Technological change and the consequences of job loss
What is this research about and why did you do it?
We examine if the introduction of new technologies (i.e., technological change) contributes to the large and persistent decline in earnings following job loss. A large literature has shown that job loss leads to substantial declines in earnings that persist for years. However, despite being documented across numerous countries and time periods, the underlying reasons for this phenomenon remain not well understood. Separately, researchers have shown that technological change has led to huge changes in the labour market in the past half-century and this motivated us to examine if the two phenomena were linked.
How did you answer this question?
To answer this question, we created a measure of how exposed each occupation is to technological change. Our measure utilizes detailed skill requirements from the near universe of online vacancies and tracks how computer and software requirements have expanded within an occupation over time. We then combined our measure of exposure to technological change with data on the earnings losses of displaced workers (i.e., workers who lost their job through no fault of their own) and information on their occupation before and after displacement. We then empirically examined how exposure to technological change in your pre-displacement occupation impacted the size of earnings losses.
What did you find?
We find that workers who are more exposed to technological change suffer larger declines in earnings after job loss. The figure below shows a binned scatterplot of the change in computer and software requirements by occupation (x-axis), which is our measure of exposure to technological change, and the average change in earnings after displacement (y-axis). The graph shows that workers who are displaced from occupations that experience the most rapid change in computer and software requirements suffer the largest earnings losses. Conversely, workers displaced from occupations where there has been less change in computer and software requirements have significantly smaller earnings losses.
Figure Notes: The figure shows a binned scatterplot of the change in computer and software requirements between 2007 and 2017 by occupation, as measured in the Burning Glass data (x-axis), and the average change in log earnings after displacement, as measured in the Displaced Worker Supplement (y-axis). Occupations are classified using four-digit SOC codes.
What implications does this have for the study (research and teaching) of wealth concentration or economic inequality?
These results highlight that there is huge heterogeneity in the size of earnings losses by the occupation that an individual is displaced from. We further show in the paper (Figure 3(b)), that our results are driven by workers switching out of the occupations that are most exposed to technological change. This suggests that technological change lowers earnings after displacement by making it so that workers no longer have the skills to work in their prior occupation.
What are the next steps in your agenda?
We view our results as suggesting that policies that encourage retaining during unemployment may play a role as part of the optimal policy for unemployed workers. We are starting to explore this in Braxton and Taska (2023b).
Citation and related resources
Braxton, C. J., and Bledi, T. (2023). "Technological Change and the Consequences of Job Loss." American Economic Review, 113 (2), pp279-316.
- Braxton, C. J., and Bledi, T. (2023b). "Technological Change and Insuring Job Loss." Unpublished manuscript.