The Coordinating State in a Fragmented Production Network: A Coordination–Network DSGE Model for the Vietnamese economy.

In the modern economy, growth is rarely about a single factory building a better product or a single farmer harvesting a larger crop. It is about how thousands of different pieces fit together. A business in one city needs parts from another; a researcher needs data from a government database; a farmer needs a loan to buy seeds that a bank in a different province might provide. When these connections work smoothly, the whole system becomes more productive. But when they are clogged by bureaucracy, incompatible computer systems, or gaps in responsibility, the economy slows down. This is not just a matter of having enough money or good ideas; it is a matter of coordination. For decades, economists have tried to measure how well a government helps these pieces connect, often treating the quality of institutions as a fixed background condition, like the weather. However, a new study suggests that the ability of a state to coordinate these complex networks is not a static trait, but a dynamic resource that can be built up, worn down, or even broken by shocks, much like a physical machine.

This study focuses on Vietnam, a country deeply integrated into global supply chains and currently undergoing rapid digital and industrial transformation. The researchers, led by Giang Thi Thuy Le, developed a complex computer simulation to understand how the government can best act as a "coordinator" in a fragmented production network. They did not simply ask if the government should spend more money. Instead, they built a model where the economy is a vast web of 63 provinces and various industries, all linked by the flow of goods, data, and knowledge. In this model, the government's role is to manage specific "states" or assets: the capacity to coordinate between different agencies, the level of institutional fragmentation (or how broken the system is), the availability of public data, and the existence of shared research infrastructure. These are not just abstract concepts; in the model, they are things that can accumulate over time, depreciate, or be damaged by shocks, directly affecting how easily businesses can trade and innovate with one another.

The researchers used data from 2020 to 2024, covering all 63 provinces and cities of Vietnam at the time, to test their ideas. They combined official economic statistics with governance indicators, such as the Provincial Competitiveness Index, which measures business perceptions of transparency and time costs, and the Provincial Digital Transformation Index. They constructed a new measure called the Coordination Capacity Index to see if better coordination actually leads to better economic outcomes. However, the study is careful to distinguish between what it has proven and what it has only suggested. The data available so far is too short to definitively prove that specific new policies enacted in 2024 and 2025 have already changed the economy. Instead, the study uses this data to build a framework for understanding how the system should work and to run simulations of what might happen under different scenarios.

The model proposes five theoretical outcomes regarding how coordination might work. First, the value of a government's ability to coordinate is not the same everywhere. It is much higher in the parts of the economy that are central to the network—places or industries that many other businesses rely on. Improving coordination in a central hub could ripple through the entire system, boosting productivity far beyond that single location, whereas improving coordination in a peripheral area might have a smaller effect. Second, the study suggests that simply throwing money at research and development is not enough. Seed capital for innovation works best when it is paired with other public assets, such as shared research labs, accessible data, and a government that can effectively manage the process. These elements act as complements; one without the others is far less effective.

The researchers also identified a significant risk in how coordination is organized. While having a single, powerful central hub to manage the network can be efficient, it creates a "single point of failure." If that central hub lacks proper data, auditing, or backup systems, a shock to that one point could spread through the entire network, causing widespread disruption. The model suggests that the best approach is not to return to a fragmented system where everyone works alone, but to maintain a central responsibility while ensuring that the data, monitoring, and execution mechanisms are robust and transparent. This prevents the central hub from becoming a bottleneck that collapses the whole system.

In terms of what the data actually shows, the results are cautious. When the researchers looked at the relationship between their new Coordination Capacity Index and economic growth, they found that three exploratory regressions did not identify non-zero coefficients; the link was not statistically significant. The data did not provide strong evidence that better coordination automatically leads to higher growth in the short term, nor did it prove that specific new government resolutions have already boosted the economy. The study explicitly states that it cannot yet claim to have measured the causal impact of recent policies. Instead, the value of this work lies in the architecture it provides. It offers a way to measure the invisible "friction" between different parts of the economy and a method to simulate how different policy choices might play out before they are implemented.

The study proposes a new way for policymakers to think about spending. Rather than allocating resources based solely on the size of a region's economy or the number of projects, the model suggests prioritizing investments where they will have the most network-wide impact. This involves looking at how central a sector is, how sensitive it is to coordination problems, and how efficiently the government can turn budget money into useful assets without loss or waste. For example, in the Mekong Delta, the model suggests that policies should focus on the entire agricultural network—from climate data and logistics to finance and processing—rather than just increasing the yield of a single crop. In the northern mountainous regions, the focus should be on shared infrastructure that connects isolated markets, though the model warns that such investments must be balanced with equity to ensure that poorer areas are not left behind.

Ultimately, this research reframes the role of the state. It argues that the government's most valuable contribution is not necessarily in controlling production or picking winners, but in building the invisible infrastructure that allows the private sector to connect. This includes creating compatible standards, sharing data, funding open research, and managing the risks of innovation. The study concludes that the success of economic reform should not be measured by how much the government spends, but by how well it connects the resources of society into a productive, resilient whole. While the current data does not yet offer a definitive answer on the impact of recent Vietnamese policies, the model provides a clear, testable framework for understanding the complex machinery of a modern, networked economy. It suggests that the path forward lies in treating coordination not as a fixed background condition, but as a dynamic asset that requires constant investment, monitoring, and careful design to avoid the pitfalls of centralization and fragmentation.