Thursday, 7 June 2018

What Digital Policy Restrictions to Focus on?

Now that the DTRI is out, I have received the same question various times from different policy officers in the field: what digital policy restrictions should we focus on?

The DTRI covers a wide spectrum of policy categories ranging from tariffs in ICT goods to regulations in services and investment to cross-border data flows restrictions. The obvious answer to a trade economist like me would therefore be: well it depends. That’s a cliché, but since economics is about scarcity, decisions need to be made about trade-offs.

Personally, I am more interested in the future of trade, namely (digital) services, data and ideas and other intangibles. I think that’s where world trade is heading to and where new large productivity gains will have to come from. That’s not to disregard trade in ICT goods or e-commerce. Precisely if a choice needs to be made, I would argue for these flows to focus on in future policy negotiations.

Now, to focus which policies then: the OECD recently released an interesting report in which services trade by modes of supply are estimated. 

This is a huge step since before we only had a rough idea whether services trade took place through foreign affiliates after FDI was established (i.e. Mode 3), or through the internet (i.e. cross-border called Mode 1). The figure below shows both items in the form of a ratio: Mode 3 over Mode 1 in two points in time, namely 2000 (white dots) and 2014 (blue bars).

The figure tells me that the higher the blue bar, the more trade through establishments were important in 2010. When the blue bar falls below the white dot, trade over the internet has become more important over the years to 2014 – and vice versa.  

Source: Andrenelli et al. (2018), page 21; Analytical AMNE database Note: Exports of foreign affiliates have been removed from cross-border exports.

When looking at this figure, some extremely interesting conclusions become visible. One, for some services such as Distribution, Publishing activities, Computer and information services, and Financial and Insurance services, and possibly for Professional and Scientific services, this ratio decreased over time. 

This means that the internet as a vehicle for trade in these sectors has become more important. In other words, more trade of these sectors has been traded over the internet rather than through foreign establishments.

Other sectors such as Construction, Postal services, Warehousing and Transport support, Administrative services or even Telecom, trade through a foreign establishment has become more important over time as the blue bar falls above the white dots (though Telecom already had a high internet-trade ratio).

This pattern is very relevant to decide to focus on in terms of policy in services. The services sectors that have seen an increase of internet-based trade are also the ones that Ferracane et al. (2018) have assessed as very data-intensive. These are sectors that use a lot of software and data. These are also the sectors that are mostly affected by data-related policies such as data localization.

That’s visible in the figure below. The sectors which are colored in green are the ones which are most data-intensive; the sectors in red are least data-intensive. If a country has comparative advantage in data-intensive sectors, the policy restrictions covered by Cluster C in the DTRI are most relevant: data localization, data retention, intermediate liability and policies related to content access. 

Source: Author; Note: Data & Software intensities based on US Census data.

Conversely, if your country has comparative advantage in say Postal services, Transport, Health or Construction, then it makes sense for policy makers to focus on policies covered under Cluster B of Establishment restrictions. 

Sure, some trade in these sectors are still traded over the internet, but since complementarities exists between the two Modes of trade, prioritizing establishment restrictions will nonetheless have a knock-on effect on trade in these services over the internet.


Ultimately, therefore, choices on what policy to focus on will have to be based on your country’s comparative advantage: where is your country good at in exporting – and how it is traded.  

Ferracane, M., J. Kren and E. van der Marel (2018) “Do Data Policy Restrictions Impact the Productivity Performance of Firms?”, DTE ECIPE Working Paper Series No. 1, ECIPE, Brussels, forthcoming.

Thursday, 3 May 2018

Who is going to win the race for AI: China or US?

Who will win the race in Artificial Intelligence (AI)? Recent articles have focused on this big battle between China and the US. Arguments appear in favour of both countries. China has huge amounts of data and a more relaxed framework regarding privacy, whereas the US attracts a vast amount of talent to develop software computing and has the good climate to let digital firms with new ideas flourish.

Frankly, I think this issue is slightly more nuanced.

Arguments on both sides in favor of a strong AI based does not year, however, make clear who of the two will profit best from long-run economic benefits using AI. For instance, it is often pointed out that China has more data than the US and therefore would win the race. To me, it seems that both countries own large amount of data. And due to strong network effects, the fact of having large sets of data just simply points out that the focus is on these two countries, and not for instance on the EU. Hence, both can win.

The question who will reap greater long-run benefits from AI is a bit more complicated to answer. Who of the two will have sustainable specialization patterns in AI depends on what economists call comparative advantage. China may be big and therefore have large network effects in AI, but that does not yet mean comparative advantage. For that to determine, one needs to have a look at what sectors are most amenable to AI.

The figure below sets out the most AI-intensive sectors based on current usage of data and software out of a much wider range of industry and services sectors. The figure reveals some interesting insights. For instance, it shows that mainly services appear to be most open to AI such as internet and software services, finance and insurance, computer systems, and logistics. But a couple of manufacturing sectors are also stands out such as chemicals, motor vehicles, computer and electronics and electrical equipment. 

Source: author’s calculations; US Census; US BLS. 

Wednesday, 11 April 2018

Is your country boosting digital services exports?

A couple of weeks ago, together with my colleague Philipp Lamprecht, I presented a webinar on digital services trade. Digital services are the fastest growing component of all types of trade flows since 1995. The gap of growth between exports in digital services and in goods or traditional services has become bigger and bigger over the years.

What explains this observation? Well, some countries are just good at exporting digital services over the internet because of their friendly policies and good digital network environment. If you want to know who is good at exporting digital services, who is under-performing its potential, and why, then watch and listen to this webinar.




For instance, some countries, such as France and Germany, are laying behind their potential to export digital services while others, such as Romania and Ireland are leading in some digital services. Of course, some countries are naturally placed to export services. Yet some of them are precisely still under-performing in digital services relative to their potential.

The webinar explains all these issues. The webinar follows our work on digital services export that we did for the Bertelsmann Foundation. My co-authors were Philipp Lamprecht, Hanna Deringer and Fredrik Erixon. We plan to make a follow-up study to see more clearly what exactly drives this pattern.


Enjoy watching!

Thursday, 22 March 2018

Economic Impact of LCRs in BRICS at the WTO

Last week, ECIPE was invited at the WTO to speak on our work on local content requirements in BRICS countries. It was a great pleasure to be there, and the discussion I had with the WTO secretariat was a most interesting. See below for the slides.





Of note, this study takes a broader definition of LCRs into account in the sense that they can relate to public procurement, investments, business operations or market access. Yet in all cases they have a clear requirement for local content classifying them as an LCR. 

Happy read!

Wednesday, 28 February 2018

Supply chain complexity, ICT and trade

A recent McKinsey report notes that the intensity with which firms employ Information and Communication Technologies (ICT) depends on four factors: (1) size of the firm; (2) supply chain complexity; (3) skill levels inside the firm; and finally (4) threat of competition. 

Deployment of ICT facilitates reaches higher productivity levels inside the firm. The four determinants therefore seem reasonable as the firm-level literature shows that generally these four points are indeed factors that are strongly associated with greater firm performance. However, in my view, the second factor of supply chain complexity merits some elaboration and refinement, especially with regards to international trade.

It seems intuitive at first sight to think that ICT tools and instruments smoothen the supply chain network. And so, the more complex this network becomes the more this chain uses ICT to solve complex hold-up problems related to trade. This is because ICT allows for geographically dispersed production and management activities. However, the data tells something different: some supply chain trade, which requires greater ICT, may actually be related to lower supply chain complexity. 

This can be seen by a measure that computes the “length” of the value chain by accounting for the number of production stages (Fally, 2012). The more stages of production involved, the lengthier the chain becomes, the more complex one can assume the supply chain is. The figure below plots the average of this indicator of supply chain complexity across a like-minded set of OECD economies for each sector on the vertical axis. The horizontal axis plots the ICT-intensity indicator from van der Marel et al (2016) to see for any meaningful pattern.


Source: US BEA; OECD TiVA. Sector numbers follow ISIC Rev 3.

Thursday, 15 February 2018

FDI Screening: Implications for the Future

Two weeks ago, I spoke at an event where I commented on a very interesting paper from Copenhagen Economics (CE). The paper assesses the economics of EU Foreign Direct Investments (FDI) and the need to undertake screening in the EU. Economists are generally weary of setting up screening measures for foreign investments as it increases substantial costs for little economic reason.

Therefore, the recent proposal of the EC to introduce screening measures has more to do with political economy, wider geopolitical or even security reasons. For instance, one of the biggest concerns is that some of the recent FDI coming into the EU is from various emerging economies such as China, Kazakhstan and Russia, which still have many State-Owned Enterprises (SOEs) in their economies.

This according to economists is a problem, but a simple back-of-the-envelope calculation tells me that the share of this FDI in some sensitive sectors flowing from these countries with much SEOs involved is at most 3.5 percent of EU’s total incoming FDI. Hence, a first question appears: does that warrant an overall screening measure? Some trade-offs are involved as obviously economics doesn’t stand on its own here.

However, my intervention was about the future of foreign investments. In particular, the changing nature of investments the EU has received in recent years, namely investments in digital sectors. This should force policy makers to think about when proposing investment regulations, including screening.

First, although it is right to state that FDI brings along a great “footprint” as economist say, this is in fact much lower for digital investments. With “footprint” we mean economic activity such as employment, value-added and greater output. Footprints for tangible investments such as manufacturing are known to be high, but the recent UNCTAD (2017) report shows that this footprint from multinational digital companies is actually much lower.

This can be seen in the figure below that measures this footprint by taking the ratio from foreign sales over foreign assets. For manufacturing and telecoms this ratio is actually 1:1 (in case of telecoms the assets are high because of infrastructure investments). Yet for FDI coming from digital multinationals, foreign assets are typically lower leading to a much higher ratio (because of higher sales), which means a lower footprint.


Source: UNCTAD WIR (2017)

Thursday, 25 January 2018

LCRs versus tariffs: The see-saw of trade barriers?

You may have not noticed it, but the use of local content requirements (LCRs) has gone up for years. They are used by developed as well as developing countries. LCRs aim to promote the use of local inputs. They also serve the purpose of fostering domestic industries. BRICS and many other emerging countries are frequent users of LCRs, together with the US.

However, LCRs can be highly damaging for the economy. While LCRs might have the perceived benefit of creating industry activity and local employment, these gains or often generated in the short-term. LCRs are most likely to have a damaging economic impact that is wider in the long run. This harmful impact therefore evolves over time, which eventually outweighs any specific short run gain they can create.

ECIPE’s new study with undersigned contribution estimates the damaging impacts of LCRs for BRICS countries. Our team has translated their negative effects into so-called ad valorem equivalents (AVE). This is a methodological concept that allows one to readily compare the adverse impact of any non-tariff barrier (NTB) such as an LCR with a tariff. Our study has taken LCRs in the heavy vehicle sector as a case in point.

The results are presented in the figure below. They show that Brazil and Russia apply the most distortive LCRs for heavy vehicles. The two countries have an estimated increase of their import price of 15.6 and 11.1 percent respectively. China and South Africa both show low AVEs of 4.5 and 3.3 percent respectively. India’s LCRs are least harmful as it shows an AVE estimate of 2.2 percent.
 
Source: ECIPE calculations, based on ECIPE LCR BRICS database; WITS/UNCTAD TRAINS

Tuesday, 16 January 2018

Are services really not helping the economy?

Bullocks! That is the conclusion that came to me when I participated in the roundtable conference on services and economic development in Tokyo last month. This conference was organized by the Asian Development Bank Institute (ADBI) and discussed with experts the positive role of services in the world economy.

It’s high time for this discussion. In recent years there has been a slight backlash against services as a contributing factor to the economy, particularly for developing countries. One reason for this set-back against services stems from a recent article by Dani Rodrik in which he drives the point that many developing countries are de-industrializing faster than before and therefore moving into services too quickly.

This premature de-industrialization, he argues, prevents them from using the manufacturing sector as a tool for rapid economic growth. According to Rodrik, in large part this is due to globalization and trade itself because globalization has produced changes in relative prices in advanced countries. This can have serious negative consequences for developing countries’ growth potential because, in the future, they would be much less able to capitalize on manufacturing exports.

There are, however, a couple of remarks that in my view must be made here.

First, these conclusions are most probably based on aggregate figures between countries’ services and manufacturing activities, which in great part mask the fact that many manufacturing sectors have already become “servicified” to a high degree. This means that a lot of gains by manufacturing firms are earned though services, not manufacturing. For instance, is Zara a garment manufacturer nowadays or just a retailer? Most probably it comes close to the latter. Such servicification is not yet picked up and properly classified in aggregate figures – and this is true also for developing countries.

Second, there are surely economic meaningful sectors in a country’s services economy. The old way to look at services is that they don’t show a great economic role as they are not receptive to productivity improvements. That assertion seems to be out of date. Even though productivity for services is hard to measure, European micro-level data suggest that productivity varies hugely across services, and therefore their contribution to the overall economy is also varied. (see figure below). This should also be the case for non-developed economies.

Source: Data taken from Van der Marel, Kren and Iootty (2015) "Services in the European Union: What Kinds of Regulatory Policies Enhance Productivity?", World Bank Policy Research Working Paper No. 7919, World Bank, Washington DC.


Friday, 12 January 2018

Productivity, Manufacturing and Trade

A very interesting piece by Robert Lawrence on manufacturing productivity and trade. Three very interesting conclusions come out: 

(a) that not trade is the major contributor to a decline in the share of manufacturing employment, but faster productivity; 

(b) that therefore productivity growth is in large part the factor that has contributed to losses of manufacturing jobs (together with our habit to not consume more goods but more services when we get richer); and 

(c) that there seems to be a trade-off in recent times between the share of manufacturing employment and productivity growth: more of the one is less of the latter -- or reverse. 

For economists dealing with this topic, the first two conclusions are not entirely new. In fact, this is how I have learned it from my textbook economics. The latter is new to me and very interesting. 

The last conclusions also raises some questions. For instance, is the historical leveling off of economic growth we have seen in the past (i.e. previous wave of globalization) related to this fact? Can new technologies in other sectors such as services we currently seeing reduce this trade-off? 

Some questions to think about.