If you were to buy a computer in, say, the US, there’s a good chance your machine will have been made in China – or at least that’s what the label will say.
In truth, your computer’s microprocessor may have been made in a high-tech hub like Singapore, while its casing was manufactured in Shenzhen. Each production stage could have been carried out in a different country, adding a different amount of value to the final article each time. The combining of intermediate goods to create a final product, often for export, is known as vertical specialization.
Vertically specialized trade is an important part of China’s business with the world. According to a working paper by researchers Judith Dean, K.C. Fung and Wang Zhi, the degree of vertical specialization in computer exports topped 50% in 2002. That is, more than half of the computer’s value was contributed by imported components, and neighboring Southeast Asia is one of China’s biggest suppliers.
The problem is, even the experts on the matter – economists who watch the region – have trouble quantifying the precise amount of vertical specialization that takes place in a given part of the supply chain.
“That’s the million-dollar question,” said Tim Condon, head of Asia research for ING Fianncial Markets.
What’s needed, economists say, is a certain set of data, known as input-output tables, for the region. These tables will show the amount of value added for each class of product in every country, which is crucial to understanding supply chains in the region because they cross so many boundaries.
Comprehensive input-output tables for the region would allow economists to track the components across its supply chain, thus allowing them to determine how much value a country contributed.
“The ultimate answer from an input-output table is: What comes in, who supplies it, what they do to that and, ultimately, where it ends up,” Condon said.
The missing link
It’s not that input-output data doesn’t exist. Many Asian countries, China included, compile comprehensive data, economists say. The data, however, is not coming out quickly enough. For example, what’s recognized as the most comprehensive set of tables for Asia is produced by the Institute of Developing Economies (IDE), a research institute linked to the Japan External Trade Organization (Jetro). But it only produces these figures twice a decade. The result is that the most recent data from IDE-Jetro available to economists now are figures from 2000.
Input-output tables updated once every five years are not peculiar to Asia. Indeed, developed economies like the US compile these figures at similar intervals. That works fine for economies with relatively stable growth rates. The sophistication of technologies used to generate economic output also tends to change little in these countries.
“For most countries, the technology coefficient does not change so dramatically. For example, if you produce a car and you need 7,000 kilograms of steel, the steel does not change by the year,” said Wang Zhi.
But five years is a long time in a place like China, where double-digit GDP growth is the norm. Therefore, these intervals are particularly ill-suited for understanding Asia’s fast-growing economies.
“For countries like China, the technology changes quite dramatically,” Wang said. “[But] people using input-output tables assume the technology coefficient doesn’t change over five years.”
If the problem in the Asian context is timeliness, why does IDE-Jetro update its input-output table so infrequently? The Japanese project collates the data from up to 24 sectors across nine countries, relying on various institutions in these countries – statistical bureaus, central banks, academic bodies – to feed it data.
Because each country constructs its own input-output tables slightly differently (for example, the number of sectors tracked; how transactions are valued), IDE-Jetro has to edit a large amount of information to conform to its own international input-output table. And, to compound the problem, these countries don’t construct their national input-output tables at the same time, which leads to synchronization problems.
“Compiling the Asian International Input-output Table is highly constrained by the compilation schedule of member countries,” said Hiroshi Kuwamori of IDE-Jetro.
Kuwamori says an update of the Asian input-output table, containing the figures from 2005, will be published in March 2011.
So economists have to make do. Trade data compiled by the IMF and the UN, for example, tend to be both timely and detailed, providing analysts with ways of working around the lack of fresh input-output tables. Others use even more novel means to get at the underlying trade flows.
“There’s an industry at this point [in workarounds],” said Condon.
The IMF’s Direction of Trade (DOT) statistics, for example, are updated monthly. The DOT figures can tell a reader where products from one country are being shipped to, but the drawback is it doesn’t say whether those goods are for final consumption. The UN’s much-referenced set of country data, on the other hand, does allow analysts to make decent assumptions about how a good is used. But both these data sets can only tell part of the story when it comes to tracking how value is created in a regional supply chain.
“The UN data and the [IMF] data are not 100% compatible,” said Park Cyn-Young, a senior economist at Asian Development Bank. “You get part of the picture from the UN data, and part of the picture from the [IMF] data. But it doesn’t really give you the complete picture of input-output structure for the region.”
The issue of the missing input-output tables could point to a deeper problem with international trade statistics. Existing metrics may not be sensitive enough to be used to understand complex supply chains fragmented across national boundaries – a situation exemplified by China’s intermediate-goods trade with Southeast Asia.
“The statistical systems that measure international commerce were designed for an earlier era, long before production chains had become so complicated,” said Lee Branstetter, an associate professor of economics and politics at Carnegie Mellon University who studies China.
“But the problems are especially severe for China and Southeast Asia because vertical trade is really important in these countries.”