logo
The G7 has once again put multinationals' profits over the interests of people

The G7 has once again put multinationals' profits over the interests of people

The Guardian3 days ago
The US Treasury just made a deal with the other G7 countries that global minimum taxes that were already agreed upon will not apply to American companies. The G7 governments caved under intense pressure from President Donald Trump and lobbying from multinationals in Washington, London, Brussels, and beyond – just as India, and now, sadly, Canada have caved on digital taxation.
Years ago, the international community recognised that too many global companies were not paying their fair share of taxes, and some weren't paying taxes to the country where the economic activity actually occurs. The complex agreement that emerged in 2021 at the OECD/G20 Inclusive Framework on Base Erosion and Profit Shifting comprised two pillars; only Pillar Two, a global minimum corporate tax, has been adopted. (The other pillar allocated taxation rights among countries and spurred opposition from developing countries and the US.)
While there has been a global consensus on the need for such a minimum, the version the US adopted during Trump's first presidential term was different, and weaker, than that of the rest of the world, allowing multinationals to 'make up' for what they didn't pay in tax havens with the 'extra' they paid in the US or other high-tax jurisdictions.
While far from perfect, Pillar Two was a first attempt to ensure a minimum tax rate of 15% on the profits of multinationals everywhere, a crucial step to end harmful tax competition between countries.
There were, of course, some carve-outs and exemptions, which lowered the effective rate somewhat below 15%. And the 15% rate was already lower than the rate imposed by many developing countries; it should have been higher, and the carve-outs smaller. Still, the Pillar Two deal halted the race to the bottom, whereby countries offered lower tax rates to attract businesses to their jurisdictions. For the world as a whole, this race didn't generate much new investment; the real winners were the rich corporations who pocketed the savings from paying almost no taxes at all in some countries.
But once again, G7 governments have decided to put multinationals' interests before the interests of developing countries, small and medium-size businesses (which can't avail themselves of the shenanigans that multinationals have found so profitable), and their own citizens – who, as a consequence, will pay higher taxes. By exempting US multinationals from Pillar Two, this deal will allow some to continue to benefit from zero or near-zero taxes on profits they book in low-tax jurisdictions or tax havens such as Puerto Rico and the Cayman Islands. This will make them more competitive than non-US multinationals. Because modern multinational corporations are willing to move their nominal headquarters to wherever they get the most favourable tax treatment (and other goodies), with the real economic activity occurring elsewhere, giving US companies preferential treatment incentivises companies to move their official headquarters to the US. This is another sad example of a race to the bottom.
By acceding to US demands, the G7 deal risks undermining the worldwide implementation of the minimum tax. It also makes a mockery of the inclusiveness of the OECD/G20 Inclusive Framework.
There was a pretence that the new global framework was crafted by more than 140 countries working together. To be sure, many developing countries complained this was an unfair agreement for them and that powerful countries did not listen to their concerns. Now that facade has crumbled. The non-G7 countries, including dozens of emerging markets and developing countries, are now being asked to rubber-stamp a decision imposed on them by just one country.
Pillar Two should be strengthened, not gutted. It currently applies only to large multinationals (with a global turnover at or above €750m), and the global minimum tax rate of 15% is set very low. The Independent Commission for the Reform of International Corporate Taxation has always advocated a minimum rate of at least 25%.
According to some estimates, Pillar Two's minimum tax would have yielded between $155bn and $192bn (£112bn-£140bn) annually in additional global corporate income tax revenue. While this is a significant amount, a minimum rate of 25% could generate more than $500bn a year in additional revenue. In a world facing converging crises of inequality, climate change, and underfunded public services, leaving such substantial resources on the table is fiscally irresponsible and morally indefensible.
Pillar Two represented a starting point – a global floor on corporate taxation that could have curbed the race to the bottom and restored some degree of tax justice. The G7's decision to let US multinationals off the hook weakens even that modest floor and sends the wrong message to the rest of the world.
Just two weeks ago at the UN, there was a global consensus about the need to strengthen international tax cooperation and to implement progressive tax systems, and a large majority of countries voted for and support ongoing negotiations toward a UN framework convention on international tax cooperation. But the US government recently walked away from the UN negotiations, stating that the goals of the proposed UN convention 'are inconsistent with US priorities and represent an unwelcome overreach'.
In the adoption of the 'Compromiso de Sevilla,' the outcome document of this week's UN Fourth International Conference on Financing for Development (FfD4), the US was the only major country that was absent. Allowing the US to bypass the already modest Pillar Two rules not only undermines multilateralism; it also flies in the face of the commitments that have been made, and further deepens the inequity in global tax governance.
The members of the OECD/G20 Inclusive Framework should reject the deal made at the G7. The US must not be allowed to dictate global policy. It is powerful, but still represents less than 20% of global GDP.
Countries meeting in Seville for FfD4 can either accept the US undermining every effort to ensure multinationals pay their fair share, or redouble efforts to create a new international tax system at the UN that works for all. For the sake of the world economy and people everywhere, they should do the latter.
Joseph E Stiglitz is a Nobel laureate in economics, a university professor at Columbia University and a former chief economist of the World Bank.
José Antonio Ocampo is professor at Columbia University and former finance minister of Colombia.
Jayati Ghosh is professor of economics at University of Massachusetts Amherst.
Ⓒ Project Syndicate
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

The 10 moments that defined Keir Starmer's first year in office
The 10 moments that defined Keir Starmer's first year in office

The Independent

time33 minutes ago

  • The Independent

The 10 moments that defined Keir Starmer's first year in office

Sir Keir Starmer celebrates his first year as prime minister with the largest parliamentary majority in a quarter of a century — a commanding mandate that has brought both opportunity and scrutiny. His tenure so far has been defined by cautious reform and major international resets. From securing a landmark US-UK trade deal with Donald Trump and a string of policy U-turns that have tested party discipline, here are the 10 most significant moments to reflect on his first year.

Two popular apps loved by millions to shut down forever in DAYS – and you might be owed some cash
Two popular apps loved by millions to shut down forever in DAYS – and you might be owed some cash

The Sun

time40 minutes ago

  • The Sun

Two popular apps loved by millions to shut down forever in DAYS – and you might be owed some cash

TWO popular apps used by millions of mobile users are set to shut down next week - and it might mean you're owed some cash. Mobile apps Pocket and Glitch will both close on 8 July. 3 Pocket was removed from app stores in late May, when sales of subscriptions also stopped. But Pocket's Premium subscribers may be owed refunds for the remaining time left on their subscriptions. A subscription for the read-it-later app costs $4.99 per month or $44.99 (£32.99) per year. So if you paid for an annual membership that will not be fulfilled, you should expect some money to be deposited back onto the card you used for payments. Mozilla, which owns the app, said annual subscribers will receive their refunds after 8 July. No action is necessary - just wait for the money to land. It's worth noting that while the app will shut down next week, users will have until 8 October to download their saved data before all Pocket data is permanently deleted. The Glitch app, aimed at web developers, is also handing out refunds to its subscribers. Those who forked out $96 (£70) for an annual Glitch Pro subscription and have paid time left will receive a refund. Users have until the end of the year to download their projects before all the data is wiped for good. Huge Global Data Breach: 16 Billion Accounts at Risk 3

Model Context Protocol (MCP) Explained With Code Examples)
Model Context Protocol (MCP) Explained With Code Examples)

Geeky Gadgets

timean hour ago

  • Geeky Gadgets

Model Context Protocol (MCP) Explained With Code Examples)

What if the way AI agents interact with tools and resources could be as seamless as browsing the web? Imagine a world where developers no longer wrestle with custom-built adapters or fragmented integration processes, and AI systems effortlessly connect with APIs, databases, and services. Enter the Model Context Protocol (MCP)—a new framework poised to redefine how AI agents handle complex tasks. By introducing a standardized approach to integration, MCP eliminates inefficiencies, enhances scalability, and simplifies workflows. It's not just a technical upgrade; it's a paradigm shift that could transform the future of AI-driven automation. In this exploration, Assembly AI unpack the core principles and benefits of MCP, revealing how it addresses long-standing challenges in AI-agent integration. From reducing developer workloads to allowing seamless interoperability, MCP offers a unified solution to some of the most pressing issues in AI development. You'll also discover real-world examples that illustrate its potential, such as AI agents performing intricate workflows without the need for custom code. Whether you're a developer, a service provider, or simply curious about the next evolution in AI, this journey into MCP will leave you rethinking what's possible in the realm of automation and innovation. Overview of Model Context Protocol Purpose and Vision of MCP MCP was developed to standardize the interaction between AI agents and external tools or services. Traditionally, developers have relied on custom-built adapters for each service, which increases complexity, maintenance demands, and development time. MCP eliminates these challenges by introducing a unified protocol that enables seamless connections between AI agents and diverse resources. This standardization not only reduces inefficiencies but also unlocks new opportunities for automation and task execution. By providing a consistent framework, MCP enables developers to focus on creating innovative AI solutions rather than managing the intricacies of integration. It also fosters a more collaborative ecosystem where service providers can design tools that are inherently compatible with AI agents, further enhancing the potential for automation and scalability. Challenges Addressed by MCP AI agents, particularly those powered by large language models (LLMs), often encounter significant hurdles when performing multi-step, high-precision, or complex tasks. Existing integration methods exacerbate these challenges by requiring bespoke solutions for each tool or service. This fragmented approach introduces several key obstacles: High Maintenance Demands: Custom-built adapters for individual tools require ongoing updates and troubleshooting. Custom-built adapters for individual tools require ongoing updates and troubleshooting. Security Risks: Inconsistent integration methods can create vulnerabilities and expose sensitive data to potential threats. Inconsistent integration methods can create vulnerabilities and expose sensitive data to potential threats. Limited Scalability: Fragmented processes hinder the ability to scale AI systems efficiently across diverse applications. MCP addresses these issues by shifting the integration responsibility to service providers and offering a standardized framework. This reduces the need for translation layers, simplifies development workflows, and enhances the overall reliability of AI-agent interactions. Understanding Model Context Protocol (MCP) Watch this video on YouTube. Stay informed about the latest in Model Context Protocol (MCP) by exploring our other resources and articles. How MCP Operates MCP employs a client-server architecture to assist interactions between AI agents and external resources. In this model, service providers manage MCP servers that expose tools and capabilities, while AI agents act as clients accessing these tools through the protocol. This architecture introduces several key advantages: Reduced Developer Workload: Service providers handle MCP server management, allowing developers to focus on building AI functionalities. Service providers handle MCP server management, allowing developers to focus on building AI functionalities. Decoupled Implementation: AI agents are abstracted from the specific technical details of external services, allowing more flexible integrations. AI agents are abstracted from the specific technical details of external services, allowing more flexible integrations. Minimized Translation Layers: Traditional API interactions often require complex translation layers, which MCP eliminates or significantly reduces. For example, an AI agent using MCP can interact with a document creation tool without requiring a custom adapter for each API. This abstraction ensures that the integration process is not only more efficient but also more reliable, allowing AI agents to perform tasks with greater precision and consistency. Core Benefits of MCP The adoption of MCP offers several fantastic benefits that enhance the development, functionality, and scalability of AI systems: Simplified Integration: Developers no longer need to create and maintain custom adapters for individual tools or services, reducing time and resource investment. Developers no longer need to create and maintain custom adapters for individual tools or services, reducing time and resource investment. Enhanced Interoperability: MCP enables seamless interaction between AI agents and a wide range of external resources, fostering a more connected ecosystem. MCP enables seamless interaction between AI agents and a wide range of external resources, fostering a more connected ecosystem. Improved Scalability: By decoupling functionality, MCP supports the growth of scalable AI ecosystems capable of handling diverse and complex tasks. By decoupling functionality, MCP supports the growth of scalable AI ecosystems capable of handling diverse and complex tasks. Flexible Tool Composition: AI agents can combine multiple tools to execute intricate workflows more effectively, enhancing their versatility. AI agents can combine multiple tools to execute intricate workflows more effectively, enhancing their versatility. Increased Security: Standardized protocols reduce vulnerabilities associated with ad hoc integration methods, making sure safer interactions. These advantages position MCP as a critical enabler for the future of AI-driven automation and tool integration, making it an essential framework for developers, service providers, and organizations alike. Applications and Real-World Use Cases MCP has already demonstrated its potential in various practical scenarios, showcasing its ability to simplify tool usage and enhance AI-agent capabilities. For instance: Document Management: An AI agent using MCP can integrate with the Google Docs API to generate, edit, and upload documents. By abstracting API interactions, MCP allows the agent to focus on the task rather than the technical details of integration. An AI agent using MCP can integrate with the Google Docs API to generate, edit, and upload documents. By abstracting API interactions, MCP allows the agent to focus on the task rather than the technical details of integration. Mathematical Operations: An AI agent can perform arithmetic calculations by accessing a calculation tool via MCP. This eliminates the need for custom code to handle each mathematical function, streamlining the process. An AI agent can perform arithmetic calculations by accessing a calculation tool via MCP. This eliminates the need for custom code to handle each mathematical function, streamlining the process. Data Analysis: MCP enables AI agents to connect with data visualization tools, allowing them to generate insights and create visual reports without requiring specialized adapters. These examples highlight how MCP simplifies the integration process, making AI agents more versatile and effective in handling diverse tasks across industries. Broader Implications and Future Potential MCP has the potential to become a foundational protocol for AI-agent ecosystems, much like HTTP/HTTPS serves as the backbone of the web. Its standardization encourages widespread adoption by service providers, fostering robust and scalable interactions between AI agents and external resources. As organizations such as OpenAI and others embrace MCP, it is likely to emerge as the universal standard for AI-agent resource integration. Looking ahead, MCP's ability to streamline integration, enhance security, and enable complex task execution positions it as a cornerstone of AI-driven automation. Its evolution will likely play a pivotal role in advancing AI capabilities, shaping interoperable ecosystems, and driving innovation across industries. Media Credit: AssemblyAI Filed Under: AI, Top News Latest Geeky Gadgets Deals Disclosure: Some of our articles include affiliate links. If you buy something through one of these links, Geeky Gadgets may earn an affiliate commission. Learn about our Disclosure Policy.

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into a world of global content with local flavor? Download Daily8 app today from your preferred app store and start exploring.
app-storeplay-store