Sitting on the outskirts of the small Japanese town of Hitachinaka, 75 miles north of Tokyo, is a compact slate-colored plant that, six years ago, nearly brought the world auto industry to a grinding halt.
Owned by the semiconductor giant Renesas Electronics, the factory primarily makes microcontrollers, tiny computer chips as light as a dime that typically cost only a dollar or so. Yet when production was suspended in the wake of the powerful earthquake that rattled Japan the afternoon of March 11, 2011, much of the country’s car production slowed to a crawl, and it wasn’t long before the United States and Europe began to feel the squeeze. In just three months the company lost approximately $156 million, much of it due to the Hitachinaka breakdown.
The effects were so far-reaching that business partners and even some of Renesas competitors rushed to its aid. Its customers, including major automakers like Toyota, dispatched hundreds of workers to help repair the plant’s cracked walls and crippled machinery and later contributed some $145 million to save Renesas from default. Competing companies, too, realized that the collapse of a key industry supplier could wreck their business and diverted equipment to the factory to forestall a shutdown.
Renesas eventually recovered, but its story illustrates just how fragile and complex today’s business networks can be. Economists have long known that natural disasters can send powerful shocks rippling through the global economy, but so far many have struggled to quantify their strength or determine just how far they can travel.
A new working paper examining the aftermath of the 2011 Japanese earthquake sheds fresh light on that question by measuring how the shock spread through the Japanese economy to both direct partners of disaster-hit companies like Renesas and firms further away in the supply chain. It found remarkably strong impacts that percolated through supply chains to sellers and customers alike, disrupting a wide range of industries.
“The presence of these effects is no longer just theoretical, but something we can actually measure,” said one of the study’s authors, Alireza Tahbaz-Salehi, an associate professor at Columbia Business School. The earthquake was both exogenous and highly localized, he explained, which offered a particularly advantageous opportunity to study how disasters like this spread through the economy, according to Tahbaz-Salehi, who conducted the study with researchers from the University of Cambridge; the Japanese Ministry of Finance; and the Research Institute of Economy, Trade and Industry, a Japanese policy think tank.
To study the earthquake’s effects, the authors mapped out concentric networks of firms located upstream and downstream from disaster-hit companies on the supply chain. Customers that relied on impacted firms directly — as Toyota did on Renesas — predictably felt the greatest impact; compared to a control group of firms with no direct or indirect exposure to the disaster, their growth rate was on average 2 percentage points smaller in the year after the disaster. The upstream supplier firms felt a smaller jolt, but it was still significant, with sales growth that was 1.2 percentage points smaller than that of the control group in the following year.
But firms did not need to have direct business partners in disaster areas to be affected, Tahbaz- Salehi said. In fact, much of the economic damage was indirect. Tight business relationships among firms meant that supply chains transmitted the shocks further and further away, to firms’ customers’ customers and suppliers’ suppliers. In fact, downstream firms up to four business relationships removed from disaster-hit companies still experienced a noticeable drop in sales growth, equal to 1.1 percentage points. That effect was 0.1 percentage point for similarly removed upstream firms.
A map of the geographic distribution of headquarters locations of firms located in the diaster-stricken prefectures of Aomori, Fukushima, Iwate, and Miyagi, and the headquarters of firms one (bright red), two (orange), three (green), and four (blue) links away on the supply chain, either upstream or downstream. The blue curve represents the boundary of the four disaster-stricken prefectures.
© Vasco M. Carvalho, Makoto Nirei, Yukiko U. Saitom, and Alireza Tahbaz-Salehi
Summing up all those effects offers a more complete picture of the national economic hangover the earthquake created. The authors found that supply chain disruptions caused by the quake may have knocked as much as 1.2 percentage points off Japan’s gross output in the following year, an effect far greater than the economic output of the disaster-hit region would suggest.
For all their manufacturing activity, the four coastal prefectures that suffered the heaviest damage (Aomori, Fukushima, Iwate, and Miyagi) actually produce a rather modest share of Japan’s aggregate output — only 4.7 percent or so. The earthquake cut their GDP growth rate from 1.4 percent in 2010 to -1.7 percent in 2011, a reduction of 3.1 percentage points. Based solely on the four prefectures’ economic weight, the shock should have translated to a mere 0.15 percentage-point decline in aggregate GDP growth. The actual overall decline, however, was much greater: around 0.8 percentage points.
That difference indicates that supply-chain linkages can amplify the economic damage of events like natural disasters by as much as a factor of five, Tahbaz-Salehi explained. “The question was whether these disruptions had a broader indirect effect on the Japanese economy at the macro level. And the answer seems to be yes,” he said. To the same extent that an interconnected economy can propel growth, then, it can also hamper it.
The Global Factory
Eighty years ago, cars were built in giant, vertically integrated assembly plants on the scale of Ford’s legendary River Rouge factory in Michigan. Even raw materials that weren’t directly produced on site, like rubber, were often sourced nearby. Now, however, hundreds of specialized manufacturers like Renesas make thousands of parts that are funneled through global supply networks of immense complexity.
This change has driven down prices, as suppliers of components have specialized, gained expertise, and achieved economies of scale. Renesas’ microcontrollers, for instance, now make up the nerve network of many new cars, responsible for firing fuel injectors, engaging antilock brakes, and powering windows. The rise of lean, or “just-in-time,” manufacturing has made automakers and other manufacturers even more reliant on predictable, uninterrupted flows of intermediate parts to fuel their operations.
These delicate chains are made even more fragile by minor rigidities created by firm-specific technology and designs. Because car giants like Toyota and Nissan employ proprietary software customized to work only with specific chips, they tend to depend heavily on individual hardware suppliers, making it challenging for them to swap in times of crisis. As a result, events like the Japanese earthquake can create severe bottlenecks, which bubble up to the wider industry.
In this respect, quantifying the effects of supply-chain shocks provides useful reference for both firms and governments, Tahbaz-Salehi explained. For firms, it’s a reminder of the magnitude of new risks to their business models and the importance of managing them. Japanese automakers, for instance, have introduced better crisis-response systems after the 2011 crisis, from the very basic — Renesas has further earthquake-proofed its clean rooms — to the more complex, like the rigorous tracking of parts and weak links on the supply chain. Making such precautions a worldwide industry standard, however, remains a long way off.
On the broader policy level, the study underscores the need for multifaceted disaster recovery plans. “When it comes to government bailouts, for example, it’s important to consider the damage to the entire industrial ecosystem,” said Tahbaz-Salehi. “Severe disruptions may not remain confined to the firms directly impacted. They can also have large, and sometimes distant, indirect effects.”
The exact scope of government response, of course, depends on the exact nature of supply-chain breakdown, and the ways in which those breakdowns can occur are swiftly multiplying.
As the climate changes, natural disasters on the scale of Hurricane Katrina (which damaged 19 percent of American oil production) are growing more common. The threat of a paralyzing terrorist attack, like a targeted strike on a major port or waterway, meanwhile, has loomed over the global energy trade for years. Increasingly powerful cyberattacks hold the potential to take down the US electrical grid, bringing the global economy to its knees. All of this to say nothing of the inevitable pressures that have come with globalization, like financial contagion, sector-wide bankruptcies, and defaults.
The good news is that the tightness of supply chains can be a source of resilience as well as vulnerability. In Japan and elsewhere, some car manufacturers have forged close supportive relationships with their suppliers, including taking equity stakes in their businesses to signal a greater commitment.
Such partnerships were key in helping Renesas secure a bailout deal, led by Toyota and other manufacturers in exchange for equity, in December 2012, the rest came from a government public restructuring fund, the Innovation Network Corporation of Japan. The firm has since managed to retain its 40-percent share of the world’s market in microcontrollers and has recovered enough to plan a major takeover of an American rival. One can only hope the next company that comes close to crashing a vital global supply chain gets equally lucky.