Trends and measures of global trade growth
Is there more that the G20 can do to restore trade and economic growth and, if so, what could the T20 think tanks do to help? I would suggest that T20 think tanks could help by enlarging the analysis in two ways:
Look beyond trade
We must first look past trade to discern what factors may be shortening global value chains (GVCs) and slowing trade growth.
Since 2008 central banks have been implementing very loose monetary policies, including by providing financing at a loss (employing negative interest rates). Interest rates have never been so low. The longest records on interest rates are those from the Netherlands, which go back 500 years and clearly demonstrate that interest rates are currently at record low levels. Ultra-low interest rates could be entirely justified for reasons that I do not intend to address in this article, but subsidising capital can only invite industries to replace labour for capital.
If, in the 90s, trade growth outpaced GDP growth by a factor of 2.5 and, in the first years of this century (up to the crisis), trade growth was still outpacing output by a factor of 2.0, it was to a large extent because manufacturing industries moved their labour-intensive production to countries where wages were cheaper. Since 2008 the main central banks (including China’s) have engaged in extra-loose monetary policies. Subsidising capital for such a prolonged period of time may be reversing the economics of that equation (compounding the trend set by technological developments of substituting labour for capital). This could help to explain some in-shoring of jobs back to advanced economies, some of the increased capital intensity of production in countries where work-force is still abundant (I understand that this is the case in China) and the slowing down (or shortening) of GVCs.
Interestingly, the last International Monetary and Financial Committee (IMFC) meeting called for “strong, sustainable, inclusive and job-rich growth”. But, since ultra-cheap capital makes labour-substitution more attractive, it is not clear that job-rich growth could be achievable without squeezing wages even more, hence compromising inclusiveness and sustainable growth.
The Fund and others are looking closely at the potential consequences of loose monetary policies, particularly the aforementioned negative interest rates. But the main concern is about spillovers on emerging markets with open capital accounts but shallow financial markets and weak regulations. What macro-prudential measures are necessary to avoid spikes of financial volatility? Are financial assets overpriced? Should central banks look beyond the Consumer Price Index (CPI)? However, the costs of capital and labour are intrinsically linked and much of the explosive growth of GVCs was related to the interest in bringing labour-intensive production to economically more developed countries (EMDCs) where labour is abundant and wages lower. I have not seen any research on the possible consequences on labour-substitution (and hence on GVCs and trade) of subsidising capital for a prolonged period of time.
Look beyond “volume”
About a month ago we, the IMF Executive Board (EB), met to discuss the latest World Economic Outlook (WEO). This is the Fund’s flagship publication and it falls under the responsibility of staff, more precisely, the Fund’s Research Department. Yet staff always shares the draft with the EB beforehand, granting us the opportunity to express our views and sometimes ask for factual and other cosmetic corrections.
At this particular meeting, the Fund’s chief economist, Maurice Obstfeld, was explaining that, for the forth time in a row, the Fund had to lower its GDP projections and, to illustrate this, among other things, he noted that trade growth was barely matching that of world GDP.
While this is strictly true, I asked him if he was referring to trade growth measure in volume. He nodded. Then I noted that volume could be telling us only part of the story about trade growth and that trade could indeed be growing much faster than what volume measurement suggested. Volume is a well-adapted measurement for economies that are dominated by production of and trade in physical goods but, arguably, the most dynamic sector of the world’s economy is immaterial, hence has no “volume”.
According to the McKinsey Global Institute, data flows already account for a larger share of global output growth than trade in goods. It is difficult to say how much trade is digitally delivered, but McKinsey finds that the use of cross-border bandwidth has grown 45 times larger since 2005 and is projected to grow by another 9 times in the next five years. Of course not all of these trans-border byte-flows could be categorised as trade, as they include exchange of personal photos, videos and documents. But the US Department of Commerce, focusing only on flows of data traded between a seller and a buyer at market price (i.e. missing commercial data exchanged between businesses and digitally-enabled services delivered to end-users at no price), found that, in 2011, trans-border data flows represented over 60 percent of US services exports and about 34 percent of US total export value.
So it seems that an increasingly significant part of global trade is delivered digitally, just under our noses. Yet it goes mostly undetected because the screens of our “trade-radars” are still showing mostly (not only, but mostly) trade in volume, namely in goods and some of their associated services (including transportation and trade payments). To use a metaphor, our radars are efficient at detecting atoms (goods and the associated traditional services) but poor at detecting waves, that is, what our countries are increasingly exchanging.
This makes progress in measuring digital trade very challenging. Why is this so challenging?
Firstly, because the value of some data flows cannot be easily equated to its price. Say when we download an application for our smart phone or tablet, or when we use entertainment from YouTube, or we obtain information of commercial or professional value by using one of the web search engines, we may not pay any discernible price for these valuable services.
Secondly, measuring digital trade is a challenge because it is difficult to determine where digital data is coming from as it is so ubiquitous and it may be stemming from different places. I am not saying that it can’t be done. I discussed this with professionals and they all agree that it could be, but in order to identify the location where digital data is stored and, therefore, where it is coming from, they may need to interfere with the flow of data. Just as with waves, digital data could be present in several places at the same time and, in order to pin-down one specific point in the map, it may be necessary to interrupt the flow.
Not surprisingly, one of the most contentious issues for international digital trade regulation will be whether, and when, governments have access to that information and whether they require that servers be localised in their discrete national jurisdictions.
To make things even more challenging, producing atoms (meaning goods) has a marginal cost (which may be, or rather should be, equal to the market price). However, the marginal cost of data flows is virtually zero (the cost difference between sending 100 bytes or one trillion bytes may be negligible). Not surprisingly, owning atoms implies enjoying the capacity to exclude all others from possession. The same could be said for intellectual property rights and for conventional services (if I pay for a business class seat, it means that I can exclude all other passengers from occupying that discrete space in a given aircraft). However, in the digital economy, possessing digital data does not necessarily exclude others from “having” (and using them) at the same time.
The good news is that the WTO is already working to improve our capacity to measure, or rather estimate, digital flows. Coming back to my original anecdote, the Fund’s chief economist acknowledged that we may be under-estimating the world’s international trade (and also global output) and decided to include a trade chapter in the next WEO. I am not expecting any groundbreaking proposal regarding how we might overcome our “volume-bias,” but IMF staff from the Research Department is already in contact with WTO statisticians, which could lead to some joint work that could help us to understand the dynamics of this immaterial new form of trade.
Note: Some technological innovations are already (albeit incipiently) reshaping production and trade patterns. For instance, machine-to-machine communication (M2M) could have profound effects on job creation, and also on GDP and trade measurements. Since the marginal costs of delivering data is zero (or very close to), even an explosion of trans-border bytes could be difficult to detect through banking transactions. In addition, 3D printing (i.e. building by adding layers rather than carving by scrapping material) could reduce trade (and of course jobs) as the process could be moved to consumption markets and result in production based on electronically delivered M2M instructions. 3D printers could also actually reduce trade by saving waste and therefore reducing demand for raw materials.
Hector Rogelio Torres is Executive Director at the International Monetary Fund (IMF).