ACADEMIC ARTICLE SUMMARY
AI and the Economy
Article Source: NBER Working Paper 24689, 2018
Publication Date:
Time to Read: 2 minute readSearch for the full article on Bing
ARTICLE SUMMARY
Summary:
Artificial intelligence (AI) and robotics may boost productivity, but might disrupt the labor market. New regulatory and antitrust policies might mitigate harmful effects.
POLICY RELEVANCE
Policy Relevance:
New policies might help ameliorate problems in the labor market caused by AI.
KEY TAKEAWAYS
Key Takeaways:
- Growth in advanced economies was slower from 2006 to 2016 than from 1996 to 2006; the male labor force participation rate fell from 98 percent in the 1950s to 89 percent in 2016, suggesting that some struggle to find new jobs.
- Some are concerned that AI will precipitate rapid change, such that there are long periods of time in which large segments of the population are not working.
- Many economists rely on theoretical models to predict productivity growth related to AI and robotics, due to a lack of data about the use of AI and robotics at the firm level; some early firm-level studies find that robots do increase labor productivity.
- So far, most AI is used to lower the cost of making predictions through machine learning; this suggests that AI will not have an unprecedented effect on the entire economy, but will disrupt some sectors and workers.
- Antitrust agencies recognize that the control of large datasets by dominant tech firms might create barriers to competition; requiring that customers be allowed to move their data from one provider to another ("data portability") could help ensure that new firms can enter the market.
- Universal basic income programs, which provide enough cash to stop people from falling below the poverty line, could address mass joblessness, but might require doubling of basic tax rates.
- Other proposals to address large-scale unemployment would include expanded employment subsidies (payments to supplement the income of those earning low wages) or guaranteed federal employment.
- Policymakers could create a new agency to regulate AI, either to advise other agencies or with its own enforcement powers.