My repository of thoughts on interesting ideas.
New Policy in the Age of AI
Final part in an analysis of the macroeconomic effects of AI.
Why We Need Policy and Government Intervention
Private sector investment and innovations in automation are not necessarily aligned with the goal of Pareto improvement and social optimization (Acemoglu, 2021). As such, to ensure a fair market of employment and wages for society, policymakers may have to implement robust policies.
Policy #1
Policymakers should first invest in training and job transition services to assist employees displaced by AI to reintegrate into the workforce (Hoynes & Rothstein, 2019). This can be achieved through subsidization of intermediaries, such as public employment services, outplacement offices, and temporary help agencies (U.S.-E.U. TTC, 2022). Policy promoting such programs could be effective in increasing earnings and job mobility (Katz et al., 2020).
Policy #2
Since risk markets are imperfect, “losers” emerge from technological progress (Korinek & Stiglitz, 2017). As such, Korinek and Stiglitz provide a series of second-best devices to initiate transfers between innovators and workers, which include changes to intellectual property rights and capital taxation. They use the intuition that by shortening the period of time required for innovation to enter the public domain, the incentive to create worker-automating capital decreases. Furthermore, a tax imposed on capital, provided the supply elasticity is low, can countervail the losses incurred by displaced workers. While this may conflict with elements of pure shareholder capitalism, system reformations are needed to promote wage growth and benefits for low-wage workers (Autor, 2022).
Policy #3
By publicly funding research on AI through an increased federal R&D budget, governments can direct the development of the technology to augment rather than automate work (U.S.-E.U. TTC, 2022). The council suggests the exploration of new AI methods; the impact of AI on workers’ wages and employment, anti-competitive behavior, and biases; and the development of AI ethics. This research could promote socially optimal and sustainable AI growth, limiting negative externalities.
Final Remarks
As AI revolutionizes work in modern economies, technology has the potential to entrench the differences between wealthy and median workers as well as high-skilled labor and low-skilled workers. In order to limit these effects, policies aimed at alleviating the adverse effects of automation faced by struggling populations must be implemented. By doing so, we progress towards a future where the transformative power of AI fosters prosperity for all.
