Two major forces, namely rapid advancements in technology and changes in global supply chains, can independently affect how labor markets operate. Little is known about the impact of these intertwined forces in combination. This study extends Frey and Osborne's (2017) machine learning approach that explores this issue. We incorporate the features that make certain jobs difficult or impossible to perform using offshore labor markets as well as local experts' opinions to estimate domestic employment risk for each occupation. We apply our models to Thailand – a major destination for outsourced operations with ongoing technological restructuring. Our results reveal that clerical jobs face the highest risk. However, as most workers are employed in the skilled agricultural and service sectors, these occupations stand to suffer most in terms of the number of job losses. Under the assumption of no voluntary adjustments (i.e., the worst-case scenario), more than half of existing jobs could be at risk. Our approach offers a useful tool for countries facing similar structural changes to identify employment risk and develop appropriate mitigation strategies.

两种主要力量，即技术的快速进步和全球供应链的变化，可以独立影响劳动力市场的运作方式。人们对这些相互交织的力量结合起来的影响知之甚少。本研究扩展了 Frey 和 Osborne (2017) 探索该问题的机器学习方法。我们结合使用离岸劳动力市场以及当地专家的意见来估计每个职业的国内就业风险，从而使某些工作难以或不可能完成的特征。我们将我们的模型应用于泰国——一个正在进行技术重组的外包业务的主要目的地。我们的结果表明，文书工作面临的风险最高。然而，由于大多数工人受雇于熟练的农业和服务部门，就失业人数而言，这些职业遭受的损失最大。在没有自愿调整的假设下（即最坏的情况），超过一半的现有工作岗位可能面临风险。我们的方法为面临类似结构变化的国家提供了有用的工具，以识别就业风险并制定适当的缓解策略。