What should a Doha Round SSM look like?

1

• What should a Doha Round SSM look like?
• Discussion on SSM triggers and remedy
• Ramesh Sharma
• Trade and Markets Division
• FAO, Rome
• ICTSD/World bank Institute seminar on Breaking
  the Deadlock on the Special Safeguard Mechanism
• Geneva, 1 December 2008

2
Inputs to discussion

• Some numbers behind the proposals and positions
• Mostly, on volume-based SSM triggers and remedy,
  in view of the intense discussions on this
  subject
• Very briefly, price-based SSM trigger and remedy

3
Volume-based SSM

• Trigger main issue what would be the
  appropriate level of threshold for the SSM
  trigger?
• Analysis is based on frequency of triggers
  (counting number of triggers for various
  thresholds using historical import data)
• A total of 131 cases (countries/commodities)
• Implicit assumption historical import trends
  are good guides for future trends

4
Data used - 131 cases (9 products and 23
countries, import trends for 1992-2004 period,
FAOSTAT)
5
Pattern 1- flat import trend with 1-2 surges
(i.e. 10-20 of the time) (MA-3 with 10
threshold)Tanzania, soy oil imports (2 triggers)
6
Pattern 2- some trend, with larger
fluctuations3-4 surges (i.e. 30-40 of the time)
(MA-3 with 10 threshold)Philippines, beef and
veal imports (4 triggers)
7
Pattern 3- steady and strong positive import
trend 7-9 surges (i.e. 70-90 of the time)
(MA-3 with 10 threshold)Honduras, rice
imports (9 triggers)
8
Import surge how a threshold determines the
number of times a SSM is triggered (shown as
of times)
9
How assumption about holiday changes the
number of triggers for various levels of
thresholds(numbers in table are times a SSM
is triggered)
10
Quantifying remedy levels for volume-based SSM

• Very difficult problem to solve .....
• But there is a way to get some rough guidance

11
How much extra (SSM) tariff is required to
contain a given level of surge?This depends on
import demand elasticity for a product in question
12
The corresponding formula ....

• ?t/t0 - (1t0)/t0)1/ ?m ?M/M0
• Relationship between change in tariff (?t/t0)
  and intensity of a surge (?M/M0) is not
  one-to-one, but is determined by
• current applied tariff (t0) and
• import demand elasticity (?m).

13
Remedy .. very difficult problem to solve ...

• Depends on import elasticity, which is product
  and country-specific (there are 90 or so products
  in 90 or so developing countries)
• Depends on initial applied tariff could range
  from 0 to say 80
• Also size of surge matters therefore, desirable
  to have tiers say, two tiers, if not three, for
  different levels of surge (e.g. 10-30 surge, and
  above 30 surge)

14
Yet, some guidelines for remedy level and caps
from model simulations

• The first option for SSM duty x of bound rate
  is useful only for high bound rates (gt80)
• With import elasticity in the range of 1-2,
• 15 percentage points of extra tariff (on current
  bound rates) mostly covers a surge of 20 (table)
• 30 percentage points of extra tariff (on current
  bound rates) mostly covers a surge of 40 (table)
• As these extras are needed, these are also the
  caps (i.e. max current bound extra)
• Results fairly close to proposals in July 10 text

15
15 percentage points of extra tariff (on current
bound rates) mostly cover the needs for a surge
of about 20Numbers on right are 15 points
bound rates)
16
30 percentage points of extra tariff (on current
bound rates) mostly cover the needs for a surge
of about 40Numbers on right are 30 points
bound rates)
17
Price-based SSM

• Threshold and triggers

18
Price-based SSM - Counting number of triggers
using world market prices and MA-3 (analysis
for 19 years, 1986-2004, for 10 products
poultry, beef, dairy SMP, dairy WMP, sugar raw,
sugar refined, palm oil, soya oil, rice and
wheat) Example refined sugar, MA-3 and 10
threshold
19
Price SSM how a threshold determines the number
of times a SSM is triggered (shown as of
times, MA-3 reference) (e.g. 10 threshold
triggers SSM 29 of the time)
20
Finally ........ real holiday - in Kathmandu