Everything You Ever Wanted to Know about Trade Facilitation and GVCs: Evidence from a Global Model�Trade Policy Research Forum

Ben Shepherd, Principal.

1

Why Should We Care About GVCs?

• We all know that the iPhone is “made in the world”.

​

• But so too are many other products, across a wide range of sectors.

​

• Here’s the Nutella GVC:
• In the US, it is imported from Canada and Mexico.
• But value added comes from all of these other areas, including the US itself.

​

• Standard trade data don’t fully capture this reality because they are in gross value terms (“invoice price).

​

• How can we do better at capturing the realities of global trade in 2022?

2

Source: https://econlife.com/2014/01/nutellas-global-supply-chain/.

Conceptualizing GVCs from an Economic Perspective

• Think of a GVC as an infinite chain of input-output relationships:
• Final goods producers use inputs.
• Intermediate goods producers use inputs.
• Those input producers use inputs.
• And so on and so forth.

​

• The essence of a GVC is that this process is spread across multiple countries, so the addition is across geography and sectors.

3

Conceptualizing GVCs from an Economic Perspective

• Blue boxes originate in the country of final consumption.
• “Domestic Value Added” (DVA).

​

• Red boxes originate in other countries.
• “Foreign Value Added”.

​

• From an accounting perspective, the value of the final product is the sum of all the elements of DVA, all the elements of FVA, and any pure double counting (PDC) due to moving goods and services across borders during production.
• X = DVA + FVA + PDC

4

How Much Does GVC Trade Matter?

• About 30% of global exports is FVA + PDC = GVC trade.

​

• Strong increase in the early 2000s, then levels off and fluctuates.

​

• Some increase again from 2015 onwards.

​

• What is driving these changes?

5

Determinants of GVC Trade

• Changes in trade costs have multiple effects in a GVC world:
• Typical trade theory: Change consumer prices.
• GVC trade theory: Also change input prices.

​

• So changes in trade policy that drive changes in trade costs can affect firm choices about:
• What to source from overseas?
• Where to source it?

​

• So there is a direct line from changes in trade costs to changes in the distribution of trade across DVA, FVA, and PDC.

6

Determinants of GVC Trade

• One possible source of changes in trade costs between 2015 and 2019 is improvement in trade facilitation.
• Customs and border procedures (as per the WTO definition).
• WTO TFA entered into force in 2017, but crystallized improvements already underway.
• Partial implementation for most LMICs.

​

• There’s a positive correlation between changes in TF performance (measured by OECD TFI) and changes in GVC integration (measured as [FVA+PDC]/Exports).

​

• But can we do better and actually quantify the link?

7

A New Quantitative Trade Model with GVCs

• To formally model the potential links between improved TF and more GVC trade, we need the following:
• An econometric model to provide an estimate of the elasticity of bilateral trade with respect to TF (structural gravity).
• A computational model (NQTM) where the input is a set of changes in trade costs based on some assumption about changes in TF: perhaps observed changes between 2015 and 2019, using 2015 as the baseline?

​

• This kind of modeling is data intensive:
• Multi-region input-output table, for calculation of GVC trade (ADB here).
• Estimates of structural parameters for the computational model (taken from a WTO working paper by Egger et al.).

​

• The model lives in thousands of lines of code, but all the math does is capture the basic logic of general equilibrium thinking about trade, then carve out a subset of information for calculating changes in GVC trade specifically.

8

A New Quantitative Trade Model with GVCs

9

Empirical Findings

Gravity Model – Changes in Trade Costs

NQTM – Changes in GVC Trade

10

Empirical Findings

• Changes in trade costs resulting from improvements in TF between 2015 and 2019 are typically modest, a few percent.

​

• But they nonetheless made a substantial contribution to the increase in GVC trade over the same period: around 30% of the total.

​

• Results vary by sector, both due to the size of the shock (gravity parameters) and the correspondence between observed and counterfactual changes.

11

Conclusion and Policy Implications

1. NQTMs bring the best of gravity modeling and CGE (GTAP) together, and can be used to answer policy questions:
• Empirical fit and simplicity of gravity.
• Parsimony and transparency relative to CGE.

​

• GVC trade is relatively sensitive to changes in trade costs, in particular in input markets. Lowering these costs can help promote production sharing across borders.

​

• Trade facilitation is one set of policies that have generally helped reduce trade costs over recent years. It has thereby helped GVC trade grow.

​

• But there is much left to do to achieve full implementation of the WTO TFA – which means there is also much more GVC trade to unlock!

12