Thanks Simon! These words mean a lot to me!

🔵 𝗪𝗵𝗲𝗿𝗲 can you read more about this paper and my research?

👉 Check out my website: fabrizioleone.github.io

And do get in touch if you have questions!

Thanks for reading!

(8/8)

🔵 𝗪𝗵𝗲𝗻 thinking about policies, what should policymakers do?

👉 Consider the responses of 🤖 sellers. The optimal 🤖 tax/subsidy can change when doing so

👉 Be aware of the spillover effects induced by unilateral policy changes

(7/8)

🗝️𝗞𝗲𝘆 𝗿𝗲𝘀𝘂𝗹𝘁𝘀🗝️

1⃣ 𝗔𝗺𝗽𝗹𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻: MNEs' responses amplify by 40% the aggregate & distributional effects of policy interventions

2⃣ 𝗦𝗽𝗶𝗹𝗹𝗼𝘃𝗲𝗿: Local policies can produce global outcomes if markets are linked via MNEs' global sales networks

(6/8)

I quantify the effects of taxing/subsidizing 🤖 adoption under  2 scenarios:

👉𝗕𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸: MNEs 𝗰𝗮𝗻'𝘁 change market entry and pricing choices after the policy change

👉𝗡𝗲𝘄: MNEs 𝗰𝗮𝗻 change market entry and pricing choices after the policy change 🆕

(5/8)

🔵 𝗪𝗵𝗮𝘁 do I do and find?

👉Collect new data on MNEs producing 🤖 and their global sales

👉Develop and estimate a multi-country general equilibrium model of 🤖 adoption➕oligopolistic MNEs competing to sell 🤖

👉Simulate currently debated policy counterfactuals

(4/8)

🔵 𝗪𝗵𝗼 are these MNEs?

👉 They have HQ mainly in 🇯🇵 and central 🇪🇺, where >50% of new 🤖 are produced every year

👉 They have local branches that sell 🤖 ➕ "integration services" (e.g., customization) to users worldwide

👉 Stable set of MNEs over time

(3/8)

This paper: 𝙨𝙪𝙥𝙥𝙡𝙮-𝙨𝙞𝙙𝙚 responses can change what we know about the effects of these policies‼️

I study the global market of 🤖. Ideal setting:

👉🤖 are a leading type of automation technology

👉𝟭𝟬 multinational enterprises (MNEs) hold 𝟵𝟬% of global sales

(2/8)

🔵 𝗪𝗵𝘆 do we care?

Automation (e.g., 🤖 and AI) is on the rise 📈

Trade-off: productivity gains vs job displacement ➡️ debate about policy interventions (tax or subsidies)

Current debate focuses on adopters, i.e., the 𝙙𝙚𝙢𝙖𝙣𝙙 𝙨𝙞𝙙𝙚 of the automation industry

(1/8)