Fears abound that automation and other advanced technologies will lead to job losses for lower-skilled workers in emerging economies and exacerbate inequality. Each new wave of technological progress is met with dire predictions. The most critic argue that the unprecedented pace of technological change today will have more dramatic effects on the future of work as new technologies (including robots and artificial intelligence) are increasingly replacing more educated workers and more cognitive and analytical work. At the same time, many economists argue that technology adoption will significantly increase firm productivity and result in job expansion, at least in the medium run under certain policy conditions. The impacts of technology adoption on overall employment and on the skills composition of occupations are ultimately an empirical question.
New evidence to inform this debate exploits firm-level data for Chile on the adoption of complex software, such as client management, business or production software packages. This type of software can plan production levels, price products, estimate costs along the production process or forecast agricultural production – all complex tasks usually performed by professional workers. Past studies of the effects of technology adoption on employment and skills have mostly considered the use of computers and access to internet and assessed impacts on aggregate employment categories. This novel study focuses on the adoption of new software and assesses impacts on the task content of occupations considering four types of tasks -abstract, routine-cognitive, routine-manual, and non-routine manual tasks (available in the PIAAC survey) and four broad occupations – managers, administrative workers, skilled and unskilled production workers (available in the ELE survey).
The hypothesis tested in the study is that, within firms, the adoption of complex software is likely to replace complex routine-cognitive and non-routine analytical tasks, typically performed by skilled workers using this technology. We test this hypothesis by looking at the patterns of changing demand for occupations and changing firm-level task indexes following the adoption of complex software for firms across all sectors of the Chilean economy over the six-year interval 2007-2013.
In Chile, the adoption of complex software has been inclusive, as adopting firms expand significantly their employment of administrative and unskilled production workers. Within firms, the mix of workers also changed in the medium run, as firms reallocated away from skilled production workers. Consistent with these employment shifts, the adoption of complex software leads also to an increase in firms’ use of routine and manual tasks, and to a reduction in firms’ use of abstract tasks, which are now arguably being performed by technology. The findings are driven by the adoption of complex software in manufacturing, wholesale and retail trade, all sectors with a relatively less educated workforce and lower productivity.
The results of this study bear important policy implications. Adoption of complex software can lead, in the medium run, to an “inclusive” employment expansion at least in some sectors. Over a six-year horizon as firms are able to overcome short-term input rigidities, and education and training systems can help promote such technology adoption by supporting the development of digital skills among employers. After all, fears that technological change will lead to mass unemployment have not materialized. On the contrary, in Chile, the adoption of complex software is offering a pathway to more inclusive growth by increasing firms’ employment of less skilled workers and more routine tasks.