What makes an innovation cluster successful? Some regions rely on research universities and venture capital, while others draw power from established industries, manufacturing networks, and government coordination. Knowing these differences is essential for universities and companies seeking international research partnerships.

At a recent seminar hosted by the Stanford Next Asia Policy Lab (SNAPL) at Stanford Asia-Pacific Research Center, three experts from the Swedish Foundation for International Cooperation in Research and Higher Education (STINT) – Executive Director  Andreas Göthenberg, Science Fellow in Japan and South Korea Laura Barbieri, and Representative in China and APAC Erik Forsberg – discussed two leading innovation regions: Tokyo-Yokohama and China’s Greater Bay Area. Drawing on new reports and interviews with regional stakeholders, they shared findings on how each ecosystem developed, where its strengths lie, and how international partners can engage successfully.

Their core insight is that no universal formula exists for scaling innovation. Recognizing how each ecosystem works is the first step toward aligning engagement strategies for successful, high-impact collaboration.

Sign up for APARC newsletters to receive insights and analysis from our scholars and guest speakers > 

Look Beyond Rankings to Understand Regional Strengths

Tokyo-Yokohama has built its position through more than a century of industrial and scientific expertise. Its innovation ecosystem connects major corporate research laboratories, universities, and national research institutes, supported by substantial corporate and public investment.

For years, this region topped the World Intellectual Property Organization’s (WIPO) global innovation cluster rankings. Its move to second place in 2025, due to the introduction of venture capital activity as an indicator, reflected differences in funding methods rather than indicating a decline in research quality.

China’s Greater Bay Area presents a contrasting model. This megacity region systematically combines nine key cities in Guangdong Province along with the Special Administrative Regions of Hong Kong and Macao. Its innovation system developed alongside manufacturing, with companies and industrial supply chains driving technological development before large-scale university research emerged as a defining feature. Shenzhen illustrates this advantage: the city’s electronics markets and supplier networks enable entrepreneurs to easily source components and quickly develop prototypes.

Instead of relying on ranked lists, universities and companies should assess regions according to the specific capabilities they need. While Tokyo-Yokohama offers deep institutional research capacity, China’s Greater Bay Area combines industrial scale with rapid product development.

Connect Research to Industries That Can Scale It

In Tokyo-Yokohama, established institutions and corporate research and development play crucial roles in driving innovation. The environment sustains continuity and long-term growth, but complex governance and limited venture capital can slow commercialization, particularly in deep technology and life sciences.

On the other hand, the Greater Bay Area uses a combination of research and manufacturing. Shenzhen and Guangzhou have robust industrial capacity and an expanding research infrastructure, while Hong Kong brings internationally connected universities and research networks. This evolving network of university campuses and research centers is fostering additional links between these resources.

However, it is important to note that the Greater Bay Area is not a single, uniform ecosystem. Hong Kong, Macao, and mainland Guangdong operate under different legal, educational, funding, and administrative systems. These differences create practical barriers to moving people, money, data, and technology across borders.

The key is to design partnerships around a pathway from research to real-world application. In Japan, this could mean building stronger industry-university connections and developing industrial doctoral programs. In the Greater Bay Area, it means identifying the right jurisdiction and considering cross-border constraints from the start.

Build Relationships Before Expecting Results

Barbieri emphasized that international collaboration in Tokyo-Yokohama is concentrated among leading institutions. Approximately 55% of scientific publications from the cluster involve international co-authors, while around 7% of patent filings include foreign partners or inventors. This suggests that, while international research collaboration is more established, cross-border commercial partnerships are still developing.

Platforms such as Tokyo Innovation Base and SusHi Tech Tokyo can help international partners make those crucial initial connections. Still, sustained collaboration generally requires relationships with universities, research institutes, and individual research groups. Tying projects to Japan’s mission-oriented strategies, including Society 5.0 and Green Innovation, can also assist in finding common goals and potential financial support.

Institutional agreements matter, but they are not enough. Memoranda of understanding can provide legitimacy and facilitate cooperation; trusted relationships between researchers are what make joint projects work.

It is important to combine institutional assistance with researcher-level engagement. International partners should identify specific collaborators, set shared research objectives, and dedicate time to build trust rather than pursuing broad, surface-level engagement.

Treat Internationalization as a Long-Term Strategy

Both regions face challenges with talent mobility, governance, and geopolitics. Japan aims to balance openness with protecting key technologies, all while managing language barriers and a shortage of specialized researchers.

In China, growing research capabilities coexist with elevated security concerns and restrictions that limit some researchers’ international travel. Comprehending this policy landscape is key to successfully navigating collaborations.

Göthenberg’s broader argument was that international engagement must be guided by science intelligence. It requires a systematic, evidence-based analysis to help institutions grasp global research trends, pinpoint opportunities, and anticipate risks.

For universities and companies, this calls for moving beyond instinct or familiar networks when choosing international partners. It requires knowing where expertise lies, how funding and governance operate, and whether a partnership can deliver sustained mutual benefit.

Tokyo-Yokohama and China’s Greater Bay Area demonstrate two distinct routes to innovation at scale. The most effective international strategies recognize these differences, connect complementary strengths, and prioritize focused and durable partnerships instead of seeking broad connections alone.

View STINT's recent reports on global science and innovation clusters in its Outlook library: 

• On the Guangdong-Hong Kong-Macao Greater Bay Area >
• On Tokyo-Yokohama >

Scott Rozelle, Senior Fellow at the Freeman Spogli Institute for International Studies and Faculty Director of the Rural Education Action Program at the Stanford Center on China's Economy and Institutions, has spent decades studying China's rural and low-income populations. Author of Invisible China: How the Urban-Rural Divide Threatens China’s Rise, he examines why China's remarkable economic achievements have not translated into better opportunities for everyone, and what that means for the country's future.

1. How large is China’s low-income population, who are they, and what does economic life actually look like for them?

I think one of the things people outside China often don’t appreciate is just how many people are still struggling economically. We hear a lot about China’s growing middle class, and rightly so. But in 2020, China’s then-premier Li Keqiang pointed out that some 600 million Chinese had an average monthly income of just 1,000 yuan, or about $140. As Li himself observed, that’s barely enough to rent a room in a medium-sized Chinese city.

Who are these people? Many come from rural China. They tend to have relatively little formal education, and they work in construction, manufacturing, and increasingly, informal services. Ride-hailing drivers are one example. Many work 12-hour days or longer, amid growing competition and pressure on earnings.

The big issue is education across this workforce. One study we published shows only about 30% of China’s workforce had been to high school. Now compare that with countries that successfully moved from middle- to high-income status: on average, 72% of their workers had completed high school when they were still middle-income economies. Even countries that got stuck at middle-income levels averaged 36%. So China is starting with a less-educated workforce than many countries that never made the leap.

That’s a huge challenge for China. As the economy becomes more sophisticated, jobs increasingly require skills in math, science, computers, and language. Where are hundreds of millions of workers without those skills going to find decent jobs?

2. How have ordinary people’s views in China about opportunity and getting ahead changed?

For a long time, there was tremendous optimism in China, and for good reason. From the early 1990s through the early 2010s, the economy was growing at roughly 10% a year. Manufacturing and construction were booming, companies needed workers, and wages were rising. People could move from the countryside to the city, find work, and give their children opportunities they never had.

Here’s something interesting about inequality in China. For a long time, the gap between rich and poor was getting wider, but almost everyone was still getting ahead. Research by some of my Stanford colleagues found that between 1988 and 2018, the incomes of the poorest 10% of China’s adults increased fourfold. That’s extraordinary. Of course, people at the very top did much better—their incomes increased something like 30 times over the same period. But even people at the bottom were seeing their lives improve.

I think that helps explain why people tolerated so much inequality. If your own life is improving, you may not worry so much about how much richer someone else is getting. But what happens when that progress slows or stops?

In recent research with Martin Whyte at Harvard, we found a substantial increase in pessimism about future incomes and living standards, especially among people with less education.

Part of the explanation is what’s happened to the labor market. As wages rose, companies increasingly turned to contractors and informal workers to reduce costs. Our research found that informal employment rose from 33% of total employment in 2004 to 56% in 2017. That means fewer benefits, less job security, and more uncertainty.

Today, many less-educated workers are competing for jobs in delivery, ride-hailing, and other informal services. And when more people chase the same jobs, earnings come under pressure.

For decades, people assumed that if they worked hard, their lives would get better. Increasingly, they’re no longer so sure.

3. Are China’s high-tech successes and the struggles of its low-income population separate stories—or two sides of the same economic model?

I would say they’re both. Some of the problems facing low-income workers started long before the current push into high technology. But China’s emphasis on advanced manufacturing has intensified some of those challenges.

Think about what drove China’s growth for the first several decades of reform. China opened up to trade and foreign investment and became the world’s factory. Then came an enormous property boom, supported in part by local governments that depended on land sales for revenue and wanted to meet growth targets.

All of that created tremendous demand for relatively unskilled labor. You didn’t need a college degree to work in a factory or on a construction site. But over time, companies increasingly relied on informal labor, and wages for less-skilled workers began to stagnate.

Now add automation, robotics, and AI.

This past spring, I took a group of Stanford students to visit factories in China. We saw factories producing more than they had five years earlier, but with fewer than 20% as many workers. And nearly everyone on the factory floor had a vocational college or university degree.

We even visited a so-called dark factory. The production floor doesn’t need lights because there aren’t any workers there. Robots and AI do the manufacturing, while a handful of college graduates monitor everything from a control room.

It’s remarkable. China is becoming incredibly good at producing sophisticated goods. But these factories don’t need many workers, and the workers they do need are highly educated.

So China’s technological success is real. But it doesn’t necessarily translate into better opportunities for the hundreds of millions of people who helped build the old economy.

4. As China’s era of property-led growth ends, Beijing is betting on advanced manufacturing, green tech, pharma, and other new industries. Are they creating jobs for workers from the old economy?

The short answer is no—not nearly enough, and certainly not for the workers who need those jobs most.

China’s manufacturing output continues to grow, but manufacturing employment began declining in the mid-2010s. That’s what happens when factories become more automated: you can produce more with fewer people.

But the much bigger shock has come from the collapse of the property sector. For years, construction was an enormous source of employment for workers with relatively little education. Now that sector has contracted dramatically. Official figures indicate that roughly 10 million migrant workers left construction in 2022 and 2023 alone.

So where do these workers go?

You might think they could return to the countryside and start farming again. But nearly 80% of rural workers under 40 have never farmed. They don’t know how, and they have no intention of doing so.

Instead, many are moving into the informal economy—delivering food, driving cars, doing whatever work they can find. As more people enter these sectors, competition increases and earnings come under pressure.

Meanwhile, the new industries China is investing in—electric vehicles, semiconductors, pharmaceuticals—need relatively few workers, and many of those jobs require specialized skills.

So there’s a huge mismatch. China is investing heavily in the industries of the future, but those industries aren’t creating enough opportunities for workers displaced from the industries of the past.

And that matters enormously. If hundreds of millions of people see their incomes stagnate or decline, that’s going to affect consumption, economic growth, and potentially social stability.

5. Can China sustain its economic success without substantially improving the prospects of “the rest”?

I think this is really the central question.

China has accomplished extraordinary things. It’s a global leader in electric vehicles, batteries, solar panels, and increasingly robotics and AI. Nobody should underestimate those achievements.

But there’s a difference between having world-class industries and becoming a high-income economy.

Look at countries that successfully made that transition—South Korea, Singapore, Ireland, and others. They had relatively well-educated workforces. They weren’t just producing sophisticated goods; they were creating economies in which large shares of the population had the skills to participate in higher-productivity work.

China faces a much harder challenge because hundreds of millions of its workers have relatively little education.

And there’s another issue: China’s government does much less than governments in other countries to redistribute income. One study found that taxes and transfers reduce the income share of China’s richest 10% by only about 4%. In the United States, that reduction is closer to 20%, and in France it’s about 25%.

And redistribution isn't just about fairness. Lower-income households tend to spend a larger share of any additional income they receive, while wealthier households are more likely to save it. That means putting more money in the hands of poorer families could also help China boost household consumption

So could China do more through its tax system, social insurance, and public services, rather than relying so heavily on growth to lift people’s incomes?

If a large share of the population isn’t seeing meaningful income growth, household consumption will remain constrained. That makes it harder to sustain growth through domestic demand. And growing pessimism about the future creates social and political challenges as well.

The World Bank has shown that relatively few countries successfully escape the middle-income trap. Is China headed in that direction? Not inevitably. It has enormous strengths, including its technological capabilities, infrastructure, and manufacturing base.

But I don’t think China can simply innovate its way out of this problem. It also has to improve the education, skills, and economic opportunities of the hundreds of millions of people who aren’t participating in the high-tech boom.

China’s future won’t be determined just by how successful its best companies become. It will also depend on what happens to the rest of its population.

This article was edited for clarity and brevity with support from ChatGPT. All outputs were reviewed by humans and the final expression was produced by the authors.