Key Takeaways
- • Dean Baker of the Center for Economic and Policy Research (CEPR) opposes Bernie Sanders's proposed AI sovereign wealth fund, calling it "wrong-headed from every angle."
- • Baker argues the fund would channel more public money to billionaires such as Elon Musk and Mark Zuckerberg, whom he says have already received enough.
- • Baker cites weak recent productivity growth as evidence that AI-driven job losses are overstated, and contends the AI sector may be in a bubble, noting Nvidia's peak market capitalization was roughly 20% of U.S. GDP versus under 6% for Microsoft at the late-1990s tech-bubble peak.
- • Baker proposes alternatives: a higher corporate income tax paid partly in non-voting shares, stronger antitrust enforcement, and shorter workweeks with higher overtime premiums.
Summary
Economist Dean Baker, writing for the Center for Economic and Policy Research (CEPR), publicly disagrees with Senator Bernie Sanders's proposal for an AI sovereign wealth fund. Baker says he admires Sanders's record in pushing American politics leftward, but argues the proposal is flawed on political, economic, and distributional grounds.
On the political side, Baker argues that the Trump administration illustrates the risks of direct federal involvement in private businesses, claiming it uses government power to benefit the president and his family and to pressure private companies over criticism. He says that because voters have elected a figure like this twice, the risk of a similar outcome cannot be dismissed.
On the economics, Baker argues that the job-displacement effect of AI is more prominent in political debate than in the data. He notes that productivity growth, which he treats as the measure of job destruction, has been "extremely weak" over the last two quarters, while acknowledging that quarterly data are erratic. He also argues the AI sector is likely in a bubble, comparing Nvidia's peak market capitalization of roughly 20% of U.S. GDP with Microsoft's peak of less than 6% of GDP during the late-1990s tech bubble. He expects a sovereign wealth fund would mainly transfer value to large AI shareholders, naming Elon Musk and Mark Zuckerberg.
Baker instead recommends three measures if AI displacement materializes: a higher corporate income tax applied to all companies, with firms issuing non-voting shares equal to the tax rate (for example, 25% of shares for a 25% rate); stronger antitrust enforcement, including protection against protectionism that would shield U.S. AI firms from likely lower-cost Chinese competition; and stronger labor standards, such as reducing the standard workweek from 40 to 32 hours or lower and doubling the overtime premium from 50% to 100%. Baker concludes that the tools to manage an AI productivity boom already exist and that the main obstacle is political will.
On the political side, Baker argues that the Trump administration illustrates the risks of direct federal involvement in private businesses, claiming it uses government power to benefit the president and his family and to pressure private companies over criticism. He says that because voters have elected a figure like this twice, the risk of a similar outcome cannot be dismissed.
On the economics, Baker argues that the job-displacement effect of AI is more prominent in political debate than in the data. He notes that productivity growth, which he treats as the measure of job destruction, has been "extremely weak" over the last two quarters, while acknowledging that quarterly data are erratic. He also argues the AI sector is likely in a bubble, comparing Nvidia's peak market capitalization of roughly 20% of U.S. GDP with Microsoft's peak of less than 6% of GDP during the late-1990s tech bubble. He expects a sovereign wealth fund would mainly transfer value to large AI shareholders, naming Elon Musk and Mark Zuckerberg.
Baker instead recommends three measures if AI displacement materializes: a higher corporate income tax applied to all companies, with firms issuing non-voting shares equal to the tax rate (for example, 25% of shares for a 25% rate); stronger antitrust enforcement, including protection against protectionism that would shield U.S. AI firms from likely lower-cost Chinese competition; and stronger labor standards, such as reducing the standard workweek from 40 to 32 hours or lower and doubling the overtime premium from 50% to 100%. Baker concludes that the tools to manage an AI productivity boom already exist and that the main obstacle is political will.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.