at the impact of trade liberalization between China and Indonesia, in the form of the ASEAN-China Free Trade Agreement, and finds that it had no effect on the probability of Indonesian workers being wage-workers, except among the more educated, and no significant impact on the employment and value-added of large manufacturing firms. The findings underscore the need for cautious optimism regarding the potential gains in terms of job quality from further trade liberalization, particularly in a global environment where tariffs are already low.
One important challenge identified to fight extreme poverty in low- and middle-income countries (LMICs) is the creation of stable and sufficiently well-paid jobs by private firms (Lee 2013). However, LMICs often lack promising private firms and when they host some, these firms often face important challenges to grow and create jobs (Rijkers et al. 2015). International trade is sometimes pointed as a catalyst for better jobs. Pavcnik and McCaig (2015) show, for example, that in the case of Vietnam new export opportunities to the US in the years after 2000 promoted transitions from the informal to the formal sector, favouring aggregate productivity growth. In the case of Brazil, Ulyssea et al. (2021) demonstrate that productivity gains generated by trade liberalisation are stronger when we account for the informal sector, as trade reduces the share of the informal economy. Similarly, trade also increases the share of better paid jobs and reduces wage inequality.
Still, the evidence regarding the impact of trade liberalisation on the quality of jobs available in LMICs remains scarce. Filling this gap is all the more important given that the relative merits of international trade are being hotly debated after the disruptions in the global value chains caused by the COVID-19 pandemic and in the context of increased international geopolitical tensions (Campos et al. 2023).
In a recent paper (Delesalle et al. 2025), we make progress in this direction by estimating the impact of the trade liberalisation between China and ASEAN countries, in the form of the ASEAN-China Free Trade Agreement (ACFTA), on the probability of Indonesian workers being wage-earners, a proxy for occupying more stable and better-paid jobs in the Indonesian context. To understand how this trade liberalisation episode affected the labour market, we also examine its impact on the job creation process in big manufacturing firms.
Being a wage-worker – a measure of job quality
In LMICs, many working-age people are self-employed (often engaged in informal subsistence entrepreneurship), or work as unpaid workers (employed without a regular contract, mostly working in family businesses). In 2010, more than 50% of the working-age population was in this situation in Indonesia. They earn much less than regular wage-workers (workers employed with a regular contract in formal firms): we find that in Indonesia, in 2003, the income of a wage-worker was, all else equal, 10% higher than the income of a self-employed and 40% higher than the income of an occasional worker in the agricultural sector. Moreover, wage-workers are less likely than the others to be looking for another job, suggesting they are more satisfied with their current job. In the context of LMICs, being a wage-worker is thus a first-order dimension of job quality.
The ASEAN-China Free Trade Agreement
In 2002, the members of the Association of Southeast Asian Nations (ASEAN) signed the ACFTA with a view to eliminate tariffs on 90% of imports by 2010. As a member of ASEAN, Indonesia granted substantial tariff reductions to China, which were progressively strengthened over the period 2005-2010. As shown in part (a) of Figure 1, this translated into a decrease in the average tariff applied by Indonesia to Chinese imports from above 8% in 2000 to below 2% in 2010. This is a moderate decrease compared to the fall in tariffs observed after the entry of Indonesia in WTO, from an average of 20.88% to an average of 8.44% (Amiti and Konings 2007), but this is the case for most recent trade liberalisation episodes. Part (b) of Figure 1 shows that the tariff cuts were comprehensive and covered a wide range of sectors, with some of them such as the textile and footwear products experiencing substantial tariff cuts.
Figure 1 Evolution of Indonesian import tariffs
a) Indonesian import tariffs by country of origin
b) Indonesian tariffs on imports from China by type of products
Over the period 2000-2010, China’s share of Indonesian trade increased markedly. In 2000, China accounted for about 7.6% of Indonesian imports and 4.7% of its exports, ranking fifth for both types of flows. By 2010, these numbers had grown to 15.7% and 9.5%, respectively, making China the largest source of Indonesian imports and its second largest export destination. The Chinese share of Indonesian trade flows kept growing after 2010 to reach 22.8% for imports and 9.6% for exports in 2015. The evolution of these shares is depicted in Figure 2. For both imports and exports, the Chinese share of Indonesian trade grew faster than its share in world trade, especially after 2005. Hence, the implementation of ACFTA was associated with an increase in Chinese exports to Indonesia that went beyond its rising importance in world trade over the same period. Regression results show that the surge in Indonesian imports from China was stronger for those products where Indonesia decreased its import tariff on China more, and similarly, Indonesian exports increased relatively more for products where China decreased its import tariff on Indonesia more. All of this shows that for Indonesia, the ACFTA caused an increase of international trade with China relatively to its other trade partners.
Figure 2 Share of China in Indonesian and world trade
a) Imports
b) Exports
Methodology
To asses the impact of the ACFTA on the probability of being a wage-worker, we focus on trade liberalisation in the manufacturing sector (tariffs already being quite low on the agricultural sector in Indonesia in 2000) We consider separately the impact of input and output tariff cuts granted by Indonesia to China, as well as the tariff cuts granted by China to Indonesian exporters. For the purpose of identification, we exploit geographical variations in the exposure to the trade liberalisation with China across 224 Indonesian districts. These variations come from the fact that output and input tariff reductions vary across sectors, and districts have different mixes of industries at the beginning of the period under study. To address the usual challenges related the clean identification of causal impacts of trade liberalisation, we use the fact that tariff cuts were more important for products that initially benefited from higher protection as an instrumentation strategy.
If anything, a positive impact for the more educated only
We find that the trade liberalisation between China and ASEAN countries had, on average, no significant impact on the probability of being a wage-worker in Indonesia. This remains true for various subgroups of the working-age population defined according their sector of activity, their gender, or their age. The only exception is the group of more educated workers (those with more than primary education), who benefit on average from overall tariff cuts. However, they account for only 25% of the workforce. Therefore, if anything, the impact of the ACFTA on access to better-paying jobs was regressive. Additional results show that the districts which were more exposed to the tariff variations imposed by the ACFTA have not experienced major changes in either the size of their active population nor in the composition of their workforce in terms of the various types of jobs or migration status.
Consistent with this lack of average effect at the worker level, we do not find any significant impact of the ACFTA on the employment and value-added of big (20+ employees) manufacturing firms in Indonesia, big firms being the main providers of jobs for wage-workers.
Conclusion
Our findings depart from Kis-Katos and Sparrow (2015), who report evidence of a positive effect of tariff reduction on formal employment in Indonesia during the period 1993 to 1996. However, in their case, the tariff cuts were those related to the entry of Indonesia into WTO; hence, they were both multilateral and larger. They also depart from the literature on the China shock that evaluates the impact of the entry of China on third-country labour markets (e.g. Price et al. 2014 on the US, Utar 2018 on Denmark).
We see two main reasons why our results are different. First, we are evaluating the effect of a FTA that came into force on top of the preferential access China already benefited from since its entry in the WTO. When signed between WTO members, the effects of such agreements may be more modest as the trade shock they induce is arguably smaller. Understanding the effects of such FTAs with China however remains important as many others are under negotiation and generate heated debates. Second, we find that the correlation between the structure of Chinese exports and Indonesian exports in terms of product in 2000 is smaller than the one measured between China and the US, Mexico or Germany. This suggests that at least for the import side, competition between China and Indonesia may have been less intense, mitigating the possible adverse effects on Indonesian local labour markets.
While our results may be context-specific, they underscore the need for cautious optimism regarding the potential gains in terms of job quality from further trade liberalisation in LMICs, particularly in a global environment where tariffs are already low. These findings suggest that, in addition to trade liberalisation agreements, policymakers should consider complementary policies to enhance labour market outcomes, especially for the least-educated, who account for the lion-share of the working-age population.
Source : VOXeu
