Network-Based Industrial Policy: Structural Bottlenecks and Policy Transmission Channels in Türkiye's Production Network

Economic policy often feels like a game of pinball, where governments aim a reform at a specific industry, hoping it will bounce through the economy and create jobs or lower prices. For decades, the standard approach has been to pick a target—like manufacturing, technology, or agriculture—and pour resources into it, assuming the benefits will naturally spread. This method treats the economy as a collection of separate rooms, where fixing the plumbing in the kitchen doesn't necessarily affect the bedroom. However, modern production is far more interconnected. Factories do not operate in isolation; they rely on a constant stream of inputs from other industries, from the electricity that powers their machines to the financial services that fund their expansion. When a policy changes conditions in one sector, the effects travel along these invisible threads, sometimes amplifying, sometimes getting stuck, and sometimes vanishing entirely before reaching the intended destination. Understanding these connections is crucial because a reform that looks perfect on paper can fail if the network of relationships around the target sector is broken or blocked.

A new study by economist Tugrul Temel shifts the focus from the target itself to the web of connections surrounding it. Instead of asking simply which sector needs help, the research asks how that sector is wired into the rest of the economy. The study uses a detailed map of Turkey's production network from 2018, treating the economy as a directed system where arrows show the flow of goods and services from suppliers to users. By tracing these paths, the author identifies where policy effects are likely to travel, where they might get trapped, and which specific links hold the entire structure together. The central finding is that the success of an industrial policy depends less on the target sector's own performance and more on the health of the sectors it relies on and the sectors that rely on it. If a key supplier is inefficient or heavily regulated, a reform in the manufacturing sector may never fully take hold, no matter how well-designed the intervention is.

The researcher applied this framework to Turkey's manufacturing sector, specifically the group of industries that excludes petroleum and refining. They found that this sector is deeply entangled with a handful of other industries that act as critical gateways. The analysis revealed that finance and agriculture are potential bottlenecks; if these sectors are constrained, the effects of manufacturing reforms struggle to pass through. Conversely, construction and energy emerged as important absorbing sectors, meaning they play a vital role in taking in the outputs of manufacturing and distributing them further. The study mapped these relationships into three distinct layers of connection. The first layer consists of immediate neighbors like finance and agriculture, which are directly linked to manufacturing. The second layer includes sectors like mining and transport, which are one step removed. The third layer contains more distant links, such as professional services. This layering matters because a shock or a policy change in a distant sector must pass through the intermediate layers to reach the target. If those middle layers are weak or rigid, the signal gets lost before it arrives.

One of the most significant discoveries is that the economy is not a single, uniform mass but is divided into distinct communities or clusters of industries that interact more intensely with each other than with the outside world. In Turkey's network, the researcher identified three such groups. One cluster includes manufacturing, construction, and energy; another groups together mining, agriculture, and oil refining; and a third links transport and finance. The connections between these groups are few and fragile. The study calculated a measure of structural resilience, which turned out to be moderate, with a score of 0.43. This number suggests that while the network is not entirely dependent on a single bridge, it lacks enough alternative routes to easily absorb a major disruption. If the links connecting these communities were to break, the network would become significantly less connected, with a resilience score dropping to 0.27 in a hypothetical scenario where intermediate layers failed. This highlights that the economy's ability to withstand shocks depends as much on the bridges between communities as it does on the strength of the communities themselves.

The paper also challenges the traditional way of thinking about competition and regulation. It argues that a competitive manufacturing sector can still be held back if it relies on a regulated or concentrated upstream supplier, such as a utility or a financial institution. Even if the manufacturer is efficient and innovative, high costs or limited access in the upstream sector can absorb the gains, preventing them from reaching the wider economy. Conversely, a reform in a seemingly peripheral sector, like improving competition in finance, could unlock productivity gains across many downstream industries, even if finance itself is not the primary target of industrial policy. The study does not claim that every regulated sector is a problem or that every concentrated market is harmful. Instead, it suggests that policymakers must look at the structural position of a sector. A small sector can become a massive bottleneck if many important pathways depend on it, while a large sector might have little impact if the connections around it are weak.

The research relies on data from the Organization for Economic Cooperation and Development, which tracks the flow of goods between 15 broad sectors in Turkey. The author used a method to filter out trivial connections, keeping only the most significant links that account for at least 20 percent of the relevant economic flow. This approach allowed them to see the skeleton of the economy without getting lost in the noise of minor transactions. They also looked at how these connections changed over time, from 2005 to 2018, finding that the relationships between manufacturing, finance, and agriculture have remained stable and persistent. This stability suggests that the bottlenecks identified are not temporary glitches but structural features of the Turkish economy that will likely persist unless addressed.

The implications for policy are profound. The study suggests that governments should stop treating industrial policy as a series of isolated interventions. Instead, they should view it as a process of network coordination. If a country wants to boost manufacturing, it must simultaneously assess the conditions in finance, energy, transport, and agriculture. A reform in one area might fail if a bottleneck in another area blocks the transmission of benefits. The author proposes a diagnostic approach where policymakers first map the network, identify the critical pathways, and then coordinate reforms across the different sectors that form those pathways. This does not mean intervening in every connected market, which could create new distortions, but rather focusing on the specific links that carry the most weight. By understanding the network, policymakers can distinguish between a policy that failed because the tool was wrong and a policy that failed because the transmission channel was broken.

Ultimately, the study offers a new way to think about economic resilience. It shows that protecting a single sector is not enough; the entire web of connections must be maintained. The research suggests that the ability of an economy to recover from a shock depends on having alternative routes and flexible connections. In Turkey's case, the moderate resilience score indicates that while the network is functional, it is vulnerable to disruptions in the specific bridges that connect its different communities. The author emphasizes that their findings are a starting point for deeper investigation. They do not prove that a specific regulation caused a specific outcome, but they provide a map of where to look for those causes. By turning the production network into a policy object, the study invites a more nuanced, connected, and effective approach to economic management, one that recognizes that in a modern economy, nothing happens in isolation.