The analysis of structural change in a period of rapid economic growth – contributions from Vietnamese experience

After the 1980s transition ‘from plan to market’, from 1992 Vietnamese economic growth accelerated significantly. Strategy continued to stress industrialisation; yet current price GVA/worker in manufacturing, initially relatively high, was by around 2010 below average. Services therefore came to dominate growth, not industry. Effective capital and labour markets, which had come to prominence in the 1980s, developed further, and higher rewards paid to labour and capital in services drove structural change, despite policy favouring manufacturing. This account, though not generally accepted, follows the global pattern and contradicted donor advice. However, like the 1980s, by the end of the 2010s Vietnamese ‘lived experience’ came to contradict donor-supported strategy and extant Party doctrine, and policy shifted to support servicisation, or more accurately structural change that supported relatively high actual or potential GVA/worker rather than an ‘industrialisation’ vision. The paper suggests that we can learn from this, and analyses structural change with an unorthodox analytical framework that requires defending as it is anathema to mainstream approaches. It does this by building on recently published work analysing the causes of confirmation bias in most economic analyses that supported the incorrect donor advice of the necessity of industrialisation. It blames this upon the reliance of standard analyses upon constant price sectoral GDP data to empirically found production function-based modelling (Arrow 1974, Fforde 2024c). This then legitimises the paper’s use of current price sectoral GVA/worker (GDP/worker) to analyse structural change during Vietnam’s ongoing Economic Miracle. The results show important aspects of servicising growth: the absence of a sub-sector that can be conceptualised as a ‘leading sector’ (the role played by manufacturing in standard incorrect narratives); and the striking failure of the very high relative GVA/worker in manufacturing in the 1990s to drive industrialisation, as by the 2010 this was below average and contributing little to GDP growth; and the importance of understanding the operative incentive pattern (which so far we do not) and the causes of variations in GVA/worker.