, Lecture 5 notes (+ Harvard Business Review article)
● Data limitations in UN Comtrade data
○ We only see border-crossing goods (not functions or services)
○ Within assembly, there is more value added (rather than selling raw cotton, for
example). Downstream activity is more valuable than the upstream activity
○ Exports/imports are used as proxies>we use exports/imports within UN
Comtrade instead of “domestic production sold abroad” and “foreign
production used at home”, which is not optimal (but for the former two, there
is not good enough UN Comtrade data)
■ Example: in export is also counted what is exported from a harbour,
but no value is added there
● Why theory and [UN Comtrade] data don’t perfectly match
○ theory level: micro/meso>firms, suppliers, lead firms, specific chains
○ data level: macro> country-product trade flows on UN Comtrade
○ We don’t have global, corporate firm-level GVC data
○ Macro as a “shadow” micro: country exports/imports are an aggregate
measure (if a country has one really big firm, it looks like the entire country is
importing/exporting a certain product, while this does not have to be the case)
● Lecture 3 recap/continued: how crises (non-policy shocks) push GVCs away from
pure efficiency, towards more resilience>more hierarchy or more relational ties? Do
responsibility
○ Non-policy shocks=exogenous disturbance to economic activity not caused
by deliberate government policy actions.
● Why does this matter?
○ exogenous: hard to anticipate/control>many companies went bankrupt during
COVID
○ network breaks: critical events can remove GVC links>raw cotton suppliers
can be integrated, for example
○ propagation: breaks can cascade upstream/downstream>problem with cotton
supply can eventually travel downstream to the T-shirt seller
○ disturbance vs disruption: is it a dip or a broken link in the GVC?
■ Examples: Suez Canal blocked by a crashed ship, Dieselgate,
Japanese earthquake increases electronic prices, Mild winter lowers
demand for gas
● Supply shock vs. demand shock
○ Supply shock= changes productive capacity or costs without a policy move
■ factory/plant outage, disease among employees, input scarcity, energy
outage, transport bottleneck, cyberattack on supplier
○ demand shock= shift private demand/preferences uncertainty or global
spending independent of policy
■ sudden preference shifts, weather-driven demands (with gas),
reputation shock (brand scandal)
● Data limitations in UN Comtrade data
○ We only see border-crossing goods (not functions or services)
○ Within assembly, there is more value added (rather than selling raw cotton, for
example). Downstream activity is more valuable than the upstream activity
○ Exports/imports are used as proxies>we use exports/imports within UN
Comtrade instead of “domestic production sold abroad” and “foreign
production used at home”, which is not optimal (but for the former two, there
is not good enough UN Comtrade data)
■ Example: in export is also counted what is exported from a harbour,
but no value is added there
● Why theory and [UN Comtrade] data don’t perfectly match
○ theory level: micro/meso>firms, suppliers, lead firms, specific chains
○ data level: macro> country-product trade flows on UN Comtrade
○ We don’t have global, corporate firm-level GVC data
○ Macro as a “shadow” micro: country exports/imports are an aggregate
measure (if a country has one really big firm, it looks like the entire country is
importing/exporting a certain product, while this does not have to be the case)
● Lecture 3 recap/continued: how crises (non-policy shocks) push GVCs away from
pure efficiency, towards more resilience>more hierarchy or more relational ties? Do
responsibility
○ Non-policy shocks=exogenous disturbance to economic activity not caused
by deliberate government policy actions.
● Why does this matter?
○ exogenous: hard to anticipate/control>many companies went bankrupt during
COVID
○ network breaks: critical events can remove GVC links>raw cotton suppliers
can be integrated, for example
○ propagation: breaks can cascade upstream/downstream>problem with cotton
supply can eventually travel downstream to the T-shirt seller
○ disturbance vs disruption: is it a dip or a broken link in the GVC?
■ Examples: Suez Canal blocked by a crashed ship, Dieselgate,
Japanese earthquake increases electronic prices, Mild winter lowers
demand for gas
● Supply shock vs. demand shock
○ Supply shock= changes productive capacity or costs without a policy move
■ factory/plant outage, disease among employees, input scarcity, energy
outage, transport bottleneck, cyberattack on supplier
○ demand shock= shift private demand/preferences uncertainty or global
spending independent of policy
■ sudden preference shifts, weather-driven demands (with gas),
reputation shock (brand scandal)
